Synthesis of peptide homo- and heterodimers as potential mimics of platelet-derived growth factor BB

Article abstract of DOI:10.1002/pep2.24150

Pericyte loss is correlated with blood-brain barrier leakage in neurological disorders such as Alzheimer's disease. The platelet-derived growth factor receptor β (PDGFRβ)/platelet-derived growth factor BB (PDGF-BB) signalling pathway is key to the regulation of pericyte survival and proliferation. A series of peptide dimers mimicking the ligand PDGF-BB were prepared in the hope of stimulating PDGFRβ internalisation and activation of this pathway. Copper-catalysed azide-alkyne cycloaddition of peptide monomers with PEGylated linkers of varying length afforded the desired peptide dimers. Evaluation of the peptide dimers in human brain pericyte assays revealed no effect on PDGFRβ internalisation nor cell proliferation at concentrations <10 μM. The peptide dimers also did not act as antagonists at PDGFRβ at concentrations <10 μM.

Full text of DOI:10.1002/pep2.24150

Received: 28 October 2019
DOI: 10.1002/pep2.24150
Revised: 12 February 2020
Accepted: 14 February 2020
F U L L P A P E R
Synthesis of peptide homo- and heterodimers as potential
mimics of platelet-derived growth factor BB
Louise A. Stubbing1 | Harveen Kaur1 | Sheryl X. Feng2,3 | Miranda Aalderink2,3
Michael Dragunow2,3 | Margaret A. Brimble1
|
1School of Chemical Sciences, The University
Abstract
of Auckland, Auckland, New Zealand
2Department of Pharmacology and Clinical
Pharmacology, The University of Auckland,
Private Bag 92019, Auckland, New Zealand
Pericyte loss is correlated with blood-brain barrier leakage in neurological disorders
such as Alzheimer's disease. The platelet-derived growth factor receptor
β
3Centre for Brain Research, The University of
Auckland, Auckland, New Zealand
(PDGFRβ)/platelet-derived growth factor BB (PDGF-BB) signalling pathway is key to
the regulation of pericyte survival and proliferation. A series of peptide dimers mim-
icking the ligand PDGF-BB were prepared in the hope of stimulating PDGFRβ
internalisation and activation of this pathway. Copper-catalysed azide-alkyne cyclo-
addition of peptide monomers with PEGylated linkers of varying length afforded the
desired peptide dimers. Evaluation of the peptide dimers in human brain pericyte
assays revealed no effect on PDGFRβ internalisation nor cell proliferation at concen-
trations <10 μM. The peptide dimers also did not act as antagonists at PDGFRβ at
concentrations <10 μM.
Correspondence
Margaret A. Brimble, School of Chemical
Sciences, The University of Auckland,
23 Symonds St, Auckland 1010, New Zealand.
Email: m.brimble@auckland.ac.nz
Funding information
Health Research Council of New Zealand;
Hugh Green Foundation; Lottery Health
Research Fellowship
K E Y W O R D S
CuAAC, PDGF-BB, PDGFRβ, peptide dimer, pericyte
1
|
INTRODUCTION
This pathway is activated by binding of the PDGF-BB homodimer
to two PDGFRβ receptors resulting in receptor dimerization, trans-
Pericytes are perivascular cells that line all brain capillaries.[1] Pericyte
loss occurs in Alzheimer's disease (AD) and loss correlates with blood-
brain barrier (BBB) leakage.[2–5] A key signalling pathway in pericytes
that regulates their survival and proliferation is the platelet-derived
growth factor receptor β (PDGFRβ) pathway.[6] The main ligand
for this receptor is platelet-derived growth factor-BB (PDGF-BB),
a homodimer of the 109 residue PDGF-B protein (Figure 1).[7,8]
phosphorylation and subsequent activation of downstream signal trans-
duction pathways that result in stimulation of cellular proliferation,
motility, and angiogenesis.[9] We aimed to activate this pathway in peri-
cytes and thereby restore impaired BBB function by targeting PDGFRβ
with peptide mimics based on the structure of PDGF-BB.
Targeting of the PDGFRβ/PDGF-BB pathway with synthetic
peptides based on the primary structure of PDGF-B has been
reported previously (Figure 2).[10–16] One approach concentrated on
residues from loop III with added cysteine residues, yielding pep-
tides 1 and 2 that were shown to induce proliferation in fibroblasts.
Interestingly, head-to-tail cyclisation of “P6” (2) afforded the cyclic
analogue “P7” (3), which did not induce proliferation.[11] Peptide
3 was instead found to have an inhibitory effect on PDGF-BB
induced mitogenesis in the same cells, and later, further studies rev-
ealed that peptide 3 induced apoptosis in exponentially growing
human fibroblasts.[12]
Abbreviations: 6-Cl-HOBt, 6-chloro-1-hydroxybenzotriazole; AA, amino acid; BSA, bovine
serum albumin; CuAAC, copper-catalysed azide-alkyne cycloaddition; DIPEA, N,N-
diisopropylethylamine; DMEM/F-12, Dulbecco's modified Eagle medium/nutrient mixture
F-12; DMSO, dimethyl sulfoxide; DNA, deoxyribonucleic acid; DODT, 3,6-dioxa-1,8-octane-
dithiol; EDCꢀHCl, N-(3-dimethylaminopropyl)-N0-ethylcarbodiimide hydrochloride; EDTA,
ethylenediaminetetraacetic acid; EdU, 5-ethynyl-20-deoxyuridine; ESI, electrospray
ionisation; FBS, fetal bovine serum; GnꢀHCl, guanidine hydrochloride; HATU, 1-[bis
(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxid hexafluorophosphate;
HFIP, 1,1,1,3,3,3-hexafluoro-2-propanol; HSA, human serum albumin; INFγ, interferon
gamma; NMM, 4-methylmorpholine; NMP, 1-methyl-2-pyrrolidinone; PBST, phosphate-
buffered saline with 0.2% Triton X-100; PFA, paraformaldehyde; PyAOP, (7-azabenzotriazol-
1-yloxy)tripyrrolidinophosphonium hexafluorophosphate; TFA, trifluoroacetic acid;
TIPS, triisopropylsilane.
Engström et al. used a different approach in their efforts towards
identifying antagonists of PDGFs.[14] At the time of the research, the
Pept Sci. 2020;e24150.
wileyonlinelibrary.com/peptidesci
© 2020 Wiley Periodicals, Inc.
1 of 15
https://doi.org/10.1002/pep2.24150

2 of 15
STUBBING ET AL.
X-ray crystal structure of the dimeric PDGFRβ/PDGF-BB complex
had not been reported. The authors instead synthesised a wide vari-
ety of peptides based on fragments of the PDGF-B primary structure
to identify epitopes involved in binding to PDGFRs. “Peptide 16” (4),
incorporating residues from what would later become known as the
loop I and loop III regions, was found to inhibit PDGFR dimerisation
and autophosphorylation.
Loop I residues have also been used by Beljaars et al. as the basis
for the design of a PDGFRβ-targeting drug carrier, pPB-HSA (5).[15,17]
Earlier site-directed mutagenesis studies of PDGF-B had identified
loop I residues R27 and I30 as being critical for receptor affinity and cell
activation,[18,19] so they, and their neighbouring residues, were incor-
porated into a cyclic octapeptide targeting domain (“pPB”). The
targeting domain was then derivatised at the N-terminus and coupled
to human serum albumin (HSA) to give pPB-HSA (5). pPB-HSA was
then shown to decrease binding of 125I-labelled PDGF-BB in fibro-
blasts at concencentrations of ≥1 nM, while either pPB or HSA alone
had no effect at concentrations up to 300 μM. pPB-HSA was also
found to reduce PDGF-BB-induced cell proliferation in fibroblasts.
pPB-HSA (5) has since been used as a targeting drug carrier for cancer
(doxorubicin to PDGFRβ-expressing tumour cells)[20] and liver fibrosis
(IFNγ to hepatic stellate cells).[21]
FIGURE 1 Dimeric PDGFRβ/PDGF-BB receptor complex
(PDGFRβ, yellow, purple; PDGF-B, green, blue)[8]
= inserted cysteine residue
= PDGF-B loop I residues
= PDGF-B loop III residues
= heparin-binding domain residues
= residue unrelated to PDGF-B loop region
Loop III analogues: Brennand et al.
73
81
CO2H
PDGF-B
residue
73
81
73
R
81
C
C
R
K
C
K
K
H2N
K
K
I
K
K
76
K
R
I
R
K
E
R
R
77
CO2H
V
V
I
E
I
I
V
I
E
H2N
3: "P7"
2: "P6"
1: "P4"
Combined loop I/loop III analogue: Engström et al.
PDGFRb-targeting peptide based on loop I residues: Beljaars et al.
"pPB"
82
P
S
S
CO2H
31
F
L
K
CO2H
C
35
N
40
K
V
C
HN
76
E
A
D
I
R
W
V
O
S
I
H2N
O
S
R
N
L
N
27
O
O
29
4: "Peptide 16"
human serum albumin
N
H
5: pPB-HSA
Targeted loop III analogue: Lin et al.
I
V
R
E
81
K
I
K
K
72
R
C
S
V
H2N
R
A
O
81
N
H
NH2
S
I
K
72
V
HN
K
R
C
H2N
R
R
E
L
K
V
I
I
E
K
O
N
H
R
K
H
N
R
6: PBA2-1c
O
FIGURE 2 Previously reported peptides based on the primary structure of PDGF-B targeting PDGFRβ

