Citation
Bradykinin B1/B2 antagonist linked to a human neutrophil elastase inhibitor

Material Information

Title:
Bradykinin B1/B2 antagonist linked to a human neutrophil elastase inhibitor a heterodimer for the treatment of inflammatory bowel disease
Portion of title:
Heterodimer for the treatment of inflammatory bowel disease
Creator:
Leimer, Axel H
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
xi, 79 leaves : illustrations (1 folded), forms ; 29 cm

Subjects

Subjects / Keywords:
Inflammatory bowel diseases -- Treatment ( lcsh )
Leucocyte elastase ( lcsh )
Inflammatory bowel diseases -- Treatment ( fast )
Leucocyte elastase ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 76-79).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Basic Science
Thesis:
Department of Chemistry
General Note:
Department of Medicine
Statement of Responsibility:
by Axel H. Leimer.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
37756966 ( OCLC )
ocm37756966
Classification:
LD1190.L44 1996m .L45 ( lcc )

Downloads

This item has the following downloads:


Full Text
BRADYKININ B1/B2 ANTAGONIST LINKED TO A HUMAN NEUTROPHIL
ELASTASE INHIBITOR; A HETERODIMER FOR THE TREATMENT OF
INFLAMMATORY BOWEL DISEASE.
by
Axel H. Leimer
B.A., University of Colorado at Denver, 1993
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Basic Science
1996


1996 by Axel H. Leimer
All rights reserved.


The thesis for the Master of Basic Science
degree by
Axel H. Leimer
has been approved
by
/ /ru
j Da


Leimer, Axel H. (M.B.S., Basic Science)
Bradykinin B1/B2 Antagonist linked to a Human Neutrophil Elastase
Inhibitor; a Heterodimer for the Treatment of Inflammatory Bowel
Disease.
Thesis directed by Professor Douglas F. Dyckes.
ABSTRACT
Inflammatory bowel disease (IBD) is a common disorder in developed
countries. Although it has some of the characteristics of a general
inflammatory disorder, its etiology is still unknown. Because inflammatory
disorders typically have a very complex etiology and involve a multitude of
mediators a multi-active drug approach is proposed. In certain chronic
inflammatory conditions, like IBD, we believe that bradykinin (BK) and
human neutrophil elastase (HNE) are co-operatively involved. The present
study describes single compounds designed to incorporate HNE inhibitory
and BK B1/B2 antagonist activity. A proprietory elastase inhibitor (CP-955)
was directly linked via amide bond formation to the BK antagonist B-9430
(D-Arg-Arg-Pro-Hyp-Gly-L-lgl-Ser-D-lgl-Oic-Arg-OH), which incorporates
both BK B1 and B2 antagonist activity and two analogs designed
IV


specifically for this study. The three resulting conjugates synthesized,
were purified and analyzed for in vitro antagonist activity and HNE
inhibition. For all three compounds (compound I, II and III), B2 receptor
binding in human cloned receptors was reduced tenfold. In a B2 antagonist
functional assay on guinea pig ileum, compounds I and II had equivalent
potency to B-9430. The BK B1 antagonist activity was significantly less
than that of B-9430 for all three compounds. Compound I showed a
fourfold improvement in the inhibition constant against human neutrophil
elastase, as compared to CP-955, while compounds II and III were not
inhibitory. This study clearly shows that it is possible to retain BK B2
antagonism and HNE inhibition by combining CP-955 and B-9430.
This abstract accurately represents the content of the candidate's thesis.
I recommend its publication.
Signed
v


DEDICATION
This work is dedicated to Trisha, Stefan, Jessica and Christian.


ACKNOWLEDGMENT
My thanks to my thesis committee for its support and advice throughout this
project. Special thanks to Heather B. Kroona for her direct supervision of
my laboratory activities and continued encouragement.
Additionally, I want to thank Douglas L. Gernert for the synthesis of the
CP-955 used in this study and Lyle W. Spruce for the initial synthesis of
CP-955 and our discussions on the 2-indaneglycine synthesis. Thanks are
also due to John M. Asztalos for amino acid analysis and Steven K. Wolk
for assistance with Nuclear Magnetic Resonance work on 2-indaneglycine.
I am also grateful to John S. Zuzack and Michael R. Burkard for their work
with the pharmacological assays, Sherman E. Ross for obtaining elastase
inhibition constants, and Val S. Goodfellow for the molecular modeling.
For the many years of previous research on bradykinin antagonists and
their utilization in heterodimers I am indebted to John C. Cheronis, Eric T.
Whalley and John M. Stewart.


CONTENTS
Chapter
1. Introduction............................................... 1
1.1 Inflammatory Bowel Disease................................ 1
1.2 Human Neutrophil Elastase................................. 2
1.3 Bradykinin Antagonists as Therapeutics....................4
1.4 Rationale of Drug Design..................................7
1.5 Structure-Activity Relationship of B-9430................. 8
2. Experimental............................................... 12
2.1 Materials and Instrumentation............................. 12
2.2 Methods................................................... 15
2.2.1 Amino Acid Analysis...................................... 15
2.2.2 2-lndaneglycine Synthesis............................... 16
2.2.2.1 2-Bromoindane.......................................... 16
2.2.2.2 Ethyl-a-acetamido-a-cyano-2-indaneacetate............. 18
2.2.2.3 (D,L)-2-lndaneglycine................................. 19
viii


2.2.2.4 A/-Acetyl-(D,L)-2-indaneglycine..........................20
2.2.3 Resolution of (D,L)-2-lndaneglycine........................ 21
2.2.4 GITC-Derivatization........................................ 27
2.2.5 Te/t-butyloxycarbonyl-lgl.................................. 29
2.3 B-9430 Analogs................................................ 31
2.3.1 Solid Phase Peptide Synthesis...............................31
2.3.2 Cleavage of Peptide from Resin and Post-cleavage Workup. 35
2.3.3 Purification and Characterization.........................36
2.4 Conjugation of CP-955 to the B-9430 Analogs ................. 42
2.5 Biological Assays............................................45
2.5.1 HNE Inhibition Assay.......................................45
2.5.2 Bradykinin B2 Guinea Pig Ileum Receptor Binding Assay..... 45
2.5.3 Bradykinin B2 Human Receptor Clone Binding Assay.......... 46
2.5.4 Bradykinin B1 Human Lung Fibroblast Receptor
Binding Assay.............................................. 46
2.5.5 Bradykinin B2 Antagonist Functional Testing on Guinea Pig
Ileum.....................................................46
3. Results and Discussion....................................... 48
3.1 Chemistry....................................................48
3.2 Biology.....................................................51
ix


Appendix
A. Animal Use Certification..................................... 64
B. Animal Use Protocol...........................................65
References....................................................... 76
x


FIGURES
Figure
1.1 Bradykinin................................................ 5
1.2 B-9430.................................................... 6
1.3 Amino Acids in B-9430.....................................6
1.4 The Human Neutrophil Elastase Inhibitor CP-955...............8
2.1 2-lndaneglycine Synthesis.................................. 17
2.2 A/-Acetyl-(D,L)-2-indaneglycine.............................21
2.3 Resolution of (D,L)-2-lndaneglycine........................ 23
2.4 Proton NMR of (L)-2-lndaneglycine.......................... 24
2.5 Magnitude COSY Spectrum of (L)-2-lndaneglycine............. 25
2.6 A/-Acetyl-(D)-2-indaneglycine...............................27
2.7 GITC-Derivatization of (L)- and (D)-2-lndaneglycine.........28
2.8 Terf-butyloxycarbonyl-lndaneglycine........................ 30
2.9 Purification of B-9430..................................... 40
2.10 Mass Analysis for B-9430................................... 41
2.11 Mass Analysis of Compound 1.................................44
XI


3.1 Effect of Dimethyl Sulfoxide on BK Response Curve.52
3.2 Effect of B-9430 on BK Response Curve.............54
3.3 Effect of Compound I on BK Response Curve.........55
3.4 Effect of Compound II on BK Response Curve........56
3.5 Effect of Compound III on BK Response Curve.......57


TABLES
1.1 Compounds Synthesized and Tested........................ 10
2.1 Peptides Synthesized by Solid Phase Peptide Synthesis....31
2.2 Amino Acid Analysis of Stempeptide...................... 34
2.3 Yields of Peptide Synthesis..............................37
2.4 Characterization of Peptides.............................37
2.5 Mass Analysis of Peptides............................... 39
2.6 Mass Analysis of Conjugates..............................43
3.1 Summary of Biological Data...............................59


1. INTRODUCTION
1.1 Inflammatory Bowel Disease
Chronic inflammatory bowel disease (IBD) is defined as a spectrum
of inflammatory bowel disorders with overlapping clinical, epidemiological
and pathological findings, but without defined etiology1. Patients with IBD
can be divided into two groups, Crohns Disease (CD, Regional Enteritis,
Granulomatous Ileitis or Ileocolitis) and Ulcerative Colitis (UC), although
differentiation is often difficult and many subgroups based on symptoms
exist. Etiology of IBD is mostly unknown and the disease manifests itself
through chronic diarrhea, abdominal pain, fever, anorexia, perianal fistulae,
and inflammation of the gastrointestinal tract. Inflammatory bowel disease
is a vascular disease with the typical symptoms of inflammation, including
transendothelial migration of leukocytes into inflamed tissue and
subsequent tissue damage. In particular for CD, the only effective therapy
is often the removal of the inflamed section of the bowel for temporary
alleviation of the symptoms1. Postoperative recurrence is between 60 and
95%1.
1


As part of the inflammatory response, white blood cells such as
polymorphonuclear neutrophils (PMNs) and macrophages invade the tissue
of the bowel and release various proteinases, including human neutrophil
elastase (HNE), which cause general edema.
Current drug therapy is limited to general immunomodulators,
aminosalicylates and steroids2. Unfortunately, many drugs in use,
especially those indicated for UC, exacerbate diarrhea and are generally
not very effective therapies. The potential drug therapy pursued in the
present study, was intended to inhibit the action of tissue and matrix-
degrading enzymes, in particular the action of elastase released from the
neutrophil. The attempt was made to incorporate a human neutrophil
elastase inhibitor (HNEi) and a BK antagonist into a single multi-active
drug.
1.2 Human Neutrophil Elastase
Although the exact etiology of IBD is still unknown, PMNs are
thought to play an important role in the disease. It is therefore helpful to
2


look at the role of PMNs in the model of general inflammation.
In inflammation, various factors, including lipopolysaccharides (LPS),
tumor necrotizing factor alpha (TNF-a), and interleukin 1 (IL-1), cause
increased expression of certain classes of membrane glycoproteins on
endothelial cells (EC)3 of postcapillary venules4. These membrane proteins
act as receptors for ligands expressed on blood cells, in particular PMNs,
promoting PMNs adherence to the EC. The PMNs are generally the first
cells to bind to the activated EC and extravasate (i.e. migrate across
endothelial layer into tissue). The activated EC express receptors of the
selectin family (E-, P-, and L-selectin), which interact with carbohydrate
ligands on the PMNs. Transient binding of the PMNs to the EC occurs,
resulting in 'rolling' of the PMNs over the EC surface. Most of this rolling is
mediated by E- and P-selectins on the EC and their corespondent ligand on
the PMNs, which is Sialyl Lewis X (SLewx; NeuAca2-3Gal31-4[Fuca1-
3]GlcNAc)5. Platelet Activating Factor (PAF) expressed on the EC
activates the PMNs during this rolling, causing an increase of adhesion of
the PMNs, corresponding to an increased expression of the integrin
receptor leukocyte function-associated antigen-1 (LFA-1). The LFA-1
binds to intercellular adhesion molecules (ICAMS) on the EC and holds the
PMNs attached to the EC despite the shear forces of the blood flow5. Now
3


extravasation occurs5. In acute inflammation, the leukocyte infiltrate is
composed of 90-95% PMNs. Following extravasation, the PMNs are
activated and self-destruct, releasing the contents of intracellular granules.
Most of the damage caused by the PMNs is due to elastase and toxic
radicals released in this respiratory burst. The recruitment of PMNs
reaches a maximum about 4 hr after the initial insult, after which
macrophages begin to accumulate at the site.
Synthetic, low-molecular-weight HNEi are of great interest to many
pharmaceutical companies. Indications for HNEi include inflammatory
diseases like Adult Respiratory Distress Syndrome (ARDS), Myocardial
Reperfusion Injury, Rheumatic Arthritis, Cystic Fybrosis, Emphysema, and
Inflammatory Bowel Disease.
1.3 Bradykinin Antagonists as Therapeutics
Bradykinin (BK) research began as early as 1909, when French
surgeons observed a fall in blood pressure following injections of human
urine fractions9. Because kinins are associated with the regulation of the
cardiovascular system and are mediators of inflammation and pain6,7,8, the
4


present study incorporates a BK antagonist into a multi-active drug. The
pharmacology of BK has been described in detail elsewhere9,10.
It is thought that BK B2 receptors are expressed constitutively in
most animals, while BK B1 receptors are induced during chronic
inflammation. However, it is not yet known what role the BK B1 receptor
plays in pain and inflammation.
Bradykinin is a nonapeptide which binds to the BK B2 receptor
(Figure 1.1).
Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg-OH
123456789
Figure 1.1: Bradykinin
Upon the action of various enzymes (Carboxypeptidases, Kininase I,
etc.), BK is converted to the [desArg9]-BK, which is the ligand for BK B1
receptors. A key synthetic transformation occurs when L-Pro7 (agonist) is
replaced with D-Phe7 (antagonist). Many hundreds of peptides have been
synthesized in extensive structure-activity relationship studies in the search
for more potent antagonists9. In the past, heterodimers, consisting of two
5


linked peptides with BK B1 and BK B2 antagonist activity respectively, have
been designed to block both receptors with one compound11.
Compound B-9430, a single decapeptide developed in Professor
John Stewarts laboratory at the Health Science Center of the University of
Colorado, is a BK B2 and B1 antagonist (Figure 1.2). This peptide is
unique because it has BK B1 antagonist activity without being a [desArg9]-
BK derivative and because it is the first to combine BK B1 and B2
antagonist activity in one peptide.
D-Arg-Arg-Pro-Hyp-Gly-L-lgl-Ser-D-lgl-Oic-Arg-OH
01 234 56 789
Figure 1.2: B-9430
In Figure 1.2 Hyp is 4-trans-hydroxyproline, Igl is 2-indaneglycine,
and Oic is octahydro-indole-2-carboxylic acid (Figure 1.3).
6


It has been found that the affinity for the BK B1 receptor generally
increases by either removing the arginine residue in position 9 (as in
[desArg9]-BK) or by adding a positive charge at the amino terminus (lysine
or arginine). While extensive structure activity relationships have been
reported previously for BK B1 and B2 agonists and antagonists12,13,14,15, it is
unclear to what extent the results can be applied to B-9430.
1.4 Rationale of Drug Design
Given the importance of bradykinins and proteinases, in particular
HNE, in inflammation, the decision was made to link a proprietary elastase
inhibitor to B-9430, a combination BK B1/B2 antagonist, which would yield
a heterodimer with possible dual or triple activity. Because of the
complexity of inflammatory disorders, it is hoped that a multi-active drug
would show increased efficacy in treatment. Additionally, dual or triple
activity could increase the number of indications for which the drug might
be used, and thus, make the drug more marketable. Finally, as both
bradykinin receptors and elastase are important and perhaps essential
targets in anti-inflammatory drug design, the combined approach allows for
the development of one rather than two or three drugs. Cortech has
7


developed several HNE inhibitors to date. The present study involved the
use of HNEi CP-955 (Figure 1.4), which was developed in collaboration
with Marion Merrill Dow.
HO
o
Human Neutrophil Elastase Inhibitor
Figure 1.4: The Human Neutrophil Elastase Inhibitor CP-955.
The carboxylic acid moiety in CP-955 serves mainly to increase
solubility. When bound to the active site of HNE, the carbonyl oxygen of an
ester moiety extends into the oxy-anion binding pocket of the enzyme. The
carboxylic acid is believed to extend from the active site cleft out into bulk
solvent in the unprimed direction16 (internal communications).
1.5 Structure-Activity Relationship of B-9430
Several analogs of B-9430 were investigated for their retention of
8


both BK B1 and B2 antagonist activity. CP-955 was linked directly to
various positions of the three B-9430 analogs via amide bond formation.
Previous studies (unpublished data, John Asztalos, 1996) have shown that
directly linking a HNEi to a substrate or polymeric carrier, even in sterically
crowded environments, does not interfere with its activity (internal
communications, 1995). This can be explained by the fact that HNE has a
very shallow substrate binding-pocket, which allows short extensions of the
substrate or inhibitor, in particular those toward the unprimed end, to
extend into bulk solvent. Prior to this study, no data were available on the
effects of conjugating another molecule to B-9430. It was thus unknown
whether a B-9430 containing conjugate would still be able to bind to BK
receptors. Since a complete SAR study was not within the scope of the
present study, three different types of conjugates were synthesized based
on reasons outlined below. The three compounds synthesized and tested
are shown in Table 1.1. For compound I, CP-955 was conjugated via its
carboxyl group to the N-terminal amino group (D-Arg) of B-9430 resulting
in an amide bond linkage and a linear conjugate. Conjugation was directed
by the use of orthogonal protecting groups as described in the Methods
section. Direct conjugation, without linkers, was chosen to keep the overall
size of the molecule to a minimum. In compound II, an N-terminal
9


extension consisting of a lysyl residue (position -1) preserved the number
of positive charges on the peptide upon conjugation, which may be
important in binding to the BK B1 receptor.
Table 1.1 Compounds Synthesized and Tested
-1 01234 56 789 D-Arg-Arg-Pro-Hyp-Gly-L-lgl-Ser-D-lgl-Oic-Arg-OH CP-955 linked at D-Arg N-term. amine. (I)
H-Lys-D-Arg-Arg-Pro-Hyp-Gly-L-lgl-Ser-D-lgl-Oic-Arg-OH CP-955 linked at Lys'1 N-term. amine. (II)
D-Lys-Arg-Pro-Hyp-Gly-L-lgl-Ser-D-lgl-Oic-Arg-OH CP-955 linked at D-Lys N-term. (I")
Compound III was designed to be essentially the same as
compound I except that the D-Arg was replaced by a D-Lys residue. This
substitution preserved the overall charge, while helping to elucidate the
importance of the Arg residue in the zero position.
Optimal positions for linking BK antagonists to other molecules with
retention of antagonist activity have previously been explored at Cortech17.
10


These studies involved the linking together of two antagonists by various
linking moieties for the express purpose of generating a dual-action
heterodimer. Although no linkers were used in the present study, the data
from the previous Cortech study yielded valuable information regarding
those positions that could be altered in a typical BK B1 or B2 antagonist
without significant loss of activity. This data base served as the starting
point for the design of the new conjugates explored here. Specifically,
linking at positions 0 and 1 has been shown to be optimal for preserving BK
B1 antagonism in [des Arg9]-BK. Linking at those same positions generally
does not reduce BK B2 antagonism of BK B2 antagonist peptides
significantly25. Linking to the amino acids in positions 5 and 6 has been
shown to severely impair the ability of the peptide to bind to the BK B1
receptor25. For these reasons the HNEi CP-955 was linked to the
N-terminus of the B-9430 peptide or analog in all three compounds. The
resulting three compounds differed only in positions -1 and 0 and could
therefore be synthesized from a common 'stem peptide'.
Although the data from previous Cortech studies guided the design
of the conjugates synthesized here, it was uncertain at the outset how well
these data would extrapolate to the behavior of B-9430, which is not a
11


[desArg9]-BK B1 antagonist. Furthermore the part of the B-9430 peptide
that actually confers the ability to bind to the BK B1 receptor is not known.
It was therefore unclear whether BK B1 antagonism could be preserved
and, additionally, whether BK B1 antagonism would be beneficial in any
drug indicated for inflammatory disease. However, both of these issues
constitute pressing questions for BK-antagonist research and drug design,
and their answers, be they positive or negative, are of importance.
2. EXPERIMENTAL
2.1 Materials and Instrumentation
All chemicals were purchased from Sigma, St. Louis, MO, unless
otherwise indicated.
Analytical Reversed-Phase HPLC (RP-HPLC) was performed on a
Hewlett Packard Series 1050 instrument using a Vydac C-18 column (0.46
12


x 15 cm); particle size 5 pm. All chromatography was monitored at both
254 and 214 nm. Solvent A was water containing 0.1% trifluoroacetic acid
(TFA, v/v) as a modifier. Solvent B was acetonitrile (CH3CN) containing
0.1% TFA (v/v).
Preparative RP-HPLC was performed on a Waters PrepLC 4000
System using a Vydac C-18 column (5 x 25 cm), particle size 5 pm, and a
Waters 486 detector. All chromatograms were obtained 'off-peak' at 230
nm to ensure that the absorbance signals were on scale. Solvents A and B
were the same as described for the analytical HPLC.
Nuclear Magnetic Resonance (NMR) spectroscopy was performed
on a Varian Gemini 300 MHz instrument by the author. Additional NMR
work (COSY) was done by Dr. Steven K. Wolk at Cortech. The magnitude
COSY spectrum of (D)- and (L)-2-indaneglycine was acquired on a Bruker
DRX400 (400MHz) NMR spectrometer using the proton channel of an
inverse detection probe. The standard pulse sequence (Ernst et al., 1987)17
was used, along with two 90 pulses (12.4 msec), a sweep width of 3453.0
Hz in each dimension, 4096 complex points in t2, and 512 t1 values (32
scans for each). The data were apodized with sine bell apodization
13


functions, and zero-filled in the t1 dimension prior to the 2D Fourier
transformation to yield a 2048x2048 data set.
Mass spectrometry was performed on a Finnigan Mat Lasermat
Mass Analyzer (Matrix Assisted Laser Desorption Ionization Time Of
Flight method, MALDI-TOF). All measurements were made with two
internal calibration standards to bracket the sample, unless otherwise
indicated. Additional mass spectra were obtained through the analytical
services of M-Scan, Inc., West Chester, PA, using Fast Atom
Bombardment Mass Spectroscopy (FAB) on a VG Analytical ZAB 2-SE
high field mass spectrometer. A cesium ion gun was used to generate ions
for the acquisition of mass spectra, which were recorded using a PDP 11-
250J data system. Mass calibration was performed using cesium iodide.
Melting points were determined using an Electrothermal 9100 (not
corrected).
Polarimetry was performed using a DigiPol Automatic Polarimeter,
Model DP1A31, by Rudolph Instruments.
14


