Citation
A novel tool for in vitro bone fusion techniques through mechanical stimulation

Material Information

Title:
A novel tool for in vitro bone fusion techniques through mechanical stimulation
Creator:
Hout, Kevin Ryan ( author )
Language:
English
Physical Description:
1 electronic file (56 pages) : ;

Subjects

Subjects / Keywords:
Mesenchymal stem cells ( lcsh )
Bioreactors ( lcsh )
Biomechanics ( lcsh )
Bones -- Growth ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Orthopedic bone grafts and bone graft substitutes are currently used to facilitate bone regeneration and new bone formation in patients needing orthopedic structural support and union of bony defects (22). Bone void healing in post-operative cases is generally improved by the use of grafts, but differs from patient to patient, requiring improvements in current and future graft products. A number of different bone graft types exist, encompassing autografts, allografts, synthetic materials, and biologics; however, each possesses limitations due to variabilities with respect to bone healing. The disadvantages of these products can be attributed to necrosis of harvested bone, rejections, diseases, inflammatory responses, and structural instability (23,24). According to Transparency Market Research, the global market for bone grafts and bone graft substitutes will have an annual compound growth of 4.5%, reaching $3.48 billion in 2023 (21).
Review:
The goal of this experiment is to develop an in vitro system capable of testing these products so their osteoinductive, osteoconductive, and osteogenic properties can be enhanced at reduced cost and experimental time. If the in vitro system is successful, then it could be further implemented within in vivo and clinical applications.
Review:
Currently, no in vitro approaches to bone healing using bone grafts or bone graft substitutes have been published. However, multiple in vivo bone healing studies have shown mechanical stimulation is an important variable with respect to enhancing osteogenesis.
Review:
Using a bioreactor to apply a mechanical load, a novel system can be generated to induce osteogenesis between two revitalized-cadaveric-bone-allografts in vitro. Finding the optimal mechanical forces, time of load, and cycling for osteogenesis are the first parts in many steps when trying to configure the tools for in vitro bone healing studies.
Thesis:
Thesis (M.S.)--University of Colorado Denver.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
by Kevin ryan Hout.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
997296324 ( OCLC )
ocn997296324

Downloads

This item has the following downloads:


Full Text
A NOVEL TOOL FOR IN VITRO
BONE FUSION TECHNIQUES THROUGH MECHANICAL STIMULATION
by
KEVIN RYAN HOUT B.S., University of Kentucky, 2011 M.S., University of Kentucky, 2013
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 Master of Science Bioengineering Program
2017


This thesis for the Master of Science degree by Kevin Hout
Has been approved for the Bioengineering Program by
Kendall Hunter, Chair Vikas Patel, Advisor Karin Payne
May 13, 2017


Hout, Kevin (M.S., Bioengineering)
A Novel Tool for In vitro Bone Fusion Techniques Through Mechanical Stimulation Thesis directed by Professor Vikas Patel
ABSTRACT
Orthopedic bone grafts and bone graft substitutes are currently used to facilitate bone regeneration and new bone formation in patients needing orthopedic structural support and union of bony defects (22). Bone void healing in postoperative cases is generally improved by the use of grafts, but differs from patient to patient, requiring improvements in current and future graft products. A number of different bone graft types exist, encompassing autografts, allografts, synthetic materials, and biologies; however, each possesses limitations due to variabilities with respect to bone healing. The disadvantages of these products can be attributed to necrosis of harvested bone, rejections, diseases, inflammatory responses, and structural instability (23,24). According to Transparency Market Research, the global market for bone grafts and bone graft substitutes will have an annual compound growth of 4.5%, reaching $3.48 billion in 2023 (21).
The goal of this experiment is to develop an in vitro system capable of testing these products so their osteoinductive, osteoconductive, and osteogenic properties can be enhanced at reduced cost and experimental time. If the in vitro system is successful, then it could be further implemented within in vivo and clinical applications.
Currently, no in vitro approaches to bone healing using bone grafts or bone graft substitutes have been published. However, multiple in vivo bone healing


studies have shown mechanical stimulation is an important variable with respect to enhancing osteogenesis.
Using a bioreactor to apply a mechanical load, a novel system can be generated to induce osteogenesis between two revitalized-cadaveric-bone-allografts in vitro. Finding the optimal mechanical forces, time of load, and cycling for osteogenesis are the first parts in many steps when trying to configure the tools for in vitro bone healing studies.
The form and content of this abstract are approved. I recommend its publication.
Approved: Vikas Patel
IV


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION AND GOALS...................................1
Background...............................................3
Adaption of Bone Cells in Non-Fracture Studies.....4
In vivo Bone Fracture Healing......................8
MSC Differentiation and Osteogenic Potential through
Mechanical
Stimulation...............................................
............................11
Classifying Bone Healing through Cellular Assays..........12
Bone Healing Histology....................................13
Goals and Hypotheses......................................14
II. MATERIALS AND METHODS..........................................15
Collecting Bone Core Samples..............................15
Finding the Stress-Strain Curves for the Bone Cores.......15
Splitting the Bone Cores..................................17
Culturing Bone Marrow Mesenchymal Stems Cells.............18
Seeding Bone Marrow MSCs onto Bone Cores..................18
Live/Dead Staining of Scaffolds...........................18
Using the Bioreactor......................................19
Mechanically Loading Bone Core Samples....................21
AlamarBlue................................................23
v


ALP Activity & PicoGreen.............................24
Histology............................................26
Statistical Analysis.................................26
III. RESULTS....................................................27
Stress-Strain Curves for the Bone Cores..............27
Live/Dead Staining...................................29
alamarBlue...........................................30
ALP Activity & Pico Green............................32
Hematoxylin & Eosin and Osteocalcin Histology........35
IV. CONCLUSIONS................................................43
REFERENCES.............................................................47
APPENDIX
A. Figures.......................................................50
vi


CHAPTER I
INTRODUCTION AND GOALS
According to the 2012 Surgeon Generals Report on Bone Health and Osteoporosis, by 2020 nearly half of the American population over 50 will have weak bones and are at a higher risk for fractures. Not only is the aging populations bone health of concern, but it also places a heavy burden on the American economy with an estimated cost of $18 billion per year (1). In addition, over 3 million bone graft procedures are performed annually which require the use of autografts, allografts, and synthetic biomaterials as bone void fillers (19,20). Within these bone void fillers, there is a push to enhance bone healing in patients by incorporating biologies, which could potentially increase osteogenesis. The orthopedic markets use of biologies and bone graft substitutes has recently increased due to the advancements made within the past 20 years (23). While bone grafts, bone graft substitutes, and biologies have their advantages, there are associated negatives that can result in failures and rejections. An autograft, which is bone harvested from the same patient, IS considered the gold standard when it comes to bone grafts due to their biocompatibility with the patient. However, harvesting autologous bone tissue has limitations due to increased infection potential, bone necrosis, and limited donor supplies (23).
Another popular bone graft is the allograft, which is harvested from human cadaveric bone. Allografts have become increasingly popular due to their relative accessibility. The nature of the allografts extra-cellular-matrix increases its
1


osteogenic potential. Unfortunately, 30-60% of allografts have issues associated with failures and rejections when evaluated 10 years post insertion (23,27).
More recently, the use of biologies and bone graft substitutes have become popular tools for bone void filling. Bone morphogenic proteins (BMPs) and vascular endothelial growth factors (VEGFs) are two examples of biologies that can promote osteogenesis and angiogenesis through osteoinduction. Osteoinduction is the process in which bone forming cells are recruited to promote bone formation, typically done in conjunction with bone graft substitutes. Bone graft substitutes, like calcium phosphate, have shown to have 4-1 Ox more compressive strength when compared to cancellous bone (23). Biologies and bone graft substitutes have their advantages, but there are shortcomings for these technologies. There are limitations with respect to bone remodeling & limited shear strength (bone graft substitutes), decreased concentrations at the site of implementation & limited human studies (biologies), and cost (23).
The development of an in vitro bone healing system could aid in analyzing new bone tissue formation using various bone grafts, bone graft substitutes, and biologies. Furthermore, an in vitro bone healing system could potentially contribute to the advancements of orthopedic solutions in a non-invasive, cost reduced, and timely manner. Effectively, the broader impacts for this system would include improvements in bone grafts, synthetic bone grafts, biologies, and tissue engineering capabilities. The in vitro method developed would be able to improve bone graft modeling and help advance their implementation within in vivo animal studies and potentially enhance treatments for those needing bone graft procedures.
2


Background
Wolffs law is the basis for understanding how bone adapts due to internal changes and states: Every change in the function of a bone is accompanied by certain changes in the internal architecture and external form of the bone according to mechanical laws (2). In order to establish a bone healing process in vitro, it is important to understand the biological and mechanical principles that occur in vivo. There are biomechanical and cellular processes of bone repair that play significant roles in coordinating and regulating the healing process. With respect to the cellular process of bone repair, four types of bone cells contribute to the architecture of bone: osteoblasts, osteoclasts, osteocytes, and bone lining cells. Osteoblasts, osteocytes, and bone lining cells arise from the differentiation of mesenchymal stem cells (MSCs), while osteoclasts arise from the differentiation of hematopoietic stem cells (HSCs). The primary determinant of healthy bone is dictated by the communication between the osteocytes, osteoblasts, osteoclasts, and is defined by new bone formation (2,3). With respect to biomechanical variables, mechanotransduction is an essential bone healing process that is effectively initiated when a mechanical force is being applied within bone. Mechanotransduction is the process of converting mechanical stimulation signaling into biochemical signals, and it is essential for MSC differentiation and osteogenesis.
Combining an applied mechanical force with osteogenic differentiation media (ODM) could increase osteogenic potential with respect to MSC differentiation and new bone tissue formation. MSC differentiation in vitro can be divided into three separate stages. The first stage occurs in days 1 to 4, which shows a peak in the
3


number of cells. Secondly, Days 5 to 14 show an increase in alkaline phosphatase (ALP) activity, which is an enzyme expressed during the beginning stages of cell differentiation and osteogenic activity. Finally, days 14 to 28 show high expressions of the proteins: osteopontin and osteocalcin, which also results in decreased ALP expression (26,18). One of the main objectives for developing an in vitro bone healing system protocol is establishing a loading frequency and compressive strain that will increase ALP activity, show new tissue formation, and exhibit increased osteocalcin. Establishing these methods will allow for an understanding of how compression and frequency of loads can stimulate and affect new bone tissue formation in vitro. Once an acceptable strain and frequency are established, the mechanical stimulus will be applied using a pre-fabricated bone bio-reactor system in order to optimize the bone healing process. Finally, various cellular assays and staining techniques will be implemented to evaluate the cellular viability, ALP activity, and histological characteristics, which will assist in understanding the osteogenic processes occurring. Establishing these methods and techniques can aid in optimizing future in vitro bone healing studies.
Adaption of Bone Cells in Non-Fracture Studies
Adaption of bone will occur, in a typical individual, where activity and mechanical forces can be detected (2,6). The detection of mechanical forces and transcribing them into bone forming signals is a process known as mechanotransduction. In bone healing studies, dynamic loading is a necessary variable for osteogenesis, while also preventing suppression of bone growth. Dynamic loads generated are critical for bone formation because of their ability to
4


generate an adaptive response (8). In addition, the stress and strain generated from the mechanical forces influence cell differentiation and are mediated by cyclic loading. Fluid can only be moved through the canaliculi by cyclic loading, which makes it an important component in bone healing and bone formation (2,4).
Burr, Robling, and Turner conducted several in vivo studies with respect to biomechanical stresses on bones and how these applications affect bones responses (4,5). Their aim was to look at bone response, instead of bone healing, in order to understand the healthy response of bone. Under static loads it has been found that bone does not undergo the same mechanotransduction responses as bone that is under dynamic load. The difference in responses can be attributed to the fluid movement generated between bouts of cyclic loading and relaxation. In addition, the fluid and shear stresses from the cyclic loading act on the cilia of the osteocytes, which are responsible for initiating the beginning stages of osteogenesis. Osteogenic responses occur rapidly, but the cells require a recovery period to reestablish their mechanical sensitivity in order to communicate cellular signaling following another mechanical stimulation. Rubin and Lanyons research implied that the response of cells can occur quickly and effectively, and does not require high cyclic loads (25). Additionally, this suggested that bones may become desensitized to stimuli if all of the cycles occur at one time (7). This introduced the idea that recovery periods in between cycles are vital for cell recovery and restoring cell sensitivity to mechanical stimulation. If the cells are subjected to continuous stimulation, their recovery time may be hindered, which could lead to a decrease in osteogenesis. In their 2001 study, Burr, Robling, and Turner analyzed six different
5


sham bending test groups on rat tibia that were under various time loading regimens (5). They were separated accordingly into 0, 0.5, 1,2, 4, and 8-hour recovery groups. At the conclusion of their experiment, they were able to find that the 8-hour recovery groups experienced a significantly higher relative bone formation rate, measured by fluorescence photomicrographs, in the loaded limb than the non-loaded limb. Overall, the 8-hour recovery group had a 125% higher bone formation rate than the 0-hour groups and was also 102% higher than the 0.5-hour groups. These findings helped to assert that recovery time was needed to restore cell sensitivity to mechanical stimuli. Furthermore, they also looked into the effects that time loading between cycles had on bone formation. The time intervals of interest were separated into 0.5, 3.5, 7, and 14-second groups, which experienced four bouts of 90 cycles/bout. Relative mineralization-surface and bone formation rates, measured by histomorphometry, were significantly higher in the 14-second groups than in the other groups (p<0.001). This indicated that a high recovery period in combination with a high interval between bouts could generate the most effective form of osteogenesis with respect to new bone formation (5).
Using this data, an analysis of loading bouts, cycles, and rest periods was conducted to understand the effects that dynamic loading in combination with recovery time has on osteogenesis. A 4-point bending model applied a dynamic load with given bouts, cycles, and recovery periods. Four interest groups were used to assess how cyclic loading has an effect on bone formation rates. The four groups were separated into 360 cycles x 1 bout/day, 60 cycles x 6 bouts/day (2hr recovery), 90 cycles x 4 bouts/day (3hr recovery), and 180 cycles x 2 bouts/day (6hr recovery).
6


A positive osteogenic effect was found for the groups with recovery periods between 2 & 3 hours (60x6 and 90x4 groups) and showed approximately 80% higher bone formation rates than the 360 cycles/day group. This further developed the idea that managing recovery periods can increase the effectiveness of loading and osteogenesis. In the same study, the model was used to determine the optimal recovery time needed between each bout. Using the 90 cycles x 4 bouts group, loading intervals between bouts were separated in 0, 0.5, 1,2,4, and 8 hour groups. Again, the recovery time between bouts proved to be the most effective at exhibiting the greatest bone formation rate. The 4 and 8 hour groups had significantly greater relative bone formation rates than the group that was allowed no recovery time. This data suggested that the sensitivity of the cells was re-established in between 4 to 8-hour recovery periods. Furthermore, the study continued to analyze time intervals between cycles for 0.5, 3.5, 7, and 14-second groups at 36 cycles/day. As expected, the 14-second group was significantly higher, with respect to bone formation rates, than the other three groups, which proved to show recovery intervals are vital for producing osteogenic responses (4).
These two studies indicated that recovery periods are needed in order to optimize osteogenesis with respect to relative mineralization and bone formation rates (4,5). Establishing recovery periods could play a significant role in cellular communication, differentiation, and gene expression with respect to osteogenesis. Bone cells can adapt quickly to the environments they are subjected to, which is why it is important to manage the mechanical stimuli and avoid de-sensitization.
7


