Synthesis, solution structure and biological activity of Val-Val-Pro-Gln, a bioactive elastin peptide

Article abstract of DOI:10.1002/ejoc.200400510

Val-Val-Pro-Gln (valyl-valyl-prolyl-glutamine) is a small but highly conserved sequence present in all elastins. We describe its synthesis by mixed anhydride solution chemistry as an alternative to solid-phase peptide synthesis (SPPS). The molecular struc

Full text of DOI:10.1002/ejoc.200400510

FULL PAPER
Synthesis, Solution Structure and Biological Activity of Val-Val-Pro-Gln,
a Bioactive Elastin Peptide[‡]
Caterina Spezzacatena,[a] Antonietta Pepe,[a] Lora M. Green,[b–d] Lawrence B. Sandberg,[c,d]
Brigida Bochicchio,[a] and Antonio M. Tamburro*[a]
Keywords: Peptides / Elastin / Biological activity
Val-Val-Pro-Gln (valyl-valyl-prolyl-glutamine) is a small but
highly conserved sequence present in all elastins. We de-
scribe its synthesis by mixed anhydride solution chemistry as
an alternative to solid-phase peptide synthesis (SPPS). The
molecular structure of the tetrapeptide in solution was inves-
tigated by classical spectroscopy, such as circular dichroism
(CD), nuclear magnetic resonance (NMR) and Fourier Trans-
form Infrared Spectroscopy (FTIR). The biological activity of
Val-Val-Pro-Gln was evaluated by a bromodeoxyuridine
(BrdU) incorporation assay with normal human dermal fibro-
blasts. This small peptide may play a critical role in control
of matrix metabolism through its release from the elastin
polypeptide chain during periods of tissue breakdown and
remodelling.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2005)
ine contents of any known protein (approaching 15 % in
some species).[1]
Introduction
Elastin is a ubiquitous protein in all vertebrates. Elastin
and type I collagen are among the most highly expressed
proteins in mammalian systems.[1] Elastin is found in vir-
tually all tissues, providing elasticity as well as contributing
to hormonal control over matrix expression through its pro-
teolysis and release of small peptides during remodelling.[2]
One such small peptide is valyl-valyl-prolyl-glutamine (Val-
Val-Pro-Gln), the subject of this report.
Elastin is produced from a single copy gene as a precur-
sor protein, tropoelastin (750–800 residues),[3] which is de-
posited extracellularly and rendered insoluble through the
formation of a unique tetrafunctional desmosine cross
link.[4] The protein elastin is particularly abundant in blood
vessels, lung tissue, skin and certain ligaments such as the
paraspinous ligaments and the ligamentum nuchae of rumi-
nants. It is a highly hydrophobic protein composed of over
90 % nonpolar amino acids and has one of the highest val-
Val-Val-Pro-Gln is a small but highly significant se-
quence of the elastin polypeptide chain. This sequence, lo-
cated in the exon 8 coded domain, a lysine–proline-rich
(KP) crosslinking domain, is highly conserved in all mam-
malian tropoelastins. It represents less than 1 % of the mass
of the precursor chain, yet it appears to have hormonal ef-
fects, which control one or more aspects of matrix expres-
sion. It is presumably released from the elastin polypeptide
chain through the action of matrix metalloproteinases
(MMPs),[5,6] which are active in tissue remodelling. In puri-
fied elastin preparations it can be released through the ac-
tion of the protease thermolysin, which cleaves on the N-
terminal sides of amino acids with large hydrophobic side
chains such as leucine, isoleucine and valine.[7]
Previous studies have identified important biological ac-
tivities exhibited by small elastin peptides derived from
thermolysin attack. A heptapeptide (LGAGGAG) and a re-
lated nonapeptide (LGAGGAGVL)[8] stimulate mRNA
and cytokine production in fibroblast tissue cultures,[9] and
other biological activities performed by these small peptides
have also been identified. For instance, in vitro studies uti-
lizing such thermolytic digests of elastin, containing Val-
Val-Pro-Gln and other small peptides of elastin, have been
found to increase the level of bromodeoxyuridine (BrdU)
incorporation in normal human dermal fibroblasts
(NHDF).[2] This increase in BrdU strongly suggests that sti-
mulation of NHDF by Val-Val-Pro-Gln has mitogenic ac-
tivity. It is probable that this response of cells to the elastin
thermolysin-produced peptide mixture has important impli-
[‡] Peptide VVPQ and certain compositions containing this pep-
tide are covered by US Patent numbers 6069129, 6777389,
6794362 and worldwide counterparts.
