Citation
Right ventricle failure in pulmonary hypertension is associated with mislocalization of gap junction Connexin-43 and changes in expression of contractile [alpha]-Actinin

Material Information

Title:
Right ventricle failure in pulmonary hypertension is associated with mislocalization of gap junction Connexin-43 and changes in expression of contractile [alpha]-Actinin
Creator:
Lohani, Ozus ( author )
Language:
English
Physical Description:
1 electronic file (126 pages) : ;

Subjects

Subjects / Keywords:
Pulmonary hypertension ( lcsh )
Heart -- Right ventricle ( lcsh )
Heart failure ( lcsh )
Heart -- Diseases ( lcsh )
Heart -- Diseases ( fast )
Heart failure ( fast )
Heart -- Right ventricle ( fast )
Pulmonary hypertension ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Abstract:
Pulmonary hypertension (PH) is a disorder of multiple etiologies and one without a definite cure, and high mortality. Little is known about the right ventricular proteomics in human heart failure, including possible differences in comparison to the left ventricle. Connexins are known to modulate inflammatory signaling in other systems and thus, we hypothesized that connexins may be responsible for increased lymphedema in the right ventricle. We show that in the right ventricular failure model of monocrotaline induced PH, connexin43 and alpha-actinin expression are significantly decreased and alpha-actinin is further decreased in Sugen4516-Hypoxic model of right ventricular failure PH whereas connexin43 shows no significant change. We used Western blot to confirm the protein expression and immunofluorescence staining for quantitative analysis of these proteins. With western blot we found that connexin43 is decreased by 64% and alpha-actinin is decreased by 63% 4 weeks after monocrotaline administration and furthermore, alpha-actinin is down by 84% in the sugen5416-hypoxic model. Lateralization of gap junctional proteins such as connexin43 is observed in ventricles of multiple heart diseases like cardiomyopathy and myocarditis. Along with the presence of gap junctional connexins in the lateral surfaces, their expression is usually found to be down-regulated. We can add monocrotaline-treated preclinical models to the growing list of cardiac diseases where the connexin43 is down-regulated. However, lateralization in the right ventricle was not observed. Patches of connexin43 are seen in which otherwise normally arrayed plaques are fewer or absent. The decrease in alpha-actinin in the preclinical models is a novel finding. Along with presence of ventricular fibrosis, increased edema in the right ventricle, these results suggest that pulmonary hypertension induced by pressure overload is associated with disorganization of gap junction distribution and the failing heart could be associated with changes in alpha-actinin expression.
Thesis:
Thesis (M.S.) - University of Colorado Denver.
Bibliography:
Includes bibliographic references
System Details:
System requirements: Adobe Reader
General Note:
Department of Bioengineering
Statement of Responsibility:
by Ozus Lohani.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
946537491 ( OCLC )
ocn946537491
Classification:
LD1193.E56 2015m L85 ( lcc )

Downloads

This item has the following downloads:


Full Text
RIGHT VENTRICLE FAILURE IN PULMONARY HYPERTENSION IS
ASSOCIATED WITH MISLOCALIZATION OF GAP JUNCTION CONNEXIN-43
AND CHANGES IN EXPRESSION OF CONTRACTILE a-ACTININ
by
OZUS LOHANI
B.S., Colorado School of Mines, 2009
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Masters of Science
Bioengineering
2015


This thesis for the Masters of Science degree by
Ozus Lohani
has been approved for the
Department of Bioengineering
by
Michael E. Yeager, Advisor
Richard Benninger, Chair
Carlyne D. Cool
November


Lohani, Ozus (M.S., Bioengineering)
Right Ventricle Failure in Pulmonary Hypertension is Associated with Mislocalization
of Gap Junction Connexin-43 and Changes in Expression of Contractile a-Actinin
Thesis directed by Assistant Professor Michael E. Yeager
ABSTRACT
Pulmonary hypertension (PH) is a disorder of multiple etiologies and one with-
out a definite cure, and high mortality. Little is known about the right ventricular
proteomics in human heart failure, including possible differences in comparison to
the left ventricle. Connexins are known to modulate inflammatory signaling in other
systems and thus, we hypothesized that connexins may be responsible for increased
lymphedema in the right ventricle. We show that in the right ventricular failure
model of monocrotaline induced PH, connexin43 and a-actinin expression are signif-
icantly decreased and a-actinin is further decreased in sugen4516-hypoxic model of
right ventricular failure PH but connexin43 shows no significant change. We used
Western blot to confirm the protein expression and immunofluorescence staining for
quantitative analysis of these proteins. With western blot we found that connexin43
is decreased by 64% and a-actinin is decreased by 63% 4 weeks after monocrotaline
administration and furthermore, a-actinin is down by 84% in the sugen5416-hypoxic
model. Lateralization of gap junctional proteins such as connexin43 is observed in
ventricles of multiple heart diseases like cardiomyopathy and myocarditis. Along with
the presence of gap junctional connexins in the lateral surfaces, their expression is
usually found to be down-regulated. We can add monocrotaline-treated preclinical
models to the growing list of cardiac diseases where the connexin43 is down-regulated.
However, lateralization in the right ventricle was not observed. Patches of connexin43
are seen in which otherwise normally arrayed plaques are fewer or absent. The de-
crease in Q!-actinin in the preclinical models is a novel Ending. Along with presence


of ventricular fibrosis, increased edema in the right ventricle, these results suggest
that pulmonary hypertension induced by pressure overload is associated with disor-
ganization of gap junction distribution and the failing heart could be associated with
changes in a-actinin expression.
The form and content of this abstract are approved. I recommend its publication.
Approved: Michael E. Yeager


DEDICATION
This file is dedicated to my parents: Om and Sarita Lohani, whose support and
dedication has made this thesis possible.
v


ACKNOWLEDGMENT
Without the continued support and mentorship of Kelley Colvin and Michael Yeager,
the completion of the project wouldnt have been possible.
Additionally, I would like to give thanks to the following individuals for their positive
contribution:
Lindsay Quandt
Greg Glazner
Radu Moldovan
Jane Parr
Aleena Notary
Mason McClatchey
Matthew Westacott
vi


TABLE OF CONTENTS
Tables......................................................................... x
Figures ...................................................................... xi
Chapters
1. Introduction................................................................ 1
1.1 Research Question.................................................... 1
1.2 Pulmonary Hypertension............................................... 2
1.3 Scientific and Clinical Justification of the Project................. 7
1.4 Hypothesis........................................................... 8
1.4.1 Right Ventricle................................................ 8
1.5 Gap Junctions....................................................... 10
1.5.1 Connexin-43 .................................................. 11
1.6 Role of Cx43 in Support of the Working Hypothesis: Vascular Leak
and Increased Interstitial Fluid..................................... 14
1.7 Connexins in the Myocardium......................................... 16
1.7.1 Fibroblast-Fibroblast Communication .......................... 17
1.7.2 Myocyte-Fibroblast interaction................................ 18
1.8 Normal Expressions: Location and Appearance......................... 18
1.9 Aberrations ........................................................ 20
1.10 Cell Specific and Lymphatic Markers ................................ 21
1.10.1 Cardiac Myocyte Markers: cu-Actinin and Dystrophin............ 21
1.10.2 Cardiac Fibroblast: Narrowing Down a Specific Marker for Fi-
brosis and Cardiac Remodeling........................................ 22
1.10.3 Apoptotic and Proliferation Markers: Caspase-3, Ki-67 .... 24
1.10.4 Markers for Lymphangiogenesis and Angiogenesis: Lyve-1,
Podoplanin, Proxl............................................. 25
2. Methods and Materials...................................................... 28
vii


2.1 Animal Models..................................................... 28
2.2 Obtaining Tissue.................................................. 29
2.3 Western Blot Protein Assay........................................ 30
2.4 Immunofluorescence Staining....................................... 32
2.4.1 IF and Antibody Staining Techniques ..................... 33
2.5 Confocal Microscopy............................................... 35
2.5.1 Rationale: Out of the Static Studies of Thin Sections..... 35
2.5.2 Why 3D over 2D: Whorling Effect/Stereology................ 37
3. Results................................................................. 38
3.1 Assessment of PH in Preclinical Models............................ 38
3.2 Immunofluorescence Staining: Qualitative Results................. 43
3.2.1 Connexin-43 Organization and Expression in the Normal Heart 43
3.2.2 Gap Junction Remodeling in RV Failure PH................... 43
3.2.3 Down-regulation of Connexin43 and cu-Actinin in MCT model
of PH as observed through immunofluorescence................ 45
3.3 Immunofluorescence Staining: Quantitative Results................. 47
3.4 Western Blot...................................................... 68
3.5 Assessment of Fibrosis............................................ 74
3.6 Limitations and Recommendations................................... 75
4. Discussion.............................................................. 78
4.1 Strengths and Weaknesses: The Tales from our Data and the Effects
on out Hypothesis................................................. 80
4.2 Future Directions ................................................ 82
4.3 Concluding Remarks................................................ 83
References................................................................. 84
Appendix
A. Custom MATLAB Codes.................................................... 103
viii


B. Supplementary Figures
109
IX


TABLES
Table
1.1 Updated clinical classification of pulmonary hypertension, 2013*............ 3
1.2 Marker profile speculations and interpretations in preclinical models ... 9
2.1 Antibodies used for Western Blot Assay..................................... 32
2.2 Primary antibodies and their optimum dilutions............................. 34
3.1 Zeiss LSM 780: Pixel size measurement reference for the different objec-
tives and zoom factors used during image acquisition...................... 51
3.2 Summary of western blot findings for control and diseased groups. Values
presented are relative as they are normalized to cell count............... 63
3.3 Summary of quantified western immunoblot protein assay..................... 72
x


FIGURES
Figure
1.1 The layers of the vessel wall........................................... 4
1.2 Composition of vascular adventitia: an overlooked nexus................. 5
1.3 Illustrative representation of progression of vascular remodeling in the set-
ting of pulmonary hypertension. Increased pulmonary vascular resistance
is observed due to unmonitored proliferation and growth of smooth mus-
cle cells and endothelial cells, as a result, leading to increased pulmonary
arterial pressure...................................................... 6
1.4 Thin-section electron microscopy showing the organization of gap junction
with other adherent junctional complexes in the rat RV. (A) Step-like
features of discs (arrowheads) observed at the end of the isolated cell.
(B) Fasciae adherent junctions occupying the electron rich region. Gap
junction appear to be connecting with fasciae adherent junctions (arrows). 12
1.5 Illustration of Connexin-43 arrangement in the plasma membrane.......... 13
1.6 Functional consequences of abnormal Cx43 organization and expression
seen in remodeled heart. It is unclear whether the remodeling of the heart
preceeds the disorganization of Cx43 or if it is causal. Dotted lines repre-
sent potential causal interplay, solid lines represent established associations. 15
1.7 Organization of connexin subunits to form the hemichannel subunit con-
nexon. Connexons have multiple connexin types forming either a het-
eromeric or a heterotypic channel...................................... 19
1.8 Colocalization of Podoplanin and Proxl in the tertiary lymphoid tissue
(BALTs) showing positive correlation between the two lymphatic markers. 26
xi


2.1 Heart tissue obtained from the preclinical models include the right and
left ventricles. The right and left atria are removed prior to cutting the
tissue desired for our assays. The whole heart on the other hand is utilized
to prepare homogenates for Western blot experiments. The Western blot
assay performed for this study used RV homogenates..................... 30
2.2 (A) Immunofluorescence stain of ci-actinin with longitudinal orientation
with the z-discs of the sarcomere distinguishable whereas (B) helps explain
the difficulty encountered when dealing with heart tissue. The complex
stereology of heart tissue causes the observation view to change from lon-
gitudinal (yellow arrow) to transversal sections (green arrow)......... 36
3.1 Macroscopic representation of morphology of the heart from preclinical
rat models. This figure shows cross section of (A) control and (B) tissue
4 week post MCT administration......................................... 38
3.2 Macroscopic representation of pentachrome stain on single 5 /im thin sec-
tions of the heart. This figure shows cross section of control, (A) and 3
week post MCT administration, (B) ................................ 39
3.3 Measurements of myocardial edema for RVs and LVs in various preclinical
models of PH. Values are reported as mean St. Dev................. 40
3.4 T2-weighted MRI showing edematous RV for MCT-treated Rat (right)
compared to the control (left). Images acquired using Bruker 4.7 Tesla MR. 41
3.5 Massons trichrome stain assessing hbrotic change in control (top left) vs.
hypoxic (top right) and MCT (bottom row). Hypoxic rats demonstrate a
similar pattern of modest collagen content in the RV as control. Elevated
collagen deposition in the RV of MCT-treated rats correlates positively
with the presence of EVF............................................... 42
3.6 Correlation between EVF and Fibrosis supports our speculations about
edema preceeding fibrosis in PH........................................ 42
xii


3.7 Distribution pattern of Cx43 in control ventricular myocardium (A) Lon-
gitudinal section from rat RV (B) Z-projection of confocal images show
apparent gap junction immuno-labeling at the lateral surfaces. These pro-
teins could be considered components of extended IDs (see Figure 1.4).
Most Cx43 gap junctions are observed at the periphery of the discs. ... 44
3.8 Down-regulation of immunodetectable Cx43 and a-actinin is observed in
rat right ventricle of MCT induced RV failure PH (B) compared to control
(A). 40x magnification................................................. 46
3.9 Representative image at a lower magnification showing decreased levels
of and Cx43 and a-actinin in MCT-treated PH (B) compared to control
(A). 20x magnification................................................. 47
3.10 Cx43 and a-actinin seem to share area coverage at the IDs. Decrease in
a-actinin in the MCT-treated rats (B), is qualitatively proportionate with
the amount of Cx43 present at the discs................................ 48
3.11 Schematic representation of quantitative image processing.............. 49
3.12 Intensity histogram for the raw image shown earlier (refer to Figure
3.7(A)). The gray-scale map on the x-axis represents the pixel intensity
value where 0 is black and 255 is white............................. 50
3.13 Images representing the steps taken for image processing. Cx43 on the
left column and nuclei on the right column were processed sequentially.
Scale bar=50 micrometer ............................................... 52
3.14 Representative image for the analyzed image used to provide the quanti-
tative index of Cx43 and nuclei count. A similar approach was taken for
obtaining the pixel count for a-actinin and dystrophin.............. 53
3.15 Box-whisker and scatter plot for a-actinin representing the total pixel area
normalized to cell count. n=5 control,; n=4 MCT; p<0.01................ 54
xiii


3.16 Box-whisker scatter plot for Cx43 representing the total area normalized
to cell count, p<0.01................................................. 55
3.17 Box-whisker and scatter plot for Cx43* correlates positively with data
performed with different clone of the antibody p<0.01................. 55
3.18 Clear demarcation of Cx43 plaques is observed in the immunofluorescent
image of RV at 80x magnification. Modified images for quantification
extraction is provided in next figure.............................. 56
3.19 (A) Thresholded image for Cx43 where blue pixels and red pixels are
excluded to provide us with the image in (B). Blue and red pixels in (A)
represent small areas and clusters that are indistinguishable as individual
plaques, respectively.................................................... 57
3.20 Bar plot summarizing plaque size measurements for Cx43 in control vs.
MCT-treated rat RVs. p<0.05.............................................. 58
3.21 Plaque size distribution distinguishing small and large Cx43 in control
and MCT groups........................................................... 59
3.22 Image of fluorescent beads (A) Raw image acquired using the same pa-
rameters used to image the Cx43 images that output plaque sizes. (B)
Post-processed binarized image that was labeled and used to extract the
length of the beads...................................................... 60
3.23 Dystrophin co-stained with Cx43 in control and 4 week MCT rat RVs.
Dystrophin labels the peripheral region of the cardiac myocyte and is
absent at the IDs containing terminus of myocytes. Highlighted regions
show individual cardiac myocytes......................................... 61
3.24 Bar plot showing that quantitative a-actinin and Cx43 correlates with
qualitative deductions; p<0.05. No significant change in dystrophin be-
tween the control and MCT-treated groups........................... 62
xiv


3.25 Cell death is a prominent feature early on during the progression of PH as
observed through CASP3 immunofluorescence staining. Cell-types mostly
going through apoptosis are in the myocardium but also in the blood
vasculature as shown by sma co-localization with CASP3...... 64
3.26 Bar plot showing active caspase3 in the RV myocardium of control and 1,
2, and 3 weeks post MCT treatment................................... 65
3.27 Nuclei counted for multiple staining sets show no significant decrease in
total number of cells from control to MCT-treated models............ 66
3.28 Dystrophin co-stained with Cx43 in control and 4 week MCT rat RVs.
Dystrophin labels the peripheral region of the cardiac myocyte and is
absent at the IDs containing terminus of myocytes. Highlighted regions
show individual cardiac myocytes...................................... 67
3.29 Summary of length and width analysis performed on control and MCT-
treated cardiac myoctes of the RV. p<0.01........................... 68
3.30 Western blot for a-Actinin (left) performed for rat RV homogenates.
GAPDH (right) used for normalization purposes. Significant down-
regulation of a-actinin in protein levels of MCT and Su5416-hypoxia
groups are observed................................................... 70
3.31 Quantification of Western blot for a-actinin shows 63% decrease for MCT
and 83% decrease for Su5416-hypoxia models............................ 70
3.32 Downregulation of Cx43 in protein levels is observed for the MCT RV fail-
ure model but not in Su5416-hypoxia PH rats. Quantification of Western
blot for Cx43 shows 64% decrease for MCT induced RV failure PH model.
p<0.01................................................................ 71
3.33 Cx43 in SuHx RV failure model. Immunofluorescence image acquired to
verify western immunoblot assay. Arrows show mislocalization of Cx43. . 73
xv


3.34 Box-whisker and scatter plot showing Cx43 protein levels in control vs.
MCT- and SuHx-treated PH rats.......................................... 73
3.35 Co-immunostaining of cardiac myocyte marker dystrophin (green) and
cardiac fibroblast marker (P4H) presented in a montage................. 75
3.36 Collagen deposition in the MCT-treated rats is indicative of fibrosis. Ad-
ditionally, the blue-green stain seen in the diseased image represents im-
mature collagen..................................................... 76
3.37 3-Dimensional reconstruction of a rat RV focused on Cx43 at the IDs. 80x
magnification; reconstructed using ImageJ................................ 77
B.l Custom MATLAB code applied for beads at 20x magnification with ac-
curate results for bead size output .................................... 109
B.2 Custom MATLAB code applied for beads at 40x magnification with accu-
rate results for bead size output. Thus, the code can be applied for images
of different magnification................................................... 110
xvi


1. Introduction
1.1 Research Question
Pulmonary hypertension (PH) is a disease characterized by progressive cardiopul-
monary remodeling and reduced right ventricular (RV) function. The progression of
the disease eventually leads to right heart failure. This mechanism is not fully un-
derstood in part because PH is associated with a rather large spectrum of conditions
and diseases that range from autoimmune disorders and immune-insufficiency syn-
dromes to systemic and metabolic disorders [1]. The functional status of pulmonary
circulation, the levels of pulmonary vascular resistance (PVR), and pulmonary arte-
rial pressure are key factors that help determine the outcome and therapeutic treat-
ments of patients with PH. Highlighting other aspects of pathobiology of PH broadly
classifies the disorder into mild/moderate vs. severe categories based on pulmonary
artery pressures, ventricular performance and overall mortality [2], Intimal thickening
and medial hypertrophy that are observed in chronic obstructive pulmonary disease
(COPD) [3] and other non-neoplastic lung diseases [4] present with mild to moder-
ate PH. Endothelial cell proliferative lesions (plexiform lesions for instance), intimal
fibrosis, smooth muscle cell (SMC) growth in the intimal and medial region are sus-
pected to cause severe cases of PH. Additionally, alterations in vascular reactivity are
observed alongside the arterial vascular remodeling.
Heart function is strongly reliant on water balance where even a minor increase
in myocardial water content impairs the contractile force [5]. Pericardial effusion and
presence of edema in the right ventricle are important pathological findings in the
setting of PH. Cell-cell junctional proteins play pivotal roles in intercellular commu-
nications between the various cell types in the heart and the lungs. Studies, discussed
in detail in section 1.6, suggest the role of gap junctions in maintaining vascular bar-
rier function. It is also speculated that gap junction inhibitors promote vascular
leak [6]. The accumulation of fluid in the heart and the lungs increases mortality in
1


patients with PH. Understanding the origin of edema and deducing why the homeo-
static fluid flow is disturbed in PH could be transformative findings. Fine-tuning our
understanding of the role of gap junction proteins in preclinical models of PH and
how they correlate to the condition in humans should lead to the development of new
therapies. This study aims to characterize the dysregulation of cell-cell junctions and
correlate the findings to presence of edema contributing to RV failure.
1.2 Pulmonary Hypertension
A disorder of multiple etiologies, PH is a condition where the mean pulmonary
arterial blood pressure (mPAP) is > 25 mmHg as determined by right heart catheter-
ization (RHC) [7]. Pulmonary arterial hypertension (PAH) reserves some of the well-
defined forms of PH and is further defined as an mPAP of > 25 mmHg in the presence
of pulmonary arterial wedge pressure (PAWP) of < 15 mmHg, or in other words, nor-
mal left-sided filling pressures [8]. The current classification of PH is based on the
Evian/Venice classification proposed during the 1998 second World Symposium of
Pulmonary Hypertension and modified during the third Word Symposium held in
Venice, Italy in 2003 [1]. Based on the classification, PH is categorized into five
groups where each group cohesively defines PH according to similar clinical, pathobi-
ological, etiological and therapeutic characteristics. This general categorization has
been maintained through the fourth and fifth World Symposium held at Dana Point,
California (2008) and Nice, Prance (2013) respectively. Some modifications have been
made and the latest updated clinical classification is provided in Table 1.1.
The animal models generated for the purpose of this study fall under Group 1
of PH. Aside from increased mPAP being the hallmark of PH, biopsies of the heart
and lung tissue reveal widely accepted characterizations: sustained vasoconstriction
and vascular remodeling [9], and cardiac myocyte hypertrophy and RV fibrosis [10].
Vessel walls are comprised of three heterogeneous layers as seen in Figure 1.1: the
tunica intima, tunica media and tunica externa.
2


Table 1.1: Updated clinical classification of pulmonary hypertension, 2013*
1. PAH
1.1 Idiopathic PAH
1.2 Heritable PAH
1.2.1 BMPR2
1.2.2 ALK-1, ENG, SMAD9, CAV1, KCNK3
1.2.3 Unknown
1.3 Drug- and toxin-induced
1.4 Associated with:
1.4.1 Connective tissue disease
1.4.2 HIV infection
1.4.3 Portal hypertension
1.4.4 Congenital heart disease
1.4.5 Schistosomiasis
1 Pulmonary veno-occlusive disease and/or pulmonary capillary hemangiomatosis
1 Persistent pulmonary hypertension of the newborn
2. Pulmonary hypertension due to left heart disease
2.1 Left ventricular systolic dysfunction
2.2 Left ventricular diastolic dysfunction
2.3 Valvular disease
2.4 Congenital/acquired left heart inflow/outflow tract obstruction and congenital cardiomyopathies
3. Pulmonary hypertension due to lung disease and/or hypoxia
3.1 Chronic obstructive pulmonary disease
3.2 Interstitial lung disease
3.3 Other pulmonary diseases with mixed restrictive and obstructive pattern
3.4 Sleep-disordered breathing
3.5 Alveolar hypoventilation disorders
3.6 Chronic exposure to high altitude
3.7 Developmental lung diseases
4. Chronic thromboembolic pulmonary hypertension
5. Pulmonary hypertension with unclear multifactorial mechanisms
5.1 Hematologic disorders: chronic hemolytic anemia, myeloproliferative disorders, splenectomy
5.2 Systemic disorders: sarcoidosis, pulmonary histiocytosis, lymphangioleiomyomatosis
5.3 Metabolic disorders: glycogen storage disease, Gaucher disease, thyroid disorders
5.4 Others: tumoural obstruction, fibrosing mediastinitis, chronic renal failure, segmental pulmonary hypertension
a *
^Main modifications to the previous Dana Point classification (2008) are in bold. ALK-1, activin receptor-like kinase 1 gene;
BMPR2, bone morphogenic protein receptor type II; CAVl, caveolin-1; ENG, endoglin; KCNK, potassium channel subfamily K gene;
PAH, pulmonary arterial hypertension; SMAD9, SMAD family member 9.
*Fifth World Symposium on Pulmonary Hypertension, Nice 2013.
Haddad et al. Q


ARTERY * 'Lumen of artery
Smooth
muscle
a
Figure 1.1: The layers of the vessel wall.
Pearson Education, Incorporated
The outermost layer, tunica externa, is often referred to as tunica adventitia
or adventitia. Working our way from the inner to the outermost wall layer, we
have: monolayer endothelial lining, connective tissue and elastic fiber composing
the tunica intima; smooth muscle tissue sheath comprising the thickest middle layer
(more common in the arteries), separated from the adventitia by a thin band of elastic
membrane; and the outermost layer that forms a connective tissue sheath around the
vessels. The adventitia is more complex than originally speculated. As illustrated
in Figure 1.2 it is thought to consist of an extracellular matrix (ECM) framework
composed of nerves, lymphatic and blood vessels, progenitor cells, immune cells, and
fibroblasts [11]. Many preclinical animal models have consistently revealed extreme
adventitial remodeling in the setting of hypertension [12, 13].
While the exact causes of the progression remain unclear and under investigation,
the remodeling seen in PH is characterized to various standards by thickening of
intimal and medial layers of muscular vessels [9, 14], thickened adventitia [15], and
4


Lymphocyte
Endothelial Cells
Smooth Muscle Cells
- Fibroblasts
Red Blood Cells
Macrophage
Resident Endothelial Progenitor Cells
Dendritic Cells
Endothelial Colony Forming Cells
Circulating Endothelial Progenitor Cells
Adipocyte
a
Figure 1.2: Composition of vascular adventitia: an overlooked nexus.
Stenmark et al.
appearance/survival of vascular smooth muscle cells [16], leading to development of
vaso-occlusive lesions and varying degrees of inflammation [17]. Figure 1.3 illustrates
the putative pathogenesis of PH: progression of muscularization of small, peripheral
pulmonary arteries, and the histological image at the center shows heavy staining from
smooth muscle marker cu-smooth muscle actin (q'-SMA). Proliferation and migration
of pulmonary arterial smooth muscle cells (PASMCs) [14] and myofibroblasts [18],
endothelial dysfunction leading to compromising vascular tone [19], and vaso-occlusive
lesions involving PASMCs and endothelial cells (ECs) [20] occur in severe cases of PH.
However, studies have shown that fibroblasts residing in the adventitia may be
the most suitable cell type able to sense hypertensive states [21], and remodeling in
the systemic vasculature is characterized by increased fibroblast proliferation which
5


Healthy Lung Pulmonary Hypertension
a
Figure 1.3: Illustrative representation of progression of vascular remodeling in the set-
ting of pulmonary hypertension. Increased pulmonary vascular resistance is observed
due to unmonitored proliferation and growth of smooth muscle cells and endothelial
cells, as a result, leading to increased pulmonary arterial pressure.
Modified from Davies et al. with permission
precede and exceed ECs and PASMCs proliferation [22, 23]. Fibroblasts, dubbed
the sentinel cell by Smith et al, have been suggested to activate and differentiate
into myofibroblasts [24, 25]. Myofibroblasts participate in would healing but the
sustained survival of this cell type contributes to remodeling [18] and a chronic state
of inflammation via forward feedback mechanism [26]. Therefore, the adventitial
layer, the peri-adventitia surrounding the major vessels, and the contents within this
dynamic vascular wall are key players in remodeling of the bronchovascular area.
In addition to the pulmonary vascular resistance resulting from changes in small
arterioles and rarefaction of the vessels, increased stiffness (decreased compliance) in
the proximal pulmonary arteries contribute to increased afterload in the right ventricle
[27, 28, 29],
6


1.3 Scientific and Clinical Justification of the Project
Knowledge about the role of right ventricle in health and disease has often lagged
behind and overlooked when compared to the left ventricle (LV). This is perhaps due
to speculations that the RV is: less muscular, restricted to pumping blood to the lungs,
and due to its less frequent and nonobvious involvement in pathology. Historically,
RV has been a bystander in pathophysiological studies involving the cardiovascular
system. RV is adversely affected, most notably in PH caused by variety of pulmonary
vascular disorders. The RV experiences pressure overload regardless of any of the
causes listed in Table 1.1 and the responses are myocardial hypertrophy [30], dilation
of the chamber, rising filling pressures as a result of contractile dysfunction [31], and
diastolic dysfunction [32], RV function is a key determinant in survival of patients
with PH [33, 34, 35].
One of the findings in the setting of PH is pericardial effusion and presence of
edema in the RV and the lungs. Pericardial effusion and RV interstitial fluid are im-
portant and have serious clinical complications. Patients with pericardial effusion had
1- and 3- and 5-year survival rate of 80% and 20% and 0% respectively [36]. The home-
ostatic mechanism to regulate the fluid in the RV is greatly influenced by the fluid
movement into the system through microvascular control, the movement out of the
system by the myocardial lymphatic network, and the fluid efflux into the pericardial
space and its eventual drainage into the larger lymphatic system. The composition of
the fluid present in the RV is either intracellular, extracellular, lymphatic or blood.
Myocardial hypertrophy is one of the early compensatory mechanisms adapted by the
RV in response to pressure overload in PH [30]. The microvascular filtration is in-
creased in PH and fluid flow in the lymph system is decreased resulting in myocardial
edema [37]. Central venous pressure (CVP), defined as the pressure of blood in the
thoracic vena cava entering the right atrium, is elevated in PH regardless of etiology
[38]. CVP is often used to reflect the amount of blood returning to the heart. The
7


edema, now prevalent in the myocardium, depresses cardiac output (CO) [39, 5] as
a result of increased filling pressures, complicates contractility functions, and causes
right heart to fail due to systolic and diastolic dysfunction. Focusing on the RV could
further our understanding of the clinical issues and increase our knowledge of the
RV to be on comparative par with LV. This is of pivotal importance as we have a
chance to add to the growing knowledge of the progression of disease and increase
survival rates among patients. With this study, we devote attention to the RV so we
may develop ways to detect and measure RV dysfunction, and how the deterioration
progresses in PH.
1.4 Hypothesis
The working hypothesis for the project is based on observations of increased inter-
stitial fluid in failing RV in the setting of PH with the presence of increased vascular
resistance in diseased state than in controls. We hypothesize that the mislocal-
ization of gap junctions in the right ventricle contributes to increased RV
lymphedma which positively correlates with increased inflammation and
fibrosis, and decreased RV function. The hypothesis was tested with PH mod-
els that develop RV failure 4 to 5 weeks after administration of either Monocrotaline
(MCT) or Sugen5416; please see Chapter 2.
1.4.1 Right Ventricle
Myocardial fluid balance is maintained in large part by the lymphatics in heart
along with coronary venous circulation [5, 39, 40]. The lymphatics run alongside
blood vessels [41] and are comprised of myocardial, subendocardial and subepicardial
lymphatic plexuses [40], and lymphatic collecting duct, or the lymphatic trunk re-
sponsible for creating a pathway for fluid to leave the heart [42], Several driving forces
are involved in controlling the lymph flow in the heart, all of which fall into either
the active or the passive lymphatic pumping category. A long standing hypothesis
8


[43] had the pericardium playing an important role in draining lymph fluid from the
heart. But due to lack of evidence supporting the hypothesis and studies showing the
inability to induce severe lymphedema post pericardial removal [42], both in animals
and humans, it is suggested that the major driving force regulating the lymph flow in
the heart could come from subepicardial muscle contractions. Nevertheless, cardiac
lymphatics are critical for maintenance of fluid balance in the heart and aberrations
from homeostatic environment causes serious clinical complications that include peri-
cardial effusion [44], inflammation and fibrosis [45, 46], and drop in cardiac output
[47] which proves fatal in many cases.
Table 1.2: Marker profile speculations and interpretations in preclinical models
Marker Control RV Failure PH Interpretations
a-Actinin Control - Number of cardiomyocytes
Connexin-43 Control Mis-localization Mechanical failure marker
Control - Change in total abundance
Caspase-3 + X Apoptotic marker
Ki-67 + X Proliferation marker
LYVE-1 Control Decreased lymphatics
Podoplanin Control Positively correlates with LYVE-1 (lymph vessels)
Prolyl-4-hydrpxylase Control ++ RV Fibrosis
Prox-1 Control LECs
TCF21 Control ++ RV fibrosis
a
LYVE, lymphatic vessel endothelial hyaluronan receptor; LEC, Lymphatic endothelial cells;
Prox, Prospero homeobox; TCF, Transcription factor; +, ++, -, and signify varying degrees
of prevalence or remodeling in contrast to controls, x signifies no significant change (Ki-67, a
proliferation marker and Caspase-3, an apoptotic maker show no/minimum activity in 4 weeks RV
failure model of PH). Interpretations stated alongside clarify their role and reflection in the diseased
state.
9


One of the specific aims of this study is to characterize the dysregulation of
cell-cell junctions in the right ventricle of irreversible PH models and link the findings
to edema and RV failure. The spotlight is on the RV but the magnifying glass is on
Connexin-43 and a-Actinin along with lymphatic markers such as Lyve-1, Proxl, and
Podoplanin. Additional background on cell-cell junctions and the markers listed are
provided in detail in the subsequent sections. Determining whether these proteins
are up- or down-regulated and their contribution towards RV failure will aid in un-
derstanding the state of fluid stasis in the heart. Additionally, the findings will be
transformative as therapeutic measures taken to correct the mislocalization of cell-
cell junctions could improve and reverse PH by restoring fluid homeostasis in the
heart. Table 1.2 summarizes speculative interpretations of the markers in the setting
of Control vs. RV failure model of PH.
While most of the focus on understanding remodeling in PH has been over ECs,
and PASMCs; incorporating the focus on gap junctions and understanding their role
in vascular leak and interstitial fluid build-up in the cardiopulmonary area could
provide insight into novel therapeutic measures of tackling PH.
1.5 Gap Junctions
Intercellular communication is crucial to not only maintain homeostasis and
proper function within a multicellular organism, but also essential for its survival
[48, 49]. For example, cardiac function is dependent upon appropriate cell-cell com-
munication either directly through gap junctions, or indirectly via paracrine signal-
ing, i.e. soluble chemokines, extracellular vesicles, and autocrine signaling. While
gap junctions allow for electrical tether and metabolic coupling between neighbor-
ing cells in the heart, desmosomes and adherens junctions are anchoring junctions
that provide mechanical continuity [50]. The direct communication between cells is
established through passage of small solutes, metabolites and ions [51, 52, 53, 54],
Increasing evidence indicates that gap junction-mediated intercellular communication
10


facilitates not only diffusion of ions in neighboring cells but also embryonic develop-
ment, controlled cell growth and differentiation, transendothelial cell migration, and
pathogenesis of atherosclerosis [52, 53, 55]. In the heart, gap junctional proteins no-
toriously contribute to impulse propagation in both uniform and non-uniform tissue
[56]. In the lungs, gap junctions can contribute to calcium signaling between cili-
ated epithelial cells to coordinate ciliary beating [57], regulate secretion of pulmonary
surfactant by type II epithelial cells [58], and regulate propagation of calcium waves
along pulmonary vessels [59, 60].
The organization of the junctions at the intercalated discs (IDs) have been metic-
ulously investigated. Gap junctions at the IDs are organized together with desmo-
somes and fasciae adherents junctions [61, 62], The specialty of IDs is to integrate
communication from cell to neighboring cell and help maintain the electro-mechanical
function. As seen in Figure 1.4, fasciae junctions are located in vertical steps with
grooves present between them and desmosomes are situated in horizontal portions of
the IDs. Cx43 are located mostly in these horizontal regions of the discs.
1.5.1 Connexin-43
Gap junction channels are composed of proteins in the connexin gene family.
There are now over two dozen connexins identified in the mammalian gene pool [55];
they are abbreviated Cx followed by a numerical identification, eg. Cx40, Cx43,
Cx45, etc. The number signifies the mass of the protein in kilodaltons. Twenty one
members of the connexin gene family are found in human and 19 of them can be
grouped as orthologue pairs with connexin members found in murines [53]. Con-
nexins are integral membrane proteins consisting of four transmembrane domains,
two extracellular loops, one cytoplasmic loop, and intracellularly located C- and N-
terminal tails [52, 53, 55]. As illustrated in Figure 1.5, the C-terminal domain varies
in length, contains a large range of phosphorylation sites, and as a result, is thought
to play important roles in providing sites for protein-protein interaction. Cx43 is
11


a
Figure 1.4: Thin-section electron microscopy showing the organization of gap junction
with other adherent junctional complexes in the rat RV. (A) Step-like features of discs
(arrowheads) observed at the end of the isolated cell. (B) Fasciae adherent junctions
occupying the electron rich region. Gap junction appear to be connecting with fasciae
adherent junctions (arrows).
Severs, NJ et al. 2008
among the three main gap junctional proteins found in the heart; the other two being
Cx40 and Cx45 [63]. The C-terminal domain of Cx43 may represent the primary site
for phosphorylation, however, there are other gap junction proteins such as Cx36 and
Cx56 that show the cytoplasmic loops as their phosphorylation sites [64], It is possi-
ble that the different isoforms of Cx43 be responsible for exacerbating the severity of
PH or even play a role in causal aspects of PH. The data supporting this hypothesis
is severely lacking and investigations towards it would be beneficial for improving the
12


therapeutic approaches for PH.
a
Figure 1.5: Illustration of Connexin-43 arrangement in the plasma membrane.
EL, Extracellular loop; PM, Plasma membrane; CL, Cytosolic loop; CT, C-Terminal; NT, N-
Terminal
Although the illustration shows multiple sites that allow for different signaling pathways, we did not
evaluate them. The focus on the carboxyl terminal is used to show multiple phosphorylation sites.
Omasits et al., 2014
Six Cx subunits come together in the plasma membrane to form a hemi-channel,
dubbed connexon, that can dock onto another connexon situated in the plasma mem-
brane of a neighboring cell. This forms a complete gap junction channel. Connexin
isoforms can combine to form either a heterotypic or heteromeric gap junctional com-
plex, for instance: Cx43 and Cx40 form a heteromeric bond [65]. Gap junction
channels composed of a single connexin type are defined as homomeric. The com-
bination of connexins involved and their ability to interact to form connexons could
13


suggest varying degree of regulation between adjacent cells. Cx43 is now recognized
as the most expressed connexin, identified in at least 34 tissues and 46 cell types [55].
Because of its predominant presence over other connexins in most cell lines, more
is known about Cx43 biology. Normal expression in various cell types and tissues,
and aberrations leading to a variety of pathophysiological conditions are also known.
Gap junction proteins are continually forming and degrading. Such dynamism is evi-
denced by reports of connexin half-lives being less than 5 hours on cultured specimens
[66, 67, 68, 69, 70]. This suggests that the turnover along with the assembly of gap
junctions are likely essential for intercellular communication [55]. One focus of the
current study is Cx43 expression in the right ventricle as a mechanical failure marker
in the setting of PH.
1.6 Role of Cx43 in Support of the Working Hypothesis: Vascular Leak
and Increased Interstitial Fluid
So, where does gap junction CxjS fit in with respect to myocardial edema and
eventual right heart failure in PH?
As aforementioned, pericardial effusion and fluid build-up in the failing RV is one
of the key findings in pulmonary hypertension. Patients with PH, regardless of etiol-
ogy or any of the causes listed in Table 1.1, present with increased CVP [38]. Elevated
CVP adversely affects fluid balance in the myocardium as myocardial lymph outflow
is now experiencing an increased pressure. Increased filling pressure depresses the
CO; a 40% decrease in CO is reported with corresponding 3.5% increase in myocar-
dial fluid content [5, 39, 71]. The clinical significance of RV edema and pericardial
effusion is steadily gaining recognition but the data explaining the origin, mainte-
nance, and increasing severity caused by fluid imbalance is missing. This particular
area is of cardinal interest to both clinicians and researchers because of its association
with higher risks of morbidity and significance mortality [72], The prognosis is not
only poor but rather difficult to accurately assess. Elevated plasma brain natriuretic
14


peptide (BNP) levels, increased right atrial pressures, and low cardiac index are some
of the factors for poor prognosis of PH [72, 73, 34]. Additionally, RV dysfunction and
presence of pericardial effusion generally worsens the prognosis [74, 75, 76].
Figure 1.6: Functional consequences of abnormal Cx43 organization and expression
seen in remodeled heart. It is unclear whether the remodeling of the heart preceeds
the disorganization of Cx43 or if it is causal. Dotted lines represent potential causal
interplay, solid lines represent established associations.
There is no mechanistic data present for the RV relating edema to Cx43
mis-localization. The role of gap junctional proteins in inflammation is indicated
in that gap junction inhibitors promote vascular leak and transvascular cell migration
[6]. This suggests that gap junctions help in maintaining vascular barrier function.
Aside from being linked to arrhythmias, the functional consequences of abnormal
Cx43 organization in the RV has not been exhausted. It is speculated [77, 78] that
the connexins found in the lateral membranes of cardiac myocytes are likely to be in
the form of closed hemichannels. Decreased electrical coupling has been observed and
reported as a result of loss of gap junctions in ischemic and damaged cells, although
dye transfer of lucifer yellow between the cells was observed not to be decreased
15


