Design of low molecular weight hematoregulatory agents from the structure-activity relationship of a dimeric pentapeptide

Article abstract of DOI:10.1021/jm9702443

We report herein, a new class of simple hematoregulatory semipeptides, formally derived from the cystine-dimerized peptide pGlu-Glu-Asp-Cys-Lys-OH, where the disulfide bond has been replaced by an isosteric dicarba bridge. The structure-activity relationship (SAR) of a series of analogues incorporating replacements at positions 1 and 2 of peptide 1 led to the design of active conformationally constrained cyclic peptides (12, 13). Ring closure was achieved by cyclization of the N-terminal amino groups at position 2 of peptide 2 using pyrazine-2,3-dicarboxylic acid. Subsequent excision of the putative C-terminal scaffold domain from the active cyclic peptides resulted in the discovery of a new class of low molecular weight hematoregulatory agents exemplified by compound 16. This semipeptide analogue, comprising two D-Ser residues connected via amide bonds to the acid groups of pyrazine-2,3-dicarboxylic acid, had comparable biological activity to the lead peptide 1. The stereochemical requirements for the observed biological activity of these novel compounds were examined. Furthermore, the hematopoietic synergistic activity induced by compound 16 in stromal cell cultures was blocked by an antibody known to neutralize the hematoregulatory effect of 1, indicating a common mechanistic end point. Compounds of the class typified by 16 may form the basis for the development of novel therapeutic agents within the area of immunoregulation.

Full text of DOI:10.1021/jm9702443

2
876
J . Med. Chem. 1997, 40, 2876-2882
Design of Low Molecu la r Weigh t Hem a tor egu la tor y Agen ts fr om th e
Str u ctu r e-Activity Rela tion sh ip of a Dim er ic P en ta p ep tid e
Alan S. Cuthbertson,* Mette Husbyn, May Engebretsen, Michael Hartmann, Meinolf Lange, J essie Sandosham,
Peter M. Fischer, Hege Fjerdingstad, and Dagfinn Løvhaug
Nycomed Imaging AS, Bioreg Research, Gaustadall e´ en 21, N-0371 Oslo, Norway
X
Received April 11, 1997
We report herein, a new class of simple hematoregulatory semipeptides, formally derived from
the cystine-dimerized peptide pGlu-Glu-Asp-Cys-Lys-OH, where the disulfide bond has been
replaced by an isosteric dicarba bridge. The structure-activity relationship (SAR) of a series
of analogues incorporating replacements at positions 1 and 2 of peptide 1 led to the design of
active conformationally constrained cyclic peptides (12, 13). Ring closure was achieved by
cyclization of the N-terminal amino groups at position 2 of peptide 2 using pyrazine-2,3-
dicarboxylic acid. Subsequent excision of the putative C-terminal scaffold domain from the
active cyclic peptides resulted in the discovery of a new class of low molecular weight
hematoregulatory agents exemplified by compound 16. This semipeptide analogue, comprising
two D-Ser residues connected via amide bonds to the acid groups of pyrazine-2,3-dicarboxylic
acid, had comparable biological activity to the lead peptide 1. The stereochemical requirements
for the observed biological activity of these novel compounds were examined. Furthermore,
the hematopoietic synergistic activity induced by compound 16 in stromal cell cultures was
blocked by an antibody known to neutralize the hematoregulatory effect of 1, indicating a
common mechanistic end point. Compounds of the class typified by 16 may form the basis for
the development of novel therapeutic agents within the area of immunoregulation.
In tr od u ction
As part of a general structure-activity relationship
SAR) study surrounding peptide 1, a picture emerged
(
The process known as hematopoiesis, whereby stem
cells proliferate and differentiate into mature blood cells,
of a pharmacophore region comprising the two dipeptide
1
2
sequences pGlu -Glu shown in Figure 1. The amino
acid residues at positions 3-5 were believed to form a
salt-bridged framework with the two pharmacophore
halves protruding from the scaffold-like structure.
has a critical role in protection against pathogenic
invasion.1 Several proteinaceous factors termed colony-
stimulating factors (CSFs), including G-CSF and GM-
CSF, have already demonstrated clinical utility by
enhancement of host defense mechanisms.
In 1982, Paukovits isolated a thiol-containing com-
pound from mature granulocytes with potent myelo-
poietic inhibitory activity. The active component was
1
2
2
Bhatnagar et al. reported that substitution at position
with picolinic acid and at position 2 with Ser resulted
1
in a peptide, 2, with significantly improved potency.
In this study we report on how the existing SAR
surrounding peptide 2 was extended by additional
substitutions at position 1 and how this information was
used to design conformationally constrained cyclic ana-
logues. The observed biological activity of cyclic pep-
tides 12 and 13 (Table 2), where the pharmacophore
region had considerably reduced mobility, showed that
it was possible to lock the critical functional elements
into a bioactive conformation. Subsequent removal of
the now redundant scaffold region, known to be es-
sential in noncyclic analogues, led to the design of the
novel class of semipeptides shown in Table 3, with full
HSF-inducing activity.
3
presumed, following analysis and chemical synthesis,
to be a pentapeptide of amino acid sequence pGlu-Glu-
Asp-Cys-Lys.4 In direct contrast to the inhibitory effect
of the monomeric peptide, dimerization through disul-
fide formation generated a peptide with potent stimu-
latory activity.5 Subsequent replacement of the disul-
fide bond with an isosteric ethylene spacer resulted in
the novel nonreducible hematoregulatory peptide 1
Figure 1).6 This peptide demonstrated potent anti-
(
7
infective activity in models of bacterial sepsis and
Candida albicans infection, significantly increasing the
8
survival rates of treated animals given a lethal chal-
lenge of pathogen. It has been shown to stimulate
indirectly hematopoiesis9 and enhance effector cell
function by the induction of a proteolytically modified
form of the chemokine KC from murine bone marrow
Ch em istr y
Peptides 1-13 were synthesized using solid-phase
techniques based on the 9-fluorenylmethoxycarbonyl
1
0
13
stromal cells.