STUBBING ET AL.
3 of 15
More recently, Lin et al. reported a targeting peptide conjugate
In this work, we aimed to promote dimerisation of PDGFRβ and
thereby receptor internalisation and activation of the PDGFRβ path-
way by mimicking the dimeric nature of PDGF-BB. Peptide monomers
based upon the sequence of the loop I and III regions in PDGF-B are
prepared via Fmoc-SPPS (Figure 3, peptides 7a, 8a). An alkyne handle
for dimerisation via copper-catalysed azide-alkyne cycloaddition
(CuAAC) is installed at the N-terminus (peptides 7c, 8c), and the
corresponding N-terminal acetylated monomers (peptides 7b, 8b) are
also prepared to assess whether dimerisation through this site is detri-
mental to activity. A set of longer loop I analogues 9a-9c are also pre-
pared as the linear analogues of cyclic monomers 10a and 10b.
Peptide monomers 7c, 8c, 9c, and 10b are then covalently linked as
strategy in their work towards developing PDGF agonists.[16] PDGF-
BB has been shown to bind to heparin and heparan sulfate proteogly-
cans without affecting affinity for PDGFRs.[22] This was shown to
enhance the effect of the PDGF-BB at PDGFRα, possibly due to
location of the growth factor to the cell surface.[23] Thus, peptide con-
jugate PBA2-1c (6) was designed with a heparin-binding domain
linked via three aminohexanoic acid (Ahx) units to a dimeric version
of a loop III peptide reported in earlier studies by Brennand et al. (i.e.,
CVRKIEIVRKK).[11] PBA2-1c (6) was found to increase cell proliferation
and migration at 1 μg/mL (ca. 0.2 μM). Additionally, the conjugate was
found to have a similar affinity for PDGFRα and PDGFRβ.
38
85
I
PDGF-B
residue
25
F
P
F
K
K
L
N
70
CO2H
CO2H
K
A
Q
V
R
S
R
R
RHN
V
R
I
RHN
N
K
L
V
T
I
I
I
R
E
D
Loop I analogues
7a: R = H
7b: R = Ac
Loop III analogues
8a: R = H
8b: R = Ac
7c: R = 4-pentynoyl
8c: R = 4-pentynoyl
NHR
P
L
V
F
S
I
R
W
RHN
R
17
T
T
42
R
K
N
L
P
T
42
K
L
F
V
E
F
I
A
17
P
E
P
W
D
N
V
R
CO2H
T
R
V
F
N
T
E
A
S
R
R
E
9a: R = H
9b: R = Ac
9c: R = 4-pentynoyl
I
N
L
T
I
R
D
10a: R = H
10b: R = 4-pentynoyl
N3
O
O
N3
cyclic K17-P42
V70-K85 analogue 8c
I25-L38 analogue 7c
K17-P42 analogue 9c
n
analogue 10b
11a: n = 1
11b: n = 2
11c: n = 4
CuAAC
O
O
O
I25-L38 analogue
V
70-K85 analogue
K17-P42 analogue
N
O
N
N
N
N
N
N
N
N
O
O
N
N
N
N
N
N
N
N
N
O
O
O
I25-L38 analogue
n
V70-K85 analogue
n
K17-P42 analogue
O
O
O
n
12a: n = 1
12b: n = 2
12c: n = 4
13a: n = 1
13b: n = 2
13c: n = 4
14a: n = 1
14b: n = 2
14c: n = 4
O
O
I25-L38 analogue
cyclic K17-P42
analogue
N
N
N
N
N
N
O
O
N
N
N
N
N
O
O
N
V70-K85 analogue
O
n
O
n
cyclic K17-P42
analogue
15a: n = 1
15b: n = 2
15c: n = 4
16a: n = 1
16b: n = 2
16c: n = 4
FIGURE 3 Structure of proposed PDGF-B loop I- and loop III-based homo- and heterodimers

4 of 15
STUBBING ET AL.
homodimers 12, 13, 14, and 16 or heterodimers 15 via CuAAC with
hydrophilic PEGylated bis-azide linkers of varying length 11a-11c. The
ability of the PDGF mimics to stimulate proliferation in human peri-
cytes is then evaluated.
solvent (δ = 3.31) for CD3OD for 1H spectra and residual solvent
(CDCl3, δ = 77.0; CD3OD, δ = 49.0) for 13C spectra. 1H NMR values are
reported as chemical shift, multiplicity (s = singlet, d = doublet, t = triplet,
q = quartet, m = multiplet, dd = doublet of doublets), coupling constant,
integral.
2
|
MATERIALS AND METHODS
| General details
2.2
|
Peptide synthesis
2.1
C-Terminal residues were loaded onto 2-chlorotrityl chloride resin
(Chempep, Wellington, Florida; 0.77 mmol g−1) by adding a solution
of 1.2 eq. Fmoc-AA and 5 eq. DIPEA in CH2Cl2 (2 mL) to preswollen
resin (0.1 mmol) and agitated for 4 hours. The resin was washed with
CH2Cl2 (5 × 5 mL) and the coupling repeated with fresh reagents. The
resin was washed with CH2Cl2 (5 × 5 mL) and any unreacted sites
capped by treating with a solution of CH2Cl2/MeOH/DIPEA (80:15:5,
2 × 5 mL, 2 × 30 minutes). All peptides were elongated according to
Fmoc-SPPS conditions on a Tribute synthesiser (Protein Technologies,
Inc.). The Nα-Fmoc group was removed using 20% piperidine in DMF
or 5% piperazine + 0.1 M 6-Cl-HOBt/DMF (3 mL, 2 × 5 minutes)
followed by DMF washes (2 mL, 5 × 0.5 minute). Couplings were per-
formed using a mixture of the Fmoc-protected amino acid (5 eq.),
HATU (4.9 eq.) and NMM (10 eq.) in DMF (3.5 mL, 1 hour), followed
by DMF washes (2 mL, 5 × 0.5 minute). A capping cycle (20% Ac2O/
DMF, 2 × 3 mL, 15 minutes) followed by DMF washes (2 mL,
5 × 0.5 minute) was performed after each coupling cycle. The com-
pleted linear peptides were cleaved from the resin in their sidechain-
protected form using HFIP/CH2Cl2 (1:4 v/v, 2 × 5 mL, 30 minutes).
The resin was filtered, washed with CH2Cl2 (5 mL), and the volatiles
removed from the combined filtrates under a stream of N2. The residue
was diluted with MeCN/H2O (1:1 + 0.1% TFA, 15 mL) and lyophilised
to give the crude protected peptides. Completed linear peptides were
cleaved from the resin, and/or sidechain deprotection (i.e., Boc, Pbf,
tBu, Trt) was performed using TFA/TIPS/H2O/DODT (94:1:2.5:2.5 v/
v/v/v, 2 × 5 mL, 2 hours). The resin was washed with TFA (2 mL), and
the combined filtrates were concentrated under a stream of N2, then
triturated with cold Et2O (35 mL). The solids were isolated by centrifu-
gation (4000 rpm, 4 minutes) and the supernatant discarded. Additional
cold Et2O (35 mL) was added to the pellet, mixed, and the isolation
repeated. The pellet was dissolved in MeCN/H2O (1:1 + 0.1% TFA,
15 mL) and lyophilised to give the crude deprotected peptide.
All reagents were used as supplied. Solvents for RP-HPLC were
purchased as HPLC grade and used without further purification.
2-Chlorotrityl chloride resin was purchased from Chempep (Wellington,
Florida). Fmoc-protected amino acids and HATU were purchased from
GL Biochem (Shanghai, China). DIPEA, DMSO, Hoechst 33258, NMM,
NMP, and Triton X-100 were purchased from Sigma Aldrich (St Louis,
Missouri). DMF (AR grade) and MeCN (HPLC grade) were purchased
from Thermo Fisher (Hampshire, NH). CH2Cl2 (AR grade) was pur-
chased from ECP Limited (Auckland, New Zealand). BSA Fraction V
IgG Free (Fatty Acid Poor), DNAse I, DMEM/F-12 medium, FBS,
Hibernate-A medium, penicillin-streptomycin-glutamine, 0.25% trypsin-
EDTA, goat anti-mouse secondary antibodies, Click-iT EdU Cell Prolif-
eration Kit, and cell culture plates and flasks were purchased from
Thermo Fisher Scientific (Waltham, Massachusetts). PDGF-BB was
purchased from PeproTech (Rocky Hill, New Jersey). Papain was pur-
chased from Worthington Biochemical Corporation (Lakewood, New
Jersey). Mouse anti-PDGFRβ antibody (catalogue number 7460-3104)
was purchased from Bio-Rad Laboratories (Hercules, California). Falcon
cell strainers were purchased from In Vitro Technologies (Auckland,
New Zealand). Analytical RP-HPLC was performed on a Thermo Scien-
tific Dionex Ultimate 3000 UHPLC equipped with a four-channel UV
detector at 210, 225, 254, and 280 nm using a Waters XTerra C18
(125 Å, 150 mm × 4.6 mm, 5 μm) column at a flow rate of 1 mL min−1
and a gradient of 5% B to 65% B over 20 minutes was used, where sol-
vent A was 0.1% TFA in H2O and B was 0.1% TFA in acetonitrile.
Semi-preparative RP-HPLC was performed on a Waters 1525 binary
HPLC system equipped with a Waters 2489 UV/Vis detector at either
214 and 254 nm or 214 and 280 nm using either a Waters XTerra
OBD C18 (125 Å, 300 mm × 19.0 mm, 10 μm) column at a flow
rate of 10 mL min−1 or
a Phenomenex Gemini C18 (110 Å,
250 mm × 10.0 mm, 5 μm) column at a flow rate of 4 mL min−1. A suit-
ably adjusted gradient of 1% B to 80% B was used, where solvent
A was 0.1% TFA in H2O and B was 0.1% TFA in acetonitrile. Low-
resolution mass spectra were obtained using a Waters Quattro micro
API Mass Spectrometer in ESI positive mode. High-resolution mass
spectra were obtained on a Bruker micrOTOF-Q mass spectrometer
using electrospray ionisation in positive mode with a nominal accelerat-
ing voltage of 70 eV. Optical rotation was determined at the sodium D
line (589 nm) using an Autopol IV instrument; concentrations are given
in g/100 mL. NMR spectra were recorded at room temperature on a
Bruker AVANCE 400 spectrometer (1H, 400 MHz; 13C, 100 MHz). All
chemical shifts are reported in parts per million (ppm) and calibrated to
internal standard tetramethylsilane (δ = 0.0) for CDCl3 or residual
2.3
|
Solution-phase peptide cyclisation
The peptide (1 eq) was dissolved in sufficient DMF to give a 1 mM
solution, then cooled to 0 ꢁC in an icebath. PyAOP (3 eq) was then
added, followed by DIPEA (6 eq), and the resulting pale yellow solu-
tion stirred at 0 ꢁC for 4 hours. The mixture was diluted with ice-cold
water (×4) and extracted with CH2Cl2 (×3). The combined organic
extracts were washed with water (×3), brine (×2), dried over Na2SO4,
and concentrated under reduced pressure. The crude residue was
diluted with MeCN/H2O (1:1 + 0.1% TFA, 15 mL) and lyophilised.