In vivo Bone Fracture Healing
Several in vivo animal studies including mice and rabbits have been modeled to show how bone healing is affected by mechanical loading compared to control groups without loading (10,11,12). Mechanical loadings in each study were measured based on specific amplitudes, frequencies, strains, and/or cycles. In each of these studies, the animals tibias were osteotomized and subjected to various compressive forces. It should be noted that producing an osteotomy instead of a real fracture could have resulted in biased data with respect to the healing process {1,9). The main approaches to analyzing bone healing in these studies were represented through radiographs and bending tests. Variables assessed within these approaches included callus volumes, osteoid presence, mineralization, and mechanical properties post-osteotomy.
Michael Gardner, et al. have produced an extensive amount of in vivo research with respect to bone healing. Similar to the studies that looked at the mechanical response of cells in healthy bones, Gardner et al. established a modeling system and incorporated mechanical stress effects on bone (10,12). Their modeling systems looked into the failure moments and callus volumes of osteotomized mice tibias after a specific load and amplitude were applied. In their first study, they analyzed the effects delayed loading post osteotomy has on the in vivo model. Five sample groups had an applied axial compression in addition to a delay and non-delay: 0.5N/0-day delay, 1 N/O-day delay, 0.5N/4-day delay, 1 N/4-day delay, 2N/4-daydelay. These five sample groups were compared to a control group, which was not subjected to mechanical stimulations. Each group was analyzed after
8


a cyclic load had been applied for 2-weeks. The results indicated that the callus strength and bending stiffness improved with a 0.5N/4-day delay when compared to the control group and other delayed groups. In addition, the groups that did not receive a delay (Od) showed a 68% decrease in callus strength even though their callus volumes were significantly larger. This indicated callus volume cannot be a direct indication of bone strength, and should not be considered when analyzing bone healing. Although the callus volumes were significantly higher in the 0-day delay groups, the 4-day delay groups produced greater failure moments overall (12). With respect to delayed onset and establishing recovery periods, bone healing mechanisms perform consistently with the information found in the non-fracture bone studies.
A study following the results of their previous analysis was conducted in 2008 by Michael Gardner, et al., which analyzed the effects of bone healing in vivo through pause-inserted-cyclic-loading (10). Eighty mice underwent tibia osteotomies and were separated into four groups based on loading procedures. Three cyclic loading groups had a 0.5N force with a 1N amplitude applied, while the control group did not experience a load. The three cyclic loading groups were subjected to a repetitive load (100 cycles for 100 seconds), 9-second pause inserted- time equivalent load (100 seconds), or 9-second pause inserted-cyclic equivalent load (1000-seconds). Loading was applied for two weeks post-osteotomy, which was followed by the bone healing analyses. The percent osteoid and mineralization were analyzed using micro-CT and histology. Overall, the loaded groups produced a larger osteoid and percent mineralization than the control group. However, the callus
9


volume in the control group was larger than the loaded groups. Although callus volume was larger in the control group, the failure moment in the loaded groups was greater, which is consistent with their previous findings. Again, callus volume does not directly correlate to the strength of healed bone. When the four groups were compared based on stiffness and maximum bending moments, the pause inserted-cycle-equivalent group (100-seconds) showed an overall advantage amongst the other groups (10). Again, a trend could be seen within groups that have an applied pause/rest period between bouts and/or compressions. This indicates that bone is healed optimally when cells can re-establish their signaling sensitivity through a given rest period.
Similar to the previous study, a 2009 study conducted by Shadmehr, et al. assessed bone healing through frequency and amplitudes of specific strains (11). The right tibias of 16 rabbits were osteotomized and an external fixator was used to apply specific frequencies and amplitudes. Four groups were formulated based on the durations of mechanical stimulations and their respective controls. The durations of mechanical stimulations continued for 1 and 2 weeks, in which a 1 Hz (1 cycle per second) frequency was applied 15 minutes per day. In addition, the stimulation sessions increased their exerted forces each day in a somewhat linear progression. After the mechanical stimulations and analyses were complete, the radiological parameters and mechanical evaluations were significantly higher in the 1 and 2 week groups compared to the respective control groups. In addition, the 2-week stimulated group had higher overall values for callus volume and stiffness when compared to the 1 -week group. This study also showed greater callus volume
10


area within the cyclic loaded groups than the previous study, which can reflect some variance based on species (11). In addition, this study further expanded upon the idea that inserting a dynamic load to a healing bone can optimally enhance the bone strength.
All three studies and their respective data proved to be consistent with one-another. It is important to find a non-destructive axial load and strain in order to optimize the bone strength and bone formation. In the first study, the 4N and 2N amplitudes may have been too great, which may have contributed to a decrease in callus volume and failure moments. The modulus of elasticity of bone for different animals will vary, which is why it is important to apply the correct load for a specific strain of interest. In addition to applying mechanical compression, allowing a pause in between stimulations has shown to provide an optimal environment for enhancing both fracture and non-fracture bone. Allowing a sufficient recovery period for the cells is an important process in the re-organization, growth, and strength of healthy bone.
MSC Differentiation and Osteogenic Potential through Mechanical Stimulation
MSCs play an integral part in osteogenesis as they are the progenitors to osteoblasts, osteocytes, and bone lining cells. Stimulating the MSCs can maximize their osteogenic potential by helping to initiate their differentiation into their cell lineages. Osteogenic markers and new bone matrix have been shown to increase within in vitro MSC-seeded-scaffolds where cyclic loading is applied (28,29,30). In addition, compressive loading of MSC-seeded-polymer scaffolds upregulated bone matrix forming genes, which in turn increased bone-like mineralization. Furthermore,
11


only 2-hours of loading per day for 5-days was needed to differentiate MSCs, which outperformed continuously loaded groups by approximately 50% with respect to calcium deposition (28). It is unlikely MSCs experienced direct compression in these studies, but it is more likely that they experienced the secondary effects of the mechanical stimulation through fluid flow and scaffold bending (28). The mechanical applications for initiating MSC differentiation in vitro are important, which re-iterates the notion that loading cycles are needed for optimizing osteogenesis. Although this study did not address osteogenesis through the use of bone, their polymer applications are important when considering MSC differentiation in bone scaffolds. Classifying Bone Healing through Cellular Assays
Osteogenic biomarker assays can help to measure osteogenic differentiation of MSCs, which has been used in publications and studies by Birmingham, et al (18). According to Birmingham, et al. there are three stages of osteogenesis in vitro. The first stage of osteogenesis in vitro occurs within the first four days, which will show a maximum population of MSCs. Following a peak in MSC population, the second stage begins between days five and fourteen. Within this stage, early osteogenic cell differentiation begins and can be characterized by a rise in ALP. The third and final stage of osteogenesis in vitro occurs between days 14 and 28, which expresses a high amount of osteocalcin and osteopontin. Calcium phosphate deposition begins in the later portion of stage three following the rise in osteocalcin and osteopontin. Transcription and protein expression of ALP is an important biomarker in the osteogenic differentiation of MSCs, which can be adequately quantified by ALP activity assays (18).
12


Bone Healing Histology
Identifying and defining the bone healing process through histology is a crucial step in assessing studies experimental methodologies and procedures.
Using various stainings on bone can assist in the evaluations and understandings of the processes that are taking place at the time of the experiment. A couple of modalities that can be used to classify bone structure, new tissue formation, and relative osteocalcin present are Hematoxylin & Eosin (H&E) and Osteocalcin Immuno-Histo-Chemistry (IHC) (13).
H&E stainings give a general overview of the new and old tissue present within the bone. This method in particular helps to delineate bone, cartilage, and connective tissue based on the color intensity of the staining. Although this is described as a general staining procedure, H&E stainings generate a working knowledge of the tissue structure and possibly new bone formation. In addition to HE staining, IHC can provide supportive information about the development of bone tissue with respect to the development of in vitro osteogenesis. In specific, osteocalcin staining can be performed as it pertains to a late stage marker for osteogenesis. As previously mentioned, high levels of osteocalcin are expressed in vitro between days 14 and 28, which is of particular interest due to the 28-day loading regimens applied within this study (18). The inclusion of H&E and IHC stainings are crucial in providing a novel understanding of the healing processes, which may be occurring during the loading.
13


Goals and Hypotheses
The goal of this study is to provide basic tools for approaching in vitro bone healing through mechanical loading applications. By applying a mechanical load to bone cores placed in ODM and seeded with MSCs, this could conceivably optimize osteogenesis. This study attempts to compare and contrast loaded versus unloaded seeded-bone cores through various assays and stainings, which could provide further knowledge for adjusting future in vitro experiments.
Classifying both early and late osteogenesis will be needed to analyze the two groups of interests, which can be conducted through cellular assays and histology. Based on previous in vivo and in vitro osteogenesis research and the mechanical processes contained within these studies, it is hypothesized the loaded groups will have more osteogenic potential when compared with the unloaded group.
14


CHAPTER II
MATERIALS AND METHODS Collecting Bone Core Samples
The following bone core sampling procedure was adapted from Effects of Initial Seeding Density and Fluid Perfusion Rate on Formation of Tissue Engineered Bone by Grayson, Warren L. et al. (14). The bone cores, each approximately 4.8mm in width and 10mm in height, were obtained from the femoral head of cadaveric human bone using a coring tool developed by Todd Baldini in the Department of Orthopedics. The cadaver was a 55-year-old white male. The core samples were then cleansed under high pressure steamed water for one hour to remove bone marrow. After the cleansing process, the cores were then washed in cleaning solutions of 0.1 % EDTA in PBS for one hour at room temperature, a hypotonic buffer of 0.1 % EDTA in 10 mM Tris overnight at 4C, detergent comprised of 0.5% SDS in 10mM Tris for 24 hours at room temperature, and an enzymatic solution of 50 U/ml DNase, 1 U/ml RNase in 10mM Tris for 3-6 hours at 37C. The washed samples were then freeze-dried, and sterilized in 70% ethanol for one hour at room temperature in a ventilated hood. This process prepared the bone cores to be conditioned in osteogenic media overnight before seeding with cells.
Finding the Stress-Strain Curves for the Bone Cores
Configuring the stress-strain curves was an important process in finding the ultimate strength of the collected bone cores. From this information, a strain value (mm/mm) at max stress was found by correlating its value to the ultimate strength (N/mm2) for the bone core of interest. This strain value was used to understand if the
15


custom-built compression bioreactor, provided by Dr. Stephanie Bryants Chemical and Biological Engineering Lab, could apply the needed strain to bone cores of interest. Overloading the bioreactor was a preliminary concern due to the relative stiffness of the bone compared with the hydrogels the bioreactor typically handles. The following procedure was used to generate the data for the stress-strain curves.
Twenty bone cores were selected to be analyzed on a universal testing machine (UTM) provided by Todd Baldinis Biomechanics Lab. Before testing the bone cores individually, the lengths (mm), widths (mm), and weights (g) were measured in order to generate their respective stress-strain curves. Each bone core was found to have varying volumes (mm3), weights (g), and therefore densities (g/mm3). The densities ranged from 4.08e-4 to 1,19e-3 g/mm3, which was relatively large, but was expected due to the variance in bone properties during the collection process. The specimen properties and stress-strain values can be seen in Table 3.1 and Table 2.2, respectively. After analyzing, measuring, and applying the testing parameters, only 17 bone cores could be used in the final analysis due to some errors in loading and data collection. The specimens highlighted in red in Tables 3.1 and 3.2 were not used in the final analysis.
Using the UTM, a linear displacement of 1 mm/sec was applied to each bone core, recording the information at 20Hz, which was then output into a text file. Strain (mm/mm) and stress (N/mm2) values were generated into two columns and were used to format the stress strain curves. Noisy data was removed from the beginning of each test, which is the subject of the UTM administering the beginning stages of the compression. The data was re-formatted to reflect these changes, and
16


the stress-strain data for each bone core began where the stress value was zero. If a zero value for stress was not found, then the first negative stress integer and its respective strain value were used. Following the formatting of each bone core loading data, the ultimate strengths and their associated strain values were calculated. Using these values, understanding the proper utilization of the bioreactor, and taking into account the most suitable mechanical environment to be established, an applied strain for the loaded bone cores was generated. Reviewing the literature and having discussions with Dr. Stephanie Bryant, Dr. Vikas Patel, and Dr. Karin Payne, a 0.1% (10 micron) strain was used for the mechanical loading. This strain was applicable to generate enough mechanical stimuli to optimize osteogenesis in vitro.
Splitting the Bone Cores
In order to properly model bone healing in vitro, the bone cores needed to be split adequately in half so they could be aligned when the loading was established. A Buehler Isomet 1000 Precision Saw, provided by Dr. Clifton Carey with the School of Dental Medicine, was used to cut the bone cores and evenly split the specimen into two bone core scaffolds. The process was generated by gently placing each 4.8mm x 10mm bone core medially onto the mechanical saw rotating at approximately 100 rpm. The mechanical saw then cut through each bone core at a moderate pace until full separation had been initiated. After the bone cores had been adequately cut, they were sterilized in 70% ethanol for 24 hours and dried in a cell culture hood for another 24 hours. Once completely dry, the cores were then placed in 20C until they could be seeded.
17


Culturing Bone Marrow Mesenchymal Stems Cells
The MSCs were purchased from Texas A&M University College of Medicine Institute for Regenerative Medicine, where they were harvested from human bone marrow of a 22-year-old male patient. Cells were removed from cryopreservation, thawed, plated in a T-225 flask (225 cm2), and cultured in 30 ml_ of complete culture media (CCM, MEM-alpha (Life Technologies, Carlsbad, CA, USA)) supplemented with 16.5% FBS (Atlanta biologicals, Lawrenceville, GA, USA), 2 mM L-glutamine (Corning, Manassas, VA, USA), and 1% penicillin streptomycin (Thermo Scientific, Logan, UT, USA). Cells were then incubated at 37C and 5% CO2 until they were confluent. After the cells had reached confluency, they were then passaged, split, and plated at a density of 100 cells/cm2 The CCM was changed every 48-72 hours until an adequate number of cells could be used to seed at 1.2x106 cells per bone core scaffold.
Seeding Bone Marrow MSCs onto Bone Cores
The MSCs in CCM were suspended at passage 4 and aliquoted onto a blot-dried bone scaffold using a 10-100 pL pipette. Every 15 minutes for 1 hour, each scaffold was rotated 90 and 5pL of osteogenic differentiation media was added directly to the scaffold. This process aided in facilitating uniform distribution of cells within each scaffold. The scaffolds were then incubated for 24 hours at 37C and 5% CO2 before mechanical loading was applied.
Live/Dead Staining of Scaffolds
Live/dead staining was performed to validate cell attachment within the scaffolds using Life Technologies LIVE/DEAD Cell Viability/Cytotoxicity Kit. Each
18


scaffold was washed with 1X PBS before performing the assay. 2ml_ of the LIVE/DEAD assay reagents were applied to each scaffold with the lights off due to the sensitivity of the staining. The scaffolds were then incubated for 45 minutes at room temperature and protected from light by wrapping each plate in foil. Following incubation, the stain was aspirated and each scaffold was washed with 1X PBS three times. The scaffolds were then transferred to 35 mm petri dishes and covered with 1X PBS before being analyzed. Each sample was analyzed using a Nikon confocal microscopy with the settings EGFP: FIV55, offset 2, 488 2.00; ALX 594: FIV 80, offset 2, 561 2.00; 1024 px size, 5.3 px dwell, 10X; Z-stacks of 10pm steps; filters: turret.
Using the Bioreactor
The custom bioreactor, provided by Dr. Stephanie Bryant, uses a specific program developed by a former student to apply specific strains, with given time intervals, to scaffolds of interest (15). This in-house program communicates with a bioreactor-controller and the bioreactor, which enables specific input frequencies, input strains, scaffold heights, and time intervals. Figure 2.1 shows an overall schematic of the bioreactor and its various components. The loading pins, which rest on top of the scaffolds, are located above the 24-well plate and are connected to a loading platform. This loading platform is connected to a stepper motor, which is controlled by the program via wired connections. The stepper motor is the component of the machine that applies the specific strain, which is applied equally across all scaffolds due to the loading pin and loading platform connection. The
19


entire bioreactor and its connections were placed in an incubator, which is where the loading regimens occurred.
Figure 2.1 Bioreactor and its various components.
Figure 2.2 shows an excel spreadsheet that is used in conjunction with the program to specify the given parameters of the loading regimen. The figure does not represent the actual values used within the experiments, but it was formatted to reflect the loading regimens applied, which is addressed within the Mechanically Loading Bone Core Samples section.
20


Figure 2.2 Excel Spreadsheet that is used to input values and parameters for a
given loading cycle.
Mechanically Loading Bone Core Samples
Similar techniques for the bone loading applications were adapted from Villanueva et al. (15). Using the custom bioreactor, a specific strain of 0.1% (10 microns) was applied to the bone cores of interest. In a 24-well non-tissue culture plastic plate, each well encompassed two seeded bone core scaffolds with BM-MSCs, which were cut from the same specimen. Stabilization was initiated using PYREX Cloning Cylinders, which provided a casted enclosure for the scaffolds. Furthermore, the cylinders would allow flow of media through the top and bottom of each scaffold due to their hollow design.
The scaffolds were organized into loaded and non-loaded groups within each row of the 24-well plate; each well was sequentially numbered to avoid confusion.
21