The truncated form of this protein,
(Fmoc) amino protection methodology employing resin
consisting of residues 5-72, was isolated from super-
natants of a marine stromal cell line following stimula-
tion with peptide 1. Furthermore, KC[5-72] had no
effect on GM-CFC colony formation alone but was a
synergistic factor along with either M-CSF, G-CSF, GM-
linkers cleavable with trifluoroacetic acid (TFA). N,N′-
Di-Fmoc-protected L,L-2,7-diaminosuberic acid (Sub)
was prepared by Kolbe electrolysis of Boc-Glu-OBn
using the experimental conditions described by Nutt et
al. Modifications to the standard assembly protocols
were necessary to obtain the dimeric target structures.
Acylation of the H-Lys(Boc) resin with 0.33 mole equiv
of N,N′-di-Fmoc-protected L,L-2,7-diaminosuberic acid
was performed followed by an additional coupling cycle
1
4
1
1
CSF, or IL-3.
KC[5-72] has therefore been termed
hematopoietic synergistic factor (HSF).
X
Abstract published in Advance ACS Abstracts, August 1, 1997.
S0022-2623(97)00244-6 CCC: $14.00 © 1997 American Chemical Society

A New Class of Hematoregulatory Semipeptides
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 18 2877
F igu r e 2. ‘Bell-shaped’ dose-response curve obtained when
HSF-containing supernatants from C6.4 cells stimulated with
1
µg/mL peptide 1 were added to GM-CFC cultures in the
presence of a suboptimal amount of L929 cell-conditioned
medium.
commercial peptide synthesis resin substituted with
t
Fmoc-L-Ser(Bu ). The diacid component was first coupled
in large excess to the deprotected resin using 0.5 mole
equiv (with respect to acid groups) of the coupling
reagent O-(7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyl-
1
6
uronium hexafluorophosphate (HATU). After washing
with DMF, a second 0.5 mole equiv of HATU was added
and preactivation of resin-bound acid groups allowed
to proceed for 5 min before addition of an excess of H-D-
t
t
Ser(Bu )-OBu . Semipeptides containing two Ser resi-
dues of the same chirality were synthesized in solution
using the appropriate aromatic diacids and tert-butyl-
protected (side chain and carboxyl group) L- or D-Ser
derivatives.
All compounds submitted for biological testing were
purified to homogeneity by preparative reverse-phase
HPLC. Final purity and structural integrity were
assessed by analytical HPLC, quantitative amino acid
analysis, NMR, fast atom bombardment mass spectrom-
etry (FAB-MS), high-resolution mass spectrometry (HR-
MS), and polarimetry, as appropriate. Diastereomeric
F igu r e 1. Structure and numbering system used for peptide
1
and its more potent analogue 2.
1
7
18
with dicyclohexylcarbodiimide (DCC) and 1-hydroxy-
benzotriazole (HOBt) in order to ensure anchoring of
both carboxyl groups of the Sub residue to the Lys
amino groups on the solid support. Excess resin-bound
amino groups were then capped by acetylation prior to
continuation of chain extension. On-resin N-terminal
cyclization in the case of peptides 11-13 was ac-
peptides and semipeptides were additionally ana-
lyzed by HPLC cochromatography.
Biology
In vitro biological activity was related to the amount
of HSF induced following incubation of test compound
with the established bone marrow-derived stromal cell
t
t
12
complished as follows: [Fmoc-Ser(Bu )Asp(OBu )]2-Sub-
Lys(Boc)]2 peptidyl resin was Fmoc deprotected with
line C6.4.
The HSF contained in the cell super-
[
natants, together with a suboptimal amount of CSFs
from L929 cell-conditioned medium, significantly in-
creased the number of colonies formed in a GM-CFC
assay compared to CSF alone. In practice, the activity
was calculated from a dose-response curve using a
dilution series of conditioned medium from the peptide-
stimulated cell line. The activity induced by compound
1 in this assay, shown in Figure 2, followed a bell-
shaped dose-response curve as previously reported.12
The HSF activity was calculated by dividing the in-
crease in colony number over background by the small-
est volume (mL) of supernatant giving rise to a signifi-
cant increase in colonies. The values, given in HSF
units/mL of supernatant, are an average of several
experiments where the mean value obtained from each
piperidine in N,N-dimethylformamide (DMF) and re-
acted with 0.5 mole equiv (with respect to resin-bound
amino groups) of pyrazine-2,3-dicarboxylic acid, acti-
vated with 1 mole equiv of benzotriazol-1-yloxytris-
(
pyrrolidino)phosphonium hexafluorophosphate (Py-
1
5
BOP). This was followed by a second step using only
coupling reagent to effect cyclization. After complete
chain assembly, peptides were simultaneously side
chain deprotected and cleaved from the solid support
followed by precipitation of the crude material from
diethyl ether. The crude peptides were purified by
preparative reverse-phase HPLC.
Semipeptide analogues containing one L- and one
D-Ser residue (15, 20) were synthesized starting with a

2
878 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 18
Cuthbertson et al.
Ta ble 1. Extended SAR of Hematoregulatory Activity at
Ta ble 2. SAR of a Series of Cyclic Peptides
Positions 1 and 2
no.