STUBBING ET AL.
5 of 15
2.4
procedure
|
Synthesis of bis-azide linkers—general
1H NMR (400 MHz, CDCl3): δ 3.69-3.66 (m, 20H), 3.39 (t,
J = 5.2 Hz, 4H); 13C NMR (100 MHz, CDCl3): δ 70.07, 70.68, 70.63,
70.59, 70.0, 50.7. Spectra were consistent with previously reported
values.[27]
Bis-PEGn-azides 11a-11c were prepared according to the procedure
reported by Wang et al.[24]
Triethylamine (3 eq.) was added to
a
solution of the
poly(ethyleneglycol) (1 eq.) in CH2Cl2 (2 mL/mmol glycol) at 0 ꢁC,
followed by methanesulfonyl chloride (3 eq.), and the mixture was
allowed to stir at rt for 4 hours. Aq. HCl (1 M, 5 mL/mmol glycol) was
added and the organic layer was removed. The aqueous layer was further
extracted with CH2Cl2 (4 × 20 mL), and the combined organic extracts
were washed with saturated aq. NaHCO3 (100 mL), dried over Na2SO4,
and concentrated under reduced pressure to give the crude bis-mesylate
as a pale yellow oil. The crude bis-mesylate was thoroughly dried under
high vacuum to remove all traces of CH2Cl2, then diluted with DMF
(3 mL/mmol glycol) and then sodium azide (3.5 eq) was added. The mix-
ture was heated at 60 ꢁC for 22 hours and then allowed to cool to rt and
poured onto ice/water (1:1, 100 mL). The mixture was extracted with
ethyl acetate (4 × 50 mL) and the combined organic extracts were
washed with brine (4 × 100 mL), dried over Na2SO4, and concentrated
under reduced pressure to give the crude bis-azide as a yellow oil.
The crude material was purified via flash column chromatography (silica
gel, eluent as indicated) to give the pure bis-PEGn-azide 11a-11c.
2.5
|
Synthesis of modified lysine building block 30
It was prepared according to the procedure reported by Kanai et al.[28]
N-Hydroxysuccinimide (1.17 g, 10.2 mmol) was added to a solu-
tion of 4-pentynoic acid 29 (0.500 g, 5.10 mmol) and EDC-HCl
(1.95 g, 10.2 mmol) in CH2Cl2/THF (2:1, 51 mL) at 0 ꢁC, and the mix-
ture was allowed to stir at rt overnight. Saturated aqueous NaHCO3
(100 mL) was added, and the mixture was extracted with ethyl acetate
(3 × 70 mL). The combined organic extracts were washed with brine,
dried over MgSO4, and concentrated under reduced pressure to give
the crude as a yellow oil. The residue was diluted with 1,4-dioxane
(18 mL) and water (18 mL), and Fmoc-Lys-OH (3.00 g, 8.15 mmol) and
NaHCO3 (0.685 g, 8.15 mmol) were added. The cloudy yellow mixture
was allowed to stir at rt overnight. Aqueous HCl (1 M, 50 mL) was
added and the mixture was extracted with ethyl acetate (3 × 70 mL).
The combined organic extracts were washed with brine (100 mL),
dried over Na2SO4, and concentrated under reduced pressure to give
the crude as a yellow oil. The crude material was purified via flash col-
umn chromatography (silica gel, ethyl acetate/petroleum ether 1:1
+ 1% AcOH as eluent) to give a colourless oil; this material was then
dissolved in EtOAc (ca. 80 mL) and triturated with petroleum ether
(ca. 100 mL). The resulting white suspension was cooled to 0 ꢁC and
the solids were collected via filtration. The solids were then
lyophilised to afford building block 30 (1.25 g, 55%) as a colourless
2.4.1
|
Bis-azide linker 11a
It was prepared from triethylene glycol (2.00 g, 13.3 mmol) according
to the general procedure described earlier and purified via flash
column chromatography (silica gel, ethyl acetate/pet. ether 1:9 as elu-
ent) to give 11a (2.33 g, 88%) as a colourless oil.
amorphous solid.
1H NMR (400 MHz, CDCl3): δ 3.71-3.68 (m, 8H), 3.40 (t, J = 5.2 Hz,
4H); 13C NMR (100 MHz, CDCl3): δ 70.7, 70.1, 50.7. Spectra were
consistent with previously reported values.[25]
αD
;
1H NMR (CD3OD, 400 MHz): δ 7.79 (d, J = 7.5 Hz, 2H),
[22]
7.67 (t, J = 7.0 Hz, 2H), 7.38 (t, J = 7.4 Hz, 2H), 7.30 (td, J = 1.0,
7.5 Hz, 2H), 4.39-4.31 (m, 2H), 4.22 (t, J = 7.0 Hz, 1H), 4.13 (dd,
J = 4.7, 9.2 Hz, 1H), 3.19 (t, J = 6.6 Hz, 2H), 2.48-2.43 (m, 2H),
2.37-2.33 (m, 2H), 2.25 (t, J = 2.5 Hz, 1H), 1.90-1.81 (m, 1H),
1.75-1.65 (m, 1H), 1.59-1.37 (m, 4H); 13C NMR (CD3OD, 100 MHz): δ
176.1, 174.1, 158.9, 145.5, 145.3, 142.7, 128.3, 126.4, 121.1, 83.7,
70.5, 68.1, 55.4, 40.3, 36.2, 32.5, 30.0, 24.4, 15.9. Spectra were con-
sistent with previously reported values.[28]
2.4.2
|
Bis-azide linker 11b
It was prepared from tetraethylene glycol (1.20 g, 6.18 mmol)
according to the general procedure described earlier and purified via
flash column chromatography (silica gel, ethyl acetate/petroleum
ether 1:4 as eluent) to give 11b (1.38 g, 91%) as a pale yellow oil.
1H NMR (400 MHz, CDCl3):
(t, J = 5.2 Hz, 4H); 13C NMR (100 MHz, CDCl3): δ 70.7, 70.0, 50.7.
Spectra were consistent with previously reported values.[26]
δ
3.70-3.66 (m, 12H), 3.39
2.6
azide linkers
|
CuAAC dimerisation of peptides using bis-
CuAAC reactions were performed according to the previously
reported procedure by Brimble et al.[29]
2.4.3
|
Bis-azide linker 11c
GnꢀHCl (5.73 g, 60.0 mmol) and Na2HPO4 (284 mg, 2.00 mmol)
were dissolved in H2O (10 mL) and sparged with Ar for 30 minutes.
An aliquot (3 mL) of the buffer solution was then taken and
CuSO4ꢀ5H2O (67 mg, 0.27 mmol) and TCEP (69 mg, 0.24 mmol) were
added to the aliquot. The copper-buffer solution was vortexed to
dissolve the solids and the pH was adjusted between 7.0-7.4 using
It was prepared from hexa(ethylene glycol) (1.50 g, 5.31 mmol)
according to the general procedure described earlier and purified
via flash column chromatography (silica gel, ethyl acetate/petroleum
ether 1:1 as eluent) to give 11c (0.596 g, 34%) as a pale yellow oil.