[a] Dipartimento di Chimica, Università della Basilicata,
Via N. Sauro 85, 85100 Potenza, Italy
Fax: +39-0971-202242
E-mail: tamburro@unibas.it
[b] Radiobiology Program – Department of Radiation Medicine,
Loma Linda University,
Shun Pavilion Room A-1010, 11175 Campus Street, Loma
Linda, CA 92354, U.S.A.
[c] Loma Linda University Medical School, Department of Rheu-
matology,
Loma Linda, CA 92354, U.S.A.
[d] Loma Linda University Graduate School, Loma Linda Univer-
sity,
11085 Campus Street, MT231, Loma Linda, CA 92354, U.S.A. cations with respect to elastin synthesis in vivo.
1644
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejoc.200400510
Eur. J. Org. Chem. 2005, 1644–1651

Val-Val-Pro-Gln, a Bioactive Elastin Peptide
FULL PAPER
Accordingly, it is important to check the possibility that SPPS becomes an extremely costly and time-consuming en-
Val-Val-Pro-Gln is indeed able to increase mitogenic ac- deavour, without any guarantee of significant purity.
tivity in normal human dermal fibroblasts. Therefore, we Furthermore, the necessity to produce gram amounts of
decided to synthesize Val-Val-Pro-Gln and to investigate its this peptide for in-depth biological testing and industrial
conformational properties.
applications prompted us to develop a solution synthesis
Furthermore, the molecular structure of the tetrapeptide procedure. For these reasons, mixed anhydride solution syn-
in solution has been investigated by classical spectroscopy, thesis, as described in this report, with use of Fmoc protec-
such as circular dichroism (CD), nuclear magnetic reso- tion of amine groups, was developed as an alternative to
nance (NMR), and Fourier transform infrared Spec- SPPS for the synthesis of Val-Val-Pro-Gln.
troscopy (FTIR). Finally the mitogenic activity of Val-Val-
Pro-Gln has been evaluated by bromodeoxyuridine (BrdU)
incorporation assay in normal human dermal fibroblasts
(NHDFs).
Structural Studies
The CD spectra of Val-Val-Pro-Gln in water and in TFE
at 0 °C and 25 °C are shown in Figure 2. In water the spec-
tra at both temperatures are typical of dominantly unor-
dered conformations. However, in the organic solvent there
is the presence of a negative band at 222–224 nm and a
positive band at 204–208 nm, which can be explained by
several different interpretations. For example, different
folded structures may be proposed, ranging from γ-
turns[11,12] to β-turns.[13] In any case, because the spectra are
also not significantly dependent on the temperature in this
case, one may suggest that the turns are rather stable. By
the way, it became necessary to implement conformational
Results and Discussion
Peptide Synthesis
The chemical structure of Val-Val-Pro-Gln is shown in
Figure 1.
1
analysis by means of 2D H NMR spectroscopy in TFE-
d3/H2O (80:20).
The resonance assignment of the peptide was straightfor-
wardly achieved by standard sequential assignment pro-
cedures[14] by use of combined analysis of 2D-TOCSY and
2D-NOESY spectra. TOCSY[15] spectra were used to ident-
ify spin systems, while NOESY[16] allowed the assignment
Figure 1. Chemical structure of Val-Val-Pro-Gln.
of resonances to individual amino acids through sequential
1
Solid-phase peptide synthesis (SPPS) of Val-Val-Pro-Gln NOE connectivities. From the monodimensional H NMR
by Fmoc chemistry was tried, but found to be difficult for spectrum the presence of three different species was evident,
several reasons. The penultimate proline is highly suscepti- pointing to cis- and trans-proline, because of an isomeriza-
ble to diketopiperazine formation, resulting in reduced effi- tion equilibrium, together with a third minor species as-
ciency and very low yields.[10] In addition, when we tried signed to the partial racemization of Val2 during the 2+2
to synthesize the tetrapeptide by SPPS, deprotection of the peptide coupling strategy. This species, according to NMR
glutamine made recovery difficult without extensive HPLC peak integration, represents 6 % of the total and may be
purification schemes. Thus, synthesis of Val-Val-Pro-Gln by neglected in the subsequent conformational analysis and
Figure 2. CD spectra of Val-Val-Pro-Gln recorded in water (a) and in TFE (b) at different temperatures: (᭿) 273 K, (᭡) 298 K.