Each well received 2 mL of osteogenic differentiation media (ODM, CCM supplemented with 10 nM dexamethasone (Sigma, St. Louis, MO, USA)), 20 mM (3-glycerophosphate (Sigma, St. Louis, MO, USA), and 50 pM L-ascorbic acid 2-phosphate (Fisher Scientific, Fair Lawn, NJ, USA)). The bioreactor and cores were incubated at 37C and 5% CO2, and the ODM was changed every 48-72 hours. The seeded cores were divided into three groups of interests: un-loaded cores with ODM at 28 days, loaded cores with ODM at 28 days, and cores at Day 0.
The loading regimens for the loaded groups experienced a 0.1% strain (10 microns) at 1 Hz, which was then followed by a 9 second pause. This process was repeated for 90 cycles, once per day, for 28 days within the first two runs, which equated to a loading time of 15 minutes per day. The third run experienced the same strain and frequency, but the regimen increased to a total of 360 cycles per day for 28 days, which equated to 60 minutes of loading. Figure 2.3 below depicts the loading cycle applied to each scaffold, and was adapted from Gardner, et. al. (2). After the 28-day loading period was complete, the samples were then removed from the bioreactor, processed, and analyzed based on their designated assay.
Load (N)
0.5 -
/V
Nina tat ond raft par Id d
H----1----1---1---1----I---I---1----1---1---1----r
Figure 2.3 Loading cycle with 1 Hz load and 9 second pause applied to bone scaffolds. Note: the loading regimens were not 0.5 to 1 N, which is reflected in this figure. Instead the loading was 0 to 0.1 % strain.
22


AlamarBlue
The alamarBlue cell viability assay was performed using Thermo Scientifics AlamarBlue Cell Viability Reagent. This reagent is a redox indicator that yields a colorimetric change and a fluorescent signal in response to metabolic activity. The redox indicator detects oxidation during cellular respiration, which is a process that releases waste products due to ATP conversion. The amount of fluorescence is proportional to the metabolic activity of the cells. Therefore, cells not proliferating or decreased activity will have a lower fluorescence (16). Detection of the dye was performed using a Promega Microplate Fluorometer, which measured excitation at 530-560 nm and emission at 590 nm.
A positive control was generated by autoclaving 5 ml_ ODM medium and 500 pL alamarBlue reagent in a glass bottle. Once autoclaved, the positive control was added to one well of a controls plate with 100 pL of ultrapure water. For the negative control, 1 ml_ of ODM only was added within the controls plate. The bone cores received 1 ml_ of fresh ODM media and 100 pL of alamarBlue reagent was added to each well. The cores and controls were then incubated for 2.5 hours at 37C and 5% CO2. 100 pL of each sample were added in triplicate to a 96-well plate and the plate was wrapped in foil due to the light sensitivity of this assay. The plate was then taken to the Promega Microplate Fluorometer where the fluorescene readings were generated. This process was repeated at days 0 and 28 of the experiment to analyze cell viability between pre-loaded, un-loaded, and loaded bone cores.
23


ALP Activity & PicoGreen
The ALP Activity & PicoGreen assays were performed using Sigma-Aldrichs SIGMAFAST p-Nitrophenyl phosphate (pNPP) Tablets and DNA quantification kit. The pNPP tablets are soluble substrates used for the detection of ALP activity, which occurs when the pNPP reacts with ALP and generates a yellow para-nitrophenol and is typically measured in umol pNPP/mL (17).
ODM was aspirated from each well containing the loaded bone cores and were then sequentially washed with 1X PBS until the media from each core was removed. The samples were then left at 4C for 30 minutes and transferred to 1.5 mL tubes. Following the transfer, 750 pL of lysis buffer (0.1% Triton X-100 in distilled water) was added to each 1.5 mL tube. The samples were then frozen at -20C until the assay was initiated. The samples were removed, thawed, and centrifuged once the ALP Activity assay was ready to be performed. The pNPP solution was prepared using one pNPP tablet and one Tris tablet, which were added to 20 mL of distilled water and vortexed until dissolved. A 96-well plate was then prepared by adding 100 pL of pNPP solution, 90 pL of distilled water, and 10 pL of the sample. Each sample was plated in triplicate. In addition, a triplicate of lysis buffer alone was used to calculate the background value. Once the sample plating was complete, the 96-well plate was wrapped in foil and incubated for 30 minutes at room temperature. Optical density readings were taken by a BioTek microplate reader. The following equation was used to calculate the ALP Activity for each sample:
ALP =
(ODblank from cell lysis buffer) x (total volume)x (dilution) 18.45 x (sample volume)
= umol pNPP/mL
24


After the samples ALP Activities were quantified, a PicoGreen Assay was performed to normalize these values to their DNA content. PicoGreen is a fluorescent nucleic acid stain for quantifying dsDNA (31). The Quant-iT PicoGreen dsDNA Assay Kit (Life Technologies #P7589) was used to perform this assay. First, a 1X TE buffer was prepared by diluting 1:20 of 20X stock TE buffer in DNase-free water. Following the TE buffer preparation, a standard curve was prepared by taking 2 pg/mL of working solution from 100 pg/mL of stock solution. Afterwards, serial dilutions were used to create curve concentrations according to the manufacturers instructions.
Following the curve concentration and serial dilution process, the bone core samples were prepared in a 1:10 sample dilution by adding 25 pL of the bone core samples to 225 pL of diluent. 100 pL of each sample were added in duplicate to the 96-well plate. Next, 100 pL of PicoGreen working solution was added to each well containing the curve concentrations and samples. The 96-well plate was wrapped in foil, incubated for 5 minutes, and read on the Promega Microplate Fluorometer. The readings from the curve concentrations generated a standard curve and produced a linear equation. The PicoGreen data was calculated by taking the samples fluorescence readings, inputting them into the linear equation, and multiplying this value by its dilution ratio of 10. These values, which were measured in ng/mL, were used to normalize the ALP Activity to its DNA content by dividing the calculated ALP Activity by their respective PicoGreen values. The normalized ALP Activity values were measured in nmol pNPP/mg.
25


Histology
The histology and immunohistochemistry methods were adapted from Comprehensive histological evaluation of bone implants by Rentsch et al. (13). The bone cores were fixed in formalin for 5 days, decalcified in Immunocal for approximately 3 weeks, processed by a Leica TP 1020 Tissue Processor, and mounted in cassettes using paraffin. 5 pm sections were cut along the longitudinal plane using a Microm HM 355 S rotary microtome. The 5 pm tissue sections were placed on 25x75x1 mm micro slides (VWR) following hydration in a tissue flotation water bath (VWR). The micro slides were then dried on a hotplate (Leica) at 45C for approximately two hours. Following the histological sectioning, the slides were stained with H&E and osteocalcin. H&E stainings for tissue evaluation were adapted from the Payne Regenerative Orthopedics lab. Osteocalcin stainings were performed based on the protocols from abcams ab198228 and ab6721 antibodies.
In addition, a DAB HRP substrate (VectorLabs) was used within the osteocalcin staining for antigen labeling, which produced a brown color. After each staining protocol was performed, the stained tissue samples were then evaluated and imaged using a Nikon H600 L microscope.
Statistical Analysis
A one-way ANOVA and Tukey post hoc analysis was used. All analyses were performed with SigmaPlot v.11.2. Significant differences are reported with a p-value < 0.05.
26


CHAPTER III
RESULTS
Stress-Strain Curves for the Bone Cores
Figure 3.1 shows the stress-strain curve for Core 5, which was typical across all tested bone cores. In this example, the max stress was 1.9631 N/mm2, while the highest strain at the max stress was 0.04623 mm/mm. The stress-strain values were used to generate a range of data for the group of tested bone cores. The median, minimum, and maximum values of the strains-at-maximum-stress (mm/mm) were 0.0388, 0.0335, and 0.0874, respectively. In addition, tables 3.1 and 3.2 exhibit the overall testing profiles of the bone cores, which show the calculated densities, maximum strains, and maximum stress values. A correlation could not be determined between the densities and the strain-at-max stress or max-stress values.
The minimum value of the strains-at-max-stress was used to determine if the bioreactor could handle this type of loading due to the stiffness of the material. After discussions and reviewing the literature, it was determined that a strain of 0.1 % (10 microns) should be used. This was estimated to generate a sufficient load on the bone cores, yet it would not be significant enough to compromise the bioreactor. In each of the runs conducted throughout this experiment, the strain of 0.1% (10 microns) remained constant.
27


Core 5
0 0.02 0.04 0.06 0.08 0.1
Strain (mm/mm)
Figure 3.1 Example Stress-Strain Curve Generated from UTM data
Table 3.1 Dimensions of bone core specimens
Specimen length (mm) width (mm) X-Area (mmA2) Weight (g) Density (g/mmA3)
1 9.59 4.79 18.020254 0.20 1.16E-03
2 10.73 4.46 15.62282611 0.20 1.19E-03
3 9.56 4.82 18.24668429 0.22 1.26E-03
4 10.22 4.67 17.12867001 0.17 9.71 E-04
5 9.96 5.15 20.83072279 0.12 5.78E-04
6 9.89 4.56 16.33125525 0.16 9.91 E-04
7 9.9 4.72 17.49741444 0.09 5.20E-04
8 10.05 4.8 18.09557368 0.14 7.70E-04
9 10.27 4.9 18.8574099 0.21 1.08E-03
10 9.63 5.09 20.34817416 0.08 4.08E-04
11 9.93 4.97 19.40004149 0.18 9.34E-04
12 9.02 5.18 21.07411768 0.12 6.31 E-04
13 10.67 4.68 17.20210473 0.13 7.08E-04
14 10.11 5.15 20.83072279 0.15 7.12E-04
15 9.86 4.52 16.04599864 0.09 5.69E-04
16 9.48 5.11 20.50839538 0.14 7.20E-04
17 9.2 5.04 19.95036999 0.16 8.72E-04
18 9.65 5.19 21.15556347 0.09 4.41 E-04
19 9.86 4.83 18.32247521 0.17 9.41 E-04
20 9.46 4.97 19.40004149 0.17 9.26E-04
28


Table 3.2 Max strain and max stress values of each bone core specimen
Specimen Strain @ Max Stress (mm/mm) Max Stress (N/mm2)
1 0.036861314 0.541923549
2 0.042050326 2.640988238
3 0.056433054 6.850554216
4 0.087377691 0.791059668
5 0.0462249 1.963133993
6 0.036683519 4.253065605
7 0.037585859 1.653421429
8 0.038875622 3.696732757
9 0.043466407 5.741843687
10 0.037673936 3.869403682
11 0.033726083 7.745777247
12 0.047949002 7.628623055
13 0.038791003 5.86859001
14 0.056587537 3.826645902
15 0.050603363 0.521546953
16 0.037774262 3.023722668
17 0.037913043 5.421162618
18 0.038082902 4.356442698
19 0.033488844 9.866898325
20 0.060475687 6.764186563
Live/Dead Staining
Live/dead staining was performed on the cores to see if cell attachment had occurred after harvesting and seeding the cores with the MSCs. Figure 3.2 depicts the imaging generated from the confocal microscopy. The green fluorescence dictates the live cells while the red fluorescence indicates dead cells within each bone scaffold. As Figure 3.2 shows, from a qualitative approach, the scaffold had relatively optimal attachment and the scaffolds primarily contained live cells.
29


Figure 3.2 2D Live/Dead Imaging of seeded bone scaffolds; green (live) red (dead). Images are of two different seeded bone cores.
alamarBlue
The alamarBlue assay was conducted for the 2nd and 3rd runs of the experiment, which aided in understanding the overall cell viability and cell proliferation before and after the loading regimens. The loading regimen of the 2nd run, which was loaded for 15 minutes per day, differed from the loading regimen of the 3rd run, which was loaded for 60 minutes per day.
Within the 2nd run, the overall cell viability, based on percent reduction, was higher within the Day 0 group (35.44% 1.59%) than Day 28 Unloaded (12.55%
1.27%) and Day 28 Loaded (10.43% 1.72%) groups. The Day 0 group was significantly higher (P<0.05) when compared with the loaded and unloaded groups, but there was no statistical significance between the loaded and unloaded groups. Although no statistical significance could be shown, the Day 28 Unloaded groups
30


overall cell viability was higher than the Day 28 Loaded group. Figure 3.3 shows the averages and standard errors between each group.
alamarBlue (2nd Run)
40%
35%
30%
O 25% I
u
=> 20% O
LU
= 15% 10% 5% 0%
Day 0
Day 28 Unloaded
Day 28 Loaded
Figure 3.3- alamar Blue percent reduction readings for Day 0 seeded, Unloaded Day 28 seeded, and Loaded Day 28 seeded cores from Run 2 (*P<0.05)
The alamarBlue of the 3rd run of the experiment produced similar results to
the 2nd run. Again, the Day 0 group had more cell viability with respect to percent
reduction (48.77% 4.61%) and was significantly higher (P<0.05) when compared
with the loaded and unloaded groups. However in this run, the Day 28 loaded group (17.04% 3.73%) had greater cell viability with respect to percent reduction than the
Day 28 Unloaded group (11.11% + 0.70%), but no statistical significance could be generated between these two groups. Figure 3.4 shows the averages and standard errors between each group.
31


60%
50%
c 40% o
"-M
£ 30%
(D
cn
^ 20% 10%
0%
alamarBlue (3rd Run)
Day 0
Day 28 Unloaded
Day 28 Loaded
Figure 3.4- alamarBlue fluorescence readings for Day 0 seeded, Unloaded Day 28 seeded, and Loaded Day 28 seeded cores from Run 3 (*P<0.05)
ALP Activity & Pico Green
ALP and Pico Green Assays were performed after the 1st and 2nd runs of the experiment, which experienced loading for 15 minutes per day (figures 3.5 and 3.6). The results of this assay aided in developing an understanding if the cores were experiencing early osteogenic cell differentiation through normalized ALP activity. The normalized ALP activity was generated for each group, which was derived from the 1st runs ALP activity and Pico Green assays. The data, seen in figure 3.5c, shows the Day 28 Loaded group (6.32e4 9.81 e3 nmol pNPP/mg) had the highest overall normalized ALP activity between the three measured groups. This normalized ALP activity analysis could point to a potential advantage with respect to osteogenic cell differentiation within the loaded group. Although there was variance between the normalized ALP activities, no statistical significance could be generated between the three groups.
32


a)
b)
ALP activity
0.40
0.30
___I
£ 1
S: o.2o
Q.
z
- 0.10
o I
£
3 0.00
DayO Day 28 Unloaded Day 28 Loaded
Pico green
8,000
7.000
6.000
5.000
4.000 ao
c 3,000
2.000 I
1.000 I 0
DayO Day 28 Unloaded Day 28 Loaded
_ Normalized ALP Activity
CUD
c) ^ 80,000
CL
Q- 60,000
o
£ 40,000
<
q 20,000
a2
< 0
DayO Day 28 Unloaded Day 28 Loaded
Figure 3.5- ALP (a), Pico Green (b), and Normalized ALP activity (c) from Run 1 The 2nd run produced similar results (see Figure 3.6), however due to issues with cell expansion, only Day 28 Unloaded and Day 28 Loaded groups could be generated within the ALP activity and Pico Green assays. Using the same process, the normalized ALP activity was calculated for the 2nd run. Again, the averages of the Day 28 Loaded group (5.62e4 8.57e3 nmol pNPP/mg) produced a higher
33


overall normalized ALP activity when compared with the Day 28 Unloaded group (4.17e4 2.99e3 nmol pNPP/mg). This could further lead to an assumption that osteogenic cell differentiation is occurring at a more pronounced rate within the loaded group than the unloaded group. Although it can be assumed this is occurring, no statistical significance could be generated between the two groups.
a)
b)
Normalized ALP
80,000 E
^ 60,000 o. '
Z
40,000 o '
E
S 20,000 < z
Q 0
CL
_l
<
I
I
Day 28 Unloaded Day 28 Loaded
Pico Green
1,200 1,000 800 -E 600
ClO
c 400 200 0
I
Day 28 Unloaded
I
Day 28 Loaded
C)
CuO
Q.
Q.
O
E
C
o
Q. _I
<
80,000
60,000
40.000
20.000 0
Normalized ALP
I
I
Day 28 Unloaded Day 28 Loaded
Figure 3.6- ALP (a), Pico Green (b), and Normalized ALP activity (c) from Run 2
34


Hematoxylin & Eosin and Osteocalcin Histology
The H&E stainings gave indications of new tissue development after 28 days between the Loaded and Unloaded groups, which are depicted in figures 3.7 and 3.8 respectively. Within the Loaded group, new tissue can be seen at both the intersection-site and along the outside of the scaffolds. Figures 3.7a and 3.7c are representative of the 15-minute Loaded group, while 3.7e and 3.7g are representative of the 60-minute Loaded group. An overall depiction of the Loaded scaffolds at 2x zoom is given within these figures. To further understand the new tissue development, 4x images were taken to enhance the resolution at the intersections of the scaffolds, which can be seen in figures 3.7b & 3.7d (15-minute Loaded group) and 3.7f & 3.7h (60-minute Loaded group). Qualitatively, it appears that slightly more tissue developed along the outside of the scaffolds within the 15-minute Loaded group than the 60-minute loaded group. Although new tissue prevalence was greater overall in the 15-minute Loaded group, both groups exhibited a relatively noteworthy amount at the points of intersection. This could potentially validate the possibility of more pronounced osteogenic development within mechanically stimulated environments.
35