[79, 80]. Transfer of fluid has been observed through lateralized gap junctions, which
suggests that Cx43 could be contributing to edema in the RV, along with interstitial
fluid stasis, pericardial effusion resulting in ineluctable failure of the heart (Figure
1.6). It is unclear when in the progression of PH that Cx43 lateralization may be
important. If it occurs, preliminary data, discussed in Chapter 3, shows that there is
extensive redistribution of Cx43 in preclinical animal models with severe PH but not
in models with mild PH. The course and extent of Cx43 redistribution is yet to be
determined.
The experimental model illustrated in Figure 1.6 follows the understanding that
Cx43 leads to arrhythmogenic events in the myocardium [63]. Additionally, decreased
plaque size and down-regulation of immunodetectable Cx-43 have been reported in
various human cardiac diseases [61, 81, 82, 83, 84], There is a paucity in data re-
garding incidence, mechanism, prognostic significance, and therapeutic treatments of
arrhythmias in patients with PH [85]. Since arrhythmias are not common in PH and
incidence may vary according to etiology, whether arrhythmias are consequence or
cause of RV failure in PH is still up for speculation. But as aforementioned, Cx43
dysfunction has been linked to arrhythmias in various cardiac diseases. When the
contractile apparatus of the ventricles malfunction due to tachyarrhythmias, as seen
in PH /citeRajdev2012, the heart is not able to effectively drain fluid affecting the
myocardial lymph outflow. We speculate that, as a result, myocardial interstitial fluid
is increased leading to prevalence of edema and pericardial effusion.
1.7 Connexins in the Myocardium
Cardiac myocytes, cardiac fibroblasts, endothelial cells and vascular smooth mus-
cle cells are the major cellular constituents of the heart. Initial studies in determining
the cell types, performed by Zak [86] and Nag [87] in 1973 and 1980 respectively,
quantified cell populations based on morphological characteristics of left ventricles of
rats. Recent studies have now included whole hearts and the evidence suggests that
16


the human heart comprises of 60-70% nonmyocytes and 30-40% cardiac myocytes
[88, 89]. Even though there are physiological differences between human heart and
the preclinical animal models, such as heart rate and total collagen [90, 91], the rat
model has become a standard preclinical model organism to study cardiac function
along with structural and functional changes that occur in pulmonary hypertension.
Cardiac myocytes are arranged in 2 to 5 cell layer thicknesses in a laminae sur-
rounded by a network of endomysial collagen [92, 93]. Studies that have investigated
the organization of cardiac fibroblasts have documented their presence within the
endomysial network [94, 95]. It has been well established that the cardiac myocytes
communicate amongst adjacent myocytes through gap junctions. However, little is
known about the cross-talk that occurs between the myocytes and the cardiac fibrob-
lasts. Similarly, little is known about the gap junctions that are present between
cardiac fibroblasts and other cell types found in the heart.
1.7.1 Fibroblast-Fibroblast Communication
Although the research documenting the distribution of fibroblasts in the heart is
in short supply, there is some evidence that fibroblasts are not uniformly distributed
throughout the heart [89, 96]. Multiple investigations have shown the regulation of
fibroblasts to varying degrees depending on pathophysiological conditions [94, 95, 97].
For example, fibroblast proliferation and tissue remodeling in general with spread and
infiltration of fibroblasts into unaffected tissue are characterization of multiple cardiac
pathologies. In cases of cardiac remodeling, as seen in pulmonary hypertension and
in post myocardial infarction, fibroblast numbers appear to be dramatically increased
[98]. The intercellular communications that occur between fibroblast-fibroblast con-
nection is presumed to be regulated by connexins, cadherins, and other unknown
molecules [55, 99]. However, the literature is unclear on whether connexins play a
major role in being the primary means of communication between fibroblasts.
17


1.7.2 Myocyte-Fibroblast interaction
Kohl et al. have studied fibroblasts in the sino-atrial (SA) node where fibrob-
lasts outnumber the myocytes. They found that fibroblasts communicate with the
impulse-generating specialized myocytes in the SA node and act as either capacitors
or conductors. This suggests that fibroblasts could affect the behavior of cardiac my-
ocytes in other regions of the heart. Direct interactions between cardiac myocytes and
fibroblasts have been shown to be regulated via communications between Cx40, Cx43,
and Cx45 [88, 89, 100, 101, 102], Connexins have been extensively studied in car-
diac myocytes and recent findings suggest that signals within the microenvironment
may regulate cellular hypertrophy [103]. Extending such studies to cardiac fibroblasts
would garner data that could suggest involvement of fibroblasts in three-dimensional
signaling.
Observations that fibroblasts are in contact with the myocytes does not neces-
sarily imply that there are direct signals being sent and/or received between the two
cell types. Cardiac fibroblasts are key players in normal heart development as well
as in diseased states. They contribute to structural, mechanical and electrical prop-
erties of the myocardium [100, 104], In diseased states of the cardiovascular system,
fibroblasts play a central role in myocardial remodeling by proliferating, migrating to
sub-endothelial areas of the tissue, and contributing to the change in the composition
of the ECM [89, 105, 106]. Cardiac myocytes are the contractile regulators of the
heart and thus, understanding the cross-talk that occurs between the two cell-types
could be a pivotal finding and could provide steps in obviating RV failure PH.
1.8 Normal Expressions: Location and Appearance
Gap junctions, adherens junctions, occluding junctions and tight junctions all
preferentially reside at the IDs [50]. IDs are also populated by non-junctional
molecules. The non-junctional fraction of these proteins include molecules that are
not involved directly in providing a physical continuum between adjacent cells. The
18


non-junctional fraction is comprised of cytosolic loop, n-terminal and c-terminal com-
ponents (cytoplasmic Cx43), and Cx43 en route to either assembly or degradation
[107].
Closed
Open
gap junction
transmembrane domains (M1-M4)
extracellular loops (El and E2)
cytoplasmic loop (CL)
terminal domains (N and Q
Connexln structure
a
Figure 1.7: Organization of connexin subunits to form the hemichannel subunit con-
nexon. Connexons have multiple connexin types forming either a heteromeric or a
heterotypic channel.
^Public Domain Image
Connexin-43 is characterized via immunofluorescence by punctate localization at
the IDs of cardiac myocytes. Gap junctions are normally expressed at the IDs in a
polarized manner and have relatively low density at the lateral sides [81, 108, 109].
The morphology of the myocyte and the ring like structural organization of gap
junctions [110] contribute greatly to the low electrical resistance and as a result,
conduction of action potential can propagate in both longitudinal and transverse
direction.
19


1.9 Aberrations
Connexin-43 has been studied in the settings of hypertrophic cardiomyopathy
(HCM), dilated cardiomyopathy (DCM), and ischemic cardiomyopathy (ICM) among
other end-stage congestive heart failures. In these cases the most prominent fea-
tures of gap junction remodeling included down-regulation of Cx43, reduction in
gap junctional plaque size, and increased heterogeneity of gap junction distribution
[107, 111, 63, 112, 113]. Other studies have reported neoformation of gap junctions
at locations other than the IDs [50, 114, 115, 116]. Most of the studies conducted
on connexins and their functions have been with respect to arrhythmias and other
electrical aspects/events. Thus, the process of gap junctional remodeling has been
described as a potential catalyst for cardiac arrhythmias [114, 83]. It can be argued,
however, that mis-localization may help maintain action potential propagation. The
pattern of remodeling simulated in computer modeling studies predict that reduc-
tion of Cx43 upto 40% in content is unlikely to have a major change in conduction
velocity [117]. Conduction velocity is defined as the speed of an electrical impulse
transmitted through excitable tissue, as in movement of an action potential through
the His-Purkinje fibers of the heart. The remodeling of cardiac myocyte gap junc-
tions characterized by decreased Cx43 in the IDs, and its redistribution throughout
the plasma membrane of a cardiac myocyte is often known as lateralization.
It is well established that organization and distribution of gap junctions is re-
markably different across various cases of cardiac disorders [61, 82], Sasano and
collaborators have studied the expression of Cx43 in hypertrophied right ventricles of
rats with pulmonary hypertension. They reported internalization and dephosphory-
lation of Cx43 in their preclinical animal models [118]. As seen in Figure 1.7, there
are multiple sites on the C-terminal of Cx43 where phosphorylation and dephospho-
rylation can occur. However, it can be speculated that the connexin domains found in
the cytoplasm are the precursors of the gap junctional complexes on their way to the
20


plasma membrane, a process that could be correlated with diseased state of the heart.
Since inflammation increases Cx43, as shown by Navab et al. [119], immunoreactive
connexins found in the cytoplasm could suggest neoformation of Cx43. Cx43 as a
mechanical marker is highly understudied in the RV. Mechanical marker is a term
used to infer function of the rat RV through experiments conducted for this project,
since we are not directly measuring the function. The functional consequences of ab-
normal Cx43 expression and organization in the right heart could include promotion
of vascular leak and interstitial fluid build-up that induces hypertrophic growth of
individual myocytes. A prolonged state of hypertrophy increases risks of heart failure
[120, 121] a condition seen in pulmonary hypertension.
1.10 Cell Specific and Lymphatic Markers
The cardiac cell specific, fibroblast specific markers, and lymphatics specific mark-
ers described in the following subsections, are chosen to aid in further strengthening
the hypothesis. By investigating each of the following markers in their respective tar-
get tissue, we attempt to answer the question of decreased lymphatics in the heart,
the degree of fibrosis, and activity of apoptosis and proliferation, all leading to the
problem of edema and mechanical failure of the right heart. Their potential contri-
bution to edema are listed in the respective subsections. Use of a single marker could
lead to misinterpretations, for instance, markers that label lymphatic vessels such
as LYVE-1 also label blood vessel and macrophages. To avoid this pitfall, multiple
markers are used for IF imaging of heart and lung tissues. The antibodies listed in the
following sections are among the most useful markers to investigate their respective
proteins of interest.
1.10.1 Cardiac Myocyte Markers: o-Actinin and Dystrophin
o-Actinin and Dystrophin both belong to the spectrin superfamily, a conserved
family of actin-binding proteins (ABP) that maintain plasma membrane integrity and
21


cytoskeletal structure [122], Of the four a-actinin genes found in humans [123, 124],
a-actinin-2 is expressed in the heart, more specifically in the sarcomeric Z-discs and
analogous dense bodies [125]. a-actinin binds to actin in a manner that is calcium
ion independent and also localize in the IDs of cardiac myocytes [125, 126]. a-actinin
crosslinks filamentous actin and titin molecules from adjoining sarcomeres at Z-discs.
Thus, a-actinin contributes to the stability of the sarcomeric structure.
Dystrophin is a peripheral membrane protein that identifies cardiac myocyte and
is known to show continuous distribution at the lateral plasma membrane of the cell
population [62], Although dystrophin is a membrane protein for cardiac myocytes
and contribute in forming part of the transverse tubules penetrating into the cell,
they are not present in the IDs.
Through immunofluorescence staining, a-actinin and dystrophin will inform us
of a couple different things. First, paired with DAPI staining that labels nuclei, the
markers will allow us to visually localize cardiac myocytes from other cell types in
the myocardium. Second, the prevalence of these proteins will allow us to quantify
the number of cardiac myocytes and how the numbers differ between our preclinical
models, a-actinin and dystrophin are representative of the contractile apparatus in
the myocardium. Disturbances in these proteins in the diseased models would suggest
aberrations in the myocardial contractile function that would lead to accumulation
interstitial fluid and therefore edema.
1.10.2 Cardiac Fibroblast: Narrowing Down a Specific Marker for
Fibrosis and Cardiac Remodeling
Cardiac fibroblasts comprise the majority of the non-myocyte population in the
heart [88, 89]. Fibroblasts help maintain the structural, mechanical, biochemical, and
electrical properties during normal cardiac function [127]. However, relatively little is
known about the genetic pathway and other factors that contribute to the differentia-
tion along this particular cell population [128]. It is suggested that cardiac fibroblasts
22


are a heterogeneous population that originate from multiple sources in the body [129],
for instance bone-marrow derived cells, endothelial cells, and the epicardium. As a re-
sult, it has been difficult to narrow down a marker that specifically represents cardiac
fibroblasts. It can be rather difficult to asses RV fibrosis and the role of fibroblasts
in the pathological conditions when the assessment is impeded by absence of spe-
cific markers. For this study, fibroblast specific protein (FSP), transcription factor-21
(TCF21), and prolyl-4-hydroxylase (P4H) were used to distinguish fibroblasts in con-
trast to the cardiac myocytes and attempt to asses fibrosis with respect to prevalence
of fibroblast in regions of the RV.
Fibroblast specific protein (FSP), also known as S100A-4, labels filament associ-
ated calcium binding proteins expressed by fibroblasts in hbrotic tissues [130, 131].
Through differential hybridization, FSP has ben identified in fibroblasts but not in
epithelial or embryonic endoderm [132], However, over the years the specificity of
FSP towards fibroblasts has been challenged suggesting that other infiltrative cell
types during injury also express FSP. Some of these cell types are dendritic cells
[133], vascular smooth muscle cells [134], lymphocytes [135] and macrophages [136].
Transcription factor-21 (TCF21) is a class II basic helix-loop-helix (bHLH) tran-
scription factor that is known to bind deoxyribonucleic acid (DNA) through the con-
sensus E box sequence (CANNTG) as a heterodimer [137]. TCF21 has been exten-
sively used in identification of epicardial and proepicardial cells [138], however, its
function is less known in cardiac development. What makes TCF21 potentially a
good marker for cardiac fibroblast is the evidence showing that TC21 null mice fail to
develop cardiac fibroblasts [138]. Acharya et al. show that TCF21 expressing epicar-
dial cells are multipotent and are able to develop into either coronary vascular smooth
muscle cells (cVSMCs) or cardiac fibroblast with TCF21 becoming restricted to the
cardiac fibroblast lineage over time. Additionally, through lineage tracing shows that
TCF21 is excluded from the cardiac myocyte population [138].
23


Prolyl-4-hydroxylases (P4H) are enzymes that catalyze irreversible posttransla-
tional modification reactions to either alter protein confirmation and protein-protein
interaction, or enable further modification [139]. The enzymes P4H reside in the
endoplasmic reticulum and play a critical role in biosynthesis of collagen [140]. An-
tibodies against P4H identify fibroblasts because fibroblasts produce collagen [141].
P4H are able to label portion of procollagen molecule cleaved before secretion by
fibroblast in response to injury, thus suggesting that the marker is best utilized in
labeling fibroblasts that are actively dividing and not the subpopulation of this cell
type that are quiescent [141].
1.10.3 Apoptotic and Proliferation Markers: Caspase-3, Ki-67
Current and past studies suggest that caspases are essential regulators of prote-
olytic cascades that occur in response to a cell death stimulus, with caspases having
roles in regulation as well as execution of apoptosis. Caspase-3 (CASP3) is the best
characterized member among the 10-member subfamily of caspases [142], Apoptosis
has been shown to contribute to loss of cardiac myocytes and progressive decline in LV
and RV function in congestive heart disease[143, 144, 145]. Apoptosis is a type of cell
death that is distinct from cell necrosis [146, 147], associated with not only CASP3
[148] but also with the cleavage of protein kinase C [149], polyADP-ribose polymerase
(PARP) [150], and certain other proteins [151]. CASP3 is detectable by immunoflu-
orescence in cardiac myocytes [152], Apoptosis has been linked to PH, mainly in
idiopathic pulmonary fibrosis (IPF)-associated PH. Farkas et al. have shown that EC
apoptosis releases vascular smooth muscle cells (VSMC) growth factors in hbrotic
regions that may contribute to augmented pulmonary arterial muscluarization [153].
Caspase-3 activity will inform us of apoptotis in our preclinical models. This is im-
portant since EC apoptosis is shown to be detrimental in organs such as the lungs
and the heart and are potential contributors to PH [153].
24


Ki-67, also referred to as MKI67, is a protein that is associated with cell prolif-
eration [154], Evidence showing that Ki-67 is present in all of active stages of cell
cycle (Gi, S, G2, and mitosis) but is absent from the resting phase (Go) makes it
an excellent marker in determining the growth-fraction of cell population [155]. Pro-
liferation, like apoptosis, contributes to formation of lesions due to structural and
functional changes to the architecture of the walls of the pulmonary arteries. In
PH, this leads to increased muscularization of the pulmonary arteries and peripheral
vessels, formation of neointima and formation of plexiform lesions [156].
These specific markers were chosen to qualitatively and quantitatively assess the
varying degree of cell death and proliferation in the progression of PH. Furthermore,
any down-regulation or up-regulation in proteins of interest can be correlated with
whether the increase or decrease in cell growth-fraction is contributing to their ex-
pression levels.
1.10.4 Markers for Lymphangiogenesis and Angiogenesis: Lyve-1,
Podoplanin, Proxl
Lymphangiogenesis and angiogenesis are essential for normal development and
physiological processes [157]. More recently, the pathological processes in the car-
diopulmonary region have garnered interests in lymphangiogenesis and angiogenesis
aspects and as a result, been recognized as potential therapeutic targets for diseases.
The difficulty in using markers for lymphatic and/or blood vessels arises from cells
that make up the respective vessels are not homogeneous and show various pheno-
typic variation depending on location and functional state (inflammation vs normal
growth for instance) [157].
Lyve-1 stands for lymphatic vessel endothelial hyaluronan receptor 1 and is one
of the best characterized markers for lymphatic endothelial cells (LECs) [158]. Lyve-
1 strongly identifies on the entire luminal and abluminal surface of the LECs as
shown by Baluk and McDonald. The protein acts as a receptor and binds to both
25


immobilized and soluble hyaluronan suggesting that Lyve-1 may function in lymphatic
hyaluronan transport [159].
Podoplanin is expressed in most lymphatic vessels where it may contribute in cell
adhesion [160]. Podoplanin is a protein detected in kidney podocytes [161], and is
also known to be expressed in epithelial Type I cells in the lungs [162], The specific
function of Podoplanin has not been determined but it has been proposed as a marker
for lung injury. Podoplanin is found in skin carcinomas [163] among other pathologic
situations, suggesting that the distribution of this marker is diverse in human tissue.
However, we have data, Figure 1.8, showing co-localization of Podoplanin and Proxl
(another lymphatic marker discussed next) which suggests that Podoplanin is an
adequate choice for a lymphatic marker, at least in the region that our study is
interested in.
Figure 1.8: Colocalization of Podoplanin and Proxl in the tertiary lymphoid tissue
(BALTs) showing positive correlation between the two lymphatic markers.
Proxl, also known as Prospero homeobox 1, is a transcription factor that labels
LECs in early stages of embryonic development [164, 165]. Proxl labels the nuclei
of LECs and is speculated to be a master control gene influencing the expression of
26


other lymphatic markers [164, 165]. Proxl has been found to consistently expressed
in adult lymphatic endothelium [166] and its expression is fairly correlated with that
of Podoplanin as evidenced in Figure 1.8. However, a small set of capillary and
vasculature with irregular smooth muscle profile lining have been shown to express
endoethelial Proxl [165, 166].
The overall function of the lymphatics is to collect leaked plasma and interstitial
fluid for their return to blood vasculature. The lymphatics are not only responsible for
transport of fluid but also the uptake and degradation of dissolved macromolecules,
one of which is hyaluronan [167, 168]. We expect to see decreased lymphatics in the
RV of PH rats. As a result, the drainage apparatus of the lymph outflow is disturbed
causing interstitial fluid buildup and edema. Additionally, hyaluronan homeostasis
is affected. Hyaluronan undergoes constant turnover, a process which comprises of
removing the macromolecule from the tissue into the afferent lymph, degradation
with the lymph nodes, and removal through efferent lymph [169]. Because hyaluronan
degradation releases products that are potentially pro-inflammatory [170], inefficient
removal of hyaluronan could be contributing to the progression of PH. We are not
looking into hyaluronan metabolism for this project but it should be kept in mind of
their role in the progression of PH with respect to lymphatics in the RV.
27


2. Methods and Materials
2.1 Animal Models
Sprague Dawley and Wistar rats were used to harvest the lungs and hearts inves-
tigated for the study. A control group of rats were acclimated to Denvers elevation
of 5,285 feet. The control group was contrasted to three diseased groups of PH,
of which two models studied have irreversible and ineluctable RV failure. The dis-
eased models are: 1) reversible PH induced due to chronic hypoxia from exposure to
hypobaric chamber stimulation of an elevation of 18,000 feet, 2) PH induced due to
monocrotaline (MCT) administration, and 3) PH induced with Sugen5416 followed by
exposure to chronic hypoxia. The stimulation of high elevation approximately mimics
pressure value of 0.50 bar or 380 torr. Control group of rats experience pressure of
0.83 bar or 621 torr.
MCT, an 11-membered macrocyclic pyrrolizidine alkaloid (PA) derived from the
seeds of Crotalaria spectabilis plant, is activated into a reactive pyrrole in the liver
by cytochrome P-450 [171, 172], The reactive pyrrole metabolite dehydromono cro-
taline (MCTP) injures pulmonary endothelial cells [173] causing them to develop
megalocytosis, displacement of endothelial nitric oxide (NO) synthase, and decreased
cell-surface/caveolar NO [174], MCT induced loss of membrane proteins stimulate
proliferation, activation of anti-apoptotic factors, and dysregulation of NO signaling
[175]. MCTP acts as a vessel obliteration agent, obstructing the pulmonary arteries
and increasing vascular resistance. The MCT induced endothelial cell damage, in-
terstitial pulmonary fibrosis [176], and pulmonary arterial medial hypertrophy [177]
among other characteristics lead to changes in the lung vasculature and irreversible
remodeling in this experimental model of PH. Studies show significant changes in
pulmonary artery pressure, medial thickness of the small pulmonary arteries, and RV
hypertrophy occurring 3rd and 4th weeks post MCT administration [178, 179]. Thus,
we investigated rats at those time points for out RV failure models. Additionally, the
28


MCT syndrome induces myocarditis in both the left and right ventricles, acute lung
injury, interstitial pulmonary fibrosis, and hepatic venooclusive disease [177].
Sugen5416 (more commonly known as Sugen) is a small molecule receptor ty-
rosine kinase (RTK) inhibitor that was discovered through a screening process de-
signed to identify compounds that inhibit vascular endothelial growth factor (VEGF)-
stimulated endothelial proliferation [180]. Sugen5416 acts to inhibit VEGF receptorl
[181] and VEGF receptor2 [180], and prompts pulmonary endothelial cell apoptosis
and loss of small lung vessels [182], Sugen5416, when combined with chronic hypoxic
environment, causes severe angio-obliterative PH and right heart failure [183]. The
VEGF receptor blockade induces endothelial cell apoptosis which leads to progressive
proliferative endotheliopathy [184], This Sugen-hypoxia (SuHx) model is an impor-
tant preclinical model of PH [185, 183] as it causes angio-obliterative lesions in the
pulmonary arterioles that are similar to the plexiform lesions found in human cases
of idiopathic PH [186].
For the purpose this research, preliminary data include SuHx and MCT models
in contrast to controls but later studies incorporate only the MCT preclinical model
of RV failure PH juxtaposed to the control groups.
2.2 Obtaining Tissue
For staining purposes, the heart of the rats were harvested from the preclinical
models. The right and the left atria are excised out, followed by the base and the apex
being separated from the main body by cutting the desired region of the heart, as
shown by the cut marks in Figure 2.1. The apex is included in with the desired region
at times when the rat heart is relatively small and fits into the cryomold dimensions.
The partition that remains is then placed into the OCT solution and is allowed to
freeze. OCT stands for optimal cutting temperature. The OCT block containing the
tissue is then frozen and stored in -80C. The face on the block seen in the figure is
sliced and collected on a frosted glass slide. The thickness of the heart section cut is
29


in the range of 5 mm to 7 mm. The cut section when sliced and placed on to a glass
slide contains both the LV and RV with the former being the circular larger structure
seen in the figure. The orientation of the specimen placed on the OCT block is such
that, when sliced onto a glass slide at desired thickness range of 5 to 150 fim and
multiple subsequent slides are imaged, the observer is making their way from the base
to the apex, i.e superior to inferior.
Rieht Ventricle
-Left Ventricle
Tissue partition as laid out on an OCT
Block
a
Figure 2.1: Heart tissue obtained from the preclinical models include the right and
left ventricles. The right and left atria are removed prior to cutting the tissue desired
for our assays. The whole heart on the other hand is utilized to prepare homogenates
for Western blot experiments. The Western blot assay performed for this study used
RV homogenates.
Lohani, 2015
2.3 Western Blot Protein Assay
Western blot, also referred to as the protein immunoblot, is a widely utilized
analytical assay for the purpose of detecting specific proteins in a sample of tissue
homogenate. This particular technique was performed for heart tissues in control and
RV failure models of PH. Samples taken were from right ventricle tissue only. Solid
30


tissue were first broken down mechanically using a homogenizer. During this step,
the tissue samples were handled in extreme cold temperatures to prevent any dena-
turing of proteins or specimen degradation. Protease and phospatase inhibitor tablets
(PhosSTOP, cOmplete Mini; Roche Holding AG) were added to prevent digestion of
the homogenate and preserve phosphorylation state of proteins. Although addition of
phosphotase inhibitor to the homogenate prevents phosphorylation of proteins, dif-
ferent isoforms of proteins could already be present at the time of tissue retrieval.
The proteins in the sample were then separated by molecular weight (in kiloDaltons:
kDa) via gel-electrophoresis. The running buffer comprised of MES-NuPAGE-SDS
(polyacrylamide gel electrophoresis sodium dodecyl sulfate) that maintains polypep-
tides in denatured forms, as a result, allowing for separation of proteins by molecular
weight. The setting used for this step of the western blot assay was: 1 hour run-time,
200 volts. Samples were loaded into a 10-well gel with triplicates of control, MCT and
Su5416-hypoxia samples loaded in well slots 1 to 9. The 10th well was loaded with
a ladder that showed colored bands signifying proteins of different molecular weight
for convenient deduction of sizes in kDa. The proteins, once separated, were then
transfered onto a polyvinylidene diflouride (PVDF) membrane to allow for antibody
detection via a process called electro-blotting. The transfer was confirmed with Pon-
ceau staining of the membrane. To prevent non-specific interactions and bindings, a
blocking solution of 5% non-fat dry milk in Tris-buffered saline (TBS) with Tween-20
was used. Blocking was performed for 1 hour followed by multiple wash cycles before
application of primary antibody. The antibodies used along with their appropriate
dilutions are provided in Table 2.1.
A two-step incubation procedure was adapted for probing of proteins in the mem-
brane: primary antibody incubation for the first step, followed by rinsing of mem-
brane for any unbound primary antibodies and secondary antibody incubation for
the second step. Both incubation intervals were performed for 1 hour each. The
31


secondary antibody used was conjugated with horseradish peroxidase (HRP) allow-
ing for multiple secondary antibodies binding to one primary antibody and cleave a
chemiluminescent agent to produce luminescence that is proportionate to the amount
of protein present. Enhanced version of the chemiluminescence reaction (ECL)
was used to detect the light emitted when the luminol reagent is oxidized by HRP
present in the membrane.
Table 2.1: Antibodies used for Western Blot Assay
Antibody Target Tissue Vendor Species Clone Concentrations
Primary Antibodies
a-Actinin Heart Sigma Mouse A7811 1:3,000
Connexin-43 Heart Novus Rabbit NBP 1-67530 1:500
GAPDH Heart Abeam Rabbit ab37168 1:10,000
Secondary Antibodies
anti-mouse HRP Heart Millipore Goat AP-124P 1:10,000
anti-rabbit HRP Heart Santa Cruz Goat sc-2004 1:10,000
GAPDH, Glyceraldehyde 3-phosphate dehydrogenase; HRP, Horseradish peroxidase
The membranes used to probe for the proteins of interest were stripped in
Restore stripping buffer (Thermo-Scientific) then incubated with Glyceralde-
hyde 3-phosphate dehydrogenase (GAPDH). GAPDH is an enzyme of 37 kDa and
was used for normalization during the quantification step of western blot. The blot
was quantified using ImageJ. The built-in Gel Analyzer function was used to mea-
sure the density of the of immunoblots.
2.4 Immunofluorescence Staining
Histological sections of the rat hearts were stained with fluorescent-labeled anti-
bodies to investigate the antibody-antigen complexes in control and diseased condi-
tions. The right ventricle of the rat heart was the primary area of interest.
32


2.4.1 IF and Antibody Staining Techniques
The histological sections range from a thickness of 5 micrometers to 200 microm-
eters. The thin sections were standards for immunofluorescence (IF) staining when
optimizing concentration of antibodies and obtaining preliminary data. Thicker sec-
tions facilitated the use of confocal microscopy and utilized the advantageous feature
of optical sectioning. Hence, confocal provided us with better anatomical context
of the tissue as well as superior images over a conventional IF microscope. These
sections were cut using HM 505 E Microm cryostat. Tissue samples, once sliced
onto glass slides, were fixed with 1:1 solution of Methanol:Acetone for 10 minutes
at -20C. Tissue sections with thickness > 20/xm were fixed for 2 hours minimum.
After fixation, the slides were allowed to air-dry followed by re-hydration using IX
PBS (Phosphate-buffered saline). The slides were then blocked with 1:1 solution of
FBS:PBS (Fetal bovine serum: Phosphate-buffered saline). Incubation time varied
with thickness of the tissue sample: 30 minutes for sections ranging from 5 to 20
micrometers, 2 hours for 50 /ml sections and approximately 8 hours for sections north
of 100 /mi. The varying incubation times, especially for the thicker sections, were
incorporated into the staining protocol to allow for adequate and even staining of
the heart tissues. Because we were labeling muscularized tissue, longer incubation
times allowed for deeper penetration of the antibodies throughout the thickness of
the tissue. The primary antibodies, diluted in 5% FBS in PBS solution, were used
to label the respective target proteins. The primary antibodies used along with their
appropriate dilutions and whether they were used as identifying markers for the heart
or the lungs are provided in Table 2.2.
The secondary antibodies used were Alexa Fluor 488 and Alexa Fluor 594 at
dilutions of 1:750 in 5% FBS:PBS solution. The sections were incubated overnight at
4 degree Celsius for primary and secondary incubations. For thicker sections, how-
ever, an additional dose of primary antibodies was administered post wash following
33


Table 2.2: Primary antibodies and their optimum dilutions
Antibody Target Tissue Vendor Species Clone Concentrations
a-Actinin Heart Sigma Mouse A7811 1:250
Aquaporin-1 Heart Claudin-5 Heart Connexin-43 Heart Mlllipore Mouse MAB3068 1:200
Heart Dystrophin Heart Abeam Rabbit abl5277 1:300
Ki-67 Heart LYVE-1 Lung RXD Systems Sheep AF7939 1:20
Podoplanin Heart P4H Heart Acris Mouse AF5110-1 1:100
PROX-1 Heart TCF21 Heart Novus Rabbit NB 11037257 1:500
LYVE, lymphatic vessel endothelial hyaluronan receptor; P4H, Prolyl-4-hydroxylase; PROX,
Prospero homeobox; TCF, Transcription factor
Some of the markers were investigated in the lung tissues as well (data not shown).
the overnight incubation to allow for adequate saturation of the tissue with the im-
munolabels. This increased the incubation time with primary antibodies to 24 hours.
Performing similar steps with the secondary resulted in the non-specific binding. In
other words, the secondary antibodies were labeling proteins in the entirety of the
section and not just the target tissues labeled with primary antibodies, or oversat-
urating the protein on interest. For sections 5 nm to 20 nm, we found that it was
unnecessary to incubate the tissue overnight for secondary antibodies. Results were
optimal with the specimen incubated with primary antibodies overnight regardless of
the thickness of the tissue. The slides were thoroughly washed with IX PBS mul-
tiple times between the incubation step of primary antibodies and the introduction
34


step of secondary antibodies. Immunolabeled sections were then followed with an
additional step of washing post incubation of secondary anibodies and mounted with
VectaShield mounting solution with DAPI (Vector Laboratories Incorporated). The
sections were then examined under a Zeiss fluorescent microscope. The confocal im-
ages were acquired using the Zeiss LSM780 and ZEN digital imaging system (Carl
Zeiss Incorporated, Thornwood, NY). For the control and the RV failure MCT pre-
clinical models, IF staining was performed in at least triplicates, on the panel of five
control and four MCT models.
2.5 Confocal Microscopy
Confocal microscopy technique offers several advantages over wide-held mi-
croscopy. Zeiss 780 LSM was used to obtain high-resolution images in contrast to
conventional fluorescent microscopy. The confocal microscopy allows for production
of thin serial optical sections in the range of 0.5 to 1.5 /mi through tissue that have
a thickness of 50 /mi or greater. Additionally, the image information gathered is re-
stricted to a well-defined plane, contrast is dramatically improved as the background
fluorescence and auto-fluorescence are reduced, and signal-to-noise ratio (SNR) is im-
proved. Serial optical sections were acquired at the interval of 1 /mi, 40x objective
magnification, and 1024 x 1024-pixel image size. This yields a pixel size of 0.21 /mi.
Additionally, images where a zoom-factor of 2 was applied in ZEN 2012 software,
1024 x 1024-pixel image size with similar settings yielded a pixel size of 0.104 /mi.
2.5.1 Rationale: Out of the Static Studies of Thin Sections
Confocal microscopy has turned out to be a popular technique to be utilized
in recent years, due in part to the convenience of obtaining high-quality images for
tissue samples prepared for conventional fluorescent microscopy. For several reasons,
we wanted to have a global perspective when looking at the harvested tissue. First,
it is difficult to examine heart tissues from humans, as they are not readily available.
Adding to that is the limitation of examining the specimen in in vivo settings. In vivo
35


studies are emerging and heart tissues are being engineered for investigation in close to
natural conditions. Conventional methods for immunofluorescence staining typically
include thin sections of specimen. We stained heart tissues of thickness ranging upto
200 pixi with efficient staining of 40% of the tissue, approximately 80p.m. The primary
reason for choosing thick sections was due to the size of cardiac myocytes. A healthy
adult cardiac myocyte is 100-150 pm long and around 10-25 pm [62], We wanted
to be able to capture as much information as possible from the RV and since the
primary location of interest is the cell-cell junctions between either cardiomyocytes
or cardiomyoctyes to fibroblasts, it was imperative of the team to attempt staining
thicker sections of the specimen.
Figure 2.2: (A) Immunofluorescence stain of a:-actinin with longitudinal orientation
with the z-discs of the sarcomere distinguishable whereas (B) helps explain the diffi-
culty encountered when dealing with heart tissue. The complex stereology of heart
tissue causes the observation view to change from longitudinal (yellow arrow) to
transversal sections (green arrow).
36


2.5.2 Why 3D over 2D: Whorling Effect/Stereology
Even with thick sections, the stereology and the orientation of the heart tissue
cut onto glass slides were unpredictable. Figure 2.2 shows the different orientations
encountered when observing the heart tissue. Figure 2.2 (a) shows the textbook
representation of heart tissue where the cardiomyocytes are seen longitudinally. The
stain of o-Actinin clearly shows that the cardiomyocytes are oriented longitudinally.
Figure 2.2 (b) on the other hand represents an orientation that shows cross-sectional
view of the cardiomyocytes in a transversal plane. To keep the observation and the
plane of visualization consistent on thin 2D sections is rather difficult if not impossible
as we visually observe all the different planes in between. With 3D reconstruction of
the anatomy of the right ventricle, we are determined to explore the cellular makeup
of the complex structure of the RV, junctional lateralization and dysfunction, water
channels, and up- or down-regulation of lymphatics. With 3D reconstructed images,
we wold be able to characterize gap junction plaques along the myocytes terminus.
Furthermore, with 3D reconstruction we would be able to get visual perspectives from
different planes otherwise limited by 2D images.
37


3. Results
3.1 Assessment of PH in Preclinical Models
Figure 3.1 shows the macroscopic representation of the morphology of heart tissue.
Without any further quantitative analysis, we are able to observe that the RV is
presenting with characteristics of hypertrophy in heart tissue from PH rats. The
thickness in the RV free wall is increased, and a change in the LV wall thickness
is observed. The cross sections of the tissue seen in the figure were taken between
the base and the apex as illustrated in Figure 2.1 in Chapter 2. The MCT-treated
heart harvested from the preclinical models is severely muscularized where the LV
chambers volume capacity appears to be affected due to hypertrophy. These images
were obtained during the sectioning phase of experiment when the heart tissue is
sliced onto a glass slide. The microscopic histological images of the RV stained with
immunofluorescence antibodies follow in the coming subsections.
Control
A
a
Figure 3.1: Macroscopic representation of morphology of the heart from preclinical
rat models. This figure shows cross section of (A) control and (B) tissue 4 week post
MCT administration
IVS, interventricular septum
38


Additionally, pentachrome stains were performed on 5 /tm thin sections of control
and 3 week MCT groups. Figure 3.2 shows a control slice juxtaposed to a 3 week
MCT tissue slide.
Figure 3.2: Macroscopic representation of pentachrome stain on single 5 /mi thin
sections of the heart. This figure shows cross section of control, (A) and 3 week post
MCT administration, (B)
Myocardial edema for MCT- and SuFLx-treated rats was measured in the RV
as described by Laine et al. The amount of myocardial edema, also referred to as
the extravascular fluid (EVF) was obtained from the unitless blood-free ventricles
[71]. The total tissue water content comprised of vascular water, interstitial water
and cellular water. A spectrophotometric correction was applied for blood volume
since the blood volumes variability may significantly change measurement parameters
throughout the experiment. Additionally, myocardial edema in the LV and RV of rats
in the other two diseased models were measured as well. The formula for the EVF is
as follows:
EVF
(wet weight dry weight)
dry weight
The presence of fluid in the RVs of MCT-treated preclinical models was signifi-
cantly higher than those of the controls. An increase in myocardial interstitial fluid
39


from a control EVF of 2.9 to 3.7 represents a change from water content of 75% to
79%. Similar results were found for the LV of MCT-treated rats where the LVs of
MCT-treated rats had significant amount of edema in comparison to hypoxic and
Su5416-hypoxia rats. The RV of rats treated with Su5416-hypoxia also showed sig-
nificantly more edema than controls. The findings are summarized in Figure 3.3.
o'
a
Figure 3.3: Measurements of myocardial edema for RVs and LVs in various preclinical
models of PH. Values are reported as mean St. Dev
p<0.05, statistically significant; St.Dev, standard deviation; mPAP, mean pulmonary arterial
pressure
The presence of edema in the RVs of the MCT-treated rats was also determined
via T2-weighted magnetic resonance imaging (MRI). As shown in Figure 3.4, the 4
week MCT-treated rats show higher prevalence of fluid in the RV. The T2-weighter
MR images shown here are precursors for the EVF data provided in Figure 3.3. EVF
data was compiled from the same animals from which we obtained the T2-weighted
MRI images.
40


T2-weighted MRI findings and EVF measurements positively correlate. A pos-
itive advantage of using T2-weighted MRI is the non-invasive and non-lethal route
of obtaining data. Therefore, this method can be used on not only rats, but also
humans. This is a strong advantage over using EVF to quantify the presence of my-
ocardial edema in our preclinical models as the tissue taken from rats is not viable
for additional experiments following measurement of EVF.
Control
*
V, i 1 2 *
t W2S00 LIZW 4 £)
lV
Scan 3 Q J S11 50/t SO rmi
Slice 12/20 FOV 640 cm
Echo i/i MTX 2S6
FlSP (pv*i) Pd* *1 IS nm F
TR 3 0 nw Car**/Control 1
TE: ISfW FA 60 0 deg TA DttS22*0Q1eii WEX TG FiSP.RaLHeartCinoained). 3 1 Crane 1
4wk MCT on
' mb W2622 i
L 1*11 ^ T
i 0 -
Scan 3 Si 1 son SO mn
Slice 12/24 FOV B40
Echo 1/1 MTX 256
FlSP (pvm) Pos 6 03 nm H
TR 3$ hypoxic 1
TE: 19 *tt FISPJtaLHeaitCmodiftetfi, 3 1 *njr* 1 FA 66 0 deg TA 0h6m27sl21iw NEX 1G
Figure 3.4: T2-weighted MRI showing edematous RV for MCT-treated Rat (right)
compared to the control (left). Images acquired using Bruker 4.7 Tesla MR.
Since interstitial fluid stasis is associated with inflammation and fibrosis in some
diseased settings, our lab group proceeded to asses hbrotic changes in the RV of
hypertensive rats. Massons trichrome staining was performed and extensive collagen
deposition was observed in the MCT-treated rats. The increased transmural collagen
found in the RV positively correlated with the presence of EVF. We firmly believe
that edema preceeds fibrosis. As shown in Figure 3.5, the RV of the 4 week MCT is
hbrotic compared to the control. The distinction is present between the rat RVs in
control and the MCT-induced PH models: the ventricle is edematous, as shown by
T2-weighted MR images and confirmed via EVF measurement, and fibrosis is present
in tje PH ventricles. The correlation is present between edema and fibrosis, leading
us to believe that the extensive collagen dposition in the MCT-treated RVs follows
41


accumulation of interstitial fluid and myocardial edema, Figure 3.6.
Figure 3.5: Massons trichrome stain assessing fibrotic change in control (top left)
vs. hypoxic (top right) and MCT (bottom row). Flypoxic rats demonstrate a similar
pattern of modest collagen content in the RV as control. Elevated collagen deposition
in the RV of MCT-treated rats correlates positively with the presence of EVF.
>
A tA 4>
c
w ft) c
t/> "D 3
O "5
-D X OJ
Ll. Q.
0) 15
L- 03
00
15'
10-
R2 = 0.7102
p = 0.0006
*
5-