Amino acid analysis was performed at Cortech by John Asztalos
with a Waters Pico-Tag Amino Acid Analysis System, using a single
wavelength UV detector set at 254 nm.
2.2 Methods
2.2.1 Amino Acid Analysis
Amino acid analysis was performed by John Asztalos as described
in detail elsewhere18. Briefly, the free or protected peptides (diluted
aliquots mixed with internal standard) were hydrolyzed in vacuo with vapor
phase 6 N HCI for 24 hr at 100 C, then dried, neutralized with triethylamine
and dried again in vacuo on a Pico-Tag Work Station. The sample was
then allowed to react for 30 min with phenylisothiocyanate (PITC) to form
the phenylthiocarbamyl (PTC) derivative of the amino acids. Subsequent
chromatographic analysis was performed by C-18 RP-HPLC, detecting
peaks at 254 nm. Peak area integration with Baseline software (Waters)
was compared to standards run separately, and the internal standard in the
analysis sample was used to calculate percent recovery. The standard
Pico-Tag 1 gradient was extended isocratically for 3 min to recover the
15


PITC derivative of 2-indaneglycine (Igl).
2.2.2 2-lndaneglycine Synthesis19 (Figure 2.1)
2.2.2.1 2-Bromoindane
To 2-indanol (1, 200 g, 1.48 mol) in pyridine (30 mL, 0.37 mol) and
chloroform (648 mL) at -15 C was added phosphorus tribromide (160 mL,
1.69 mol) over 45 min. The reaction mixture was stirred at room
temperature for 24 hr. The mixture was then extracted with 450 mL
chloroform and 500 g ice. The organic layer was washed with water (2 x
400 mL) and dried over anhydrous sodium sulfate (Na2S04). The solvent
was evaporated in vacuo leaving a brown semi-solid. The product was
distilled in vacuo and fractionated, yielding 121 g of a colorless liquid (2-
bromoindane19, 2, 121 g, 0.61 mol, 41.2%, bp 92-95 C at 5 mmHg, nD21 =
1.6).
16


Figure 2.1 2-lndaneglycine Synthesis
17


2.2.2.2 Ethyl-a-acetamido-a-cyano-2-indaneacetate
To sodium ethoxide (40 g, 0.6 mol) suspended in 500 ml_ dry
dimethyl sulfoxide (DMSO) a solution of ethylacetamidocyanoacetate (100
g, 0.57 mol) in 500 ml_ dry DMSO was added with vigorous stirring. The 2-
bromoindane (2, 121 g, 0.61 mol) was added dropwise over a one-hour
period with continued vigorous stirring. The brown solution was stirred for
24 hr at room temperature, then the solution was stirred for 7 hr at 50 C to
drive the reaction to completion. The mixture was evaporated in vacuo and
the residue treated with 600 mL cold water and extracted with ethyl acetate
(2 x 600 mL). Then the extract was dried over magnesium sulfate (MgS04)
and evaporated in vacuo leaving a brown oil. The first recrystallisation from
water yielded a yellowish brown solid (50 g); recrystallisation and extraction
with ethyl acetate from the mother-liquor resulted in another 75 g of
yellowish white solid. Both solids were combined and recrystallized from
toluene to yield 90 g of a white solid (ethyl-a-acetamido-a-cyano-2-
indaneacetate, 3, 50.8%, mp 159-160 C19) which showed a single peak on
RP-HPLC.
(RP-HPLC, single peak, gradient 0-40%B in 40 min, retention time 12.24
min)
18


2.2.2.3 (D,L)-2-lndaneglycine
A solution containing ethyl-a-acetamido-a-cyano-2-indaneacetate
(3, 90 g, 0.32 mol) in 1600 mL of 10% sodium hydroxide (NaOH) was
refluxed for 20 hr. The solution was cooled and adjusted to pH 6.5 with
concentrated HCI and chilled. The precipitate was washed with cold water
(2 x 100 mL) and methanol (2 x 50 mL). The solid was then refluxed in
1200 mL 6N hydrochloric acid (HCI) for 12 hr. The solution was then
chilled in an ice bath and adjusted to pH 6.5 with 6N NaOH. The
precipitate was collected (14 g). Because of the low yield in the
precipitation, the remaining aqueous mixture was evaporated in vacuo
leaving a white solid mixture of 2-indaneglycine and salt. This second solid
was carefully washed with cold water to remove most of the salt. Both
solids were combined (4, (D,L)-2-indaneglycine, ~15 g, 0.066 mol, 21%,
mp 322-324 C19).
(1H-NMR in D20 and NaOD, 300 MHz; 5 7.4-7.14 (m, 4H);3.54 (d, 1H,
J= 7.5 Hz); 3.05-2.9 (m, 4H); 2.55-2.54 (m, 1H); RP-HPLC single peak,
gradient 0-40 % B in 40 min, retention time 13.24 min at 215 nm)
19


2.2.2A A/-Acetyl-(D,L)-2-indaneglycine
(D,L)-2-lndaneglycine (4, ~15 g, 0.066 mol, Figure 2.2) was
dissolved in 200 mL water and 250 mL 2N NaOH, then chilled in an ice-
water bath and stirred. Acetic anhydride (5.6 mL) was added, followed by
eight successive additions of 56 mL 2N NaOH and 5.6 mL acetic
anhydride. The pH was maintained at pH 10-11. The solution was stirred
at room temperature overnight and the progress of the reaction monitored
by RP-HPLC (on a gradient of 0-40% B in 40 min, the (D,L)-2-
indaneglycine has the retention time of 16 min and the acetylated amino
acid has retention time 24.1 min). After 48 hr the reaction was 97%
complete as judged by disappearance of the peak for the free amino acid.
The solution was then chilled and adjusted to pH 3 with 6N HCI. After
standing in the cold for several hr the precipitate was collected and washed
with a small amount of cold water. The product was suspended in water
and extracted with ethyl acetate. The organic layer was then evaporated in
vacuo leaving a slightly yellowish solid (5, A/-acetyl-(D,L)-2-indaneglycine,
12 g, 0.05 mol, 76%; single peak on RP-HPLC with gradient 0-40% B in 40
min., retention time 24 min; mp 201-202 C; 1H-NMR, in DMSO, 300 MHz;
5 8.29 (d, 1H, J= 8.25 Hz); 7.22-7.11 (m, 4H); 4.32 (m, 1H); 2.94-2.74 (m,
20


5H); 2.52 (m, 1H); 1.89 (s, 3H).
Figure 2.2 A/-Acetyl-(D,L)-2-indaneglycine
2.2.3 Resolution of (D,L)-2-lndaneglycine (Figure 2.3)
/V-Acetyl-(D,L)-2-indaneglycine (5, 12 g, 0.05 mol) was suspended in
500 mL of phosphate-buffered saline solution (PBS) at pH 7.4 and stirred at
room temperature. 50 mg porcine kidney acylase I (E.C. 3.5.1.14; 1,430
units / mg solid; 2,125 units / mg protein) was added and the pH monitored
continuously and readjusted as needed with dilute NaOH. After 5 hr
21


another 50 mg acylase I was added. The hydrolysis reaction was allowed
to run for 72 hr and then another 50 mg acylase I was added. Reaction
progress was monitored with RP-HPLC for the disappearance of the peak
corresponding to the acetylated amino acid and reappearance of the peak
at 16 min corresponding to the free Igl. After 5 days the peaks at 215 nm
integrated as 48% and 51%, respectively, and the reaction was stopped.
Specifically, 300 mL of ice water and 600 ml_ ethyl acetate were added to
the solution and the mixture was carefully acidified to pH 0.75 with 6N HCI.
The two layers were filtered through Celite and separated. The water
phase was extracted twice with ethyl acetate. The aqueous solution was
evaporated in vacuo and the residue dissolved in 30 mL water and 15 mL
6N HCI. The solution was then adjusted to pH 5.5 with concentrated
ammonium hydroxide (NH4OH). The precipitate was collected and washed
with cold water over a medium-pore glass fritt. Filtrate was again
evaporated in vacuo and the solid retained and recrystallized from water to
give white flakes (6, (L)-2-indaneglycine19, 3.8 g, 0.019 mol, 73%; mp 296-
297 C, [a]25D= 50.35, c=1; Single peak 1H-NMR, in D20 and NaOD, 300
MHz (Figure 2.4); 6 7.31-7.20 (m, 4H); 3.28 (d, 1H, J= 7.29); 3.02-2.95 (m,
2H); 2.85-2.6 (m, 3H); Two dimensional 1H-NMR in DMSO and NaOD, 400
MHz (Figure 2.5)).
22


Figure 2.3 Resolution of (D,L)-2-lndaneglycine
h3<
H
23


S'.
Figure 2.4 Proton NMR of (L)-2-lndaneglycine
(L)-2-lgl in D20 with NAOD
EXPl PULSE SEQUENCE; S2PUL
DATE 01 -21-96 17; 19: 29
SOLVENT D20
FILE H
ACQUISITION DEC. fi YT
TN 1.000 DN 1.000
SM 4S00.5 DO -450.0
AT 2.496 DM MM
NP 22464 £H> 1.0
PN S.O OLP 20
PI 0 TOKO N
D1 1.000
D2 0 PROCESSING
TO 0 NATH I
NT 64
CT 64 DISPLAY
PN90 17.0 SP -90.1
TIN N NP 3090.7
FB 2250 VS 1765
BS 64 SC 2
SS 0 NC 396
IL N IS 617
IN T RFL 1231.7
DP T RFP 0
AL0CX N TH 36
INS 1.000
*1
5

CO o
or cn
CU 03



oj m
o
cm
5 5
J




/ n.
I I I'T'I
TT
I | I I I I | I I I l | I I t I | I I
9 8
l 1 l
L-r-J
40.0
ill!
24
a

wJ

*yJ

k
Tl'l | I IT I | I I I I
3.2
t----.----1 L
3.0
TTTT
2.8
J L
TJTTTT
2.6 PPM
___I
10.2
21.4
33.3

iiM >'** i
J
T I | I I I I |
I l I I | I I I I | I I I I | I I I I p
3 2
L-r-J 1j1>-------'
21.4
*33^3
I I I T
i 1 1 |
T I I T
i i i i | i r
0 PPM
10.2
-0.0106


to
U1
Igi in 020/Na00 (TSP ref)
mag cosy, higher res
for leimer sw/md drx7 2
4/3/96
jUl
o o
0f]j


jjMK
0 fj0 o 0
0 h0 0
O0. o o
I
3 4
0
00 '
0
I
3 0
tifU'Hlll
2 a
2 6
Current Dele Perea
NAME onl
EXPNO ?
PHOCNO I
rt
H
Deli.
li*e
INS1HUM
pHoato
PU.PHOG
2 6 id
SOI VENT
NS
OS
Si
I HIV S
AO
nc
OH
oc
IE
-2 0 00
01
PI
OC
SI01
NUC!
Pi 1
I NO
Acguitilion Pereaeterk
960403
IS ?2
5 Mult inw
COS|90
4096
C0CI3
3?
3?
34S) 019 Hi
0 8410.'/ lli
0 5911,00 vet
tOl
144 800 utcc
4 /9 usee
300 0 k
0 00000)0 tec
I 00000000 sec
I? 40 usee
4 /9 usee
400 ISOS//? ten
IH
0 00 oB
0 00028960 sec
Fl Acquisition pereeeiers
NOO l
-3 0 T0 512
Si 01 400 1506 wi/
Finrvs 6 744?i6 m
Sk 6 629 oo*
3 2
SI
SF
MOD
SSB
18
Ut
PC
F? Processing pereacters
2040
400 1486405 MMi
SINE
0
0 00 Hi
U
1 00
SI
MC2
SF
Off
ssa
18
08
Processing pereaeters
2040
OF
400 1491054 MHi
SINE
0
0 00 Hi
0
-3 4
pp*
CX2
CXI
F2PiO
2l0
(2PH1
I *11
I lit U
I uo
i IPHI
Mill
rppPMCN
F2H/CM
f 11PMfM
2D NMFI
plot psrewlers
14 00 CP
14 00 ca
3 503 ppe
1401 86 Hi
2 496 ppe
9911 H9 lli
3 604 11pa
140? 06 IU
2 500 ppp
1000 37 lli
0 0/193 ppa/CP
28 78335 Hi/cp
Figure 2.5 Magnitude COSY Spectrum of (L)-2-lndaneg!ycine


The ethyl acetate phase (single peak on RP-HPLC) was dried over
MgS04 and evaporated in vacuo. The solid residue (7, A/-acetyl-(D)-2-
indaneglycine, mp 204-204.6 C) was refluxed in 150 mL 6N HCI overnight
(Figure 2.6). The solution was evaporated in vacuo and then taken up in
50 mL 6N HCI and 50 mL water, after which the pH was adjusted to pH 5.5
with concentrated NH4OH and a white precipitate collected (RP-HPLC
>98% pure at 215 nm, gradient 0-40% B in 40 min, retention time 15.6 min,
small impurity at 14.25 min/254 nm). The remaining liquid was adjusted to
pH < 1 and extracted with ethyl acetate. The ethyl acetate was evaporated
in vacuo and the white solid collected (RP-HPLC same as previous
precipitate). Both solids (flakes) were combined and analyzed (8, (D)-2-
indaneglycine, 3.5 g, 0.017 mol, 67%; [a]25D = -22.32, c=1 in 1N HCI; 1H-
NMR, in DMSO and NaOD, 300 MHz; 6 7.3-7.2 (m, 4H); 3.28 (d, 1H, J= 7.3
Hz); 3.012-2.95 (m, 2H); 3.85-2.6 (m, 3H)). Although, the rotation was
found to be in the opposite direction, it was not of equal intensity as that of
the L-lgl, which may in part be explained by the small amount of L-lgl
contamination in the sample (see next section).
26


Figure 2.6 A/-Acetyl-(D)-2-indaneglycine
6 N HCI
reflux 8 h.
(7)
2.2.4 GITC-Derivatization (Figure 2.7)
The chiral agent 2,3,4,6-tetra-O-acetyl-p-D-glucopyranosyl-
isothiocyanate (9, GITC, 2.9 mg, 7.4 pmol, FLUKA) was dissolved in 1.5
ml_ CH3CN. D-lgl or L-lgl (5 mg, 0.025 mmol) was dissolved in 10 mL
acetonitrile/water/triethylamine (50:50:0.4; v/v/w). GITC (50 pL, 0.25 pmol)
was added to 50 pL of each Igl solution(0.125 pmol), the solutions were
vortexed and aliquots analyzed via RP-HPLC on a gradient of 10-50% B in
40 min. A racemic mixture of Igl had been derivatized and separated
previously on RP-HPLC to establish resolution. The D-lgl derivative (10,
27


Figure 2.7 GITC Derivatization of (L)- and (D)- 2-lndaneglycine. Both, (L)-
and (D)-lgl, were derivatized with GITC and injected onto analytical HPLC
separately and as a mixture. Frame A shows the mixture, which was run
as a standard to establish separation. Frame B shows the chromatogram
of (L)-lgl and frame C that of (D)-lgl.
(6.8)
28


retention time 33.0 min) contained <3% of L-lgl as a contaminant and the
L-lgl derivative (retention time 31.7 min) contained no detectable D-lgl.
2.2.5 7erf-butyloxycarbonyl-lgl (Figure 2.8)
The tert-butyloxycarbonyl (Boc) group was introduced according to
the method of Moroder et al.20. D-lgl (8, 3 g, 0.013 mol) dissolved in a
mixture of 26 ml_ dioxane, 13 mL water and 13 mL 1N NaOH was stirred in
an ice-water bath. Di-fert-butylpyrocarbonate (3.12 g, 0.0143 mol) was
added and stirring continued for 0.5 hr at room temperature. The reaction
was monitored via RP-HPLC by appearance of a peak at 40.4 min on a
gradient of 0-40% B in 40 min. When the reaction was judged to be -80%
complete, the reaction mixture was concentrated in vacuo to -13 mL,
cooled in an ice-water bath and covered with a layer of ethyl acetate (30
mL). The mixture was then acidified to pH 2-3 with 6N HCI. The phases
were separated and the aqueous phase was extracted twice with ethyl
acetate. The ethyl acetate extracts were pooled, washed twice with 30 mL
water each, dried over anhydrous Na2S04 and then evaporated in vacuo to
give a white solid (11,2,2 g Boc-D-lgl; single peak on RP-HPLC, >99%
29


purity, for both; 1H-NMR, in DMSO, 300 MHz; 5 7.2-7.1 (m, 4H); 3.6 (s,
1H); 2.9-2.75 (m, 5H); 1.41 (s, 9H)). The Boc-L-lgl (12) was prepared in
the same manner (2,3 g; single peak on RP-HPLC, >99%purity; 1H-NMR, in
DMSO, 300 MHz; 5 7.19-7.1 (m, 4H); 3.59 (s, 1H); 2.94-2.74 (m, 5H); 1.4
(s, 9H)).
Figure 2.8 Te/t-butyloxycarbonyl-lndaneglycine
+ co-.
30


2.3 B-9430 Analogs
2.3.1 Solid Phase Peptide Synthesis
Table 2.1 shows the peptides which were synthesized using
standard Boc chemistry.
Table 2.1 Peptides Synthesized by Solid Phase Peptide Synthesis
B-9430 D-Arg-Arg-Pro-Hyp-Gly-L-lgl-Ser-D-lgl-Oic-Arg-OH
Peptide A H-Lys-D-Arg-Arg-Pro-Hyp-Gly-L-lgl-Ser-D-lgl-Oic-Arg-OH
Peptide B D-Lys-Arg-Pro-Hyp-Gly-L-lgl-Ser-D-lgl-Oic-Arg-OH
The N-terminal amino group of peptide A and the e-amino group of
D-Lys in peptide B were protected with a fluorenylmethoxycarbonyl
(FMOC) group. The FMOC /V-protecting group was chosen because it is
stable to HF-cleavage and thus allows the directed coupling of CP-955 to
the remaining unprotected amine following HF-cleavage.
All three peptides were completed from the common 'stempeptide'
Boc-Pro-Hyp-Gly-L-lgl-Ser(0-Bzl)-D-lgl-Oic-Arg(Tos)-PAM-resin, according
to Merrifield as described by Stewart and Young21. Amino acids (except D-
and L-lgl, which were prepared as described above), resin and coupling
31


reagents were purchased from Advanced ChemTech and Bachem
California. All solid phase chemistry was performed manually in a 250 ml_
glass reaction vessel (ACE Labglass) with a medium-pore glass fritt at the
bottom. Mixing was achieved by bubbling nitrogen gas in from the bottom
of the reaction vessel. The Boc-Arg(Tos)-Pam-resin (5 g, 0.44 mmol/ g,
total 2.2 mmol) was placed in the reaction vessel and washed four times
with dimethylformamide (DMF) and methylene chloride (DCM, 10-15 min
each), and the solvents then removed. Deprotection was carried out by
adding 80 ml_ deprotection reagent, 50% trifluoroacetic acid (TFA) in DCM,
and mixing for 2 min and then repeating with fresh deprotection reagent
while mixing for 30 min. Following deprotection, the resin was washed
extensively (3 x DCM, 3 x DMF, 3 x DCM; 80 ml_ for 30 sec each) and then
neutralized with 10% /V/V-diisopropylethylamine (DIEA) in DCM (3 x 50 mL
for 1 min each). The resin was again washed (3 x DCM, 2 x DMF, 2 x
20%, 1-methyl-2-pyrrolidinone [NMP] in DMF; 80 mL for 30 sec each) and
a small resin sample was removed from the reaction vessel for a qualitative
ninhydrin test according to Kaiser, et al.27. After a positive ninhydrin test for
free amines the preactivated amino acid (Boc-Oic) was added in DMF and
coupling allowed to proceed overnight (all subsequent couplings were
allowed to proceed for 4-8 hr). All couplings were carried out with a 3-fold
32


molar excess of activated amino acids (6.6 mmol).
Activation of amino acids employed the carbodiimide method
described previously28. Diisopropylcarbodiimide (DIC, 2mL per coupling,
7.7 mmol) with addition of A/-hydroxybenzotriazole (HOBT, 1.2 g, 7.7
mmol) were dissolved in 20 mL 20% NMP in DMF, and the amino acid (6.6
mmol) was added for in situ activation. After 20 min, the mixture was
added to the resin and another 30 mL of 20% NMP in DMF was added.
After each coupling step the resin was washed (3 x DMF, 3 x DCM; 80 mL
for 30 sec each) and a ninhydrin test performed. With a negative ninhydrin
test the deletion peptides were capped by reaction with acetic anhydride
(10% acetic anhydride, 10%DIEA, 80% DCM; v:v:v) for 10 min. In fact, no
double couplings were required. The peptide resin was then washed and
deprotected as described above and the cycle repeated for the next amino
acid.
Following the coupling of the prolyl residue, the stem-peptide on
resin was analyzed for amino acid composition (Table 2.2).
33