36


Figure 3.7- H&E stainings of the 15-minute Loaded at 2x (a,c) & 4x (b,d). H&E stainings of 60-minute loading at 2x (e,f) & 4x (g,h). 4x images show the area of
intersection between the two cores.
37


Similar to Figure 3.7, Figure 3.8 gives an overall depiction of the Unloaded group at 2x (3.8a & 3.8c) and 4x zoom (3.8b & 3.8d). Based on the FI&E stainings between the Loaded and Unloaded groups, from a strict observation standpoint, the Unloaded group did not show as much new tissue. When compared with the 15-minute and 60-minute Loaded groups, it appears that there was slightly less tissue developed at the points of intersection in the Unloaded group. In addition, it could also be proposed that more tissue developed along the edges of the scaffolds within the 15-minute Loaded group than the Unloaded group. This observational analysis could further validate the assumption that less tissue, and therefore less potential induced osteogenesis was produced within the Unloaded scaffolds. The assumptions of increased osteogenic potential within the Loaded groups could be further validated with osteocalcin stainings.
38


Figure 3.8- H&E stainings of Unloaded bone cores at 2x (a,c) and 4x (b,d). 4x images show the area of intersection between the two cores.
39


To further expand on the results of the histology within the Loaded and Unloaded groups, an osteocalcin stain was performed. Osteocalcin is a late stage marker in osteogenesis and should provide additional information with respect to the formation of tissue within the bone scaffolds. The stainings within the Loaded and Unloaded groups are provided by figures 3.9 and 3.10, respectively. The images were taken at 2x zoom in order to provide an overall depiction of the scaffolds.
Based on the overall depictions of the scaffolds it does not appear that much, if any, difference can be deduced. Blindly looking at the Loaded and Unloaded groups, it appears that osteocalcin is prevalent throughout each scaffold, which could point to the possibility of osteogenesis. However, it is important to note that residual proteins could have been left within the bone scaffolds after the washing process, which could have led to a misrepresentation of the stain. Because of this occurrence, it is difficult to assess whether or not osteocalcin is being expressed or if it is a result of lingering proteins.
40


Figure 3.9 Osteocalcin stainings of the 15-minute Loaded bone cores at 2x (a,b) and 60-minute Loaded bone cores at 2x (c,d).
41


b)
Figure 3.10 Osteocalcin staining of the Unloaded bone cores at 2x (a,b).
42


CHAPTER IV
CONCLUSIONS
Beginning with in vivo and limited in vitro studies, the information available was limited with respect to developing an in vitro bone healing testing apparatus using bone allografts. The assays performed within this experiment provided insight into the future implementation of tools and techniques needed to optimize this system. Through the information gathered from live/dead, AlamarBlue, normalized ALP Activity, H&E, and osteocalcin stainings, an understanding of the potential osteogenic processes could be formulated.
The live/dead stainings showed that cells had attached and survived when seeded onto the scaffolds, which was important to show before loading regimens began. It was interesting comparing the Day 0 groups with the Loaded and Unloaded groups, because there was a significant difference in their cell viability.
The cells at the beginning of the experiment were showing greater metabolic activity than the cells at the end of the experiment. This difference could show that a majority of the cells, after 28 days in ODM, had been removed during media exchanges and were therefore less likely to express relatively high measurements in metabolic activity. This reasoning is further validated by the PicoGreen assay, which shows a decrease in DNA content between Day 0 and Day 28. In future studies, it would be advantageous to measure cell viability at multiple time points to understand these variances. Measuring at intermediate times of 0, 3, 7, 14, 21, and 28 days will provide a general overview of when cells may be lost and what is occurring throughout the experiment.
43


When compared with the Day 0 and Unloaded groups, the Loaded groups had an overall greater output with respect to normalized ALP activity. Although this difference was not significant, this trend could point to the notion of greater potential osteogenesis being experienced within the loaded group. The difference in normalized ALP expression could have been a product of the dynamic environment within the Loaded group. Based on in vivo and in vitro osteogenesis research, this was to be expected due to the enhanced environment mechanical stimulation provides for cellular communication and differentiation. Again, measuring at intermediate times of 0, 3, 7, 14, 21, and 28 days could provide additional information in understanding the differences between the loaded and unloaded groups.
The H&E and osteocalcin stainings provided an in-depth understanding of the differences between the scaffolds, as a qualitative assessment could determine what is happening at the tissue and cellular levels. Within the Loaded groups, there appeared to be a greater amount of new tissue development at the intersections and around the scaffolds. While the new tissue development was greater, the osteocalcin stainings could not provide a qualitative difference between the loaded and unloaded groups. Although osteocalcin is expressed within each group, the information is not conclusive enough to point to greater osteogenesis within the Loaded group. Unfortunately, the lack of difference in the late-stage osteogenic marker after 28 days of loading could not provide further insight into development of new bone.
Although the two bone scaffolds did not fuse together, which would ultimately define bone healing, the results enhanced the current understanding of an in vitro
44


bone healing model and how it can be improved with additional implementations. Because this type of in vitro analysis is in its infancy, there are solutions and directions that can be implemented such as including missing cellular variables. Osteoclasts, which are part of the hematopoietic cell lineage, play a significant role in the restructuring and remodeling of bone due to their communication with osteoblasts and osteocytes. Introducing osteoclasts into the model through coculturing and perfusion could generate micro blood vessels and enhance the modeling of in vitro bone healing. In addition to the inclusion of osteoclasts, recent studies have shown that co-culturing mesenchymal stem cells with osteoblasts and osteocytes can influence osteogenesis within in vitro models (18). On the cellular level, more research will need to be conducted in order to understand what coculturing environment will be optimal for enhancing osteogenic potential in vitro.
While enhancing the cellular environment of the model could be beneficial, adjusting the loading regimens of the model could ameliorate the bone healing system. For instance, increasing the load incrementally each day from 0.1% to 0.2% strain may produce greater osteogenic results, which was validated within the in vivo rabbit study (11). Although the custom bioreactor was limited to only implementing one bout per day, it would be advantageous to experiment with multiple bouts per day, as the increase in bouts per day proved to be relatively successful in vivo with respect to bone healing variables.
Implementing more test groups through the inclusion of cellular and mechanical variables will aid in understanding which variables have the greatest influence on osteogenesis. Furthermore, it will be advantageous to find other assays
45


outside of the H&E and osteocalcin stainings, which can better delineate the differences between the Loaded and Unloaded groups with respect to late stage osteogenesis. Previous in vivo studies had included Mason-Goldner Trichrome, Mason-Goldner Modified Trichrome, and Movats Pentachrome stainings to give added dimension to their histological evaluations. With these stainings implemented, they were able to decipher between osteoblasts, osteocytes, new bone formations, and new osteoids. According to these comprehensive evaluations, these stainings have greater detailed information with respect to bone healing and its various cell types. In addition, these stainings can also provide quantitative tools for comparing and contrasting overall osteogenic markers, which this study did not provide (13).
The enhancement of osteogenesis in the in vitro bone healing model can be complex due to the cellular communication, extra-cellular environments, and precise mechanical stimulations needed. Moving forward, it will be important to maintain and expand upon the dynamic environment for in vitro bone modeling in addition to including co-cultures. The results of this analysis should be used to optimize future studies and give a strong indication that mechanical stimulation is an important factor when discussing bone healing. While much knowledge still needs to be attained, the insight provided within this study can have a significant impact in future analyses and can hopefully generate enhancements within the bone graft, bone graft substitute, and bone biologic markets.
46


REFERENCES
1) U.S. Department of Health and Human Services. The Surgeon Generals Report on Bone Health and Osteoporosis: What It Means To You. U.S. Department of Health and Human Services, Office of the Surgeon General, 2012.
2) Ulstrup, Anton K. "Biomechanical Concepts of Fracture Healing in Weightbearing Long Bones." Acta Orthop 74.3 (2008): 291-302. Web.
3) Bab, Itai A., and Jona J. Sela. "Cellular and Molecular Aspects of Bone Repair." Principles of Bone Regeneration (2012): 11-41. Web.
4) Burr, D.b., A.g. Robling, and C.h. Turner. "Effects of Biomechanical Stress on Bones in Animals." Bone 30.5 (2002): 781-86.
5) Robling AG, Burr DB, Turner CH (2001). Recovery periods restore mechanosensitivity to dynamically loaded bone. J. Exp. Biol. 204, 3389-3399
6) Turner C. Toward a mathematical description of bone biology : The principle of cellular accomodation. Calcif Tissue Int 1999 ; 65 : 466-471.
7) Rubin, C. T., Ph.D., & Lanyon, L. E., M.R.C.V.S, Ph.D. (1984). The Journal of Bone and Joint Surgery. Regulation of bone formation by applied dynamic loads,66(3), 397-402.
8) OConnor, J. A., Lanyon, L. E., and MacFie, H. The influence of strain rate on adaptive bone remodeling. J Biomech 15:767-781; 1982.
9) Pan W, Einhorn T. The biochemistry of fracture healing. Curr Orthop (1992); 6 : 207-213.
10) Gardner, Michael J., MD et al. "Pause Insertions During Cyclic In Vivo Loading Affect Bone Healing." Clin Orthop Relat Res (2008): 1232-238.
11) Shadmehr, Azadeh, AN Esteki, Gholam R. Oliaie, Giti Torkaman, and Alireza Sabbaghian. "Augmentation of Bone Healing by Specific Frequency and Amplitude Compressive Strains." Orthopedics 32.3 (2009): 1-6.
12) Gardner, Michael J., MD et al. "In Vivo Cyclic Axial Compression Affects Bone Healing in the Mouse Tibia." Journal of Orthopaedic Research J.
Orthop. Res. 24.8 (2006): 1679-686.
13) Rentsch, Claudia, Wolfgang Schneiders, Suzanne Manthey, Barbe Rentsch, and Stephan Rammelt. "Comprehensive Histological Evaluation of Bone Implants." Biomatter 4.1 (2014). Web.
47


14) Grayson, Warren L., Sarindr Bhumiratana, Christopher Cannizzaro, P.-H. Grace Chao, Donald P. Lennon, Arnold I. Caplan, and Gordana Vunjak-Novakovic. "Effects of Initial Seeding Density and Fluid Perfusion Rate on Formation of Tissue-Engineered Bone." Tissue Engineering Part A 14.11 (2008): 1809-820. Web.
15) Villanueva, I., D.s. Hauschulz, D. Mejic, and S.j. Bryant. "Static and Dynamic Compressive Strains Influence Nitric Oxide Production and Chondrocyte Bioactivity When Encapsulated in PEG Hydrogels of Different Crosslinking Densities." Osteoarthritis and Cartilage 16.8 (2008): 909-18. Web.
16) AlamarBlue Cell Viability Reagent, (n.d.). Retrieved January 30, 2017, from https://www.thermofisher.com/order/catalog/product/DAL1025
17) SIGMAFAST p-Nitrophenyl phosphate Tablets, (n.d.). Retrieved January 30, 2017, from
http://www.sigmaaldrich.com/catalog/product/sigma/n1891 ?lang=enion=US &cm_sp=lnsite-_-prodRecCold_xorders-_-prodRecCold2-1
18) Birmingham, E., Niebur, G., Mchugh, P., Shaw, G., Barry, F., & Mcnamara,
L. (2012). Osteogenic differentiation of mesenchymal stem cells is regulated by osteocyte and osteoblast cells in a simplified bone niche. European Cells and Materials,23, 13-27. doi:10.22203/ecm.v023a02
19) A. Alex Jahangir, MD; Ryan M. Nunley, MD; Samir Mehta, MD; Alok Sharan, MD; and the, Washington Health Policy Fellows. Bone-graft subsitutes in orthpaedic surgery [Internet], AAOS; 2008 Jan. Available from: http://www.aaos.org/news/aaosnow/jan08/reimbursement2.asp
20) Brooks, B. D., Sinclair, K. D., Grainger, D. W., & Brooks, A. E. (2015). A Resorbable Antibiotic-Eluting Polymer Composite Bone Void Filler for Perioperative Infection Prevention in a Rabbit Radial Defect Model. PLoS ONE, 10(3), e0118696. http://doi.orq/10.1371/iournal.oone.0118696
21) Research, T. M. (2015, September 07). Bone Grafts and Substitutes Market expected to reach USD 3.48 Billion Globally in 2023: Transparency Market Research. Retrieved March 24, 2017, from http://qlobenewswire.com/news-release/2015/09/07/766429/10148433/en/Bone-Grafts-and-Substitutes-Market-expected-to-reach-USD-3-48-Billion-Globallv-in-2023-Transparencv-Market-Research.htm l#sthash.kpAvDPqn.dpuf
22) Kinaci A, Neuhaus V, Ring D. Trends in Bone Graft Use in the United States. ORTHOPEDICS. 2014; 37: e783-e788. doi: 10.3928/01477447-20140825-54
48


23) Roberts, T. T., & Rosenbaum, A. J. (2012). Bone grafts, bone substitutes and orthobiologics: The bridge between basic science and clinical advancements in fracture healing. Organogenesis, 8(4), 114-124.
http://doi.Org/10.4161 /org.23306
24) Oryan, A., Alidadi, S., Moshiri, A., & Maffulli, N. Bone regenerative medicine: classic options, novel strategies, and future directions. Journal of Orthopaedic Surgery and Research, (2014). 9, 18. http://doi.Org/10.1186/1749-799X-9-18
25) Rubin, C. T. and Lanyon, L. E. Regulation of bone formation by applied dynamic loads. J Bone Jt Surg 66-A (1984). 397-402
26) Golub, E. E., & Boesze-Battaglia, K. (2007). The role of alkaline phosphatase in mineralization. Current Opinion in Orthopaedics, 18(5), 444-448.
doi: 10.1097/bco.0b013e3282630851
27) Sharmin, F., Adams, D., Pensak, M., Dukas, A., Lieberman, J., & Khan, Y. (2015). Biofunctionalizing devitalized bone allografts through polymer-mediated short and long term growth factor delivery. Journal of Biomedical Materials Research Part A,103(9), 2847-2854. doi:10.1002/jbm.a.35435
28) Delaine-Smith, R. M., & Reilly, G. C. (2012). Mesenchymal stem cell responses to mechanical stimuli. Muscles, Ligaments and Tendons Journal, 2(3), 169-180.
29) Wagner DR, Lindsey DP, Li KW, Tummala P, Chandran SE, Smith RL, et al. (2008). Hydrostatic pressure enhances chondrogenic differentiation of human bone marrow stromal cells in osteochondrogenic medium. Ann Biomed Eng. 36(5):813-820.
30) Liu J, Zhao ZH, Li J, Zou L, Shuler C, Zou YW, et al. (2009) Hydrostatic Pressures Promote Initial Osteodifferentiation With ERK1/2 Not p38 MAPK Signaling Involved. J Cell Biochem. 107(2):224-232.
31) Quant-iT PicoGreen dsDNA Assay Kit. (n.d.). Retrieved April 23, 2017, from https://www.thermofisher.com/order/catalog/product/P7589
49