2.0
2.5
3.0
-i--------r
3.5 4.0
Extravascular fluid
(W-D)/D
i
4.5
Figure 3.6: Correlation between EVF and Fibrosis supports our speculations about
edema preceeding fibrosis in PF1.
42


3.2 Immunofluorescence Staining: Qualitative Results
3.2.1 Connexin-43 Organization and Expression in the Normal Heart
The characteristic distribution of gap junctional proteins Cx43 in the left ventric-
ular myocardium is well established. Cx43 are characterized by punctate localization
at the IDs. As seen in Figure 3.7, the contractile myocytes of the right ventricle are
extensively interconnected by clustures of Cx43 (green fluorescence) containing junc-
tions. The longitudinal section seen in Figure 3.7 shows gap junctional protein Cx43
appearing in rows, corresponding to edge-on viewed IDs. Red fluorescence represents
o-actinin and DAPI represent nuclei. Cx43 appear to be well demarcated at the IDs,
Figure 3.7 (B). In Figure 3.7 (A), we observe Cx43 expression at locations other than
the IDs. As shown in Figure 1.4, the isolated rat RVs have vertical step-like feature.
Severs et al. have shown via thin-section electron microscopy that at these locations
in the periphery of the cardiac myocytes, there are components of gap junctions and
desmosomes connected with fasciae adherent junctions. These are referred to as ex-
tended components of the IDs [61]. Thus, the immunodetectable Cx43 observed at
what seem like lateral surfaces could be gap junctions present at the extended IDs.
3.2.2 Gap Junction Remodeling in RV Failure PH
Disease related alterations in myocardial gap junction organization have gained
wide attention, especially in the connexins contribution to abnormal propagation
of action potentials and arrhythmia [82, 84], These contributions could be due to
decreased expression, decreased plaque size, and lateralization. However, it is of
absolute importance to note that arrhythmias are multi-factorial, comprising of not
only gap-junctional coupling but also membrane excitability and architectural block
making up the tissue [187]. Thus, the origin of arrhythmogenic events cannot be
associated with connexin remodeling alone.
Lateralization is a term describing a prominent feature of the gap-junctions that
are found throughout the plasma membrane of cardiac myocytes. Lateral disposition
43


Figure 3.7: Distribution pattern of Cx43 in control ventricular myocardium (A) Lon-
gitudinal section from rat RV (B) Z-projection of confocal images show apparent gap
junction immuno-labeling at the lateral surfaces. These proteins could be considered
components of extended IDs (see Figure 1.4). Most Cx43 gap junctions are observed
at the periphery of the discs.
of Cx43 was not observed in the IF images of control and MCT-treated rat RVs.
The Cx43 immunolabels are mostly found at the IDs albeit in smaller areas and
44


in patches in which they are fewer or absent, otherwise normal. The amount of
immunodetectable Cx43 in the RV of MCT-treated preclinical models show decreased
levels of the gap junction protein, observed qualitatively. The reduction of Cx43 is
shown widely irrespective of whether the heart failure is due to idiopathic dilated
cardiomyopathy, ischemic heart disease, or aortic stenosis [111, 113, 188]. We, like
the studies before us, observed Cx43 expression to be spatially heterogeneous.
3.2.3 Down-regulation of Connexin43 and o-Actinin in MCT model of
PH as observed through immunofluorescence
Studying the immunofluorescent images of control and 4 week post MCT-
treatment, it is clear that there is reduction in Cx43 levels as indicated by the rep-
resentative pattern seen in RV failure model vs. its presence seen in control group,
Figures 3.8 and 3.9. The gap junction protein is confined to the IDs of both the
control and MCT-treated rat groups. However, carefully observing the longitudinal
sections, the MCT-treated rat RVs show disturbances in the organizational pattern
in comparison to the control RVs. Figure 3.8 (B) shows decreased or absent Cx43
proteins at the IDs compared to control (A). Figure 3.9 (B) shows a higher degree of
disorganization (arrows) where Cx43 plaques are situated at locations outside the con-
fined space of the IDs. There are fewer immunodetectable Cx43 in the MCT-treated
rat RVs. Whether the reduction in expression occurs with progressing severity of PH
is yet to be determined. Obtaining images as a function of time at 2 weeks and 3
weeks along with western data supporting the immunofluorescent images would aid
in answering the question. The finding that Cx43 level is decreased at 4 weeks was
verified with western blot and the results are provided in the Section 3.4.
Fluorescently immunodetectable a-actinin was observed to be dramatically down-
regulated in the MCT-treated RV failure model of PH. The discrepancy in expression
between the control and the diseased model is illustrated in Figure 3.8. The morphol-
ogy of the cardiac myocytes are visually altered where the coverage of the contractile
45


a-actinin is lesser in the MCT-treated RVs. The immunofluorescent images of the
control MCT-treate groups show distinct staining of myofibrilar Z-bands, but the
registration of the antibodies is decreased in the later group. Additionally, cc-actinin
staining shows the array of Z-bands in the control RV to be organized in parallel but
the disarrayed myofibrils in the MCT-treated RV lose that organizational pattern and
in some areas show zig-zag Z-line irregularity. These observations are not location
specific. Figure 3.8 shows 40x magnification images of the RV tissues but Figure 3.9
is of representative images at 20x magnification showing the pattern of ct-actinin (and
Cx43) in control and MCT-treated rat RVs.
Figure 3.8: Down-regulation of immunodetectable Cx43 and cu-actinin is observed in
rat right ventricle of MCT induced RV failure PH (B) compared to control (A). 40x
magnification
The qualitative findings for cu-actinin are consistent for longitudinal sections of
the RV labeled with a-actinin. a-actinin labels the sarcomeric Z-discs in the cardiac-
myocyte. Mutations in the cu-actinin gene are associated with cardiomyopathy [189,
190], but its role in pulmonary hypertension is yet to be determined. There isnt any
data present linking Cx43 and Q'-actinin localization patterns in the myocardium.
46


Control
v

\
uactlnin Cx43 D^PJ
50 pm
MCT
/

-actin irrtx43 DAPI
50 jjm
Figure 3.9: Representative image at a lower magnification showing decreased levels
of and Cx43 and cc-actinin in MCT-treated PH (B) compared to control (A). 20x
magnification
Observing the control and the MCT-treated RVs we see that Cx43 is contiguous with
a-actinin at the borders of cardiac myocytes where the IDs are present. Highlighted
in Figure 3.10, we observed that the area occupied by both Cx43 and a-actinin could
be proportional to each other. The change in morphology of cardiac myocytes seen via
distinguishable disarray of cc-actinin staining in MCT-treated rat RVs is accompanied
by decreased a-actinin staining at the IDs and colocalizing Cx43 at the discs.
3.3 Immunofluorescence Staining: Quantitative Results
Figure 3.11 illustrates the steps taken in order to quantify and analyze the IF
images. The dynamic range of the pixels was adjusted to remove background as
well as any unnecessary elements. This step in the code is implemented to limit
the dynamic range of an 8-bit grayscale image so that deductions made through a
thresholding algorithm later in the quantification step is ameliorated with minimum
iterations. An additional step for improving the signal to noise ratio can be added
to the custom MATLAB code at this point in image processing by removing the
47


A > t v i f : . .73 Control "V / - B 4 t MCT
! > \ O * i '1, ' t-A -
. ; . , N. * ' v ",
- Cx43 O'. IP! , ; -j 50 pm a.: i: Cx43 DAPI - 50 Mm
Figure 3.10: Cx43 and a-actinin seem to share area coverage at the IDs. Decrease in
a-actinin in the MCT-treated rats (B), is qualitatively proportionate with the amount
of Cx43 present at the discs.
variance caused by pixels in the tail ends of the intensity range. For instance, Figure
3.12 shows an intensity histogram for a raw image. The red line represents a threshold
value that could be applied to the image such that any intensity value above that line
is now converted to a maximum intensity value of 255 (as a 8-bit grayscale image can
have 256 pixel values; 0 being minimum or black and 255 being maximum intensity or
white). Otsu thresholding, discussed next, is applied and the image is then binarized
to allow for automatic counting by MATLAB or by ImageJ.
Otsu thresholding method, established in 1979, is based upon a simple idea: find
a threshold value from a gray-level histogram that minimizes the weighted within-
class variance [191]. In other words, the algorithm chooses a threshold value that
minimizes the intra-class variance of the thresholded black and white pixels in an
image. The thresholding method assumes that the image and the histogram are
bimodal, meaning that the image consists of two classes of pixels: foreground and
background. It then calculates the optimum threshold value separating the two classes
so that the combined spread is minimal. An advantage of using the Otsy thresholding
algorithm is that the optimal threshold value is computed automatically and stably,
based on the integration of the histogram rather than based on differentiation. In
48


Image Acquisition
Microscopy
;>>
ZEISS LSM 780
Preparation
Binarization
Post processing for
optimization
Particle Analysis
Binarized Image
Channel
-Actinia Cx43 1 Nuclei
r n Original Image (Raw data) r y Original Image (Raw data) r y Original Image (Raw Data) 1024x1024 pixels ZEN 2012
Adjust dynamic range of signal and background
r y Prepared Image (Raw data) v. ^ r y Prepared Image (Raw data) > Prepared Image (Raw data)
Apply Otsu thresholding
Binarized Image
Channel
Binarized Image
Channel
Dilate: Erode
unnecessary pixels:
r N Postprocessed channel v .7 Postprocessed channel r y Postprocessed Nuclei channel
\
r y Analyzed image r y Analyzed image r y Analyzed image V
Quantitative Index Total area (normalized
to nuclei count)
Total area (normalized
to nuclei count)
Number of nuclei
Figure 3.11: Schematic representation of quantitative image processing
other words, the global property is incorporated rather that a local property of the
histogram, for instance, a valley between the peaks of an image histogram.
The acquisition of images followed a stochastic manner where the location of the
image in the RV cannot be determined precisely, in other words, areas of interest
were picked randomly. A minimum of three images per sample, and three samples
per group per immunolabel marker were the basis for staining and obtaining data. It
quickly became apparent that to correctly analyze the ct-actinin stains, the section
required to have been imaged in the longitudinal plane. Longitudinal sections of the
RVs show clear demarcated Z-bands of the sarcomere, and is a true representation of
the ct-actinin structure over stains seen in cross-sectional views (refer back to Figure
2.2. The purpose of immunofluorescence images were to not only localize the protein
49


0 50 100 150 200 250
Figure 3.12: Intensity histogram for the raw image shown earlier (refer to Figure
3.7(A)). The gray-scale map on the x-axis represents the pixel intensity value where
0 is black and 255 is white.
of interest in the myocardium but also study the organizational patterns in control
vs. diseased states. We were able to extract reliable information only form the
longitudinal sections with respect to a-actinin immunofluorescent stains. However,
this difficulty did not factor into analysis of Cx43, dystrophin or other markers.
Representative images for the steps taken to process each immunofluorescent image
following the schematic (Figure 3.11) are shown in Figure 3.13. Table 3.1 consists
of values for length in /mi designated per pixel for respective magnifications used
to acquire the images. The analyzed and quantified indexes return values of area,
length, etc. in pixels. The table provides conversion factor to obtain the characteristic
50


parameters in units of //m.
Table 3.1: Zeiss LSM 780: Pixel size measurement reference for the different objectives
and zoom factors used during image acquisition
Objective Zoom factor Pixel count Distance conversion facor [//m/pixel]
40x 2x 1 0.104
40x lx 1 0.208
40x 0.6x 1 0.346
2 Ox lx 1 0.415
2 Ox 0.6x 1 0.692
Quantitative analysis performed on the immunofluorescence images are in support
of the qualitative deductions. Figure 3.15 shows the scatter plot for the total area
covered by cu-actinin in control vs. the MCT RV failure model of PH. Sample sizes for
control and MCT-treated models were n=5 and n=4 respectively. We are, for the first
time, presenting data showing that cu-actinin level is decreased in the MCT-treated
rat RVs in the setting of PH. The result for the immunofluorescence quantification
is presented in a box-whisker and scatter plot manner as the data points are skewed.
The edge of the boxes represent 1st and 3rd quartile values, the red line represents
median of the data set, the whiskers represent minimum and maximum data points,
and lastly, the data point outside of the whiskers represent suspected outliers. The
outliers are not excluded from the data set since results for cu-actinin havent been
presented in such a manner before. Additionally, robust quantification methods and
repetitions would be required to verify the outlier characteristic of the data points.
The quantified immunodetectable cu-actinin is significantly decreased in the MCT-
treated rat RV compared to the control group. There is a decrease in protein that
51


(b) Modified image
(c) Binarized image
Figure 3.13: Images representing the steps taken for image processing. Cx43 on the
left column and nuclei on the right column were processed sequentially. Scale bar=50
micrometer
52


1
Figure 3.14: Representative image for the analyzed image used to provide the quan-
titative index of Cx43 and nuclei count. A similar approach was taken for obtaining
the pixel count for cc-actinin and dystrophin.
maintains the structural integrity of the myocyte cytoplasm and the plasma mem-
brane. ct-actnin isoform found in the myocardium is calcium-insensitive which is an
advantage as the maintenance of structural integrity should not be calcium dependent
in an organ where the flux is changing constantly allowing for transduction of elec-
trical impulses. Since members of the spectrin family are actin binding proteins, the
decrease in cu-actinin could suggest degradation of the contractile apparatus in the
myocardium. Whether the decreased levels is due to a problem in a-actinin itself or
the protein it binds to is yet to be determined in PH. We are also quantifying protein
levels at week 4 post MCT administration. Histological sections obtained at 2 and
3 week time points would be needed to observed and analyzed in order to determine
the trend of Q'-actinin depletion in the MCT preclinical model of PH.
53


Figure 3.15: Box-whisker and scatter plot for cc-actinin representing the total pixel
area normalized to cell count. n=5 control,; n=4 MCT; p<0.01.
Similar steps were taken to quantify the Cx43 protein levels in controls and MCT-
treated rat groups. Surely enough, immunodetectable Cx43 was found to be decreased
in the MCT-treated rats compared to the control group (p<0.01). The box-whisker
and scatter plot for connexin plaques expression is presented in Figures 3.16 and 3.17.
We had two subsets of data for Cx43. This was due to utilization of two different
clones of the antibody allowing us to co-stain Cx43 with other markers. Cx43* seen
throughout the results is representative of data subset that uses a different antibody
compared to data shown using Cx43. But regardless of the clone of the antibody
used, the result is consistent: decreased levels of Cx43 in the MCT-treated rat RVs
compared to controls. The values are normalized but the discrepancy in the range
between the two figures may have risen from the use of two different antibodies used
to probe for Cx43. (Please refer back to Table 2.2 for information on the different
subsets of antibodies used).
54


'53
o
u
c
o
O
TO
TO
o
c
rt
c
X
TO
c
c
o
O
TO
TO
TO
O
H
1 2
Group 1 = Control; Group 2 = MCT
Figure 3.16: Box-whisker scatter plot for Cx43 representing the total area normalized
to cell count, p<0.01.
TO
o
D
C
o
Q
TO
O
C
c
X
TO
C
c
o
o
TO
TO
1 2
Group 1 = Control; Group 2 = MCT
Figure 3.17: Box-whisker and scatter plot for Cx43* correlates positively with data
performed with different clone of the antibody p<0.01.
55


As aforementioned, gap junction remodeling presents with features such as down-
regulation of connexins and decreased plaque sizes in multiple heart diseases. Through
immunofluorescence we found that immunodetectable Cx43 is down-regulated in PH
rats. We proceeded to further characterize the gap junction remodeling in the setting
of PH by quantifying the plaque size in the controls and 4 week MCT-treated rats.
Figure 3.18 shows an immunofluorescent image acquired at 80x magnification where
the Cx43 plaques are discernible and clearly demarcated.
aactitiin Cx43 20
Figure 3.18: Clear demarcation of Cx43 plaques is observed in the immunofluorescent
image of RV at 80x magnification. Modified images for quantification extraction is
provided in next figure.
With additional modification in the custom MATLAB code to adjust for and
exclude indistinguishable plaques (due to high intensity profile at certain areas), we
56


extracted plaque sizes for Cx43 in control and diseased RVs. The additional modifica-
tion of the code also excluded small pixels that werent contiguous in the thresholded
image. Figure 3.19 shows the representative images of image processing for plaque
size quantification. We quantified 406 and 507 individual Cx43 plaques across multi-
ple images to get the length and area, respectively, for the control group. Similarly
295 and 386 individual Cx43 plaques were quantified to output length and area, re-
spectively, for the MCT-treated rat RVs.
a
Figure 3.19: (A) Thresholded image for Cx43 where blue pixels and red pixels are
excluded to provide us with the image in (B). Blue and red pixels in (A) represent
small areas and clusters that are indistinguishable as individual plaques, respectively.
Scale bar: 20 fxm
We found that Cx plaque size decreased significantly (p<0.05) from controls to
4 week MCT-treated rat RVs. The length of Cx43 plaques measured orthogonal to
myocytes long axis (reported as meanSD) was computed to be 0.7440.0189 fini
for control and 0.6760.0162 lira. The areas of these plaques were computed to be
1.4700.0579 fira and 1.2970.0558 fim for control and MCT groups respectively.
57


The graphical representation of the findings is presented in Figure 3.20.
1.8
on
2
I
I
to
CO
CD
a
N
* M
CO
CD
-2
"E,
Id
U
1.6
1.4
1.2
0.8
0.6
0.4
0.2
0
Plaque Length Area
Figure 3.20: Bar plot summarizing plaque size measurements for Cx43 in control vs.
MCT-treated rat RVs. p<0.05
Intercellular communication has been shown to be disrupted when gap junction
plaque sizes are reduced [112, 115, 116]. The conduction velocity is reported to be
slower and remodeling with characterization of decreased plaque sizes have been linked
to arrhythmias. Additionally Tan et al. have shown that decreased Cx43 plaque size
and disturbed organization may be associated with RV hypertrophy. Whether the
loss of Cx43 plaques and decrease in plaque size occur right after MCT administration
is yet to be determined. This void in the current research could be filled when data for
the 1 and 2 weeks is acquired for immunofluorescence as well as western blot protein
assay. Thus, at this time we arent able to deduce whether the loss and decrease in
58


size of Cx43 plaques is associated with progression of MCT-induced PH.
The quantification of plaque sizes was further characterized to see the distribution
of small and large Cx43 plaques. Histogram of the plaque sizes for control and MCT-
treated groups are provided in Figure 3.21. In both control and 4 week MCT-treated
rat groups, there is a higher frequency of smaller gap junction Cx43 plaques.
Distribution of Cx43 Plaque Size, Control
60 -----------------.--------.--------.--------.-------
0 0.5 1 1.5 2 2.5 3 3.5
Cx43 plaque length
Distribution of Cx43 Plaque Size, MCT
02 0.4 0.6 0.8 1 12 1.4 1.6
Cx43 plaque length
Figure 3.21: Plaque size distribution distinguishing small and large Cx43 in control
and MCT groups.
59


To verify that the MATLAB code is returning accurate values for plaque size
in terms of length and area, fluorescent beads of known dimensions were quanti-
fied using the custom code. The fluorescent beads were imaged using the same
parameters used to acquire immunofluorescent images of our proteins of interest.
FocalCheckFluorescent Microspheres from Molecular Probes of 6/im diameter were
used. Figure 3.22 (A) shows the raw image of the fluorescent beads taken at 80x mag-
nification and 1024x1024 resolution. Figure 3.22 (B) shows the analyzed binary image
that was then labeled and the length was computed. The code returned a mean value
of 56.894 pixels for the diameter of the bead. The known distance of 1 pixel at these
specific acquisition parameters is known: 0.104 /xm. This gives us the length of the
bead as 5.915 /mi. With only 1.417% difference in between the computed and actual
size, we feel strongly about the numbers reported for Cx43 plaque sizes.
Figure 3.22: Image of fluorescent beads (A) Raw image acquired using the same pa-
rameters used to image the Cx43 images that output plaque sizes. (B) Post-processed
binarized image that was labeled and used to extract the length of the beads.
Additionally, images were taken at magnifications of 20x and 40x. The custom
MATLAB code returned average values of 5.959 /mi and 5.920 /mi respectively. With
60


the minimum errors in size measurements for the fluorescent beads we are confident of
the custom code generated for this study. The representative images for the respective
magnifications are provided in Appendix B.
The findings that there is reduction in immunodetectable protein levels in MCT-
treated rat RVs raises a question of whether the reduction is due to cell death in the
diseased preclinical models. To answer this, we proceeded to quantify the prevalence
of cardiac myocytes in the RV of both the control and MCT groups. The a-actinin
stain, although specific for cardiac myocytes, does not provide adequate informa-
tion through immunofluorescene to discern the amount of cardiac myocytes in the
myocardium. Therefore, we turned to dystrophin, a peripheral membrane protein
specific to cardiac myocytes.
Figure 3.23: Dystrophin co-stained with Cx43 in control and 4 week MCT rat RVs.
Dystrophin labels the peripheral region of the cardiac myocyte and is absent at the
IDs containing terminus of myocytes. Highlighted regions show individual cardiac
myocytes.
Figure 3.23 shows dystrophin and Cx43 co-staining presented in the longitudinal
section. The cardiac myocytes can now be traced in these immunofluorescent images
61


as dystrophin lines up the lateral membrane of the myocytes and Cx43 indicate the
terminus of these cells. The image analyzing MATLAB code used for quantifying
a-actinin and Cx43 was applied to 20 control samples and 12 MCT samples (3 im-
ages per sample). There was no significant difference between the immunodetectable
expression level of dystrophin in the plasma membranes of rat cardiac myocyte in
control vs MCT-treated models. With this finding, we suggest that the prevalence
of cardiac myocyte is not different between the control and 4 week MCT-treated
rat RVs. However, this does not mean that there is no activity representing change
in number of cardiac myocyte during the progression of PH. Data generated from
analyzing immunofluorescent images representing total area normalized to nuclei for
a-actinin, Cx43, and dystrophin is summarized in Table 3.2 followed by the graphical
representation shown in Figure 3.24.
120 r
a-actinm
Cx43
Control MCT
*
Cx43* Dystrophin
Figure 3.24: Bar plot showing that quantitative cr-actinin and Cx43 correlates with
qualitative deductions; p<0.05. No significant change in dystrophin between the
control and MCT-treated groups.
62


Table 3.2: Summary of western blot findings for control and diseased groups. Values
presented are relative as they are normalized to cell count.
Marker Control MCT
Mean 83.212 69.814
St. Dev 7.352 14.785
cu-actinin SEM 2.325 4.268
p-value - 2.178e-02
Mean 88.868 74.383
St. Dev 18.273 12.125
Cx43 SEM 7.460 4.950
p-value - 1.019e-03
Mean 95.890 58.275
St. Dev 7.260 11.415
Cx43* SEM 4.191 6.590
p-value - 6.809e-03
St. Dev, standard deviation; SEM, standard error of the mean
To further support our suggestion that the loss of cr-actinin and Cx43 in the RV
of MCT-treated rats is not due to cell death, we assessed prevalence of CASP3 active
cells in the course of progression of PH and quantified the findings. Figure 3.25 shows
in control, and in 1, 2, and 3 weeks of the MCT-treated rat RVs, the cells participating
in apoptosis. CASP3 active cells are in a cascade pathway and are certain to fulfil
apoptosis. We observe that there is abundance of cell death in 1 week after MCT
administration and the apoptotic activity starts to taper off at 2 weeks. By 3 weeks,
apoptosis is virtually absent. With c-staining of smooth muscle actin (sma) present,
we are able to deduce that vascular endothelial cells are among the cell types that are
CASP3 active. But most of CAP3 active cells early on in the progression of PH are
63


present in the myocardium (arrows, Figure 3.25). These could be cardiac myocyte,
cardiac fibroblasts, or other non-myocyte cell types in the RV. Without a cell-type
specific marker co-stained with CASP3, we are not able to ascertain the cell type
undergoing apoptosis at this time.
Figure 3.25: Cell death is a prominent feature early on during the progression of PH
as observed through CASP3 immunofluorescence staining. Cell-types mostly going
through apoptosis are in the myocardium but also in the blood vasculature as shown
by sma co-localization with CASP3.
Our custom MATLAB code was applied to these images to quantify apoptosis
as PH progresses after the administration of MCT. The result is presented in Figure
64


3.26 below.
Control 1 wk MCT 2 wk MCI 3 wk MCT
Figure 3.26: Bar plot showing active caspase3 in the RV myocardium of control and
1, 2, and 3 weeks post MCT treatment.
It is interesting to observe that the cell death is much more robust in the early
stages of PH but is largely absent in the end-stages of PH in rats. Additionally, we
looked at analyzing prevalence of nuclei in controls and 4 week MCT to see whether
there is a major difference in cell count between the groups. We found no significant
changes in the number of nuclei present which further confirms out data and sug-
gestion that the depletion in proteins of interest is not occurring due to cell death.
The loss of Q'-actinin and Cx43 is important in the setting of PH and the fact that
it is occurring independent of cell death supports our hypothesis that organizational
and expressional changes of these proteins are promoting intercellular disturbances,
changes in contractile function leading to edema and RV failure. The nuclei count
analysis is summarized in Figure 3.27. The experimental sets comprise of all of the
acquired images separated properly into respective control and MCT-treated groups
65


but in a randomized manner. In other words, the protein of interest didnt play a
role in this analysis since we were only looking at DAPI channel.
Group 1 = Control; Group 2 = MCT
(a) Imaging set 1
1 2
Group 1 = Control; Group 2 = MCT
(b) Imaging set 2
(c) Imaging set 3
(d) Summary of nuclei count per staining set
Figure 3.27: Nuclei counted for multiple staining sets show no significant decrease in
total number of cells from control to MCT-treated models.
When examining the dystrophin images in the cross-sectional view, as shown in
Figure 3.28, it was apparent that there were discrepancy in cardiac myocyte cell size
between the control and the MCT-treated rat RVs. The length of cardiac myocytes
66


were measured using longitudinal sections (Figure 3.23. The width was computed
using coss-sectional sections as seen in the figure below.
> YV \S
A Vh/,
f < Sr S. -t S '
j.^V ? / Y r
If i A .
Control i
't / v
, t * f , -ik. t x *
\ // V 'V i gA
--.v .
tf(}P>£WMxx
} */
/7v:: wT-'
o. '-i/! Z7/'v' X':~
r ^ '3 iA .-'V ^ 3T;-
j 1 A . -y- A Y V /-l...
/,' f,,*. 'Wi.V v V
, /, A 4-
i t ^,
\ A- I CA/ A W4. / V" > / J^I

i/Vv:/ Z^s, ( v-h
>,V; nV-/_
\X" :^pfjy ;'vit..h/>A/AT ,/v-.
*3* K -Atr r i ;i
(SSS~ :rn '-r^ :/.
'{\

/ r ' '
. H -J ''-1
Mr V

f(50A- px^3pysjro^l'npOT '; . 50r
Figure 3.28: Dystrophin co-stained with Cx43 in control and 4 week MCT rat RVs.
Dystrophin labels the peripheral region of the cardiac myocyte and is absent at the
IDs containing terminus of myocytes. Highlighted regions show individual cardiac
myocytes.
The length was measured from the IDs at each of the terminal edges of the
myocyte cells labeled by Cx43. We processed the image and computed width by
measuring the distance along the axis parallel to the IDs (shown by Cx43 stains in
red) and along the axis perpendicular to the IDs. We found no significant difference
in doing the measurement two ways hence, the width measurements represent the
diameter of the cardiac myocyte cells. For n = 43 individual cardiac myocyte for
controls and n = 33 for MCT we found no significant changes in the length of the cells.
For n = 46 control and n 37 MCT-treated cardiac myocytes we found a significance
increase in cell width from 17.1113.862 /in1 to 28.1957.616 /m1. The length stays
relatively unchanged but the increase in width suggests increase in cardiac myocyte
cell volume. With the presence of edema in the myocardium the cells maybe taking
67


up fluid either by passive or active transport. Additional experiments would be
needed to verify whether the water channels are disturbed and whether the cells
taking up volume is compensatory or not depending on the type of transport (passive
vs. active). Since there is not change in dystrophin expression, as shown earlier, the
plasma membranes integrity may be unchanged. The bar graph showing length and
width analysis is provided in Figure 3.29.
70 r
n = 43 n = 33
Control MCT
n = 46 n = 37
Length
Width
Figure 3.29: Summary of length and width analysis performed on control and MCT-
treated cardiac myoctes of the RV. p<0.01
3.4 Western Blot
As one of the assays to verily the findings presented in the previous subsection,
we looked to western blot protein assay to validate the loss of protein expression in
the MCT-treated rat RVs in PH. Qualitative analysis on the IF images during early
stages of the study led to deductions that immunodetectable Cx43 and cr-actinin
were down-regulated in the MCT RV failure model of PH. The quantitative analysis
68


performed on the images further supported those deductions. Nine samples of RV
homogenates were loaded into a 10-well gel, and comprised of triplicates for control, 4
week MCT, and 4 week Su5416-hypoxia groups. The 10th well, loaded with a ladder,
is used for size reference (kDa).
Before the immunoblots were quantified, qualitatively assessing the a-actinin IF
images, it was very apparent that the proteins expression is down-regulated in the
MCT-treated group. The western blot results presented in Figure 3.30 positively
correlate with the qualitative findings from the immunofluorescence staining images.
The extra bands seen at the 98 and 62 kDa levels were background. A negative control
assay for the western blot was performed using just the secondary antibody: goat-anti-
mouse conjugated with HRP (EMD Millipore). It was observed that the secondary
antibody would bind to proteins around 98 and 62 kDa regions as seen in Figure 3.30
with the same patter. Since the protein of interest, cu-actinin, lies in the 107 kDa,
the background bands produced by the secondary antibody do not compromise the
integrity of the assay, and the bands, therefore, are ignored. Quantification of the
western blot show that the decrease in cu-actinin expression is highly significant in
both the diseased models of RV failure PH when compared to control groups. Figure
3.31 demonstrates the quantified results (normalized by GAPDH). Two-sample t-test
on the sets of control and MCT groups and sets of control and Su5416-hypoxia were
performed with a of 5%.
There is variance within the triplicates of the same groups. Thus, there is pop-
ulation heterogeneity present for the amount of proteins being expressed among the
population within the respective diseased groups. This is true for humans as well, not
everyone expresses the same protein at exactly the same levels. The data presented
is a strong representation of the proteins being expressed among the triplicates as the
loading control in Figure 3.30 is even throughout each of the 10-well plates.
69


Nx MCT SuHx Nx MCT SuHx
a
Figure 3.30: Western blot for a-Actinin (left) performed for rat RV homogenates.
GAPDH (right) used for normalization purposes. Significant down-regulation of a-
actinin in protein levels of MCT and Su5416-hypoxia groups are observed.
Nx, Normoxic; MCT, Monocrotaline; SuHx, Sugen5416-hypoxia; GAPDH, Glyceraldehyde 3-
phosphate dehydrogenase
3 9
Is
8 V
s ^
CQ
S
U 2
e* B
0.25
**
* *

Control MCT
SuHx
Figure 3.31: Quantification of Western blot for cc-actinin shows 63% decrease for
MCT and 83% decrease for Su5416-hypoxia models.
70


Figure 3.32 shows decreased expression of Cx43 protein levels in the MCT-treated
model of PH but there is no significant difference seen in the Su5416-hypoxia group
in comparison to the control. The protein levels for MCT group is decreased by 64%
in contrast to control. Multiple bands seen around 43 kDa could be the different
isoform of Cx-43 being picked up by the antibody used (Figure 1.7). The band across
98 kDa might suggest that Cx-43 are present as dimers. Phosphatase inhibitors were
administered during the tissue homogenization process but the status of proteins
before the inhibitor administration is not controlled. There is a possibility of Cx43
undergoing phosphorylation prior to or even during the time organs were harvested.
Cx43 being able to form a heteromeric bond with other connexins, mainly Cx45 in
the heart, could also suggest the band at 98 kDa. Quantifying the western blot reveal
that Cx43 is significantly decreased in the MCT model. The variance for the Su5416-
hypoxia model is greater than for the MCT group. Due to population heterogeneity,
the high variance seen in SuHx could be due to different rats expressing protein at
different levels.
1.5
125
Figure 3.32: Downregulation of Cx43 in protein levels is observed for the MCT RV
failure model but not in Su5416-hypoxia PH rats. Quantification of Western blot for
Cx43 shows 64% decrease for MCT induced RV failure PH model. p<0.01
71


Negative correlation between the expression of Cx43 among the two diseased
RV failure models of PH could be attributed to the different mechanism of distinct
preclinical models. Additionally, Cx43 is shown to have a response to hypoxia where
there is increased expression in rat carotid body [195, 194], We focused mostly on
MCT model for this project. However, obtaining data for SuHx would prove useful in
understanding Cx43 expression across multiple diseased models of PH. Additionally,
we could verify the hypoxia response viaa chronic hypoxia model for PH.
Table 3.3: Summary of quantified western immunoblot protein assay.
Marker Control MCT SuHx
Mean 1 0.376 0.165
St. Dev 0.118 0.174 0.151
cu-actinin SEM 0.068 0.101 0.087
p-value - 0.00719 0.00145
Mean 1 0.361 1.068
St. Dev 0.202 0.127 0.541
Cx43 SEM 0.116 0.073 0.312
p-value - 0.00204 0.867
St. Dev, standard deviation; SEM, standard error of the mean. The p-values presented are with
respect to control groups. There is no significant difference between the MCT- and SuHx-treated
groups for either a-actinin or Cx43.
Recognizing that Cx43 expression has a negative correlation among different mod-
els that lead to the same end point, RV failure, is interesting. We went back and ac-
quired SuHx immnofluorescent images to validate our western findings. Figure 3.33
shows Cx43 immunofluorescence stain for SuHx-treated rat RV. Three sets of samples
with three images per sample were acquired to produce the data shown in the box-
whisker and scatter plot presented in Figure 3.34. There was no significant change in
72


immunodetectable Cx43 expression between controls and SuHx-treated rat RVs.
a. 1
f "s W % V**
V ^
* v i ** f , XV- ** V r" ^ '-'1 t
1 1 4 wy':- ^ /'i
1 V / : i < -v; -V A V i ~ ^ 1 1 < ^ 1 ^ 't* **** & '
At \ * , A v >
x * * l* V
Cx43.. t **' 50 pm v'1 v.--' v % ;
Figure 3.33: Cx43 in SuHx RV failure model. Immunofluorescence image acquired to
verify western immunoblot assay. Arrows show mislocalization of Cx43.
1 2 3
Group 1 = Control; Group 2 = MCT; Group 3 = SuHx
Figure 3.34: Box-whisker and scatter plot showing Cx43 protein levels in control vs.
MCT- and SuHx-treated PH rats.
73


3.5 Assessment of Fibrosis
We believe that edema in the myocardium preceeds fibrosis. As mentioned in
Section 3.1, our lab group has been successful in positively correlating the presence
of EVF with elevated collagen deposition in the RV. With immunofluorescence we
wanted to assess the change in number of fibroblasts that either reside in the my-
ocardium or migrate there after introduction of MCT to further strengthen the corre-
lation and better our understanding about the role of fibroblasts, their cross-talk with
cardiac myocytes and the surrounding ECM. But it is well established that a marker
specific for fibroblast has been difficult to nail down. This remains one of the main
reasons why our understanding of fibroblast in severely lacking. We pursued TCF21,
FSP and P4H to try and distinguish the cardiac myocytes from the fibroblasts. We
had successful staining for all three markers but the pitfall that limits our data is
that these markers do not co-localize for the same cell type. This is due to presence
of different populations of fibroblasts in the myocardium [88, 131, 138] and in part
due to FSP also labeling inflammatory subpopulation of macrophages [136].
Figure 3.35 shows effective immuno labeling of cardiac myocytes plasma mem-
brane with dystrophin and fibroblasts with P4H. This co-staining combination could
be used in determining information about fibroblasts with respect to the following
aspects: 1) their prevalence in control vs. their in diseased state correlating a change
in expression due to injury and progression of disease, and 2) their location with re-
spect to diseased models; we know that their distribution is non-uniform throughout
the ventricle [96]. Utilizing this technique will allow us to visually place fibroblasts
in the regions of the RV. As a result, we could potentially determine the locations of
the RV that is severely injured compared to other locations. For instance, assessing
preliminary pentachrome stains of the RV, as shown in Figure ?? we were able to see
elevated deposition of collagen (yellow stains) in the RV free wall compared to the
myocardium. Combining the pentachrome stain, assessing the collagen deposition, to
74


- ^
\ J-. V \ .
.J
'
\.
/ ^
' " v j.
*v 4 < ; ^ *> N
P4K .Dystrophin

n* ; " V: \
JL \ \ V

V
7* t r .. -
- *" V
4 -A 'h' \
DAPi Merge
Figure 3.35: Co-immunostaining of cardiac myocyte marker dystrophin (green) and
cardiac fibroblast marker (P4H) presented in a montage.
the immunofluorescence images, assessing the location of fibroblasts, could lead us to
narrow down the regions of the RV most affected by PH.
3.6 Limitations and Recommendations
The ability to assess fibrosis using immunofluorescence is hindered by missing
specific markers for fibroblasts. Discoidin domain-containing receptor 2 (DDR2),
also known as CD 167b is a protein that is associated with fibrosis and cancer [95, 94],
DDR2 is also known to regulate fibroblast proliferation and migration through the
ECM. It is recommended that the use of DDR2 be incorporated into the mix of the
P4H and TCF21 to target fibroblasts through the hypertensive heart. Immunofluo-
rescence labeling of Cx43 has provided valuable information in the setting of multiple
heart diseases. Studies have made headway about interaction between connexins and
other adhesion junction molecules. IF imaging is a resourceful tool that allows us
75


Figure 3.36: Collagen deposition in the MCT-treated rats is indicative of fibrosis.
Additionally, the blue-green stain seen in the diseased image represents immature
collagen.
to visualize and to some extent analyze the colocaliztion of these junctional pro-
teins at the IDs. To fully understand the interaction of these discrete structures, it
is recommended that we go beyond the use of immunofluorescence. This technique
doesnt have the resolution to tell us whether the apparent overlap of fluorescence
signals are true representation of interaction and association at the molecular level.
We could, however take confocal microscopy to reconstruct the myocardial tissue in
3-dimensional. Figure 3.37 shows an attempt at a reconstruction of IDs in the RV.
Figure 3.37 has advantages over 2D images. With a 3-D reconstruction we would
be able to quantify Cx43 plaque size with increased accuracy. Additionally, we would
be able to characterize the distribution and organization of gap junction Cx43 at the
IDs with an added view of the transverse plane. We could also investigate how the
organization changes within the IDs. There are however limitations that prevent us
from acquiring such images. The first of them is the heart tissue. It is complex and
is highly muscularized which makes it difficult to effectively stain thicker sections.
Optimization of the a staining protocol would be required that do not jeopardize
the integrity of the tissue structure so that we are quantifying our data correctly.
76


Figure 3.37: 3-Dimensional reconstruction of a rat RV focused on Cx43 at the IDs.
80x magnification; reconstructed using ImageJ
Failure to optimize parameters to acquire images is a disadvantage. This is one of the
limitation we ran into during our study. For images used to quantify Cx43 plaques,
such as Figure 3.20, there were some regions that had to be excluded from the data
pool because the intensity profile was too hight to get a discernable Cx43 plaque(s).
One way to overcome this problem is to either turn down the power or the digital gain
so that the intensity profile is adequately measure throughout the imaging plane.
77


4. Discussion
The underlying cellular mechanisms governing the pathogenesis of pulmonary
hypertension still remain to be fully explored and elucidated. A contributing factor
could be due to the pathogenesis of PH differing depending on the type of etiology.
For this reason, our lab group focused on studies in different preclinical models of PH:
hypoxia, monocrotaline and sugen5416-hypoxia. We conducted our research mostly in
the MCT model of RV failure PH due to the time-frame and the scope of this project
being a two-year Masters thesis. Additionally, there is no mechanistic data present
for gap junction remodeling in the RV in the context of RV failure PH. Therefore,
a large part of the project reflected on learning about gap junction remodeling in
various other diseased settings and organs, and try to incorporate those findings to
our preclinical models whether they overlapped or not.
Emerging studies suggest that connexins may regulate inflammatory signals, such
as infiltration of leukocytes and adhesion molecule expression [192], We dont have
evidence of this for the project since we didnt look into leukocyte distribution in
the RV myocardium. Another example is a study that linked Cx43 depletion to re-
duced numbers of inflammatory cells and atherosclerotic lesions [193]. At this point
in the study, we can only speculate that the inflammation and fibrosis seen in the
RV correlates to the dramatic depletion of Cx43 and their disorganization. This is
largely in due to out data reflecting time points at control and 4 week post MCT-
administration. Additional data points with inflammatory markers and at shorter
time intervals as PH progresses would allow us to extract more information with re-
spect to fibrosis, inflammation and of course Cx43 and o-actinin. As outlined in the
Future Directions, Section 4.2, the functional consequences of altered Cx43 expression
need to be investigated with robust experiments that answer specific questions such
as: What other components/networks interact with Cx43 in failing heart?, what are
the nature and mechanisms of gap junction remodeling?, how does the expression in
78