6 of 15
STUBBING ET AL.
10 M NaOH/5 M HCl. The mixture was then heated with a heat gun
for 5 minutes and then centrifuged. The alkynyl peptide (1 eq) was
then dissolved in enough of the supernatant in a glass microwave
reaction vessel to give a 3 mM solution. A solution of the bis-PEGn-
azide 11a-11c (20 mg mL−1 in 6.0 M GnꢀHCl/0.2 M Na2HPO4,
0.55 eq) was then added, the vessel flushed with Ar, and sealed. The
mixture was subjected to microwave heating (25 W, 50 ꢁC) for
3 hours, taking aliquots (7 μL) at 1 hour intervals to monitor reaction
progress. Upon completion, the vessel was removed from the micro-
wave and allowed to cool to rt aqueous HCl (5 M, 0.5 mL) was added
and the mixture was diluted 10 times with H2O. The solution was
loaded onto a LP C18 SPE cartridge (AllTech) and eluted with 5%
MeCN/H2O (10 mL), followed by 80% MeCN/H2O (10 mL). The SPE
fractions were analysed by RP-HPLC and those containing the desired
peptide lyophilised to afford the crude peptide dimers 12a-14c and
16a-16b.
pellet was resuspended in fresh complete DMEM, and returned to the
flask for a further 24 hour incubation, at which point debris was
removed and discarded. Culture medium was replaced twice a week
until confluency, at which point cells were passaged by trypsinisation
with 0.25% trypsin-EDTA. Cultures were grown out to at least pas-
sage five before use in experiments to ensure that nonproliferating
cells (microglia and astrocytes) were diluted out. For experiments,
5000 cells per well were seeded into uncoated 96-well cell culture
plates and allowed to attach for at least 48 hours before initiation of
treatments.
2.8.2
|
EdU cell proliferation assay
Incorporation of 5-ethynyl-20-deoxyuridine (EdU) into DNA was used
to assess cell proliferation. Pericytes were treated with the indicated
peptide mimics, PDGF-BB (10 ng/mL), or vehicle (0.1% DMSO) for
48 hours before EdU (10 μM) was added. Cells were cultured for a
further 24 hours to allow incorporation of EdU and then fixed with
4% paraformaldehyde (PFA) pH 7.4 and washed with phosphate-
buffered saline (PBS) with 0.2% triton X-100 (PBST). Cells were then
washed with 3% BSA in PBS and EdU incorporation was visualised
using the Click-iT EdU Cell Proliferation Kit as per the manufacturer's
instructions. Cell nuclei were counterstained by incubation with
Hoechst 33258 for 20 minutes. Images were acquired using an
ImageXpress Micro XLS (Molecular Devices, California) high-
throughput fluorescent microscope. Quantitative analysis of cell
staining was performed using the Multi-Wavelength Cell Scoring
module on MetaXpress software version 6.2.3 (Molecular Devices,
California).
Heterodimers 15a-15c were prepared in a similar manner by
step-wise CuAAC reaction of peptide 7c (1 eq.) with bis-PEGn-azide
11a-11c (1 eq), followed by a second CuAAC reaction with peptide
7c (1 eq).
2.7
|
Conversion of peptide TFA salts to HCl
salts[30]
The peptide was dissolved in MeCN/H2O to give a 1 mgmL−1 peptide
solution. Aqueous HCl (0.1 M) was added to give a final acid concen-
tration of between 2 and 10 mM. The mixture was allowed to stand
at rt for 10 minutes and then lyophilised. The HCl treatment/
lyophilisation was repeated 3-4 times.
2.8
|
Biological evaluation
2.8.3
|
PDGFRβ internalisation assay
2.8.1
|
Mixed glial culture from brain tissue
Internalisation of PDGFRβ was examined using immunocytochemis-
try. To examine agonist-induced PDGFRβ internalisation, pericytes
were treated with the indicated peptide mimics, PDGF-BB (10 ng/mL),
or vehicle (0.1% DMSO) for 1 hour before being fixed with 4% PFA.
To examine antagonistic activity of peptide mimics, pericytes were
sequentially treated with mimic or vehicle for 2 hours and then with
PDGF-BB (10 ng/mL) for 1 hour before being fixed with 4% PFA.
Following fixation, cells were washed with PBST then incubated
with mouse anti-PDGFRβ antibody diluted 1:200 in immunobuffer
(1% goat serum, 0.2% triton X-100, 0.04% thimerosal in PBS) over-
night at 4 ꢁC. Cells were washed in PBST, then incubated with a
fluorescently conjugated goat anti-mouse secondary antibody
diluted 1:500 in immunobuffer for 2 hours at room temperature.
Cells were washed with PBST and then incubated with Hoechst
33258 for 20 minutes to stain cell nuclei. Images were acquired
using an ImageXpress Micro XLS (Molecular Devices, California)
high-throughput fluorescent microscope. Quantitative analysis of
cell staining was performed using MetaXpress software version
6.2.3 (Molecular Devices, California).
Samples of human middle temporal gyrus were obtained from patients
undergoing surgical treatment for intractable temporal lobe epilepsy,
with written patient consent and approval from the Northern Regional
Ethics Committee (New Zealand). Mixed glial cells were cultured
as described by Gibbons et al. with minor modifications.[31] Briefly, tis-
sue was diced to pieces <1 mm3 in size and then enzymatically disso-
ciated using 2.5 U/mL papain and 10 U/mL DNase I prepared in
Hibernate-A medium, with trituration after 15 and 30 minutes to
improve dissociation. Digestion was stopped by the addition of com-
plete DMEM (DMEM/F-12 medium supplemented with 10% FBS and
1% penicillin-streptomycin-glutamine), and then the cell suspension
was passed through a 100-μm nylon cell strainer. Cells were collected
by centrifugation at 160g for 10 minutes, then resuspended in com-
plete DMEM and transferred to an uncoated T75 cell culture flask.
Cells were cultured for 24 hours in an incubator at 37 ꢁC with 5%
CO2 and high humidity, before debris and culture media were
removed and collected by centrifugation at 160g for 10 minutes. The