Eur. J. Org. Chem. 2005, 1644–1651
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1645

A. M. Tamburro et al.
FULL PAPER
biological assays. The identification of the other two species plication of the Karplus equation.[18] The CYANA calcula-
was accomplished by NOE pattern analysis. The trans-pro- tion was performed with 50 randomized starting structures,
line conformation shows typical sequential dαδ(Pro-1, Pro) and the 10 resulting conformers with the lowest target func-
NOEs, while the cis conformation presents dαα(Pro-1, Pro) tion were chosen for conformational analysis by the MOL-
NOEs. Because the trans conformation is predominant MOL program.[19]
(85 %), the following conformational analysis was carried
Figure 4 shows the backbone superposition of ten con-
out only for the trans-proline peptide. The chemical shifts formers with the lowest target function. The results indi-
are shown in Table 1, while a summary of other important cated an S-shaped folded structure with an inverse γ-turn-
NMR parameters is shown in Figure 3.
like conformation in the V2–Q4 region. All the conformers
show dihedral angle values in the range expected for an
inverse gamma turn for the Pro residue (φ = –75° 10°; ψ
= 65° 10°). This conformation is stabilized not by a main-
chain-to-main-chain hydrogen bond, but more probably by
a main-chain-to-side-chain hydrogen bond involving the
side chain NHε proton of Q4 . The reduced number of
NOEs involving the Q4 side chain protons precludes us
from determining the preferred conformation for this side
chain, so the identification of the H-bond acceptor was not
possible.
Table 1. Assignments of proton resonances of the peptide Val-Val-
Pro-Gln in TFE-d3/H2O (80:20, v/v) at 298 K.
Chemical shifts of proton resonances (ppm)
Residue
NH
Hα
Hβ
Others
V1
V2
P3
–
7.92
–
3.85
4.59
4.44
4.45
2.23
2.16
2.27/2.02
2.25/2.03
γ 1.03
γ 1.00/1.05
γ 2.12 δ 3.86/3.73
γ 2.42 Nε 7.25/6.41
Q4
7.81
Figure 3. Summary of the NMR parameters of Val-Val-Pro-Gln
peptide recorded in TFE-d3/H2O (80:20) at 283 K. The summary
includes the temperature coefficients (–Δδ/ΔT, ppb K–1, with the
values for the Nε amide protons of the glutamine side chain indi-
cated in brackets), the coupling constants (3JHN–Hα, Hz), and se-
quential and medium range NOEs. The NOE intensities are re-
flected by the thickness of the line
The NOE analysis suggested the presence of a folded
conformation, as evidenced by nonsequential dαN(i,i+2) and
d
βN(i,i+2) NOE connectivities between V2 and Q4. Because
Figure 4. Structural model of Val-Val-Pro-Gln. Superposition of
the backbone atoms (N, Cα, CЈ) of ten conformers with the lowest
target function of peptide Val-Val-Pro-Gln. The backbone atoms
are represented by grey ribbons, while the side chain atoms are
represented by grey lines. (RMSD: 0.39 ( 0.22) Å for backbone
heavy atoms; 1.41 ( 0.76) Å for all heavy atoms)
of the dimensions of the peptide, only β-turns and γ-turns
can be regarded as folded conformations. Both types of sec-
ondary structures are usually, but not necessarily, stabilized
by a hydrogen bond involving the backbone atoms. Never-
theless, the temperature coefficients for V2 and Q4 ruled
out the presence of hydrogen bonds between the backbone
The FTIR absorption spectrum (Figure 5) appears to be
amide groups. Interestingly, one proton of the side chain in agreement with the NMR structural model, provided
amide protons of glutamine showed a reduced temperature that the γ-turn is assumed to be solvated by trifluoro-
coefficient (Δδ/Δt = –4.1 ppb K–1), suggesting the presence ethanol. As a matter of fact, in the amide I region the band
of a rather weak H-bond involving the Q4 side chain. How- at 1633 cm–1 is assigned to the Val–Pro tertiary amide
ever, while the donor of the hydrogen bond is easily deter- bond,[20] the lowering in frequency most probably being as-
mined by temperature coefficient analysis, the acceptor is sociated with hydrogen bonding to the solvent. Additional
harder to identify.