Xaa
Yaa
HSF units/mLa
0
1
1
1
1
2
3
L-Ser
L-Ser
D-Ser
L-Ser
D-Ser
D-Ser
5
5.7 × 10
6
2.3 × 10
a
Cyclic Peptides tested at a dose of 100 ng/mL.
and the activity in the HSF assay determined (Table
). Peptides containing the position 1 substitutions
1
orotic acid, 5, 2-pyrazinecarboxylic acid, 6, 6-N-acetyl-
picolinic acid, 7, 3-hydroxypicolinic acid, 8, and 2-pyr-
rolecarboxylic acid, 10, were all found to be active, while
the larger heterocyclic rings of quinaldic acid, 3, 2-in-
dolecarboxylic acid, 4, and 1-isoquinolinecarboxylic acid,
9
, had no detectable activity. These results showed that
a diverse range of substitutions was tolerated provided
the steric bulk at position 1 was not significantly
increased. Furthermore, the presence of a nitrogen
atom R to the carboxyl group of position 1 was not
enough to overcome the apparent loss of biological
activity caused by the increase in bulk. These oberva-
tion were in agreement with the previously published
1
2
data for peptide 1 and may reflect the shape of the
receptor binding pocket of these compounds.
In an attempt to design a conformationally con-
strained cyclic peptide, we postulated that the structural
requirements described above would be maintained in
the two pharmacophore halves were replaced by a
bifunctional, monomeric mimetic. Knowing that the
pyrazine rings of peptide 6 were tolerated, we chose
pyrazine-2,3-dicarboxylic acid as a possible candidate
to effect ring closure of the position 2 Ser residues via
amide bond formation, as illustrated by the general
structure in Table 2. In this way it was possible to
retain both nitrogen atoms R to the dicarboxylic acid
while satisfying conditions of size and symmetry. To
our surprise peptide 11, the all L-Ser peptide, had no
detectable activity, while the peptides 12, containing one
D- and one L-Ser, and 13, containing two D-Ser residues,
had levels of activity comparable to peptide 1.
We rationalized from these results that the scaffold
region (positions 3-5) in the cyclic peptide series no
longer played a significant role in determining the
correct orientation at positions 1 and 2. Due to the
constrained nature of the pyrazine ring system, it
seemed likely that the pharmacophore mimetic alone
should possess the necessary functionality and confor-
mational integrity required for biological activity. It
was anticipated that the scaffold region of the cyclic
peptide had now become a redundant feature of the
molecule.
a
All peptides tested at a dose of 1000 ng/mL.
experimental group was compared to control using the
two-tailed Student’s t-test.
Polyclonal IgG antibodies, raised against the synthetic
peptide fragment 38-72 of the chemokine KC, were
added to the supernatants collected from peptide-
stimulated C6.4 cells immediately before addition to the
GM-CFC assay. Incubation with this antibody had no
effect on the background colony numbers produced in
the assay by the CSFs contained in the L929 cell-
conditioned medium.
1
9
Biologica l Resu lts a n d Discu ssion
The aim of this work was firstly to examine more
closely the SAR at position 1 of the highly potent peptide
. This, it was hoped, would provide further evidence
for the developing pharmacophore hypothesis and ulti-
mately permit the design of simpler, low molecular
weight hematoregulatory agents.
Substitution of the picolinoyl moiety of peptide 2 with
a variety of heterocyclic carboxylic acids was carried out
2
Semipeptides, representing the new constrained phar-
macophore (Table 3), were synthesized and tested for
their ability to induce HSF. As for the dimer peptides,
the activity was found to depend on the stereochemistry

A New Class of Hematoregulatory Semipeptides
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 18 2879
Ta ble 3. SAR for a Series of 2,3-Disubstituted Dicarboxylic
Acids Containing Combinations of D- and L-Ser
Ta ble 4. Neutralization of Colony-Stimulating Activity from
Active Supernatants on Incubation with Polyclonal IgG
Antibodies Raised against KC[38-72]
b
active supernatant,
HSF units/mL
active supernatant + antibody,
no.
HSF units/mL
1a
2
3
15
6
1.7 × 10
5
0
0
0
0
0
5
1
1
5.7 × 10
6
2.3 × 10
5
no.
X
Y
R1
R2
HSF units/mLa
0
1.6 × 10
5
1
1.9 × 10
1
1
1
1
1
1
2
2
4
5
6
7
8
9
0
1
N
N
N
CH
CH
CH
CH
CH
N
N
N
N
L-Ser-OH
L-Ser-OH
D-Ser-OH
L-Ser-OH
D-Ser-OH
L-Ser-OH
L-Ser-OH
D-Ser-OH
L-Ser-OH
D-Ser-OH
D-Ser-OH
L-Ser-OH
D-Ser-OH
L-Ser-OH
D-Ser-OH
D-Ser-OH
a
b
5
Test dose 1000 ng/mL. Others tested at 100 ng/mL. Antibody
1.6 × 10
5
used at 1:10000 dilution.
1.9 × 10
0
6
disubstituted ring system, as well as the presence of the
constraining amide bonds between the ring and the
D-Ser residues. Finally, analogue 24, derived from
pyrazine-2-carboxylic acid and D-Ser, showed that two
Ser residues were required for activity.
N
10 .2 × 10
0
CH
CH
CH
5
1.3 × 10
a
Semipeptides tested at a dose of 100 ng/mL.
The level of in vitro activity for each of the test
compounds was expressed in HSF units/mL. This
measure of activity relates directly to the amount of
HSF produced by the C6.4 cells on stimulation with a
given dose of test compound. For the dimer analogues
listed in Table 1, a dose of 1000 ng of peptide/mL of
stromal cell culture medium was found to be optimal,
giving the typical bell-shaped curve shown in Figure 2.