STUBBING ET AL.
7 of 15
3
|
RESULTS AND DISCUSSION
The final third of the peptidyl resin 19 was coupled to 4-pentynoic acid
using HATU and DIPEA in DMF, followed by deprotection and resin
cleavage to afford CuAAC partner 7c. The peptides were then purified
via semi-prep HPLC to afford 7a-c in 32-39% yield.
K17-P42 linear analogues 9a-9c were prepared in an analogous
manner by automated Fmoc-SPPS from peptidyl resin 20 (Scheme 1).
All three peptides were then cleaved from the resin with TFA/TIPS/
H2O/DODT (94:1:2.5:2.5) and purified via semi-prep RP-HPLC to give
9a, 9b, and 9c in 11-14% yield.
Loop III analogues 8a-8c (i.e., PDGF-B residues V70-K85) were
similarly prepared from common precursor 23, itself prepared from
peptidyl resin 22 (Scheme 1). Peptides 8a-8c were obtained from pep-
tidyl resin 23 in the same manner as 7a-7c and 9a-9c (vide supra) to
afford 8a-8c in 29-39% yield after purification by RP-HPLC.
Synthesis of PDGF-BB mimics began with loop I analogues 7a-7c (i.e.,
PDGF-B residues I25-L38, Scheme 1). Fmoc-Leu-OH was loaded onto
2-chlorotrityl chloride resin (17) and the peptidyl chain elongated via
automated Fmoc-SPPS to give the linear sequence 19. HATU was used
as coupling agent and NMM as base in DMF. Nα-Fmoc group deblocking
was accomplished with 20% piperidine/DMF, and 20% Ac2O/DMF was
used as a capping agent after coupling of each amino acid residue.
The resin was then split into three portions and one third subjected to
resin cleavage and global deprotection with TFA/TIPS/H2O/DODT
(94:1:2.5:2.5, v/v/v/v) to give linear analogue 7a. Another third of
peptidyl resin 19 was acetylated, then deprotected and cleaved from
the resin under the same conditions to give N-acetyl linear analogue 7b.
NHBoc
a
Cl
a
O
FmocHN
Cl
O
O
FmocHN
17
O
17
18
22
Fmoc-SPPS
b, c, d
Fmoc-SPPS
b, c, d
Trt
A
Boc
Boc
F
tBu
Pbf
R
L
I
N
Pbf
R
F
P
Trt
Pbf
Boc
K
K
Trt
S
K
R
Boc
I
Q
N
V
R
L
RHN
V
T
K
I
R
RHN
V
Pbf
D
tBu
I
I
E
Pbf
tBu
tBu
19
R =H
23
R =H
e
d, e
f, e
e
d, e
f, e
7c
R = 4-pentynoyl
7a
R =H
7b
R = Ac
8c
8a
R =H
8b
R = Ac
R = 4-pentynoyl
a
Cl
O
= PDGF-B loop I residues
= PDGF-B loop III residues
N
Fmoc
O
17
20
= PDGF-B residues not part of
loops I and III
Fmoc-SPPS
b, c, d, b
e
9a
R =H
Cl
17
Cl
Cl
tBu
Pbf
R
Pbf
Boc
Trt
L
V
F
S
W
R
H2N
Boc
I
N
d, e
tBu
R
L
P
2-chlorotrityl chloride
polystyrene resin
9b
R = Ac
K
E
I
A
P
F
D
T
N
tBu
V
T
R
tBu
T
Trt
E
tBu
Pbf
Pbf
tBu
tBu
21
R =H
f, e
9c
R = 4-pentynoyl
SCHEME 1 Reagents and conditions: A, Fmoc-AA-OH, DIPEA, CH2Cl2, rt, 1 × 6 hours, 1 × 12 hours; B, 20% piperidine/DMF, rt,
2 × 5 minutes; C, Fmoc-AA-OH, HATU, 0.5 M NMM/DMF, rt, 1 hour; D, 20% Ac2O/DMF, rt, 2 × 15 minutes; E, TFA/TIPS/H2O/DODT
(94:1:2.5:2.5, v/v/v/v), rt, 2 hours; F, 4-pentynoic acid, HATU, DIPEA, DMF, rt, 1 hour; 7a 33%, 7b 39%, 7c 32%; 8a 37%, 8b 39%, 8c 29%; 9a
12%; 9b 14%; 9c 11% after purification by RP-HPLC

8 of 15
STUBBING ET AL.
3.1
|
Dimerisation of linear peptides
unreacted (see time-course RP-HPLC, Figure S15). There is only a
slight increase in the amount of 24b after 2 hours at rt. After leaving
the reaction overnight at rt, a third peak appears, corresponding to
the desired homodimer 12b (i.e., both desired CuAAC reactions have
occurred). However, the majority of 7c remains unreacted, with both
mono-coupled species 24b and desired homodimer 12b as minor
components of the reaction mixture (4% and 14%, respectively). No
further change to the ratio of products was observed after this time,
presumably due to oxidation of the copper(I) catalyst. Subsequently,
CuAAC reactions were performed with increased catalyst loading
(80 mM) and mild heating as described in an earlier publication[29]
(microwave, 25 W, 50 ꢁC) to accelerate the reaction (ca. 69% 12b and
24% 24b after 1 hour, see Supporting Information). Separation of any
unreacted alkynyl peptide 7c from the homodimers was difficult due
The CuAAC homodimerisation was first attempted with I25-L38 ana-
logue 7c and bis-azide linker 11b using conditions similar to those
previously used within our research group for CuAAC reactions with
peptides (Scheme 2).[29] Thus, catalytic Cu(II) (20 mM) was reduced to
Cu(I) with TCEP (20 mM) in degassed GnHCl-Na2HPO4 buffer at
pH 7.2. The alkynyl peptide 7c was then added, followed by the bis-
azide 11b, and the reaction was allowed to proceed at room tempera-
ture. Aliquots of the reaction mixture were analysed by RP-HPLC and
ESI-MS at regular intervals.
After 1 hour at rt, a small amount (<1%) of the monocoupled
species (i.e., 24b, Scheme 3) is present (i.e., one of two desired CuAAC
reactions has occurred), while the majority of the starting material is
N3
O
F
O
N3
L
N
n
CO2H
A
11a: n = 1
11b: n = 2
11c: n = 4
O
S
R
R
I
N
L
L
F
L
T
N
2 x
I
N
R
R
CO2H
N
D
S
A
F
S
R
O
L
a
N
R
N
I
N
CO2H
L
A
N
N
H
O
T
I
O
R
R
I
D
N
T
I
O
N
R
N
7c
D
n
12a: n = 1
12b: n = 2
12c: n = 4
2 x
O
L
V
F
S
R
W
R
HN
I
N
A
L
P
K
E
I
P
F
D
T
N
V
R
CO2H
T
R
T
E
9c
L
V
F
S
R
W
R
I
O
N
L
P
K
E
I
A
P
F
D
T
N
N
V
R
CO2H
T
R
N
11a - 11c
T
E
N
a
O
L
V
F
S
R
W
R
N
N
I
N
O
N
L
P
O
K
n
E
I
A
P
F
D
T
N
V
R
CO2H
T
R
T
E
13a: n = 1
13b: n = 2
13c: n = 4
I
F
P
K
K
2 x
CO2H
K
Q
V
O
R
V
R
K
N
V
H
I
I
I
F
P
E
K
K
CO2H
P
K
O
8c
Q
V
R
V
R
K
K
I
V
11a - 11c
N
I
N
E
I
F
N
K
K
a
CO2H
K
O
Q
V
O
R
V
R
N
V
O
I
N
I
N
E
n
14a: n = 1
14b: n = 2
14c: n = 4
SCHEME 2 Reagents and conditions: A, CuSO4ꢀ5H2O, TCEP, 6.0 M GnꢀHCl/0.2 M Na2HPO4, 25 W, 50 ꢁC, 3 hours; 12a 38%, 12b 40%, 12c
41%; 13a 28%, 13b 28%, 13c 26%; 14a 32%, 14b 35%, 14c 31% after purification by RP-HPLC