Table 2.2 Amino Acid Analysis of Stem- Peptide
Amino acid Solution cone, (pico ole) Calculated ratio Theoretical ratio
Arg 580.092 0.6 1
Hyp 1168.396 1.2 1
Ser 1034.503 1.08 1
Gly 1115.232 1.16 1
Pro 1385.629 1.44 1
ABA 234.407 N/A >99%
Oic 873.913 .91 1
igi 1531.18 1.59 2
TOTAL: 7688.945 7.98 8
Table 2.2 shows the calculated and theoretical ratios of amino acids.
Ratios were calculated as follows: The total amounts in picomoles
detected are summed (here 7343.26 pmol), excluding contributions from
background peaks and breakdown products. The sum was divided by the
total number of amino acids expected to be present in an integral amount
per peptide sample. Then the picomole amount found for each amino acid
was divided by n times the average integral amount, where n is equal to the
number of times the particular amino acid is theoretically present in the
34


peptide. The result was subsequently compared to the theoretical value.
The results in Table 2.2 show a large deviation from the expected value in
the case of proline and indaneglycine. As these data were obtained
directly from a peptide-resin sample, without prior cleavage or removal of
protecting side chain groups, some deviation may be expected due to
increased experimental error. The stem peptide was split into four batches
(-0.5 mmol each) from which the three final peptides were made. After the
coupling of the final amino acid each to each peptide, the N-terminal Boc
group was removed as described above. All three peptides were then
again analyzed as peptide-resins by amino acid analysis (data not shown).
The amino acid compositions found for the final peptide-resins on resin
were in good agreement with the theoretical composition of the peptides.
2.3.2 Cleavage of Peptide from Resin and Post-cleavage Workup
Cleavage of the peptides from the resin was achieved with
anhydrous HF and anisole (10:1, v:v) at 5C for 2 hr, according to Stewart
and Young21. Next, the cleavage mixture was filtered through a sintered
glass fritt and the residue washed with small amounts of glacial acetic acid
35


in water (30:70 v:v). Cold ethyl ether was added to the filtrate until the
peptide precipitated. The precipitate was collected on a sintered glass fritt
and washed with additional cold ethyl ether. The solid was redissolved in
water and small amounts of glacial acetic acid (10-25% by volume), and
the solution filtered in preparation for HPLC purification. In each case, the
peptide solution was then immediately taken onto purification.
2.3.3 Purification and Characterization
All peptides were purified by RP-HPLC as described in Materials and
Methods section. Gradients for the crude extracts were optimized with
small aliquots on the analytical system prior to loading onto the preparative
HPLC system. Fractions collected were checked by analytical RP-HPLC
for single peak content and by mass spectral analysis (MS) for compound
identification. Fractions deemed to be identical based on mass and
retention time (on RP-HPLC) determination of the dissolved solute were
pooled and lyophilized. Final yields are shown in Table 2.3. The yield for
B-9430 was expected to be higher compared to peptides A and B, as its
retention time was known. Although peptide A was only 95% pure as
36


Table 2.3 Yields of Peptide Si /nthesis
Compound Actual Yield (mg) % Purity by RP-HPLC (at 215 nm)
B-9430 116.8 >98%
Peptide A 7 95%
Peptide B 13.6 >99%
Table 2.4 Characterization of Peptides
B-9430 Peptide A Peptide B
Amino acid Calculated ratio Theoretical ratio Calculated ratio Theoretical ratio Calculated ratio Theoretical ratio
Hyp 1.21 1 1.06 1 1.18 1
Ser 1.05 1 0.86 1 0.94 1
Gly 1.05 1 0.97 1 0.99 1
Arg 3.33 3 2.96 3 2.16 2
Pro 0.72 1 0.82 1 0.94 1
Oic 0.59 1 0.81 1 0.86 1
Lys 0 0 0.74 1 0.93 1
igi 2.05 2 1.79 2 1.99 2
Total 10 10 11 11 10 10
37


judged by HPLC peak integration, it was not repurified due to the small
amount of material recovered. Amino acid analysis (Table 2.4) and mass
spectral analysis (Lasermat and FAB, Table 2.5) were obtained for each
final peptide. Amino acid analysis for B-9430 showed that the proline and
octahydro-indole-2-carboxylic acid ratios were significantly lower than
expected. Since B-9430 had been characterized by coelution with a fully-
characterized standard (Figure 2.9; standard provided by Ved Srivastava)
and the mass was found to be in good agreement with the theoretical value
(see Table 2.5), the deviation is believed to be due to experimental error or
incomplete hydrolysis. The lysine value of peptide A was also found to be
low. All three peptides, B-9430, peptide A and peptide B, were subjected
to mass spectrometry using both, LASERMAT (MALDI-TOF) and fast atom
bombardment techniques (Table 2.5). The results from both techniques
confirmed the expected M+H mass within experimental error for both
peptides A and B. Both instruments had been calibrated as described in
the Materials and Instrumentation section above. Figure 2.9 shows the
chromatograms and data obtained in the purification of B-9430. The mass
analysis data for B-9430 are shown in Figure 2.10 (here only one standard
at mass 1600.2 was used for calibration).
38


Table 2.5 Mass Analysis of Peptides
Compound Found Mass LASERMAT Found Mass FAB Theoretical Mass + H
B-9430 1341 1339 1338.74
Peptide A 1690.3 1689 1690.2
Peptide B 1534.5 1533 1534.8
39


Figure 2.9 Purification of B-9430


Figure 2.10 Mass Analysis for B-9430


2.4 Conjugation of CP-955 to the B-9430 Analogs
CP-955 (20 mg, 45 pmol), which had been synthesized and
characterized by Douglas L. Gernert, was activated with O-benzotriazole-
A/.A/./V./V-tetramethyl-uronium-hexafluorophosphate (HBTU, 20.5 mg, 54
pmol; Advanced ChemTech) and A/,A/-diisopropylethylamine (DIEA, 12 pL,
67 pmol) for 30 min. Peptide A (13, 30 mg, 22.4 pmol) was added and the
reaction mixture was stirred overnight at room temperature. The reaction
product was then precipitated with 15 ml_ cold diethyl ether, centrifuged
and the resulting supernatant decanted. The pellet was resuspended in 50
mL 30% acetic acid in water and loaded onto a preparative RP-HPLC
column. The fractions collected were checked by both analytical RP-HPLC
and mass spectral analysis (Lasermat). Fractions 6 and 7 eluting at -67%
B, both displayed a single peak, with identical retention time, and a M+H
mass of 1768.6 (theoretical 1768.5). Both fractions were pooled and
lyophilized. The final lyophilized material was again analyzed by RP-HPLC
(compound I, single peak, gradient 1-100% B in 25 min, retention time
15.3 min, Figure 2.11) and submitted for mass spectral analysis with FAB
(M+1 found: 1768, plus Na peaks, Figure 2.11). The total yield was 24.5
mg ( 62 %, 14 pmol).
42


Peptides A and B were coupled to CP-955 in the same manner as
described for B-9430, using the same molar equivalents. The resulting
compounds were labeled compound II and compound III, respectively. The
results for all three conjugates are listed in Table 2.6. Masses found for
compounds I, II and III using both, the LASERMAT (MALDI-TOF) and fast
atom bombardment techniques (not showing M+sodium peaks) indicate
that the molecular ions were in agreement with theoretical calculations.
Purity determination was made by RP-HPLC monitored at 215 nm and was
based on the integrated area under peaks as a percentage of total peak
area. Since only 2.4 mg were recovered for compound II, no attempt was
made to repurify the material.
Table 2.6 Mass Analysis of Conjugates HPLC- purity Yield
Compound Found Mass LASERMAT Found Mass FAB Theoretical Mass +H at 215 nm (mg)
Compound I 1768.6 1768 1768.9 95% 24.5
Compound II no data 1896 1896.6 85% 2.4
Compound III 1740 1740 1741 95% 3.9
43


3-OCT-95 ll:28B:H:<2
Figure 2.11 Mass Analysis of Compound I
44


2.5 Biological Assays
2.5.1 HNEi Assay
The inhibition constants for human neutrophil elastase were
determined by Sherman E. Ross using the method described by
Cunningham et al.22 with some minor modifications.
(All pharmacological assays following hereafter were performed by
John S. Zuzack and Michael R. Burkard, Department of Pharmacology,
Cortech, Inc.)
2.5.2 Bradykinin B2 Guinea Pig Ileum Receptor Binding Assay
This assay was performed with binding assay conditions similar to
those previously described (Manning et al., 1986)23. Guinea pig ileum (gpi)
membranes were prepared by Analytical Biological Services, Inc.
45


2.5.3 Bradykinin B2 Human Receptor Clone Binding Assay
This assay was performed as described previously24, using a human
bradykinin B2 receptor, expressed in CHO-K1 (ATCC) cells.
2.5.4 Bradykinin B1 Human Lung Fibroblast Receptor Binding Assay
This assay was performed as described for the BK2 human receptor
clone binding assay24 with some modifications. Human lung fibroblasts
IMR-90 cells (ATCC) were used, which express the BK1 human receptor.
The cells were incubated with IL-1P (200 pg / mL) for 3 hr. Tritiated
desArg10 kallidin was used as the ligand.
2.5.5 Bradykinin B2 Antagonist Functional Testing on Guinea Pig
Ileum
Tissues were prepared by a method similar to that described by
Cheronis et al.25. Briefly, 1cm long strips of ileum from male Hartley guinea
pigs (350-400g; SASCO) were prepared as previously described. One end
46


of each strip was tied to a support rod and the other to a Grass FT03 force
displacement transducer. Tissues were allowed to equilibrate in siliconized
tissue baths for 45 min under 2 g resting tension at 37C in gassed (95%
02 and 5% C02) Kreb's solution. The tissues were initially primed with 1
pM BK followed by a 45-min wash period. The potency of the B2 agonist,
BK, was measured by recording the contractile response to a range of BK
concentrations and was expressed as the EC50 (the concentration
producing half maximal contraction). To determine antagonist potencies
cumulative BK concentration-response curves were repeated in the same
tissue prior to and following a 15-min exposure to the antagonist. Three
antagonist concentrations were tested in the same tissue. Data were
calculated as percent of the maximum contraction in the control curve.
Shifts in the ECS0 were used to calculate the pA2 (negative log of the molar
concentration of antagonist causing a two-fold increase in the agonist EC50)
for the antagonist by the method of Schild (1947)26. As a control, the effect
of DMSO, which was used to solubilize the antagonists, on the BK
response curve was established.
47


3. RESULTS AND DISCUSSION
3.1 Chemistry
The overall yields in the synthesis of (D,L)-2-lgl (with exception of
the final step) were similar to those cited in the literature19. Due to the
relatively low yield in synthesizing the 2-bromoindane using PBr3 an
alternative route may be explored in the future. The resolution of the (D,L)-
2-lgl gave good results, although the time required for the reaction to go to
completion was longer than expected. It is likely that much of the initially
added acylase denatured because of the pH change that accompanied the
generation of acetic acid. A fairly large volume of a buffer with an ionic
strength similar to physiological conditions would have to be used to
completely buffer such a change in pH. High-ionic-strength buffer may be
a another solution. Although other synthetic routes using chiral reagents
may be used to arrive at optically pure isomers, the reagents required are
typically much more expensive than the acylase. Additionally, the acylase
is an ideal enzyme for future scale up of the method, as it can be covalently
bound to solid resins without significant loss of activity29,30.
48


Chirality assignment for the L and D enantiomers of Igl was based
on the enzymatic reaction. It was assumed that Acylase I preferentially
hydrolyzes L-amino acid derivatives31. The fact that complete resolution
was achieved, as shown by the GITC-derivatization, further supports the
assignment because incomplete resolution would have been expected if
the enzyme was not completely specific for the L-enantiomer. Further
support comes from the fact that the B-9430 standard made in this study
had the expected activity. All BK B1/2 antagonists explored to date require
an aromatic D-amino acid in position 7. An inadvertent exchange between
the L-lgl5 and D-lgl7 would likely result in an inactive peptide. To date no
absolute assignment of the enantiomers has been made in published
studies.
To facilitate more complete assignments in the proton NMR
spectrum of (L)-2-indaneglycine, a COSY spectrum was obtained (in
DMSO and NaOD, 400 MHz; Figure 2.5). A COSY spectrum is a
two-dimensional NMR spectrum in which the one-dimensional spectrum
lies along the diagonal of a two-dimensional frequency map, and an
off-diagonal cross-peak at (a)a, cob) indicates that resonances at (oa and a)b
(i.e., the diagonal peaks at (coa, u)a) and (a)b, d)b), respectively) are
49


J-coupled. A magnitude COSY spectrum is symmetric about the diagonal.
J-coupling is usually only observed between protons that are two or three
bonds apart. Hence, this information can be used to determine which
structural groups are adjacent in the molecule. The off-diagonal cross-
peak (3.28,2.65) in Figure 2.5 shows coupling between the a-proton
(3.28,3.28) and the sextet (2.65,2.65), which can then be assigned as the
3-proton. The remaining two multiplets can be assigned to the y-protons
which are cis and trans to the 3-proton. Due to the chirality of the amino
acid and the limited rotation of the bulky side chain, all four y-protons are
magnetically nonequivalent. An absolute assignment to match coupling
constants to the cis and trans y-protons cannot be made from this data.
For such an assignment a NOESY experiment would be required.
The standard B-9430, compound I and compound III were obtained
in final purities of at least 95% as judged by HPLC peak integration of
absorbance peaks obtained at 215 nm. Compound II was only -85 %
pure, as judged by HPLC. Compound II was synthesized from the two
principal components, Peptide A and CP-955, which were 95% and 99%
pure respectively (by HPLC). The retention time of the contaminant
50


associated with compound II did not match the retention times of either
unreacted CP-955 or Peptide A. The contaminants are therefore believed
to be due to coupling of CP-955 at the side chain of the D-Arg residue
rather than at the N-terminal amine. This is further supported by the fact
that only one mass-ion peak was found in the MS, which corresponds to
compound II.
3.2 Biology
As DMSO was used to make stock solutions of the various
compounds tested in the BK B2 antagonist functional assay on gpi, a
DMSO dose response curve was established (Figure 3.1). The DMSO
concentrations here are the same as the final DMSO concentrations in the
samples tested for activity. The first two sets, 0.01% DMSO (by volume)
for 1 hr and 0.1% DMSO for 2 hr, did not significantly change the dose
response curve as compared to the control with no DMSO. The highest
concentration, 1% for 3 hr, did result in a depression of the dose response
curve, which reflects a decrease in the percent of maximum contraction.
The same effect can be seen with all compounds tested, as the maximum
51


Figure 3.1 Effect of DMSO on BK Dose Response Curve.
Effect of DMSO on bradykinin dose response curve in guine
ileum
c
o
o
g
c
o
o
E
3
E
X
o
E
o
Control
-.01%. 1 hr
-.1%. 2hr
-1%. 3hr
Recovery
------Series7
------Series8
Series9
Series 10
-----Series 12
-----Seriesl3
- Seriesl4
Series! 5
-----Seriesl 7
-----Series 18
- Seriesl 9
- Series20
- Serles22
- Series23
52


contractile response is reduced at the highest concentrations. This is
quantitatively expressed as a significant reduction of the slope in the Schild
plots of all compounds. A slope of 1, which corresponds to the retention of
the maximum contractile response, is interpreted as reversible competitive
binding32. A reduction of the slope in the Schild plot, corresponding to a
reduction of maximum contractile response achieved at high
concentrations, usually indicates noncompetitive binding32.
In the case of the compounds discussed below, the slope reduction
is most likely due to the effect of DMSO, and the binding is still believed to
be of reversible, competitive nature. In support of this interpretation is the
fact that B-9430, which has previously been shown to be a competitive
ligand33, using an aqueous solvent, had the most severe reduction of
maximum contraction in this study (Schild plot slope 0.67 0.11; Figure
3.2). The pA2 values for compounds I and II (Figures 3.3 and 3.4
respectively) are essentially the same as that of the B-9430 standard,
within experimental error. The pA2 values are 8.320.38 for B-9430, 7.95
0.25 for compound I, 8.39 0.24 for compound II, and 7.62 0.28 for
compound III (Figure 3.5). This means that, in this functional assay,
compound I retained full activity on the BK B2 receptor compared to
53


Figure 3.2 Effect of B-9430 on BK Dose Response Curve
Effect of B-9430 on BK does response curve in
guinea pig ileum
-log brcdykJnin ccnc. (M)
-20.00
Control
-1.0CE-07
-1.0CE-C6
- 1.C0E-G5 '
Recovery i
- Sertes7
-SeriesS
Series?
Senes 10
-Series 12
-Series 13
Series 14
Series 15
-Series! 7
-Series 18
Series 19
Series20
- Series22
- Series23
54


Figure 3.3 Effect of Compound I on BK Dose Response Curve
Effect of Compound I on BK does response curve in
guinea pig ileum
u
g
c
o
o
E
3
6
X
O
£
o
tf
Contra
1.0CE-07
1.0CE-06
1.CCE-C5
Recovery
Series7
SeriesS
Series?
Series 10
-Series! 2
-Sertesl3
Series! 4
Series 15
-Series 17
-Series 18
Seriesl9
Series20
-Series22
-Series23
Schild plot of Compound I antagonism of bradykinin effect in
guinea pig ileum
55


Figure 3.4 Effect of Compound II on BK Dose Response Curve.
Control
Effect of Compound II on BK does response curve in guinea pig ileum o 1.00E-07 ; 1.00E-06 1 o 1.0CE-O5 i
n-3 p r t r 100.00 Recovery' Series?
/ / - 80.00 Series8
/ / / 2 C Series?
/ | if - 60.00 g Senes 10
lit O o 40.00 | Seriesl2 Sertesl3
/ /" / J £ X a - Sertesl4 Senesl5 '
20.00 1 O ** o ro Seriesl7 i Series! 8 '
' ^ ^ 2 - Seriesl9 1
12.00 11.00 10.00 9.00 8.CC 7.C0 6.00 5.00 4.00 J- -20.00 -log bradylu'nin eonc. (M) - Sertes20 Sertes22 Sertes23 1
56


Figure 3.5 Effect of Compound III on BK Dose Response Curve.
Effect of Compound III on BK does response curve in
guinea pig ileum
log bradykinln cone. (M)
a
a
c
o
a
E
3
e
x
O
£
o

Control
o l.OCE-07
1.00E-06
-o 1.00E-05
a - Recovery'
----Series7
----Ser!es8
Series9
Series 10
-Seriesl2
-Series 13
Series! 4
Series 15
-Series 17
-Series 18
Series 19
Series20
-Series22
- Series23
Schild plot of Compound III antagonism of bradykinin effect in
guinea pig ileum
57


B-9430 and compound II had equivalent activity. The pA2 for compound III
was lower by approximately a half logarithmic unit, suggesting some loss of
activity, although it is not known what activity is inherent in the parent
peptides for compounds II and III.
The pKb values (pKb = -log{[antagonist concentration] / (IC50
antagonist) / (IC50)} ) were determined because of the observed depression
of the contractile response. The pKb values confirm that the activity was
essentially retained in compounds I (pKb=7.64 0.08) and II (pKb=7.97
0.08), with a slight improvement in compound II, as compared to the B-
9430 standard (pKb=7.58 0.10). Compound III (pKb=6.83 0.15) again
showed a small decrease in activity. Table 3.1 is a summary of the
biological data obtained on the compounds, including receptor binding and
functional assay data, as well as the HNE inhibition constants.
In the BK B1 receptor binding assay B-9430 had a plC50 of 8.07,
whereas compound I was significantly less active with a plC50 of 6.74.
Compounds II and III had plC50 values of 7.04 and 6.09, respectively. In
the BK B2 receptor binding assay, the B-9430 standard had a plC50 of
9.64, and plC50 values of 8.42 for compound I, 8.00 for compound II, and
58