The active cyclic peptides and semipeptide analogues
gave a similar curve (data not shown) but at an optimal
dose of 100 ng of compound/mL of stromal cell culture
medium. The more rigid analogues therefore possessed
a specific activity 10 times greater than their more
flexible dimeric precursors. Furthermore, incubation of
supernatants from stromal cells stimulated with com-
pounds 1, 12, 13, 15, and 16 with an antibody raised
1
9
against KC[38-72] completely blocked synergistic ac-
tivity, implicating this chemokine product in the mech-
anism of both peptide and semipeptide analogues (Table
F igu r e 3. Inactive D-Ser analogues demonstrating the im-
portance of both the 2,3-disubstituted pyrazine ring system
and the constraining amide bonds.
4
).
We have shown therefore in this SAR study that it
was possible to design, in a rational manner, active
hematoregulatory agents of much reduced complexity
to peptide 1. The simplicity of this new class of
hematoregulatory semipeptides may preclude the dis-
covery of bioavailable analogues of increased utility.
Although the receptors have not yet been identified, a
broadening knowledge of the downstream molecular
events triggered by these compounds may pave the way
for the development of new therapeutic agents.
of the Ser residues. The L-Ser analogues 14, 17, and
1
9 were inactive, while a comparison of compounds 15
and 20, comprising one D- and one L-Ser residue coupled
to the pyrazine and phthalic acid rings, respectively,
demonstrated an activity requirement for the ring
nitrogen atom. In contrast, analogues 16, 18, and 21,
comprising two D-Ser residues in combination with the
pyrazine-2,3-carboxylic, quinolinic, and phthalic acid
ring systems, respectively, retained full activity, with
no obvious heteroatom dependence R to the carboxyl
groups.
The compound 16 was chosen as a model structure
to investigate more closely the effect on activity when
other functional groups were removed or modified. The
â-hydroxyl groups of the Ser side chains appeared to
be important in the semipeptide series. Analogues
containing either two D-Ala, two Gly, or one D-Ser and
one Gly residue were devoid of activity. The D-Ser
carboxylic acid termini in the pyrazine series were also
important since the C-terminal diamide analogue of 16
was inactive (unpublished results).
Con clu sion
We have further extended the SAR data of peptide 2,
incorporating a range of substituents at position 1.
From a series of tolerated substitutions, the pyrazine
ring system was chosen as a suitable template for the
synthesis of conformationally constrained cyclic pep-
tides. The fully cyclic analogues, 11-13, were prepared
using pyrazine-2,3-dicarboxylic acid as a monomeric
position 1 pharmacophore mimetic. The activity of
cyclic peptides 12 and 13 revealed a stereochemical
requirement for at least one D-Ser residue at position
2. Furthermore, excision of the conformationally con-
strained pharmacophore mimetic from the putative
scaffold region led to a series of active semipeptides
exhibiting the same stereochemical requirement as the
fully cyclic structures.
The three analogues 22-24, shown in Figure 3, were
designed to answer other crucial questions on semipep-
tide SAR. Compound 22, where two D-Ser residues were
incorporated into the pyrazine-3,6-dicarboxylic acid ring,
was inactive, as was the reduced amide bond analogue
The results reported here provide further evidence for
a model consisting of a scaffold region in peptide 1
2
3. This demonstrated a requirement for the 2,3-

2
880 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 18
Cuthbertson et al.
furnished by positions 3-5, which allow for the correct
orientation of the position 1 and 2 pharmacophore
halves. This model has been used with success in the
development of analogues of considerably reduced chemi-
cal complexity. The low molecular weight of the active
semipeptides described herein may lead to the develop-
ment of new classes of readily synthesizable and poten-
tially bioavailable immunoregulators.
NMM (0.2 mL, 2 mmol) in DMF (10 mL), was coupled to the
deprotected resin for 2 h. Following extensive washing with
DMF, a further Fmoc deprotection cycle was carried out.
Coupling of 1 mmol of pyrazine-2-carboxylic acid, preactivated
with PyBOP (0.52 g, 1 mmol), HOBt (0.13 g, 1 mmol), and
NMM (0.2 mL, 2 mmol) in DMF (10 mL), was then performed.
22
The coupling was monitored using the Kaiser ninhydrin test
and found to be complete after 3 h. The peptide resin was
then extensively washed with DMF, CH Cl , and Et O before
2
2
2
drying in a stream of nitrogen. Simultaneous deprotection and
cleavage of the peptide from the solid support was carried out
using 5% aqueous TFA. After a reaction time of 2 h the filtered
mixture was evaporated in vacuo and the crude peptide
Exp er im en ta l Section
P ep tid e Syn th esis. All reagents and solvents for peptide
synthesis were of reagent grade and were used without further
purification. The peptides were synthesized by solid-phase
techniques using a manual nitrogen bubbler apparatus.20 All
amino acid derivatives were purchased from Novabiochem and
heterocyclic acid derivatives used as supplied by Fluka.
Peptides were synthesized using 2-methoxy-4-alkoxybenzyl
alcohol resin (SASRIN) preloaded with Fmoc-Lys(Boc) as
supplied by Bachem. Amino acid couplings, except for di-
aminosuberic acid, were carried out in a stepwise manner
using a 10-fold excess of protected amino acid preactivated in
DMF with 1 equiv of PyBOP, 1 equiv of HOBt, and 2 equiv
N-methylmorpholine (NMM), unless otherwise stated. Fmoc-
2
precipitated and washed with Et O before air-drying. Follow-
ing purification by preparative reverse-phase HPLC (5-20%
MeCN gradient over 120 min) and lyophilization, pure peptide
6
(19 mg, 10%) was obtained.