RV differ from LV? lungs?, and most importantly, how does this connect us to human
heart failure? These are but few questions that remain to be addressed. In addition
to including more time-points in our data acquisition, we could answer the aforemen-
tioned questions by investigating the IDs at a molecular level via electron microscopy,
check for and understand the interactions between gap junctions and other junctional
complexes within the IDs nexus, and further characterize gap junction remodeling.
The perturbed localization and down-regulation of connexins and other junctional
complexes like claudins (a tight junction) give rise to a variety of developmental de-
fects. Arrhythmias are one of the well-established functional consequence of Cx43
disturbance, but these defects are just not limited to the electrical coupling in heart.
They extend to deafness, skin diseases and demyelination [110]. What could be the
cause of the reduction in numbers of Cx43? Interaction with other junctional com-
plexes and mutations in them could be a cause of downregulation in Cx43 expression
in the RV. The turnover of Cx43 is rapid (1.5 to 2 hours) /citeBeardsleel998, thus,
mutations in Cx43 due to endoplasmic reticulum stress could lead to a decrease in
expression as well as plaque size reduction. The organization of the Cx43 is known
to be at the intercalated discs. However, some claims of lateral immunostaining have
been reported. This can be explicated through the organization of the intercalated
discs and interaction of gap junctions with other junctional complexes located at the
IDs. For as long as IDs have been known the information acquired through research
has not been linear with respect to time. Studies of IDs in diseased states could
likely add to our knowledge about its formation and cross-talk among its various
components.
Some could argue that the depletion of Cx43 and a-actinin seen in the diseased
models is due to cell death. Thus, suggesting that the accompanying decrease in
numbers of targeted proteins in RV failure model of PH positively correlates to fewer
cells present post 4 week MCT administration. But we have shown that decreased
79


Cx43 expression could rise independent of cell death. It is interesting to note that
the RV of the PH rats at 3 and 4 weeks recover from robust cell death and apoptosis
in the 1st couple of weeks post MCT-administration (please refer back to Figures
3.25, 3.26 and 3.27). During the early time-points when there is an abundance of
cell death as shown by heavy active CASP3 staining, we speculate that Cx43 and
a-actinin start to drop out immediately. Image acquisitions and western assay at
early weeks post MCT-administration would be used to validate the speculations.
But further investigation, as to why the Cx43 and a-actinin loss dont recover at 4
week post MCT-administration, would be required.
We observed no significant change in Cx43 in the Sugen5416-hypoxic model in our
study. As aforementioned, this result could be attributed to the different models of
PH. There are studies that show up-regulation of Cx43 in hypoxic models [194, 195].
Loosely follows the data seen in the Western blot performed for Su5416-hypoxic. The
variance is greater and the mean is slightly higher, albeit insignificantly, than control.
4.1 Strengths and Weaknesses: The Tales from our Data and the Effects
on out Hypothesis
The strength in this study is the positive correlation seen between multiple assays
and different quantihcation/analysis done on those assays revealing the same results.
The qualitative and the quantitative analysis on the IF images and the quantitative
results from Western blot output similar findings. We hypothesized that disturbances
in organizational and expressional changes in gap junction contributes to increased
myocardial edema and decreased RV function. The specific aim was directed towards
characterization of Cx43 in the RV of irreversible PH models. We have shown that
there is a loss of immunodetectable Cx43 at the IDs located at the terminus of cardiac
myocytes. These are situated perpendicular to the long axis of the myocytes. Com-
pared to an arrayed organization for the controls, the MCT-treated rat RVs showed
loss of Cx43 plaques. We further characterized plaque sizes differences in controls
80


vs. diseased model and found that there is a significant decrease in the individual
plaque sizes of Cx43. Further investigation is required, however, since Cx43 is able
to form homomeric bonds with other Cx43 junctional complex. We seem it necessary
to verify whether the larger plaque sizes represent a single Cx43 plaque or whether
they ar aggregates of multiple Cx43 plaques. Additional data that would support
the hypothesis would be to acquire electric coupling data. Arrhythmias are capable
of causing heart failure and since disturbances in Cx43 has been linked to decreased
conduction velocity and arrhythmogenic events in multiple cardiomyopathies, it is im-
portant to assess electric and conductive apparatus of the myocardium in the setting
of PH.
One of the major drawbacks of studying a disorder associated with failing hearts
and lungs is the inability or the difficulty to obtain human tissue. Thus, depending
on the type of funding and part of the world the research is being conducted, we are
limited to preclinical animal models. As mentioned throughout the thesis, cell types
are species dependent and thus correlating the findings and applying the results to
the humans takes time and the outcomes are not always favorable. Performing assays,
similar to those conducted in this study, on human tissue would be transformative
and pivotal if findings positively correlate between species.
A pitfall encountered during staining of thicker sections was the precipitation
fixation method used prior to IF staining. It turns out that this method is not very
useful when using thick sections for confocal microscopy. As mentioned in Chapter 2, a
1:1 ratio of methanol:acetone was used. Although, this fixation type may have helped
in permeabilizing the tissue for effective penetration and labeling of the antibodies, it
did not help preserve the 3-dimensional organization for our specimen. Information
gathered could lead to misconceptions.
81


4.2 Future Directions
The first and foremost follow up will include verifying whether the findings from
our preclinical models regarding the expression of Cx43 and o-actinin correlate with
human patients with PH. Our expert collaborators in France who study PH in human
tissue will aid in making the connection between our preclinical models and human
tissue.
The findings of this study very much set up rather heavy questions towards un-
derstanding the pathogenesis of PH. In terms of connexins and gap junctional pro-
teins role in vascular permeability and/or lymphedema, a study could be concocted
whereby a permeability response is assessed in an in vitro model to inflammatory
stimuli followed by inhibition of Cx43 function. Will this study then show that there
is down-regulation of Cx43 to an inflammatory response. This will also answer the
question of whether Cx43 down-regulation is causal to the progression of PH or is
it a consequence. The follow up will be to further investigate the changes in Cx43
and o-actinin as the disease progresses. Then moving onto SuHx to verify whether
the changes are uniform throughout multiple models of the same disease or not. This
will aid in targeted therapeutic measures that would ameliorate pathophysiology of
PH. For instance, we found that in both MCT- and SuHx-treated rats, o-actinin
expression is decreased. A future study must be incorporated to support and further
investigate the expressional changes in o-actinin. Very little in known about the pro-
tein expressed abundantly in the sarcomeric z-discs and research catered to figuring
out the functional consequences could provide novel avenues of assessment, prognosis
and treatment of pulmonary hypertension.
82


4.3 Concluding Remarks
It is firmly established that gap junctional remodeling and connexin expression is
a consistent characteristic observed and recorded in diseased ventricles. This is evi-
denced through comparative studies conducted on various preclinical models. There-
fore, gap junctions and connexin are increasing considered as therapeutic targets to
potentially assess and treat pulmonary hypertension. With more studies focusing on
the electrical coupling functionality of gap junctions, we could benefit from studies
on these protein complexes from a mechanical point of view. On the other hand, a
new protein of interest might be of importance in understanding the progression of
PH to right heart failure and could prove to be a valuable marker in helping us un-
derstand the unknown mechanisms of the disease, o-actinin is dramatically depleted
in RV failure models of PH. Seeing as how the contractile myocytes of the RV are
dealing with depletion of a major component required for proper function, it could
lead to transformative novel therapies. The prognosis associated with pericardial ef-
fusion and edema in the RV is very poor and despite this negative impact, very little
is known about te etiology, consequences and management of edema in PH. With
our fin dings, we anticipate future studies to incorporate investigations that reveal
practical potential therapeutic targets in effective assessment and treatment of PH.
83


REFERENCES
[1] Gerald Simonneau, Michael A. Gatzoulis, Ian Adatia, David Celermajer, Chris
Denton, Ardeschir Ghofrani, Miguel Angel Gomez Sanchez, R. Krishna Kumar,
Michael Landzberg, Roberto F. Machado, Horst Olschewski, Ivan M. Robbins,
and Rogiero Souza. Updated clinical classification of pulmonary hypertension.
In Journal of the American College of Cardiology, volume 62, 2013.
[2] Rubin M. Tuder, Carlyne D. Cool, Michael Yeager, Laimute Taraseviciene-
Stewart, Todd M. Bull, and Norbert F. Voelkel. The pathobiology of pulmonary
hypertension: Endothelium. Clinics in Chest Medicine, 22(3):405418, 2001.
[3] R Kessler, M Faller, E Weitzenblum, a Chaouat, a Aykut, a Ducolone,
M Ehrhart, and M Oswald-Mammosser. Natural history of pulmonary hy-
pertension in a series of 131 patients with chronic obstructive lung disease.
American journal of respiratory and critical care medicine, 164(2):219-24, 2001.
[4] Charlie Strange and Kristin B Highland. Pulmonary hypertension in interstitial
lung disease. Current opinion in pulmonary medicine, 11(5)452-5, 2005.
[5] U Mehlhorn, H J Geissler, G A Laine, and S J Allen. Myocardial fluid balance.
Eur J Cardiothorac Surg, 20(6)4220-1230, 2001.
[6] Stefan Zahler, Anke Hoffmann, Torsten Gloe, and Ulrich Pohl. Gap-junctional
coupling between neutrophils and endothelial cells: a novel modulator of
transendothelial migration. Journal of leukocyte biology, 73(1)418-126, 2003.
[7] Vallerie V McLaughlin, Stephen L Archer, David B Badesch, Robyn J Barst,
Harrison W Farber, Jonathan R Lindner, Michael A Mathier, Michael D Mc-
Goon, Myung H Park, Robert S Rosenson, Lewis J Rubin, Victor F Tapson,
and John Varga. ACCF/AHA 2009 expert consensus document on pulmonary
hypertension a report of the American College of Cardiology Foundation Task
Force on Expert Consensus Documents and the American Heart Association de-
veloped in collaboration with the American College of. Journal of the American
College of Cardiology, 53(17)4573-619, apr 2009.
[8] Ruba N Haddad and Lisa M Mielniczuk. An evidence-based approach to screen-
ing and diagnosis of pulmonary hypertension. The Canadian journal of cardi-
ology, 31(4):38290, apr 2015.
[9] Larissa A Shimoda and Steven S Laurie. Vascular remodeling in pulmonary
hypertension. Journal of Molecular Medicine, 91(3):297 309, 2013.
84


Full Text

PAGE 1

RIGHTVENTRICLEFAILUREINPULMONARYHYPERTENSIONIS ASSOCIATEDWITHMISLOCALIZATIONOFGAPJUNCTIONCONNEXIN-43 ANDCHANGESINEXPRESSIONOFCONTRACTILE -ACTININ by OZUSLOHANI B.S.,ColoradoSchoolofMines,2009 Athesissubmittedtothe FacultyoftheGraduateSchoolofthe UniversityofColoradoinpartialfulllment oftherequirementsforthedegreeof MastersofScience Bioengineering 2015

PAGE 2

ThisthesisfortheMastersofSciencedegreeby OzusLohani hasbeenapprovedforthe DepartmentofBioengineering by MichaelE.Yeager,Advisor RichardBenninger,Chair CarlyneD.Cool November20,2015 ii

PAGE 3

Lohani,OzusM.S.,Bioengineering RightVentricleFailureinPulmonaryHypertensionisAssociatedwithMislocalization ofGapJunctionConnexin-43andChangesinExpressionofContractile -Actinin ThesisdirectedbyAssistantProfessorMichaelE.Yeager ABSTRACT PulmonaryhypertensionPHisadisorderofmultipleetiologiesandonewithoutadenitecure,andhighmortality.Littleisknownabouttherightventricular proteomicsinhumanheartfailure,includingpossibledierencesincomparisonto theleftventricle.Connexinsareknowntomodulateinammatorysignalinginother systemsandthus,wehypothesizedthatconnexinsmayberesponsibleforincreased lymphedemaintherightventricle.Weshowthatintherightventricularfailure modelofmonocrotalineinducedPH,connexin43and -actininexpressionaresignificantlydecreasedand -actininisfurtherdecreasedinsugen4516-hypoxicmodelof rightventricularfailurePHbutconnexin43showsnosignicantchange.Weused Westernblottoconrmtheproteinexpressionandimmunouorescencestainingfor quantitativeanalysisoftheseproteins.Withwesternblotwefoundthatconnexin43 isdecreasedby64%and -actininisdecreasedby63%4weeksaftermonocrotaline administrationandfurthermore, -actininisdownby84%inthesugen5416-hypoxic model.Lateralizationofgapjunctionalproteinssuchasconnexin43isobservedin ventriclesofmultipleheartdiseaseslikecardiomyopathyandmyocarditis.Alongwith thepresenceofgapjunctionalconnexinsinthelateralsurfaces,theirexpressionis usuallyfoundtobedown-regulated.Wecanaddmonocrotaline-treatedpreclinical modelstothegrowinglistofcardiacdiseaseswheretheconnexin43isdown-regulated. However,lateralizationintherightventriclewasnotobserved.Patchesofconnexin43 areseeninwhichotherwisenormallyarrayedplaquesarefewerorabsent.Thedecreasein -actinininthepreclinicalmodelsisanovelnding.Alongwithpresence iii

PAGE 4

ofventricularbrosis,increasededemaintherightventricle,theseresultssuggest thatpulmonaryhypertensioninducedbypressureoverloadisassociatedwithdisorganizationofgapjunctiondistributionandthefailingheartcouldbeassociatedwith changesin -actininexpression. Theformandcontentofthisabstractareapproved.Irecommenditspublication. Approved:MichaelE.Yeager iv

PAGE 5

DEDICATION Thisleisdedicatedtomyparents:OmandSaritaLohani,whosesupportand dedicationhasmadethisthesispossible. v

PAGE 6

ACKNOWLEDGMENT WithoutthecontinuedsupportandmentorshipofKelleyColvinandMichaelYeager, thecompletionoftheprojectwouldn'thavebeenpossible. Additionally,Iwouldliketogivethankstothefollowingindividualsfortheirpositive contribution: ... LindsayQuandt GregGlazner RaduMoldovan JaneParr AleenaNotary MasonMcClatchey MatthewWestacott vi

PAGE 7

TABLEOFCONTENTS Tables........................................x Figures.......................................xi Chapters 1.Introduction...................................1 1.1ResearchQuestion............................1 1.2PulmonaryHypertension........................2 1.3ScienticandClinicalJusticationoftheProject...........7 1.4Hypothesis................................8 1.4.1RightVentricle..........................8 1.5GapJunctions..............................10 1.5.1Connexin-43...........................11 1.6RoleofCx43inSupportoftheWorkingHypothesis:VascularLeak andIncreasedInterstitialFluid.....................14 1.7ConnexinsintheMyocardium.....................16 1.7.1Fibroblast-FibroblastCommunication.............17 1.7.2Myocyte-Fibroblastinteraction.................18 1.8NormalExpressions:LocationandAppearance............18 1.9Aberrations...............................20 1.10CellSpecicandLymphaticMarkers.................21 1.10.1CardiacMyocyteMarkers: -ActininandDystrophin.....21 1.10.2CardiacFibroblast:NarrowingDownaSpecicMarkerforFibrosisandCardiacRemodeling.................22 1.10.3ApoptoticandProliferationMarkers:Caspase-3,Ki-67....24 1.10.4MarkersforLymphangiogenesisandAngiogenesis:Lyve-1, Podoplanin,Prox1........................25 2.MethodsandMaterials.............................28 vii

PAGE 8

2.1AnimalModels..............................28 2.2ObtainingTissue.............................29 2.3WesternBlotProteinAssay.......................30 2.4ImmunouorescenceStaining......................32 2.4.1IFandAntibodyStainingTechniques.............33 2.5ConfocalMicroscopy...........................35 2.5.1Rationale:OutoftheStaticStudiesofThinSections.....35 2.5.2Why3Dover2D:WhorlingEect/Stereology.........37 3.Results......................................38 3.1AssessmentofPHinPreclinicalModels................38 3.2ImmunouorescenceStaining:QualitativeResults..........43 3.2.1Connexin-43OrganizationandExpressionintheNormalHeart43 3.2.2GapJunctionRemodelinginRVFailurePH..........43 3.2.3Down-regulationofConnexin43and -ActinininMCTmodel ofPHasobservedthroughimmunouorescence........45 3.3ImmunouorescenceStaining:QuantitativeResults..........47 3.4WesternBlot...............................68 3.5AssessmentofFibrosis..........................74 3.6LimitationsandRecommendations...................75 4.Discussion....................................78 4.1StrengthsandWeaknesses:TheTalesfromourDataandtheEects onoutHypothesis............................80 4.2FutureDirections............................82 4.3ConcludingRemarks...........................83 References ......................................84 Appendix A.CustomMATLABCodes............................103 viii

PAGE 9

B.SupplementaryFigures.............................109 ix

PAGE 10

TABLES Table 1.1Updatedclinicalclassicationofpulmonaryhypertension,2013*.....3 1.2Markerprolespeculationsandinterpretationsinpreclinicalmodels...9 2.1AntibodiesusedforWesternBlotAssay..................32 2.2Primaryantibodiesandtheiroptimumdilutions..............34 3.1ZeissLSM780:Pixelsizemeasurementreferenceforthedierentobjectivesandzoomfactorsusedduringimageacquisition...........51 3.2Summaryofwesternblotndingsforcontrolanddiseasedgroups.Values presentedarerelativeastheyarenormalizedtocellcount.........63 3.3Summaryofquantiedwesternimmunoblotproteinassay.........72 x

PAGE 11

FIGURES Figure 1.1Thelayersofthevesselwall..........................4 1.2Compositionofvascularadventitia:anoverlookednexus..........5 1.3Illustrativerepresentationofprogressionofvascularremodelinginthesettingofpulmonaryhypertension.Increasedpulmonaryvascularresistance isobservedduetounmonitoredproliferationandgrowthofsmoothmusclecellsandendothelialcells,asaresult,leadingtoincreasedpulmonary arterialpressure................................6 1.4Thin-sectionelectronmicroscopyshowingtheorganizationofgapjunction withotheradherentjunctionalcomplexesintheratRV.AStep-like featuresofdiscsarrowheadsobservedattheendoftheisolatedcell. BFasciaeadherentjunctionsoccupyingtheelectronrichregion.Gap junctionappeartobeconnectingwithfasciaeadherentjunctionsarrows.12 1.5IllustrationofConnexin-43arrangementintheplasmamembrane.....13 1.6FunctionalconsequencesofabnormalCx43organizationandexpression seeninremodeledheart.Itisunclearwhethertheremodelingoftheheart preceedsthedisorganizationofCx43orifitiscausal.Dottedlinesrepresentpotentialcausalinterplay,solidlinesrepresentestablishedassociations.15 1.7Organizationofconnexinsubunitstoformthehemichannelsubunitconnexon.Connexonshavemultipleconnexintypesformingeitheraheteromericoraheterotypicchannel.......................19 1.8ColocalizationofPodoplaninandProx1inthetertiarylymphoidtissue BALTsshowingpositivecorrelationbetweenthetwolymphaticmarkers.26 xi

PAGE 12

2.1Hearttissueobtainedfromthepreclinicalmodelsincludetherightand leftventricles.Therightandleftatriaareremovedpriortocuttingthe tissuedesiredforourassays.Thewholeheartontheotherhandisutilized topreparehomogenatesforWesternblotexperiments.TheWesternblot assayperformedforthisstudyusedRVhomogenates............30 2.2AImmunouorescencestainof -actininwithlongitudinalorientation withthez-discsofthesarcomeredistinguishablewhereasBhelpsexplain thedicultyencounteredwhendealingwithhearttissue.Thecomplex stereologyofhearttissuecausestheobservationviewtochangefromlongitudinalyellowarrowtotransversalsectionsgreenarrow.......36 3.1Macroscopicrepresentationofmorphologyoftheheartfrompreclinical ratmodels.ThisgureshowscrosssectionofAcontrolandBtissue 4weekpostMCTadministration......................38 3.2Macroscopicrepresentationofpentachromestainonsingle5 mthinsectionsoftheheart.Thisgureshowscrosssectionofcontrol,Aand3 weekpostMCTadministration,B....................39 3.3MeasurementsofmyocardialedemaforRVsandLVsinvariouspreclinical modelsofPH.Valuesarereportedasmean St.Dev..........40 3.4T2-weightedMRIshowingedematousRVforMCT-treatedRatright comparedtothecontrolleft.ImagesacquiredusingBruker4.7TeslaMR.41 3.5Masson'strichromestainassessingbroticchangeincontroltopleftvs. hypoxictoprightandMCTbottomrow.Hypoxicratsdemonstratea similarpatternofmodestcollagencontentintheRVascontrol.Elevated collagendepositionintheRVofMCT-treatedratscorrelatespositively withthepresenceofEVF...........................42 3.6CorrelationbetweenEVFandFibrosissupportsourspeculationsabout edemapreceedingbrosisinPH.......................42 xii

PAGE 13

3.7DistributionpatternofCx43incontrolventricularmyocardiumALongitudinalsectionfromratRVBZ-projectionofconfocalimagesshow apparentgapjunctionimmuno-labelingatthelateralsurfaces.TheseproteinscouldbeconsideredcomponentsofextendedIDsseeFigure1.4. MostCx43gapjunctionsareobservedattheperipheryofthediscs....44 3.8Down-regulationofimmunodetectableCx43and -actininisobservedin ratrightventricleofMCTinducedRVfailurePHBcomparedtocontrol A.40xmagnication............................46 3.9Representativeimageatalowermagnicationshowingdecreasedlevels ofandCx43and -actinininMCT-treatedPHBcomparedtocontrol A.20xmagnication............................47 3.10Cx43and -actininseemtoshareareacoverageattheIDs.Decreasein -actininintheMCT-treatedratsB,isqualitativelyproportionatewith theamountofCx43presentatthediscs...................48 3.11Schematicrepresentationofquantitativeimageprocessing........49 3.12IntensityhistogramfortherawimageshownearlierrefertoFigure 3.7A.Thegray-scalemaponthex-axisrepresentsthepixelintensity valuewhere0isblackand255iswhite...................50 3.13Imagesrepresentingthestepstakenforimageprocessing.Cx43onthe leftcolumnandnucleiontherightcolumnwereprocessedsequentially. Scalebar=50micrometer..........................52 3.14RepresentativeimagefortheanalyzedimageusedtoprovidethequantitativeindexofCx43andnucleicount.Asimilarapproachwastakenfor obtainingthepixelcountfor -actininanddystrophin...........53 3.15Box-whiskerandscatterplotfor -actininrepresentingthetotalpixelarea normalizedtocellcount.n=5control,;n=4MCT;p < 0.01.........54 xiii

PAGE 14

3.16Box-whiskerscatterplotforCx43representingthetotalareanormalized tocellcount,p < 0.01..............................55 3.17Box-whiskerandscatterplotforCx43*correlatespositivelywithdata performedwithdierentcloneoftheantibodyp < 0.01...........55 3.18CleardemarcationofCx43plaquesisobservedintheimmunouorescent imageofRVat80xmagnication.Modiedimagesforquantication extractionisprovidedinnextgure.....................56 3.19AThresholdedimageforCx43wherebluepixelsandredpixelsare excludedtoprovideuswiththeimageinB.BlueandredpixelsinA representsmallareasandclustersthatareindistinguishableasindividual plaques,respectively..............................57 3.20BarplotsummarizingplaquesizemeasurementsforCx43incontrolvs. MCT-treatedratRVs.p < 0.05........................58 3.21PlaquesizedistributiondistinguishingsmallandlargeCx43incontrol andMCTgroups................................59 3.22ImageofuorescentbeadsARawimageacquiredusingthesameparametersusedtoimagetheCx43imagesthatoutputplaquesizes.B Post-processedbinarizedimagethatwaslabeledandusedtoextractthe lengthofthebeads..............................60 3.23Dystrophinco-stainedwithCx43incontroland4weekMCTratRVs. Dystrophinlabelstheperipheralregionofthecardiacmyocyteandis absentattheIDscontainingterminusofmyocytes.Highlightedregions showindividualcardiacmyocytes.......................61 3.24Barplotshowingthatquantitative -actininandCx43correlateswith qualitativedeductions;p < 0.05.NosignicantchangeindystrophinbetweenthecontrolandMCT-treatedgroups.................62 xiv

PAGE 15

3.25CelldeathisaprominentfeatureearlyonduringtheprogressionofPHas observedthroughCASP3immunouorescencestaining.Cell-typesmostly goingthroughapoptosisareinthemyocardiumbutalsointheblood vasculatureasshownbysmaco-localizationwithCASP3.........64 3.26Barplotshowingactivecaspase3intheRVmyocardiumofcontroland1, 2,and3weekspostMCTtreatment.....................65 3.27Nucleicountedformultiplestainingsetsshownosignicantdecreasein totalnumberofcellsfromcontroltoMCT-treatedmodels.........66 3.28Dystrophinco-stainedwithCx43incontroland4weekMCTratRVs. Dystrophinlabelstheperipheralregionofthecardiacmyocyteandis absentattheIDscontainingterminusofmyocytes.Highlightedregions showindividualcardiacmyocytes.......................67 3.29SummaryoflengthandwidthanalysisperformedoncontrolandMCTtreatedcardiacmyoctesoftheRV.p < 0.01.................68 3.30Westernblotfor -ActininleftperformedforratRVhomogenates. GAPDHrightusedfornormalizationpurposes.Signicantdownregulationof -actinininproteinlevelsofMCTandSu5416-hypoxia groupsareobserved..............................70 3.31QuanticationofWesternblotfor -actininshows63%decreaseforMCT and83%decreaseforSu5416-hypoxiamodels................70 3.32DownregulationofCx43inproteinlevelsisobservedfortheMCTRVfailuremodelbutnotinSu5416-hypoxiaPHrats.QuanticationofWestern blotforCx43shows64%decreaseforMCTinducedRVfailurePHmodel. p < 0.01.....................................71 3.33Cx43inSuHxRVfailuremodel.Immunouorescenceimageacquiredto verifywesternimmunoblotassay.ArrowsshowmislocalizationofCx43..73 xv

PAGE 16

3.34Box-whiskerandscatterplotshowingCx43proteinlevelsincontrolvs. MCT-andSuHx-treatedPHrats.......................73 3.35Co-immunostainingofcardiacmyocytemarkerdystrophingreenand cardiacbroblastmarkerP4Hpresentedinamontage..........75 3.36CollagendepositionintheMCT-treatedratsisindicativeofbrosis.Additionally,theblue-greenstainseeninthediseasedimagerepresentsimmaturecollagen................................76 3.373-DimensionalreconstructionofaratRVfocusedonCx43attheIDs.80x magnication;reconstructedusingImageJ.................77 B.1CustomMATLABcodeappliedforbeadsat20xmagnicationwithaccurateresultsforbeadsizeoutput.....................109 B.2CustomMATLABcodeappliedforbeadsat40xmagnicationwithaccurateresultsforbeadsizeoutput.Thus,thecodecanbeappliedforimages ofdierentmagnication...........................110 xvi

PAGE 17

1.Introduction 1.1ResearchQuestion PulmonaryhypertensionPHisadiseasecharacterizedbyprogressivecardiopulmonaryremodelingandreducedrightventricularRVfunction.Theprogressionof thediseaseeventuallyleadstorightheartfailure.ThismechanismisnotfullyunderstoodinpartbecausePHisassociatedwitharatherlargespectrumofconditions anddiseasesthatrangefromautoimmunedisordersandimmune-insuciencysyndromestosystemicandmetabolicdisorders[1].Thefunctionalstatusofpulmonary circulation,thelevelsofpulmonaryvascularresistancePVR,andpulmonaryarterialpressurearekeyfactorsthathelpdeterminetheoutcomeandtherapeutictreatmentsofpatientswithPH.HighlightingotheraspectsofpathobiologyofPHbroadly classiesthedisorderintomild/moderatevs.severecategoriesbasedonpulmonary arterypressures,ventricularperformanceandoverallmortality[2].Intimalthickening andmedialhypertrophythatareobservedinchronicobstructivepulmonarydisease COPD[3]andothernon-neoplasticlungdiseases[4]presentwithmildtomoderatePH.Endothelialcellproliferativelesionsplexiformlesionsforinstance,intimal brosis,smoothmusclecellSMCgrowthintheintimalandmedialregionaresuspectedtocauseseverecasesofPH.Additionally,alterationsinvascularreactivityare observedalongsidethearterialvascularremodeling. Heartfunctionisstronglyreliantonwaterbalancewhereevenaminorincrease inmyocardialwatercontentimpairsthecontractileforce[5].Pericardialeusionand presenceofedemaintherightventricleareimportantpathologicalndingsinthe settingofPH.Cell-celljunctionalproteinsplaypivotalrolesinintercellularcommunicationsbetweenthevariouscelltypesintheheartandthelungs.Studies,discussed indetailinsection1.6,suggesttheroleofgapjunctionsinmaintainingvascularbarrierfunction.Itisalsospeculatedthatgapjunctioninhibitorspromotevascular leak[6].Theaccumulationofuidintheheartandthelungsincreasesmortalityin 1

PAGE 18

patientswithPH.UnderstandingtheoriginofedemaanddeducingwhythehomeostaticuidowisdisturbedinPHcouldbetransformativendings.Fine-tuningour understandingoftheroleofgapjunctionproteinsinpreclinicalmodelsofPHand howtheycorrelatetotheconditioninhumansshouldleadtothedevelopmentofnew therapies.Thisstudyaimstocharacterizethedysregulationofcell-celljunctionsand correlatethendingstopresenceofedemacontributingtoRVfailure. 1.2PulmonaryHypertension Adisorderofmultipleetiologies,PHisaconditionwherethemeanpulmonary arterialbloodpressuremPAPis 25mmHgasdeterminedbyrightheartcatheterizationRHC[7].PulmonaryarterialhypertensionPAHreservessomeofthewelldenedformsofPHandisfurtherdenedasanmPAPof 25mmHginthepresence ofpulmonaryarterialwedgepressurePAWPof < 15mmHg,orinotherwords,normalleft-sidedllingpressures[8].ThecurrentclassicationofPHisbasedonthe Evian/Veniceclassicationproposedduringthe1998secondWorldSymposiumof PulmonaryHypertensionandmodiedduringthethirdWordSymposiumheldin Venice,Italyin2003[1].Basedontheclassication,PHiscategorizedintove groupswhereeachgroupcohesivelydenesPHaccordingtosimilarclinical,pathobiological,etiologicalandtherapeuticcharacteristics.Thisgeneralcategorizationhas beenmaintainedthroughthefourthandfthWorldSymposiumheldatDanaPoint, CaliforniaandNice,Francerespectively.Somemodicationshavebeen madeandthelatestupdatedclinicalclassicationisprovidedinTable1.1. TheanimalmodelsgeneratedforthepurposeofthisstudyfallunderGroup1 ofPH.AsidefromincreasedmPAPbeingthehallmarkofPH,biopsiesoftheheart andlungtissuerevealwidelyacceptedcharacterizations:sustainedvasoconstriction andvascularremodeling[9],andcardiacmyocytehypertrophyandRVbrosis[10]. VesselwallsarecomprisedofthreeheterogeneouslayersasseeninFigure1.1:the tunicaintima,tunicamediaandtunicaexterna. 2

PAGE 19

Table1.1:Updatedclinicalclassicationofpulmonaryhypertension,2013* 1.PAH 1.1IdiopathicPAH 1.2HeritablePAH 1.2.1BMPR2 1.2.2ALK-1,ENG, SMAD9,CAV1,KCNK3 1.2.3Unknown 1.3Drug-andtoxin-induced 1.4Associatedwith: 1.4.1Connectivetissuedisease 1.4.2HIVinfection 1.4.3Portalhypertension 1.4.4Congenitalheartdisease 1.4.5Schistosomiasis 1`Pulmonaryveno-occlusivediseaseand/orpulmonarycapillaryhemangiomatosis 1Persistentpulmonaryhypertensionofthenewborn 2.Pulmonaryhypertensionduetoleftheartdisease 2.1Leftventricularsystolicdysfunction 2.2Leftventriculardiastolicdysfunction 2.3Valvulardisease 2.4Congenital/acquiredleftheartinow/outowtractobstructionandcongenitalcardiomyopathies 3.Pulmonaryhypertensionduetolungdiseaseand/orhypoxia 3.1Chronicobstructivepulmonarydisease 3.2Interstitiallungdisease 3.3Otherpulmonarydiseaseswithmixedrestrictiveandobstructivepattern 3.4Sleep-disorderedbreathing 3.5Alveolarhypoventilationdisorders 3.6Chronicexposuretohighaltitude 3.7Developmentallungdiseases 4.Chronicthromboembolicpulmonaryhypertension 5.Pulmonaryhypertensionwithunclearmultifactorialmechanisms 5.1Hematologicdisorders: chronichemolyticanemia ,myeloproliferativedisorders,splenectomy 5.2Systemicdisorders:sarcoidosis,pulmonaryhistiocytosis,lymphangioleiomyomatosis 5.3Metabolicdisorders:glycogenstoragedisease,Gaucherdisease,thyroiddisorders 5.4Others:tumouralobstruction,brosingmediastinitis,chronicrenalfailure, segmentalpulmonaryhypertension a a MainmodicationstothepreviousDanaPointclassicationareinbold.ALK-1,activinreceptor-likekinase1gene; BMPR2,bonemorphogenicproteinreceptortypeII;CAV1,caveolin-1;ENG,endoglin;KCNK,potassiumchannelsubfamilyKgene; PAH,pulmonaryarterialhypertension;SMAD9,SMADfamilymember9. *FifthWorldSymposiumonPulmonaryHypertension,Nice2013. Haddad etal. 3

PAGE 20

a Figure1.1:Thelayersofthevesselwall. a c PearsonEducation,Incorporated Theoutermostlayer,tunicaexterna,isoftenreferredtoastunicaadventitia oradventitia.Workingourwayfromtheinnertotheoutermostwalllayer,we have:monolayerendotheliallining,connectivetissueandelasticbercomposing thetunicaintima;smoothmuscletissuesheathcomprisingthethickestmiddlelayer morecommoninthearteries,separatedfromtheadventitiabyathinbandofelastic membrane;andtheoutermostlayerthatformsaconnectivetissuesheatharoundthe vessels.Theadventitiaismorecomplexthanoriginallyspeculated.Asillustrated inFigure1.2itisthoughttoconsistofanextracellularmatrixECMframework composedofnerves,lymphaticandbloodvessels,progenitorcells,immunecells,and broblasts[11].Manypreclinicalanimalmodelshaveconsistentlyrevealedextreme adventitialremodelinginthesettingofhypertension[12,13]. Whiletheexactcausesoftheprogressionremainunclearandunderinvestigation, theremodelingseeninPHischaracterizedtovariousstandardsbythickeningof intimalandmediallayersofmuscularvessels[9,14],thickenedadventitia[15],and 4

PAGE 21

a Figure1.2:Compositionofvascularadventitia:anoverlookednexus. a Stenmark etal. appearance/survivalofvascularsmoothmusclecells[16],leadingtodevelopmentof vaso-occlusivelesionsandvaryingdegreesofinammation[17].Figure1.3illustrates theputativepathogenesisofPH:progressionofmuscularizationofsmall,peripheral pulmonaryarteries,andthehistologicalimageatthecentershowsheavystainingfrom smoothmusclemarker -smoothmuscleactin -SMA.Proliferationandmigration ofpulmonaryarterialsmoothmusclecellsPASMCs[14]andmyobroblasts[18], endothelialdysfunctionleadingtocompromisingvasculartone[19],andvaso-occlusive lesionsinvolvingPASMCsandendothelialcellsECs[20]occurinseverecasesofPH. However,studieshaveshownthatbroblastsresidingintheadventitiamaybe themostsuitablecelltypeabletosense"hypertensivestates[21],andremodelingin thesystemicvasculatureischaracterizedbyincreasedbroblastproliferationwhich 5

PAGE 22

a Figure1.3:Illustrativerepresentationofprogressionofvascularremodelinginthesettingofpulmonaryhypertension.Increasedpulmonaryvascularresistanceisobserved duetounmonitoredproliferationandgrowthofsmoothmusclecellsandendothelial cells,asaresult,leadingtoincreasedpulmonaryarterialpressure. a ModiedfromDavies etal. withpermission precedeandexceedECsandPASMCsproliferation[22,23].Fibroblasts,dubbed thesentinelcell"bySmith etal. ,havebeensuggestedtoactivateanddierentiate intomyobroblasts[24,25].Myobroblastsparticipateinwouldhealingbutthe sustainedsurvivalofthiscelltypecontributestoremodeling[18]andachronicstate ofinammationviaforwardfeedbackmechanism[26].Therefore,theadventitial layer,theperi-adventitiasurroundingthemajorvessels,andthecontentswithinthis dynamicvascularwallarekeyplayersinremodelingofthebronchovasculararea. Inadditiontothepulmonaryvascularresistanceresultingfromchangesinsmall arteriolesandrarefactionofthevessels,increasedstinessdecreasedcompliancein theproximalpulmonaryarteriescontributetoincreasedafterloadintherightventricle [27,28,29]. 6

PAGE 23

1.3ScienticandClinicalJusticationoftheProject Knowledgeabouttheroleofrightventricleinhealthanddiseasehasoftenlagged behindandoverlookedwhencomparedtotheleftventricleLV.Thisisperhapsdue tospeculationsthattheRVis:lessmuscular,restrictedtopumpingbloodtothelungs, andduetoitslessfrequentandnonobviousinvolvementinpathology.Historically, RVhasbeenabystanderinpathophysiologicalstudiesinvolvingthecardiovascular system.RVisadverselyaected,mostnotablyinPHcausedbyvarietyofpulmonary vasculardisorders.TheRVexperiencespressureoverloadregardlessofanyofthe causeslistedinTable1.1andtheresponsesaremyocardialhypertrophy[30],dilation ofthechamber,risingllingpressuresasaresultofcontractiledysfunction[31],and diastolicdysfunction[32].RVfunctionisakeydeterminantinsurvivalofpatients withPH[33,34,35]. OneofthendingsinthesettingofPHispericardialeusionandpresenceof edemaintheRVandthelungs.PericardialeusionandRVinterstitialuidareimportantandhaveseriousclinicalcomplications.Patientswithpericardialeusionhad 1-and3-and5-yearsurvivalrateof80%and20%and0%respectively[36].ThehomeostaticmechanismtoregulatetheuidintheRVisgreatlyinuencedbytheuid movementintothesystemthroughmicrovascularcontrol,themovementoutofthe systembythemyocardiallymphaticnetwork,andtheuideuxintothepericardial spaceanditseventualdrainageintothelargerlymphaticsystem.Thecompositionof theuidpresentintheRViseitherintracellular,extracellular,lymphaticorblood. Myocardialhypertrophyisoneoftheearlycompensatorymechanismsadaptedbythe RVinresponsetopressureoverloadinPH[30].ThemicrovascularltrationisincreasedinPHanduidowinthelymphsystemisdecreasedresultinginmyocardial edema[37].CentralvenouspressureCVP,denedasthepressureofbloodinthe thoracicvenacavaenteringtherightatrium,iselevatedinPHregardlessofetiology [38].CVPisoftenusedtoreecttheamountofbloodreturningtotheheart.The 7

PAGE 24

edema,nowprevalentinthemyocardium,depressescardiacoutputCO[39,5]as aresultofincreasedllingpressures,complicatescontractilityfunctions,andcauses righthearttofailduetosystolicanddiastolicdysfunction.FocusingontheRVcould furtherourunderstandingoftheclinicalissuesandincreaseourknowledgeofthe RVtobeoncomparativeparwithLV.Thisisofpivotalimportanceaswehavea chancetoaddtothegrowingknowledgeoftheprogressionofdiseaseandincrease survivalratesamongpatients.Withthisstudy,wedevoteattentiontotheRVsowe maydevelopwaystodetectandmeasureRVdysfunction,andhowthedeterioration progressesinPH. 1.4Hypothesis TheworkinghypothesisfortheprojectisbasedonobservationsofincreasedinterstitialuidinfailingRVinthesettingofPHwiththepresenceofincreasedvascular resistanceindiseasedstatethanincontrols. WehypothesizethatthemislocalizationofgapjunctionsintherightventriclecontributestoincreasedRV lymphedmawhichpositivelycorrelateswithincreasedinammationand brosis,anddecreasedRVfunction .ThehypothesiswastestedwithPHmodelsthatdevelopRVfailure4to5weeksafteradministrationofeitherMonocrotaline MCTorSugen5416;pleaseseeChapter2. 1.4.1RightVentricle Myocardialuidbalanceismaintainedinlargepartbythelymphaticsinheart alongwithcoronaryvenouscirculation[5,39,40].Thelymphaticsrunalongside bloodvessels[41]andarecomprisedofmyocardial,subendocardialandsubepicardial lymphaticplexuses[40],andlymphaticcollectingduct,orthelymphatictrunkresponsibleforcreatingapathwayforuidtoleavetheheart[42].Severaldrivingforces areinvolvedincontrollingthelymphowintheheart,allofwhichfallintoeither theactiveorthepassivelymphaticpumpingcategory.Alongstandinghypothesis 8