Table 3.1 Summary of Biological Data
Receptor binding assay Guinea pig functional assay Inhibition I
Compound BKB1 (human) BK B2 (human) BK B2 (gpi) B2(gpi) functional pA2 B2(gpi) functional pKb Schild's slope HNE Ki I (nM) 8
B-9430 8.07 9.64 9.13 8.32 0.38 7.58 0.10 0.67 0.11 N/A I
-1 01234 56 789 Q-Arg-Arg-Pro-Hyp-Gly-L-lgl-Ser-Q-lgl-Oic-Arg-OH (1) CP-955 linked at Arg N-term. amine. 6.74 8.42 7.67 7.95 0.25 7.64 0.08 0.87 0.10 2.4 |
-1 01234 56 789 H-Lys-Q-Arg-Arg-Pro-Hyp-Gly-L-lgl-Ser-Q-lgl-Oic-Arg-OH (II) CP-955 linked at Lys' e-amine. 7.04 8.00 7.49 8.39 0.24 7.97 0.08 0.82 0.08 not I active D
-1 01234 56 789 B-Lys-Arg-Pro-Hyp-Gly-L-lgl-Ser-Q-lgl-Oic-Arg-OH (III) CP-955 linked at Lys0 N-term. amine. 6.09 7.10 6.75 7.62 0.28 6.83 0.15 0.61 0.09 not D active 1
i CP-955 N/A N/A N/A N/A N/A N/A 10.5 I


7.10 for compound III. Although in the receptor binding study all
conjugates had more than one logarithmic unit less activity as compared to
B-9430, it is important to note here that activity was higher against the
human clone BK B2 receptor compared to the guinea pig ileum BK B2
receptor with all three conjugates. Compound I has the highest retention of
BK B2 antagonism in receptor binding.
Compound II had the highest retention of BK B1 antagonism. This
correlates to the observation made in other studies that retention of BK B1
activity is directly linked to retention of the positive charges on the N-
terminal end of the peptide. Here, the Lys'1 extension provides the
functional group for conjugation, thus preserving the positive charges on
the arginine residues in positions 0 and 1. The receptor-binding data
indicate that there is a general selectivity for the BK B2 receptor over the
BK B1 receptor, especially in the case of compound I, which has nearly a
hundredfold selectivity.
In the BK B2 functional assay, all three conjugates essentially had
the same activity as B-9430, with a possible slight improvement seen with
compound II. Since, functionally, the BK B2 antagonism of the conjugates
60


on gpi was equivalent to B-9430, compared to severe losses in the BK B2
gpi receptor assay, it would be of interest to know what the functional
human data for BK B2 might look like. Further testing, especially of
compound I, in a human functional assay is therefore indicated.
Compound I showed a fourfold improvement of the inhibition
constant toward neutrophil elastase, resulting in a Kj of 2.4 nM compared to
a Kj of 10.5 nM for CP-955. Similar improvements have been seen in other
studies when HNE inhibitors were conjugated to various molecules (Axel H.
Leimer, internal communications, 1996). Several studies have shown that
extensions on the P unprimed end of HNE inhibitors have little effect on the
K| (internal communications, 1996). These extensions are thought to be in
solution, outside the shallow binding-pocket of HNE. There is no obvious
explanation for the loss of HNE inhibition with compounds II and III, in
particular for compound III, which differs from compound I only by the D-
arginine0 to D-lysine0 substitution. In both compounds the inhibitor is linked
at the N-terminus. Preliminary modeling of these compounds and energy
minimization in the elastase binding pocket gave no obvious indication as
to why these compounds might be inactive toward the enzyme (modeling
conducted by Val S. Goodfellow). The heterodimer is very soluble in
61


aqueous solution compared to the highly insoluble CP-955.
Overall, the results of the study suggest that it is possible to link
HNE inhibitors to B-9430 and retain both the HNE inhibition and at least BK
B2 antagonism, and thus arrive at a heterodimer with dual activity. Since
the role of the BK B1 receptor is still uncertain, BK B1 antagonism may not
be advantageous in a drug. Compound I, which is essentially selective for
the BK B2 receptor over the BK B1 receptor and has complete retention of
activity in the BK B2 functional assay, as well as an improved HNE Kj,
would be a good starting point for a drug that does not target BK B1
receptors. Should the BK B1 receptor prove to be an important target in
drug therapy, compound II data indicates that it is possible to somewhat
improve BK B1 antagonism, and through further SAR studies, possibly
retain BK B1 antagonism. As further SAR work is indicated by the results
obtained from the present study, it should also include various linkers
between B-9430 and CP-955.
A polypharmacological agent, such as compound I, may be capable
to intervene at multiple points in the inflammatory process of IBD. As both,
the kinin system and tissue degradation by elastase, are important
62


components of chronic inflammatory diseases, compound I or similar
heterodimers could be expected to be potent therapeutic agents. In IBD
one would hope to see significant reduction of the inflammatory response
and tissue edema.
As we begin to appreciate the complexity of many disease states, in
particular inflammation, we see the inherent difficulty in treating patients
with a single drug, which is active only on one target. Multiple-action drugs
are therefore likely to be regarded with increasing interest. The prersent
study has shown that it is possible to design and synthesize single-
molecule drugs, which combine both BK B2 antagonism and HNE
inhibition. In addition to the possible therapeutic advantages there is a
direct cost benefit with polypharmaceutical agents. The multi-active
heterodimer allows for the development and safety profiling of one drug, as
compared to combination therapy, where both drugs would have to be
profiled individually and as a mixture.
63


APPENDIX A
Animal Use Certification
T
April 3, 1996
University of Colorado, Denver
RE: Cortech Protocol #46-93-NP
To Whom it May Concern:
The purpose of this letter is to inform you that the animal studies performed to
support Axel Leimers Masters thesis were done in our USDA licensed animal
facility under the auspices of all applicable federal regulations.
The methods described in the protocol were reviewed and approved by the
Cortech Animal Care and Use Committee prior to the initiation of this project.
Please contact me if I can be of any further assistance in this matter.
Sincerely,
Jeffrey A. Solomon
Manager, Preclinical Projects
Director, Animal Research Center
Cortscft. inc.
9050 NorVt BroKtawv
Dtrwir Co*OfdO 00221
64


APPENDIX B
Animal Use Protocol
The following is a copy of Cortech's Institutional Animal Care and Use
Committee Protocol which was followed for the functional assay contained in
this thesis:
Protocol # 46-93-NP
ANIMAL USE PROTOCOL
Submit completed protocol form to the EPU Director.
DATE FILED: 6/16/93
INVESTIGATOR AND RESEARCH ASSOCIATES:Dr. Eric Whalley, Dr.
William Selig, John Zuzack, David Cuadrado, Michael Burkard
TITLE OF PROTOCOL: In vitro Pharmacological Studies Using Animal
Tissues
CHECK ONE: Initial review Start date 6/16/93
XXX Renewal End date 6/16/96
Pilot project
I. ABSTRACT OF PLANNED USE OF ANIMALS; Write a brief yet
complete description of the planned use of animals.
Animals (mice, rats, hamsters, guinea pigs, rabbits) will be humanely
euthanized and selected tissues will be dissected out. The female rats
65


will be pre-treated 12-18 hours before euthanasia with 100 mg/kg of di-
ethyl stilbestrol subcutaneously. These tissues will include vas
deferens, thoracic aorta, mesenteric artery, jugular vein, pulmonary
artery, uterus, ileum, lungs, trachea, and bladder. Where possible
multiple tissue types will be used from the same animal. The tissues will
be prepared and used for in vitro pharmacological assays. Other tissues
will be used as required.
II. A. Please state your rationale for the use of living animals in the
study. Also include a discussion why an alternative non-living model wasn't
chosen.
The tissues chosen are known to contain specific receptors which are
essential for assay with specific classes of compounds being developed
by Cortech. Measurement of the magnitude of response of these
tissues to test compounds will allow selected compounds to be chosen
for assay in in vivo models and ultimately be developed for use as
therapeutic agents in man. Non-living alternative models which yield
these biological responses are not available.
66


B. Explain the appropriateness of the species chosen.
Literature reports have established the above stated tissues from these
species as excellent models for various receptors because they have a
high degree of specificity for one receptor type and have an easily
measured response to receptor activation.
C. Justify the numbers of animals planned for this study. Please attach
an experimental plan outlining the numbers in each treatment/control groups.
Document the statistical relevance.
While it is difficult to accurately predict the numbers of these
experiments required in the future, the following is a reasonable
estimate: On a yearly basis there may be 150 new compounds
synthesized by our chemists. Given the polytherapeutic nature of our
drug design strategy, each of these compounds must be tested for
multiple activities. This necessitates testing each compound on
approximately 3 tissue types. Efforts are made to take more than one
tissue type from each animal. Depending on the tissue type we can
obtain approximately 4 pieces of tissue per animal. It is necessary to
repeat each experiment a minimum of four times to reduce the variance
of response between tissues. Thus, one animal is required per
67


compound per receptor type (tissue variety). These facts can be used
to estimate the animal requirements for these studies: 150 compounds
X 3 tests (tissue type) X 4 tissues / 4 tissues per animal = 450 animals.
III. ANIMAL USE
A. Species
Sprague-Dawley rat, Syrian hamster, Hartley guinea pig, New
Zealand White rabbit, ICR mice
B. Number: per year
200 rats, 75 guinea pigs, 30 hamsters, 75 rabbits, 100 mice
entire study same numbers
housed at any one time 20 rats, 12 guinea pigs, 10 hamsters, 8 rabbits,
10 mice
C. Sex and age (or weight range): either sex for all species, any size
D. Source (vendor): Sasco, Harlan Sprague-Dawley, R&R Rabbitry
68


E. Describe any expected health changes in the animals due to your
experimental procedures, e.g. weight loss, food intake and/or behavioral
changes, and indicate treatment.
none
F. Will the animals suffer any pain and/or distress during the experiment
(excluding survival surgery procedures)?
Yes No XXX
If yes, explain.
G. Describe abnormal environmental conditions that may be imposed.
This includes temperature, lighting, etc.
none
H. Describe and justify the use of any restraint devices employed:
none
69


I. Will special housing, conditioning, diet or other conditions be required:
Yes No XXX If yes, explain:
J. Does the protocol involve animal surgery? Yes No XXX
Survival Non-survival
If survival surgery is involved, provide the following information:
Surgeon's name and qualifications:
Anesthetist's name and qualifications:
III. K. Complete description of the surgical procedures:
L. For anesthesia procedures, provide the following information:
Pre-anesthesia: Drug:
Dose:
Anesthesia: Drug:
Dose:
How is anesthesia level monitored?
What are the provisions for post-surgical care and include the name(s) of the
70


person(s) responsible for care?
What drugs are to be used to alleviate pain or distress and what are their
dosages? If analgesics are intentionally withheld, please justify in writing.
M. If your procedures require animal immunizations and/or bleeding,
provide the following information:
Antigen(s):
Adjuvant(s):
Injection volume, frequency and site(s):
Blood collection methods, volume and frequency:
N. What form of euthanasia will be used and when? Give agent, dose and
method.
Animals will be euthanized by rapid asphyxiation with carbon dioxide.
In animals where the lungs are to be used an overdose of pentobarbital
(100 mg/kg i.v. or i.p.) or urethane (3 g/kg i.p.).
O. Who will be euthanizing the animals? Give name and qualifications.
71


Eric Whalley, Ph.D. 20+ yrs experience in academic/industrial labs
William Selig, Ph.D approx. 15 yrs experience in pharmaceutical
companies' research labs
John Zuzack, M.S. -10 years experience in pharmaceutical companies'
research labs
David Cuadrado, B.S. 6 mo training in Cortech pharm lab
Michael Burkhard, B.S. 5 yrs experience in pharmaceutical companies'
research labs
IV. TOXIC AND HAZARDOUS SUBSTANCES
A. Will any of the following be used in the animals?
1. Infectious agents
2. Radioisotopes
3. Toxic chemicals or carcinogens
4. Recombinant DNA
5. Experimental drugs
6. Malignant cells or hybridomas
If yes, describe all procedures necessary for personnel and animal
safety including biohazardous waste and carcass disposal and cage
72


decontamination:
All experimental compounds of unknown toxicity are treated as
though they were toxic in compliance with OSHA regulations.
Extreme care will be taken to avoid contamination of personnel
with these chemicals.
B. If transplantable tumors or hybridoma cells are to be injected into the
animals, have the tissues/cells been tested for inadvertent contamination by
viruses or mycoplasma? Yes No
If yes, give results of the test:
V. CATEGORIES OF ANIMAL EXPERIMENTATION BASED UPON
LEVEL OF MANIPULATION AND PAIN:
XXX NP No pain. This category includes any procedure that would
be conducted on a human being without analgesia. Although a slight
degree of discomfort may be associated with the procedure, in general it
will not produce any escape behavior beyond that of psychological
distress. This category also includes animals humanely killed without any
73


treatments, manipulations, etc., but will be used to obtain tissue, cells,
sera, etc.
MP Minor pain. This category includes any procedures that main
cause the animal mild pain or distress, either directly caused by the
procedure or occurring after time. These procedures include ascites
production, most toxicity studies, etc. *****
D Pain alleviated with drugs. Any protocol which calls for an invasive
procedure, or will cause more than momentary or slight distress will be
performed with appropriate analgesia, sedation, or anesthesia.
Appropriate medications will be determined with the coordination of a
veterinarian. It should be remembered that muscle relaxants and paralytic
drugs such as succinylcholine or curariform drugs are not analgesic nor
anesthetic. They cannot be used alone for surgical restraint. *****
P Those protocols which necessitate painful procedures that cannot
be conducted with the use of analgesics, anesthetics or tranquilizers
because the use of such agents would defeat the purpose of the research.
Such procedures must be justified in writing, approved by the committee
74


and be conducted under the direct supervision of the investigator.
**
*****special section*****
Please fill out this section if these procedures will cause more than slight
or momentary pain or distress to the animals. Explain why these
procedures are necessary and unavoidable. Provide a written narrative of
alternative procedures/protocols considered and rejected. Include a
description of the methods and sources (Medline, Animal Welfare
Information Center) used, or logic used to make your determination.
We require that the principal investigator conduct a literature search for
two purposes: 1) to determine that the study isn't duplicative; 2) to
determine if the study can be done less painfully or on non-whole animal
models.) A literature search has been conducted, and to the best of my
knowledge, this study is not duplicative and other methods are
inappropriate or unsuitable for the purpose of the study.
75


REFERENCES
1. In The Merck Manual of Diagnosis and Therapy, 16th edition, Merck
Research Laboratories, 1992.
2. Loftus, Edward V. Jr., Sandborn, W. J. Drug Therapy for
Inflammatory Bowel Disease. Contemp. Int. Med., 1995, 7, 21-34.
3. In Cellular Pathophysiology, Sayeed, M., Ed.; CRC Press, 1989.
4. Springer, T. A. Adhesion Receptors of the Immune System, Nature,
1990, 346, 425-434.
5. Osborn, L. Leukocyte Adhesion to endothelium in Inflammation,
Cell, 1990, 62, 3.
6. Maling, H. M., Webster,M. E., Williams, M. A., Saul, W., Anderson,
W. Inflammation Induced by Histamine, Serotonin, Bradykinin and
Compound 48/80 in the Rat: Antagonists and Mechanisms of
Action. J. Pharmacol. Exp. Ther., 1974, 191: 300-310.
7. Miles, A. A., and Wilhelm, D. L. The Activation of Endogenous
Substances Inducing Pathological Increases of Capillary
Permeability. In The Biochemical Response to Injury, Stoner, H. B.,
Eds.; Blackwell, Oxford, 1960, 57-79.
8. Wilhelm, D. L., Mill, P. J., Sparrow, E. M., McKay, M. E., Miles, A.
A. Enzyme-like globulin from Serum Reproducing the Vascular
Phenomena of Inflammation. IV. Activable Permeability Factor and
its Inhibition in the Serum of the Rat and the Rabbit. Brit. J. Exp.
Path., 1958, 39, 228-250.
9. Regoli, D., Barabe, J. Pharmacology of Bradykinin and Related
Kinins, Pharmacological Reviews, 1980, 32, 1-45.
76


20. Moroder, L., Hallett, S., Wunsch, E., Keller, O., Wersin, G. Hoppe
Seyler,sZ. Physiol. Chem., 1976, 357, 1651.
21. Stewart, J.M., Young, J.D. In Solid Phase Peptide Synthesis,
Freeman, San Fransisco, 1969.
22. Cunningham, R. T., Mangold, S. E., Spruce, L. W., Ying, Q. L.,
Simon, S. R., Wieczorek, M., Ross, S., Cheronis, J. C.,
Kirschenheuter, G. P. Synthesis and Evaluation of CE-0266: A New
Human Neutrophil Elastase Inhibitor. Bioorg. Chem., 1992, 20,
345-355.
23. Manning, D. C., Vavrek, R., Stewart, J. M., Snyder, S.H. Two
Bradykinin Binding Sites with Picomolar Affinities. J. Pharmacol.
Exp. Then, 1986, 237, 504-512.
24. Goodfellow, V. S., Marathe, M. V., Kuhlmann, K. G., Fitzpatrick, T.
D., Cuadrado, D., Hanson, W., Zusack, J. S., Ross, S. E.,
Wieczorek, M., Burkhard, M., Whalley, E. T. Bradykinin Receptor
Antagonists Containing N-Substituted Amino Acids: In Vitro and In
Vivo B2 and B1 Receptor Antagonist Activity. J. Med. Chem., 1996,
in press.
25. Cheronis, J. C., Whalley, E. T., Nguyen, K. T., Eubanks, S. R.,
Allen, L., Duggan, M. J., Loy, S. D., Bonham, K. A., Blodgett, J. K.
A New Class of Bradykinin Antagonists: Synthesis and In Vitro
Activity of Bissuccinimidoalkane Peptide Dimers. J. Med. Chem.,
1992, 35, 1563-1572.
26. Schild, H.O. pA2, A New Scale for the Measurement of Drug
Antagonism, Br. J. Pharmacol., 1947, 2,189.
27. Kaiser, E. T., Colescott, R. L., Bossinger, C.D., Cook, P. I. Color test
for detection of free terminal amino groups inthe solid-phase
synthesis of peptides. Anal. Biochem., 1970, 34, 595-598.
28. Atherton, E., Sheppard, R. C. In Solid Phase Peptide Synthesis; A
Practical Approach] IRL Press, Oxford, 1989.
78


10. Stewart, J. M. In The Kinin System in Inflammation, in Proteases,
Protease Inhibitors and Protease-Derived Peptides, Cheronis,
Repine Eds.; Birkhauser, 1993.
11. Cheronis, J. C., Whalley, E.T., Allen, L.G., Loy, S.D., Elder, M.W.,
Duggan, M.J., Gross, K.L., and Blodgett, J.K. .Design, Synthesis
and in Vitro Activity of Bissuccinimidohexane Peptide Heterodimers
with Combined BK1 and BK2 Antagonist Activity, J. Med. Chem.
1994, 37, 348.
12. Regoli, D., Barabe, J., Park, W. K. Receptors for bradykinin in rabbit
aorta. Can. J. Physiol. Pharmacol., 1977, 55, 855-867.
13. Park, W. K., St-Pierre, S., Barabe, J., Regoli, D. Synthesis of
Peptides by the Solid Phase Method. III. Bradykinin: Fragments
and Analogues. Can. J. Biochem., 1978, 56, 92-100.
14. Schroder, E. Structure-activity relationships of kinins. Handb. Exp.
Pharmacol., 1970, 25,324-350,1970.
15. Stewart, J.M. Chemistry and Biologic Activity of Peptides Related
to Bradykinin. Handb. of Exp. Pharmacol.25(supp.), 1979, 227-272.
16. In the convention used here, the amino acid residues, starting at the
scissile bond and going out towards the carboxyl terminal of the
peptide, are labelled 1' (one prime), 2', 3', etc.. The residues,
starting from the scissile bond and going toward the amino terminal,
are labelled 1,2,3, and so on (unprimed).
17. Ernst, R. R., Bodenhausen, G., Wokaun, A. In Principles of Nuclear
Magnetic Resonance in One and Two Dimensions; Clarendon
Press, Oxford, 1987.
18. Bidlingmeyer, B. A., Chen, S. A., Tarvin, T. L. Rapid Analysis of
Amino Acids Using Pre-column Derivatization, J. Chrom., 1984,
336, 93-104.
19. Porter, T.H., Shive, W. DL-2-lndaneglycine and DL-(3-
Trimethylsilylalanine, J. Med. Chem., 1968, 11, 402.
77


29. Rosell, C. M., Fernandez-Lafuente, R., Guisan, J. M. Resolution of
Racemic Mixtures by Synthesis Reactions Catalyzed by
Immobilized Derivatives of the Enzyme Penicillin G Acylase. J.
Molec. Cat., 1993, 84, 365.
30. Guisan, J. M., Alvaro, G., Fernandez-Lafuente, R. Stabilization of
Heterodimeric Enzyme by Multipoint Covalent Immobilization.
Biotech & Bioeng., 1993, 42, 455.
31. Bommarius, A. S., Drauz, K., Klenk, H. Operational Stability of
Enzymes: Acylase-catalyzed Resolution of A/-Acetyl Amino Acids to
Enantiomerically Pure L-Amino Acids. Annals., 1992, 672, 126.
32. Palmer, T., Paul, A. I. In Principals of Drug Action] Pratt, W. T.,
Taylor, P. Eds.; Churchille Livingston: New York, 1990; 60-63.
33. Lajos, G., Stewart, J. M., Whatley, E., Hanson, W., McCullough, R.
New Bradykinin Antagonists Having High Potency at both B1 and
B2 Receptors. Proceedings 14th Am. Pept. Symp.; Escom: Leiden,
1995.
79


Full Text

PAGE 1

BRADYKININ B1/B2 ANTAGONIST LINKED TO A HUMAN NEUTROPHIL ELASTASE INHIBITOR; A HETERODIMER FOR THE TREATMENT OF INFLAMMATORY BOWEL DISEASE. by Axel H. Leimer B.A., University of Colorado at Denver, 1993 A thesis submitted to the University of Colorado at Denver in partial fulfillment of the requirements for the degree of Master of Basic Science 1996

PAGE 2

1996 by Axel H. Leimer All rights reserved.