three bands are present at 1674, 1663 and 1616 cm–1. The
In order to produce a model of the conformation of Val- first is assigned to free C=O groups[21,22] and the second to
Val-Pro-Gln in solution, the NMR-derived constraints were the solvated amide carbonyls (mostly hydrogen-bonded to
used as input data for a simulated annealing structure cal- trifluoroethanol) γ-turn.[21] Finally, the third one could be
culation as implemented in a standard protocol in the CY- assigned to the absorption of water, as suggested by an
ANA program.[17] NOEs were translated into interproton anonymous referee, although absorption of water is mostly
distances and were used as the upper limit constraints, while found between 1640–1650 cm–1.[23] However, because of the
3JHN–Hα values were converted into angle constraints by ap- small intensity of the band, it could also originate from an
1646
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 1644–1651

Val-Val-Pro-Gln, a Bioactive Elastin Peptide
FULL PAPER
Figure 5. FT-IR absorption spectrum of Val-Val-Pro-Gln. Amide I region deconvoluted spectrum in TFE
artefact of decomposition, as suggested by another referee, blasts. It is easily seen that Val-Val-Pro-Gln increased the
and in any case is not significant for the assignment of the BrdU levels sixfold in relation to nonstimulated cells,
peptide conformation. Going on, these findings fit the whereas total HEP increased the BrdU levels threefold.
NMR and CD data well in that they are not incompatible
A more impressive image of BrdU incorporation in fibro-
with the presence of an open γ-turn, devoid of any intra- blasts is given in Figure 7. The stimulation by total HEP
molecular H-bond, although solvated by trifluoroethanol.
and, even more so, by Val-Val-Pro-Gln is quite clear. HEP
is a mixture of 42 discrete small elastin fragments, all of less
than 1000 Da in mass.[2] These have been complete charac-
terized as to composition (average size is five amino acids),
but not for biologic activity, except for Val-Val-Pro-Gln as
reported in this paper. By weight, Val-Val-Pro-Gln repre-
sents less than 1 % of the total peptides in HEP. Increased
BrdU incorporation in stimulated cells, in relation to un-
stimulated controls indicates an increase in DNA synthesis.
The increase in S-phase is coupled to an increase in cell
division. This increased incorporation may indicate that
Val-Val-Pro-Gln and HEP (partly through the action of its
Biological Activity Studies
Figure 6 shows the Val-Val-Pro-Gln and HEP-induced
cellular activity as determined by the bromodeoxyuridine
(BrdU) incorporation assay, tested in human dermal fibro-
Figure 6. BrdU incorporation in normal human dermal fibroblasts.
The bar graph represents the relative levels of BrdU incorporation
after a 3-hour stimulation with the total mixture of hydrophilic
elastin peptides (HEP) and Val-Val-Pro-Gln relative to the level of
BrdU in nonstimulated controls. Val-Val-Pro-Gln increased BrdU
levels by a factor of six in relation to nonstimulated levels, whereas
total HEP increased the BrdU levels threefold. Six replicate values
were averaged and the levels are shown plus/minus standard error.
The differences were found to be significant by Student t-tests, p
Ͻ 0.001.
Figure 7. Image of BrdU incorporation in normal human dermal
fibroblasts. The three-panel composite shows representative images
of normal human dermal fibroblasts photographed after the 3-hour
BrdU incubation period. The nuclei were stained red with propid-
ium iodide (PI) and the BrdU labelled green with Alexa-488. The
BrdU label and PI combined colour is yellow-orange in the nucleus,
the cytoplasmic green is autofluorescence from the constitutive pro-
duction of tropoelastin. Panel a) is nonstimulated cells, panel b) is
Val-Val-Pro-Gln, and panel c) is total HEP.