PAGE 25

[43]hadthepericardiumplayinganimportantroleindraininglymphuidfromthe heart.Butduetolackofevidencesupportingthehypothesisandstudiesshowingthe inabilitytoinduceseverelymphedemapostpericardialremoval[42],bothinanimals andhumans,itissuggestedthatthemajordrivingforceregulatingthelymphowin theheartcouldcomefromsubepicardialmusclecontractions.Nevertheless,cardiac lymphaticsarecriticalformaintenanceofuidbalanceintheheartandaberrations fromhomeostaticenvironmentcausesseriousclinicalcomplicationsthatincludepericardialeusion[44],inammationandbrosis[45,46],anddropincardiacoutput [47]whichprovesfatalinmanycases. Table1.2:Markerprolespeculationsandinterpretationsinpreclinicalmodels MarkerControlRVFailurePHInterpretations -ActininControl-Numberofcardiomyocytes Connexin-43 ControlMis-localizationMechanicalfailuremarker Control-Changeintotalabundance Caspase-3+xApoptoticmarker Ki-67+xProliferationmarker LYVE-1Control--Decreasedlymphatics PodoplaninControl--PositivelycorrelateswithLYVE-1lymphvessels Prolyl-4-hydrpxylaseControl++RVFibrosis Prox-1Control--LECs TCF21Control++RVbrosis a a LYVE,lymphaticvesselendothelialhyaluronanreceptor;LEC,Lymphaticendothelialcells; Prox,Prosperohomeobox;TCF,Transcriptionfactor;+,++,-,and--signifyvaryingdegrees ofprevalenceorremodelingincontrasttocontrols.xsigniesnosignicantchangeKi-67,a proliferationmarkerandCaspase-3,anapoptoticmakershowno/minimumactivityin4weeksRV failuremodelofPH.Interpretationsstatedalongsideclarifytheirroleandreectioninthediseased state. 9

PAGE 26

Oneofthe specicaims ofthisstudyisto characterizethedysregulationof cell-celljunctionsintherightventricleofirreversiblePHmodelsandlinkthendings toedemaandRVfailure .ThespotlightisontheRVbutthemagnifyingglassison Connexin-43and -ActininalongwithlymphaticmarkerssuchasLyve-1,Prox1,and Podoplanin.Additionalbackgroundoncell-celljunctionsandthemarkerslistedare providedindetailinthesubsequentsections.Determiningwhethertheseproteins areup-ordown-regulatedandtheircontributiontowardsRVfailurewillaidinunderstandingthestateofuidstasisintheheart.Additionally,thendingswillbe transformativeastherapeuticmeasurestakentocorrectthemislocalizationofcellcelljunctionscouldimproveandreversePHbyrestoringuidhomeostasisinthe heart.Table1.2summarizesspeculativeinterpretationsofthemarkersinthesetting ofControlvs.RVfailuremodelofPH. WhilemostofthefocusonunderstandingremodelinginPHhasbeenoverECs, andPASMCs;incorporatingthefocusongapjunctionsandunderstandingtheirrole invascularleakandinterstitialuidbuild-upinthecardiopulmonaryareacould provideinsightintonoveltherapeuticmeasuresoftacklingPH. 1.5GapJunctions Intercellularcommunicationiscrucialtonotonlymaintainhomeostasisand properfunctionwithinamulticellularorganism,butalsoessentialforitssurvival [48,49].Forexample,cardiacfunctionisdependentuponappropriatecell-cellcommunicationeitherdirectlythroughgapjunctions,orindirectlyviaparacrinesignaling,i.e.solublechemokines,extracellularvesicles,andautocrinesignaling.While gapjunctionsallowforelectricaltetherandmetaboliccouplingbetweenneighboringcellsintheheart,desmosomesandadherensjunctionsareanchoringjunctions thatprovidemechanicalcontinuity[50].Thedirectcommunicationbetweencellsis establishedthroughpassageofsmallsolutes,metabolitesandions[51,52,53,54]. Increasingevidenceindicatesthatgapjunction-mediatedintercellularcommunication 10

PAGE 27

facilitatesnotonlydiusionofionsinneighboringcellsbutalsoembryonicdevelopment,controlledcellgrowthanddierentiation,transendothelialcellmigration,and pathogenesisofatherosclerosis[52,53,55].Intheheart,gapjunctionalproteinsnotoriouslycontributetoimpulsepropagationinbothuniformandnon-uniformtissue [56].Inthelungs,gapjunctionscancontributetocalciumsignalingbetweenciliatedepithelialcellstocoordinateciliarybeating[57],regulatesecretionofpulmonary surfactantbytypeIIepithelialcells[58],andregulatepropagationofcalciumwaves alongpulmonaryvessels[59,60]. TheorganizationofthejunctionsattheintercalateddiscsIDshavebeenmeticulouslyinvestigated.GapjunctionsattheIDsareorganizedtogetherwithdesmosomesandfasciaeadherentsjunctions[61,62].ThespecialtyofIDsistointegrate communicationfromcelltoneighboringcellandhelpmaintaintheelectro-mechanical function.AsseeninFigure1.4,fasciaejunctionsarelocatedinverticalstepswith groovespresentbetweenthemanddesmosomesaresituatedinhorizontalportionsof theIDs.Cx43arelocatedmostlyinthesehorizontalregionsofthediscs. 1.5.1Connexin-43 Gapjunctionchannelsarecomposedofproteinsintheconnexingenefamily. Therearenowovertwodozenconnexinsidentiedinthemammaliangenepool[55]; theyareabbreviatedCx"followedbyanumericalidentication,eg.Cx40,Cx43, Cx45,etc.Thenumbersigniesthemassoftheproteininkilodaltons.Twentyone membersoftheconnexingenefamilyarefoundinhumanand19ofthemcanbe groupedasorthologuepairswithconnexinmembersfoundin murines [53].Connexinsareintegralmembraneproteinsconsistingoffourtransmembranedomains, twoextracellularloops,onecytoplasmicloop,andintracellularlylocatedC-andNterminaltails[52,53,55].AsillustratedinFigure1.5,theC-terminaldomainvaries inlength,containsalargerangeofphosphorylationsites,andasaresult,isthought toplayimportantrolesinprovidingsitesforprotein-proteininteraction.Cx43is 11

PAGE 28

a Figure1.4:Thin-sectionelectronmicroscopyshowingtheorganizationofgapjunction withotheradherentjunctionalcomplexesintheratRV.AStep-likefeaturesofdiscs arrowheadsobservedattheendoftheisolatedcell.BFasciaeadherentjunctions occupyingtheelectronrichregion.Gapjunctionappeartobeconnectingwithfasciae adherentjunctionsarrows. a Severs,NJ etal. 2008 amongthethreemaingapjunctionalproteinsfoundintheheart;theothertwobeing Cx40andCx45[63].TheC-terminaldomainofCx43mayrepresenttheprimarysite forphosphorylation,however,thereareothergapjunctionproteinssuchasCx36and Cx56thatshowthecytoplasmicloopsastheirphosphorylationsites[64].ItispossiblethatthedierentisoformsofCx43beresponsibleforexacerbatingtheseverityof PHorevenplayaroleincausalaspectsofPH.Thedatasupportingthishypothesis isseverelylackingandinvestigationstowardsitwouldbebenecialforimprovingthe 12

PAGE 29

therapeuticapproachesforPH. a Figure1.5:IllustrationofConnexin-43arrangementintheplasmamembrane. a EL,Extracellularloop;PM,Plasmamembrane;CL,Cytosolicloop;CT,C-Terminal;NT,NTerminal Althoughtheillustrationshowsmultiplesitesthatallowfordierentsignalingpathways,wedidnot evaluatethem.Thefocusonthecarboxylterminalisusedtoshowmultiplephosphorylationsites. Omasits etal. ,2014 SixCxsubunitscometogetherintheplasmamembranetoformahemi-channel, dubbed connexon ,thatcandockontoanotherconnexonsituatedintheplasmamembraneofaneighboringcell.Thisformsacompletegapjunctionchannel.Connexin isoformscancombinetoformeitheraheterotypicorheteromericgapjunctionalcomplex,forinstance:Cx43andCx40formaheteromericbond[65].Gapjunction channelscomposedofasingleconnexintypearedenedashomomeric.Thecombinationofconnexinsinvolvedandtheirabilitytointeracttoformconnexonscould 13

PAGE 30

suggestvaryingdegreeofregulationbetweenadjacentcells.Cx43isnowrecognized asthemostexpressedconnexin,identiedinatleast34tissuesand46celltypes[55]. Becauseofitspredominantpresenceoverotherconnexinsinmostcelllines,more isknownaboutCx43biology.Normalexpressioninvariouscelltypesandtissues, andaberrationsleadingtoavarietyofpathophysiologicalconditionsarealsoknown. Gapjunctionproteinsarecontinuallyforminganddegrading.Suchdynamismisevidencedbyreportsofconnexinhalf-livesbeinglessthan5hoursonculturedspecimens [66,67,68,69,70].Thissuggeststhattheturnoveralongwiththeassemblyofgap junctionsarelikelyessentialforintercellularcommunication[55].Onefocusofthe currentstudyisCx43expressionintherightventricleasamechanicalfailuremarker inthesettingofPH. 1.6RoleofCx43inSupportoftheWorkingHypothesis:VascularLeak andIncreasedInterstitialFluid So,wheredoesgapjunctionCx43tinwithrespecttomyocardialedemaand eventualrightheartfailureinPH? Asaforementioned,pericardialeusionanduidbuild-upinthefailingRVisone ofthekeyndingsinpulmonaryhypertension.PatientswithPH,regardlessofetiologyoranyofthecauseslistedinTable1.1,presentwithincreasedCVP[38].Elevated CVPadverselyaectsuidbalanceinthemyocardiumasmyocardiallymphoutow isnowexperiencinganincreasedpressure.Increasedllingpressuredepressesthe CO;a40%decreaseinCOisreportedwithcorresponding3.5%increaseinmyocardialuidcontent[5,39,71].TheclinicalsignicanceofRVedemaandpericardial eusionissteadilygainingrecognitionbutthedataexplainingtheorigin,maintenance,andincreasingseveritycausedbyuidimbalanceismissing.Thisparticular areaisofcardinalinteresttobothcliniciansandresearchersbecauseofitsassociation withhigherrisksofmorbidityandsignicancemortality[72].Theprognosisisnot onlypoorbutratherdiculttoaccuratelyassess.Elevatedplasmabrainnatriuretic 14

PAGE 31

peptideBNPlevels,increasedrightatrialpressures,andlowcardiacindexaresome ofthefactorsforpoorprognosisofPH[72,73,34].Additionally,RVdysfunctionand presenceofpericardialeusiongenerallyworsenstheprognosis[74,75,76]. Figure1.6:FunctionalconsequencesofabnormalCx43organizationandexpression seeninremodeledheart.Itisunclearwhethertheremodelingoftheheartpreceeds thedisorganizationofCx43orifitiscausal.Dottedlinesrepresentpotentialcausal interplay,solidlinesrepresentestablishedassociations. ThereisnomechanisticdatapresentfortheRVrelatingedematoCx43 mis-localization. Theroleofgapjunctionalproteinsininammationisindicated inthatgapjunctioninhibitorspromotevascularleakandtransvascularcellmigration [6].Thissuggeststhatgapjunctionshelpinmaintainingvascularbarrierfunction. Asidefrombeinglinkedtoarrhythmias,thefunctionalconsequencesofabnormal Cx43organizationintheRVhasnotbeenexhausted.Itisspeculated[77,78]that theconnexinsfoundinthelateralmembranesofcardiacmyocytesarelikelytobein theformofclosedhemichannels.Decreasedelectricalcouplinghasbeenobservedand reportedasaresultoflossofgapjunctionsinischemicanddamagedcells,although dyetransferofluciferyellowbetweenthecellswasobservednottobedecreased 15

PAGE 32

[79,80].Transferofuidhasbeenobservedthroughlateralizedgapjunctions,which suggeststhatCx43couldbecontributingtoedemaintheRV,alongwithinterstitial uidstasis,pericardialeusionresultinginineluctablefailureoftheheartFigure 1.6.ItisunclearwhenintheprogressionofPHthatCx43lateralizationmaybe important.Ifitoccurs,preliminarydata,discussedinChapter3,showsthatthereis extensiveredistributionofCx43inpreclinicalanimalmodelswithseverePHbutnot inmodelswithmildPH.ThecourseandextentofCx43redistributionisyettobe determined. TheexperimentalmodelillustratedinFigure1.6followstheunderstandingthat Cx43leadstoarrhythmogeniceventsinthemyocardium[63].Additionally,decreased plaquesizeanddown-regulationofimmunodetectableCx-43havebeenreportedin varioushumancardiacdiseases[61,81,82,83,84].Thereisapaucityindataregardingincidence,mechanism,prognosticsignicance,andtherapeutictreatmentsof arrhythmiasinpatientswithPH[85].SincearrhythmiasarenotcommoninPHand incidencemayvaryaccordingtoetiology,whetherarrhythmiasareconsequenceor causeofRVfailureinPHisstillupforspeculation.Butasaforementioned,Cx43 dysfunctionhasbeenlinkedtoarrhythmiasinvariouscardiacdiseases.Whenthe contractileapparatusoftheventriclesmalfunctionduetotachyarrhythmias,asseen inPH/citeRajdev2012,theheartisnotabletoeectivelydrainuidaectingthe myocardiallymphoutow.Wespeculatethat,asaresult,myocardialinterstitialuid isincreasedleadingtoprevalenceofedemaandpericardialeusion. 1.7ConnexinsintheMyocardium Cardiacmyocytes,cardiacbroblasts,endothelialcellsandvascularsmoothmusclecellsarethemajorcellularconstituentsoftheheart.Initialstudiesindetermining thecelltypes,performedbyZak[86]andNag[87]in1973and1980respectively, quantiedcellpopulationsbasedonmorphologicalcharacteristicsofleftventriclesof rats.Recentstudieshavenowincludedwholeheartsandtheevidencesuggeststhat 16

PAGE 33

thehumanheartcomprisesof60-70%nonmyocytesand30-40%cardiacmyocytes [88,89].Eventhoughtherearephysiologicaldierencesbetweenhumanheartand thepreclinicalanimalmodels,suchasheartrateandtotalcollagen[90,91],therat modelhasbecomeastandardpreclinicalmodelorganismtostudycardiacfunction alongwithstructuralandfunctionalchangesthatoccurinpulmonaryhypertension. Cardiacmyocytesarearrangedin2to5celllayerthicknessesinalaminaesurroundedbyanetworkofendomysialcollagen[92,93].Studiesthathaveinvestigated theorganizationofcardiacbroblastshavedocumentedtheirpresencewithinthe endomysialnetwork[94,95].Ithasbeenwellestablishedthatthecardiacmyocytes communicateamongstadjacentmyocytesthroughgapjunctions.However,littleis knownaboutthecross-talkthatoccursbetweenthemyocytesandthecardiacbroblasts.Similarly,littleisknownaboutthegapjunctionsthatarepresentbetween cardiacbroblastsandothercelltypesfoundintheheart. 1.7.1Fibroblast-FibroblastCommunication Althoughtheresearchdocumentingthedistributionofbroblastsintheheartis inshortsupply,thereissomeevidencethatbroblastsarenotuniformlydistributed throughouttheheart[89,96].Multipleinvestigationshaveshowntheregulationof broblaststovaryingdegreesdependingonpathophysiologicalconditions[94,95,97]. Forexample,broblastproliferationandtissueremodelingingeneralwithspreadand inltrationofbroblastsintounaectedtissuearecharacterizationofmultiplecardiac pathologies.Incasesofcardiacremodeling,asseeninpulmonaryhypertensionand inpostmyocardialinfarction,broblastnumbersappeartobedramaticallyincreased [98].Theintercellularcommunicationsthatoccurbetweenbroblast-broblastconnectionispresumedtoberegulatedbyconnexins,cadherins,andotherunknown molecules[55,99].However,theliteratureisunclearonwhetherconnexinsplaya majorroleinbeingtheprimarymeansofcommunicationbetweenbroblasts. 17

PAGE 34

1.7.2Myocyte-Fibroblastinteraction Kohl etal. havestudiedbroblastsinthesino-atrialSAnodewherebroblastsoutnumberthemyocytes.Theyfoundthatbroblastscommunicatewiththe impulse-generatingspecializedmyocytesintheSAnodeandactaseithercapacitors orconductors.Thissuggeststhatbroblastscouldaectthebehaviorofcardiacmyocytesinotherregionsoftheheart.Directinteractionsbetweencardiacmyocytesand broblastshavebeenshowntoberegulatedviacommunicationsbetweenCx40,Cx43, andCx45[88,89,100,101,102].Connexinshavebeenextensivelystudiedincardiacmyocytesandrecentndingssuggestthatsignalswithinthemicroenvironment mayregulatecellularhypertrophy[103].Extendingsuchstudiestocardiacbroblasts wouldgarnerdatathatcouldsuggestinvolvementofbroblastsinthree-dimensional signaling. Observationsthatbroblastsareincontactwiththemyocytesdoesnotnecessarilyimplythattherearedirectsignalsbeingsentand/orreceivedbetweenthetwo celltypes.Cardiacbroblastsarekeyplayersinnormalheartdevelopmentaswell asindiseasedstates.Theycontributetostructural,mechanicalandelectricalpropertiesofthemyocardium[100,104].Indiseasedstatesofthecardiovascularsystem, broblastsplayacentralroleinmyocardialremodelingbyproliferating,migratingto sub-endothelialareasofthetissue,andcontributingtothechangeinthecomposition oftheECM[89,105,106].Cardiacmyocytesarethecontractileregulatorsofthe heartandthus,understandingthecross-talkthatoccursbetweenthetwocell-types couldbeapivotalndingandcouldprovidestepsinobviatingRVfailurePH. 1.8NormalExpressions:LocationandAppearance Gapjunctions,adherensjunctions,occludingjunctionsandtightjunctionsall preferentiallyresideattheIDs[50].IDsarealsopopulatedbynon-junctional" molecules.Thenon-junctionalfractionoftheseproteinsincludemoleculesthatare notinvolveddirectlyinprovidingaphysicalcontinuumbetweenadjacentcells.The 18

PAGE 35

non-junctionalfractioniscomprisedofcytosolicloop,n-terminalandc-terminalcomponentscytoplasmicCx43,andCx43enroutetoeitherassemblyordegradation [107]. a Figure1.7:Organizationofconnexinsubunitstoformthehemichannelsubunitconnexon.Connexonshavemultipleconnexintypesformingeitheraheteromericora heterotypicchannel. a PublicDomainImage Connexin-43ischaracterizedviaimmunouorescencebypunctatelocalizationat theIDsofcardiacmyocytes.GapjunctionsarenormallyexpressedattheIDsina polarizedmannerandhaverelativelylowdensityatthelateralsides[81,108,109]. Themorphologyofthemyocyteandtheringlikestructuralorganizationofgap junctions[110]contributegreatlytothelowelectricalresistanceandasaresult, conductionofactionpotentialcanpropagateinbothlongitudinalandtransverse direction. 19

PAGE 36

1.9Aberrations Connexin-43hasbeenstudiedinthesettingsofhypertrophiccardiomyopathy HCM,dilatedcardiomyopathyDCM,andischemiccardiomyopathyICMamong otherend-stagecongestiveheartfailures.Inthesecasesthemostprominentfeaturesofgapjunctionremodelingincludeddown-regulationofCx43,reductionin gapjunctionalplaquesize,andincreasedheterogeneityofgapjunctiondistribution [107,111,63,112,113].Otherstudieshavereportedneoformationofgapjunctions atlocationsotherthantheIDs[50,114,115,116].Mostofthestudiesconducted onconnexinsandtheirfunctionshavebeenwithrespecttoarrhythmiasandother electricalaspects/events.Thus,theprocessofgapjunctionalremodelinghasbeen describedasapotentialcatalystforcardiacarrhythmias[114,83].Itcanbeargued, however,thatmis-localizationmayhelpmaintainactionpotentialpropagation.The patternofremodelingsimulatedincomputermodelingstudiespredictthatreductionofCx43upto40%incontentisunlikelytohaveamajorchangeinconduction velocity[117].Conductionvelocityisdenedasthespeedofanelectricalimpulse transmittedthroughexcitabletissue,asinmovementofanactionpotentialthrough theHis-Purkinjebersoftheheart.TheremodelingofcardiacmyocytegapjunctionscharacterizedbydecreasedCx43intheIDs,anditsredistributionthroughout theplasmamembraneofacardiacmyocyteisoftenknownaslateralization. Itiswellestablishedthatorganizationanddistributionofgapjunctionsisremarkablydierentacrossvariouscasesofcardiacdisorders[61,82].Sasanoand collaboratorshavestudiedtheexpressionofCx43inhypertrophiedrightventriclesof ratswithpulmonaryhypertension.TheyreportedinternalizationanddephosphorylationofCx43intheirpreclinicalanimalmodels[118].AsseeninFigure1.7,there aremultiplesitesontheC-terminalofCx43wherephosphorylationanddephosphorylationcanoccur.However,itcanbespeculatedthattheconnexindomainsfoundin thecytoplasmaretheprecursorsofthegapjunctionalcomplexesontheirwaytothe 20

PAGE 37

plasmamembrane,aprocessthatcouldbecorrelatedwithdiseasedstateoftheheart. SinceinammationincreasesCx43,asshownbyNavab etal. [119],immunoreactive connexinsfoundinthecytoplasmcouldsuggestneoformationofCx43.Cx43asa mechanicalmarkerishighlyunderstudiedintheRV.Mechanicalmarkerisaterm usedtoinferfunctionoftheratRVthroughexperimentsconductedforthisproject, sincewearenotdirectlymeasuringthefunction.ThefunctionalconsequencesofabnormalCx43expressionandorganizationintherightheartcouldincludepromotion ofvascularleakandinterstitialuidbuild-upthatinduceshypertrophicgrowthof individualmyocytes.Aprolongedstateofhypertrophyincreasesrisksofheartfailure [120,121],aconditionseeninpulmonaryhypertension. 1.10CellSpecicandLymphaticMarkers Thecardiaccellspecic,broblastspecicmarkers,andlymphaticsspecicmarkersdescribedinthefollowingsubsections,arechosentoaidinfurtherstrengthening thehypothesis.Byinvestigatingeachofthefollowingmarkersintheirrespectivetargettissue,weattempttoanswerthequestionofdecreasedlymphaticsintheheart, thedegreeofbrosis,andactivityofapoptosisandproliferation,allleadingtothe problemofedemaandmechanicalfailureoftherightheart.Theirpotentialcontributiontoedemaarelistedintherespectivesubsections.Useofasinglemarkercould leadtomisinterpretations,forinstance,markersthatlabellymphaticvesselssuch asLYVE-1alsolabelbloodvesselandmacrophages.Toavoidthispitfall,multiple markersareusedforIFimagingofheartandlungtissues.Theantibodieslistedinthe followingsectionsareamongthemostusefulmarkerstoinvestigatetheirrespective proteinsofinterest. 1.10.1CardiacMyocyteMarkers: -ActininandDystrophin -ActininandDystrophinbothbelongtothespectrinsuperfamily,aconserved familyofactin-bindingproteinsABPthatmaintainplasmamembraneintegrityand 21

PAGE 38

cytoskeletalstructure[122].Ofthefour -actiningenesfoundinhumans[123,124], -actinin-2isexpressedintheheart,morespecicallyinthesarcomericZ-discsand analogousdensebodies[125]. -actininbindstoactininamannerthatiscalcium ionindependentandalsolocalizeintheIDsofcardiacmyocytes[125,126]. -actinin crosslinkslamentousactinandtitinmoleculesfromadjoiningsarcomeresatZ-discs. Thus, -actinincontributestothestabilityofthesarcomericstructure. Dystrophinisaperipheralmembraneproteinthatidentiescardiacmyocyteand isknowntoshowcontinuousdistributionatthelateralplasmamembraneofthecell population[62].Althoughdystrophinisamembraneproteinforcardiacmyocytes andcontributeinformingpartofthetransversetubulespenetratingintothecell, theyarenotpresentintheIDs. Throughimmunouorescencestaining, -actininanddystrophinwillinformus ofacoupledierentthings.First,pairedwithDAPIstainingthatlabelsnuclei,the markerswillallowustovisuallylocalizecardiacmyocytesfromothercelltypesin themyocardium.Second,theprevalenceoftheseproteinswillallowustoquantify thenumberofcardiacmyocytesandhowthenumbersdierbetweenourpreclinical models. -actininanddystrophinarerepresentativeofthecontractileapparatusin themyocardium.Disturbancesintheseproteinsinthediseasedmodelswouldsuggest aberrationsinthemyocardialcontractilefunctionthatwouldleadtoaccumulation interstitialuidandthereforeedema. 1.10.2CardiacFibroblast:NarrowingDownaSpecicMarkerfor FibrosisandCardiacRemodeling Cardiacbroblastscomprisethemajorityofthenon-myocytepopulationinthe heart[88,89].Fibroblastshelpmaintainthestructural,mechanical,biochemical,and electricalpropertiesduringnormalcardiacfunction[127].However,relativelylittleis knownaboutthegeneticpathwayandotherfactorsthatcontributetothedierentiationalongthisparticularcellpopulation[128].Itissuggestedthatcardiacbroblasts 22

PAGE 39

areaheterogeneouspopulationthatoriginatefrommultiplesourcesinthebody[129], forinstancebone-marrowderivedcells,endothelialcells,andtheepicardium.Asaresult,ithasbeendiculttonarrowdownamarkerthatspecicallyrepresentscardiac broblasts.ItcanberatherdiculttoassesRVbrosisandtheroleofbroblasts inthepathologicalconditionswhentheassessmentisimpededbyabsenceofspecicmarkers.Forthisstudy,broblastspecicproteinFSP,transcriptionfactor-21 TCF21,andprolyl-4-hydroxylaseP4Hwereusedtodistinguishbroblastsincontrasttothecardiacmyocytesandattempttoassesbrosiswithrespecttoprevalence ofbroblastinregionsoftheRV. FibroblastspecicproteinFSP,alsoknownasS100A-4,labelslamentassociatedcalciumbindingproteinsexpressedbybroblastsinbrotictissues[130,131]. Throughdierentialhybridization,FSPhasbenidentiedinbroblastsbutnotin epithelialorembryonicendoderm[132].However,overtheyearsthespecicityof FSPtowardsbroblastshasbeenchallengedsuggestingthatotherinltrativecell typesduringinjuryalsoexpressFSP.Someofthesecelltypesaredendriticcells [133],vascularsmoothmusclecells[134],lymphocytes[135]andmacrophages[136]. Transcriptionfactor-21TCF21isaclassIIbasichelix-loop-helixbHLHtranscriptionfactorthatisknowntobinddeoxyribonucleicacidDNAthroughtheconsensusEboxsequenceCANNTGasaheterodimer[137].TCF21hasbeenextensivelyusedinidenticationofepicardialandproepicardialcells[138],however,its functionislessknownincardiacdevelopment.WhatmakesTCF21potentiallya goodmarkerforcardiacbroblastistheevidenceshowingthatTC21nullmicefailto developcardiacbroblasts[138].Acharya etal. showthatTCF21expressingepicardialcellsaremultipotentandareabletodevelopintoeithercoronaryvascularsmooth musclecellscVSMCsorcardiacbroblastwithTCF21becomingrestrictedtothe cardiacbroblastlineageovertime.Additionally,throughlineagetracingshowsthat TCF21isexcludedfromthecardiacmyocytepopulation[138]. 23

PAGE 40

Prolyl-4-hydroxylasesP4Hareenzymesthatcatalyzeirreversibleposttranslationalmodicationreactionstoeitheralterproteinconrmationandprotein-protein interaction,orenablefurthermodication[139].TheenzymesP4Hresideinthe endoplasmicreticulumandplayacriticalroleinbiosynthesisofcollagen[140].AntibodiesagainstP4Hidentifybroblastsbecausebroblastsproducecollagen[141]. P4Hareabletolabelportionofprocollagenmoleculecleavedbeforesecretionby broblastinresponsetoinjury,thussuggestingthatthemarkerisbestutilizedin labelingbroblaststhatareactivelydividingandnotthesubpopulationofthiscell typethatarequiescent[141]. 1.10.3ApoptoticandProliferationMarkers:Caspase-3,Ki-67 Currentandpaststudiessuggestthatcaspasesareessentialregulatorsofproteolyticcascadesthatoccurinresponsetoacelldeathstimulus,withcaspaseshaving rolesinregulationaswellasexecutionofapoptosis.Caspase-3CASP3isthebest characterizedmemberamongthe10-membersubfamilyofcaspases[142].Apoptosis hasbeenshowntocontributetolossofcardiacmyocytesandprogressivedeclineinLV andRVfunctionincongestiveheartdisease[143,144,145].Apoptosisisatypeofcell deaththatisdistinctfromcellnecrosis[146,147],associatedwithnotonlyCASP3 [148]butalsowiththecleavageofproteinkinaseC[149],polyADP-ribosepolymerase PARP[150],andcertainotherproteins[151].CASP3isdetectablebyimmunouorescenceincardiacmyocytes[152].ApoptosishasbeenlinkedtoPH,mainlyin idiopathicpulmonarybrosisIPF-associatedPH.Farkas etal. haveshownthatEC apoptosisreleasesvascularsmoothmusclecellsVSMCgrowthfactorsinbrotic regionsthatmaycontributetoaugmentedpulmonaryarterialmuscluarization[153]. Caspase-3activitywillinformusofapoptotisinourpreclinicalmodels.ThisisimportantsinceECapoptosisisshowntobedetrimentalinorganssuchasthelungs andtheheartandarepotentialcontributorstoPH[153]. 24

PAGE 41

Ki-67,alsoreferredtoasMKI67,isaproteinthatisassociatedwithcellproliferation[154].EvidenceshowingthatKi-67ispresentinallofactivestagesofcell cycleG 1 ,S,G 2 ,andmitosisbutisabsentfromtherestingphaseG 0 makesit anexcellentmarkerindeterminingthegrowth-fractionofcellpopulation[155].Proliferation,likeapoptosis,contributestoformationoflesionsduetostructuraland functionalchangestothearchitectureofthewallsofthepulmonaryarteries.In PH,thisleadstoincreasedmuscularizationofthepulmonaryarteriesandperipheral vessels,formationofneointimaandformationofplexiformlesions[156]. Thesespecicmarkerswerechosentoqualitativelyandquantitativelyassessthe varyingdegreeofcelldeathandproliferationintheprogressionofPH.Furthermore, anydown-regulationorup-regulationinproteinsofinterestcanbecorrelatedwith whethertheincreaseordecreaseincellgrowth-fractioniscontributingtotheirexpressionlevels. 1.10.4MarkersforLymphangiogenesisandAngiogenesis:Lyve-1, Podoplanin,Prox1 Lymphangiogenesisandangiogenesisareessentialfornormaldevelopmentand physiologicalprocesses[157].Morerecently,thepathologicalprocessesinthecardiopulmonaryregionhavegarneredinterestsinlymphangiogenesisandangiogenesis aspectsandasaresult,beenrecognizedaspotentialtherapeutictargetsfordiseases. Thedicultyinusingmarkersforlymphaticand/orbloodvesselsarisesfromcells thatmakeuptherespectivevesselsarenothomogeneousandshowvariousphenotypicvariationdependingonlocationandfunctionalstateinammationvsnormal growthforinstance[157]. Lyve-1standsforlymphaticvesselendothelialhyaluronanreceptor1andisone ofthebestcharacterizedmarkersforlymphaticendothelialcellsLECs[158].Lyve1stronglyidentiesontheentireluminalandabluminalsurfaceoftheLECsas shownbyBalukandMcDonald.Theproteinactsasareceptorandbindstoboth 25

PAGE 42

immobilizedandsolublehyaluronansuggestingthatLyve-1mayfunctioninlymphatic hyaluronantransport[159]. Podoplaninisexpressedinmostlymphaticvesselswhereitmaycontributeincell adhesion[160].Podoplaninisaproteindetectedinkidneypodocytes[161],andis alsoknowntobeexpressedinepithelialTypeIcellsinthelungs[162].Thespecic functionofPodoplaninhasnotbeendeterminedbutithasbeenproposedasamarker forlunginjury.Podoplaninisfoundinskincarcinomas[163]amongotherpathologic situations,suggestingthatthedistributionofthismarkerisdiverseinhumantissue. However,wehavedata,Figure1.8,showingco-localizationofPodoplaninandProx1 anotherlymphaticmarkerdiscussednextwhichsuggeststhatPodoplaninisan adequatechoiceforalymphaticmarker,atleastintheregionthatourstudyis interestedin. Figure1.8:ColocalizationofPodoplaninandProx1inthetertiarylymphoidtissue BALTsshowingpositivecorrelationbetweenthetwolymphaticmarkers. Prox1,alsoknownasProsperohomeobox1,isatranscriptionfactorthatlabels LECsinearlystagesofembryonicdevelopment[164,165].Prox1labelsthenuclei ofLECsandisspeculatedtobeamastercontrolgeneinuencingtheexpressionof 26

PAGE 43

otherlymphaticmarkers[164,165].Prox1hasbeenfoundtoconsistentlyexpressed inadultlymphaticendothelium[166]anditsexpressionisfairlycorrelatedwiththat ofPodoplaninasevidencedinFigure1.8.However,asmallsetofcapillaryand vasculaturewithirregularsmoothmuscleprolelininghavebeenshowntoexpress endoethelialProx1[165,166]. Theoverallfunctionofthelymphaticsistocollectleakedplasmaandinterstitial uidfortheirreturntobloodvasculature.Thelymphaticsarenotonlyresponsiblefor transportofuidbutalsotheuptakeanddegradationofdissolvedmacromolecules, oneofwhichishyaluronan[167,168].Weexpecttoseedecreasedlymphaticsinthe RVofPHrats.Asaresult,thedrainageapparatusofthelymphoutowisdisturbed causinginterstitialuidbuildupandedema.Additionally,hyaluronanhomeostasis isaected.Hyaluronanundergoesconstantturnover,aprocesswhichcomprisesof removingthemacromoleculefromthetissueintotheaerentlymph,degradation withthelymphnodes,andremovalthrougheerentlymph[169].Becausehyaluronan degradationreleasesproductsthatarepotentiallypro-inammatory[170],inecient removalofhyaluronancouldbecontributingtotheprogressionofPH.Wearenot lookingintohyaluronanmetabolismforthisprojectbutitshouldbekeptinmindof theirroleintheprogressionofPHwithrespecttolymphaticsintheRV. 27

PAGE 44

2.MethodsandMaterials 2.1AnimalModels SpragueDawleyandWistarratswereusedtoharvestthelungsandheartsinvestigatedforthestudy.AcontrolgroupofratswereacclimatedtoDenver'selevation of5,285feet.ThecontrolgroupwascontrastedtothreediseasedgroupsofPH, ofwhichtwomodelsstudiedhaveirreversibleandineluctableRVfailure.Thediseasedmodelsare:1reversiblePHinducedduetochronichypoxiafromexposureto hypobaricchamberstimulationofanelevationof18,000feet,2PHinduceddueto monocrotalineMCTadministration,and3PHinducedwithSugen5416followedby exposuretochronichypoxia.Thestimulationofhighelevationapproximatelymimics pressurevalueof0.50baror380torr.Controlgroupofratsexperiencepressureof 0.83baror621torr. MCT,an11-memberedmacrocyclicpyrrolizidinealkaloidPAderivedfromthe seedsof Crotalariaspectabilis plant,isactivatedintoareactivepyrroleintheliver bycytochromeP-450[171,172].ThereactivepyrrolemetabolitedehydromonocrotalineMCTPinjurespulmonaryendothelialcells[173]causingthemtodevelop megalocytosis,displacementofendothelialnitricoxideNOsynthase,anddecreased cell-surface/caveolarNO[174].MCTinducedlossofmembraneproteinsstimulate proliferation,activationofanti-apoptoticfactors,anddysregulationofNOsignaling [175].MCTPactsasavesselobliterationagent,obstructingthepulmonaryarteries andincreasingvascularresistance.TheMCTinducedendothelialcelldamage,interstitialpulmonarybrosis[176],andpulmonaryarterialmedialhypertrophy[177] amongothercharacteristicsleadtochangesinthelungvasculatureandirreversible remodelinginthisexperimentalmodelofPH.Studiesshowsignicantchangesin pulmonaryarterypressure,medialthicknessofthesmallpulmonaryarteries,andRV hypertrophyoccurring3rdand4thweekspostMCTadministration[178,179].Thus, weinvestigatedratsatthosetimepointsforoutRVfailuremodels.Additionally,the 28

PAGE 45

MCTsyndromeinducesmyocarditisinboththeleftandrightventricles,acutelung injury,interstitialpulmonarybrosis,andhepaticvenooclusivedisease[177]. Sugen5416morecommonlyknownasSugenisasmallmoleculereceptortyrosinekinaseRTKinhibitorthatwasdiscoveredthroughascreeningprocessdesignedtoidentifycompoundsthatinhibitvascularendothelialgrowthfactorVEGFstimulatedendothelialproliferation[180].Sugen5416actstoinhibitVEGFreceptor1 [181]andVEGFreceptor2[180],andpromptspulmonaryendothelialcellapoptosis andlossofsmalllungvessels[182].Sugen5416,whencombinedwithchronichypoxic environment,causessevereangio-obliterativePHandrightheartfailure[183].The VEGFreceptorblockadeinducesendothelialcellapoptosiswhichleadstoprogressive proliferativeendotheliopathy[184].ThisSugen-hypoxiaSuHxmodelisanimportantpreclinicalmodelofPH[185,183]asitcausesangio-obliterativelesionsinthe pulmonaryarteriolesthataresimilartotheplexiformlesionsfoundinhumancases ofidiopathicPH[186]. Forthepurposethisresearch,preliminarydataincludeSuHxandMCTmodels incontrasttocontrolsbutlaterstudiesincorporateonlytheMCTpreclinicalmodel ofRVfailurePHjuxtaposedtothecontrolgroups. 2.2ObtainingTissue Forstainingpurposes,theheartoftheratswereharvestedfromthepreclinical models.Therightandtheleftatriaareexcisedout,followedbythebaseandtheapex beingseparatedfromthemainbodybycuttingthedesiredregionoftheheart,as shownbythecutmarksinFigure2.1.Theapexisincludedinwiththedesiredregion attimeswhentheratheartisrelativelysmallandtsintothecryomolddimensions. ThepartitionthatremainsisthenplacedintotheOCTsolutionandisallowedto freeze.OCTstandsforoptimalcuttingtemperature.TheOCTblockcontainingthe tissueisthenfrozenandstoredin-80 o C.Thefaceontheblockseeninthegureis slicedandcollectedonafrostedglassslide.Thethicknessoftheheartsectioncutis 29

PAGE 46

intherangeof5mmto7mm.Thecutsectionwhenslicedandplacedontoaglass slidecontainsboththeLVandRVwiththeformerbeingthecircularlargerstructure seeninthegure.TheorientationofthespecimenplacedontheOCTblockissuch that,whenslicedontoaglassslideatdesiredthicknessrangeof5to150 mand multiplesubsequentslidesareimaged,theobserverismakingtheirwayfromthebase totheapex,i.esuperiortoinferior. a Figure2.1:Hearttissueobtainedfromthepreclinicalmodelsincludetherightand leftventricles.Therightandleftatriaareremovedpriortocuttingthetissuedesired forourassays.Thewholeheartontheotherhandisutilizedtopreparehomogenates forWesternblotexperiments.TheWesternblotassayperformedforthisstudyused RVhomogenates. a Lohani,2015 2.3WesternBlotProteinAssay Westernblot,alsoreferredtoastheproteinimmunoblot,isawidelyutilized analyticalassayforthepurposeofdetectingspecicproteinsinasampleoftissue homogenate.Thisparticulartechniquewasperformedforhearttissuesincontroland RVfailuremodelsofPH.Samplestakenwerefromrightventricletissueonly.Solid 30