STUBBING ET AL.
9 of 15
to their similar retention time; therefore a slight excess of the bis-
azide linker (0.55 eq.) was employed in the CuAAC as separation of
the homodimer from the corresponding mono-coupled azido peptide
was more facile. Homodimers 12a-12c were thus obtained in 38-41%
yield after purification.
The longer K17-P42 analogue homodimers 13a-13c were prepared
in the same manner as alkynyl peptide 9c and bis-azide linkers 11a-
11c (Scheme 2). Similarly, loop III homodimers 14a-14c were also pre-
pared in a straightforward manner according to the same CuAAC pro-
tocol using alkynyl peptide 8c and bis-azide linkers 11a-11c
(Scheme 2).
SPE cartridge, followed by lyophilisation. The crude azido peptides
24a-24c were then subjected to a second CuAAC with alkynyl peptide
8c to afford the desired heterodimers 15a-15c. Homodimers 12a-12c,
14a-14c were also observed during the heterodimerisation reaction
(see Figures S25-S27); however, these were easily separable from the
desired heterodimers 15a-15c.
3.3
|
Cyclic peptide analogues
Peptide cyclisation is a strategy often used for the optimisation of
new peptide therapeutics.[32,33] It is thought that the loss of flexibility
“locks in” active conformations, enhancing interactions with target
(macro)molecules, while also increasing resistance to enzymatic degra-
dation, thereby increasing the half-life of the peptide.[32,33] Thus,
head-to-tail cyclisation (i.e., cyclisation along the peptide backbone
Figure 4) of K17-P42 peptide 25 was proposed to enhance potential
binding of the loop I peptide analogue to PDGFRβ. Selective removal
of the lysine Dde protecting group on resin enables installation of an
alkyne handle for CuAAC dimerisation with linkers 11a - 11c.
3.2
|
Loop I/loop III CuAAC-linked peptides
A series of analogues combining both loop I and loop III binding
regions were also prepared using linkers 11a-11c via stepwise CuAAC
reaction with alkynyl peptides 7c and 8c (Scheme 3).
The same CuAAC reaction conditions were employed as for the
previously synthesised homodimers; however, the stoichiometry was
adjusted in this case (1:1 M ratio 7c:11a-c) to ensure complete con-
sumption of alkynyl peptide 7c. The crude azido peptides 24a-24c
were isolated from the reaction mixture by passage through a C18
2-Chlorotrityl chloride polystyrene resin 17 was loaded with
Fmoc-Glu(tBu)-OH to give peptidyl resin 26 (Scheme 4). Elongation of
the peptidyl chain via Fmoc-SPPS was performed manually. Couplings
N3
O
O
N3
F
L
n
N
CO2H
11a: n = 1
11b: n = 2
11c: n = 4
A
O
S
R
R
I
N
L
T
7c
I
N
R
N
D
a
N
O
24a: n = 1
24b: n = 2
24c: n = 4
N3
O
n
F
L
N
CO2H
A
O
S
R
R
I
N
K
L
T
I
N
R
N
I
D
F
P
8c
N
K
K
b
CO2H
K
O
Q
V
O
R
V
R
V
N
O
I
N
I
N
E
n
15a: n = 1
15b: n = 2
15c: n = 4
SCHEME 3 Reagents and conditions: A, CuSO4ꢀ5H2O, TCEP, 6.0 M GnꢀHCl/0.2 M Na2HPO4, 25 W, 50 ꢁC, 3 hours; B, CuSO45H2O, TCEP,
6.0 M GnꢀHCl/0.2 M Na2HPO4, 25 W, 50 ꢁC, 2 hours; 15a 29%, 15b 33%, 15c 28%

10 of 15
STUBBING ET AL.
N
T
N
N
O
O
NHR
P
HN
n
O
T
R
T
K
R
K
T
P
L
V
L
V
E
N
N
N
P
W
E
P
W
F
F
CuAAC
V
V
N
N
F
F
A
A
S
R
S
R
R
E
R
E
I
HN
K
N
I
O
N
L
L
T
T
I
R
T
I
R
R
D
D
T
E
P
L
V
P
W
F
10a: R = H
10b: R = 4-pentynoyl
V
F
E
N
A
S
R
R
1. selective removal of lysine Dde protecting group
2. derivatisation of lysine residue (4-pentynoic acid)
3. N-terminal Fmoc deblocking
I
N
L
16a: n = 1
16b: n = 2
16c: n = 4
T
I
R
D
4. resin cleavage and cyclisation
Trt
A
F
tBu
Pbf
L
tBu
N
tBu
Pbf
R
V
Trt
Boc
P
tBu
R
F
V
S
R
E
E
W
I
N
L
T
FmocHN
T
P
I
R
D
tBu
K
T
Pbf
Pbf
tBu
tBu
Dde
25
FIGURE 4 Synthetic strategy toward cyclic peptides 10a and 10b and corresponding homodimers 16a–16c
were performed using HATU as coupling agent and DIPEA as base in
DMF for 1 hour. Fmoc deblocking was accomplished using 20% piper-
idine/DMF, and capping was performed at selected residues using
20% Ac2O/DMF. Upon completion of the full linear peptide sequence,
treatment of the peptidyl resin 25 with 20% HFIP/CH2Cl2 afforded
the sidechain-protected cyclisation precursor peptide 27 after
lyophilisation.
Dde group. An alternative method for the selective removal of Dde
groups in the presence of Fmoc on 2-chlorotrityl-bound dipeptides
using sodium borohydride was reported by Lokey et al. in 2013,
though there are no reports of its use on longer or more complex sub-
strates.[35] Treatment of the Dde-protected peptidyl resin 25 with
sodium borohydride in ethanol for 30 minutes unfortunately also
failed to affect the complete removal of the lysine Dde-protecting
group, resulting in a messier RP-HPLC profile. Repeated treatment
with sodium borohydride and/or longer reaction times again resulted
in concomitant undesired removal of the N-terminal Fmoc group.
Given our initial inability to selectively unmask the lysine sidechain
amine for further modification, our synthetic approach towards cyclic
CuAAC precursor 10b was revised.
Head-to-tail cyclisation of 27 was accomplished in solution using
PyAOP and DIPEA in DMF (1 mM peptide concentration) at 0 ꢁC for
4 hours. Removal of the Dde protecting group with hydrazine,
followed by global deprotection afforded the cyclic K17-P42 analogue
10a in 5% yield overall.
With the cyclic analogue 10a in hand, we next synthesised the
corresponding CuAAC dimers 16a-16c. This required attachment of
an alkyne handle to the lysine sidechain amino group to give cyclic
CuAAC partner 10b (Figure 4). Initially we planned to modify the
peptidyl resin 25 synthesised previously by selectively removing the
lysine Dde protecting group in the presence of the N-terminal
Fmoc group, followed by coupling of the unmasked amino group to
4-pentynoic acid to give peptidyl resin 28 (Scheme 5). Thus peptidyl
resin 25 was treated with hydroxylamine hydrochloride and imidazole
in NMP/CH2Cl2 (5:1) for 3 hours.[34] Mini-cleavage of the resin and
analysis by RP-HPLC/MS revealed that the deprotection was not
complete; however, repeated exposure of the peptidyl resin to
these deprotection conditions and/or longer reaction times resulted
in undesired removal of the N-terminal Fmoc group in addition to the
The requisite alkyne handle for CuAAC was instead installed using
a building block approach to access the requisite alkynyl peptide 28.
Alkynyl building block 30 was prepared in a two-step process previ-
ously reported by Shinoda et al. (Scheme 6).[28] 4-Pentynoic acid 29
was coupled with N-hydroxysuccinimide in the presence of carbo-
diimide reagent EDCꢀHCl to give the activated NHS ester; this was
then reacted with Fmoc-Lys-OH in basic solution to afford the desired
building block 30 in good yield over two steps.
The building block 30 was then incorporated into the automated
Fmoc-SPPS elongation of peptidyl resin 26 using HATU as coupling
agent, 0.5 M NMM as base, 5% piperazine +0.1 M 6-Cl-HOBt/DMF
as Fmoc deblocking agent, and 20% Ac2O/DMF as capping agent
(Scheme 7). Upon completion of the linear sequence 28, the N-