P r oced u r e for th e Syn th esis of Cyclic P ep tid es 11 a n d
1
3. These peptides were obtained by a general method
exemplified by the description of peptide 11: [Fmoc-Asp-
t
(
OBu )]
Fmoc deprotected and washed. Fmoc-L-Ser(Bu )-OH (0.38 g,
mmol), preactivated with PyBOP (0.52 g, 1 mmol), HOBt
0.13 g, 1 mmol), and NMM (0.2 mL, 2 mmol) in DMF (10 mL),
2 2
-Sub-[Lys(Boc)] peptidyl resin (0.55 g, 0.2 mmol) was
t
1
(
was then coupled for 2 h. After renewed Fmoc deprotection
and washing, pyrazine-2,3-dicarboxylic acid (17 mg, 0.1 mmol),
preactivated with PyBOP (52 mg, 0.1 mmol), HOBt (13 mg,
t
t
t
L-Glu(OBu )-OH, Fmoc-L-Asp(OBu )-OH, and Fmoc-L-Ser(Bu )-
OH were used for stepwise solid-phase chain assembly, while
0
.1 mmol), and NMM (20 µL, 0.2 mmol) in DMF (10 mL), was
t
t
the H-L- or H-D-Ser(Bu )-OBu ‚HCl derivatives were used for
coupled for 10 h and then washed well with DMF. A second
portion of PyBOP (52 mg, 0.1 mmol), HOBt (13 mg, 0.1 mmol),
and NMM (20 µL, 0.2 mmol) in DMF (8 mL) was then added
to the resin, and coupling continued for a further 5 h. After
solution synthesis.
Ch r om a togr a p h y. Crude peptides 1-13 were dissolved
in H O containing 0.1% TFA, loaded onto a preparative HPLC
2
column (Vydac 218TP1022, 2.2 × 25 cm), and eluted at a flow
2 2 2
extensive washing with DMF, CH Cl , and Et O, the air-dried
rate of 5 mL/min using linear gradients of increasing MeCN
peptidyl resin was cleaved from the solid support in 5%
aqueous TFA. After 2 h the TFA solution was evaporated in
vacuo and the crude peptide precipitated and washed with
Et O before air-drying. The crude peptide was purified by
2
preparative HPLC (0-15% MeCN gradient over 40 min, 9 mL/
min) and lyophilized yielding 7.2 mg (7.8%) of pure peptide
concentrations in H
Fractions were collected and monitored by analytical HPLC
Vydac 218TP54, 0.46 × 25 cm column; 1 mL/min, gradient
elution typically from 0% to 30% MeCN in H O, 0.1% TFA over
0 min, λ ) 215 nm). Those fractions containing target peptide
2
O (constant concentration of TFA at 0.1%).
(
2
2
with purity > 98% were pooled and lyophilized. Final yields,
after chromatography, ranging from 7% to 15% pure peptide
wee typically obtained.
1
1.
P r oced u r e for th e Syn th esis of Cyclic P ep tid e 12.
t
[
2 2
Fmoc-Asp(OBu )] -Sub-[Lys(Boc)] peptidyl resin (0.55 g, 0.2
Semipeptides 14-24 were purified similarly using Vydac
mmol) was Fmoc deprotected, washed, and reacted with
compound 15 (34 mg, 0.1 mmol), preactivated with PyBOP (52
mg, 0.1 mmol), HOBt (13 mg, 0.1 mmol), and NMM (20 µL,
2
01HS1022 (2.2 × 25 cm) and Vydac 201HS54 (0.46 × 25 cm)
columns for preparative and analytical work, respectively. The
columns were developed by isocratic elution with H O contain-
2
0
.2 mmol) in DMF (8 mL). After a reaction time of 10 h, a
second portion of PyBOP (52 mg, 0.1 mmol), HOBt (13 mg,
.1 mmol), and NMM (20 µL, 0.2 mmol) in DMF (8 mL) was
ing 0.1% TFA. Those fractions containing target semipeptide
with purity > 98% were pooled and lyophilized. Final yields,
after chromatography, ranging from 60% to 70% pure semi-
peptide were typically obtained.
0
added to the resin, and coupling continued for a further 3 h.
Cleavage and purification procedures were identical to those
outlined above. Final yield after chromatography of pure
peptide 12 was 11 mg (11%).
Gen er a l P r oced u r e for th e P r ep a r a tion of [F m oc-Asp -
t
(
OBu )]
2
-Su b-[Lys(Boc)]
2
P ep tid yl Resin . Fmoc-Lys(Boc)-
SASRIN resin (5 g, 0.6 mmol/g), preswollen in DMF (5 mL),
was treated with 20% piperidine in DMF for 15 min to remove
P r oced u r e for th e Syn th esis of Sem ip ep tid es 14, 16-
19, 21, 22, a n d 24. These were obtained by a general method
exemplified by the synthesis of 16: Pyrazine-2,3-dicarboxylic
acid (1 g, 6 mmol) was activated with diisopropylcarbodiimide
(DIC) (0.93 mL, 6 mmol) in DMF (15 mL) for 10 min to form
the anhydride in situ. To this was added a solution of H-D-
the Fmoc group. After repeated DMF washing, Fmoc
2
-Sub-
(
OH) (648 mg, 1 mmol) was coupled to the deprotected resin
2
using DCC (1.03 g, 5 mmol) and HOBt (765 mg, 5 mmol) in
DMF (10 mL) for 10 h. After repeated DMF washing, a further
coupling cycle with similar quantities of DCC and HOBt in
DMF was performed for 5 h. Unreacted amino groups were
t
t
Ser(Bu )-OBu ‚HCl (1.7 g, 6.6 mmol) and NMM (0.69 mL, 6.6
mmol) in DMF (10 mL). The mixture was stirred for 4 h. A
second portion of DIC (0.93 mL, 6 mmol), along with HOBt
(0.81 g, 6 mmol), was then added followed, after 10 min, by
2
then capped by reaction with 10% (v/v) Ac O/DMF (25 mL)
for 15 min. Following a further Fmoc removal cycle, acylation
t
with the pentafluorophenol ester of Fmoc-Asp(OBu ) (5.78 g,
t
t
1
(
0 mmol) and HOBt (1.53 g, 10 mmol) was carried out in DMF
10 mL) for 2 h followed by extensive washing with DMF,
CH Cl , and Et O. After drying in vacuo, the peptidyl resin
H-D-Ser(Bu )-OBu ‚HCl (1.7 g, 6.6 mmol) and NMM (0.69 mL,
6.6 mmol) in DMF (10 mL). The mixture was stirred over-
night. Solvent was removed in vacuo and the residue redis-
2
2
2
was found to contain 0.36 mmol/g amino groups according to
spectrophotometric measurement of Fmoc release from a
weighed aliquot of resin.