PAGE 3

-The thesis for the Master of Basic Science degree by Axel H. Leimer has been approved by Date

PAGE 4

Leimer, Axel H. (M.B.S., Basic Science) Bradykinin B1/B2 Antagonist linked to a Human Neutrophil Elastase Inhibitor; a Heterodimer for the Treatment of Inflammatory Bowel Disease. Thesis directed by Professor Douglas F. Dyckes. ABSTRACT Inflammatory bowel disease (lBO) is a common disorder in developed countries. Although it has some of the characteristics of a general inflammatory disorder, its etiology is still unknown. Because inflammatory disorders typically have a very complex etiology and involve a multitude of mediators a multi-active drug approach is proposed. In certain chronic inflammatory conditions, like lBO, we believe that bradykinin (BK) and human neutrophil elastase (HNE) are co-operatively involved. The present study describes single compounds designed to incorporate HNE inhibitory and BK B1/B2 antagonist activity. A proprietary elastase inhibitor (CP-955) was directly linked via amide bond formation to the BK antagonist B-9430 (D-Arg-Arg-Pro-Hyp-Giy-.b-lgi-Ser-0-lgi-Oic-Arg-OH), which incorporates both BK B1 and B2 antagonist activity and two analogs designed iv

PAGE 5

specifically for this study. The three resulting conjugates synthesized, were purified and analyzed for in vitro antagonist activity and HNE inhibition. For all three compounds (compound I, II and Ill), 82 receptor binding in human cloned receptors was reduced tenfold. In a 82 antagonist functional assay on guinea pig ileum, compounds I and II had equivalent potency to 8-9430. The 8K 81 antagonist activity was significantly less than that of 8-9430 for all three compounds. Compound I showed a fourfold improvement in the inhibition constant against human neutrophil elastase, as compared to CP-955, while compounds II and Ill were not inhibitory. This study clearly shows that it is possible to retain 8K 82 antagonism and HNE inhibition by combining CP-955 and 8-9430. This abstract accurately represents the content of the candidate's thesis. I recommend its publication. v

PAGE 6

DEDICATION This work is dedicated to Trisha, Stefan, Jessica and Christian.

PAGE 7

ACKNOWLEDGMENT My thanks to my thesis committee for its support and advice throughout this project. Special thanks to Heather B. Kroona for her direct supervision of my laboratory activities and continued encouragement. Additionally, I want to thank Douglas L. Gernert for the synthesis of the CP-955 used in this study and Lyle W. Spruce for the initial synthesis of CP-955 and our discussions on the 2-indaneglycine synthesis. Thanks are also due to John M. Asztalos for amino acid analysis and Steven K. Wolk for assistance with Nuclear Magnetic Resonance work on 2-indaneglycine. I am also grateful to John S. Zuzack and Michael R. Burkard for their work with the pharmacological assays, Sherman E. Ross for obtaining elastase inhibition constants, and Val S. Goodfellow for the molecular modeling. For the many years of previous research on bradykinin antagonists and their utilization in heterodimers I am indebted to John C. Cheronis, Eric T. Whalley and John M. Stewart.

PAGE 8

CONTENTS Chapter 1. Introduction .................................................................................. 1 1.1 Inflammatory Bowel Disease..................................................... 1 1.2 Human Neutrophil Elastase ....................................................... 2 1.3 Bradykinin Antagonists as Therapeutics .................................... 4 1.4 Rationale of Drug Design ............................................................ 7 1.5 Structure-Activity Relationship of B-9430 .................................. 8 2. Experimental ................................................................................ 12 2.1 Materials and Instrumentation ................................................... 12 2.2 Methods ..................................................................................... 15 2.2.1 Amino Acid Analysis............................................................... 15 2.2.2 2-lndaneglycine Synthesis ...................................................... 16 2.2.2.1 2-Bromoindane... ... ......... ........... ... ... ....... .. ...... .. .. .. .... ... ........ 16 2.2.2.2 Ethyl-a-acetamido-a-cyano-2-indaneacetate ............... : ...... 18 2.2.2.3 (D,L)-2-Indaneglycine .......................................................... viii

PAGE 9

2.2.2.4 N-Acetyi-(D,L)-2-indaneglycine ............................................ 20 2.2.3 Resolution of (D,L)-2-Indaneglycine ....................................... 21 2.2.4 GITC-Derivatization ................................................................ 27 2.2.5 Tert-butyloxycarbonyl-lgl.. ...................................................... 29 2.3 B-9430 Analogs ......................................................................... 31 2.3.1 Solid Phase Peptide Synthesis ............................................... 31 2.3.2 Cleavage of Peptide from Resin and Post-cleavage Workup. 35 2.3.3 Purification and Characterization ............................................ 36 2.4 Conjugation of CP-955 to the B-9430 Analogs .................... .. 42 2.5 Biological Assays ....................................................................... 45 2.5.1 HNE Inhibition Assay ...................................................... u ....... .45 2.5.2 Bradykinin B2 Guinea Pig Ileum Receptor Binding Assay ..... 45 2.5.3 Bradykinin B2 Human Receptor Clone Binding Assay ........... 46 2.5.4 Bradykinin B1 Human Lung Fibroblast Receptor Binding Assay ........................................................................ 46 2.5.5 Bradykinin B2 Antagonist Functional Testing on Guinea Pig lleum ........................................................................................ 46 3. Results and Discussion............................................................... 48 3.1 Chemistry ................................................................................... 48 3.2 Biology ....................................................................................... 51 ix

PAGE 10

Appendix A. Animal Use Certification .............................................................. 64 B. Animal Use Protocol. ................................................................... 65 References ....................................................................................... 76 X

PAGE 11

FIGURES Figure 1.1 Bradykinin .............................................................................. 5 1.2 B-9430.... ..... ......... ...................................... ................ ... .. .. .... 6 1.3 Amino Acids in B-9430 ........................................................... 6 1.4 The Human Neutrophil Elastase Inhibitor CP-955 .................. 8 2.1 2-lndaneglycine Synthesis ..................................................... 17 2.2 N-Acetyi-(D,L)-2-indaneglycine .............................................. 21 2.3 Resolution of (D,L)-2-Indaneglycine ...................................... 23 2.4 Proton NMR of (L)-2-Indaneglycine ....................................... 24 2.5 Magnitude COSY Spectrum of (L)-2-Indaneglycine.............. 25 2.6 N-Acetyi-(D)-2-indaneglycine ................................................. 27 2.7 GITC-Derivatization of (L)and (D)-2-Indaneglycine ............... 28 2.8 Tert-butyloxycarbonyl-lndaneglycine ..................................... 30 2.9 Purification of B-9430 ............................................................. 40 2.1 0 Mass Analysis for B-9430 ...................................................... 41 2.11 Mass Analysis of Compound I. .............................................. A4 xi

PAGE 12

3.1 Effect of Dimethyl Sulfoxide on BK Response Curve ............. 52 3.2 Effect of B-9430 on BK Response Curve ............................... 54 3.3 Effect of Compound I on BK Response Curve ....................... 55 3.4 Effect of Compound II on BK Response Curve ...................... 56 3.5 Effect of Compound Ill on BK Response Curve ..................... 57

PAGE 13

TABLES 1.1 Compounds Synthesized and Tested .................................... 10 2.1 Peptides Synthesized by Solid Phase Peptide Synthesis ....... 31 2.2 Amino Acid Analysis of Stempeptide................... ............... .. 34 2.3 Yields of Peptide Synthesis .................................................... 37 2.4 Characterization of Peptides .................................................. 37 2.5 Mass Analysis of Peptides ..................................................... 39 2.6 Mass Analysis of Conjugates ................................................. 43 3.1 Summary of Biological Data ................................................... 59

PAGE 14

1. INTRODUCTION 1.1 Inflammatory Bowel Disease Chronic inflammatory bowel disease (IBD) is defined as "a spectrum of inflammatory bowel disorders with overlapping clinical, epidemiological and pathological findings, but without defined etiology"1 Patients with IBD can be divided into two groups, Crohn's Disease (CD, Regional Enteritis, Granulomatous Ileitis or Ileocolitis) and Ulcerative Colitis (UC), although differentiation is often difficult and many subgroups based on symptoms exist. Etiology of IBD is mostly unknown and the disease manifests itself through chronic diarrhea, abdominal pain, fever, anorexia, perianal fistulae, and inflammation of the gastrointestinal tract. Inflammatory bowel disease is a vascular disease with the typical symptoms of inflammation, including transendothelial migration of leukocytes into inflamed tissue and subsequent tissue damage. In particular for CD, the only effective therapy is often the removal of the inflamed section of the bowel for temporary alleviation of the symptoms 1 Postoperative recurrence is between 60 and 95%1 1

PAGE 15

As part of the inflammatory response, white blood cells such as polymorphonuclear neutrophils (PMNs) and macrophages invade the tissue of the bowel and release various proteinases, including human neutrophil elastase (HNE), which cause general edema. Current drug therapy is limited to general immunomodulators, aminosalicylates and steroids2 Unfortunately, many drugs in use, especially those indicated for UC, exacerbate diarrhea and are generally not very effective therapies. The potential drug therapy pursued in the present study, was intended to inhibit the action of tissue and matrix degrading enzymes, in particular the action of elastase released from the neutrophil. The attempt was made to incorporate a human neutrophil elastase inhibitor (HNEi) and a BK antagonist into a single multi-active drug. 1.2 Human Neutrophil Elastase Although the exact etiology of lBO is still unknown, PMNs aFe thought to play an important role in the disease. It is therefore helpful to 2

PAGE 16

look at the role of PMNs in the model of general inflammation. In inflammation, various factors, including lipopolysaccharides (LPS), tumor necrotizing factor alpha (TNF-a), and interleukin 1 (IL-1), cause increased expression of certain classes of membrane glycoproteins on endothelial cells (EC)3 of postcapillary venules4 These membrane proteins act as receptors for ligands expressed on blood cells, in particular PMNs, promoting PMNs adherence to the EC. The PMNs are generally the first cells to bind to the activated EC and extravasate (i.e. migrate across endothelial layer into tissue). The activated EC express receptors of the selectin family (E-, P-, and L-selectin), which interact with carbohydrate ligands on the PMNs. Transient binding of the PMNs to the EC occurs, resulting in 'rolling' of the PMNs over the EC surface. Most of this rolling is mediated by E-and P-selectins on the EC and their corespondent ligand on the PMNs, which is Sialyl Lewis X (SLew"; NeuAca2-3GaiP1-4[Fuca13]GicNAc)5. Platelet Activating Factor (PAF) expressed on the EC activates the PMNs during this rolling, causing an increase of adhesion of the PMNs, corresponding to an increased expression of the integrin receptor leukocyte function-associated antigen-1 (LFA-1). The LFA-1 binds to intercellular adhesion molecules (I CAMS) on the EC and holds the PMNs attached to the EC despite the shear forces of the blood flow 5 Now 3

PAGE 17

extravasation occurs5 In acute inflammation, the leukocyte infiltrate is composed of 90-95% PMNs. Following extravasation, the PMNs are activated and self-destruct, releasing the contents of intracellular granules. Most of the damage caused by the PMNs is due to elastase and toxic radicals released in this respiratory burst. The recruitment of PMNs reaches a maximum about 4 hr after the initial insult, after which macrophages begin to accumulate at the site. Synthetic, low-molecular-weight HNEi are of great interest to many pharmaceutical companies. Indications for HNEi include inflammatory diseases like Adult Respiratory Distress Syndrome (ARDS), Myocardial Reperfusion Injury, Rheumatic Arthritis, Cystic Fybrosis, Emphysema, and Inflammatory Bowel Disease. 1.3 Bradykinin Antagonists as Therapeutics Bradykinin (BK) research began as early as 1909, when French surgeons observed a fall in blood pressure following injectionsof human urine fractions9 Because kinins are associated with the regulation otthe cardiovascular system and are mediators of inflammation and pain6 7 8 the 4

PAGE 18

present study incorporates a BK antagonist into a multi-active drug. The pharmacology of BK has been described in detail elsewhere9 10 It is thought that BK B2 receptors are expressed constitutively in most animals, while BK B1 receptors are induced during chronic inflammation. However, it is not yet known what role the BK B1 receptor plays in pain and inflammation. Bradykinin is a nonapeptide which binds to the BK B2 receptor (Figure 1.1). Arg-Pro-Pro-Giy-Phe-Ser-Pro-Phe-Arg-OH 1 2 3 4 5 6 7 8 9 Figure 1.1: Bradykinin Upon the action of various enzymes (Carboxypeptidases, Kininase I, etc.), BK is converted to the [desArg9]-BK, which is the ligand for BK B1 receptors. A key synthetic transformation occurs when L-Pro7 (agonist) is replaced with D-Phe7 (antagonist). Many hundreds of peptides have been synthesized in extensive structure-activity relationship studies in the search for more potent antagonists9 In the past, heterodimers, consisting of two 5

PAGE 19

linked peptides with 8K 81 and 8K 82 antagonist activity respectively, have been designed to block both receptors with one compound". Compound 8-9430, a single decapeptide developed in Professor John Stewart's laboratory at the Health Science Center of the University of Colorado, is a 8K 82 and 81 antagonist (Figure 1.2). This peptide is unique because it has 8K 81 antagonist activity without being a [desArg9]-8K derivative and because it is the first to combine 8K 81 and 82 antagonist activity in one peptide. D-Arg-Arg-Pro-Hyp-Giy-.1.-lgi-Ser-D-Igi-Oic-Arg-OH 0 1 2 3 4 5 6 7 8 9 Figure 1.2: 8-9430 In Figure 1.2 Hyp is 4-trans-hydroxyproline, lgl is 2-indaneglycine, and Oic is octahydro-indole-2-carboxylic acid (Figure 1.3). A H COOH COOH L-Hyp L-Oic lgl 6

PAGE 20

It has been found that the affinity for the BK B 1 receptor generally increases by either removing the arginine residue in position 9 (as in [desArg9]-BK) or by adding a positive charge at the amino terminus (lysine or arginine). While extensive structure activity relationships have been reported previously for BK B1 and B2 agonists and antagonists12 13 14 15, it is unclear to what extent the results can be applied to B-9430. 1.4 Rationale of Drug Design Given the importance of bradykinins and proteinases, in particular HNE, in inflammation, the decision was made to link a proprietary elastase inhibitor to B-9430, a combination BK B1/B2 antagonist, which would yield a heterodimer with possible dual or triple activity. Because of the complexity of inflammatory disorders, it is hoped that a multi-active drug would show increased efficacy in treatment. Additionally, dual or triple activity could increase the number of indications for which the drug might be used, and thus, make the drug more marketable. Finally, as both bradykinin receptors and elastase are important and perhaps essential targets in anti-inflammatory drug design, the combined approach allows for the development of one rather than two or three drugs. Cortech has 7

PAGE 21

developed several HNE inhibitors to date. The present study involved the use of HNEi CP-955 (Figure 1.4), which was developed in collaboration with Marion Merrill Dow. HO Human Neutrophil Elastase Inhibitor 0 Figure 1.4: The Human Neutrophil Elastase Inhibitor CP-955. The carboxylic acid moiety in CP-955 serves mainly to increase solubility. When bound to the active site of HNE, the carbonyl oxygen of an ester moiety extends into the oxy-anion binding pocket of the enzyme. The carboxylic acid is believed to extend from the active site cleft out into bulk solvent in the unprimed direction16 (internal communications). 1.5 Structure-Activity Relationship of B-9430 Several analogs of B-9430 were investigated for their retention of 8

PAGE 22

both 8K 81 and 82 antagonist activity. CP-955 was linked directly to various positions of the three 8-9430 analogs via amide bond formation. Previous studies (unpublished data, John Asztalos, 1996) have shown that directly linking a HNEi to a substrate or polymeric carrier, even in sterically crowded environments, does not interfere with its activity (internal communications, 1995). This can be explained by the fact that HNE has a very shallow substrate binding-pocket, which allows short extensions of the substrate or inhibitor, in particular those toward the unprimed end, to extend into bulk solvent. Prior to this study, no data were available on the effects of conjugating another molecule to B-9430. It was thus unknown whether a B-9430 containing conjugate would still be able to bind to BK receptors. Since a complete SAR study was not within the scope of the present study, three different types of conjugates were synthesized based on reasons outlined below. The three compounds synthesized and tested are shown in Table 1.1. For compound I, CP-955 was conjugated via its carboxyl group to the N-terminal amino group (D-Arg0 ) of B-9430 resulting in an amide bond linkage and a linear conjugate. Conjugation was directed by the use of orthogonal protecting groups as described in the Methods section. Direct conjugation, without linkers, was chosen to keep the overall size of the molecule to a minimum. In compound II, an N-terminal 9

PAGE 23

extension consisting of a lysyl residue (position -1) preserved the number of positive charges on the peptide upon conjugation, which may be important in binding to the BK B1 receptor. Table 1.1 Compounds Synthesized and Tested -1 0 1 2 3 4 5 6 7 8 9 Q-Arg-Arg-Pro-Hyp-Giy-L-Igi-Ser-0-lgi-Oic-Arg-OH CP-955 linked at 0-Argo N-term. amine. (I) H-Lys-0-Arg-Arg-Pro-Hyp-Giy-L-Igi-Ser-0-lgi-Oic-Arg-OH CP-955 linked at Lys-1 N-term. amine. (II) 0-Lys-Arg-Pro-Hyp-Giy-L-lgi-Ser-0-lgi-Oic-Arg-OH CP-955 linked at O-Lys0 N-term. (Ill) Compound Ill was designed to be essentially the same as compound I except that the O-Arg0 was replaced by a O-Lys0 residue. This substitution preserved the overall charge, while helping to elucidate the importance of the Arg residue in the zero position. Optimal positions for linking BK antagonists to other molecules with retention of antagonist activity have previously been explored at Cortech17 10

PAGE 24

These studies involved the linking together of two antagonists by various linking moieties for the express purpose of generating a dual-action heterodimer. Although no linkers were used in the present study, the data from the previous Cortech study yielded valuable information regarding those positions that could be altered in a typical 8K 81 or 82 antagonist without significant loss of activity. This data base served as the starting point for the design of the new conjugates explored here. Specifically, linking at positions 0 and 1 has been shown to be optimal for preserving 8K 81 antagonism in [des Arg9]-8K. Linking at those same positions generally does not reduce 8K 82 antagonism of 8K 82 antagonist peptides significantly25 Linking to the amino acids in positions 5 and 6 has been shown to severely impair the ability of the peptide to bind to the 8K 81 receptor5 For these reasons the HNEi CP-955 was linked to the N-terminus of the 8-9430 peptide or analog in all three compounds. The resulting three compounds differed only in positions -1 and 0 and could therefore be synthesized from a common 'stem peptide'. Although the data from previous Cortech studies guided the design of the conjugates synthesized here, it was uncertain at the outset how well these data would extrapolate to the behavior of 8-9430, which is not a 11

PAGE 25

[ desArg9]-8K 81 antagonist. Furthermore the part of the 8-9430 peptide that actually confers the ability to bind to the 8K 81 receptor is not known. It was therefore unclear whether 8K 81 antagonism could be preserved and, additionally, whether 8K 81 antagonism would be beneficial in any drug indicated for inflammatory disease. However, both of these issues constitute pressing questions for 8K-antagonist research and drug design, and their answers, be they positive or negative, are of importance. 2. EXPERIMENTAL 2.1 Materials and Instrumentation All chemicals were purchased from Sigma, St. Louis, MO, unless otherwise indicated. Analytical Reversed-Phase HPLC (RP-HPLC) was performed on a Hewlett Packard Series 1050 instrument using a Vydac C-18 column (0.46 12

PAGE 26

x 15 em); particle size 5 All chromatography was monitored at both 254 and 214 nm. Solvent A was water containing 0.1% trifluoroacetic acid (TFA, v/v) as a modifier. Solvent 8 was acetonitrile (CH3CN) containing 0.1% TFA (v/v). Preparative RP-HPLC was performed on a Waters PrepLC 4000 System using a Vydac C-18 column (5 x 25 em), particle size 5 and a Waters 486 detector. All chromatograms were obtained 'off-peak' at 230 nm to ensure that the absorbance signals were on scale. Solvents A and 8 were the same as described for the analytical HPLC. Nuclear Magnetic Resonance (NMR) spectroscopy was performed on a Varian Gemini 300 MHz instrument by the author. Additional NMR work (COSY) was done by Dr. Steven K. Walk at Cortech. The magnitude COSY spectrum of (D)and (L)-2-indaneglycine was acquired on a 8ruker DRX400 (400MHz) NMR spectrometer using the proton channel of an inverse detection probe. The standard pulse sequence (Ernst et al., 1987)17 was used, along with two 90 pulses (12.4 msec), a sweep width of 3453.0 Hz in each dimension, 4096 complex points in t2, and 512 !1 values (32 scans for each). The data were apodized with sine bell apodization 13