20
21
22
28
28
29
30
31
32
33
34
37
39
41
42
APPENDIX A
Figures
50


Full Text
xml version 1.0 encoding ISO-8859-1
DISS_submission publishing_option 0 embargo_code third_party_search Y
DISS_authorship
DISS_author type primary
DISS_name
DISS_surname Hout
DISS_fname Kevin
DISS_middle Ryan
DISS_suffix
DISS_affiliation University of Colorado at Denver
DISS_contact current
DISS_contact_effdt 03/23/2017
DISS_phone_fax P
DISS_cntry_cd 1
DISS_area_code 502
DISS_phone_num 649-6730
DISS_phone_ext
DISS_address
DISS_addrline 3531 Sandcastle Lane
DISS_city San Diego
DISS_st CA
DISS_pcode 92110
DISS_country US
DISS_email kevin.hout@ucdenver.edu
future
05/30/2017
1
502
649-6730
9217 Marlboro Circle
Louisville
KY
40222
US
kevin.hout@ucdenver.edu
DISS_citizenship US
DISS_description page_count 56 masters external_id http:dissertations.umi.comucdenver:10818 apply_for_copyright no
DISS_title A NOVEL TOOL FOR IN VITRO BONE FUSION TECHNIQUES THROUGH MECHANICAL STIMULATION
DISS_dates
DISS_comp_date 2017
DISS_accept_date 01/01/2017
DISS_degree M.S.
DISS_institution
DISS_inst_code 0765
DISS_inst_name University of Colorado Denver
DISS_inst_contact Bioengineering
DISS_processing_code N
DISS_advisor
Patel
Vikas
DISS_cmte_member
Hunter
Kendall
Payne
Karin
DISS_categorization
DISS_category
DISS_cat_code 0541
DISS_cat_desc Biomedical engineering
DISS_keyword Bioreactor, Bone Healing, Mechanical Loading, Mesenchymal Stem Cells
DISS_language en
DISS_content
DISS_abstract
DISS_para Orthopedic bone grafts and bone graft substitutes are currently used to facilitate bone regeneration and new bone formation in patients needing orthopedic structural support and union of bony defects (22). Bone void healing in post-operative cases is generally improved by the use of grafts, but differs from patient to patient, requiring improvements in current and future graft products. A number of different bone graft types exist, encompassing autografts, allografts, synthetic materials, and biologics; however, each possesses limitations due to variabilities with respect to bone healing. The disadvantages of these products can be attributed to necrosis of harvested bone, rejections, diseases, inflammatory responses, and structural instability (23,24). According to Transparency Market Research, the global market for bone grafts and bone graft substitutes will have an annual compound growth of 4.5%, reaching $3.48 billion in 2023 (21).
The goal of this experiment is to develop an in vitro system capable of testing these products so their osteoinductive, osteoconductive, and osteogenic properties can be enhanced at reduced cost and experimental time. If the in vitro system is successful, then it could be further implemented within in vivo and clinical applications.
Currently, no in vitro approaches to bone healing using bone grafts or bone graft substitutes have been published. However, multiple in vivo bone healing studies have shown mechanical stimulation is an important variable with respect to enhancing osteogenesis.
Using a bioreactor to apply a mechanical load, a novel system can be generated to induce osteogenesis between two revitalized-cadaveric-bone-allografts in vitro. Finding the optimal mechanical forces, time of load, and cycling for osteogenesis are the first parts in many steps when trying to configure the tools for in vitro bone healing studies.
DISS_supp_abstract
DISS_binary PDF Hout_ucdenver_0765N_10818.pdf
DISS_restriction
DISS_repository
DISS_version 2011-11-08 15:37:33
DISS_agreement_decision_date 2017-04-03 11:52:30
DISS_acceptance 1
DISS_delayed_release
DISS_access_option



PAGE 1

A NOVEL TOOL FOR IN VITRO BONE FUSION TECHNIQUES THROUGH MECHANICAL STIMULATION by KEVIN RYAN HOUT B.S., University of Kentucky, 2011 M.S., University of Kentucky, 2013 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 Master of Science Bioengineering Program 2017

PAGE 2

ii This thesis for the Master of Science degree by Kevin Hout Has been approved for the Bioengineering Program by Kendall Hunter, Chair Vikas Patel, Advisor Karin Payne May 13, 2017

PAGE 3

iii Hout, Kevin (M.S., Bioengineering) A Novel Tool for In vitro Bone Fusion Techniques Through Mechanical Stimulation Thesis directed by Professor Vikas Patel ABSTRACT Orthopedic bone grafts and bone graft substitutes are currently used to facilitate bone regeneration and new bone formation in patients needing orthopedic structural support and union of bony defects (22). Bone void healing in post operative cases is gener ally improved by the use of grafts, but differs from patient to patient, requiring improvements in current and future graft products. A number of different bone graft types exist, encompassing autografts, allografts, synthetic materials, and biologics; how ever, each possesses limitations due to variabilities with respect to bone healing. The disadvantages of these products can be attributed to necrosis of harvested bone, rejections, diseases, inflammatory responses, and structural instability (23,24). Accor ding to Transparency Market Research, the global market for bone grafts and bone graft substitutes will have an annual compound growth of 4.5%, reaching $3.48 billion in 2023 (21). The goal of this experiment is to develop an in vitro system capable of t esting these products so their osteoinductive, osteoconductive, and osteogenic properties can be enhanced at reduced cost and experimental time. If the in vitro system is successful, then it could be further implemented within in vivo and clinical applicat ions. Currently, no in vitro approaches to bone healing using bone grafts or bone graft substitutes have been published. However, multiple in vivo bone healing

PAGE 4

iv studies have shown mechanical stimulation is an important variable with respect to enhancing o steogenesis Using a bioreactor to apply a mechanical load, a novel system can be generated to induce osteogenesis between two revitalized cadaveric bone allografts in vitro. Finding the optimal mechanical forces, time of load, and cycling for osteogenes is are the first parts in many steps when trying to configure the tools for in vitro bone healing studies. The form and content of this abstract are approved. I recommend its publication. Approved: Vikas Patel

PAGE 5

v TABLE OF CONTENTS CHAPTER I. INTRODUCTION AND GOALS Background ................................ ................................ ............................. 3 Adaption of Bone Cells in Non Fracture Studies ........................ 4 In vivo Bone Fracture Healing ................................ ..................... 8 MSC Differentiation and Osteogenic Potential through Mechanical 11 Classifying Bone Healing through Cellular Assays ................... 12 Bone Healing Histology ................................ ............................. 13 Goals and Hypotheses ................................ .............................. 14 II. 15 Collecting Bone Core Samples ................................ ................. 15 Finding the Stress Strain Curves for the Bone Cores .............. 15 Splitting the Bone Cores ................................ ............................ 17 Culturing Bone Marrow Mesenchymal Stems Cells ................. 18 Seeding Bone Marrow MSCs onto Bone Cores ....................... 18 Live/Dead Staining of Scaffolds ................................ ................ 18 Using the Bioreactor ................................ ................................ .. 19 Mechanically Loading Bone Core Samples .............................. 21 AlamarBlue ................................ ................................ ................. 23

PAGE 6

vi ALP Activity & PicoGreen ................................ .......................... 24 Histology ................................ ................................ .................... 26 Statistical Analysis ................................ ................................ ..... 26 III. 27 Stress Strain Curves for the Bone Cores ................................ .. 27 Live/Dead Staining ................................ ................................ ..... 29 alamarBlue ................................ ................................ ................. 30 ALP Activity & Pico Green ................................ ......................... 32 Hematoxylin & Eosin and Osteocalcin Histology ...................... 35 IV. CONCLUSIONS 43 REFERENCES ................................ ................................ ................................ ............ 47 APPENDIX

PAGE 7

1 CHAPTER I INTRODUCTION AND GOALS Osteoporosis, by 2020 nearly half of the American population over 50 will have American economy with an estimated co st of $18 billion per year (1). In addition over 3 million bone graft procedures are performed annually which require the use of autografts, allografts, and synthetic biomaterials as bone void fillers (19,20) Within these bone void filler s there is a push to enhance bone healing in patients by incorporating biologics which cou l d potentially increase osteogenesis The iologics and bone graft substitutes has recently increased due to the advancements made within the past 20 years (23). While bone grafts, bone graft substitutes, and biologics have their advantages, there are associated negatives that can result in failures and rejections. An autograft which is bone harve sted from the same patient, IS considered t he gold standard when it comes to bone grafts due to their biocompatibility with the patient. However, harvesting autologous bone tissue has limitations due to increased infection potential bone necrosis, and limited donor supplies (23). Another popular bone graft is the allograft, which is harvested from human cadaveric bone. Allografts have become increasingly popular due to their relative extra cellular matrix increases its

PAGE 8

2 osteogenic potential. Unfortunate ly, 30 60% of allografts have issues associated with failures and rejections when evaluated 10 years post insertion (23,27) More rece ntly, the use of biologics and bone graft substitutes h ave become popular tools for bone void filling Bone morphogenic proteins (BMPs) and vascular endothelial growth factors (VEGFs) are two examples of biologics that can promote osteogenesis and angiogenesis through osteoinduction Osteoinduction is the process in which bone forming cells are recruited to promote bone fo rmation, typically done in conjunction with bone graft substitutes. B one graft substitutes like calcium phosphate have shown to have 4 10x more compressive strength when compared to cancellous bone (23) B iologics and bone graft substitutes have their ad vantages but there are shortcomings for these technologies. There are limitations with respect to bone remodeling & limited shear strength ( bone graft substitutes ) decre ased concentrations at the site of implementation & limited human studies ( biologics ), and cost (23). The development of a n in vitro bone healing system could aid in analyzing new bone tissue formation using various bone graft s, bone graft substitutes, and biologics. Furthermore, an in vitro bon e healing system could potentially contrib ute to the advancements of orthopedic solutions i n a non invasive, cost reduced and timely manner. Effectively the broader impacts for this system would include improvements in bone grafts, synthetic bone grafts, biologics and tissue engineering capabilities The in vitro method develop ed would be able to improve bone graft modeling and help advance their implementation within in vivo animal studies and potentially enhance treatments for those needing bone graft procedures

PAGE 9

3 Background s law is the basis for understanding how bone adapts due to internal changes and states: certain changes in the internal architecture and external form of the bone according to mechanical laws In o rder to establish a bone healing process in vitro it is important to understand the biological and mechanical principles that occur in vivo There are biomechanical and cellular processes of bone repair that play significant roles in coordinating and regu lating the healing process. With respect to the cellular process of bone repair, f our types of bone cells contribu te to the architecture of bone: osteoblasts, osteoclasts, ost eocytes, and bone lining cells. Osteoblasts, osteocytes, and bone lining cells ar ise from the differentiation of mesenchymal stem cells (MSCs ), while osteoclasts arise from the differentiation of hematopoietic stem cells (HSCs). The primary determinant of healt hy bone is d ictated by the communication between the osteocytes, osteoblasts osteoclasts, and is defined by new bone formation (2,3). With respect to biomechanical variables m echanotransduction is an essential bone healing process that is effectively initiated when a mechanical force is being applied within bone. Mechanotransduction is the process of converting mechanical stimulation signaling into biochemical signals, and it is essential for MSC differentiation and osteogenesis. Combining an applied mechanical force with osteogenic differentiation media (ODM) c ould increase osteogenic potential with respect to MSC differentiation and new bone tissue formation MSC differentiation in vitro can be divided into three separate stages. The first stage occurs in days 1 to 4, which shows a peak in the

PAGE 10

4 number of cells. Secondly, Days 5 to 14 show an increase in alkaline phosphatase ( ALP ) activity, which is an enzyme expressed during the beginning stages of cell differentiation and osteogenic activity Finally, days 14 to 28 show high expressions of the proteins: osteopontin and osteocalcin which also results in decrease d ALP expression ( 26, 18) One of the m ain objectives for developing an in vitro bone healing system protocol is establishing a loading frequency and compressive strain that will increase ALP activi ty, show new tissue formation, and exhibit increase d osteocalcin. Establishing these methods will allow for an understanding of how compression and frequency of loads can stimulate and a ffect new bone tissue formation in vitro Once an acceptable str ain an d frequency are established, the mechanical stimulus will be applied using a pre fabricated bone bio reactor system in order to optimize the b one healing process. Finally, various cellular assays and staining techniques will be implemented to evaluate the cellular viability, ALP activity, and histological characteristics which will assist in understanding the osteogenic processes occurring. Establishing these methods and techniques can aid in optimizing future in vitro bone healing studies. Adaption of Bone Cells in Non Fracture Studies Adaption of bone will occur, in a typical individual, where activity and mechanical forces can be detected (2,6). The detection of mechanical forces and transcribing them into bone forming signals is a process known as m echanotransduction In bone healing studies, d ynamic loading is a necessary variable for osteogenesis, while also preventing suppression of bone growth. Dynamic loads generate d a re critical for bone formation because of their ability to

PAGE 11

5 generate an adaptiv e response (8). In addition, the stress and strain generated from the mechanical forces influence cell differentiation and are mediated by cyclic loading. Fluid can only be moved through the canaliculi by cyclic loading, which makes it an important compone nt in bone healing and bone formation (2,4). Burr, Robling, and Turner conducted several in vivo studies with respect to biomechanic al stresses on bones and how these application s affect responses (4,5) Their aim was to look at bone response instead of bone healing in order to understand the healthy response of bone. Under static loads it has been found that bone does not undergo the same mechanotransduction responses as bone that is under dynamic load. The difference in responses can be att ributed to the fluid movement generated between bouts of cyclic loading and relaxation. In addition, the fluid and shear stresses from the cyclic loading act on the cilia of the osteocytes, which are responsible for initia ting the beginning stages of os t e o gen esis. Osteogenic response s occur rapidly, but the cells require a recovery period to re establi sh their mechanical sensitivity in order to communicate cellular signaling following another mechanical stimulation. Rubin and Lanyon implied that the response of cells can occur quickly and effectively, and does not require high cyclic loads (25) Additionally this suggested that bones may become desensitized to stimuli if all of the cycles occur at one time (7). This introduced the idea that recovery periods in between cycles are vital for cell recovery and restoring cell sensitivity to mechanical stimulation. If the cells are subjected to continuous stimulation, their re covery time may be hindered, which could lead to a decrease in osteogenes is. In their 2001 study, Burr, Robling, and Turner analyzed six different

PAGE 12

6 sham bending test groups on rat tibia that were under various time loading regimens (5) The y were separated accordingly into 0, 0.5, 1, 2, 4, and 8 hour recovery groups At the conc lusion of their experiment, they were able to find that the 8 hour recovery groups experienced a significantly higher relative bone formation rate measured by fluorescence photomicrographs in the loaded limb than the non loaded limb. Overall, the 8 hour recovery group had a 125% higher bone formation rate than the 0 hour groups and was also 102% higher than the 0.5 hour groups. These findings helped to assert that recovery time was needed to restore cell sensitivity to mechanical stimuli. Furthermore, the y also looked into the effects that time loading between cycles had on bone formation. The time intervals of interest were separated into 0. 5, 3.5, 7, and 14 secon d groups, which experienced four bouts of 90 cycles/bout. Relative mineralization surface and bone formation rates measured by histomorphometry, were significantly higher in the 14 second groups than in the other groups (p<0.001). This indicated that a high recovery period in combination with a high interval between bouts could generate the m ost effective form of osteogenesis with respect to new bone formation (5). Using th is data, an analysis of loading bouts, cycles, and rest periods was conducted to understand the effects that dynamic loading in combination with recovery time has on osteo genesis. A 4 point bending model applied a dynamic load with given bouts, cycles, and recovery periods. F our interest groups were used to assess how cyclic loading has an effect on bone formation rates. The four groups were separated into 360 cycles x 1 bo ut/ day, 60 cycles x 6 bouts/day (2hr recovery), 90 cycles x 4 bout s /day (3hr recovery), and 180 cycles x 2 bouts/day (6hr recovery).

PAGE 13

7 A positive osteogenic effect was found for the groups with recovery periods between 2 & 3 hours (60x6 and 90x4 groups) and showed approximately 80% higher bone formation rates than the 360 cycles/day group This further developed the idea that managing recovery periods can i ncrease the effectiveness of loading and osteogenesis. In the same study, the model was used to determine the optimal recovery time needed between each bout. Using the 90 cycles x 4 bout s group, loading intervals between bouts were separated in 0, 0.5, 1, 2, 4, and 8 hour groups. Again, the recovery time betwe en bouts proved to be the most effective at exhibiting the greatest bone formation rate. The 4 and 8 hour groups had significantly greater relative bone formation rates than the group that was allowed no recovery time. This data suggested that the sensitiv ity of the cells was re established in between 4 to 8 hour recovery peri ods. Furthermore, the study continued to analyze time intervals between cycles for 0.5, 3.5, 7, and 14 second groups at 36 cycles/day. As expected, t he 14 second group was significantl y higher with respect to bone formation rates, than the other three groups, which proved to show recovery intervals are vital for producing osteogenic responses (4) These two studies indicated that recovery periods are need ed in order to optimize osteo genesis with respect to relative mineralization and bone formation rates (4,5) Establishing recovery periods could play a significant role in cellular communication differentiation, and gene expression with respect to osteogenesis. Bone cells can adapt q uickly to the environments they are subjected to, which is why it is important to manage the mechanical stimuli and avoid de sensitization.