STUBBING ET AL.
11 of 15
Trt
A
F
tBu
Pbf
R
L
tBu
N
tBu
Pbf
R
CO2tBu
V
Trt
Boc
P
tBu
R
Fmoc-SPPS
b, c, d
F
V
S
E
E
a
W
I
N
L
Cl
T
O
O
FmocHN
T
P
FmocHN
I
R
D
tBu
K
17
T
Pbf
Pbf
tBu
tBu
26
Dde
25
Trt
N
Trt
A
T
R
K
T
F
P
tBu
Pbf
R
V
L
L
F
tBu
tBu
P
Pbf
E
W
V
Boc
P
tBu
R
F
V
f, g, h
S
E
b, e
V
R
E
N
A
W
I
N
T
CO2H
L
T
H2N
F
P
I
R
S
R
D
R
tBu
K
T
E
I
N
T
Pbf
Pbf
L
tBu
tBu
I
R
Dde
D
27
10a
SCHEME 4 Reagents and conditions: A, Fmoc-Glu(tBu)-OH, DIPEA, CH2Cl2, rt, 2 × 2.5 hours; B, 20% piperidine/DMF, rt, 1 × 5 minutes,
1 × 15 minutes; C, Fmoc-AA-OH, HATU, DIPEA, DMF, 1 hour; D, 20% Ac2O/DMF, rt, 2 × 15 minutes after T18, K17, P42, P41, V39, N34, I30,
I25; E, 20% HFIP/CH2Cl2, rt, 2 × 0.5 hour; F, PyAOP, DIPEA, DMF (1 mM), 0 ꢁC, 4 hours; G, 4% H2NNH2/DMF, rt, 15 minutes; H, TFA/TIPS/
H2O/DODT (94:1:2.5:2.5, v/v/v/v), rt, 2 × 2 hours, 5% yield overall after RP-HPLC purification
Trt
F
tBu
Pbf
R
L
tBu
N
tBu
Pbf
R
V
Trt
Boc
P
tBu
R
A
F
V
S
E
E
W
I
N
L
T
FmocHN
T
P
I
R
D
tBu
T
K
Pbf
Pbf
tBu
tBu
Dde
25
a or b,
then c
Trt
F
tBu
Pbf
R
L
tBu
N
tBu
Pbf
R
V
Trt
Boc
tBu
R
A
F
V
S
E
E
W
I
N
L
T
FmocHN
T
P
I
R
P
D
tBu
T
K
Pbf
Pbf
tBu
tBu
NH
O
28
SCHEME 5 Reagents and conditions: A, NH2OHꢀHCl, imidazole, NMP/CH2Cl2 (5:1), rt, 3 hours; B, NaBH4, EtOH, 0.5 hour; C, 4-pentynoic
acid, HATU, DIPEA, DMF

12 of 15
STUBBING ET AL.
terminal Fmoc group was removed and the protected peptide was
cleaved from the resin with 20% HFIP/CH2Cl2 to give cyclisation
precursor 31. Cyclisation of 31 in solution using the previously
established protocol (PyAOP, DIPEA, DMF, 0 ꢁC, 4 hours), followed
by global deprotection with TFA/TIPS/H2O/DODT (94:1:2.5:2.5,
v/v/v/v, rt, 2 hours) afforded cyclic peptide 10b. The crude peptide
was purified via semi-prep RP-HPLC to afford 10b in 9% overall
yield.
O
O
NH
OH
O
a, b
OH
29
FmocHN
30
The cyclic alkynyl peptide 10b was then subjected to homo-
dimerisation via CuAAC using bis-azide linkers 11a-c under the con-
ditions previously established for linear homodimers 16a-16c. The
CuAAC reactions proceeded smoothly, and after purification via
semi-prep RP-HPLC, 16a-16c were obtained in moderate yield and
good purity.
SCHEME 6 Reagents and conditions: A, N-hydroxysuccinimide,
EDCꢀHCl, THF/CH2Cl2, (1:2), rt, 18 hours; B, Fmoc-Lys-OH, NaHCO3,
H2O/1,4-dioxane (1:1), rt, 18 hours, 55% over two steps
Trt
F
tBu
Pbf
R
L
CO2tBu
tBu
N
tBu
Pbf
R
V
Fmoc-SPPS
b, c, d
Trt
Boc
a
tBu
R
A
F
V
S
E
E
W
Cl
O
O
I
N
FmocHN
L
T
FmocHN
T
P
I
17
R
P
D
tBu
K
T
Pbf
26
Pbf
tBu
tBu
NH
O
28
Trt
F
L
tBu
Pbf
R
tBu
N
tBu
Pbf
V
Trt
A
Boc
tBu
F
V
S
E
R
E
W
b, e
I
N
T
CO2H
L
T
H2N
P
R
I
R
P
D
tBu
K
T
Pbf
Pbf
tBu
tBu
NH
O
31
N
N
N
O
O
N3
O
HN
O
N3
HN
O
n
n
O
T
T
R
R
K
K
11a: n = 1
11b: n = 2
11c: n = 4
T
T
P
P
L
V
L
V
f, g
N
N
N
E
P
E
W
P
W
F
F
V
V
N
N
h
F
F
A
A
S
R
S
R
R
E
R
E
HN
K
I
N
I
O
N
L
L
T
T
I
T
I
R
R
R
D
D
T
P
L
V
E
P
W
F
10b
V
N
F
A
S
R
R
E
I
N
L
16a: n = 1
16b: n = 2
16c: n = 4
T
I
R
D
SCHEME 7 Reagents and conditions: A, Fmoc-Glu(tBu)-OH, DIPEA, CH2Cl2, rt, 1 × 5 hours, 1 × 18 hours; B, 5% piperazine + 0.1 M 6-Cl-
HOBt/DMF, rt, 2 × 5 minutes; C, Fmoc-AA-OH, HATU, 0.5 M NMM/DMF, rt, 1 hour; D, 20% Ac2O/DMF, rt, 2 × 15 minutes; E, 20% HFIP/CH2Cl2,
rt, 2 × 0.5 hour; F, PyAOP, DIPEA, DMF (1 mM peptide), 0 ꢁC, 1 hour; G, TFA/TIPS/H2O/DODT (94:1:2.5:2.5, v/v/v/v), rt, 2 hours, 9% after
purification by semi-prep RP-HPLC; H, CuSO45H2O, TCEP, 6.0 M GnꢀHCl/0.2 M Na2HPO4, 25 W, 50 ꢁC, 3 hours, 16a 45%; 16b 42%, 16c 42%

STUBBING ET AL.
13 of 15
FIGURE 5 EdU incorporation assay for homodimer 12a. Compounds were added to cells and then 48 hours later Edu was added; 24 hours
later, cells were fixed and EdU incorporation was visualized using the Click-iT EdU cell proliferation kit
FIGURE 6 Results of PDGFRβ internalization assay for homodimer 12a. Compounds were added to cells, and then 1 hour later, cells were
fixed and processed for PDGFRβ immunocytochemistry
3.4
pericytes
|
Evaluation of activity in human brain
The remaining peptide analogues 8a, 8b, 9a, 9b, 10a, and 13a-16c
were then tested for their ability to induce proliferation of human epi-
leptic brain-derived pericytes in the same way. Unfortunately, none of
the peptide mimics prepared in this study were able to induce cellular
proliferation in these cells at <10 μM. The use of assays to determine
antagonistic activity of the PDGF-BB peptide mimics (vs PDGF-BB)
also failed to show any activity.
All peptides were converted to biologically compatible HCl salts for
testing.[30,36] The peptides were dissolved in DMSO to make a 10 mM
stock solution and aliquots diluted as necessary for assay.
Peptide analogues 7a-7b and 12a-12c were initially tested for
induction of cell proliferation in human epileptic brain-derived peri-
cytes at concentrations of 0.1-10 μM. However, cell proliferation
(as measured by incorporation of thymidine analogue EdU over
48 hours)[37] did not appear to increase appreciably (Figure 5). In con-
trast, the positive control PDGF-BB reliably increased proliferation at
10 ng/mL (molecular weight ~25 kDa, ca. 400 nM).
4
|
CONCLUSIONS
A series of peptide mimics 7a-10b, 12a-16c based on loops I and III of
PDGF-BB have been synthesised utilising a copper-catalysed azide-
alkyne cycloaddition to assemble homo- and hetero-peptide dimers of
varying linker length. Loop I-based analogues 7a, 7b, 9a, and 9b failed to
show agonist or antagonist activity against PDGFRβ in human
brain pericyte assays. Dimerisation of the loop I-based analogues
An agonist-induced PDGFRβ internalisation assay was performed
with peptides 7a, 7b and 12a-12c. Unfortunately, it was clear from this
assay that these PDGF-BB mimics were not causing receptor inter-
nalisation and were therefore not acting as agonists at PDGFRβ (Figure 6).