PAGE 27

functions, and zero-filled in the t1 dimension prior to the 2D Fourier transformation to yield a 2048x2048 data set. Mass spectrometry was performed on a Finnigan Mat Lasermat Mass Analyzer (Matrix Assisted Laser Desorption Ionization Time Of Flight method, MALDI-TOF). All measurements were made with two internal calibration standards to bracket the sample, unless otherwise indicated. Additional mass spectra were obtained through the analytical services of M-Scan, Inc., West Chester, PA, using Fast Atom Bombardment Mass Spectroscopy (FAB) on a VG Analytical ZAB 2-SE high field mass spectrometer. A cesium ion gun was used to generate ions for the acquisition of mass spectra, which were recorded using a PDP 11250J data system. Mass calibration was performed using cesium iodide. Melting points were determined using an Electrothermal 9100 (not corrected). Polarimetry was performed using a DigiPol Automatic Polarimeter, Model DP1A31, by Rudolph Instruments. 14

PAGE 28

Amino acid analysis was performed at Cortech by John Asztalos with a Waters Pico-Tag Amino Acid Analysis System, using a single wavelength UV detector set at 254 nm. 2.2 Methods 2.2.1 Amino Acid Analysis Amino acid analysis was performed by John Asztalos as described in detail elsewhere18 Briefly, the free or protected peptides (diluted aliquots mixed with internal standard) were hydrolyzed in vacuo with vapor phase 6 N HCI for 24 hr at 100 C, then dried, neutralized with triethylamine and dried again in vacuo on a Pico-Tag Work Station. The sample was then allowed to react for 30 min with phenylisothiocyanate (PITC) to form the phenylthiocarbamyl (PTC) derivative of the amino acids. Subsequent chromatographic analysis was performed by C-18 RP-HPLC, detecting peaks at 254 nm. Peak area integration with Baseline software (Waters) was compared to standards run separately, and the internal standard in the analysis sample was used to calculate percent recovery. The standard PicoTag 1 gradient was extended isocratically for 3 min to recover the 15

PAGE 29

PITC derivative of 2-indaneglycine (lgl). 2.2.2 2-lndaneglycine Synthesis 19 (Figure 2.1) 2.2.2.1 2-Bromoindane To 2-indanol (1, 200 g, 1.48 mol) in pyridine (30 mL, 0.37 mol) and chloroform (648 mL) at -15 C was added phosphorus tribromide (160 mL, 1.69 mol) over 45 min. The reaction mixture was stirred at room temperature for 24 hr. The mixture was then extracted with 450 mL chloroform and 500 g ice. The organic layer was washed with water (2 x 400 mL) and dried over anhydrous sodium sulfate (Na2S04). The solvent was evaporated in vacuo leaving a brown semi-solid. The product was distilled in vacuo and fractionated, yielding 121 g of a colorless liquid (2bromoindane19, 2, 121 g, 0.61 mol, 41.2%, bp 92-95 C at 5 mmHg, n0 21 = 1.6). 16

PAGE 30

Figure 2.1 2-lndaneglycine Synthesis '> 0 )=O 0-{ z 0=\ 17 z 0=\ z 0 "'::c oo oii:i + +

PAGE 31

2.2.2.2 Ethyl-a-acetamido-a-cyano-2-indaneacetate To sodium ethoxide (40 g, 0.6 mol) suspended in 500 ml dry dimethyl sulfoxide (DMSO) a solution of ethylacetamidocyanoacetate (1 00 g, 0.57 mol) in 500 ml dry DMSO was added with vigorous stirring. The 2bromoindane (2, 121 g, 0.61 mol) was added dropwise over a one-hour period with continued vigorous stirring. The brown solution was stirred for 24 hr at room temperature, then the solution was stirred for 7 hr at 50 C to drive the reaction to completion. The mixture was evaporated in vacuo and the residue treated with 600 ml cold water and extracted with ethyl acetate (2 x 600 ml). Then the extract was dried over magnesium sulfate (MgS04 ) and evaporated in vacuo leaving a brown oil. The first recrystallisation from water yielded a yellowish brown solid (50 g); recrystallisation and extraction with ethyl acetate from the mother-liquor resulted in another 75 g of yellowish white solid. Both solids were combined and recrystallized from toluene to yield 90 g of a white solid (ethyl-a-acetamido-a-cyano-2indaneacetate, 3, 50.8%, mp 159-160 C19 ) which showed a single peak on RP-HPLC. (RP-HPLC, single peak, gradient 0-40%8 in 40 min, retention time 12.24 min) 18

PAGE 32

2.2.2.3 (D,L)-2-Indaneglycine A solution containing ethyl-cx-acetamido-a-cyano-2-indaneacetate (3, 90 g, 0.32 mol) in 1600 mL of 10% sodium hydroxide (NaOH) was refluxed for 20 hr. The solution was cooled and adjusted to pH 6.5 with concentrated HCI and chilled. The precipitate was washed with cold water (2 x 100 mL) and methanol (2 x 50 mL). The solid was then refluxed in 1200 mL 6N hydrochloric acid (HCI) for 12 hr. The solution was then chilled in an ice bath and adjusted to pH 6.5 with 6N NaOH. The precipitate was collected (14 g). Because of the low yield in the precipitation, the remaining aqueous mixture was evaporated in vacuo leaving a white solid mixture of 2-indaneglycine and salt. This second solid was carefully washed with cold water to remove most of the salt. Both solids were combined (4, (O,L)-2-indaneglycine, -15 g, 0.066 mol, 21 %, mp 322-324 C19). ('H-NMR in D,O and NaOD, 300 MHz; o 7.4-7.14 (m, 4H);3.54 (d, 1H, J=7.5 Hz); 3.05-2.9 (m, 4H); 2.55-2.54 (m, 1 H); RP-HPLC single peak, gradient 0-40% Bin 40 min, retention time 13.24 min at 215 nm) 19

PAGE 33

2.2.2.4 N-Acetyi-(D,L)-2-indaneglycine (D,L)-2-Indaneglycine (4, -15 g, 0.066 mol, Figure 2.2) was dissolved in 200 mL water and 250 mL 2N NaOH, then chilled in an ice water bath and stirred. Acetic anhydride (5.6 mL) was added, followed by eight successive additions of 56 mL 2N NaOH and 5.6 mL acetic anhydride. The pH was maintained at pH 1 0-11. The solution was stirred at room temperature overnight and the progress of the reaction monitored by RP-HPLC (on a gradient of 0-40% Bin 40 min, the (D,L)-2indaneglycine has the retention time of 16 min and the acetylated amino acid has retention time 24.1 min). After 48 hr the reaction was 97% complete as judged by disappearance of the peak for the free amino acid. The solution was then chilled and adjusted to pH 3 with 6N HCI. After standing in the cold for several hr the precipitate was collected and washed with a small amount of cold water. The product was suspended in water and extracted with ethyl acetate. The organic layer was then evaporated in vacuo leaving a slightly yellowish solid (5, N-acetyi-(D,L)-2-indaneglycine, 12 g, 0.05 mol, 76%; single peak on RP-HPLC with gradient 0-40% Bin 40 min., retention time 24 min; mp 201-202 C; 'H-NMR, in DMSO, 300 MHz; o 8.29 (d, 1 H, ..1=8.25 Hz); 7.22-7.11 (m, 4H); 4.32 (m, 1 H); 2.94-2,74 (m, 20

PAGE 34

5H); 2.52 (m, 1 H); 1.89 (s, 3H). NaOH + COOH (4) Figure 2.2 N-Acetyi-(D,L)-2-indaneglycine 2.2.3 Resolution of (D,L)-2-Indaneglycine (Figure 2.3) N-Acetyl-(0 ,L)-2-indaneglycine (5, 12 g, 0.05 mol) was suspended in 500 mL of phosphate-buffered saline solution (PBS) at pH 7.4 and stirred at room temperature. 50 mg porcine kidney acylase I (E.G. 3.5.1.14; 1 ,430 units I mg solid; 2,125 units I mg protein) was added and the pH monitored continuously and readjusted as needed with dilute NaOH. After hr 21

PAGE 35

another 50 mg acylase I was added. The hydrolysis reaction was allowed to run for 72 hr and then another 50 mg acylase I was added. Reaction progress was monitored with RP-HPLC for the disappearance of the peak corresponding to the acetylated amino acid and reappearance of the peak at 16 min corresponding to the free lgl. After 5 days the peaks at 215 nm integrated as 48% and 51%, respectively, and the reaction was stopped. Specifically, 300 ml of ice water and 600 ml ethyl acetate were added to the solution and the mixture was carefully acidified to pH 0.75 with 6N HCI. The two layers were filtered through Celite and separated. The water phase was extracted twice with ethyl acetate. The aqueous solution was evaporated in vacuo and the residue dissolved in 30 ml water and 15 ml 6N HCI. The solution was then adjusted to pH 5.5 with concentrated ammonium hydroxide (NH40H). The precipitate was collected and washed with cold water over a medium-pore glass fritt. Filtrate was again evaporated in vacuo and the solid retained and recrystallized from water to give white flakes (6, (L)-2-indaneglycine19 3.8 g, 0.019 mol, 73%; mp 2962970 C, [a)25 0 = 50.35, c=1; Single peak 1H-NMR, in 020 and NaOD, 300 MHz (Figure 2.4); o 7.31-7.20 (m, 4H); 3.28 (d, 1 H, ..1=7.29); 3.02-2.95 (m, 2H); 2.85-2.6 (m, 3H); Two dimensionai'H-NMR in DMSO and NaOD, 400 MHz (Figure 2.5)). 22

PAGE 36

Figure 2.3 Resolution of (D,L)-2-Indaneglycine (5) Porcine Kidney Acylase I E.C.3.5.1.14 23 + 0 + {7)

PAGE 37

' Figure 2.4 Proton NMR of (L)-2-Indaneglycine (L)-2-Igl in 020 with NAOD EXP1 PULSE SEQUENCE: S2PUL DATE 01-21-96 17: 19:29 SOLVENT 020 fiLE H IJ:OOISJTJOH DEC. 'YT Til 1.000 llN !.000 S!l 4500.5 00 -<150.0 AT 2.<1!!6 llN IHl II' IW !. 0 PW 5.0 I)J' 20 Pi 0 HllNO H 01 !.DOD 02 0 PROCSSI!ii TO 0 N.ITil I HT CT 64 DISPlAY Pll90 17.0 SP -90.! TIN N NP 3090.7 fB 2250 YS 1765 BS 64 sc 2 ss 0 IIC 396 IL N IS 6!7 IH y 1231.7 DP y 0 AI.OCl( H TH 36 INS 1.000 I I 9 I I 8 -"' 0 m"' ''"" ru ,._ ,._ I I ::; gj ,_:o "'"' I I I 7 L,.-1 .w.o 24 y y I 6 y y COOH I I 5

PAGE 38

y llljllllllllljllllllllljllllllllljiiiiJIIII 3.2 3.0 2.6 2.6 PPM 10.2 2U 33.3 I I I I I I I I I I I I 1 I 4 3 2 1 0 PI L,.J TL..,.-J !0.2 33.3 21.4

PAGE 39

Figure 2.5 Magnitude COSY Spectrum of (L)-2-Indaneglycine w c ru c e -' c a ao z '"" fa c w -== o=o == 0 = 0 == == 0 =<= """"" = = = = a Ill 0 0 -=-=. 25 i i i I <:> = 0 oco 0 ; .... .. -

PAGE 40

The ethyl acetate phase (single peak on RP-HPLC) was dried over MgS04 and evaporated in vacuo. The solid residue (7, N-acetyl-(0)-2indaneglycine, mp 204-204.6 C) was refluxed in 150 ml 6N HCI overnight (Figure 2.6). The solution was evaporated in vacuo and then taken up in 50 ml 6N HCI and 50 ml water, after which the pH was adjusted to pH 5.5 with concentrated NH40H and a white precipitate collected (RP-HPLC >98% pure at 215 nm, gradient 0-40% B in 40 min, retention time 15.6 min, small impurity at 14.25 min/254 nm). The remaining liquid was adjusted to pH < 1 and extracted with ethyl acetate. The ethyl acetate was evaporated in vacuo and the white solid collected (RP-HPLC same as previous precipitate). Both solids (flakes) were combined and analyzed (8, (0)-2indaneglycine, 3.5 g, 0.017 mol, 67%; [af5 0 = -22.32, c=1in 1N HCI; 1H NMR, in OMSO and NaOO, 300 MHz; o 7.3-7.2 (m, 4H); 3.28 (d, 1 H, .1=7.3 Hz); 3.012-2.95 (m, 2H); 3.85-2.6 (m, 3H)). Although, the rotation was found to be in the opposite direction, it was not of equal intensity as that of the L-lgl, which may in part be explained by the small amount of L-lgl contamination in the sample (see next section). 26

PAGE 41

Figure 2.6 N-Acetyi-(D)-2-indaneglycine (7) 6N HCI reflux 8 h. 2.2.4 GITC-Derivatization (Figure 2.7) 0 + (8) The chiral agent 2,3,4,6-tetra-0-acetyi-13-D-glucopyranosylisothiocyanate (9, GITC, 2.9 mg, 7.4 IJmol, FLUKA) was dissolved in 1.5 mL CH3CN. D-lgl or L-lgl (5 mg, 0.025 mmol) was dissolved in 10 mL acetonitrile/water/triethylamine (50:50:0.4; v/v/w). GITC (50 iJL, 0.25 iJmol) was added to 50 iJL of each lgl solution(0.125 IJmol), the solutions were vortexed and aliquots analyzed via RP-HPLC on a gradient of 10-50% B in 40 min. A racemic mixture of lgl had been derivatized and separated previously on RP-HPLC to establish resolution. The D-lgl derivative (10, 27

PAGE 42

Figure 2.7 GITC Derivatization of (L)and (D)2-lndaneglycine. Both, (L)and (D)-Igl, were derivatized with GITC and injected onto analytical HPLC separately and as a mixture. Frame A shows the mixture, which was run as a standard to establish separation. Frame B shows the chromatogram of (L)-Igl and frame C that of (0)-lgl. 0 (6,8} OA01 A, Slr2'J5,10 ...,._,100ol01-21-411m'ta.o501.0 mAll 1000 000 (tt) A 600 c 28

PAGE 43

retention time 33.0 min) contained <3% of L-lgl as a contaminant and the L-lgl derivative (retention time 31.7 min) contained no detectable 0-lgl. 2.2.5 Tert-butyloxycarbonyl-lgl (Figure 2.8) The tert-butyloxycarbonyl (Boc) group was introduced according to the method of Moroder et al.20 0-lgl (8, 3 g, 0.013 mol) dissolved in a mixture of 26 ml dioxane, 13 ml water and 13 ml 1 N NaOH was stirred in an ice-water bath. Oi-tert-butylpyrocarbonate (3.12 g, 0.0143 mol) was added and stirring continued for 0.5 hr at room temperature. The reaction was monitored via RP-HPLC by appearance of a peak at 40.4 min on a gradient of 0-40% B in 40 min. When the reaction was judged to be -80% complete, the reaction mixture was concentrated in vacuo to -13 ml, cooled in an ice-water bath and covered with a layer of ethyl acetate (30 ml). The mixture was then acidified to pH 2-3 with 6N HCI. The phases were separated and the aqueous phase was extracted twice with ethyl acetate. The ethyl acetate extracts were pooled, washed twice with 30 mL water each, dried over anhydrous Na2S04 and then evaporated in vacuo to give a white solid (11, 2,2 g Boc-0-lgl; single peak on RP-HPLC, 29

PAGE 44

purity, for both; 1H-NMR, in DMSO, 300 MHz; i5 7.2-7.1 (m, 4H); 3.6 (s, 1 H); 2.9-2.75 (m, 5H); 1.41 (s, 9H)). The Boc-L-Igl (12) was prepared in the same manner (2,3 g; single peak on RP-HPLC, >99%purity; 1H-NMR, in DMSO, 300 MHz; i5 7.19-7.1 (m, 4H); 3.59 (s, 1 H); 2.94-2.74 (m, 5H); 1.4 (s, 9H)). Figure 2.8 Tert-butyloxycarbonyl-lndaneglycine ;; ;; dioxane 11 H 20, OH NaOH, H2N 0 0 0 S'C 0 0 '}-(11) (6,8) OH +co, 30

PAGE 45

2.3 B-9430 Analogs 2.3.1 Solid Phase Peptide Synthesis Table 2.1 shows the peptides which were synthesized using standard Boc chemistry. Table 2.1 Peptides Synthesized by Solid Phase Peptide Synthesis B-9430 0-Arg-Arg-Pro-Hyp-Giy-.L-Igi-Ser-0-lgi-Oic-Arg-OH Peptide A H-Lys-0-Arg-Arg-Pro-Hyp-Giy-.L-Igi-Ser-0-lgi-Oic-Arg-OH Peptide B 0-Lys-Arg-Pro-Hyp-Giy-L-Igi-Ser-Q-Igi-Oic-Arg-OH The N-terminal amino group of peptide A and the e-amino group of O-Lys0 in peptide B were protected with a fluorenylmethoxycarbonyl (FMOC) group. The FMOC N-protecting group was chosen because it is stable to HF-cleavage and thus allows the directed coupling of CP-955 to the remaining unprotected amine following HF-cleavage. All three peptides were completed from the common 'stempeptide' Boc-Pro-Hyp-Giy-L-Igi-Ser(O-Bzl)-0-lgi-Oic-Arg(Tos)-PAM-resin, according to Merrifield as described by Stewart and Young21 Amino acids (except 0and L-lgl, which were prepared as described above), resin and coupling 31

PAGE 46

reagents were purchased from Advanced ChemTech and Bachem California. All solid phase chemistry was performed manually in a 250 mL glass reaction vessel (ACE Labglass) with a medium-pore glass frill at the bottom. Mixing was achieved by bubbling nitrogen gas in from the bottom of the reaction vessel. The Boc-Arg(Tos)-Pam-resin (5 g, 0.44 mmol/ g, total 2.2 mmol) was placed in the reaction vessel and washed four times with dimethylformamide (DMF) and methylene chlorid}'l (DCM, 10-15 min each), and the solvents then removed. Deprotection was carried out by adding 80 ml deprotection reagent, 50% trifluoroacetic acid (TFA) in DCM, and mixing for 2 min and then repeating with fresh deprotection reagent while mixing for 30 min. Following deprotection, the resin was washed extensively (3 x DCM, 3 x DMF, 3 x DCM; 80 ml for 30 sec each) and then neutralized with 10% NN-diisopropylethylamine (DIEA) in DCM (3 x 50 ml for 1 min each). The resin was again washed (3 x DCM, 2 x DMF, 2 x 20%, 1-methyl-2-pyrrolidinone [NMP] in DMF; 80 mL for 30 sec each) and a small resin sample was removed from the reaction vessel for a qualitative ninhydrin test according to Kaiser, et al.27 After a positive ninhydrin test for free amines the preactivated amino acid (Boc-Oic) was added in DMF and coupling allowed to proceed overnight (all subsequent couplings were allowed to proceed for 4-8 hr). All couplings were carried out with a 3-fold 32

PAGE 47

molar excess of activated amino acids (6.6 mmol). Activation of amino acids employed the carbodiimide method described previousl/8 Diisopropylcarbodiimide (DIC, 2mL per coupling, 7.7 mmol) with addition of N-hydroxybenzotriazole (HOST, 1.2 g, 7.7 mmol) were dissolved in 20 mL 20% NMP in DMF, and the amino acid (6.6 mmol) was added for in situ activation. After 20 min, the mixture was added to the resin and another 30 mL of 20% NMP in DMF was added. After each coupling step the resin was washed (3 x DMF, 3 x DCM; 80 mL for 30 sec each) and a ninhydrin test performed. With a negative ninhydrin test the deletion peptides were capped by reaction with acetic anhydride (10% acetic anhydride, 10%DIEA, 80% DCM; v:v:v) for 10 min. In fact, no double couplings were required. The peptide resin was then washed and deprotected as described above and the cycle repeated for the next amino acid. Following the coupling of the prolyl residue, the stem-peptide on resin was analyzed for amino acid composition (Table 2.2). 33

PAGE 48

Table 2.2 Amino Acid Analysis of StemPeptide Amino Solution Calculated Theoretical acid cone. (pico ratio ratio ole) Arg 580.092 0.6 1 Hyp 1168.396 1.2 1 Ser 1034.503 1.08 1 Gly 1115.232 1.16 1 Pro 1385.629 1.44 1 ABA 234.407 N/A >99% Oic 873.913 .91 1 lgl 1531.18 1.59 2 TOTAL: 7688.945 7.98 8 Table 2.2 shows the calculated and theoretical ratios of amino acids. Ratios were calculated as follows: The total amounts in picomoles detected are summed (here 7343.26 pmol), excluding contributions from background peaks and breakdown products. The sum was divided by the total number of amino acids expected to be present in an integral amount per peptide sample. Then the picomole amount found for each amino acid was divided by n times the average integral amount, where n is equal to the number of times the particular amino acid is theoretically present in the 34