PAGE 14

8 In vivo Bone Fracture Healing Several in vivo animal studies including mice and rabbits have been modeled to show how bone healing is affected by mechanical loading compared to c ontrol groups without loading (10,11,12) Mechanical loadings in each study were measured based on specific amplitudes, frequencies, strains, and/or cycles. In each of these studies, the various compressive forces It should be noted that producing an osteotomy instead of a have result ed in biased data with respect to the healing process ( 1, 9 ). The main approaches to analyzing bone healing in these studies were represented through radiographs and bending tests. Variables assessed within these approaches included callus volumes, osteoid presence, mineralization, and mechanical properties post osteotomy. Michael Gardner, et al. have produced an extensive amount of in vivo resear ch with respect to bone healing. Similar to the studies that looked at the mechanical response of cells in healthy bones, Gardner et al. established a modeling system and in corporated mecha nical stress effects on bone (10,12) Their modeling systems looked into the failure moments and callus volumes of osteoto mized mice tibias after a specific load an d amplitude were applied. In their first study, they analyzed the effects delayed loading po st osteotomy has on the in vivo model. Five sample groups had an a pplied axial compression in addition to a delay and non delay : 0.5N/0 day delay, 1N/0 day delay 0.5N/4 day delay 1N /4 day delay, 2N/4 daydelay These five sample groups were compared to a control group which was not subjected to mechanical stimulations. Each group was analyzed after

PAGE 15

9 a cyclic load had been applied for 2 weeks. The results indicated that the callus strength and bending stiffness improved with a 0.5N/ 4 day delay when compa red to the control group and other delayed groups. In addition, the groups that did not receive a delay (0d) showed a 68% decrease in callus strength even though their callus volumes were significantly larger. This indicated callus volume cannot be a direc t indication of bone strength, and should not be considered when analyzing bone healing. Although the callus volumes were significantly higher in the 0 day delay groups, the 4 day delay groups produced g reater failure moments overall (12 ). With respect to delayed onset and establishing recovery per iods, bone healing mechanisms p e r form consistently with the information found in the non fracture bone studies. A study following the res ults of their previous analysis was conducted in 2008 by Michael Gardner, et al., which analyzed the effects of bon e healing in vivo through pause inserted cyclic loading (10). Eighty mice underwent tibia osteotomies and were separated into four group s based on loading procedures. Three cyclic loading groups had a 0.5N force wit h a 1N amplitude applied, while the control group did not experience a load. The three cyclic loading groups were subjected to a repetitive load (100 cycles for 100 seconds) 9 second pause inserted time equivalent load (100 seconds ) or 9 second pause inserted cyclic equivalent load ( 1000 seconds) Loading was applied for two weeks post osteotomy, which was followed by the bone healing analyses. The percent osteoid and mineralization were analyzed using micro CT and histology Overall, the loaded group s produced a larger osteoid and percent mineralization than the control group. However, the callus

PAGE 16

10 volume in the control group was larger than the loaded groups. Although callus volume was larger in the control group the failure moment in the loaded group s was greater, which is consistent with their previous findings. Again, callus volume does not directly correlate to the strength of healed bone. When the four groups were compared based on stiffness and ma ximum bending moments, the pause inserted cycle eq uivalent group (100 seconds) showed an overall advantage amongst the other groups (10) Again, a trend could be seen within groups that have an applied pause/rest period between bouts and/or compressions. This indicates that bone is healed optimally when c ells can re establish their signaling sensitivity through a given rest period. Sim ilar to the previous study, a 2009 study conducted by Shadmehr, et al. assessed bone healing through frequency and amplitudes of specific strains (11) The right tibias of 1 6 rabbits were osteotomized and an external fixator was used to apply specific frequencies and amplitudes. Four groups were formulated based on the durations of mechanical stimulations and their respective controls. The durations of mechani cal stimulation s continued for 1 and 2 weeks, in which a 1 Hz (1 cycle per second) frequency was applied 15 minutes per day. In addition, the stimulation sessions increased their exerted forces each day in a somewh at linear progression After the mechanical stimulations and analyses were complete, t he radiological parameters and mechanical evaluations were significantly higher in the 1 and 2 week groups compared to the respective con trol groups. In addition, the 2 week stimulated group had higher overall values for callus volume and stiffness when compared to the 1 week group. This study also showed greater callus volume

PAGE 17

11 area within the cyclic loaded groups than the previous study, which can reflect some variance based on species (11). In addition, t his study furthe r expanded upon the idea that inserting a dynamic load to a healing bone can optimally enhance the bone streng th. All three studies and their respective data proved to be consistent with one another. It is important to find a non destructive axi al load and strain in order to optimize the bone strength and bone formation. In the first study, the 4N and 2N amplitudes may have been too great, which may have contributed to a decrease in callus volume and failure moments. The modulus of elasticity of bone for different animals will vary, which is why it is important to apply the correct load for a specific strain of interest. In addition to applying mechanical compression, allowing a pause in between stimulations has shown to provide an optimal environ ment for enhancing both f racture and non fracture bone. Allowing a sufficient recovery period for the cells is an important process in the re organization, growth, and strength of healthy bone. MSC Differentiation and Osteogenic Potential through Mechanica l Stimulation MSCs play an integral part in osteogenesis as they are the progenitor s to osteoblasts, osteocy t e s and bone lining cells Stimulating the MSCs can maximize their osteogenic potential by helping to initiate their differentiation into their ce ll lineages Osteogenic markers and new bone matrix have been shown to increase within in vitro MSC seeded scaffolds where cyclic loading is applied (28,29,30) In addi tion, compressive loading of MSC seeded polymer scaffolds upregulated bone matrix formin g genes, which in turn increased bone like mineralization Furthermore,

PAGE 18

12 only 2 hours of loading per day for 5 days was needed to differentiate MSCs which outperformed continuously loaded groups by approximately 50% with respect to calcium deposition (28) It is unlikely MSCs experienced direct compression in these studies, but it is more likely that they experienced the secondary effects of the mechanical stimulation through fluid flow and scaffold bending (28). The mechanical applications for initiating MSC differentiation in vitro are important which re iterates the notion that loading cycles are need ed for optimizing osteogenesis. Although this study did not address osteogenesis through the use of bone, their polymer applications are important when con sidering MSC differentiation in bone scaffolds. Classifying Bone Healing through Cellular Assays Osteogenic biomarker assays can help to measure osteogenic differentiation of MSCs, which has been used in publications and studies by Birmingham, et al (18) According to Birmingham, et al. there are three stages of osteogenesis in vitro The first stage of osteogenesis in vitro occurs within the first four days which will show a maximum population of MSCs. Following a peak in MSC population, the second stag e begins between days five and fourteen. Within this stage, early osteogenic cell differentiation begins and can be characterized by a rise in ALP The third and final stage of osteogenesis in vitro occurs between days 14 and 28, which expresses a high amo unt of osteocalcin and osteopontin. C alcium phosphate deposition begins in the later portion of stage three following the rise in osteocalcin and osteopontin Transcription and protein expression of ALP is an important biomarker in the osteogenic differentiation of MSCs which can be adequately quantified by ALP activity assay s (18 ).

PAGE 19

13 Bone Healing Histology Identifying and defining the bone healing process through histology is a experimental m ethodologies and procedures. Using various stainings on bone can assist in the evaluations and understanding s of the processes that are taking place at the time of the experiment A couple of modalities that can be used to classify bone structure, new tiss ue formation, and relative osteocalcin present are Hematoxylin & Eosin (H&E) and Osteocalcin Immuno Histo Chemistry (IHC) (13) H &E staining s give a general overview of the new and old tissue present within the bone. This method in particular helps to de lineate bone, cartilage, and connective tissue based on the color intensity of the staining. Although this is described as a general staining procedure, H & E staining s generate a working knowledge of the tissue structure and possibly new bone formation. In addition to HE staining, IHC can provide supportive information about the development of bone tissue with respect to the development of in vitro osteogenesis In specific, osteocalcin staining can be performed as it pertains to a late stage marker for osteogenesis As previously mentioned, high levels of osteocalcin are expressed in vitro between days 14 and 28, which is of particular interest due to the 28 day loading regimens applied within this study (18 ). T he inclusion of H&E and IHC stainings are crucial in providing a novel underst anding of the healing processes, which may be occurring during the loading.

PAGE 20

14 Goals and Hypotheses The goa l of this study is to provide basic tools for approach ing in vitro bone healing through mechanical loading applications. By applying a mechanical load to bone cores placed in ODM and se eded with MSCs this could conceivably optimize osteogenesis This study attempts to compare and contras t loaded versus unloaded seeded b one cores through various assays and stainings, which could provide further knowledge for adjusting future in vitro experiments Classifying both early and late osteogenesis will be needed to analyze the two groups of interests, which can be conducted th rough cellular assays and histology Based on previous in vivo and in vitro osteogenesis research and the mechanical processes contained within these studies, it is hypothesized the loaded groups will have more osteogenic potential when c ompared with the u nloaded group.

PAGE 21

15 CHAPTER II MATERIALS AND METHOD S Collecting Bone Core Samples The following bone core sampling procedure was adapted fro Ini tial Seeding Density and Fluid Perfusion Rate on Formation of Tissue Engineered bone cores each approximately 4.8mm in width and 10mm in height, were obtained from the femoral head of cadaveric human bone using a coring tool developed by Todd Baldini in the Department of Orthopedics The cadaver was a 55 year old white male. The core samples were then cleansed under high pressure steamed water for one hour to remove bone marrow. After the cleansing process, the cores were then washed in cleaning solutions of 0.1% EDTA in PBS for one hour at room temperature, a hypotonic buffer of 0.1% EDTA in 10 mM Tris overnight at 4C, detergent comprised of 0.5% SDS in 10mM Tris for 24 hours at room temperature, and an enzymat ic solution of 50 U/ml DNase, 1 U/ml RNase in 10mM Tris for 3 6 ho urs at 37C. The washed samples were then freeze dried, and sterilized in 70% ethanol for one hour at room t emperature in a ventilated hood. This process prepared the bone cores to be conditioned in osteogenic media overnight before seeding with cells. Finding the Stress Strain Curves for the Bone Cores Configuring the stress strain curves was an important process in finding the ultimate strength of the collected bone cores. From thi s information, a strain value (mm/mm) at max stress was found by correlating its value to the ultimate strength (N /mm 2 ) fo r the bone core of interest. This strain value was used to understand if the

PAGE 22

16 custom built compression bioreactor provided by Dr. Step and Biological Engineering Lab, could apply the needed strain to bone cores of interest. Overloading the bioreactor was a preliminary concern due to the relative stiffness of the bone compared with the hydrogels the bioreactor typic ally handles. The following procedure was used to generate the data for the stress strain curves. T wenty bone cores were selected to be analyzed on a universal testing machine (UTM) Biomechanics Lab Before testing the bone co res individually, the lengths (mm), widths (mm), and weights (g) were measured in order to generate the ir respective stress strain curves. Each bone core was found to have varying volumes ( weights (g), and therefore densities (g/ The densit ies ranged from 4.08e 4 to 1.19e 3 g/ which was relatively large but was expected due to the variance in bone properties during the collection process. The specimen properties and stress strain values can be seen in Table 3 1 and Table 2.2, respectively After analyzing, measuring, and applying the testing parameters, only 17 bone cores could be used in the final analys is due to some errors in loading and data collection The specimens highlighted in red in Table s 3 1 and 3 .2 were not used in the final analysis. Using the UTM, a linear displacement of 1 mm/sec was applied to each bone core, recording the information at 20Hz, which was then output into a text file. Strain (mm/mm) and stress (N /mm 2 ) values were generated into two columns and were from the beginning of each test, which is the subject of the UTM administering the beginning stages of the compression. T he data was re form atted to reflect these changes and

PAGE 23

17 the stress strain data for each bone core began where the stress value was zero. If a zero value for stress was not found then the first negative stress integer and its respective strain value were used. Following the f ormatting of each bone core loading data, the ultimate strengths and their associated strain values were calculated. Using these values, understanding the proper utilization of the bioreactor, and taking into account the most suitable mechanical environmen t to be established an applied strain for the loaded bone cores was generated. Reviewing the literature and having discussions with Dr. Stephanie Bryant, Dr. Vikas Patel, and Dr. Karin Payne, a 0.1% (10 micron) strain was used for the mechanical loading. This strain was applicable to generate enough mechanical stimuli to optimize osteogenesis in vitro Splitting the Bone Cores In order to properly model bone healing in vitro the bone cores needed to be split adequately in half so they could be aligned when the loading was established. A Buehler Isomet 1000 Precision Saw provided by Dr. Clifton Carey with the School of Dental Medicine, was used to cut the bone cores and evenly split the specimen into two bone core scaffolds. The process was generated by gently placing each 4.8mm x 10mm bone core medially onto the mechanical saw rotating at approximately 10 0 rpm The mechanical saw then cut through each bone core at a moderate pace until full separation had been initiated. After the bone cores had been ad equately cut they were sterilized in 70% e thanol for 24 hours and dried in a cell culture hood for another 24 hours. Once completely dry, the cores were then placed in 20 C until they could be seeded.

PAGE 24

18 Culturing Bone Marrow Mesenchymal Stems Cells The MSCs were purchased from Texas A&M University College of Medicine Institute for Regenerative Medicine, where they were harvested from human bone marrow of a 22 year old male patient Cells were removed from cryopreservation, thawed, plated in a T 225 flask (225 cm 2 ), and cultured in 30 mL of complete culture media (CCM MEM alpha (Life Technologies, Carlsbad, CA, USA) ) supplemented with 16.5% FBS (Atlanta biologicals, Lawrenceville, GA, USA), 2 mM L glutamine (Corning, Manassas, VA, USA), and 1% penicillin streptomycin (Thermo Scientific, Logan, UT, USA). Cells were then incubated at 37 C and 5% CO 2 until they were confluent. After the cells had reached confluency, they were then passaged, split, a nd plated at a density of 100 cells/cm 2 The CCM was changed every 48 72 hours until an adequate number of cells could be used to seed at 1.2x10 6 cells per bone core scaffold Seeding Bone Marrow MSCs onto Bone Cores The MSCs in CCM were suspended at passage 4 and aliquoted onto a blot dried bone scaffold using a 10 100 L pip ette Every 15 minutes for 1 hour, each scaffold was rotated 90 and 5 L of osteogenic differentiation media was added directly to the scaffold. This process aided in facilitating uniform distribution of cells within each scaffold. The scaffolds were then incubated for 24 hours at 37 C and 5% CO 2 before mechanical loading was applied. Live/Dead Staining of Scaffolds Live/dead staining was performed to validate cell attachment within the scaffolds using Life Tech nologies LIVE/DEAD Cell Viability/Cytotoxicity Kit. Each

PAGE 25

19 scaffold was washed with 1X PBS before performing the assay. 2mL of the LIVE/DEAD assay reagents were applied to each scaffold with the lights off due t o the sensitivity of the staining. The scaffolds were then incubated for 45 minutes at room temperature and protected from light by wrapping each plate in foil. Following incubation, the stain was aspirated and each scaffold was washed with 1X PBS three ti mes. The scaffolds were then transferred to 35 mm petri dishes and covered with 1X PBS before being analyzed. Each sample was analyzed using a Nikon confocal microscopy with the settings EGFP: HV55, offset 2, 488 2.00; ALX 594: HV 80, offset 2, 561 2.00; 1 024 px size, 5.3 px dwell, 10X; Z stacks of 10 m steps; filters: turret. Using the Bioreactor The custom bioreactor, provided by Dr. Stephanie Bryant, uses a specific program developed by a former student to apply specific strains with given time intervals to scaffolds of interest (15) Th is in house program communicates with a bioreactor controller and the bioreactor which enables specific input frequencies, input strains, scaffold heights, and time intervals Figure 2 .1 shows a n overall schematic of the bioreactor and its various components. The loading pins, which rest on top of the scaffolds, are located above the 24 well plate and are connected to a loading platform. This loading platform is connected to a stepper motor, whic h is controlled by the program via wired connections. The stepper motor is the component of the machine that applies the specific strain, which is applied equally across all scaffolds due to the loading pin and loading platform connection. The

PAGE 26

20 entire biore actor and its connections were placed in an incubator, which is where the loading regimens occurred. Figure 2 .1 Bioreactor and its various components. Figure 2 .2 shows an excel spreadsheet that is used in conjunction with the program to specify the given parameters of the loading regimen. The figure does not represent the actual values used within the experiments, but it was formatted to reflect the loading regimen

PAGE 27

21 Figure 2 .2 Excel Spreadsheet that is used to input values and parameters for a given loading cycle. Mechanically Loading Bone Core Samples Similar techniques for the bone loading applications were adapted from Villanueva et al. (15 ) Using the custom bioreactor, a specific strain of 0.1% (10 microns) was applied to the bone cores of interest. In a 24 well non tissue culture plastic plate, e ach well encomp ass ed two seeded bone core scaffolds with BM MSCs which were cut from the same specimen. Stabilization was initiated using PYREX Cloning Cylinders, which provided a casted enclosure for the scaffolds. Furthermore the cylinders would allow flow of media throug h the top and bottom of each scaffold due to their hollow design. The scaffolds were organized into loaded and non loaded groups within each row of the 24 well plate; each well was sequentially numbered to avoid confusion.