PAGE 47

tissuewererstbrokendownmechanicallyusingahomogenizer.Duringthisstep, thetissuesampleswerehandledinextremecoldtemperaturestopreventanydenaturingofproteinsorspecimendegradation.Proteaseandphospataseinhibitortablets PhosSTOP,cOmpleteMini;RocheHoldingAGwereaddedtopreventdigestionof thehomogenateandpreservephosphorylationstateofproteins.Althoughadditionof phosphotaseinhibitortothehomogenatepreventsphosphorylationofproteins,differentisoformsofproteinscouldalreadybepresentatthetimeoftissueretrieval. TheproteinsinthesamplewerethenseparatedbymolecularweightinkiloDaltons: kDaviagel-electrophoresis.TherunningbuercomprisedofMES-NuPAGE R -SDS polyacrylamidegelelectrophoresissodiumdodecylsulfatethatmaintainspolypeptidesindenaturedforms,asaresult,allowingforseparationofproteinsbymolecular weight.Thesettingusedforthisstepofthewesternblotassaywas:1hourrun-time, 200volts.Sampleswereloadedintoa10-wellgelwithtriplicatesofcontrol,MCTand Su5416-hypoxiasamplesloadedinwellslots1to9.The10thwellwasloadedwith aladderthatshowedcoloredbandssignifyingproteinsofdierentmolecularweight forconvenientdeductionofsizesinkDa.Theproteins,onceseparated,werethen transferedontoapolyvinylidenediouridePVDFmembranetoallowforantibody detectionviaaprocesscalledelectro-blotting.ThetransferwasconrmedwithPonceaustainingofthemembrane.Topreventnon-specicinteractionsandbindings,a blockingsolutionof5%non-fatdrymilkinTris-bueredsalineTBSwithTween-20 wasused.Blockingwasperformedfor1hourfollowedbymultiplewashcyclesbefore applicationofprimaryantibody.Theantibodiesusedalongwiththeirappropriate dilutionsareprovidedinTable2.1. Atwo-stepincubationprocedurewasadaptedforprobingofproteinsinthemembrane:primaryantibodyincubationfortherststep,followedbyrinsingofmembraneforanyunboundprimaryantibodiesandsecondaryantibodyincubationfor thesecondstep.Bothincubationintervalswereperformedfor1houreach.The 31

PAGE 48

secondaryantibodyusedwasconjugatedwithhorseradishperoxidaseHRPallowingformultiplesecondaryantibodiesbindingtooneprimaryantibodyandcleavea chemiluminescentagenttoproduceluminescencethatisproportionatetotheamount ofproteinpresent.EnhancedversionofthechemiluminescencereactionECL TM wasusedtodetectthelightemittedwhentheluminolreagentisoxidizedbyHRP presentinthemembrane. Table2.1:AntibodiesusedforWesternBlotAssay AntibodyTargetTissueVendorSpeciesCloneConcentrations PrimaryAntibodies -ActininHeartSigmaMouseA78111:3,000 Connexin-43HeartNovusRabbitNBP1-675301:500 GAPDHHeartAbcamRabbitab371681:10,000 SecondaryAntibodies anti-mouseHRPHeartMilliporeGoatAP-124P1:10,000 anti-rabbitHRPHeartSantaCruzGoatsc-20041:10,000 a a GAPDH,Glyceraldehyde3-phosphatedehydrogenase;HRP,Horseradishperoxidase Themembranesusedtoprobefortheproteinsofinterestwerestrippedin Restore TM strippingbuerThermo-Scientic TM thenincubatedwithGlyceraldehyde3-phosphatedehydrogenaseGAPDH.GAPDHisanenzymeof37kDaand wasusedfornormalizationduringthequanticationstepofwesternblot.Theblot wasquantiedusingImageJ.Thebuilt-inGelAnalyzer"functionwasusedtomeasurethedensityoftheofimmunoblots. 2.4ImmunouorescenceStaining Histologicalsectionsoftheratheartswerestainedwithuorescent-labeledantibodiestoinvestigatetheantibody-antigencomplexesincontrolanddiseasedconditions.Therightventricleoftheratheartwastheprimaryareaofinterest. 32

PAGE 49

2.4.1IFandAntibodyStainingTechniques Thehistologicalsectionsrangefromathicknessof5micrometersto200micrometers.ThethinsectionswerestandardsforimmunouorescenceIFstainingwhen optimizingconcentrationofantibodiesandobtainingpreliminarydata.Thickersectionsfacilitatedtheuseofconfocalmicroscopyandutilizedtheadvantageousfeature ofopticalsectioning.Hence,confocalprovideduswithbetteranatomicalcontext ofthetissueaswellassuperiorimagesoveraconventionalIFmicroscope.These sectionswerecutusingHM505EMicromcryostat.Tissuesamples,oncesliced ontoglassslides,werexedwith1:1solutionofMethanol:Acetonefor10minutes at-20 o C.Tissuesectionswiththickness > 20 mwerexedfor2hoursminimum. Afterxation,theslideswereallowedtoair-dryfollowedbyre-hydrationusing1X PBSPhosphate-bueredsaline.Theslideswerethenblockedwith1:1solutionof FBS:PBSFetalbovineserum:Phosphate-bueredsaline.Incubationtimevaried withthicknessofthetissuesample:30minutesforsectionsrangingfrom5to20 micrometers,2hoursfor50 msectionsandapproximately8hoursforsectionsnorth of100 m.Thevaryingincubationtimes,especiallyforthethickersections,were incorporatedintothestainingprotocoltoallowforadequateandevenstainingof thehearttissues.Becausewewerelabelingmuscularizedtissue,longerincubation timesallowedfordeeperpenetrationoftheantibodiesthroughoutthethicknessof thetissue.Theprimaryantibodies,dilutedin5%FBSinPBSsolution,wereused tolabeltherespectivetargetproteins.Theprimaryantibodiesusedalongwiththeir appropriatedilutionsandwhethertheywereusedasidentifyingmarkersfortheheart orthelungsareprovidedinTable2.2. ThesecondaryantibodiesusedwereAlexaFluor R 488andAlexaFluor R 594at dilutionsof1:750in5%FBS:PBSsolution.Thesectionswereincubatedovernightat 4degreeCelsiusforprimaryandsecondaryincubations.Forthickersections,however,anadditionaldoseofprimaryantibodieswasadministeredpostwashfollowing 33

PAGE 50

Table2.2:Primaryantibodiesandtheiroptimumdilutions Antibody TargetTissueVendorSpeciesCloneConcentrations -Actinin HeartSigmaMouseA78111:250 Aquaporin-1 Heart&LungNovusRabbitNB600-7491:400 Claudin-5 Heart&LungSantaCruzRabbitsc-286701:200 Connexin-43 HeartMIlliporeMouseMAB30681:200 Heart&LungNovusRabbitNBP1-675301:200 Dystrophin HeartAbcamRabbitab152771:300 Ki-67 Heart&LungAbcamRabbitab166671:20 LYVE-1 LungR&DSystemsSheepAF79391:20 Podoplanin Heart&LungNovusMouseNBP110-964231:10,000 P4H HeartAcrisMouseAF5110-11:100 PROX-1 Heart&LungAngioBioRabbit110021:200 TCF21 HeartNovusRabbitNB110372571:500 a a LYVE,lymphaticvesselendothelialhyaluronanreceptor;P4H,Prolyl-4-hydroxylase;PROX, Prosperohomeobox;TCF,Transcriptionfactor Someofthemarkerswereinvestigatedinthelungtissuesaswelldatanotshown. theovernightincubationtoallowforadequatesaturationofthetissuewiththeimmunolabels.Thisincreasedtheincubationtimewithprimaryantibodiesto24hours. Performingsimilarstepswiththesecondaryresultedinthenon-specicbinding.In otherwords,thesecondaryantibodieswerelabelingproteinsintheentiretyofthe sectionandnotjustthetargettissueslabeledwithprimaryantibodies,oroversaturatingtheproteinoninterest.Forsections5 mto20 m,wefoundthatitwas unnecessarytoincubatethetissueovernightforsecondaryantibodies.Resultswere optimalwiththespecimenincubatedwithprimaryantibodiesovernightregardlessof thethicknessofthetissue.Theslideswerethoroughlywashedwith1XPBSmultipletimesbetweentheincubationstepofprimaryantibodiesandtheintroduction 34

PAGE 51

stepofsecondaryantibodies.Immunolabeledsectionswerethenfollowedwithan additionalstepofwashingpostincubationofsecondaryanibodiesandmountedwith VectaShield R mountingsolutionwithDAPIVectorLaboratoriesIncorporated.The sectionswerethenexaminedunderaZeissuorescentmicroscope.TheconfocalimageswereacquiredusingtheZeissLSM780andZENdigitalimagingsystemCarl ZeissIncorporated,Thornwood,NY.ForthecontrolandtheRVfailureMCTpreclinicalmodels,IFstainingwasperformedinatleasttriplicates,onthepanelofve controlandfourMCTmodels. 2.5ConfocalMicroscopy Confocalmicroscopytechniqueoersseveraladvantagesoverwide-eldmicroscopy.Zeiss780LSMwasusedtoobtainhigh-resolutionimagesincontrastto conventionaluorescentmicroscopy.Theconfocalmicroscopyallowsforproduction ofthinserialopticalsectionsintherangeof0.5to1.5 mthroughtissuethathave athicknessof50 morgreater.Additionally,theimageinformationgatheredisrestrictedtoawell-denedplane,contrastisdramaticallyimprovedasthebackground uorescenceandauto-uorescencearereduced,andsignal-to-noiseratioSNRisimproved.Serialopticalsectionswereacquiredattheintervalof1 m,40xobjective magnication,and1024x1024-pixelimagesize.Thisyieldsapixelsizeof0.21 m. Additionally,imageswhereazoom-factorof2wasappliedinZEN2012software, 1024x1024-pixelimagesizewithsimilarsettingsyieldedapixelsizeof0.104 m. 2.5.1Rationale:OutoftheStaticStudiesofThinSections Confocalmicroscopyhasturnedouttobeapopulartechniquetobeutilized inrecentyears,dueinparttotheconvenienceofobtaininghigh-qualityimagesfor tissuesamplespreparedforconventionaluorescentmicroscopy.Forseveralreasons, wewantedtohaveaglobalperspectivewhenlookingattheharvestedtissue.First, itisdiculttoexaminehearttissuesfromhumans,astheyarenotreadilyavailable. Addingtothatisthelimitationofexaminingthespecimenin invivo settings. Invivo 35

PAGE 52

studiesareemergingandhearttissuesarebeingengineeredforinvestigationincloseto naturalconditions.Conventionalmethodsforimmunouorescencestainingtypically includethinsectionsofspecimen.Westainedhearttissuesofthicknessrangingupto 200 mwithecientstainingof40%ofthetissue,approximately80 m.Theprimary reasonforchoosingthicksectionswasduetothesizeofcardiacmyocytes.Ahealthy adultcardiacmyocyteis100-150 mlongandaround10-25 m[62].Wewanted tobeabletocaptureasmuchinformationaspossiblefromtheRVandsincethe primarylocationofinterestisthecell-celljunctionsbetweeneithercardiomyocytes orcardiomyoctyestobroblasts,itwasimperativeoftheteamtoattemptstaining thickersectionsofthespecimen. Figure2.2:AImmunouorescencestainof -actininwithlongitudinalorientation withthez-discsofthesarcomeredistinguishablewhereasBhelpsexplainthedicultyencounteredwhendealingwithhearttissue.Thecomplexstereologyofheart tissuecausestheobservationviewtochangefromlongitudinalyellowarrowto transversalsectionsgreenarrow. 36

PAGE 53

2.5.2Why3Dover2D:WhorlingEect/Stereology Evenwiththicksections,thestereologyandtheorientationofthehearttissue cutontoglassslideswereunpredictable.Figure2.2showsthedierentorientations encounteredwhenobservingthehearttissue.Figure2.2ashowsthetextbook representationofhearttissuewherethecardiomyocytesareseenlongitudinally.The stainof -Actininclearlyshowsthatthecardiomyocytesareorientedlongitudinally. Figure2.2bontheotherhandrepresentsanorientationthatshowscross-sectional viewofthecardiomyocytesinatransversalplane.Tokeeptheobservationandthe planeofvisualizationconsistentonthin2Dsectionsisratherdicultifnotimpossible aswevisuallyobserveallthedierentplanesinbetween.With3Dreconstructionof theanatomyoftherightventricle,wearedeterminedtoexplorethecellularmakeup ofthecomplexstructureoftheRV,junctionallateralizationanddysfunction,water channels,andup-ordown-regulationoflymphatics.With3Dreconstructedimages, wewoldbeabletocharacterizegapjunctionplaquesalongthemyocyte'sterminus. Furthermore,with3Dreconstructionwewouldbeabletogetvisualperspectivesfrom dierentplanesotherwiselimitedby2Dimages. 37

PAGE 54

3.Results 3.1AssessmentofPHinPreclinicalModels Figure3.1showsthemacroscopicrepresentationofthemorphologyofhearttissue. Withoutanyfurtherquantitativeanalysis,weareabletoobservethattheRVis presentingwithcharacteristicsofhypertrophyinhearttissuefromPHrats.The thicknessintheRVfreewallisincreased,andachangeintheLVwallthickness isobserved.Thecrosssectionsofthetissueseeninthegureweretakenbetween thebaseandtheapexasillustratedinFigure2.1inChapter2.TheMCT-treated heartharvestedfromthepreclinicalmodelsisseverelymuscularizedwheretheLV chamber'svolumecapacityappearstobeaectedduetohypertrophy.Theseimages wereobtainedduringthesectioningphaseofexperimentwhenthehearttissueis slicedontoaglassslide.ThemicroscopichistologicalimagesoftheRVstainedwith immunouorescenceantibodiesfollowinthecomingsubsections. a Figure3.1:Macroscopicrepresentationofmorphologyoftheheartfrompreclinical ratmodels.ThisgureshowscrosssectionofAcontrolandBtissue4weekpost MCTadministration a IVS,interventricularseptum 38

PAGE 55

Additionally,pentachromestainswereperformedon5 mthinsectionsofcontrol and3weekMCTgroups.Figure3.2showsacontrolslicejuxtaposedtoa3week MCTtissueslide. Figure3.2:Macroscopicrepresentationofpentachromestainonsingle5 mthin sectionsoftheheart.Thisgureshowscrosssectionofcontrol,Aand3weekpost MCTadministration,B MyocardialedemaforMCT-andSuHx-treatedratswasmeasuredintheRV asdescribedbyLaine etal. .Theamountofmyocardialedema,alsoreferredtoas theextravascularuidEVFwasobtainedfromtheunitlessblood-freeventricles [71].Thetotaltissuewatercontentcomprisedofvascularwater,interstitialwater andcellularwater.Aspectrophotometriccorrectionwasappliedforbloodvolume sincethebloodvolume'svariabilitymaysignicantlychangemeasurementparameters throughouttheexperiment.Additionally,myocardialedemaintheLVandRVofrats intheothertwodiseasedmodelsweremeasuredaswell.TheformulafortheEVFis asfollows: EVF = wetweight )]TJ/F18 11.9552 Tf 11.955 0 Td [(dryweight dryweight ThepresenceofuidintheRVsofMCT-treatedpreclinicalmodelswassignicantlyhigherthanthoseofthecontrols.Anincreaseinmyocardialinterstitialuid 39

PAGE 56

fromacontrolEVFof2.9to3.7representsachangefromwatercontentof75%to 79%.SimilarresultswerefoundfortheLVofMCT-treatedratswheretheLVsof MCT-treatedratshadsignicantamountofedemaincomparisontohypoxicand Su5416-hypoxiarats.TheRVofratstreatedwithSu5416-hypoxiaalsoshowedsignicantlymoreedemathancontrols.ThendingsaresummarizedinFigure3.3. a Figure3.3:MeasurementsofmyocardialedemaforRVsandLVsinvariouspreclinical modelsofPH.Valuesarereportedasmean St.Dev a *,p < 0.05,statisticallysignicant;St.Dev,standarddeviation;mPAP,meanpulmonaryarterial pressure ThepresenceofedemaintheRVsoftheMCT-treatedratswasalsodetermined viaT2-weightedmagneticresonanceimagingMRI.AsshowninFigure3.4,the4 weekMCT-treatedratsshowhigherprevalenceofuidintheRV.TheT2-weighter MRimagesshownhereareprecursorsfortheEVFdataprovidedinFigure3.3.EVF datawascompiledfromthesameanimalsfromwhichweobtainedtheT2-weighted MRIimages. 40

PAGE 57

T2-weightedMRIndingsandEVFmeasurementspositivelycorrelate.ApositiveadvantageofusingT2-weightedMRIisthenon-invasiveandnon-lethalroute ofobtainingdata.Therefore,thismethodcanbeusedonnotonlyrats,butalso humans.ThisisastrongadvantageoverusingEVFtoquantifythepresenceofmyocardialedemainourpreclinicalmodelsasthetissuetakenfromratsisnotviable foradditionalexperimentsfollowingmeasurementofEVF. Figure3.4:T2-weightedMRIshowingedematousRVforMCT-treatedRatright comparedtothecontrolleft.ImagesacquiredusingBruker4.7TeslaMR. Sinceinterstitialuidstasisisassociatedwithinammationandbrosisinsome diseasedsettings,ourlabgroupproceededtoassesbroticchangesintheRVof hypertensiverats.Masson'strichromestainingwasperformedandextensivecollagen depositionwasobservedintheMCT-treatedrats.Theincreasedtransmuralcollagen foundintheRVpositivelycorrelatedwiththepresenceofEVF.Wermlybelieve thatedemapreceedsbrosis.AsshowninFigure3.5,theRVofthe4weekMCTis broticcomparedtothecontrol.ThedistinctionispresentbetweentheratRVsin controlandtheMCT-inducedPHmodels:theventricleisedematous,asshownby T2-weightedMRimagesandconrmedviaEVFmeasurement,andbrosisispresent intjePHventricles.Thecorrelationispresentbetweenedemaandbrosis,leading ustobelievethattheextensivecollagendpositionintheMCT-treatedRVsfollows 41

PAGE 58

accumulationofinterstitialuidandmyocardialedema,Figure3.6. Figure3.5:Masson'strichromestainassessingbroticchangeincontroltopleft vs.hypoxictoprightandMCTbottomrow.Hypoxicratsdemonstrateasimilar patternofmodestcollagencontentintheRVascontrol.Elevatedcollagendeposition intheRVofMCT-treatedratscorrelatespositivelywiththepresenceofEVF. Figure3.6:CorrelationbetweenEVFandFibrosissupportsourspeculationsabout edemapreceedingbrosisinPH. 42

PAGE 59

3.2ImmunouorescenceStaining:QualitativeResults 3.2.1Connexin-43OrganizationandExpressionintheNormalHeart ThecharacteristicdistributionofgapjunctionalproteinsCx43intheleftventricularmyocardiumiswellestablished.Cx43arecharacterizedbypunctatelocalization attheIDs.AsseeninFigure3.7,thecontractilemyocytesoftherightventricleare extensivelyinterconnectedbyclusturesofCx43greenuorescencecontainingjunctions.ThelongitudinalsectionseeninFigure3.7showsgapjunctionalproteinCx43 appearinginrows,correspondingtoedge-onviewedIDs.Reduorescencerepresents -actininandDAPIrepresentnuclei.Cx43appeartobewelldemarcatedattheIDs, Figure3.7B.InFigure3.7A,weobserveCx43expressionatlocationsotherthan theIDs.AsshowninFigure1.4,theisolatedratRVshaveverticalstep-likefeature. Severs etal. haveshownviathin-sectionelectronmicroscopythatattheselocations intheperipheryofthecardiacmyocytes,therearecomponentsofgapjunctionsand desmosomesconnectedwithfasciaeadherentjunctions.ThesearereferredtoasextendedcomponentsoftheIDs[61].Thus,theimmunodetectableCx43observedat whatseemlikelateralsurfacescouldbegapjunctionspresentattheextendedIDs. 3.2.2GapJunctionRemodelinginRVFailurePH Diseaserelatedalterationsinmyocardialgapjunctionorganizationhavegained wideattention,especiallyintheconnexin'scontributiontoabnormalpropagation ofactionpotentialsandarrhythmia[82,84].Thesecontributionscouldbedueto decreasedexpression,decreasedplaquesize,andlateralization.However,itisof absoluteimportancetonotethatarrhythmiasaremulti-factorial,comprisingofnot onlygap-junctionalcouplingbutalsomembraneexcitabilityandarchitecturalblock makingupthetissue[187].Thus,theoriginofarrhythmogeniceventscannotbe associatedwithconnexinremodelingalone. Lateralizationisatermdescribingaprominentfeatureofthegap-junctionsthat arefoundthroughouttheplasmamembraneofcardiacmyocytes.Lateraldisposition 43

PAGE 60

Figure3.7:DistributionpatternofCx43incontrolventricularmyocardiumALongitudinalsectionfromratRVBZ-projectionofconfocalimagesshowapparentgap junctionimmuno-labelingatthelateralsurfaces.Theseproteinscouldbeconsidered componentsofextendedIDsseeFigure1.4.MostCx43gapjunctionsareobserved attheperipheryofthediscs. ofCx43wasnotobservedintheIFimagesofcontrolandMCT-treatedratRVs. TheCx43immunolabelsaremostlyfoundattheIDsalbeitinsmallerareasand 44

PAGE 61

inpatchesinwhichtheyarefewerorabsent,otherwisenormal.Theamountof immunodetectableCx43intheRVofMCT-treatedpreclinicalmodelsshowdecreased levelsofthegapjunctionprotein,observedqualitatively.ThereductionofCx43is shownwidelyirrespectiveofwhethertheheartfailureisduetoidiopathicdilated cardiomyopathy,ischemicheartdisease,oraorticstenosis[111,113,188].We,like thestudiesbeforeus,observedCx43expressiontobespatiallyheterogeneous. 3.2.3Down-regulationofConnexin43and -ActinininMCTmodelof PHasobservedthroughimmunouorescence Studyingtheimmunouorescentimagesofcontroland4weekpostMCTtreatment,itisclearthatthereisreductioninCx43levelsasindicatedbytherepresentativepatternseeninRVfailuremodelvs.itspresenceseenincontrolgroup, Figures3.8and3.9.ThegapjunctionproteinisconnedtotheIDsofboththe controlandMCT-treatedratgroups.However,carefullyobservingthelongitudinal sections,theMCT-treatedratRVsshowdisturbancesintheorganizationalpattern incomparisontothecontrolRVs.Figure3.8BshowsdecreasedorabsentCx43 proteinsattheIDscomparedtocontrolA.Figure3.9Bshowsahigherdegreeof disorganizationarrowswhereCx43plaquesaresituatedatlocationsoutsidetheconnedspaceoftheIDs.TherearefewerimmunodetectableCx43intheMCT-treated ratRVs.WhetherthereductioninexpressionoccurswithprogressingseverityofPH isyettobedetermined.Obtainingimagesasafunctionoftimeat2weeksand3 weeksalongwithwesterndatasupportingtheimmunouorescentimageswouldaid inansweringthequestion.ThendingthatCx43levelisdecreasedat4weekswas veriedwithwesternblotandtheresultsareprovidedintheSection3.4. Fluorescentlyimmunodetectable -actininwasobservedtobedramaticallydownregulatedintheMCT-treatedRVfailuremodelofPH.Thediscrepancyinexpression betweenthecontrolandthediseasedmodelisillustratedinFigure3.8.Themorphologyofthecardiacmyocytesarevisuallyalteredwherethecoverageofthecontractile 45

PAGE 62

-actininislesserintheMCT-treatedRVs.Theimmunouorescentimagesofthe controlMCT-treategroupsshowdistinctstainingofmyobrilarZ-bands,butthe registrationoftheantibodiesisdecreasedinthelatergroup.Additionally, -actinin stainingshowsthearrayofZ-bandsinthecontrolRVtobeorganizedinparallelbut thedisarrayedmyobrilsintheMCT-treatedRVlosethatorganizationalpatternand insomeareasshowzig-zagZ-lineirregularity.Theseobservationsarenotlocation specic.Figure3.8shows40xmagnicationimagesoftheRVtissuesbutFigure3.9 isofrepresentativeimagesat20xmagnicationshowingthepatternof -actininand Cx43incontrolandMCT-treatedratRVs. Figure3.8:Down-regulationofimmunodetectableCx43and -actininisobservedin ratrightventricleofMCTinducedRVfailurePHBcomparedtocontrolA.40x magnication Thequalitativendingsfor -actininareconsistentforlongitudinalsectionsof theRVlabeledwith -actinin. -actininlabelsthesarcomericZ-discsinthecardiacmyocyte.Mutationsinthe -actiningeneareassociatedwithcardiomyopathy[189, 190],butitsroleinpulmonaryhypertensionisyettobedetermined.Thereisn'tany datapresentlinkingCx43and -actininlocalizationpatternsinthemyocardium. 46

PAGE 63

Figure3.9:Representativeimageatalowermagnicationshowingdecreasedlevels ofandCx43and -actinininMCT-treatedPHBcomparedtocontrolA.20x magnication ObservingthecontrolandtheMCT-treatedRVsweseethatCx43iscontiguouswith -actininatthebordersofcardiacmyocyteswheretheIDsarepresent.Highlighted inFigure3.10,weobservedthattheareaoccupiedbybothCx43and -actinincould beproportionaltoeachother.Thechangeinmorphologyofcardiacmyocytesseenvia distinguishabledisarrayof -actininstaininginMCT-treatedratRVsisaccompanied bydecreased -actininstainingattheIDsandcolocalizingCx43atthediscs. 3.3ImmunouorescenceStaining:QuantitativeResults Figure3.11illustratesthestepstakeninordertoquantifyandanalyzetheIF images.Thedynamicrangeofthepixelswasadjustedtoremovebackgroundas wellasanyunnecessaryelements.Thisstepinthecodeisimplementedtolimit thedynamicrangeofan8-bitgrayscaleimagesothatdeductionsmadethrougha thresholdingalgorithmlaterinthequanticationstepisamelioratedwithminimum iterations.Anadditionalstepforimprovingthesignaltonoiseratiocanbeadded tothecustomMATLABcodeatthispointinimageprocessingbyremovingthe 47

PAGE 64

Figure3.10:Cx43and -actininseemtoshareareacoverageattheIDs.Decreasein -actininintheMCT-treatedratsB,isqualitativelyproportionatewiththeamount ofCx43presentatthediscs. variancecausedbypixelsinthetailendsoftheintensityrange.Forinstance,Figure 3.12showsanintensityhistogramforarawimage.Theredlinerepresentsathreshold valuethatcouldbeappliedtotheimagesuchthatanyintensityvalueabovethatline isnowconvertedtoamaximumintensityvalueof255asa8-bitgrayscaleimagecan have256pixelvalues;0beingminimumorblackand255beingmaximumintensityor white.Otsuthresholding,discussednext,isappliedandtheimageisthenbinarized toallowforautomaticcountingbyMATLABorbyImageJ. Otsuthresholdingmethod,establishedin1979,isbaseduponasimpleidea:nd athresholdvaluefromagray-levelhistogramthatminimizestheweightedwithinclassvariance[191].Inotherwords,thealgorithmchoosesathresholdvaluethat minimizestheintra-classvarianceofthethresholdedblackandwhitepixelsinan image.Thethresholdingmethodassumesthattheimageandthehistogramare bimodal,meaningthattheimageconsistsoftwoclassesofpixels:foregroundand background.Itthencalculatestheoptimumthresholdvalueseparatingthetwoclasses sothatthecombinedspreadisminimal.AnadvantageofusingtheOtsythresholding algorithmisthattheoptimalthresholdvalueiscomputedautomaticallyandstably, basedontheintegrationofthehistogramratherthanbasedondierentiation.In 48

PAGE 65

Figure3.11:Schematicrepresentationofquantitativeimageprocessing otherwords,theglobalpropertyisincorporatedratherthatalocalpropertyofthe histogram,forinstance,avalleybetweenthepeaksofanimagehistogram. Theacquisitionofimagesfollowedastochasticmannerwherethelocationofthe imageintheRVcannotbedeterminedprecisely,inotherwords,areasofinterest werepickedrandomly.Aminimumofthreeimagespersample,andthreesamples pergroupperimmunolabelmarkerwerethebasisforstainingandobtainingdata.It quicklybecameapparentthattocorrectlyanalyzethe -actininstains,thesection requiredtohavebeenimagedinthelongitudinalplane.Longitudinalsectionsofthe RVsshowcleardemarcatedZ-bandsofthesarcomere,andisatruerepresentationof the -actininstructureoverstainsseenincross-sectionalviewsreferbacktoFigure 2.2.Thepurposeofimmunouorescenceimagesweretonotonlylocalizetheprotein 49

PAGE 66

Figure3.12:IntensityhistogramfortherawimageshownearlierrefertoFigure 3.7A.Thegray-scalemaponthex-axisrepresentsthepixelintensityvaluewhere 0isblackand255iswhite. ofinterestinthemyocardiumbutalsostudytheorganizationalpatternsincontrol vs.diseasedstates.Wewereabletoextractreliableinformationonlyformthe longitudinalsectionswithrespectto -actininimmunouorescentstains.However, thisdicultydidnotfactorintoanalysisofCx43,dystrophinorothermarkers. Representativeimagesforthestepstakentoprocesseachimmunouorescentimage followingtheschematicFigure3.11areshowninFigure3.13.Table3.1consists ofvaluesforlengthin mdesignatedperpixelforrespectivemagnicationsused toacquiretheimages.Theanalyzedandquantiedindexesreturnvaluesofarea, length,etc.inpixels.Thetableprovidesconversionfactortoobtainthecharacteristic 50

PAGE 67

parametersinunitsof` m. Table3.1:ZeissLSM780:Pixelsizemeasurementreferenceforthedierentobjectives andzoomfactorsusedduringimageacquisition Objective ZoomfactorPixelcountDistanceconversionfacor[ m/pixel] 40x 2x10.104 40x 1x10.208 40x 0.6x10.346 20x 1x10.415 20x 0.6x10.692 a a Quantitativeanalysisperformedontheimmunouorescenceimagesareinsupport ofthequalitativedeductions.Figure3.15showsthescatterplotforthetotalarea coveredby -actininincontrolvs.theMCTRVfailuremodelofPH.Samplesizesfor controlandMCT-treatedmodelsweren=5andn=4respectively.Weare,fortherst time,presentingdatashowingthat -actininlevelisdecreasedintheMCT-treated ratRVsinthesettingofPH.Theresultfortheimmunouorescencequantication ispresentedinabox-whiskerandscatterplotmannerasthedatapointsareskewed. Theedgeoftheboxesrepresent1 st and3 rd quartilevalues,theredlinerepresents medianofthedataset,thewhiskersrepresentminimumandmaximumdatapoints, andlastly,thedatapointoutsideofthewhiskersrepresentsuspectedoutliers.The outliersarenotexcludedfromthedatasetsinceresultsfor -actininhaven'tbeen presentedinsuchamannerbefore.Additionally,robustquanticationmethodsand repetitionswouldberequiredtoverifytheoutliercharacteristicofthedatapoints. Thequantiedimmunodetectable -actininissignicantlydecreasedintheMCTtreatedratRVcomparedtothecontrolgroup.Thereisadecreaseinproteinthat 51

PAGE 68

aOriginalimage bModiedimage cBinarizedimage Figure3.13:Imagesrepresentingthestepstakenforimageprocessing.Cx43onthe leftcolumnandnucleiontherightcolumnwereprocessedsequentially.Scalebar=50 micrometer 52

PAGE 69

Figure3.14:RepresentativeimagefortheanalyzedimageusedtoprovidethequantitativeindexofCx43andnucleicount.Asimilarapproachwastakenforobtaining thepixelcountfor -actininanddystrophin. maintainsthestructuralintegrityofthemyocytecytoplasmandtheplasmamembrane. -actninisoformfoundinthemyocardiumiscalcium-insensitivewhichisan advantageasthemaintenanceofstructuralintegrityshouldnotbecalciumdependent inanorganwheretheuxischangingconstantlyallowingfortransductionofelectricalimpulses.Sincemembersofthespectrinfamilyareactinbindingproteins,the decreasein -actinincouldsuggestdegradationofthecontractileapparatusinthe myocardium.Whetherthedecreasedlevelsisduetoaproblemin -actininitselfor theproteinitbindstoisyettobedeterminedinPH.Wearealsoquantifyingprotein levelsatweek4postMCTadministration.Histologicalsectionsobtainedat2and 3weektimepointswouldbeneededtoobservedandanalyzedinordertodetermine thetrendof -actinindepletionintheMCTpreclinicalmodelofPH. 53

PAGE 70

Figure3.15:Box-whiskerandscatterplotfor -actininrepresentingthetotalpixel areanormalizedtocellcount.n=5control,;n=4MCT;p < 0.01. SimilarstepsweretakentoquantifytheCx43proteinlevelsincontrolsandMCTtreatedratgroups.Surelyenough,immunodetectableCx43wasfoundtobedecreased intheMCT-treatedratscomparedtothecontrolgroupp < 0.01.Thebox-whisker andscatterplotforconnexinplaquesexpressionispresentedinFigures3.16and3.17. WehadtwosubsetsofdataforCx43.Thiswasduetoutilizationoftwodierent clonesoftheantibodyallowingustoco-stainCx43withothermarkers.Cx43*seen throughouttheresultsisrepresentativeofdatasubsetthatusesadierentantibody comparedtodatashownusingCx43".Butregardlessofthecloneoftheantibody used,theresultisconsistent:decreasedlevelsofCx43intheMCT-treatedratRVs comparedtocontrols.Thevaluesarenormalizedbutthediscrepancyintherange betweenthetwoguresmayhaverisenfromtheuseoftwodierentantibodiesused toprobeforCx43.PleasereferbacktoTable2.2forinformationonthedierent subsetsofantibodiesused. 54

PAGE 71

Figure3.16:Box-whiskerscatterplotforCx43representingthetotalareanormalized tocellcount,p < 0.01. Figure3.17:Box-whiskerandscatterplotforCx43*correlatespositivelywithdata performedwithdierentcloneoftheantibodyp < 0.01. 55

PAGE 72

Asaforementioned,gapjunctionremodelingpresentswithfeaturessuchasdownregulationofconnexinsanddecreasedplaquesizesinmultipleheartdiseases.Through immunouorescencewefoundthatimmunodetectableCx43isdown-regulatedinPH rats.Weproceededtofurthercharacterizethegapjunctionremodelinginthesetting ofPHbyquantifyingtheplaquesizeinthecontrolsand4weekMCT-treatedrats. Figure3.18showsanimmunouorescentimageacquiredat80xmagnicationwhere theCx43plaquesarediscernibleandclearlydemarcated. Figure3.18:CleardemarcationofCx43plaquesisobservedintheimmunouorescent imageofRVat80xmagnication.Modiedimagesforquanticationextractionis providedinnextgure. WithadditionalmodicationinthecustomMATLABcodetoadjustforand excludeindistinguishableplaquesduetohighintensityproleatcertainareas,we 56

PAGE 73

extractedplaquesizesforCx43incontrolanddiseasedRVs.Theadditionalmodicationofthecodealsoexcludedsmallpixelsthatweren'tcontiguousinthethresholded image.Figure3.19showstherepresentativeimagesofimageprocessingforplaque sizequantication.Wequantied406and507individualCx43plaquesacrossmultipleimagestogetthelengthandarea,respectively,forthecontrolgroup.Similarly 295and386individualCx43plaqueswerequantiedtooutputlengthandarea,respectively,fortheMCT-treatedratRVs. a Figure3.19:AThresholdedimageforCx43wherebluepixelsandredpixelsare excludedtoprovideuswiththeimageinB.BlueandredpixelsinArepresent smallareasandclustersthatareindistinguishableasindividualplaques,respectively. a Scalebar:20 m WefoundthatCxplaquesizedecreasedsignicantlyp < 0.05fromcontrolsto 4weekMCT-treatedratRVs.ThelengthofCx43plaquesmeasuredorthogonalto myocyte'slongaxisreportedasmean SDwascomputedtobe0.744 0.0189 m forcontroland0.676 0.0162 m.Theareasoftheseplaqueswerecomputedtobe 1.470 0.0579 mand1.297 0.0558 mforcontrolandMCTgroupsrespectively. 57

PAGE 74

ThegraphicalrepresentationofthendingsispresentedinFigure3.20. Figure3.20:BarplotsummarizingplaquesizemeasurementsforCx43incontrolvs. MCT-treatedratRVs.p < 0.05 Intercellularcommunicationhasbeenshowntobedisruptedwhengapjunction plaquesizesarereduced[112,115,116].Theconductionvelocityisreportedtobe slowerandremodelingwithcharacterizationofdecreasedplaquesizeshavebeenlinked toarrhythmias.AdditionallyTan etal. haveshownthatdecreasedCx43plaquesize anddisturbedorganizationmaybeassociatedwithRVhypertrophy.Whetherthe lossofCx43plaquesanddecreaseinplaquesizeoccurrightafterMCTadministration isyettobedetermined.Thisvoidinthecurrentresearchcouldbelledwhendatafor the1and2weeksisacquiredforimmunouorescenceaswellaswesternblotprotein assay.Thus,atthistimewearen'tabletodeducewhetherthelossanddecreasein 58

PAGE 75

sizeofCx43plaquesisassociatedwithprogressionofMCT-inducedPH. Thequanticationofplaquesizeswasfurthercharacterizedtoseethedistribution ofsmallandlargeCx43plaques.HistogramoftheplaquesizesforcontrolandMCTtreatedgroupsareprovidedinFigure3.21.Inbothcontroland4weekMCT-treated ratgroups,thereisahigherfrequencyofsmallergapjunctionCx43plaques. Figure3.21:PlaquesizedistributiondistinguishingsmallandlargeCx43incontrol andMCTgroups. 59

PAGE 76

ToverifythattheMATLABcodeisreturningaccuratevaluesforplaquesize intermsoflengthandarea,uorescentbeadsofknowndimensionswerequantiedusingthecustomcode.Theuorescentbeadswereimagedusingthesame parametersusedtoacquireimmunouorescentimagesofourproteinsofinterest. FocalCheck TM FluorescentMicrospheresfromMolecularProbesof6 mdiameterwere used.Figure3.22Ashowstherawimageoftheuorescentbeadstakenat80xmagnicationand1024x1024resolution.Figure3.22Bshowstheanalyzedbinaryimage thatwasthenlabeledandthelengthwascomputed.Thecodereturnedameanvalue of56.894pixelsforthediameterofthebead.Theknowndistanceof1pixelatthese specicacquisitionparametersisknown:0.104 m.Thisgivesusthelengthofthe beadas5.915 m.Withonly1.417%dierenceinbetweenthecomputedandactual size,wefeelstronglyaboutthenumbersreportedforCx43plaquesizes. Figure3.22:ImageofuorescentbeadsARawimageacquiredusingthesameparametersusedtoimagetheCx43imagesthatoutputplaquesizes.BPost-processed binarizedimagethatwaslabeledandusedtoextractthelengthofthebeads. Additionally,imagesweretakenatmagnicationsof20xand40x.Thecustom MATLABcodereturnedaveragevaluesof5.959 mand5.920 mrespectively.With 60

PAGE 77

theminimumerrorsinsizemeasurementsfortheuorescentbeadswearecondentof thecustomcodegeneratedforthisstudy.Therepresentativeimagesfortherespective magnicationsareprovidedinAppendixB. ThendingsthatthereisreductioninimmunodetectableproteinlevelsinMCTtreatedratRVsraisesaquestionofwhetherthereductionisduetocelldeathinthe diseasedpreclinicalmodels.Toanswerthis,weproceededtoquantifytheprevalence ofcardiacmyocytesintheRVofboththecontrolandMCTgroups.The -actinin stain,althoughspecicforcardiacmyocytes,doesnotprovideadequateinformationthroughimmunouorescenetodiscerntheamountofcardiacmyocytesinthe myocardium.Therefore,weturnedtodystrophin,aperipheralmembraneprotein specictocardiacmyocytes. Figure3.23:Dystrophinco-stainedwithCx43incontroland4weekMCTratRVs. Dystrophinlabelstheperipheralregionofthecardiacmyocyteandisabsentatthe IDscontainingterminusofmyocytes.Highlightedregionsshowindividualcardiac myocytes. Figure3.23showsdystrophinandCx43co-stainingpresentedinthelongitudinal section.Thecardiacmyocytescannowbetracedintheseimmunouorescentimages 61

PAGE 78

asdystrophinlinesupthelateralmembraneofthemyocytesandCx43indicatethe terminusofthesecells.TheimageanalyzingMATLABcodeusedforquantifying -actininandCx43wasappliedto20controlsamplesand12MCTsamplesimagespersample.Therewasnosignicantdierencebetweentheimmunodetectable expressionlevelofdystrophinintheplasmamembranesofratcardiacmyocytein controlvsMCT-treatedmodels.Withthisnding,wesuggestthattheprevalence ofcardiacmyocyteisnotdierentbetweenthecontroland4weekMCT-treated ratRVs.However,thisdoesnotmeanthatthereisnoactivityrepresentingchange innumberofcardiacmyocyteduringtheprogressionofPH.Datageneratedfrom analyzingimmunouorescentimagesrepresentingtotalareanormalizedtonucleifor -actinin,Cx43,anddystrophinissummarizedinTable3.2followedbythegraphical representationshowninFigure3.24. Figure3.24:Barplotshowingthatquantitative -actininandCx43correlateswith qualitativedeductions;p < 0.05.Nosignicantchangeindystrophinbetweenthe controlandMCT-treatedgroups. 62