PAGE 49

peptide. The result was subsequently compared to the theoretical value. The results in Table 2.2 show a large deviation from the expected value in the case of proline and indaneglycine. As these data were obtained directly from a peptide-resin sample, without prior cleavage or removal of protecting side chain groups, some deviation may be expected due to increased experimental error. The stem peptide was split into four batches ( -0.5 mmol each) from which the three final peptides were made. After the coupling of the final amino acid each to each peptide, the N-terminal Boc group was removed as described above. All three peptides were then again analyzed as peptide-resins by amino acid analysis (data not shown). The amino acid compositions found for the final peptide-resins on resin were in good agreement with the theoretical composition of the peptides. 2.3.2 Cleavage of Peptide from Resin and Post-cleavage Workup Cleavage of the peptides from the resin was achieved with anhydrous HF and anisole (10:1, v:v) at 5C for.2 hr, according to Stewart and Young21 Next, the cleavage mixture was filtered through a sintered glass fritt and the residue washed with small amounts of glacial acetic acid 35

PAGE 50

in water (30:70 v:v). Cold ethyl ether was added to the filtrate until the peptide precipitated. The precipitate was collected on a sintered glass fritt and washed with additional cold ethyl ether. The solid was redissolved in water and small amounts of glacial acetic acid ( 1 0-25% by volume), and the solution filtered in preparation for HPLC purification. In each case, the peptide solution was then immediately taken onto purification. 2.3.3 Purification and Characterization All peptides were purified by RP-HPLC as described in Materials and Methods section. Gradients for the crude extracts were optimized with small aliquots on the analytical system prior to loading onto the preparative HPLC system. Fractions collected were checked by analytical RP-HPLC for single peak content and by mass spectral analysis (MS) for compound identification. Fractions deemed to be identical based on mass and retention time (on RP-HPLC) determination of the dissolved solute were pooled and lyophilized. Final yields are shown in Table 2.3. The yield for B-9430 was expected to be higher compared to peptides A and 8, as its retention time was known. Although peptide A was only 95% pure as 36

PAGE 51

Table 2.3 Yields of Peptide Synthesis Compound Actual %Purity by Yield (mg) RPHPLC (at 215 nm) B-9430 116.8 >98% Peptide A 7 95% Peptide B 13.6 >99% Table 2.4 Characterization of Peptides B-9430 Peptide A Peptide B Amino Calculated Theoretical Calculated Theoretical Calculated Theoretical acid ratio ratio ratio ratio ratio ratio Hyp 1.21 1 1.06 1 1.18 1 Ser 1.05 1 0.86 1 0.94 1 Gly 1.05 1 0.97 1 0.99 1 Arg 3.33 3 2.96 3 2.16 2 Pro 0.72 1 0.82 1 0.94 1 Oic 0.59 1 0.81 1 0.86 1 Lys 0 0 0.74 1 0.93 1 lgl 2.05 2 1.79 2 1.99 2 Total 10 10 11 11 10 10 37

PAGE 52

judged by HPLC peak integration, it was not repurified due to the small amount of material recovered. Amino acid analysis (Table 2.4) and mass spectral analysis (Lasermat and FAB, Table 2.5) were obtained for each final peptide. Amino acid analysis for B-9430 showed that the proline and octahydro-indole-2-carboxylic acid ratios were significantly lower than expected. Since B-9430 had been characterized by coelution with a fully characterized standard (Figure 2.9; standard provided by Ved Srivastava) and the mass was found to be in good agreement with the theoretical value (see Table 2.5), the deviation is believed to be due to experimental error or incomplete hydrolysis. The lysine value of peptide A was also found to be low. All three peptides, B-9430, peptide A and peptide B, were subjected to mass spectrometry using both, LASERMAT (MALDI-TOF) and fast atom bombardment techniques (Table 2.5). The results from both techniques confirmed the expected M+H mass within experimental error for both peptides A and B. Both instruments had been calibrated as described in the Materials and Instrumentation section above. Figure 2.9 shows the chromatograms and data obtained in the purification of B-9430. The mass analysis data for B-9430 are shown in Figure 2.10 (here only one standard at mass 1600.2 was used for calibration). 38

PAGE 53

Table 2.5 Mass Analysis of Peptides Compound Found Mass Found Mass Theoretical Mass LASERMAT FAB +H B-9430 1341 1339 1338.74 Peptide A 1690.3 1689 1690.2 Peotide B 1534.5 1533 1534.8 39

PAGE 54

>1>-0 RP-HPLC/prep Flow: 10 ml I min monitored at 230 nm +,.l lr I Gradient %8 (acetonitrile, 0.1 % TFA) h ... :: Pooled fractions 2,3 and 5 coeluted OOD with 8-9430 standard on analytical : RP-HPLC. -!t_. .20 "" MoW. -G :: {\ I 2,3 and 5 were pooled. & 1 2,3 and 5 each showed single .s.. peak on analytical AP-HPLC with same retention time and had identical molec. ion on LASERMAT.) \ -n ca c 0 -OJ '. w 0

PAGE 55

Figure 2.10 Mass Analysis for B-9430 ! II 15 i! !! "' !3 t -"'1!1 ;, ... 'Eg i!!j w, EO> "' "C = 8 i: ;: E 0-E
PAGE 56

2.4 Conjugation of CP-955 to the B-9430 Analogs CP-955 (20 mg, 45 !Jmol), which had been synthesized and characterized by Douglas L. Gernert, was activated with 0-benzotriazole N,N,N,N-tetramethyl-uronium-hexafluorophosphate (HBTU, 20.5 mg, 54 !Jmol; Advanced ChemTech) and N,N-diisopropylethylamine (DIEA, 12 !JL, 671Jmol) for 30 min. Peptide A (13, 30 mg, 22.4 !Jmol) was added and the reaction mixture was stirred overnight at room temperature. The reaction product was then precipitated with 15 ml cold diethyl ether, centrifuged and the resulting supernatant decanted. The pellet was resuspended in 50 ml 30% acetic acid in water and loaded onto a preparative RP-HPLC column. The fractions collected were checked by both analytical RP-HPLC and mass spectral analysis (Lasermat). Fractions 6 and 7 eluting at -67% B, both displayed a single peak, with identical retention time, and a M+H mass of 1768.6 (theoretical 1768.5). Both fractions were pooled and lyophilized. The final lyophilized material was again analyzed by RP-HPLC (compound I, single peak, gradient 1-100% Bin 25 min, retention time 15.3 min, Figure 2.11) and submitted for mass spectral analysis with FAB (M+ 1 found: 1768, plus Na peaks, Figure 2.11 ). The total yield was 24.5 mg ( 62%, 141Jmol). 42

PAGE 57

Peptides A and B were coupled to CP-955 in the same manner as described for B-9430, using the same molar equivalents. The resulting compounds were labeled compound II and compound Ill, respectively. The results for all three conjugates are listed in Table 2.6. Masses found for compounds I, II and Ill using both, the LASERMAT (MALDI-TOF) and fast atom bombardment techniques (not showing M+sodium peaks) indicate that the molecular ions were in agreement with theoretical calculations. Purity determination was made by RP-HPLC monitored at 215 nm and was based on the integrated area under peaks as a percentage of total peak area. Since only 2.4 mg were recovered for compound II, no attempt was made to repurify the material. Table 2.6 Mass Analysis of Conjugates HPLC Yield purity Compound Found Mass Found Theoretical at 215 nm (mg) LASERMAT Mass Mass FAB +H Compound I 1768.6 1768 1768.9 95% 24.5 Compound II no data 1896 1896.6 85% 2.4 Compound Ill 1740 1740 1741 95% 3.9 43

PAGE 58

Figure 2. 11 Mass Analysis of Compound I 1!-ii 44

PAGE 59

2.5 Biological Assays 2.5.1 HNEi Assay The inhibition constants for human neutrophil elastase were determined by Sherman E. Ross using the method described by Cunningham et al.22 with some minor modifications. (All pharmacological assays following hereafter were performed by JohnS. Zuzack and Michael R. Burkard, Department of Pharmacology, Cortech, Inc.) 2.5.2 Bradykinin B2 Guinea Pig Ileum Receptor Binding Assay This assay was performed with binding assay conditions similar to those previously described (Manning et al., 1986)23 Guinea pig ileum (gpi) membranes were prepared by Analytical Biological Services, Inc. 45

PAGE 60

2.5.3 Bradykinin B2 Human Receptor Clone Binding Assay This assay was periormed as described previously24 using a human bradykinin 82 receptor, expressed in CHO-K1 (ATCC) cells. 2.5.4 Bradykinin B1 Human Lung Fibroblast Receptor Binding Assay This assay was periormed as described for the BK2 human receptor clone binding assay24 with some modifications. Human lung fibroblasts IMR-90 cells (ATCC) were used, which express the BK1 human receptor. The cells were incubated with IL-1 [3 (200 pg I ml) for 3 hr. Tritiated desArg 10 kallidin was used as the ligand. 2.5.5 Bradykinin B2 Antagonist Functional Testing on Guinea Pig Ileum Tissues were prepared by a method similar to that described by Cheronis et al.25 Briefly, 1cm long strips of ileum from male Hartley guinea pigs (350-400g; SASCO) were prepared as previously described.One end 46

PAGE 61

of each strip was tied to a support rod and the other to a Grass FT03 force displacement transducer. Tissues were allowed to equilibrate in siliconized tissue baths for 45 min under 2 g resting tension at 37C in gassed (95% 02 and 5% C02 ) Kreb's solution. The tissues were initially primed with 1 iJM BK followed by a 45-min wash period. The potency of the 82 agonist, BK, was measured by recording the contractile response to a range of BK concentrations and was expressed as the EC50 (the concentration producing half maximal contraction). To determine antagonist potencies cumulative BK concentration-response curves were repeated in the same tissue prior to and following a 15-min exposure to the antagonist. Three antagonist concentrations were tested in the same tissue. Data were calculated as percent of the maximum contraction in the control curve. Shifts in the EC50 were used to calculate the pA2 (negative log of the molar concentration of antagonist causing a two-fold increase in the agonist EC50 ) for the antagonist by the method of Schild (1947)26 As a control, the effect of DMSO, which was used to solubilize the antagonists, on the BK response curve was established. 47

PAGE 62

3. RESULTS AND DISCUSSION 3.1 Chemistry The overall yields in the synthesis of (D,L)-2-Igl (with exception of the final step) were similar to those cited in the literature19 Due to the relatively low yield in synthesizing the 2-bromoindane using PBr3 an alternative route may be explored in the future. The resolution of the (D,L)2-Igl gave good results, although the time required for the reaction to go to completion was longer than expected. It is likely that much of the initially added acylase denatured because of the pH change that accompanied the generation of acetic acid. A fairly large volume of a buffer with an ionic strength similar to physiological conditions would have to be used to completely buffer such a change in pH. High-ionic-strength buffer may be a another solution. Although other synthetic routes using chiral reagents may be used to arrive at optically pure isomers, the reagents required are typically much more expensive than the acylase. Additionally, the acylase is an ideal enzyme for future scale up of the method, as it can be covalently bound to solid resins without significant loss of activity29 30 48

PAGE 63

Chirality assignment for the L and 0 enantiomers of lgl was based on the enzymatic reaction. It was assumed that Acylase I preferentially hydrolyzes L-amino acid derivatives31 The fact that complete resolution was achieved, as shown by the GITC-derivatization, further supports the assignment because incomplete resolution would have been expected if the enzyme was not completely specific for the L-enantiomer. Further support comes from the fact that the B-9430 standard made in this study had the expected activity. All BK B1/2 antagonists explored to date require an aromatic 0-amino acid in position 7. An inadvertent exchange between the L-lgl5 and 0-lgf would likely result in an inactive peptide. To date no absolute assignment of the enantiomers has been made in published studies. To facilitate more complete assignments in the proton NMR spectrum of (L)-2-indaneglycine, a COSY spectrum was obtained (in OMSO and NaOO, 400 MHz; Figure 2.5). A COSY spectrum is a two-dimensional NMR spectrum in which the one-dimensional spectrum lies along the diagonal of a two-dimensional frequency map, and an off-diagonal cross-peak at (w., wb) indicates that resonances at w. and wb (i.e., the diagonal peaks at (w., w.) and (wb, wb), respectively) ar.e 49

PAGE 64

J-coupled. A magnitude COSY spectrum is symmetric about the diagonal. J-coupling is usually only observed between protons that are two or three bonds apart. Hence, this information can be used to determine which structural groups are adjacent in the molecule. The off-diagonal cross peak (3.28,2.65) in Figure 2.5 shows coupling between the a-proton (3.28,3.28) and the sextet (2.65,2.65), which can then be assigned as the [3-proton. The remaining two multiplets can be assigned to they-protons which are cis and trans to the [3-proton. Due to the chirality of the amino acid and the limited rotation of the bulky side chain, all four y-protons are magnetically nonequivalent. An absolute assignment to match coupling constants to the cis and trans y-protons cannot be made from this data. For such an assignment a NOESY experiment would be required. The standard B-9430, compound I and compound Ill were obtained in final purities of at least 95% as judged by HPLC peak integration of absorbance peaks obtained at 215 nm. Compound II was only -85% pure, as judged by HPLC. Compound II was synthesized from the two principal components, Peptide A and CP-955, which were 95% and 99% pure respectively (by HPLC). The retention time of the contaminant 50

PAGE 65

associated with compound II did not match the retention times of either unreacted CP-955 or Peptide A. The contaminants are therefore believed to be due to coupling of CP-955 at the side chain of the D-Arg0 residue rather than at the N-terminal amine. This is further supported by the fact that only one mass-ion peak was found in the MS, which corresponds to compound II. 3.2 Biology As DMSO was used to make stock solutions of the various compounds tested in the BK 82 antagonist functional assay on gpi, a DMSO dose response curve was established (Figure 3.1). The DMSO concentrations here are the same as the final DMSO concentrations in the samples tested for activity. The first two sets, 0.01% DMSO (by volume) for 1 hr and 0.1% DMSO for 2 hr, did not significantly change the dose response curve as compared to the control with no DMSO. The highest concentration, 1% for 3 hr, did result in a depression of the dose response curve, which reflects a decrease in the percent of maximum contraction. The same effect can be seen with all compounds tested, as the maximum 51

PAGE 66

Figure 3.1 Effect of DMSO on BK Dose Response Curve. -control Effect of OMSO on bradykinin dose response curve in guine -o-.01%. 1 hr neum -<>-.l%.2hr :::: c .Q 80.00 g -E 12.00 11.00 10.00 9.00 8.00 7.00 6.00 5.00 4. -log brcdykinin cone. (M) 52 60.00 8 E 40.00 20.00 -20.00 l 0 .. -o-1%.3hr --.-.""Recovery --Senes7 --Senes8 Senes9 Sertes10 --Senes12 --Sertes13 Senes14 Sertes15 --Sertesl7 --Sertes18 Sertesl9 Senes20 --Sertes22 --Sertes23

PAGE 67

contractile response is reduced at the highest concentrations. This is quantitatively expressed as a significant reduction of the slope in the Schild plots of all compounds. A slope of 1, which corresponds to the retention of the maximum contractile response, is interpreted as reversible competitive binding32 A reduction of the slope in the Schild plot, corresponding to a reduction of maximum contractile response achieved at high concentrations, usually indicates noncompetitive binding32 In the case of the compounds discussed below, the slope reduction is most likely due to the effect of DMSO, and the binding is still believed to be of reversible, competitive nature. In support of this interpretation is the fact that B-9430, which has previously been shown to be a competitive ligand33 using an aqueous solvent, had the most severe reduction of maximum contraction in this study (Schild plot slope 0.67 .11; Figure 3.2). The pA2 values for compounds I and II (Figures 3.3 and 3.4 respectively) are essentially the same as that of the B-9430 standard, within experimental error. The pA2 values are 8.32.38 for B-9430, 7.95 .25 for compound I, 8.39 .24 for compound II, and 7.62 .28 for compound Ill (Figure 3.5). This means that, in this functional assay, compound I retained full activity on the BK B2 receptor compare!l to 53

PAGE 68

Figure 3.2 Effect of B-9430 on BK Dose Response Curve ,\1 e l .2 Effect of B-9430 on BK does response curve in guinea pig ileum bradykinin eone. (M) .00 e I e 0 Q E E ;; .. -+-control -a-t.OOE-o7 ----.J.OOE-06 --o-l.OOE-05 -- v -Recovery --Sertes7 --SerlesB -SerieslO --Ser1es12 --Series13 -Serles14 Serles15 --Seriesl7 --Serlesl8 -Seriesl9 Serles20 --Serles22 --Serles23 Schild plot of B-9430 antagonism of bradykinin effect in guinea pig ileum 3 2.5 2 I 1.5 0.5 4 4.5 5 5.5 0 I 6 6.5 7 7.5 -log B-9430 cone. (M) 54 8 pA2=8.32.0.38 slope=0.67.0.11 recovery=not done+% 8.5 9 9.5

PAGE 69

Figure 3.3 Effect of Compound I on BK Dose Response Curve Effect of Compound I on BK does response curve in guinea pig ileum 120.00 N=7 100.00 60.00 Q 60.00 Q E 40.00 j 0 E 20.00 "0 .. --+-Conrro! -o-1.00E-Q7 -o-l.COE..QS -- -o RecovetY --Series? --Series8 Sel'ies9 SerieslO --Serles12 --Ser!es13 -Series14 Seriesl5 --Seriesl7 --Seriesl8 Series19 12.00 11.00 10.00 9.00 6.00 7.00 6.00 5.00 4 ro.oo -5eries20 .00 log bradykinin cone. (M) --Serles22 --Series23 Schild plot of Compound I antagonism of bradykinin effect In guinea pig ileum 3 T R I pA2=7.95+0.25 si0Pe=0.87+0.10 recovery=not done+% N a g t I 1.5- .!l 0.5 0 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 log Compound I cone. (M) 55

PAGE 70

Figure 3.4 Effect of Compound II on BK Dose Response Curve. Effect of Compound II on BK does response curve in guinea pig ileum 100.00 80.00 c 0 = 60.00 u E c 0 u 40.00 E 0 12000 E 0 .. I 0.00 --o-1.00E..Q7 --o--t.OOE-QS --Recovery --Selies7 --Series8 Sertes9 SeneslO --Serlesl2 --Sertesl3 Sertes14 Ser1es15 --Seriesl7 --Serlesl8 Ser1esl9 12.00 11.00 6.00 !2000 Selies20 log bradykinin cone. (M} --Selies22 --Selie.23 Schild plot of Compound II antagonism of bradykinin effect in guinea pig ileum 35 I pA2=8.39+0.2.4 slope=0.82+0.08 31 recovery:not done+% N=3 25 I I 21 a 1.51 "' .!! 1 f 05 I o. 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 -log Compound II cone. {M) 56

PAGE 71

Figure 3.5 Effect of Compound Ill on BK Dose Response CuNe. _.,_Control Effect of Compound Ill on BK does response curve in -<>-1.ooe-o7 guinea pig ileum -+-l.OOE-D6 --o-1.ooe-os 120.00 --- -a Recovery --Series7 --Sen..a Senes9 SefieslO --Series12 --Sertes13 -Seriesl4 Serlesl5 12.00 11.00 10.00 9.00 8.00 7.00 6.00 5.00 --Seriesl7 --Series18 log bradykinin cone. (M) -Series19 Series20 --Series22 --Series23 Schild plot of Compound Ill antagonism of bradykinin effect in guinea pig ileum 1.: t pA2=7.62+0.28 g slcpe=0.6l +0.09 1.6-recovery=not done+% N=7 i "j 1 0.8 D 0.6 0.4 li a 0.2 0 4 4.5 5 5.5 6 6.5 7 7.S 8 8.5 9 9.5 .Jog Compound Ill cone. (M) 57

PAGE 72

B-9430 and compound II had equivalent activity. The pA2 for compound Ill was lower by approximately a half logarithmic unit, suggesting some loss of activity, although it is not known what activity is inherent in the parent peptides for compounds II and Ill. The pKb values (pKb = -log{[ antagonist concentration] I (IC50 ANTAGONisT) I (IC50)} ) were determined because of the observed depression of the contractile response. The pKb values confirm that the activity was essentially retained in compounds I (pKb=7.64 .08) and II (pKb=7.97 .08), with a slight improvement in compound II, as compared to the B9430 standard (pKb=7.58 .1 0). Compound Ill (pKb=6.83 .15) again showed a small decrease in activity. Table 3.1 is a summary of the biological data obtained on the compounds, including receptor binding and functional assay data, as well as the HNE inhibition constants. In the BK B1 receptor binding assay B-9430 had a piC50 of 8.07, whereas compound I was significantly less active with a piC50 of 6.74. Compounds II and Ill had piC50 values of 7.04 and 6.09, respectively. In the BK B2 receptor binding assay, the B-9430 standard had a 'piC50 of 9.64, and piC50 values of 8.42 for compound I, 8.00 for compound II, and 58