PAGE 28

22 Eac h well receive d 2 mL of oste ogenic differentiation media (ODM, CCM supplemented with 10 nM dexamethasone (Sigma, St. Louis, MO, USA) ) ascorbic acid 2 phosphate (Fisher Scientific, Fair Lawn, NJ, USA)) The bioreactor and cores were incubated at 37 C and 5% CO 2 and the ODM was changed every 48 72 hours. The seeded cores were divided into three groups of interests : un loaded cores with ODM at 28 days, loaded cores with ODM at 28 days, and cores at Day 0. The loading regimens for the loaded gr oups experience d a 0.1% strain (10 microns) at 1Hz which was then followed by a 9 second pause. This process was repeated for 90 cycles, once per day, for 28 days within the first two runs which equated to a loading time of 15 minutes per day. The third run experienced the same strain and frequency, but the regimen increased to a total of 360 cycles per day for 28 days which equated to 60 minutes of loading F igure 2 .3 below depicts the loading cycle applied to each scaffold and was adapted from Gardner, et. al. (2) After the 28 day loading period was complete, the samples were then removed from the bioreactor processed, and analyzed based on their designated assay. Figure 2 .3 Loading cycle with 1Hz load and 9 second pause applied to bone scaffolds Note: the loading regimens were not 0.5 to 1 N, which is reflected in this figure. Instead the loading was 0 to 0.1% strain.

PAGE 29

23 Alamar Blue The colorimetric change and a fluorescent signal in redox indicato r detects oxidation during cellular respiration, which is a process that releases waste products due to ATP conversion. The amount of fluorescence is proportional to the metabolic activity of the cells. Therefore, cells not proliferating or decreased acti vity will have a lower fluorescence (16). Detection of the dye was performed using a Promega Microplate Fluorometer which measured excitation at 530 560 nm and emission at 590 nm. A positive control was generated by autoclaving 5 mL ODM medium and 500 L a lamarBlue reagent in a glass bottle. Once autoclaved, the positive control was added to one well of a controls plate with 100 L of ultrapure water. For the negative control, 1 mL of ODM only was added within the controls plate. The bone cores received 1 mL of fresh ODM media and 100 L of alamarBlue reagent was added to each well. The cores and controls were then incubated for 2.5 hours at 37 C and 5% CO 2 100 L of each sample were added in triplicate to a 96 well plate and the plate was wrapped in foi l due to the light sensitivity of this assay. The plate was then taken to the Promega Microplate Fluorometer where the fluorescene readings were generated. This process was repeated at days 0 and 28 of the experiment to analyze cell viability between pre l oaded, un loaded, and loaded bone cores.

PAGE 30

24 ALP Activity & Pico Green The ALP Activity & PicoGreen assays were performed using Sigma SIGMAFAST p Nitrophenyl phosphate (pNPP) Tablets and DNA quantification kit. The pNPP tablets are soluble substr ates used for the detection of ALP activity which occurs when the pNPP reacts with ALP and generates a yellow para nitro phenol and is typically measured in umol pNPP/mL (17) ODM was aspirated from each well containing the loaded bone cores and were then sequentially washed with 1X PBS until the media from each core was removed. The samples were then left at 4 C for 30 minutes and transferred to 1.5 mL tubes. Following the transfer, 750 L of lysis buffer (0.1% Triton X 100 in distilled water) was added to each 1.5 mL tube. The samples were then frozen at 20 C until the assay was initiated. The samples were removed, thawed, and cen trifuged once the ALP Activity a ssay was ready to be performed. The pNPP solution wa s prep ared using one pNPP tablet and one Tris tablet which were added to 20 mL of distilled water and vortexed until dissolved. A 96 well plate was then prepared by adding 100 L of pNPP solution, 90 L of distilled water, and 10 L of the sample. Each sample w as plated in triplicate. In addition, a triplicate of lysis buffer alone was used to calculate the background value. Once the sample plating was complete, the 96 well plate was wrapped in foil and incubated for 30 minutes at room temperature. Optical densi ty readings were taken by a BioTek microplate reader The following equation was used to calculate the ALP Activity for each sample: umol pNPP/mL

PAGE 31

25 ALP Activities were quantified a PicoGreen Assay was performed to normalize these values to their DNA content. PicoGreen is a fluorescent nucleic acid stain for quantifying dsDNA (31) The Quant i T PicoGreen dsDNA Assay Kit (Life Technologies #P7589) was used to perform this assay. First, a 1X TE buffer was prepared by diluting 1:20 of 20X stock TE buffer in DNase free w ater. Following the TE buffer preparation, a standard curve was prepared by tak ing 2 g/mL of working solution from 100 g/mL of stock solution. Afterwards, serial dilutions were used to create curve concentrations instructions. Following the curve concentration and serial dilution process, the bone c ore samples were prepared in a 1:10 sample dilution by adding 25 L of the bone core samples to 225 L of diluent. 100 L of each sample were added in duplicate to the 96 well plate. Next, 100 L of PicoGreen working solution was added to each well containing the curve concentrations and samples. The 96 well plate was wrapped in foil, incubated for 5 minutes, and read on the Promega Microplate Fluorometer. The readings from the curve concentrations generated a standard curve and produced a linear equation. The PicoGreen data was calculated by taking the fluorescence readings, inputting them into the linear equation, and multiplying t his value by its dilution ratio of 10. These values, which were measur ed in ng/mL, were used to normalize the ALP Activity to its DNA content by dividing the calculated ALP Activity by their respective PicoGreen values. The normalized ALP Activity values were measured in nmol pNPP/mg

PAGE 32

26 Histology The histology and immunohis tochemistry methods were adapted from bone cores were fixed in formalin for 5 days decalcified in Immunocal for approximately 3 weeks processed by a Leica TP 1020 Tissue Processor, and mounted in cassettes using paraffin 5 m sections were cut along the longitudinal plane using a Microm HM 355 S rotary microtome. The 5 m tissue sections were placed on 25x75x1 mm micro slides (VWR) following hydrati on in a tissue flotation water ba th (VWR). The micro slides were then dried on a hotplate (Leica) at 45C for approximately two hours. Following the histological sectioning, the slide s were stained with H&E and osteocalcin. H&E stainings for tissue evaluat ion were adapted from the Pay ne Regenerative Orthopedics lab O steocalcin stainings were performed based on the protocols from abcam In addition, a DAB HRP substrate (VectorLabs) was used within the osteocalcin staining for antigen labeling, which produced a brown color. After each staining protocol was performed, t he stained tissue samples were then evaluated and imaged u sing a Nikon H600 L microscope. Statistical Analysis A one way ANOVA and Tu key post hoc analysis was used. All analyses were performed with SigmaPlot v.11.2. Significant differences are reported with a p value < 0.05.

PAGE 33

27 CHAPTER III RESULTS Stress Strain Curves for the Bone Cores Figure 3. 1 shows the stress strain curve for Core 5, which was typical across all tested bone cores. In this ex a mple, the max stress was 1.9631 N /mm 2 while the highest strain at the max stress was 0. 04623 mm/mm The stress strain values were used to generate a range of d ata for the group of tested bone cores. The median, minimum, and maximum values of the strains at maximum stress (mm/mm) were 0.0388, 0. 0 335, and 0.0874 respectively. In addition, t ables 3.1 and 3 .2 exhibit the overall testing p rofiles of the bone cores, which show the calculated densities, maximum strains, and maximum stress values. A correlation could not be determine d between the densities and the strain at max stress or max stress values. T he minimum value of the strains at max str ess was used to determine if the bioreactor could handle this type of loading due to the stiffness of the material. After discussions and reviewing the literature, it was de termined that a strain of 0.1% ( 10 micr ons) should be used. This was estimated to g enerate a sufficient load on the bone cores yet it would not be significant enough to compromise the bioreactor. In each of the runs conducted throughout this experiment, the strain of 0.1% (10 microns) remained constant

PAGE 34

28 Figure 3 .1 Example Stress Strain Curve Generated from UTM data Table 3 .1 Dimensions of bone core specimens Specimen length (mm) width (mm) X Area (mm^2) Weight (g) Density (g/mm^3) 1 9.59 4.79 18.020254 0.20 1.16E 03 2 10.73 4.46 15.62282611 0.20 1.19E 03 3 9.56 4.82 18.24668429 0.22 1.26E 03 4 10.22 4.67 17.12867001 0.17 9.71E 04 5 9.96 5.15 20.83072279 0.12 5.78E 04 6 9.89 4.56 16.33125525 0.16 9.91E 04 7 9.9 4.72 17.49741444 0.09 5.20E 04 8 10.05 4.8 18.09557368 0.14 7.70E 04 9 10.27 4.9 18.8574099 0.21 1.08E 03 10 9.63 5.09 20.34817416 0.08 4.08E 04 11 9.93 4.97 19.40004149 0.18 9.34E 04 12 9.02 5.18 21.07411768 0.12 6.31E 04 13 10.67 4.68 17.20210473 0.13 7.08E 04 14 10.11 5.15 20.83072279 0.15 7.12E 04 15 9.86 4.52 16.04599864 0.09 5.69E 04 16 9.48 5.11 20.50839538 0.14 7.20E 04 17 9.2 5.04 19.95036999 0.16 8.72E 04 18 9.65 5.19 21.15556347 0.09 4.41E 04 19 9.86 4.83 18.32247521 0.17 9.41E 04 20 9.46 4.97 19.40004149 0.17 9.26E 04 0 0.5 1 1.5 2 2.5 0 0.02 0.04 0.06 0.08 0.1 Stress (N/mm 2 ) Strain (mm/mm) Core 5

PAGE 35

29 Table 3 .2 Max strain and max stress values of each bone core specimen Specimen Strain @ Max Stress (mm/mm) Max Stress (N /mm 2 ) 1 0.036861314 0.541923549 2 0.042050326 2.640988238 3 0.056433054 6.850554216 4 0.087377691 0.791059668 5 0.0462249 1.963133993 6 0.036683519 4.253065605 7 0.037585859 1.653421429 8 0.038875622 3.696732757 9 0.043466407 5.741843687 10 0.037673936 3.869403682 11 0.033726083 7.745777247 12 0.047949002 7.628623055 13 0.038791003 5.86859001 14 0.056587537 3.826645902 15 0.050603363 0.521546953 16 0.037774262 3.023722668 17 0.037913043 5.421162618 18 0.038082902 4.356442698 19 0.033488844 9.866898325 20 0.060475687 6.764186563 Live/Dead Staining Live/dead staining wa s performed on the cores to see if cell attachment had occurred after harvesting and seeding the cores with the MSCs Figure 3.2 depicts the imaging generated from the confocal microscopy. The green fluorescence dictates the live cells while the red fluorescence indicates dead cells within each bone scaffold. As Figure 3.2 shows, from a qualitative approach, the scaffol d had relatively optimal attachment and the scaffolds primarily contained live cells.

PAGE 36

30 Figure 3.2 2D Live/Dead Imaging of seeded bone scaffolds; green (live) red (dead) Images are of two different seeded bone cores. a lamarBlue The a lamar Blue assay was conducted for the 2 nd and 3 rd runs of the experiment, which aided in understanding the overall cell viability and cell proliferation before and after the loading regimens. The loading regimen of the 2 nd run, which was loaded for 15 minutes per d ay, differed from the loading regimen of the 3 rd run, which was loaded for 60 minutes per day. Within the 2 nd run, the overall cell viability based on percent reduction was higher within the Day 0 group (35.44% 1.59% ) than Day 28 U nloaded (12.55% 1 .27% ) and Day 28 L oaded (10.43% 1.72% ) groups. The Day 0 group was significant ly higher (P<0.05) when compared with the loaded and unloaded groups, but there was no statistical significance between the loaded and unloaded groups. Although no statistical significance could be shown, the Day 28 Un l

PAGE 37

31 overall cell viability wa s higher than the Day 28 L oaded group. Figure 3.3 shows the averages and standard errors between each group. Figure 3.3 alamar Blue percent reduction readings for Day 0 seeded, Unloaded Day 28 seeded, and Loaded Day 28 seeded cores from Run 2 ( P<0.05) The alamarBlue of the 3 rd run of the experiment produced similar results to the 2 nd run. Again, the Day 0 group had more cell viabili ty with respect to percent reduction (48.77% 4.61%) and was significant ly higher (P<0.05) when compared with the loaded and unloaded groups However in this run the Day 28 loaded group (17.04% 3.73%) had greater cell viability with respect to percent reduction than the Day 28 U nloaded group (11.11% 0.70%), but no statistical significance could be generated between these two groups Figure 3.4 shows the averages and standard errors between each group. 0% 5% 10% 15% 20% 25% 30% 35% 40% Day 0 Day 28 Unloaded Day 28 Loaded % REDUCTION alamarBlue (2nd Run)

PAGE 38

32 Figure 3.4 alamar Blue fluorescence readings for Day 0 seeded, Unloaded Day 28 seeded, and Loaded Day 28 seeded cores from Run 3 (*P<0.05) ALP Activity & Pico Green ALP and Pico Green Assays were performed after the 1 st and 2 nd runs of the experiment, which e xperienced loading for 15 minutes per day (figures 3.5 and 3.6) The re sults of this assay aided in developing an understanding if the cores were experiencing early osteogenic cell differentiation through normalized ALP activity T he normalized ALP activity was generated for each group which was derived f rom the 1 st ALP activity and Pico Green assays The data, see n in figure 3.5c, shows the Day 28 Loaded group (6.32e4 9.81e3 nmol pNPP/m g) had the highest overall normalized ALP activity betw een the three measured gro ups. This normalized ALP activity analysis could point to a potential advantage with respect to osteogenic cell differentiation within the loaded group. Although there was variance between the normalized ALP activities, no statist ical significance could be generated between the three groups 0% 10% 20% 30% 40% 50% 60% Day 0 Day 28 Unloaded Day 28 Loaded % Reduction alamarBlue (3rd Run)

PAGE 39

33 a) b) c) Figure 3.5 ALP (a) Pico Green (b) and Normalized ALP activity (c) from Run 1 The 2 nd run produced similar results (see Figure 3.6) however due to issues with cell expansion only Day 28 Unl oad ed and Day 28 Loaded groups could be generated within the ALP activity and Pico Green assays. Using the same process, the normalized ALP activity was calculate d for the 2 nd run. Again, the averages of the Day 2 8 Loaded group ( 5.62e4 8.57e3 nmol pNPP/mg ) produced a higher 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 Day 0 Day 28 Unloaded Day 28 Loaded ng/mL Pico green 0 20,000 40,000 60,000 80,000 Day 0 Day 28 Unloaded Day 28 Loaded ALP/DNA (nmol pNPP/mg) Normalized ALP Activity 0.00 0.10 0.20 0.30 0.40 Day 0 Day 28 Unloaded Day 28 Loaded umol pNPP/mL ALP activity

PAGE 40

34 overall normalized ALP activity when compared with the Day 28 U nloade d group (4.17e4 2.99e3 nmol pNPP/m g). This could further lead to an assumption that osteogenic cell differentiation is occurring at a more pronounced rate within the loaded group than the unloaded group Although it can be assumed this is occurring, no statistical significance could be generated between the two groups. a) b) c) Figure 3. 6 ALP (a) Pico Green (b) and Normalized ALP activity (c) from Run 2 0 20,000 40,000 60,000 80,000 Day 28 Unloaded Day 28 Loaded ALP/DNA (nmol pNPP/mg) Normalized ALP 0 200 400 600 800 1,000 1,200 Day 28 Unloaded Day 28 Loaded ng/mL Pico Green 0 20,000 40,000 60,000 80,000 Day 28 Unloaded Day 28 Loaded ALP/DNA (nmol pNPP/mg) Normalized ALP

PAGE 41

35 Hematoxylin & Eosin and Osteocalcin Histology The H&E stainings gave indications of new tissue development after 28 days between the Load ed and Unl oad ed groups, w hich are depicted in figure s 3.7 and 3.8 respectively Within the Load ed group, new tissue can be seen at both the intersecti on site and along the outside of the scaffolds. Figures 3.7a and 3.7c are representative of the 15 minute Loaded group, while 3.7e and 3.7g are representative of the 60 minute Loaded group. An overal l depiction of the Loaded scaffold s at 2x zoom is given within these figures To further understand the new tissue development 4x images were taken to enhance the resolution at the intersections of the scaffolds w hich can be seen in figures 3.7b & 3.7d (15 m inute Loaded group) and 3.7f & 3.7h (60 minute Loaded group). Qualitatively, it appears that slightly more tissue developed along the outside of the scaffolds within the 15 minute Loaded group than the 60 minute loaded group. Although new tissue prevalence was greater overall in the 15 minute Loaded group, both groups exhibited a relatively noteworthy amount at the points of intersection. This could potentially validate the possibility of more pronounced osteogen ic development with in mechanically stimulated environments

PAGE 42

36 a) b) c) d)

PAGE 43

37 e) f) g ) h) Figure 3.7 H&E staining s of the 15 minute Loaded at 2x (a,c) & 4x (b,d) H&E stainings of 60 minute loading at 2x (e,f) & 4x (g,h). 4x images show the area of intersection between the two cores.