solved in CH
2 2
Cl and extracted several times with 5% aqueous
NaHCO and water. The organic layer was dried over MgSO ,
3
4
2
1
filtered, and evaporated in vacuo. The resulting oil was
chromatographed on silica gel (70:25:5 hexane/EtOAc/AcOH),
and the fractions containing pure product were pooled and
evaporated. The residue was then treated for 30 min with 5%
aqueous TFA, evaporated in vacuo, redissolved in water, and
lyophilized to give 1.6 g (76%) of title compound 16. Aliquots
of bulk material were further purified by preparative HPLC
as described above.
P r ep a r a tion of P ep tid es 1-10. These were obtained by
a general method, exemplified by the description of peptide 6:
t
The above [Fmoc-Asp(OBu )
2
-Sub-[Lys(Boc)]
2
peptidyl resin
(0.27 g, 0.1 mmol) was treated with 20% piperidine in DMF
for 15 min to remove Fmoc groups. After repeated DMF
t
washing, Fmoc-L-Ser(Bu )-OH (0.38 g, 1 mmol), preactivated
with PyBOP (0.52 g, 1 mmol), HOBt (0.13 g, 1 mmol), and

A New Class of Hematoregulatory Semipeptides
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 18 2881
P r oced u r e for th e Syn th esis of Sem ip ep tid es 15 a n d
An tibod ies to HSF . Polyclonal IgG antibodies were raised
2
0. These were obtained by a general method exemplified by
in goats against a fragment of synthetic KC polypeptide
corresponding to the sequence 38-72. The IgG fraction was
t
19
the synthesis of compound 15: Fmoc-L-Ser(Bu )-Wang resin
1 g, 0.5 mmol/g; from Novabiochem) was Fmoc deprotected
(
purified using protein G/Superose-12 (Pharmacia) and was
found to be more than 98% pure. The antibody preparation
(3.8 mg/mL) was used at 1:10000 dilution, which completely
blocked the activity of recombinant KC[5-72] as well as the
HSF activity induced by 1 µg/mL compound 1. The antibody
had no effect on the background colony numbers produced by
the suboptimal amounts of L929 cell-conditioned medium used
in the GM-CFC assay.
as previously described. After repeated DMF washing, pyr-
azine-2,3-dicarboxylic acid (0.35 g, 2 mmol), preactivated with
HATU (0.76 g, 2 mmol) and NMM (412 µL, 4 mmol) in DMF
(10 mL), was coupled for 2 h to the deprotected resin followed
by washing with DMF (10 mL). On-resin preactivation of acid
groups with HATU (0.76 g, 2 mmol) and NMM (412 µL, 4
mmol) in DMF (15 mL) was performed for 5 min before the
t
t
addition of H-D-Ser(Bu )-OBu ‚HCl (0.5 g, 2 mmol) and NMM
0.2 mL, 2 mmol) in DMF (5 mL). Coupling was continued
for a further 5 h before washing with DMF, CH Cl , and Et O.
Gr a n u locyte/Ma cr op h a ge Colon y-F or m in g Cell (GM-
CF C) Assa y. Multiple dilutions of filtrates were added to
(
2
2
2
7
(
(
5 000 femoral bone marrow cells obtained from C57bl/6 mice
Bomholt Gaard, Denmark) suspended in McCoys medium
Gibco, Paisley, Scotland) enriched with nutrients (NaHCO
After the resin was dried in a stream of nitrogen, simultaneous
deprotection and cleavage of the crude semipeptide from the
resin was carried out in 5% aqueous TFA for 2 h. The solution
was evaporated in vacuo and the crude semipeptide precipi-
tated and washed with Et O. Aliquots of bulk material were
2
purified by preparative HPLC as described above.
P r oced u r e for th e Syn th esis of Red u ced Am id e Bon d
Sem ip ep tid e 23. 2,3-Dimethylpyrazine (1.08 g, 10 mmol),
N-chlorosuccinimide (3.3 g, 25 mmol), dry benzoyl peroxide
3
,
pyruvate, amino acids, and vitamins) and containing 100 U/mL
penicillin, 100 µg/mL streptomycin, and 15% FBS. To this was
added a predetermined suboptimal amount of L929 cell-
conditioned medium, known to produce between 20% and 30%
of maximum colony number, along with 100 µL of 3.3% agar
(
Difco Laboratories, MI) to make a final volume of 1 mL.