PAGE 73

lJ1 \.0 Table 3.1 Summary of Biological Data Receptor binding assay Compound BKB1 BKB2 BKB2 (human) (human) (gpi) B-9430 8.07 9.64 9.13 -1 0 1 2 3 4 5 6 7 8 9 6.74 8.42 7.67 1)-Arg-Arg-Pro-Hyp-Giy-L-Igi-Ser-l)-lgi-Qic-Arg-OH (I) CP-955 linked at Arrf N-lenn. amine. -1 0 1 2 3 4 5 6 7 8 9 7.04 8.00 7.49 H-Lya-t!-Arg-Arg-Pro-Hyp-Giy-L -lgi-Sert!lgi-Oic-ArgOH (II) CPinked at lys1 e-amlne. 0 1 2 3 4 5 6 7 8 9 6.09 7.10 6.75 t!-Lys-Arg-Pro-Hyp-Giy-L-Igi-Ser-1)-lgi-Qic-Arg-QH (Ill) CP-9551inked at Lys0 N-lerm. amine. CP-955 NIA NIA NIA Guinea pig funcllonal assay Inhibition B2(gpi) B2(gpi) Schild's HNEKi functional functional slope (nM) pA2 pKb 8.32 7.58 0.67 N/A .38 .10 0.11 7.95 7.64 0.87 2.4 .25 .08 .10 8.39 7.97 0.82 not .24 .08 .08 active 7.62 6.83 0.61 not .28 .15 .09 active N/A N/A NIA 10.5 ------------------

PAGE 74

7.10 for compound Ill. Although in the receptor binding study all conjugates had more than one logarithmic unit less activity as compared to 8-9430, it is important to note here that activity was higher against the human clone 8K 82 receptor compared to the guinea pig ileum 8K 82 receptor with all three conjugates. Compound I has the highest retention of 8K 82 antagonism in receptor binding. Compound II had the highest retention of 8K 81 antagonism. This correlates to the observation made in other studies that retention of 8K 81 activity is directly linked to retention of the positive charges on the N terminal end of the peptide. Here, the Lys1 extension provides the functional group for conjugation, thus preserving the positive charges on the arginine residues in positions 0 and 1. The receptor-binding data indicate that there is a general selectivity for the 8K 82 receptor over the 8K 81 receptor, especially in the case of compound I, which has nearly a hundredfold selectivity. In the 8K 82 functional assay, all three conjugates essentially had the same activity as 8-9430, with a possible slight improvement seen with compound II. Since, functionally, the 8K 82 antagonism of the conjugates 60

PAGE 75

on gpi was equivalent to B-9430, compared to severe losses in the BK B2 gpi receptor assay, it would be of interest to know what the functional human data for BK B2 might look like. Further testing, especially of compound I, in a human functional assay is therefore indicated. Compound I showed a fourfold improvement of the inhibition constant toward neutrophil elastase, resulting in a K; of 2.4 nM compared to a K; of 10.5 nM for CP-955. Similar improvements have been seen in other studies when HNE inhibitors were conjugated to various molecules (Axel H. Leimer, internal communications, 1996). Several studies have shown that extensions on the P unprimed end of HNE inhibitors have little effect on the K; (internal communications, 1996). These extensions are thought to be in solution, outside the shallow binding-pocket of HNE. There is no obvious explanation for the loss of HNE inhibition with compounds II and Ill, in particular for compound Ill, which differs from compound l only by the D arginine0 to D-lysine 0 substitution. In both compounds the inhibitor is linked at the N-terminus. Preliminary modeling of these compounds and energy minimization in the elastase binding pocket gave no obvious indication as to why these compounds might be inactive toward the enzyme(modeling conducted by Val S. Goodfellow). The heterodimer is very soluble in 61

PAGE 76

aqueous solution compared to the highly insoluble CP-955. Overall, the results of the study suggest that it is possible to link HNE inhibitors to 8-9430 and retain both the HNE inhibition and at least 8K 82 antagonism, and thus arrive at a heterodimer with dual activity. Since the role of the 8K 81 receptor is still uncertain, 8K 81 antagonism may not be advantageous in a drug. Compound I, which is essentially selective for the 8K 82 receptor over the 8K 81 receptor and has complete retention of activity in the 8K 82 functional assay, as well as an improved HNE K;, would be a good starting point for a drug that does not target 8K 81 receptors. Should the 8K 81 receptor prove to be an important target in drug therapy, compound II data indicates that it is possible to somewhat improve 8K 81 antagonism, and through further SAR studies, possibly retain 8K 81 antagonism. As further SAR work is indicated by the results obtained from the present study, it should also include various linkers between 8-9430 and CP-955. A polypharmacological agent, such as compound I, may be capable to intervene at multiple points in the inflammatory process of 180. As both, the kinin system and tissue degradation by elastase, are important 62

PAGE 77

components of chronic inflammatory diseases, compound I or similar heterodimers could be expected to be potent therapeutic agents. In lBO one would hope to see significant reduction of the inflammatory response and tissue edema. As we begin to appreciate the complexity of many disease states, in particular inflammation, we see the inherent difficulty in treating patients with a single drug, which is active only on one target. Multiple-action drugs are therefore likely to be regarded with increasing interest. The prersent study has shown that it is possible to design and synthesize single molecule drugs, which combine both BK 82 antagonism and HNE inhibition. In addition to the possible therapeutic advantages there is a direct cost benefit with polypharmaceutical agents. The multi-active heterodimer allows for the development and safety profiling of one drug, as compared to combination therapy, where both drugs would have to be profiled individually and as a mixture. 63

PAGE 78

APPENDIX A Animal Use Certification April3, 1996 University of Colorado, Denver RE: Cortech Protocol #46-93-NP To Whom it May Concern: The purpose of this letter is to inform you that the animal smdies performed to support Axel Leimer's Master's thesis were done in our USDA licensed animal facility under the auspices of all applicable federal regulatior&S. The methods described in the protocol were reviewed and approved by the Cortech Animal Care and Use Committee prior to the initiation of this project. Please contact me if I can be of any further assistance in this matter. Sincerely, J:ftt. .. Manager, Preclinical Projects Director. Animal Research Center 64

PAGE 79

APPENDIX 8 Animal Use Protocol The following is a copy of Cortech's Institutional Animal Care and Use Committee Protocol which was followed for the functional assay contained in this thesis: Protocol # 46-93-NP ANIMAL USE PROTOCOL Submit completed protocol form to the EPU Director. DATE FILED: 6/16/93 INVESTIGATOR AND RESEARCH ASSOCIATES:Dr. Eric Whalley, Dr. William Selig, John Zuzack, David Cuadrado, Michael Burkard TITLE OF PROTOCOL: In vitro Pharmacological Studies Using Animal Tissues CHECK ONE: Initial review XXX Renewal Pilot project Start date 6/16/93 End date6116196 I. ABSTRACT OF PLANNED USE OF ANIMALS; Write a brief yet complete description of the planned use of animals. Animals (mice, rats, hamsters, guinea pigs, rabbits) will be humanely euthanized and selected tissues will be dissected out. The female rats 65

PAGE 80

will be pre-treated 12-18 hours before euthanasia with 100 mg/kg of di ethyl stilbestrol subcutaneously. These tissues will include vas deferens, thoracic aorta, mesenteric artery, jugular vein, pulmonary artery, uterus, ileum, lungs, trachea, and bladder. Where possible multiple tissue types will be used from the same animal. The tissues will be prepared and used for in vitro pharmacological assays. Other tissues will be used as required. II. A. Please state your rationale for the use of living animals in the study. Also include a discussion why an alternative non-living model wasn't chosen. The tissues chosen are known to contain specific receptors which are essential for assay with specific classes of compounds being developed by Cortech. Measurement of the magnitude of response of these tissues to test compounds will allow selected compounds to be chosen for assay in in vivo models and ultimately be developed for use as therapeutic agents in man. Non-living alternative models which yield these biological responses are not available. 66

PAGE 81

B. Explain the appropriateness of the species chosen. Literature reports have established the above stated tissues from these species as excellent models for various receptors because they have a high degree of specificity for one receptor type and have an easily measured response to receptor activation. C. Justify the numbers of animals planned for this study. Please attach an experimental plan outlining the numbers in each treatment/control groups. Document the statistical relevance. While it is difficult to accurately predict the numbers of these experiments required in the future, the following is a reasonable estimate: On a yearly basis there may be 150 new compounds synthesized by our chemists. Given the polytherapeutic nature of our drug design strategy, each of these compounds must be tested for multiple activities. This necessitates testing each compound on approximately 3 tissue types. Efforts are made to take more than one tissue type from each animal. Depending on the tissue type we can obtain approximately 4 pieces of tissue per animal. It is necessary to repeat each experiment a minimum of four times to reduce the variance of response between tissues. Thus, one animal is re"quired per 67

PAGE 82

compound per receptor type (tissue variety). These facts can be used to estimate the animal requirements for these studies: 150 compounds X 3 tests (tissue type) X 4 tissues /4 tissues per animal = 450 animals. Ill. ANIMAL USE A. Species Sprague-Dawley rat, Syrian hamster, Hartley guinea pig, New Zealand White rabbit, ICR mice B. Number: per year 200 rats, 75 guinea pigs, 30 hamsters, 75 rabbits, 100 mice entire study same numbers housed at any one time 20 rats, 12 guinea pigs, 10 hamsters, 8 rabbits, 10 mice C. Sex and age (or weight range): either sex for all species, any size D. Source (vendor): Sasco, Harlan Sprague-Dawley, R&R Rabbitry 68

PAGE 83

E. Describe any expected health changes in the animals due to your experimental procedures, e.g. weight loss, food intake and/or behavioral changes, and indicate treatment. none F. Will the animals suffer any pain and/or distress during the experiment (excluding survival surgery procedures)? Yes No XXX If yes, explain. G. Describe abnormal environmental conditions that may be imposed. This includes temperature, lighting, etc. none H. Describe and justify the use of any restraint devices employed: none 69

PAGE 84

I. Will special housing, conditioning, diet or other conditions be required: Yes No XXX If yes, explain: J. Does the protocol involve animal surgery? Yes No XXX Survival Non-survival If survival surgery is involved, provide the following information: Surgeon's name and qualifications: Anesthetist's name and qualifications: Ill. K. Complete description of the surgical procedures: L. For anesthesia procedures, provide the following information: Pre-anesthesia: Drug: Dose: Anesthesia: Drug: Dose: How is anesthesia level monitored? What are the provisions for post-surgical care and include the name(s) of the 70

PAGE 85

person(s) responsible for care? What drugs are to be used to alleviate pain or distress and what are their dosages? If analgesics are intentionally withheld, please justify in writing. M. If your procedures require animal immunizations and/or bleeding, provide the following information: Antigen(s): Adjuvant(s): Injection volume, frequency and site(s): Blood collection methods, volume and frequency: N. What form of euthanasia will be used and when? Give agent, dose and method. Animals will be euthanized by rapid asphyxiation with carbon dioxide. In animals where the lungs are to be used an overdose of pentobarbital (100 mg/kg i.v. or i.p.) or urethane (3 g/kg i.p.). 0. Who will be euthanizing the animals? Give name and qualifications. 71

PAGE 86

Eric Whalley, Ph.D. 20+ yrs experience in academic/industrial labs William Selig, Ph.D approx. 15 yrs experience in pharmaceutical companies' research labs John Zuzack, M.S.-10 years experience in pharmaceutical companies' research labs David Cuadrado, B.S. 6 mo training in Cortech pharm lab Michael Burkhard, B.S.5 yrs experience in pharmaceutical companies' research labs IV. A. XXX TOXIC AND HAZARDOUS SUBSTANCES Will any of the following be used in the animals? 1. Infectious agents 2. Radioisotopes 3. Toxic chemicals or carcinogens 4. Recombinant DNA 5. Experimental drugs 6. Malignant cells or hybridomas If yes, describe all procedures necessary for personnel and animal safety including biohazardous waste and carcass disposal and cage 72

PAGE 87

decontamination: All experimental compounds of unknown toxicity are treated as though they were toxic in compliance with OSHA regulations. Extreme care will be taken to avoid contamination of personnel with these chemicals. B. If transplantable tumors or hybridoma cells are to be injected into the animals, have the tissues/cells been tested for inadvertent contamination by viruses or mycoplasma? Yes No If yes, give results of the test: V. CATEGORIES OF ANIMAL EXPERIMENTATION BASED UPON LEVEL OF MANIPULATION AND PAIN: XXX NP No pain. This category includes any procedure that would be conducted on a human being without analgesia. Although a slight degree of discomfort may be associated with the procedure, in general it will not produce any escape behavior beyond that of psychological distress. This category also includes animals humanely killed_without any 73

PAGE 88

treatments, manipulations, etc., but will be used to obtain tissue, cells, sera, etc. MP Minor pain. This category includes any procedures that main cause the animal mild pain or distress, either directly caused by the procedure or occurring after time. These procedures include ascites production, most toxicity studies, etc. ***** D Pain alleviated with drugs. Any protocol which calls for an invasive procedure, or will cause more than momentary or slight distress will be performed with appropriate analgesia, sedation, or anesthesia. Appropriate medications will be determined with the coordination of a veterinarian. It should be remembered that muscle relaxants and paralytic drugs such as succinylcholine or curariform drugs are not analgesic nor anesthetic. They cannot be used alone for surgical restraint. ***** P Those protocols which necessitate painful procedures that cannot be conducted with the use of analgesics, anesthetics or tranquilizers because the use of such agents would defeat the purpose of the research. Such procedures must be justified in writing, approved by the committee 74

PAGE 89

and be conducted under the direct supervision of the investigator. **'** *****SPECIAL SECTION***** Please fill out this section if these procedures will cause more than slight or momentary pain or distress to the animals. Explain why these procedures are necessary and unavoidable. Provide a written narrative of alternative procedures/protocols considered and rejected. Include a description of the methods and sources (Medline, Animal Welfare Information Center) used, or logic used to make your determination. We require that the principal investigator conduct a literature search for two purposes: 1) to determine that the study isn't duplicative; 2) to determine if the study can be done less painfully or on non-whole animal models.) A literature search has been conducted, and to the best of my knowledge, this study is not duplicative and other methods are inappropriate or unsuitable for the purpose of the study. 75

PAGE 90

REFERENCES 1. In The Merck Manual of Diagnosis and Therapy, 16th edition, Merck Research Laboratories, 1992. 2. Loftus, Edward V. Jr., Sandborn, W. J. Drug Therapy for Inflammatory Bowel Disease. Contemp. Int. Med., 1995, 7, 21-34. 3. In Cellular Pathophysiology, Sayeed, M., Ed.; CRC Press, 1989. 4. Springer, T. A. Adhesion Receptors of the Immune System, Nature, 1990, 346, 425-434. 5. Osborn, L. Leukocyte Adhesion to endothelium in Inflammation, Cell, 1990, 62, 3. 6. Maling, H. M., Webster,M. E., Williams, M. A., Saul, W., Anderson, W. Inflammation Induced by Histamine, Serotonin, Bradykinin and Compound 48/80 in the Rat: Antagonists and Mechanisms of Action. J. Pharmacal. Exp. Ther., 1974, 191: 300-310. 7. Miles, A. A., and Wilhelm, D. L. The Activation of Endogenous Substances Inducing Pathological Increases of Capillary Permeability. In The Biochemical Response to Injury; Stoner, H. B., Eds.; Blackwell, Oxford, 1960, 57-79. 8. Wilhelm, D. L., Mill, P. J., Sparrow, E. M., McKay, M. E., Miles, A. A. Enzyme-like globulin from Serum Reproducing the Vascular Phenomena of Inflammation. IV. Activable Permeability Factor and its Inhibition in the Serum of the Rat and the Rabbit. Brit. J. Exp. Path., 1958, 39, 228-250. 9. Regoli, D., Barabe, J. Pharmacology of Bradykinin and Related Kinins, Pharmacological Reviews,, 1980, 32, 1-45. 76

PAGE 91

20. Moroder, L., Hallett, S., Wunsch, E., Keller, 0., Wersin, G. Hoppe Seyler,s Z. Physiol. Chern., 1976, 357, 1 651. 21. Stewart, J.M., Young, J.D. In Solid Phase Peptide Synthesis, Freeman, San Fransisco, 1969. 22. Cunningham, R. T., Mangold, S. E., Spruce, L. W., Ying, Q. L., Simon, S. R., Wieczorek, M., Ross, S., Cheronis, J. C., Kirschenheuter, G. P. Synthesis and Evaluation of CE-0266: A New Human Neutrophil Elastase Inhibitor. Bioorg. Chern., 1992, 20, 345-355. 23. Manning, D. C., Vavrek, R., Stewart, J. M., Snyder, S.H. Two Bradykinin Binding Sites with Picomolar Affinities. J. Pharmacal. Exp. Ther., 1986, 237, 504-512. 24. Goodfellow, V. S., Marathe, M. V., Kuhlmann, K. G., Fitzpatrick, T. D., Cuadrado, D., Hanson, W., Zusack, J. S., Ross, S. E., Wieczorek, M., Burkhard, M., Whalley, E. T. Bradykinin Receptor Antagonists Containing N-Substituted Amino Acids: In Vitro and In Vivo B2 and B1 Receptor Antagonist Activity. J. Med. Chern., 1996, in press. 25. Cheronis, J. C., Whalley, E. T., Nguyen, K. T., Eubanks, S. R., Allen, L., Duggan, M. J., Loy, S. D., Bonham, K. A., Blodgett, J. K. A New Class of Bradykinin Antagonists: Synthesis and In Vitro Activity of Bissuccinimidoalkane Peptide Dimers. J. Med. Chern., 1992, 35, 1563-1572. 26. Schild, H.O. pA2 A New Scale for the Measurement of Drug Antagonism, Br. J. Pharmacal., 1947, 2,189. 27. Kaiser, E. T., Colescott, R. L., Bassinger, C.D., Cook, P. I. Color test for detection of free terminal amino groups inthe solid-phase synthesis of peptides. Anal. Biochem., 1970, 34, 595-598. 28. Atherton, E., Sheppard, R. C. In Solid Phase Peptide Synthesis; A Practical Approach; IRL Press, Oxford, 1989. 78

PAGE 92

10. Stewart, J. M. In The Kinin System in Inflammation, in Pro teases, Protease Inhibitors and Protease-Derived Peptides, Cheronis, Repine Eds.; Birkhauser, 1993. 11. Cheronis, J. C., Whalley, E.T., Allen, L.G., Loy, S.D., Elder, M.W., Duggan, M.J., Gross, K.L., and Blodgett, J.K. ,Design, Synthesis and in Vitro Activity of Bissuccinimidohexane Peptide Heterodimers with Combined BK1 and BK2 Antagonist Activity, J. Med. Chern. 1994, 37, 348. 12. Regoli, D., Barabe, J., Park, W. K. Receptors for bradykinin in rabbit aorta. Can. J. Physiol. Pharmacal., 1977, 55, 855-867. 13. Park, W. K., St-Pierre, S., Barabe, J., Regoli, D. Synthesis of Peptides by the Solid Phase Method. Ill. Bradykinin: Fragments and Analogues. Can. J. Biochem., 1978, 56,92-100. 14. Schroder, E. Structure-activity relationships of kinins. Handb. Exp. Pharmacal., 1970, 25,324-350, i 970. 15. Stewart, J.M. Chemistry and Biologic Activity of Peptides Related to Bradykinin. Handb. of Exp. Pharmacol.25(supp.), 1979, 227-272. 16. In the convention used here, the amino acid residues, starting at the scissile bond and going out towards the carboxyl terminal of the peptide, are labelled 1' (one prime), 2', 3', etc .. The residues, starting from the scissile bond and going toward the amino terminal, are labelled 1 ,2,3, and so on (unprimed). 17. Ernst, R. R., Bodenhausen, G., Wokaun, A. In Principles of Nuclear Magnetic Resonance in One and Two Dimensions; Clarendon Press, Oxford, 1987. 18. Bidlingmeyer, B. A., Chen, S. A., Tarvin, T. L. Rapid Analysis of Amino Acids Using Pre-column Derivatization, J. Chrom., 1984, 336, 93-104. 19. Porter, T.H., Shive, W. DL-2-Indaneglycine and DL-13Trimethylsilylalanine, J. Med. Chern., 1968, 11, 402. 77

PAGE 93

29. Rosell, C. M., Fernandez-Lafuente, R., Guisan, J. M. Resolution of Racemic Mixtures by Synthesis Reactions Catalyzed by Immobilized Derivatives of the Enzyme Penicillin G Acylase. J. Malec. Cat., 1993, 84, 365. 30. Guisan, J. M., Alvaro, G., Fernandez-Lafuente, R. Stabilization of Heterodimeric Enzyme by Multipoint Covalent Immobilization Biotech & Bioeng., 1993, 42, 455. 31. Bommarius, A. S., Drauz, K., Klenk, H. Operational Stability of Enzymes: Acylase-catalyzed Resolution of N-Acetyl Amino Acids to Enantiomerically Pure L-Amina Acids. Annals., 1992, 672, 126. 32. Palmer, T., Paul, A. I. In Principals of Drug Action; Pratt, W. T., Taylor, P. Eds.; Churchille Livingston: New York, 1990; 60-63. 33. Lajos, G., Stewart, J. M., Whalley, E., Hanson, W., McCullough, R. New Bradykinin Antagonists Having High Potency at both 81 and 82 Receptors. Proceedings 14th Am. Pept. Symp.; Escom: Leiden, 1995. 79