PAGE 44

38 Similar to Figure 3.7, Figu re 3. 8 gives an overal l depiction of the Unloaded group at 2x (3.8a & 3.8c) and 4x zoom (3.8b & 3.8d) Based on the H&E stainings between the Loaded and Unloaded groups, f rom a st rict observation standpoint, the Unloaded group did not show as much new tissue W hen compared with the 15 minute and 60 minute Loaded groups it appears that there was slightly less tissue developed at the points of intersection in the Unloaded group In addition, it could also be proposed that more tissue developed along the edges of the scaffolds within the 15 minute Loaded group than the Unloaded group. This observational analysis could further validate the assumption that less tissue, and therefore less potential induced osteogenesis was produced within the Unloaded scaffolds The assumptions of increased osteogenic potential within the Loaded groups could be further validated with osteocalcin stainings.

PAGE 45

39 a) b) c) d) Figure 3.8 H&E staining s of Unloaded bone cores at 2x (a,c) and 4x (b,d) 4x images show the area of intersection between the two cores.

PAGE 46

40 To further expand on the results of the histology within the Load ed and Unl oad ed groups an osteocalcin stain was performed Osteocalcin is a late stage marker in osteogenesis and should provide additional information with respect to the formation of tissue within the bone scaffolds. The stainings within the Load ed and Unloaded groups are provided by fi gures 3.9 and 3.10, resp ectively. The images were taken at 2x zoom in order to provide an overall depiction of the scaffolds Based on the overall depictions of the scaffolds it does not appear that much, if any, difference can be deduced. Blindly looking at the Loaded and Unload ed groups, it appears that osteocalcin is prevalent throughout each scaffold, which could point to the possibility of osteogenesis. However, it is important to note that residual proteins could have been left within the bone scaffolds after the washing pro cess, which could have led to a misrepresentation of the stain. Because of this occurrence, it is difficult to assess whether or not osteocalcin is being expressed or if it is a result of lingering proteins.

PAGE 47

41 a) b) c) d) Figure 3.9 Osteocalcin staining s of the 15 minute Load ed bone cores at 2x (a ,b ) and 60 minute Loaded bone cores at 2x (c ,d).

PAGE 48

42 a) b) Figure 3.10 Osteocalcin staining of the Unloaded bone cores at 2x (a ,b)

PAGE 49

43 CHAPTER IV CONCLUSIONS Beginning with in vivo and limited in vitro studies, the information available was limited with respect to developing an in vitro bone healing testing apparatus using bone allografts. The assays performed within this experiment provided insight into the future implementation of tools and techniques needed to optimize this system. Through the information gathered from live/dead, AlamarBlue, normalized ALP Activity, H&E, and osteocalcin stainings, an understanding of the potential osteogenic processes could be formulated. The liv e/dead stainings showed that cells had attached and survived when seeded onto the scaffolds, which was important to show before loading regimens began. It was interesting comparing the Day 0 groups with the Loaded and Unloaded groups, because there was a s ignificant difference in their cell viability. The cells at the beginning of the experiment were sho wing greater metabolic activity than the cells at the end of the experiment. This difference could show that a majority of the cells after 28 days in ODM had been removed during media exchanges and were therefore less likely to express relatively high measurements in metabolic activity This reasoning is further validated by the PicoGreen assay, which shows a decrease in DNA content between Day 0 and Day 28 In future studies, it would be advantageous to measure cell viability at multiple time points to understand these variances. Measuring at intermediate times of 0, 3, 7, 14, 21, and 28 days will provide a general overview of when cells may be lost and wh at is occurring throughout the experiment.

PAGE 50

44 W hen comp ared with the Day 0 and Unl oaded groups, the L oaded group s had an overall greater output with respect to normalized ALP activity Although this difference was not significant, this trend could point to the notion of greater potential osteogenesis being experienced within the loaded group. The difference in normalized ALP expression could have been a product of the dynamic environment within the Loaded group. Based on in vivo and in vitro osteogenesis res earch, this was to be expected due to the enhanced environment mechanical stimulation provides for cellular communication and differentiation. Again, measuring at intermediate time s of 0, 3, 7, 14, 21, and 28 days co uld provide additional information in un derstanding the differences between the loaded and unloaded groups The H&E and osteocalcin stainings provided an in depth understanding of the differences between the scaffolds, as a qualitative assessment could determine what is happening at the tissue and cellular levels Within the Loaded groups, there appeared to be a greater amount of new tissue development at the intersections and around the scaffolds. While the new tissue development was greater, the osteocalcin stainings could not provide a qualitative difference betw een the loaded and unloaded groups. Although osteocalcin is expressed within each group, the information is not conclusive enough to point to greater osteo genesis within the Loaded group Unfortunately, the lack of difference in the late stage osteogenic m arker after 28 days of loading could not provide further insight into development of new bone. Although the two bone scaffolds did not fuse together, which would ultimately define bone healing, the results enhanced the current understanding of an in vitr o

PAGE 51

45 bone healing model and how it can be improved with additional implementations. Because this type of in vitro analysi s is in its infancy, there are solutions and dire ctions that can be implemented such as including missing cellular variables. Osteoclasts, which are part of the hematopoietic cell lineage, play a significant role in the restructuring and remodeling of bone due to their communication with osteoblasts and osteocytes. Introducing osteoclasts into the model through co culturing and perfusion cou ld generate micro blood vessels and enhance the modeling of in vitro bone healing. In addition to the inclusion of osteoclasts recent studies have shown that co culturing mesenchymal stem cells with osteoblasts and osteocytes can influence osteogenesis wi thin in vitro models (18). On the cellular level, more research will need to be conducted in order to understand what co culturing environment will be optimal for enhancing osteogenic potential in vitro While enhancing the cellular environment of the mo del could be beneficial, adjusting the loading regimens of the model could ameliorate the bone healing system. For instance, increasing the load incrementally each day from 0.1% to 0.2% strain may produce greater osteogenic results, which was validated wit hin the in vivo rabbit study (11). Although the custom bioreactor was limited to only implementing one bout per day it would be advantageous to experiment with multiple bouts per day, as the increase in bouts per day proved to be relatively successful in vivo with respect to bone healing variables Implementing more test groups through the inclusion of cellular and mechanical variables will aid in understand ing which variables have the greatest influence on osteogenesis. Furthermore it will be advantage ous to find other assay s

PAGE 52

46 outside of the H&E and osteocalcin staining s, which can better delineate the differences between the Loaded and Unloaded groups with respect to late stage osteogenesis. Previous in vivo studies had included Mason Goldner Trichrome, Mason Goldner Modified Trichrome entachrome stainings to give added dimension to their histological evaluations. With these stainings implemented, they were able to decipher between osteoblasts, osteocytes, new b one f ormations, and new osteoids According to th ese comprehensive evaluations these stainings have greater detailed information with respect to bone healing and its various cell types In addition, these stainings can also provide quantitative tools f or compa ring and contrasting overall osteogenic markers which this study did not provide (13). The enhan cement of osteogenesis in the in vitro bone healing model can be complex due to the cellular communication, extra cellular environments, and precise mechanica l stimulations needed. Moving forward, it will be important to maintain a nd expand upon the dynamic environment for in vitro bone modeling in addition to including co cultures. The results of this analysis should be used to optimize future studies and give a strong indication that mechanical stimulation is an important facto r when discussing bone healing. While much knowledge still needs to be attained, the insight provided within this study can have a significant impact in future analyses and can hopefully generate enhancements within the bone graft, bone graft substitute, and bone biologic markets.

PAGE 53

47 REFERENCES 1) Report on Bone Health and Osteoporosis: What It Means To You. U.S. Department of Health and Human Services, Office of the Surgeon General, 2012. 2) Ulstrup, Anton K. "Biomechanical Concepts of Fracture Healing in Weight bearing Long Bones." Acta Orthop 74.3 (2008): 291 302. Web. 3) Bab, Itai A., and Jona J. Sela. "Cellular and Molecular Aspects of Bone Repair." Principles of Bone Regeneration (2012): 11 41. Web. 4) Burr, D.b., A.g. Robling, and C.h. Turner. "Effects of Biomechanical Stress on Bones in Animals." Bone 30.5 (2002): 781 86. 5) Robling AG, Burr DB, Turner CH ( 2001 ) Recovery periods restore mechanosensitivity to dynamically loaded bone J. Exp. Biol 204 3389 3399 6) Turner C. Toward a mathematical description of bone biology : The principle of cellular accomodation. Calcif Tissue Int 1999 ; 65 : 466 471. 7) Rubin, C. T., Ph.D., & Lanyon, L. E., M.R.C.V.S, Ph.D. (1984). The Journal of Bone and Joint Surgery. Regulation of bone formation by applied dynamic loads,66 (3), 397 402. 8) adap tive bone remodeling. J Biomech 15:767 781; 1982. 9) Pan W, Einhorn T. The biochemistry of fracture healing. Curr Orthop ( 1992 ) ; 6 : 207 213. 10) Gardner, Michael J., MD et al. "Pause Insertions During Cyclic In Vivo Loading Affect Bone Healing." Clin Orthop Relat Res (2008): 1232 238. 11) Shadmehr, Azadeh, Ali Esteki, Gholam R. Oliaie, Giti Torkaman, and Alireza Sabbaghian. "Augmentation of Bone Healing by Specific Frequency and Amplitude Compressive Strains." Orthopedics 32.3 (2009): 1 6. 12) Gardner, Michael J., MD et al. "In Vivo Cyclic Axial Compression Affects Bone Healing in the Mouse Tibia." Journal of Orthopaedic Research J. Orthop. Res. 24.8 (2006): 1679 686. 13) Rentsch, Claudia, Wolfgang Schneiders, Suzanne Manthey, Barbe Rentsch, and Stephan Rammelt. "Comprehensive Histological Evaluation of Bone Implants." Biomatter 4.1 (2014). Web.

PAGE 54

48 14) Grayson, Warren L., Sarindr Bhumiratana, Christopher Cannizzaro, P. H. Grace Chao, Donald P. Lennon, Arnold I. Caplan, and Gordana Vunjak Novakovic. "Effects of Initial Seeding Density and Fluid Perfusion Rate on Formation of Tissue Engineered Bone." Tissue Engineering Part A 14.11 (2008): 1809 820. Web. 15) Villanueva, I., D.s. Hauschulz, D. Mejic, and S.j. Bryant. "Static and Dynamic Compre ssive Strains Influence Nitric Oxide Production and Chondrocyte Bioactivity When Encapsulated in PEG Hydrogels of Different Crosslinking Den sities." Osteoarthritis and Cartilage 16.8 (2008): 909 18. Web. 16) AlamarBlue Cell Viability Reagent. (n.d. ). Retrieved January 30, 2017, from https://www.thermofisher.com/order/catalog/product/DAL1025 17) SIGMAFAST p Nitrophenyl phosphate Tablets. (n.d.). Retrieved January 30, 2017, from http://www.sigmaaldrich.com/catalog/product/sigma/n1891?lang=enion=US &cm_sp=Insite prodRecCold_xorders prodRecCold2 1 18) Birmingham, E., Niebur, G., Mchugh, P., Shaw, G., Barry, F., & Mcnamara, L. (2012). Osteogenic differentiation of mesenchymal stem cells is regulated by osteocyte and osteoblast cells in a simplified bone niche. European Cells and Materials,23 13 27. doi:10.22203/ecm.v023a02 19) A. Alex Jahangir, MD; Ryan M. Nunley, MD; Samir Mehta, MD; Alok Sharan, MD; and the, Washington Health Policy Fellows. Bone graft subsitutes in orthpaedic surgery [Internet]. AAOS; 2008 Jan. Available from: http://www.aaos.org/news/aaosnow/jan08 /reimbursement2.asp 20) Brooks, B. D., Sinclair, K. D., Grainger, D. W., & Brooks, A. E. (2015). A Resorbable Antibiotic Eluting Polymer Composite Bone Void Filler for Perioperative Infection Prevention in a Rabbit Radial Defect Model. PLoS ONE 10 (3), e0118 696. http://doi.org/10.1371/journal.pone.0118696 21) Research, T. M. (2015, September 07). Bone Grafts and Substitutes Market expected to reach USD 3.48 Billion Globally in 2023: Transparency Market Research. Retrieved March 24, 2017, from http://globenewswire.com/ne ws release/2015/09/07/766429/10148433/en/Bone Grafts and Substitutes Market expected to reach USD 3 48 Billion Globally in 2023 Transparency Market Research.html#sthash.kpAvDPgn.dpuf 22) Kinaci A Neuhaus V Ring D Trends in Bone Graft Use in the United States ORTHOPEDICS 2014; 37: e 783 e 788. doi : 10.3928/01477447 20140825 54

PAGE 55

49 23) Roberts, T. T., & Rosenbaum, A. J. (2012). Bone grafts, bone substitutes and orthobiologics: The bridge between basic science and clinical advancements in fracture healing. Organogenesis 8 (4), 114 124. http://doi.org/10.4161/org.23306 24) Oryan A ., Alidadi S ., Moshiri A ., & Maffulli N Bone regenerative medicine : classic options novel strategies and future directions Journal of Orthopaedic Surgery and Research (2014). 9 18. http :// doi org /10.1186/1749 799 X 9 18 25) Rubin, C. T. and Lanyon, L. E. Regulation of bone formation by applied dynamic loads. J Bone Jt Surg 66 A (1984). 397 402 26) Golub, E. E., & Boesze Battaglia, K. (2007). The role of alkaline phosphatase in mineralization. Current Opinion in Orthopaedics,18 (5), 444 448. doi:10.1097/bco.0b013e3282630851 27) Sharmin, F., Adams, D., Pensak, M., Dukas, A., Lieberman, J., & Khan, Y. (2015). Biofunctionalizing devitalized bone allografts through polymer mediated short and long term growth factor delivery. Journal of Biomedical Materials Research Part A,103 (9), 2847 2854. doi:10.1002/jbm.a.35435 28) Delaine Smith, R. M., & Reilly, G. C. (2012). M esenchymal stem cell responses to mechanical stimuli. Muscles, Ligaments and Tendons Journal 2 (3), 169 180. 29) Wagner DR, Lindsey DP, Li KW, Tummala P, Chandran SE, Smith RL, et al. (2008). Hydrostatic pressure enhances chondrogenic differentiation of human bone marrow stromal cells in osteochondrogenic medium. Ann Biomed Eng 36 (5):813 820. 30) Liu J, Zhao ZH, Li J, Zou L, Shuler C, Zou YW, et al. (2009) Hydrostatic Pressures Promote Initial Osteodifferentiation With ERK1/2 Not p38 MAPK Signaling Involved. J Cell Biochem 107 (2):224 232. 31) Quant from https://www.thermofisher.com/order/catalog/product/P7589

PAGE 56

50 APPENDIX A Figures Figure 2.1 ................................ ................................ ................................ ................... 20 Figure 2.2 ................................ ................................ ................................ ................... 21 Figure 2.3 ................................ ................................ ................................ ................... 22 Figure 3.1 ................................ ................................ ................................ ................... 28 Table 3.1 ................................ ................................ ................................ ..................... 28 Table 3.2 ................................ ................................ ................................ ..................... 29 Figure 3.2 ................................ ................................ ................................ ................... 30 Figure 3.3 ................................ ................................ ................................ ................... 31 Figure 3.4 ................................ ................................ ................................ ................... 32 Figure 3.5 ................................ ................................ ................................ ................... 33 Figure 3.6 ................................ ................................ ................................ ................... 34 Figure 3.7 ................................ ................................ ................................ ................... 37 Figure 3.8 ................................ ................................ ................................ ................... 39 Figure 3.9 ................................ ................................ ................................ ................... 41 Figure 3.10 ................................ ................................ ................................ ................. 42