Cultures were incubated for 7 days in an incubator (Nuair) at
7 °C and 5% CO in air. Colonies were inspected using an
(
100 mg, 0.4 mmol), and CCl
until no more starting material remained (48 h) as shown by
thin layer chromatography (SiO ; 5:1 hexane/Et O). The
4
(100 mL) were refluxed together
3
2
inverted microscope (Zeiss, Germany), and those containing
more than 50 cells were scored.
2
2
reaction mixture was cooled, filtered through a pad of Celite,
evaporated to dryness, and then purified by silica gel column
Sta tistica l An a lysis. The synergistic activity was deter-
mined from the most dilute sample (i.e., smallest volume of
added filtrate in milliliters) giving rise to a significant increase
in colony numbers over background. By defining 1 unit of HSF
as the activity generating one colony above background, the
following formula was used to calculate the units of HSF/mL
of supernatant: HSF units/mL ) (colonies in test group-
colonies in background)/amount of filtrate added. The mean
value obtained from each experimental group was compared
to control using the two-tailed Student’s t-test.
chromatography using CH
of 2,3-bis(chloromethyl)pyrazine. An aliquot of this material
0.1 g, 0.57 mmol) was redissolved in CH Cl (10 mL) and
added dropwise over several minutes to a stirred solution of
2 2
Cl as eluent to afford 1.2 g (68%)
(
2
2
t
t
H-D-Ser(Bu )-OBu ‚HCl (0.58 g, 2.3 mmol) and NMM (0.35 mL,
.4 mmol). The mixture was stirred for a further 48 h and
then diluted to 30 mL with CH Cl . The organic layer was
extracted several times with 5% aqueous citric acid, H O, 5%
aqueous NaHCO , and H O, then dried over MgSO , filtered,
3
2
2
2
3
2
4
and evaporated in vacuo. The residue was treated for 30 min
with 5% aqueous TFA, evaporated in vacuo, redissolved in
water, and lyophilized. Further purification of an aliquot of
crude material by preparative HPLC was carried out as
described above yielding 20 mg (44%) of peptide 23.
Mu r in e Str om a l Cell Lin e (C6.4). A murine fibroblast-
like cell line was obtained from A. King (SmithKline Beecham
Pharmaceuticals, PA). The parental cell line, C6, was first
derived from the adherent cell layer of a long-term bone
marrow culture. Several biweekly passages of the adherent
cells and cloning by limited dilution yielded a cell line that
responded to peptide 1 by releasing a factor that synergized
with a colony-stimulating factor in a GM-CFC assay. Later
this cell line was subcloned by limiting dilution and a new
clone, C6.4, obtained which was used throughout this study
for the screening of peptide analogues.
Ack n ow led gm en t. We wish to thank all our col-
laborators at SmithKline Beecham Pharmaceuticals,
King of Prussia, PA, for the provision of KC antibody
and C6.4 cells and for helpful discussions which took
place during this study.
Refer en ces
(1) Rich, I. N. Hematopoietic-Initiating Cells. J . Perinat. Med. 1995,
2
3, 31-38.
(
2) Metcalf, D. The Colony Stimulating Factors: Discovery, Devel-
opment and Clinical Applications. Cancer 1990, 65, 2185-2195.
3) Paukovits, W. R. Isolation and Synthesis of a Hematoregulatory
Peptide. Z. Naturforsch. 1982, 37C, 1297-1300.
(4) Lærum, O. D.; Paukovits, W. R. Modulation of Murine Hemato-
poiesis In Vivo by a Synthetic Hematoregulatory Pentapeptide
(
P r ep a r a tion of Test Com p ou n d s for Biologica l Assa ys.
All test compounds were dissolved in water containing 0.1%
TFA at a concentration of 1 mg/mL and dispensed into Nalgene
cryovials before lyophilization. Each vial, containing 0.1 mg
of compound, was stored at -20 °C and diluted immediately
prior to an experiment in 1 mL of phosphate-buffered saline
(
HP-5b). Differentiation 1984, 27, 106-112.
(
5) Lærum, O. D.; Sletvold, O.; Bjerknes, R.; Eriksen, J . A.;
J ohansen, J . H.; Schanche, J . S.; Tveteras, T.; Paukovits, W. R.
The Dimer of Hemoregulatory Peptide (HP5B) Stimulates Mouse
and Human Myelopoiesis In Vitro. Exp. Hematol. 1988, 16,
274-280.
6) Pelus, L. M.; DeMarsh, P.; King, A.; Balcarek, J .; Frey, C.;
Bhatnagar, P. K.; Levin, R.; Scott, R. In Vitro and In Vivo
Hematopoietic Activity of a Novel Synthetic Hematoregulatory
Peptide. J . Cell. Biol. 1992, 16C, 87.
(PBS) containing 1% fetal bovine serum (FBS) (Hyclone
(
Laboratories, Utah), 100 U/mL penicillin, and 100 µg/µL
streptomycin (Bio Whittaker, Walkerville, MD) before addition
to in vitro cultures.
HSF In d u ction . C6.4 cells were grown to confluence in
(7) DeMarsh, P.; Wells, G. I.; Lewandowski, T. F.; Bhatnagar, P.
K.; Ostovic, E. J . Treatment of Experimental Gram-negative and
Gram-positive Bacterial Sepsis with the Hematoregulatory
Peptide SK&F 107647. J . Infect. Dis. 1996, 173, 203-211.
1
2-well tissue culture dishes (Costar) in RPMI 1640 medium
with 15% FBS, 100 U/mL penicillin, and 100 µg/µL strepto-
mycin before the addition of either PBS or test compound
solution. Analogues were tested at multiple doses in duplicate
wells; those doses yielding optimal results were selected for
further testing. Dimer peptides (1-10) were typically tested
at 1000 ng/mL, while cyclic and semipeptide analogues (11-
(
8) DeMarsh, P.; Sucoloski, S. K.; Frey, C. L.; Koltin, Y.; Actor, P.;
Bhatnagar, P. K.; Petteway, S. R. Efficacy of the Hematoregu-
latory Peptide SK&F 107647 in Experimental Systemic Candida
Albicans Infections in Normal and Immunosuppressed Mice.