14 of 15
STUBBING ET AL.
[5] A. Ben-Zvi, B. Lacoste, E. Kur, B. J. Andreone, Y. Mayshar, H. Yan,
C. Gu, Nature 2014, 509(7501), 507.
(i.e., homodimers 12a - 13c) did not improve the activity. Cyclisation of
the peptide sequence to reduce flexibility and enhance potential receptor
binding (i.e., analogues 10a, 16a-16c) also did not appear to have any
positive effect on the activity of the peptide mimic. Further cyclic ana-
logues (i.e., head-to-tail cyclised 7a or 8a and their corresponding dimers)
were therefore not prepared.
[6] J. Rustenhoven, D. Jansson, L. C. Smyth, M. Dragunow, Trends
Pharmacol. Sci. 2017, 38(3), 291.
[7] C.-H. Heldin, B. Westermark, Physiol. Rev. 1999, 79(4), 1283.
[8] A. H.-R. Shim, H. Liu, P. J. Focia, X. Chen, P. C. Lin, X. He, Proc. Natl.
Acad. Sci. 2010, 107(25), 11307.
[9] J. D. Kelly, B. A. Haldeman, F. J. Grant, M. J. Murray, R. A. Seifert,
D. F. Bowen-Pope, J. A. Cooper, A. Kazlauskas, J. Biol. Chem. 1991,
266(14), 8987.
Loop III analogues 8a and 8b likewise had no effect on
PDGFRβ or proliferation of human brain pericytes. Again,
dimerisation (i.e., homodimers 14a-14c) did not appear to improve
or impact activity of the peptide mimics. Loop I/loop III
heterodimers 15a-15c also failed to induce internalisation of
PDGFRβ or the proliferation of human brain pericytes at concen-
trations up to 10 μM.
[10] L. D. D'Andrea, A. Del Gatto, L. De Rosa, A. Romanelli, C. Pedone,
Curr. Pharm. Des. 2009, 15(21), 2414.
[11] D. M. Brennand, U. Dennehy, V. Ellis, M. F. Scully, P. Tripathi,
V. V. Kakkar, G. Patel, FEBS Lett. 1997, 413(1), 70.
[12] D. M. Brennand, M. F. Scully, V. V. Kakkar, G. Patel, FEBS Lett. 1997,
419(2), 166.
[13] G. Patel, W. Husman, A. M. Jehanli, J. J. Deadman, D. Green,
V. V. Kakkar, D. M. Brennand, J. Pept. Res. 1999, 53(1), 68.
[14] U. Engström, A. Engström, A. Ernlund, B. Westermark, C. H. Heldin,
J. Biol. Chem. 1992, 267(23), 16581.
The lack of activity observed for all analogues prepared in this
work may be due to lack of constraint of these peptides, necessitating
a high loss of conformational entropy for binding due to their flexibility
in solution. It is possible that modification of the N-terminus of peptides
7a, 8a, and 9a via CuAAC homodimerisation affects the activity of the
peptides, and it was for this reason the corresponding acetylated deriv-
atives 7b, 8b, and 9b were prepared. Similarly, it is possible that the
length of the PEGylated linkers between dimers could also affect activ-
ity of the peptide mimics - longer linker chains may allow more flexibil-
ity, or increase solubility. However, as all the peptides were inactive,
we cannot conclude whether N-terminal modification or the length of
the linkers had any effect on activity.
[15] L. Beljaars, B. Weert, A. Geerts, D. K. Meijer, K. Poelstra, Biochem.
Pharmacol. 2003, 66(7), 1307.
[16] X. Lin, K. Takahashi, Y. Liu, A. Derrien, P. O. Zamora, Growth Factors
2007, 25(2), 87.
[17] D. K. Meijer, L. Beljaars, G. Molema, K. Poelstra, J. Control Release
2001, 72(1–3), 157.
[18] J. M. Clements, L. J. Bawden, R. E. Bloxidge, G. Catlin, A. L. Cook,
S. Craig, A. H. Drummond, R. M. Edwards, A. Fallon, D. R. Green,
EMBO J. 1991, 10(13), 4113.
[19] A. Ostman, M. Andersson, U. Hellman, C.-H. Heldin, J. Biol. Chem.
1991, 266(16), 10073.
Despite this, we demonstrate here that synthetic SPPS combined
[20] J. Prakash, E. de Jong, E. Post, A. S. H. Gouw, L. Beljaars, K. Poelstra,
J. Control Release 2010, 145(2), 91.
with CuAAC was
a useful method to assemble homo- and
[21] R. Bansal, J. Prakash, M. D. Ruiter, K. Poelstra, PLoS One 2014, 9
(2), 10.
heterodimers of key peptide loop regions of PDGF-B.
[22] C. Rolny, D. Spillmann, U. Lindahl, L. Claesson-Welsh, J. Biol. Chem.
2002, 277(22), 19315.
ACKNOWLEDGMENTS
[23] A. Verrecchio, M. W. Germann, B. P. Schick, B. Kung, T. Twardowski,
J. D. S. Antonio, J. Biol. Chem. 2000, 275(11), 7701.
[24] C. Wang, D. Abegg, D. G. Hoch, A. Adibekian, Angew. Chem. Int. Ed.
2016, 55(8), 2911.
We are very grateful to the brain tissue donors and their families for
their gift of tissue for research, Dr Edward Mee, Dr Patrick Schweder,
and the neurosurgery team at Auckland Hospital. Financial support
was provided by Sir Thomas and Lady Duncan Trust; Lottery Health
Research Fellowship; Hugh Green Foundation; and Health Research
Council of New Zealand.
[25] J. Xiao, T. J. Tolbert, Org. Lett. 2009, 11(18), 4144.
[26] D. Lahav, B. Liu, R. J. B. H. N. van den Berg, A. M. C. H. van den
Nieuwendijk, T. Wennekes, A. T. Ghisaidoobe, I. Breen, M. J. Ferraz,
C.-L. Kuo, L. Wu, P. P. Geurink, H. Ovaa, G. A. van der Marel, M. van
der Stelt, R. G. Boot, G. J. Davies, A. JMFG, H. S. Overkleeft, J. Am.
Chem. Soc. 2017, 139(40), 14192.
CONFLICT OF INTEREST
[27] X. Li, S. J. H. Martin, Z. S. Chinoy, L. Liu, B. Rittgers, R. A. Dluhy, G.-
J. Boons, Chem. Eur. J. 2016, 22(32), 11180.
The authors declare no competing interests.
[28] K. Shinoda, Y. Sohma, M. Kanai, Bioorg. Med. Chem. Lett. 2015, 25
(15), 2976.
ORCID
[29] D. J. Lee, S.-H. Yang, G. M. Williams, M. A. Brimble, J. Org. Chem.
2012, 77(17), 7564.
Margaret A. Brimble
https://orcid.org/0000-0002-7086-4096
[30] V. V. Andrushchenko, H. J. Vogel, E. J. Prenner, J. Pept. Sci. 2007, 13
(1), 37.
REFERENCES
[1] M. D. Sweeney, S. Ayyadurai, B. V. Zlokovic, Nat. Neurosci. 2016, 19
(6), 771.
[31] H. M. Gibbons, S. M. Hughes, W. Van Roon-Mom,
J. M. Greenwood, P. J. Narayan, H. H. Teoh, P. M. Bergin,
E. W. Mee, P. C. Wood, R. L. M. Faull, M. Dragunow, J. Neurosci.
Methods 2007, 166(1), 89.
[2] A. P. Sagare, R. D. Bell, Z. Zhao, Q. Ma, E. A. Winkler, A. Ramanathan,
B. V. Zlokovic, Nat. Commun. 2013, 4, 2932.
[3] J. D. Sengillo, E. A. Winkler, C. T. Walker, J. S. Sullivan, M. Johnson,
B. V. Zlokovic, Brain Pathol. 2013, 23(3), 303.
[32] S.-H. Joo, Biomol. Ther. 2012, 20(1), 19.
ꢀ
[33] A. Roxin, G. Zheng, Future Med. Chem. 2012, 4(12), 1601.
[4] A. Armulik, G. Genové, M. Mäe, M. H. Nisancioglu, E. Wallgard,
C. Niaudet, L. He, J. Norlin, P. Lindblom, K. Strittmatter,
B. R. Johansson, C. Betsholtz, Nature 2010, 468(7323), 557.
[34] J. J. Díaz-Mochón, L. Bialy, M. Bradley, Org. Lett. 2004, 6(7), 1127.
[35] R. A. Turner, N. E. Hauksson, J. H. Gipe, R. S. Lokey, Org. Lett. 2013,
15(19), 5012.

STUBBING ET AL.
15 of 15
[36] J. Cornish, K. E. Callon, C. Q.-X. Lin, C. L. Xiao, T. B. Mulvey,
G. J. S. Cooper, I. R. Reid, Am. J. Physiol.Endocrinol. Metab. 1999, 277
(5), E779.
How to cite this article: Stubbing LA, Kaur H, Feng SX,
Aalderink M, Dragunow M, Brimble MA. Synthesis of peptide
homo- and heterodimers as potential mimics of platelet-
derived growth factor BB. Pept Sci. 2020;e24150. https://doi.
org/10.1002/pep2.24150
[37] A. Salic, T. J. Mitchison, Proc. Natl. Acad. Sci. 2008, 105(7), 2415.
SUPPORTING INFORMATION
Additional supporting information may be found online in the
Supporting Information section at the end of this article.