PAGE 79

Table3.2:Summaryofwesternblotndingsforcontrolanddiseasedgroups.Values presentedarerelativeastheyarenormalizedtocellcount. Marker ControlMCT -actinin Mean 83.21269.814 St.Dev 7.35214.785 SEM 2.3254.268 p-value 2.178e-02 Cx43 Mean 88.86874.383 St.Dev 18.27312.125 SEM 7.4604.950 p-value 1.019e-03 Cx43* Mean 95.89058.275 St.Dev 7.26011.415 SEM 4.1916.590 p-value 6.809e-03 a a St.Dev,standarddeviation;SEM,standarderrorofthemean Tofurthersupportoursuggestionthatthelossof -actininandCx43intheRV ofMCT-treatedratsisnotduetocelldeath,weassessedprevalenceofCASP3active cellsinthecourseofprogressionofPHandquantiedthendings.Figure3.25shows incontrol,andin1,2,and3weeksoftheMCT-treatedratRVs,thecellsparticipating inapoptosis.CASP3activecellsareinacascadepathwayandarecertaintofull apoptosis.Weobservethatthereisabundanceofcelldeathin1weekafterMCT administrationandtheapoptoticactivitystartstotaperoat2weeks.By3weeks, apoptosisisvirtuallyabsent.Withc-stainingofsmoothmuscleactinsmapresent, weareabletodeducethatvascularendothelialcellsareamongthecelltypesthatare CASP3active.ButmostofCAP3activecellsearlyonintheprogressionofPHare 63

PAGE 80

presentinthemyocardiumarrows,Figure3.25.Thesecouldbecardiacmyocyte, cardiacbroblasts,orothernon-myocytecelltypesintheRV.Withoutacell-type specicmarkerco-stainedwithCASP3,wearenotabletoascertainthecelltype undergoingapoptosisatthistime. Figure3.25:CelldeathisaprominentfeatureearlyonduringtheprogressionofPH asobservedthroughCASP3immunouorescencestaining.Cell-typesmostlygoing throughapoptosisareinthemyocardiumbutalsointhebloodvasculatureasshown bysmaco-localizationwithCASP3. OurcustomMATLABcodewasappliedtotheseimagestoquantifyapoptosis asPHprogressesaftertheadministrationofMCT.TheresultispresentedinFigure 64

PAGE 81

3.26below. Figure3.26:Barplotshowingactivecaspase3intheRVmyocardiumofcontroland 1,2,and3weekspostMCTtreatment. Itisinterestingtoobservethatthecelldeathismuchmorerobustintheearly stagesofPHbutislargelyabsentintheend-stagesofPHinrats.Additionally,we lookedatanalyzingprevalenceofnucleiincontrolsand4weekMCTtoseewhether thereisamajordierenceincellcountbetweenthegroups.Wefoundnosignicant changesinthenumberofnucleipresentwhichfurtherconrmsoutdataandsuggestionthatthedepletioninproteinsofinterestisnotoccurringduetocelldeath. Thelossof -actininandCx43isimportantinthesettingofPHandthefactthat itisoccurringindependentofcelldeathsupportsourhypothesisthatorganizational andexpressionalchangesoftheseproteinsarepromotingintercellulardisturbances, changesincontractilefunctionleadingtoedemaandRVfailure.Thenucleicount analysisissummarizedinFigure3.27.Theexperimentalsetscompriseofallofthe acquiredimagesseparatedproperlyintorespectivecontrolandMCT-treatedgroups 65

PAGE 82

butinarandomizedmanner.Inotherwords,theproteinofinterestdidn'tplaya roleinthisanalysissincewewereonlylookingatDAPIchannel. aImagingset1 bImagingset2 cImagingset3 dSummaryofnucleicountperstainingset Figure3.27:Nucleicountedformultiplestainingsetsshownosignicantdecreasein totalnumberofcellsfromcontroltoMCT-treatedmodels. Whenexaminingthedystrophinimagesinthecross-sectionalview,asshownin Figure3.28,itwasapparentthattherewerediscrepancyincardiacmyocytecellsize betweenthecontrolandtheMCT-treatedratRVs.Thelengthofcardiacmyocytes 66

PAGE 83

weremeasuredusinglongitudinalsectionsFigure3.23.Thewidthwascomputed usingcoss-sectionalsectionsasseeninthegurebelow. Figure3.28:Dystrophinco-stainedwithCx43incontroland4weekMCTratRVs. Dystrophinlabelstheperipheralregionofthecardiacmyocyteandisabsentatthe IDscontainingterminusofmyocytes.Highlightedregionsshowindividualcardiac myocytes. ThelengthwasmeasuredfromtheIDsateachoftheterminaledgesofthe myocytecellslabeledbyCx43.Weprocessedtheimageandcomputedwidthby measuringthedistancealongtheaxisparalleltotheIDsshownbyCx43stainsin redandalongtheaxisperpendiculartotheIDs.Wefoundnosignicantdierence indoingthemeasurementtwowayshence,thewidthmeasurementsrepresentthe diameterofthecardiacmyocytecells.Forn=43individualcardiacmyocytefor controlsandn=33forMCTwefoundnosignicantchangesinthelengthofthecells. Forn=46controlandn37MCT-treatedcardiacmyocyteswefoundasignicance increaseincellwidthfrom17.111 3.862 mto28.195 7.616 m.Thelengthstays relativelyunchangedbuttheincreaseinwidthsuggestsincreaseincardiacmyocyte cellvolume.Withthepresenceofedemainthemyocardiumthecellsmaybetaking 67

PAGE 84

upuideitherbypassiveoractivetransport.Additionalexperimentswouldbe neededtoverifywhetherthewaterchannelsaredisturbedandwhetherthecells takingupvolumeiscompensatoryornotdependingonthetypeoftransportpassive vs.active.Sincethereisnotchangeindystrophinexpression,asshownearlier,the plasmamembrane'sintegritymaybeunchanged.Thebargraphshowinglengthand widthanalysisisprovidedinFigure3.29. Figure3.29:SummaryoflengthandwidthanalysisperformedoncontrolandMCTtreatedcardiacmyoctesoftheRV.p < 0.01 3.4WesternBlot Asoneoftheassaystoverifythendingspresentedintheprevioussubsection, welookedtowesternblotproteinassaytovalidatethelossofproteinexpressionin theMCT-treatedratRVsinPH.QualitativeanalysisontheIFimagesduringearly stagesofthestudyledtodeductionsthatimmunodetectableCx43and -actinin weredown-regulatedintheMCTRVfailuremodelofPH.Thequantitativeanalysis 68

PAGE 85

performedontheimagesfurthersupportedthosedeductions.NinesamplesofRV homogenateswereloadedintoa10-wellgel,andcomprisedoftriplicatesforcontrol,4 weekMCT,and4weekSu5416-hypoxiagroups.The10 th well,loadedwithaladder, isusedforsizereferencekDa. Beforetheimmunoblotswerequantied,qualitativelyassessingthe -actininIF images,itwasveryapparentthattheprotein'sexpressionisdown-regulatedinthe MCT-treatedgroup.ThewesternblotresultspresentedinFigure3.30positively correlatewiththequalitativendingsfromtheimmunouorescencestainingimages. Theextrabandsseenatthe98and62kDalevelswerebackground.Anegativecontrol assayforthewesternblotwasperformedusingjustthesecondaryantibody:goat-antimouseconjugatedwithHRPEMDMillipore.Itwasobservedthatthesecondary antibodywouldbindtoproteinsaround98and62kDaregionsasseeninFigure3.30 withthesamepatter.Sincetheproteinofinterest, -actinin,liesinthe107kDa, thebackgroundbandsproducedbythesecondaryantibodydonotcompromisethe integrityoftheassay,andthebands,therefore,areignored.Quanticationofthe westernblotshowthatthedecreasein -actininexpressionishighlysignicantin boththediseasedmodelsofRVfailurePHwhencomparedtocontrolgroups.Figure 3.31demonstratesthequantiedresultsnormalizedbyGAPDH.Two-samplet-test onthesetsofcontrolandMCTgroupsandsetsofcontrolandSu5416-hypoxiawere performedwith of5%. Thereisvariancewithinthetriplicatesofthesamegroups.Thus,thereispopulationheterogeneitypresentfortheamountofproteinsbeingexpressedamongthe populationwithintherespectivediseasedgroups.Thisistrueforhumansaswell,not everyoneexpressesthesameproteinatexactlythesamelevels.Thedatapresented isastrongrepresentationoftheproteinsbeingexpressedamongthetriplicatesasthe loadingcontrolinFigure3.30iseventhroughouteachofthe10-wellplates. 69

PAGE 86

a Figure3.30:Westernblotfor -ActininleftperformedforratRVhomogenates. GAPDHrightusedfornormalizationpurposes.Signicantdown-regulationof actinininproteinlevelsofMCTandSu5416-hypoxiagroupsareobserved. a Nx,Normoxic;MCT,Monocrotaline;SuHx,Sugen5416-hypoxia;GAPDH,Glyceraldehyde3phosphatedehydrogenase Figure3.31:QuanticationofWesternblotfor -actininshows63%decreasefor MCTand83%decreaseforSu5416-hypoxiamodels. 70

PAGE 87

Figure3.32showsdecreasedexpressionofCx43proteinlevelsintheMCT-treated modelofPHbutthereisnosignicantdierenceseenintheSu5416-hypoxiagroup incomparisontothecontrol.TheproteinlevelsforMCTgroupisdecreasedby64% incontrasttocontrol.Multiplebandsseenaround43kDacouldbethedierent isoformofCx-43beingpickedupbytheantibodyusedFigure1.7.Thebandacross 98kDamightsuggestthatCx-43arepresentasdimers.Phosphataseinhibitorswere administeredduringthetissuehomogenizationprocessbutthestatusofproteins beforetheinhibitoradministrationisnotcontrolled.ThereisapossibilityofCx43 undergoingphosphorylationpriortoorevenduringthetimeorganswereharvested. Cx43beingabletoformaheteromericbondwithotherconnexins,mainlyCx45in theheart,couldalsosuggestthebandat98kDa.Quantifyingthewesternblotreveal thatCx43issignicantlydecreasedintheMCTmodel.ThevariancefortheSu5416hypoxiamodelisgreaterthanfortheMCTgroup.Duetopopulationheterogeneity, thehighvarianceseeninSuHxcouldbeduetodierentratsexpressingproteinat dierentlevels. Figure3.32:DownregulationofCx43inproteinlevelsisobservedfortheMCTRV failuremodelbutnotinSu5416-hypoxiaPHrats.QuanticationofWesternblotfor Cx43shows64%decreaseforMCTinducedRVfailurePHmodel.p < 0.01 71

PAGE 88

NegativecorrelationbetweentheexpressionofCx43amongthetwodiseased RVfailuremodelsofPHcouldbeattributedtothedierentmechanismofdistinct preclinicalmodels.Additionally,Cx43isshowntohavearesponsetohypoxiawhere thereisincreasedexpressioninratcarotidbody[195,194].Wefocusedmostlyon MCTmodelforthisproject.However,obtainingdataforSuHxwouldproveusefulin understandingCx43expressionacrossmultiplediseasedmodelsofPH.Additionally, wecouldverifythehypoxiaresponseviaachronichypoxiamodelforPH. Table3.3:Summaryofquantiedwesternimmunoblotproteinassay. Marker ControlMCTSuHx -actinin Mean 10.3760.165 St.Dev 0.1180.1740.151 SEM 0.0680.1010.087 p-value 0.007190.00145 Cx43 Mean 10.3611.068 St.Dev 0.2020.1270.541 SEM 0.1160.0730.312 p-value 0.00204 0.867 a a St.Dev,standarddeviation;SEM,standarderrorofthemean.Thep-valuespresentedarewith respecttocontrolgroups.ThereisnosignicantdierencebetweentheMCT-andSuHx-treated groupsforeither -actininorCx43. RecognizingthatCx43expressionhasanegativecorrelationamongdierentmodelsthatleadtothesameendpoint,RVfailure,isinteresting.WewentbackandacquiredSuHximmnouorescentimagestovalidateourwesternndings.Figure3.33 showsCx43immunouorescencestainforSuHx-treatedratRV.Threesetsofsamples withthreeimagespersamplewereacquiredtoproducethedatashownintheboxwhiskerandscatterplotpresentedinFigure3.34.Therewasnosignicantchangein 72

PAGE 89

immunodetectableCx43expressionbetweencontrolsandSuHx-treatedratRVs. Figure3.33:Cx43inSuHxRVfailuremodel.Immunouorescenceimageacquiredto verifywesternimmunoblotassay.ArrowsshowmislocalizationofCx43. Figure3.34:Box-whiskerandscatterplotshowingCx43proteinlevelsincontrolvs. MCT-andSuHx-treatedPHrats. 73

PAGE 90

3.5AssessmentofFibrosis Webelievethatedemainthemyocardiumpreceedsbrosis.Asmentionedin Section3.1,ourlabgrouphasbeensuccessfulinpositivelycorrelatingthepresence ofEVFwithelevatedcollagendepositionintheRV.Withimmunouorescencewe wantedtoassessthechangeinnumberofbroblaststhateitherresideinthemyocardiumormigratethereafterintroductionofMCTtofurtherstrengthenthecorrelationandbetterourunderstandingabouttheroleofbroblasts,theircross-talkwith cardiacmyocytesandthesurroundingECM.Butitiswellestablishedthatamarker specicforbroblasthasbeendiculttonaildown.Thisremainsoneofthemain reasonswhyourunderstandingofbroblastinseverelylacking.WepursuedTCF21, FSPandP4Htotryanddistinguishthecardiacmyocytesfromthebroblasts.We hadsuccessfulstainingforallthreemarkersbutthepitfallthatlimitsourdatais thatthesemarkersdonotco-localizeforthesamecelltype.Thisisduetopresence ofdierentpopulationsofbroblastsinthemyocardium[88,131,138]andinpart duetoFSPalsolabelinginammatorysubpopulationofmacrophages[136]. Figure3.35showseectiveimmunolabelingofcardiacmyocyte'splasmamembranewithdystrophinandbroblastswithP4H.Thisco-stainingcombinationcould beusedindetermininginformationaboutbroblastswithrespecttothefollowing aspects:1theirprevalenceincontrolvs.theirindiseasedstatecorrelatingachange inexpressionduetoinjuryandprogressionofdisease,and2theirlocationwithrespecttodiseasedmodels;weknowthattheirdistributionisnon-uniformthroughout theventricle[96].Utilizingthistechniquewillallowustovisuallyplacebroblasts intheregionsoftheRV.Asaresult,wecouldpotentiallydeterminethelocationsof theRVthatisseverelyinjuredcomparedtootherlocations.Forinstance,assessing preliminarypentachromestainsoftheRV,asshowninFigure ?? wewereabletosee elevateddepositionofcollagenyellowstainsintheRVfreewallcomparedtothe myocardium.Combiningthepentachromestain,assessingthecollagendeposition,to 74

PAGE 91

Figure3.35:Co-immunostainingofcardiacmyocytemarkerdystrophingreenand cardiacbroblastmarkerP4Hpresentedinamontage. theimmunouorescenceimages,assessingthelocationofbroblasts,couldleadusto narrowdowntheregionsoftheRVmostaectedbyPH. 3.6LimitationsandRecommendations Theabilitytoassessbrosisusingimmunouorescenceishinderedbymissing specicmarkersforbroblasts.Discoidindomain-containingreceptor2DDR2, alsoknownasCD167bisaproteinthatisassociatedwithbrosisandcancer[95,94]. DDR2isalsoknowntoregulatebroblastproliferationandmigrationthroughthe ECM.ItisrecommendedthattheuseofDDR2beincorporatedintothemixofthe P4HandTCF21totargetbroblaststhroughthehypertensiveheart.ImmunouorescencelabelingofCx43hasprovidedvaluableinformationinthesettingofmultiple heartdiseases.Studieshavemadeheadwayaboutinteractionbetweenconnexinsand otheradhesionjunctionmolecules.IFimagingisaresourcefultoolthatallowsus 75

PAGE 92

Figure3.36:CollagendepositionintheMCT-treatedratsisindicativeofbrosis. Additionally,theblue-greenstainseeninthediseasedimagerepresentsimmature collagen. tovisualizeandtosomeextentanalyzethecolocaliztionofthesejunctionalproteinsattheIDs.Tofullyunderstandtheinteractionofthesediscretestructures,it isrecommendedthatwegobeyondtheuseofimmunouorescence.Thistechnique doesn'thavetheresolutiontotelluswhethertheapparentoverlapofuorescence signalsaretruerepresentationofinteractionandassociationatthemolecularlevel. Wecould,howevertakeconfocalmicroscopytoreconstructthemyocardialtissuein 3-dimensional.Figure3.37showsanattemptatareconstructionofIDsintheRV. Figure3.37hasadvantagesover2Dimages.Witha3-Dreconstructionwewould beabletoquantifyCx43plaquesizewithincreasedaccuracy.Additionally,wewould beabletocharacterizethedistributionandorganizationofgapjunctionCx43atthe IDswithanaddedviewofthetransverseplane.Wecouldalsoinvestigatehowthe organizationchangeswithintheIDs.Therearehoweverlimitationsthatpreventus fromacquiringsuchimages.Therstofthemisthehearttissue.Itiscomplexand ishighlymuscularizedwhichmakesitdiculttoeectivelystainthickersections. Optimizationoftheastainingprotocolwouldberequiredthatdonotjeopardize theintegrityofthetissuestructuresothatwearequantifyingourdatacorrectly. 76

PAGE 93

Figure3.37:3-DimensionalreconstructionofaratRVfocusedonCx43attheIDs. 80xmagnication;reconstructedusingImageJ Failuretooptimizeparameterstoacquireimagesisadisadvantage.Thisisoneofthe limitationweranintoduringourstudy.ForimagesusedtoquantifyCx43plaques, suchasFigure3.20,thereweresomeregionsthathadtobeexcludedfromthedata poolbecausetheintensityprolewastoohighttogetadiscernableCx43plaques. Onewaytoovercomethisproblemistoeitherturndownthepowerorthedigitalgain sothattheintensityproleisadequatelymeasurethroughouttheimagingplane. 77

PAGE 94

4.Discussion Theunderlyingcellularmechanismsgoverningthepathogenesisofpulmonary hypertensionstillremaintobefullyexploredandelucidated.Acontributingfactor couldbeduetothepathogenesisofPHdieringdependingonthetypeofetiology. Forthisreason,ourlabgroupfocusedonstudiesindierentpreclinicalmodelsofPH: hypoxia,monocrotalineandsugen5416-hypoxia.Weconductedourresearchmostlyin theMCTmodelofRVfailurePHduetothetime-frameandthescopeofthisproject beingatwo-yearMaster'sthesis.Additionally,thereisnomechanisticdatapresent forgapjunctionremodelingintheRVinthecontextofRVfailurePH.Therefore, alargepartoftheprojectreectedonlearningaboutgapjunctionremodelingin variousotherdiseasedsettingsandorgans,andtrytoincorporatethosendingsto ourpreclinicalmodelswhethertheyoverlappedornot. Emergingstudiessuggestthatconnexinsmayregulateinammatorysignals,such asinltrationofleukocytesandadhesionmoleculeexpression[192].Wedon'thave evidenceofthisfortheprojectsincewedidn'tlookintoleukocytedistributionin theRVmyocardium.AnotherexampleisastudythatlinkedCx43depletiontoreducednumbersofinammatorycellsandatheroscleroticlesions[193].Atthispoint inthestudy,wecanonlyspeculatethattheinammationandbrosisseeninthe RVcorrelatestothedramaticdepletionofCx43andtheirdisorganization.Thisis largelyinduetooutdatareectingtimepointsatcontroland4weekpostMCTadministration.Additionaldatapointswithinammatorymarkersandatshorter timeintervalsasPHprogresseswouldallowustoextractmoreinformationwithrespecttobrosis,inammationandofcourseCx43and -actinin.Asoutlinedinthe FutureDirections,Section4.2,thefunctionalconsequencesofalteredCx43expression needtobeinvestigatedwithrobustexperimentsthatanswerspecicquestionssuch as:Whatothercomponents/networksinteractwithCx43infailingheart?,whatare thenatureandmechanismsofgapjunctionremodeling?,howdoestheexpressionin 78

PAGE 95

RVdierfromLV?lungs?,andmostimportantly,howdoesthisconnectustohuman heartfailure?Thesearebutfewquestionsthatremaintobeaddressed.Inaddition toincludingmoretime-pointsinourdataacquisition,wecouldanswertheaforementionedquestionsbyinvestigatingtheIDsatamolecularlevelviaelectronmicroscopy, checkforandunderstandtheinteractionsbetweengapjunctionsandotherjunctional complexeswithintheIDsnexus,andfurthercharacterizegapjunctionremodeling. Theperturbedlocalizationanddown-regulationofconnexinsandotherjunctional complexeslikeclaudinsatightjunctiongiverisetoavarietyofdevelopmentaldefects.Arrhythmiasareoneofthewell-establishedfunctionalconsequenceofCx43 disturbance,butthesedefectsarejustnotlimitedtotheelectricalcouplinginheart. Theyextendtodeafness,skindiseasesanddemyelination[110].Whatcouldbethe causeofthereductioninnumbersofCx43?InteractionwithotherjunctionalcomplexesandmutationsinthemcouldbeacauseofdownregulationinCx43expression intheRV.TheturnoverofCx43israpid.5to2hours/citeBeardslee1998,thus, mutationsinCx43duetoendoplasmicreticulumstresscouldleadtoadecreasein expressionaswellasplaquesizereduction.TheorganizationoftheCx43isknown tobeattheintercalateddiscs.However,someclaimsoflateralimmunostaininghave beenreported.Thiscanbeexplicatedthroughtheorganizationoftheintercalated discsandinteractionofgapjunctionswithotherjunctionalcomplexeslocatedatthe IDs.ForaslongasIDshavebeenknowntheinformationacquiredthroughresearch hasnotbeenlinearwithrespecttotime.StudiesofIDsindiseasedstatescould likelyaddtoourknowledgeaboutitsformationand'cross-talk'amongitsvarious components. SomecouldarguethatthedepletionofCx43and -actininseeninthediseased modelsisduetocelldeath.Thus,suggestingthattheaccompanyingdecreasein numbersoftargetedproteinsinRVfailuremodelofPHpositivelycorrelatestofewer cellspresentpost4weekMCTadministration.Butwehaveshownthatdecreased 79

PAGE 96

Cx43expressioncouldriseindependentofcelldeath.Itisinterestingtonotethat theRVofthePHratsat3and4weeksrecoverfromrobustcelldeathandapoptosis inthe1stcoupleofweekspostMCT-administrationpleasereferbacktoFigures 3.25,3.26and3.27.Duringtheearlytime-pointswhenthereisanabundanceof celldeathasshownbyheavyactiveCASP3staining,wespeculatethatCx43and -actininstarttodropoutimmediately.Imageacquisitionsandwesternassayat earlyweekspostMCT-administrationwouldbeusedtovalidatethespeculations. Butfurtherinvestigation,astowhytheCx43and -actininlossdon'trecoverat4 weekpostMCT-administration,wouldberequired. WeobservednosignicantchangeinCx43intheSugen5416-hypoxicmodelinour study.Asaforementioned,thisresultcouldbeattributedtothedierentmodelsof PH.Therearestudiesthatshowup-regulationofCx43inhypoxicmodels[194,195]. LooselyfollowsthedataseenintheWesternblotperformedforSu5416-hypoxic.The varianceisgreaterandthemeanisslightlyhigher,albeitinsignicantly,thancontrol. 4.1StrengthsandWeaknesses:TheTalesfromourDataandtheEects onoutHypothesis Thestrengthinthisstudyisthepositivecorrelationseenbetweenmultipleassays anddierentquantication/analysisdoneonthoseassaysrevealingthesameresults. ThequalitativeandthequantitativeanalysisontheIFimagesandthequantitative resultsfromWesternblotoutputsimilarndings.Wehypothesizedthatdisturbances inorganizationalandexpressionalchangesingapjunctioncontributestoincreased myocardialedemaanddecreasedRVfunction.Thespecicaimwasdirectedtowards characterizationofCx43intheRVofirreversiblePHmodels.Wehaveshownthat thereisalossofimmunodetectableCx43attheIDslocatedattheterminusofcardiac myocytes.Thesearesituatedperpendiculartothelongaxisofthemyocytes.Comparedtoanarrayedorganizationforthecontrols,theMCT-treatedratRVsshowed lossofCx43plaques.Wefurthercharacterizedplaquesizesdierencesincontrols 80

PAGE 97

vs.diseasedmodelandfoundthatthereisasignicantdecreaseintheindividual plaquesizesofCx43.Furtherinvestigationisrequired,however,sinceCx43isable toformhomomericbondswithotherCx43junctionalcomplex.Weseemitnecessary toverifywhetherthelargerplaquesizesrepresentasingleCx43plaqueorwhether theyaraggregatesofmultipleCx43plaques.Additionaldatathatwouldsupport thehypothesiswouldbetoacquireelectriccouplingdata.Arrhythmiasarecapable ofcausingheartfailureandsincedisturbancesinCx43hasbeenlinkedtodecreased conductionvelocityandarrhythmogeniceventsinmultiplecardiomyopathies,itisimportanttoassesselectricandconductiveapparatusofthemyocardiuminthesetting ofPH. Oneofthemajordrawbacksofstudyingadisorderassociatedwithfailinghearts andlungsistheinabilityorthedicultytoobtainhumantissue.Thus,depending onthetypeoffundingandpartoftheworldtheresearchisbeingconducted,weare limitedtopreclinicalanimalmodels.Asmentionedthroughoutthethesis,celltypes arespeciesdependentandthuscorrelatingthendingsandapplyingtheresultsto thehumanstakestimeandtheoutcomesarenotalwaysfavorable.Performingassays, similartothoseconductedinthisstudy,onhumantissuewouldbetransformative andpivotalifndingspositivelycorrelatebetweenspecies. Apitfallencounteredduringstainingofthickersectionswastheprecipitation xationmethodusedpriortoIFstaining.Itturnsoutthatthismethodisnotvery usefulwhenusingthicksectionsforconfocalmicroscopy.AsmentionedinChapter2,a 1:1ratioofmethanol:acetonewasused.Although,thisxationtypemayhavehelped inpermeabilizingthetissueforeectivepenetrationandlabelingoftheantibodies,it didnothelppreservethe3-dimensionalorganizationforourspecimen.Information gatheredcouldleadtomisconceptions. 81

PAGE 98

4.2FutureDirections Therstandforemostfollowupwillincludeverifyingwhetherthendingsfrom ourpreclinicalmodelsregardingtheexpressionofCx43and -actinincorrelatewith humanpatientswithPH.OurexpertcollaboratorsinFrancewhostudyPHinhuman tissuewillaidinmakingtheconnectionbetweenourpreclinicalmodelsandhuman tissue. ThendingsofthisstudyverymuchsetupratherheavyquestionstowardsunderstandingthepathogenesisofPH.Intermsofconnexinsandgapjunctionalproteins'roleinvascularpermeabilityand/orlymphedema,astudycouldbeconcocted wherebyapermeabilityresponseisassessedinaninvitromodeltoinammatory stimulifollowedbyinhibitionofCx43function.Willthisstudythenshowthatthere isdown-regulationofCx43toaninammatoryresponse.Thiswillalsoanswerthe questionofwhetherCx43down-regulationiscausaltotheprogressionofPHoris itaconsequence.ThefollowupwillbetofurtherinvestigatethechangesinCx43 and -actininasthediseaseprogresses.ThenmovingontoSuHxtoverifywhether thechangesareuniformthroughoutmultiplemodelsofthesamediseaseornot.This willaidintargetedtherapeuticmeasuresthatwouldamelioratepathophysiologyof PH.Forinstance,wefoundthatinbothMCT-andSuHx-treatedrats, -actinin expressionisdecreased.Afuturestudymustbeincorporatedtosupportandfurther investigatetheexpressionalchangesin -actinin.Verylittleinknownabouttheproteinexpressedabundantlyinthesarcomericz-discsandresearchcateredtoguring outthefunctionalconsequencescouldprovidenovelavenuesofassessment,prognosis andtreatmentofpulmonaryhypertension. 82

PAGE 99

4.3ConcludingRemarks Itisrmlyestablishedthatgapjunctionalremodelingandconnexinexpressionis aconsistentcharacteristicobservedandrecordedindiseasedventricles.Thisisevidencedthroughcomparativestudiesconductedonvariouspreclinicalmodels.Therefore,gapjunctionsandconnexinareincreasingconsideredastherapeutictargetsto potentiallyassessandtreatpulmonaryhypertension.Withmorestudiesfocusingon theelectricalcouplingfunctionalityofgapjunctions,wecouldbenetfromstudies ontheseproteincomplexesfromamechanicalpointofview.Ontheotherhand,a newproteinofinterestmightbeofimportanceinunderstandingtheprogressionof PHtorightheartfailureandcouldprovetobeavaluablemarkerinhelpingusunderstandtheunknownmechanismsofthedisease. -actininisdramaticallydepleted inRVfailuremodelsofPH.SeeingashowthecontractilemyocytesoftheRVare dealingwithdepletionofamajorcomponentrequiredforproperfunction,itcould leadtotransformativenoveltherapies.TheprognosisassociatedwithpericardialeffusionandedemaintheRVisverypooranddespitethisnegativeimpact,verylittle isknownaboutteetiology,consequencesandmanagementofedemainPH.With ourndings,weanticipatefuturestudiestoincorporateinvestigationsthatreveal practicalpotentialtherapeutictargetsineectiveassessmentandtreatmentofPH. 83

PAGE 100

REFERENCES [1]GeraldSimonneau,MichaelA.Gatzoulis,IanAdatia,DavidCelermajer,Chris Denton,ArdeschirGhofrani,MiguelAngelGomezSanchez,R.KrishnaKumar, MichaelLandzberg,RobertoF.Machado,HorstOlschewski,IvanM.Robbins, andRogieroSouza.Updatedclinicalclassicationofpulmonaryhypertension. In JournaloftheAmericanCollegeofCardiology ,volume62,2013. [2]RubinM.Tuder,CarlyneD.Cool,MichaelYeager,LaimuteTarasevicieneStewart,ToddM.Bull,andNorbertF.Voelkel.Thepathobiologyofpulmonary hypertension:Endothelium. ClinicsinChestMedicine ,22:405{418,2001. [3]RKessler,MFaller,EWeitzenblum,aChaouat,aAykut,aDucolone, MEhrhart,andMOswald-Mammosser."Naturalhistory"ofpulmonaryhypertensioninaseriesof131patientswithchronicobstructivelungdisease. Americanjournalofrespiratoryandcriticalcaremedicine ,164:219{24,2001. [4]CharlieStrangeandKristinBHighland.Pulmonaryhypertensionininterstitial lungdisease. Currentopinioninpulmonarymedicine ,11:452{5,2005. [5]UMehlhorn,HJGeissler,GALaine,andSJAllen.Myocardialuidbalance. EurJCardiothoracSurg ,20:1220{1230,2001. [6]StefanZahler,AnkeHomann,TorstenGloe,andUlrichPohl.Gap-junctional couplingbetweenneutrophilsandendothelialcells:anovelmodulatorof transendothelialmigration. Journalofleukocytebiology ,73:118{126,2003. [7]VallerieVMcLaughlin,StephenLArcher,DavidBBadesch,RobynJBarst, HarrisonWFarber,JonathanRLindner,MichaelAMathier,MichaelDMcGoon,MyungHPark,RobertSRosenson,LewisJRubin,VictorFTapson, andJohnVarga.ACCF/AHA2009expertconsensusdocumentonpulmonary hypertensionareportoftheAmericanCollegeofCardiologyFoundationTask ForceonExpertConsensusDocumentsandtheAmericanHeartAssociationdevelopedincollaborationwiththeAmericanCollegeof. JournaloftheAmerican CollegeofCardiology ,53:1573{619,apr2009. [8]RubaNHaddadandLisaMMielniczuk.Anevidence-basedapproachtoscreeninganddiagnosisofpulmonaryhypertension. TheCanadianjournalofcardiology ,31:382{90,apr2015. [9]LarissaAShimodaandStevenSLaurie.Vascularremodelinginpulmonary hypertension. JournalofMolecularMedicine ,91:297{309,2013. 84

PAGE 101

[10]JaumeAguero,KiyotakeIshikawa,LahouariaHadri,CarlosSantos-Gallego, KennethFish,NadjibHammoudi,AntoineChaanine,SamanthaTorquato, CharbelNaim,BorjaIbanez,DanielPereda,AnaGarca-Alvarez,Valentin Fuster,ParthoPSengupta,JaneaLeopold,andRogerJHajjar.Characterizationofrightventricularremodelingandfailureinachronicpulmonary hypertensionmodel. Americanjournalofphysiology.Heartandcirculatory physiology ,307:H1204{15,2014. [11]KurtRStenmark,MichaelEYeager,KarimCElKasmi,EvaNozik-Grayck, EvgeniaVGerasimovskaya,MinLi,SuzetteRRiddle,andMariaGFrid.The adventitia:essentialregulatorofvascularwallstructureandfunction. Annual reviewofphysiology ,75:23,2013. [12]SSartore,AChiavegato,EFaggin,RFranch,MPuato,SAusoni,and PPauletto.Contributionofadventitialbroblaststoneointimaformationand vascularremodeling:frominnocentbystandertoactiveparticipant. Circulation research ,89:1111{1121,2001. [13]KurtRStenmark,NeilDavie,MariaFrid,EvgeniaGerasimovskaya,andMita Das.Roleoftheadventitiainpulmonaryvascularremodeling. Physiology Bethesda,Md. ,21:134{145,2006. [14]NFVoelkelandRMTuder.Hypoxia-inducedpulmonaryvascularremodeling:amodelforwhathumandisease? TheJournalofclinicalinvestigation 106:733{8,sep2000. [15]C.Atkinson.PrimaryPulmonaryHypertensionIsAssociatedWithReduced PulmonaryVascularExpressionofTypeIIBoneMorphogeneticProteinReceptor,2002. [16]SPGaineandLJRubin.Primarypulmonaryhypertension. Lancet 352:719{725,1998. [17]MariusMHoeper,NazzarenoGalie,GeraldSimonneau,andLewisJRubin. Newtreatmentsforpulmonaryarterialhypertension. Americanjournalofrespiratoryandcriticalcaremedicine ,165:1209{16,may2002. [18]RachelJ.DaviesandNicholasW.Morrell.Molecularmechanismsofpulmonary arterialhypertension:Roleofmutationsinthebonemorphogeneticproteintype IIreceptor,2008. [19]RDMachado,MWPauciulo,NFretwell,CVeal,JRThomson,CVilari~no Guell,MAldred,CABrannon,RCTrembath,andWCNichols.Aphysical andtranscriptmapbaseduponrenementofthecriticalintervalforPPH1, ageneforfamilialprimarypulmonaryhypertension.TheInternationalPPH Consortium. Genomics ,68:220{8,sep2000. [20]RubinM.Tuder.Pathologyofpulmonaryarterialhypertension,2009. 85

PAGE 102

[21]JohnCMcGrath,ClareDeighan,AnaMBriones,MajidMalekzadehShafaroudi,MelissaMcBride,JeremyAdler,SilviaMArribas,ElisabetVila,and CraigJDaly.Newaspectsofvascularremodelling:theinvolvementofall vascularcelltypes. Experimentalphysiology ,90:469{475,2005. [22]JamesK.Belknap,E.ChristopherOrton,BeverlyEnsley,AlanTucker,and KurtR.Stenmark.HypoxiaIncreasesBromodeoxyuridineLabelingIndicesin BovineNeonatalPulmonaryArteries. AmericanJournalofRespiratoryCell andMolecularBiology ,16:366{371,1997. [23]BOMeyrickandLMReid.Crotalaria-inducedpulmonaryhypertension.Uptakeof3H-thymidinebythecellsofthepulmonarycirculationandalveolar walls. TheAmericanjournalofpathology ,106:84{94,1982. [24]AnnaEnzerinkandAVaheri.Fibroblastactivationinvascularinammation. JournalofThrombosisandHaemostasis ,9:619{626,2011. [25]RSSmith,TJSmith,TMBlieden,andRPPhipps.Fibroblastsassentinel cells.Synthesisofchemokinesandregulationofinammation. TheAmerican journalofpathology ,151:317{322,1997. [26]MatteoCoen,GiulioGabbiani,andMarie-LuceBochaton-Piallat. Myobroblast-MediatedAdventitialRemodelingAnUnderestimatedPlayer inArterialPathology. Arteriosclerosis,thrombosis,andvascularbiology 31:2391{2396,2011. [27]C.TjiJoongGan,JanWillemLankhaar,NicoWesterhof,J.TimMarcus,AnnemarieBecker,JosWRTwisk,AncoBoonstra,PieterE.Postmus,andAnton Vonk-Noordegraaf.Noninvasivelyassessedpulmonaryarterystinesspredicts mortalityinpulmonaryarterialhypertension. Chest ,132:1906{1912,2007. [28]SrijoyMahapatra,RickA.Nishimura,PaulSorajja,StephenCha,and MichaelD.McGoon.Relationshipofpulmonaryarterialcapacitanceandmortalityinidiopathicpulmonaryarterialhypertension. JournaloftheAmerican CollegeofCardiology ,47:799{803,2006. [29]RebeccaRVanderpool,AhRamKim,RobertMolthen,andNaomiCChesler. EectsofacuteRhokinaseinhibitiononchronichypoxia-inducedchangesin proximalanddistalpulmonaryarterialstructureandfunction. JournalofappliedphysiologyBethesda,Md.:1985 ,110:188{198,2011. [30]CarlosA.Dias,RenatoS.Assad,LuizF.Caneo,MariaCristinaDAbduch, VeraD.Aiello,AltamiroR.Dias,MiguelBarberoMarcial,andSergioA. Oliveira.Reversiblepulmonarytrunkbanding.II.Anexperimentalmodelfor rapidpulmonaryventricularhypertrophy. JournalofThoracicandCardiovascularSurgery ,124:999{1006,2002. 86

PAGE 103

[31]EPChen,DMCraig,HBBittner,RDDavis,andPVanTrigt.Pharmacologicalstrategiesforimprovingdiastolicdysfunctioninthesettingofchronic pulmonaryhypertension. Circulation ,97:1606{1612,1998. [32]EricKLouie,SteveSLin,SandraIReynertson,BruceHBrundage,Sidney Levitsky,andStuartRich.PressureandVolumeLoadingoftheRightVentricleHaveOppositeEectsonLeftVentricularEjectionFraction. Circulation 92:819{824,1995. [33]KMChin,NHKim,andLJRubin.Therightventricleinpulmonaryhypertension. CoronArteryDis ,16:13{18,2005. [34]GEDAlonzo,RJBarst,SMAyres,EHBergofsky,BHBrundage,KM Detre,APFishman,RMGoldring,BMGroves,JTKernis,andEtAl.Survivalinpatientswithprimarypulmonaryhypertension.Resultsfromanational prospectiveregistry. AnnInternMed ,115:343{349,1991. [35]VallerieVMcLaughlin,O.Sitbon,D.B.Badesch,R.J.Barst,C.Black, N.Galie,M.Rainisio,G.Simonneau,andL.J.Rubin.Survivalwithrst-line bosentaninpatientswithprimarypulmonaryhypertension. EuropeanRespiratoryJournal ,25:244{249,2005. [36]AShimony,BDFox,DLangleben,andLGRudski.Incidenceandsignicance ofpericardialeusioninpatientswithpulmonaryarterialhypertension. CanJ Cardiol ,29:678{682,2013. [37]KLDavis,UMehlhorn,GALaine,andSJAllen.Myocardialedema,left ventricularfunction,andpulmonaryhypertension. JApplPhysiol 78:132{137,1995. [38]CRGreyson.Pathophysiologyofrightventricularfailure. CritCareMed ,36 Suppl:S57{65,2008. [39]UMehlhorn,HJGeissler,GALaine,andSJAllen.Roleofthecardiaclymph systeminmyocardialuidbalance. EurJCardiothoracSurg ,20:424{427, 2001. [40]MLoukas,NAbel,RSTubbs,JGrabska,JBirungi,andRHAnderson.The cardiaclymphaticsystem. ClinAnat ,24:684{691,2011. [41]RAJohnsonandTMBlake.Lymphaticsoftheheart. Circulation ,33:137{ 142,1966. [42]YingjieCui.Theroleoflymphaticvesselsintheheart,2010. [43]OttoF.Kampmeier.Onthelymphowofthehumanheart,withreference tothedevelopmentofthechannelsandtherstappearance,distribution,and physiologyoftheirvalves. AmericanHeartJournal ,4:210{222,dec1928. 87