Immunopharmacoloy 1994, 27, 199-206.
(
9) King, A.; Bhatnagar, P. K.; Balcarek, J .; Pelus, L. M. Modulation
of Bone Marrow Stromal Cell Production of Colony Stimulating
Activity by the Synthetic Peptide, SK&F 107647. Exp. Hematol.
2
4) were tested at 100 ng/mL in a total incubation volume of
1
mL. The cultures were incubated for 18 h at 37 °C with 5%
CO
2
in air (Nuair incubator). The supernatants were har-
vested and filtered through a Centricon 30 kD a cutoff
membrane (Amicon, Beverly, MA).
1
991, 19, 481.
(10) King, A.; Frey, C. L.; Arbo, B.; Scott, M.; J ohansen, K.; Bhat-
nagar, P. K.; Pelus, L. M. Hematoregulatory Peptide, SK&F

2
882 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 18
Cuthbertson et al.
1
07647, Induced Stromal Cell Production of KC[5-72] Enhances
CFU-GM Growth and Effector Cell Function. Blood 1995, 86
Suppl. 1), 309a.
(17) Peptide 13 was analyzed by coinjection with peptide 12 on
analytical HPLC (Vydac 218TP54, 0.46 × 25 cm column).
Following elution with a linear gradient of increasing 0.1% TFA/
MeCN (0-30% over 20 min) at a flow rate of 1 mL/min, two
peaks were observed corresponding to compound 13 at 11.3 min
and compound 12 at 11.7 min.
(
(
11) King, A.; Scott, R.; Wu, D. W.; Strickler, J .; McNulty, D.; Scott,
M.; J ohansen, K.; McDevitt, D.; Bhatnagar, P. K.; Balcarek, J .;
Pelus, L. M. Characterization and Purification of a Stromal Cell-
Derived Hematopoietic Synergistic Factor Induced by a Novel
Hematoregulatory Compound, SK&F 107647. Blood 1995, 86
(
18) Semipeptides 15 and 20 containing a mixture of D- and L-Ser
were analyzed by coinjection on HPLC with all the L isomers
(Suppl. 1), 310a.
1
4
and 19, respectively. Two closely eluting peaks were
observed. Further analysis by polarimetry in DMF of both 15
and 20 gave [R] ) 0, c ) 1, for both compounds as expected. A
comparison by polarimetry of the L,L and D,D isomers was also
(
12) Bhatnagar, P. K.; Agner, E. K.; Alberts, D.; Arbo, B. E.; Callahan,
J . F.; Cuthbertson, A. S.; Engelsen, S. J .; Fjerdingstad, H.;
Hartmann, M.; Heerding, D.; Hiebl, J .; Huffman, W. F.; Husbyn,
M.; King, A. G.; Kremminger, P.; Kwon, C.; LoCastro, S.;
Løvhaug, D.; Pelus, L. M.; Petteway, S.; Takata, J . S. Structure-
Activity Relationships of Novel Hematoregulatory Peptide. J .
Med. Chem. 1996, 39, 3814-3819.
D
made; compound 14, [R]
D
) +17, c ) 1; compound 16, [R]
D
)
-
17, c ) 1; compound 19, [R]
D
) +9.1, c ) 0.31; compound 21
had [R]
D
) -9.2, c ) 0.32.
(
19) The C6.4 cell line and anti-KC antibody were supplied by A.
King, SmithKline Beecham Pharmaceuticals, King of Prussia,
PA.
(
13) Fields, G. B.; Noble, R. L. Solid Phase Peptide Synthesis Utilizing
9
-Fluorenylmethoxycarbonyl Amino Acids. Int. J . Pept. Protein
Res. 1990, 35, 161-214.
(
20) Atherton, E.; Sheppard, R. C. In Solid Phase Peptide Synthesis.
A Practical Approach; Rickwood, D., Hames, B. D., Eds.; Oxford
University Press: Oxford, 1989; pp 88-89.
(
14) Nutt, R.; Strachan, R.; Veber, D.; Holly, F. J . Useful Intermedi-
ates for Synthesis of Dicarba Analogues of Cystine Peptides:
Selectively Protected R-aminosuberic acid and R,R-Diamino-
suberic Acid of Defined Stereochemistry. J . Org. Chem. 1980,
(21) Frank, R.; Doering, R. Simultaneous Multiple Peptide Synthesis
Under Continuous Flow Conditions on Cellulose Paper Discs as
Segmental Solid Supports. Tetrahedron 1988, 44, (19), 6031-
6040.
4
5, 3078-3080.
(
15) Coste, J .; Le-Nguyen, D.; Castro, B. PyBOP: a New Coupling
Reagent Devoid of Toxic Byproducts. Tetrahedron Lett. 1990,
3
1 (2), 205-208.
(22) Kaiser, E.; Colescott, R. L.; Bossinger, C. D.; Cook, P. I. Color
Test for Detection of Free Terminal Amino Groups in the Solid-
Phase Synthesis of Peptides. Anal. Biochem. 1970, 84, 595-598.
(
16) Carpino, L. A.; El-Faham, A.; Albericio, F. Advantageous Ap-
plications of Azabenzotriazole (triazolopyridine)-Based Coupling
Reagents to Solid-Phase Peptide Synthesis. J . Chem. Soc.,
Chem. Commun. 1994, 201-203.
J M9702443