PAGE 104

[44]DMCaro,ABerjon,JTeijeira,andCGDuran.[Anatomicalstudyofthe cardiaclymphatics.Theirpathogenicroleinpericardialeusions]. AnnChir ThoracCardiovasc ,11:373{376,1972. [45]IKKline,AJMiller,RPick,andLNKatz.TheRelationshipbetween HumanEndocardialFibroelastosisandObstructionoftheCardiacLymphatics. Circulation ,30:728{735,1964. [46]SMMaurice,ASPalmer,AJMiller,andRGreene.Lymphaticdrainageof theheartinthelaboratoryrat. Lymphology ,34:145{148,2001. [47]SRUllal,THKluge,andFGerbode.Functionalandpathologicchangesinthe heartfollowingchroniccardiaclymphaticobstruction. Surgery ,71:328{334, 1972. [48]DanielA.Goodenough,JereyA.JefreyAJereyA.Goliger,andDavidL. Paul.Connexins,Connexons,andIntercellularCommunication. AnnualReviewofBiochemistry ,65:475{502,1996. [49]EwaStepniak,GlennLRadice,andValeriVasioukhin.Adhesiveandsignaling functionsofcadherinsandcateninsinvertebratedevelopment. ColdSpring Harborperspectivesinbiology ,1:a002949,2009. [50]HalinaS.Chkourko,GuadalupeGuerrero-Serna,XianmingLin,NedalDarwish, JoshuaR.Pohlmann,KeithE.Cook,JereyR.Martens,EliRothenberg,HassanMusa,andMarioDelmar.Remodelingofmechanicaljunctionsandof microtubule-associatedproteinsaccompanycardiacconnexin43lateralization. HeartRhythm ,9:1133{1140,2012. [51]AndrewL.Harris.Connexinchannelpermeabilitytocytoplasmicmolecules, 2007. [52]JuanCSaez,VivianaMBerthoud,MariaCBranes,AgustinDMartinez,and EricCBeyer.Plasmamembranechannelsformedbyconnexins:theirregulation andfunctions. Physiologicalreviews ,83:1359{1400,2003. [53]GoranSohlandKlausWillecke.Gapjunctionsandtheconnexinproteinfamily, 2004. [54]TWWhiteandDLPaul.Geneticdiseasesandgeneknockoutsrevealdiverse connexinfunctions. Annualreviewofphysiology ,61:283{310,1999. [55]DWLaird.Lifecycleofconnexinsinhealthanddisease. TheBiochemical journal ,394Pt3:527{543,2006. [56]StephanRohr.Roleofgapjunctionsinthepropagationofthecardiacaction potential. CardiovascularResearch ,62:309{322,2004. 88

PAGE 105

[57]MJSanderson,IChow,andERDirksen.Intercellularcommunicationbetween ciliatedcellsinculture. TheAmericanjournalofphysiology ,254Pt1:C63{ 74,1988. [58]LatoyaNJohnsonandMichaelKoval.Cross-talkbetweenpulmonaryinjury, oxidantstress,andgapjunctionalcommunication. Antioxidants&redoxsignaling ,11:355{367,2009. [59]KaushikParthasarathi,HideoIchimura,EijiMonma,JensLindert,Sadiqa Quadri,AndrewIssekutz,andJaharBhattacharya.Connexin43mediates spreadofCa2+-dependentproinammatoryresponsesinlungcapillaries. The Journalofclinicalinvestigation ,116:2193{200,2006. [60]XiaoyouYing,YoshihiroMinamiya,ChenzhongFu,andJaharBhattacharya. Ca2+WavesinLungCapillaryEndothelium. Circ.Res. ,79:898{908,1996. [61]NJSevers,AlexandraF.Bruce,EmmanuelDupont,andStephenRothery. Remodellingofgapjunctionsandconnexinexpressionindiseasedmyocardium, 2008. [62]NJSevers.Thecardiacmusclecell. BioEssays ,22:188{99,2000. [63]MagdaSCFontes,ToonABVanVeen,JacquesMTDeBakker,andHarold VMVanRijen.FunctionalconsequencesofabnormalCx43expressionin theheart. BiochimicaetBiophysicaActa-Biomembranes ,1818:2020{2029, 2012. [64]StephanieUrschel,ThorstenHoher,TimmSchubert,CantasAlev,GoranSohl, PhilippWorsdorfer,TakayukiAsahara,RolfDermietzel,RetoWeiler,andKlaus Willecke.ProteinkinaseA-mediatedphosphorylationofconnexin36inmouse retinaresultsindecreasedgapjunctionalcommunicationbetweenAIIamacrine cells. JournalofBiologicalChemistry ,281:33163{33171,2006. [65]NMKumarandNBGilula.Thegapjunctioncommunicationchannel. Cell 84:381{388,1996. [66]MABeardslee,JGLaing,ECBeyer,andJESatz.Rapidturnoverof connexin43intheadultratheart. CircRes ,83:629{635,1998. [67]DSCrow,ECBeyer,DLPaul,SSKobe,andAFLau.Phosphorylationofconnexin43gapjunctionproteininuninfectedandRoussarcomavirustransformedmammalianbroblasts. Molecularandcellularbiology ,10:1754{ 1763,1990. [68]DWLaird,KLPuranam,andJPRevel.Turnoverandphosphorylation dynamicsofconnexin43gapjunctionproteininculturedcardiacmyocytes. TheBiochemicaljournal ,273Pt1:67{72,1991. 89

PAGE 106

[69]PaulD.Lampe.Analyzingphorbolestereectsongapjunctionalcommunication:Adramaticinhibitionofassembly. JournalofCellBiology ,127 II:1895{1905,1994. [70]LindaS.Musil,EricC.Beyer,andDanielA.Goodenough.Expressionofthe gapjunctionproteinconnexin43inembryonicchicklens:Molecularcloning,ultrastructurallocalization,andpost-translationalphosphorylation. TheJournal ofMembraneBiology ,116:163{175,1990. [71]GALaineandSJAllen.Leftventricularmyocardialedema.Lymphow, interstitialbrosis,andcardiacfunction. CircRes ,68:1713{1721,1991. [72]RaymondL.Benza,DaveP.Miller,MardiGomberg-Maitland,RobertP. Frantz,AimeeJ.Foreman,ChristopherS.Coey,AdaaniFrost,RobynJ.Barst, DavidB.Badesch,C.GregoryElliott,TheodoreG.Liou,andMichaelD.McGoon.Predictingsurvivalinpulmonaryarterialhypertension:Insightsfrom theregistrytoevaluateearlyandlong-termpulmonaryarterialhypertension diseasemanagementREVEAL. Circulation ,122:164{172,2010. [73]MarcHumbert,OlivierSitbon,AriChaouat,MicheleBertocchi,GilbertHabib, VirginieGressin,AzzedineYaci,EmmanuelWeitzenblum,Jean-Francois Cordier,FrancoisChabot,ClaireDromer,ChristophePison,MartineReynaudGaubert,AlainHaloun,MarcelLaurent,EricHachulla,VincentCottin,Bruno Degano,XavierJas,DavidMontani,RogerioSouza,andGeraldSimonneau. Survivalinpatientswithidiopathic,familial,andanorexigen-associatedpulmonaryarterialhypertensioninthemodernmanagementera. Circulation 122:156{63,2010. [74]ALHinderliter,PWthWillis,WLong,WRClarke,DRalph,EJCaldwell, WWilliams,NAEttinger,NSHill,WRSummer,BdeBiosblanc,GKoch, SLi,LMClayton,MMJobsis,andJWCrow.Frequencyandprognostic signicanceofpericardialeusioninprimarypulmonaryhypertension.PPH StudyGroup.Primarypulmonaryhypertension. AmJCardiol ,84:481{4, A10,1999. [75]OmarBatal,OmarFKhatib,RaedaDweik,JereyPHammel,KevinMcCarthy,andOmaraMinai.Comparisonofbaselinepredictorsofprognosisin pulmonaryarterialhypertensioninpatientssurviving2yearsandthosesurviving5yearsafterbaselineright-sidedcardiaccatheterization. TheAmerican journalofcardiology ,109:1514{20,2012. [76]SSahayandARTonelli.Pericardialeusioninpulmonaryarterialhypertension. PulmCirc ,3:467{477,2013. [77]TetsujiMiura,ToshiyukiYano,KazuyukiNaitoh,MasahiroNishihara, TakayukiMiki,MasayaTanno,andKazuakiShimamoto.Delta-opioidreceptor activationbeforeischemiareducesgapjunctionpermeabilityinischemicmyocardiumbyPKC-epsilon-mediatedphosphorylationofconnexin43. American 90

PAGE 107

journalofphysiology.Heartandcirculatoryphysiology ,293:H1425{H1431, 2007. [78]UweSchwanke,InaKonietzka,AlexejDuschin,XiaokuiLi,RainerSchulz, andGerdHeusch.Noischemicpreconditioninginheterozygousconnexin43decientmice. Americanjournalofphysiology.Heartandcirculatoryphysiology ,283:H1740{H1742,2002. [79]TetsujiMiura,YoshitoOhnuma,AtsushiKuno,MasayaTanno,Yoshihiko Ichikawa,YuichiNakamura,ToshiyukiYano,TakayukiMiki,JunSakamoto, andKazuakiShimamoto.Protectiveroleofgapjunctionsinpreconditioning againstmyocardialinfarction. Americanjournalofphysiology.Heartandcirculatoryphysiology ,286:H214{H221,2004. [80]MarisolRuiz-Meana,AntonioRodrguez-Sinovas,AlbertoCabestrero,Kerstin Boengler,GerdHeusch,andDavidGarcia-Dorado.Mitochondrialconnexin43 asanewplayerinthepathophysiologyofmyocardialischaemia-reperfusion injury,2008. [81]RSepp,NJSevers,andRGGourdie.Alteredpatternsofcardiacintercellular junctiondistributioninhypertrophiccardiomyopathy. HeartBritishCardiac Society ,76:412{417,1996. [82]NJSevers,SRCoppen,EDupont,HIYeh,YSKo,andTMatsushita.Gap junctionalterationsinhumancardiacdisease. CardiovascRes ,62:368{377, 2004. [83]FadiGAkar,RobertDNass,SamuelHahn,EugenioCingolani,ManishShah, GeoreyGHesketh,DeborahDiSilvestre,RichardSTunin,DavidAKass,and GordonFTomaselli.Dynamicchangesinconductionvelocityandgapjunctionpropertiesduringdevelopmentofpacing-inducedheartfailure. American journalofphysiology.Heartandcirculatoryphysiology ,293:H1223{H1230, 2007. [84]NicholasJ.Severs,EmmanuelDupont,NeilThomas,RiyazKaba,Stephen Rothery,RevtiJain,KatherineSharpey,andChristopherH.Fry.Alterations incardiacconnexinexpressionincardiomyopathies,2006. [85]ArchanaRajdev,HasanGaran,andAngeloBiviano.Arrhythmiasinpulmonary arterialhypertension. Progressincardiovasculardiseases ,55:180{6,2012. [86]RZak.Cellproliferationduringcardiacgrowth. TheAmericanjournalof cardiology ,31:211{219,1973. [87]ACNag.Studyofnon-musclecellsoftheadultmammalianheart:ane structuralanalysisanddistribution. Cytobios ,28:41{61,1980. 91

PAGE 108

[88]TroyABaudino,WayneCarver,WayneGiles,andThomasKBorg.Cardiacbroblasts:friendorfoe? Americanjournalofphysiology.Heartandcirculatory physiology ,291:H1015{H1026,2006. [89]P.Camelliti,T.K.Borg,andP.Kohl.Structuralandfunctionalcharacterisationofcardiacbroblasts. CardiovascularResearch ,65:40{51,2005. [90]EdsonABregagnollo,LeonardoAMZorno,KatashiOkoshi,MarioSugizaki, MarcoAMestrinel,CarlosRPadovani,andAntonioCCicogna.Myocardial contractiledysfunctioncontributestothedevelopmentofheartfailureinrats withaorticstenosis. Internationaljournalofcardiology ,117:109{114,2007. [91]A.K.Wong,T.G.Osborn,J.G.Miller,andS.A.Wickline.Quanticationof ventricularremodelinginthetight-skinmousecardiomyopathywithacoustic microscopy. UltrasoundinMedicineandBiology ,19:365{374,1993. [92]TKBorgandJBCauleld.Thecollagenmatrixoftheheart. Federation proceedings ,40:2037{2041,1981. [93]A.A.Young,I.J.Legrice,M.A.Young,andB.H.Smaill.Extendedconfocal microscopyofmyocardiallaminaeandcollagennetwork. JournalofMicroscopy 192:139{150,1998. [94]EdieC.Goldsmith,AdamHoman,MaryO.Morales,JayD.Potts,RobertL. Price,AlexMcFadden,MichaelRice,andThomasK.Borg.Organizationof broblastsintheheart. DevelopmentalDynamics ,230:787{794,2004. [95]MaryOMorales,RobertLPrice,andEdieCGoldsmith.ExpressionofDiscoidinDomainReceptor2DDR2inthedevelopingheart. Microscopyandmicroanalysis:theocialjournalofMicroscopySocietyofAmerica,Microbeam AnalysisSociety,MicroscopicalSocietyofCanada ,11:260{267,2005. [96]PeterKohl.Cardiaccellularheterogeneityandremodelling. Cardiovascular research ,64:195{7,nov2004. [97]MarcHumbert,NicholasWMorrell,StephenLArcher,KurtRStenmark,MargaretRMacLean,IreneMLang,BrianWChristman,EKennethWeir,Oliver Eickelberg,NorbertFVoelkel,andMarleneRabinovitch.Cellularandmolecularpathobiologyofpulmonaryarterialhypertension. JournaloftheAmerican CollegeofCardiology ,43SupplS:13S{24S,jun2004. [98]JereyWHolmes,ThomasKBorg,andJamesWCovell.Structureandmechanicsofhealingmyocardialinfarcts. Annualreviewofbiomedicalengineering 7:223{253,2005. [99]BarryMGumbiner.Regulationofcadherin-mediatedadhesioninmorphogenesis. Naturereviews.Molecularcellbiology ,6:622{634,2005. 92

PAGE 109

[100]PCamelliti,ColinR.Green,IanLeGrice,andPeterKohl.FibroblastNetwork inRabbitSinoatrialNode:StructuralandFunctionalIdenticationofHomogeneousandHeterogeneousCellCoupling. CirculationResearch ,94:828{835, 2004. [101]GiedriusGaudesius,MicheleMiragoli,StuartP.Thomas,andStephanRohr. Couplingofcardiacelectricalactivityoverextendeddistancesbybroblastsof cardiacorigin. CirculationResearch ,93:421{428,2003. [102]IchiroManabe,TakayukiShindo,andRyozoNagai.Geneexpressioninbroblastsandbrosisinvolvementincardiachypertrophy,2002. [103]JunYoshioka,RobinNPrince,HaydenHuang,ScottBPerkins,FranciscoU Cruz,CatherineMacGillivray,DouglasALauenburger,andRichardTLee. Cardiomyocytehypertrophyanddegradationofconnexin43throughspatially restrictedautocrine/paracrineheparin-bindingEGF. ProceedingsoftheNationalAcademyofSciencesoftheUnitedStatesofAmerica ,102:10622{ 10627,2005. [104]DeidreMacKenna,SonyaR.Summerour,andFranciscoJ.Villarreal.Roleof mechanicalfactorsinmodulatingcardiacbroblastfunctionandextracellular matrixsynthesis,2000. [105]YaoSun,MohammadF.Kiani,ArnoldE.Postlethwaite,andKarlT.Weber. Infarctscaraslivingtissue,2002. [106]NeilaTurner,KarenEPorter,WilliamHTSmith,HazelLWhite,StephenG Ball,andAnthonyJBalmforth.Chronicbeta2-adrenergicreceptorstimulation increasesproliferationofhumancardiacbroblastsviaanautocrinemechanism. Cardiovascularresearch ,57:784{792,2003. [107]AlexandraF.Bruce,StephenRothery,EmmanuelDupont,andNicholasJ. Severs.Gapjunctionremodellinginhumanheartfailureisassociatedwith increasedinteractionofconnexin43withZO-1. CardiovascularResearch 77:757{765,2008. [108]RHHoyt,MLCohen,andJESatz.Distributionandthree-dimensional structureofintercellularjunctionsincaninemyocardium. Circulationresearch 64:563{574,1989. [109]MJVanKempen,CFromaget,DGros,AFMoorman,andWHLamers. Spatialdistributionofconnexin43,themajorcardiacgapjunctionprotein,in thedevelopingandadultratheart. CircRes ,68:1638{1651,1991. [110]JoellLSolanandPaulDLampe.Connexin43phosphorylation:structural changesandbiologicaleects. BiochemicalJournal ,419:261{272,2009. 93

PAGE 110

[111]EDupont,TMatsushita,RAKaba,CVozzi,SRCoppen,NKhan,RKaprielian,MHYacoub,andNJSevers.Alteredconnexinexpressioninhuman congestiveheartfailure. Journalofmolecularandcellularcardiology ,33:359{ 371,2001. [112]RRKaprielian,MGunning,EDupont,MNSheppard,SMRothery,RUnderwood,DJPennell,KFox,JPepper,PAPoole-Wilson,andNJSevers. Downregulationofimmunodetectableconnexin43anddecreasedgapjunction sizeinthepathogenesisofchronichibernationinthehumanleftventricle. Circulation ,97:651{660,1998. [113]SawaKostin,MarkusRieger,SebastianDammer,StefanHein,ManfredRichter, WolfPeterKlovekorn,ErwinP.Bauer,andJuttaSchaper.Gapjunctionremodelingandalteredconnexin43expressioninthefailinghumanheart. MolecularandCellularBiochemistry ,242-2:135{144,2003. [114]NSPeters,JCoromilas,NJSevers,andALWit.Disturbedconnexin43gap junctiondistributioncorrelateswiththelocationofreentrantcircuitsintheepicardialborderzoneofhealingcanineinfarctsthatcauseventriculartachycardia. Circulation ,95:988{996,1997. [115]JiaxiangQu,FrankM.Volpicelli,LuisI.Garcia,NefthiSandeep,JieZhang,LucreciaMarquez-Rosado,PaulD.Lampe,andGlennI.Fishman.Gapjunction remodelingandspironolactone-dependentreverseremodelinginthehypertrophiedheart. CirculationResearch ,104:365{371,2009. [116]XiaoYanTanandJianGuoHe.Theremodelingofconnexininthehypertrophiedrightventricularinpulmonaryarterialhypertensionandtheeectof adualETreceptorantagonistbosentan. PathologyResearchandPractice 205:473{482,2009. [117]HJJongsmaandRWilders.Gapjunctionsincardiovasculardisease. Circulationresearch ,86:1193{1197,2000. [118]ChiekoSasano,HaruoHonjo,YoshikoTakagishi,MahmudUzzaman,LuniEmdad,AtsuyaShimizu,YoshiharuMurata,KaichiroKamiya,andItsuoKodama. Internalizationanddephosphorylationofconnexin43inhypertrophiedright ventriclesofratswithpulmonaryhypertension. Circulationjournal:ocial journaloftheJapaneseCirculationSociety ,71:382{389,2007. [119]MahamadNavab,FengLiao,GregoryP.Hough,LoriA.Ross,BrianJ.Van Lenten,TripathiB.Rajavashisth,AidonsJ.Lusis,HillelLaks,DavisC. Drinkwater,andAlanM.Fogelman.Interactionofmonocyteswithcocultures ofhumanaorticwallcellsinvolvesinterleukins1and6withmarkedincreasesin connexin43message. JournalofClinicalInvestigation ,87:1763{1772,1991. [120]NorbertFrey,Hugoa.Katus,EricN.Olson,andJosepha.Hill.Hypertrophyof theHeart:ANewTherapeuticTarget? Circulation ,109:1580{1589,2004. 94

PAGE 111

[121]MarkHDrazner.Theprogressionofhypertensiveheartdisease. Circulation 123:327{34,jan2011. [122]BSjoblom,aSalmazo,andKDjinovic-Carugo.Alpha-actininstructureand regulation. Cellularandmolecularlifesciences:CMLS ,65:2688{701,2008. [123]AHBeggs,TJByers,JHKnoll,FMBoyce,GABruns,andLMKunkel. Cloningandcharacterizationoftwohumanskeletalmuscle -actiningenes locatedonchromosomes1and11. TheJournalofBiologicalChemistry 267:9281{9288,1992. [124]HYoussouan,MMcAfee,andDJKwiatkowski.Cloningandchromosomal localizationofthehumancytoskeletalalpha-actiningenerevealslinkagetothe beta-spectringene. AmJHumGenet ,47:62{71,1990. [125]KathleenaClark,AbigailSMcElhinny,MaryCBeckerle,andCarolCGregorio. Striatedmusclecytoarchitecture:anintricatewebofformandfunction. Annual reviewofcellanddevelopmentalbiology ,18:637{706,2002. [126]KBurridge,TKelly,andLConnell.Proteinsinvolvedintheattachment ofactintotheplasmamembrane. PhilosTransRSocLondBBiolSci 299:291{299,1982. [127]ColbyASouders,StephanieLKBowers,andTroyABaudino.Cardiacbroblast:therenaissancecell. Circulationresearch ,105:1164{76,2009. [128]ChristopherL.Smith,SeungTaeBaek,CarolineY.Sung,andMichelleD. Tallquist.Epicardial-derivedcellepithelial-to-mesenchymaltransitionand fatespecicationrequirePDGFreceptorsignaling. CirculationResearch 108:e15{e26,2011. [129]ElisabethM.ZeisbergandRaghuKalluri.Originsofcardiacbroblasts. CirculationResearch ,107:1304{1312,2010. [130]FStrutz,HOkada,CWLo,TDano,RLCarone,JETomaszewski,and EGNeilson.Identicationandcharacterizationofabroblastmarker:FSP1. TheJournalofcellbiology ,130:393{405,1995. [131]PingKong,PanagiotaChristia,AmitSaxena,YaSu,andNikolaosGFrangogiannis.Lackofspecicityofbroblast-specicprotein1incardiacremodelingandbrosis. AmericanJournalofPhysiology-HeartandCirculatoryPhysiology ,305:H1363|-H1372,2013. [132]YingTianandEdwardEMorrisey.Importanceofmyocyte-nonmyocyteinteractionsincardiacdevelopmentanddisease. Circulationresearch ,110:1023{ 34,2012. 95

PAGE 112

[133]CSBoomershine,aChamberlain,PKendall,aRAfshar-Sharif,HHuang,MK Washington,WELawson,JWThomas,TSBlackwell,andNaBhowmick. AutoimmunepancreatitisresultsfromlossofTGFbetasignallinginS100A4positivedendriticcells. Gut ,58:1267{1274,2009. [134]AnneC.Brisset,HiroyukiHao,EdoardoCamenzind,MarcBacchetta,Antoine Geinoz,JeanCharlesSanchez,ChristineChaponnier,GiulioGabbiani,and MarieLuceBochaton-Piallat.IntimalsmoothmusclecellsofporcineandhumancoronaryarteryexpressS100A4,amarkeroftherhomboidphenotypein vitro. CirculationResearch ,100:1055{1062,2007. [135]MichelLeHir,IvanHegyi,DominiqueCueni-Long,JohannesLong,and BrigitteKaissling.Characterizationofrenalinterstitialbroblast-specicprotein1/S100A4-positivecellsinhealthyandinamedrodentkidneys. Histochemistryandcellbiology ,123-5:335{46,2005. [136]ChristophH Osterreicher,MelittaPenzOsterreicher,SergeiIGrivennikov, MonicaGuma,EkaterinaKKoltsova,ChristianDatz,RomanSasik,Gary Hardiman,MichaelKarin,andDavidaBrenner.Fibroblast-specicprotein 1identiesaninammatorysubpopulationofmacrophagesintheliver. ProceedingsoftheNationalAcademyofSciencesoftheUnitedStatesofAmerica 108:308{13,2011. [137]JianrongLu,Jamesa.Richardson,andEricN.Olson.Capsulin:Anovel bHLHtranscriptionfactorexpressedinepicardialprogenitorsandmesenchyme ofvisceralorgans. MechanismsofDevelopment ,73:23{32,1998. [138]a.Acharya,S.T.Baek,G.Huang,B.Eskiocak,S.Goetsch,C.Y.Sung,S.Ban, M.F.Sauer,G.S.Olsen,J.S.Dueld,E.N.Olson,andM.D.Tallquist.The bHLHtranscriptionfactorTcf21isrequiredforlineage-specicEMTofcardiac broblastprogenitors. Development ,139:2139{2149,2012. [139]KellyLGorresandRonaldTRaines.Prolyl4-hydroxylase. Criticalreviewsin biochemistryandmolecularbiology ,45:106{24,apr2010. [140]JohannaMyllyharju.Prolyl4-hydroxylases,keyenzymesinthesynthesisof collagensandregulationoftheresponsetohypoxia,andtheirrolesastreatment targets. Annalsofmedicine ,40January:402{417,2008. [141]J.a.McDonald,T.J.Broekelmann,M.L.Matheke,E.Crouch,M.Koo,and C.Kuhn.AmonoclonalantibodytothecarboxyterminaldomainofprocollagentypeIvisualizescollagen-synthesizingbroblasts.Detectionofanaltered broblastphenotypeinlungsofpatientswithpulmonarybrosis. Journalof ClinicalInvestigation ,78November:1237{1244,1986. [142]NaThornberry.Thecaspasefamilyofcysteineproteases. Britishmedical bulletin ,53:478{90,1997. 96

PAGE 113

[143]J.Narula,P.Pandey,E.Arbustini,N.Haider,N.Narula,F.D.Kolodgie,B.Dal Bello,M.J.Semigran,A.Bielsa-Masdeu,G.W.Dec,S.Israels,M.Ballester, R.Virmani,S.Saxena,andS.Kharbanda.Apoptosisinheartfailure:Release ofcytochromecfrommitochondriaandactivationofcaspase-3inhumancardiomyopathy. ProceedingsoftheNationalAcademyofSciences ,964:8144{ 8149,jul1999. [144]CBELTRAMI,NFINATO,MROCCO,GFERUGLIO,CPURICELLI, ECIGOLA,ESONNENBLICK,GOLIVETTI,andPANVERSA.Thecellularbasisofdilatedcardiomyopathyinhumans. JournalofMolecularand CellularCardiology ,27:291{305,jan1995. [145]E.J.EichhornandM.R.Bristow.MedicalTherapyCanImprovetheBiological PropertiesoftheChronicallyFailingHeart:ANewEraintheTreatmentof HeartFailure. Circulation ,94:2285{2296,nov1996. [146]aHWyllie,JFKerr,andaRCurrie. Celldeath:thesignicanceofapoptosis. volume68.1980. [147]W.Bursch,L.Kleine,andM.Tenniswood.Thebiochemistryofcelldeathby apoptosis. BiochemistryandCellBiology ,88:1071{1074,1990. [148]TFernandes-Alnemri,GLitwack,andESAlnemri.CPP32,anovelhuman apoptoticproteinwithhomologytoCaenorhabditiseleganscelldeathprotein Ced-3andmammalianinterleukin-1beta-convertingenzyme. TheJournalof biologicalchemistry ,269:30761{4,1994. [149]YEmoto,YManome,GMeinhardt,HKisaki,SKharbanda,MRobertson, TGhayur,WWWong,RKamen,andRWeichselbaum.Proteolyticactivation ofproteinkinaseCdeltabyanICE-likeproteaseinapoptoticcells. TheEMBO journal ,14:6148{6156,1995. [150]MTewari,LTQuan,KO'Rourke,SDesnoyers,ZZeng,DRBeidler,GG Poirier,GSSalvesen,andVMDixit.Yama/CPP32beta,amammalianhomologofCED-3,isaCrmA-inhibitableproteasethatcleavesthedeathsubstrate polyADP-ribosepolymerase. Cell ,81:801{809,1995. [151]XWang,NZelenski,JYang,JSakai,MBrown,andJGoldstein.Cleavage ofsterolregulatoryelementbindingproteins f SREBPs g by f CPP32 g during apoptosis. The f EMBO g journal ,15:1012{1020,1996. [152]MarylaKrajewska,HongGangWang,StanislawKrajewski,JuanM.Zapata, AhmedShabaik,RandyGascoyne,andJohnC.Reed.Immunohistochemical analysisofinvivopatternsofexpressionofCPP32Caspase-3,acelldeath protease. CancerResearch ,57:1605{1613,1997. 97

PAGE 114

[153]LaszloFarkas,JackGauldie,NorbertFVoelkel,andMartinKolb.Pulmonary hypertensionandidiopathicpulmonarybrosis:ataleofangiogenesis,apoptosis,andgrowthfactors. Americanjournalofrespiratorycellandmolecular biology ,45:1{15,2011. [154]JGerdes,USchwab,HLemke,andHStein.Productionofamousemonoclonalantibodyreactivewithahumannuclearantigenassociatedwithcell proliferation. IntJCancer ,31:13{20,1983. [155]JGerdes,HLemke,HBaisch,HHWacker,USchwab,andHStein.Cellcycle analysisofacellproliferation-associatedhumannuclearantigendenedbythe monoclonalantibodyKi-67. JournalofimmunologyBaltimore,Md.:1950 133:1710{1715,1984. [156]TamaraTajsicandNicholasWMorrell.Smoothmusclecellhypertrophy,proliferation,migrationandapoptosisinpulmonaryhypertension. Comprehensive Physiology ,1:295{317,2011. [157]PeterBalukandDonaldMMcDonald.Markersformicroscopicimagingoflymphangiogenesisandangiogenesis. AnnalsoftheNewYorkAcademyofSciences 1131:1{12,2008. [158]DavidG.Jackson.BiologyofthelymphaticmarkerLYVE-1andapplications inresearchintolymphatictrackingandlymphangiogenesis. Apmis ,1128:526{538,2004. [159]DavidG.Jackson.Thelymphaticsrevisited:Newperspectivesfromthe hyaluronanreceptorLYVE-1. TrendsinCardiovascularMedicine ,13:1{7, 2003. [160]VivienSchacht,MariaIRamirez,Young-KwonHong,SatoshiHirakawa,Dian Feng,NatashaHarvey,MaryWilliams,AnnMDvorak,HaroldFDvorak, GuillermoOliver,andMichaelDetmar.T1alpha/podoplanindeciencydisruptsnormallymphaticvasculatureformationandcauseslymphedema. The EMBOjournal ,22:3546{56,2003. [161]SBreiteneder-Gele,ASoleiman,HKowalski,RHorvat,GAmann,EKriehuber,KDiem,WWeninger,ETschachler,KAlitalo,andDKerjaschki.Angiosarcomasexpressmixedendothelialphenotypesofbloodandlymphaticcapillaries:podoplaninasaspecicmarkerforlymphaticendothelium. TheAmericanjournalofpathology ,154:385{94,1999. [162]LelandGDobbs,MCWilliams,andRGonzalez.MonoclonalantibodiesspecictoapicalsurfacesofratalveolartypeIcellsbindtosurfacesofcultured,but notfreshlyisolated,typeIIcells. Biochimicaetbiophysicaacta ,970:146{56, 1988. 98

PAGE 115

[163]VivienSchacht,SoheilSDadras,LouiseAJohnson,DavidGJackson,YoungKwonHong,andMichaelDetmar.Up-regulationofthelymphaticmarker podoplanin,amucin-typetransmembraneglycoprotein,inhumansquamous cellcarcinomasandgermcelltumors. TheAmericanjournalofpathology 166:913{21,2005. [164]JTWigle,NHarvey,MDetmar,ILagutina,GGrosveld,MDGunn,DG Jackson,andGOliver.AnessentialroleforProx1intheinductionofthe lymphaticendothelialcellphenotype. EMBOJournal ,21:1505{1513,2002. [165]JorgWilting,MariaPapoutsi,BodoChrist,KyprosHNicolaides,ConstantinS vonKaisenberg,JorgBorges,GBjornStark,KariAlitalo,StanislavITomarev, CharlotteNiemeyer,andJochenRossler.ThetranscriptionfactorProx1isa markerforlymphaticendothelialcellsinnormalanddiseasedhumantissues. FASEBjournal:ocialpublicationoftheFederationofAmericanSocieties forExperimentalBiology ,16:1271{3,2002. [166]MarkkuMiettinenandZeng-FengWang.Prox1transcriptionfactorasamarker forvasculartumors-evaluationof314vascularendothelialand1086nonvascular tumors. TheAmericanjournalofsurgicalpathology ,36:351{9,2012. [167]TCLaurentandJRFraser.Hyaluronan. TheFASEBjournal:ocial publicationoftheFederationofAmericanSocietiesforExperimentalBiology 6:2397{2404,1992. [168]JanetYLeeandAndrewPSpicer.Hyaluronan:amultifunctional,megaDalton, stealthmolecule. CurrentOpinioninCellBiology ,12:581{586,2000. [169]JRFraser,TCLaurent,andUBLaurent.Hyaluronan:itsnature,distribution,functionsandturnover. Journalofinternalmedicine ,242:27{33,1997. [170]PWNoble,CMMcKee,MCowman,andHSShin.HyaluronanfragmentsactivateanNF-kappaB/I-kappaBalphaautoregulatoryloopinmurine macrophages. TheJournalofexperimentalmedicine ,183May:2373{2378, 1996. [171]MJReid,MWLame,DMorin,DWWilson,andHJSegall.InvolvementofcytochromeP4503Ainthemetabolismandcovalentbindingof14Cmonocrotalineinratlivermicrosomes. Journalofbiochemicalandmolecular toxicology ,12:157{166,1998. [172]DWWilson,HJSegall,LCPan,MWLame,JEEstep,andDMorin.Mechanismsandpathologyofmonocrotalinepulmonarytoxicity. Criticalreviewsin toxicology ,22-6:307{325,1992. [173]HCRosenbergandMRabinovitch.Endothelialinjuryandvascularreactivity inmonocrotalinepulmonaryhypertension. TheAmericanjournalofphysiology 255Pt2:H1484{H1491,1988. 99

PAGE 116

[174]JasonLee,ReubenReich,FangXu,andPravinBSehgal.Golgi,tracking,andmitosisdysfunctionsinpulmonaryarterialendothelialcellsexposed tomonocrotalinepyrroleandNOscavenging. Americanjournalofphysiology. Lungcellularandmolecularphysiology ,297:L715{L728,2009. [175]JingHuang,JohnHWolk,MichaelHGewitz,andRajammaMathew.Progressiveendothelialcelldamageinaninammatorymodelofpulmonaryhypertension. Experimentallungresearch ,36:57{66,2010. [176]AMolteni,WFWard,CHTs'ao,CDPort,andNHSolliday.Monocrotalineinducedpulmonaryendothelialdysfunctioninrats. Proc.Soc.Exp.Biol.Med 176:88{94,1984. [177]J.G.Gomez-Arroyo,L.Farkas,a.a.Alhussaini,D.Farkas,D.Kraskauskas, N.F.Voelkel,andH.J.Bogaard.Themonocrotalinemodelofpulmonary hypertensioninperspective. AJP:LungCellularandMolecularPhysiology 302:L363{L369,feb2012. [178]YLLai,JWOlson,andMNGillespie.Ventilatorydysfunctionprecedes pulmonaryvascularchangesinmonocrotaline-treatedrats. Journalofapplied physiologyBethesda,Md.:1985 ,70:561{566,1991. [179]DaliaUrboniene,IdithHaber,Yong-HuFang,ThenappanThenappan,and StephenLArcher.Validationofhigh-resolutionechocardiographyandmagneticresonanceimagingvs.high-delitycatheterizationinexperimentalpulmonaryhypertension. Americanjournalofphysiology.Lungcellularandmolecularphysiology ,299:L401{L412,2010. [180]T.AnnieTFong,LauraK.Shawver,LiSun,ChoTang,HaraldApp,T.Je Powell,YoungH.Kim,RandallSchreck,XueyanWang,WernerRisau,Axel Ullrich,K.PeterHirth,andGeraldMcMahon.SU5416isapotentandselective inhibitorofthevascularendothelialgrowthfactorreceptorFlk-1/KDRthat inhibitstyrosinekinasecatalysis,tumorvascularization,andgrowthofmultiple tumortypes. CancerResearch ,59:99{106,1999. [181]TakashiItokawa,HirokiNokihara,YasuhikoNishioka,SaburoSone,YukihideIwamoto,YujiYamada,JulieCherrington,GeraldMcMahon,Masabumi Shibuya,MichihikoKuwano,andMayumiOno.Antiangiogeniceectby SU5416ispartlyattributabletoinhibitionofFlt-1receptorsignaling. Molecular cancertherapeutics ,1:295{302,2002. [182]YasunoriKasahara,RubinM.Tuder,LaimuteTaraseviciene-Stewart,TimothyD.LeCras,StevenAbman,PeterK.Hirth,JohannesWaltenberger,and NorbertF.Voelkel.InhibitionofVEGFreceptorscauseslungcellapoptosis andemphysema. JournalofClinicalInvestigation ,106:1311{1319,2000. [183]LTaraseviciene-Stewart,YKasahara,LAlger,PHirth,GMcMahon,JWaltenberger,NFVoelkel,andRMTuder.InhibitionoftheVEGFreceptor2 100

PAGE 117

combinedwithchronichypoxiacausescelldeath-dependentpulmonaryendothelialcellproliferationandseverepulmonaryhypertension. TheFASEBjournal :ocialpublicationoftheFederationofAmericanSocietiesforExperimental Biology ,15:427{438,2001. [184]MarkRNicolls,ShiroMizuno,LaimaTaraseviciene-Stewart,LaszloFarkas, JennniferIDrake,AysarAlHusseini,JoseGGomez-Arroyo,NorbertFVoelkel, andHermanJBogaard.Newmodelsofpulmonaryhypertensionbasedon VEGFreceptorblockade-inducedendothelialcellapoptosis. Pulmonarycirculation ,2:434{42,2012. [185]KLColvinandMEYeager.AnimalModelsofPulmonaryHypertension: MatchingDiseaseMechanismstoEtiologyoftheHumanDisease. JPulmRespir Med ,4,2014. [186]SallyHVitali,GeorgHansmann,ChaseRose,AngelesFernandez-Gonzalez, AnnetteScheid,SAlexMitsialis,andStellaKourembanas.TheSugen5416/hypoxiamousemodelofpulmonaryhypertensionrevisited:long-termfollow-up. Pulmonarycirculation ,4:619{629,dec2014. [187]AndreGKleberandYoramRudy.Basicmechanismsofcardiacimpulsepropagationandassociatedarrhythmias. Physiologicalreviews ,84:431{488,2004. [188]SawaKostin,SebastianDammer,StefanHein,WolfP.Klovekorn,ErwinP. Bauer,andJuttaSchaper.Connexin43expressionanddistributionincompensatedanddecompensatedcardiachypertrophyinpatientswithaorticstenosis. CardiovascularResearch ,62:426{436,2004. [189]ChristineChiu,RichardD.Bagnall,JodieIngles,LauraYeates,MarinaKennerson,Jennifera.Donald,MikaJormakka,JoanneM.Lind,andChristopher Semsarian.MutationsinAlpha-Actinin-2CauseHypertrophicCardiomyopathy.AGenome-WideAnalysis. JournaloftheAmericanCollegeofCardiology 55:1127{1135,2010. [190]BhagyalaxmiMohapatra,ShinaweJimenez,JiuannHueyLin,KarlaR.Bowles, KarenJ.Coveler,JosephG.Marx,Michelea.Chrisco,RossT.Murphy,PaulR. Lurie,RobertJ.Schwartz,PerryM.Elliott,MatteoVatta,WilliamMcKenna, Jereya.Towbin,andNeilE.Bowles.MutationsinthemuscleLIMprotein and -actinin-2genesindilatedcardiomyopathyandendocardialbroelastosis. MolecularGeneticsandMetabolism ,80-2:207{215,2003. [191]NOtsu.Athresholdselectionmethodfromgraylevelhistogram. f IEEE g Trans.Systems,ManandCybernetics ,9:62{66,1979. [192]JamesJ.O'Donnell,Annaa.Birukova,EricC.Beyer,andKonstantinG. Birukov.Gapjunctionproteinconnexin43exacerbateslungvascularpermeability. PLoSONE ,9:1{11,2014. 101

PAGE 118

[193]BrendaR.Kwak,NielsVeillard,GrazianoPelli,FloreMulhaupt,RichardW. James,MarcChanson,andFrancoisMach.Reducedconnexin43expression inhibitsatheroscleroticlesionformationinlow-densitylipoproteinreceptordecientmice. Circulation ,107:1033{1039,2003. [194]XianghongWu,WenHuang,GangLuo,andLavalAndyAlain.Hypoxiainducesconnexin43dysregulationbymodulatingmatrixmetalloproteinasesvia MAPKsignaling. Molecularandcellularbiochemistry ,384-2:155{62,2013. [195]JChen,LHe,BDinger,LJStensaas,andSJFidone.Chronichypoxia upregulatesconnexin43expressioninratcarotidbodyandpetrosalganglion. JournalofappliedphysiologyBethesda,Md.:1985 ,92:1480{6,2002. 102

PAGE 119

APPENDIXA.CustomMATLABCodes 103

PAGE 120

104

PAGE 121

105

PAGE 122

106

PAGE 123

107

PAGE 124

108

PAGE 125

APPENDIXB.SupplementaryFigures FigureB.1:CustomMATLABcodeappliedforbeadsat20xmagnicationwith accurateresultsforbeadsizeoutput 109

PAGE 126

FigureB.2:CustomMATLABcodeappliedforbeadsat40xmagnicationwith accurateresultsforbeadsizeoutput.Thus,thecodecanbeappliedforimagesof dierentmagnication. 110