Eur. J. Org. Chem. 2005, 1644–1651
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1647

A. M. Tamburro et al.
FULL PAPER
troscopy; TOF-MALDI/MS: time-of-flight, matrix-assisted laser
desorption ionization mass spectrometry; Trt-: trityl; V (or Val):
valine; Z: benzyloxycarbonyl; SPPS: solid-phase peptide synthesis.
Val-Val-Pro-Gln content) are mitogenic to normal dermal
fibroblasts, causing them to increase their rate of prolifera-
tion. HEP has also been shown to increase the synthesis of
tropoelastin in dermal fibroblasts,[2] but this capacity has
not yet been tested with Val-Val-Pro-Gln. It may also be
that there are other mitogen-inducing peptides in the HEP
mixture that work with or without the action of Val-Val-
Pro-Gln, but because the Val-Val-Pro-Gln effect was signifi-
cantly more dramatic than that of HEP, this seems unlikely.
General Remarks: Amino acids were purchased from Novabi-
ochem AG (Laufelfinger, Switzerland). Purification of all synthetic
products was performed by reversed-phase HPLC on a 250
mm×4.6 mm micron Jupiter C-18 column with a binary gradient
of water (0.1 % TFA) and acetonitrile (0.1 % TFA) as eluent. Sub-
sequently the purity was ascertained by electrospray mass spec-
trometry. Further characterization was provided by 1H NMR spec-
tra recorded with a Bruker AMX 300 spectrometer. The synthesis
of the peptide was performed as indicated in Scheme 1. Briefly,
Fmoc and Z groups were used for N-terminal protection; these
were removed throughout by use of piperidine and catalytic hydro-
genolysis, respectively, while the coupling steps were accomplished
by a classical mixed anhydride procedure. Finally, the deprotection
steps were carried out.
Conclusions
The results reported in this study demonstrate that this
simple elastin-related tetrapeptide has significant biological/
mitogenic activity toward human dermal fibroblasts in vi-
tro. We believe this is diagnostic of a previously unappreci-
ated role for elastin in matrix-related tissues: the hormonal
control of matrix synthesis through release of peptides such
as Val-Val-Pro-Gln during matrix remodelling and growth.
Our hypothesis is that the open γ-turn conformation as-
certained for Val-Val-Pro-Gln in TFE is probably that
adopted by the peptide when it is interacting with cellular
receptors (the elastin binding protein?). Of course, this is
not to assume a single molecular structure for a short tetra-
peptide. Rather, although water strongly solvates the pep-
tide backbone, thereby favouring extended, unordered con-
formations, TFE nevertheless mimics a less polar environ-
ment reasonably well, therefore selecting some relatively
fixed structures, which are not discernible when exploring
its conformational space in water by experimental tech-
niques, because they are hidden within the conformational
ensemble.
Finally, we would like to emphasize that this peptide may
have useful applications in medicine (for instance, as a
wound-healing preparation) and cosmetics. To this end, the
solution synthesis reported here would be of value for large-
scale production of the tetrapeptide Val-Val-Pro-Gln or of
related peptides with the same C-terminal prolyl-glutamine
sequence.
Scheme 1.
Z-Val-Val-OMe: IBCF (13.0 mL, 99.5 mmol) was added at –13 °C
to
a solution of Z-Val-OH (25.0 g, 99.5 mmol) and NMM
(10.9 mL, 99.5 mmol) in chloroform (70.0 mL). The temperature
was maintained at –13 °C for 1 min, and Cl–H2+-Val-OMe (16.7 g,
99.5 mmol) and NMM (10.9 mL, 99.5 mmol) were then added. The
mixture was stirred at room temperature for 24 h. The organic sol-
vent was removed under reduced pressure and the oily residue was
dissolved in ethyl acetate and washed with 5 % sodium hydrogen-
carbonate, water, 5 % citric acid and water. The solution, dried with
anhydrous sodium sulfate, was concentrated to dryness and the oil
obtained was crystallized from ethyl acetate/petroleum ether to give
Experimental Section
List of Abbreviations: BrdU: bromodeoxyuridine; CCD: charge-
coupled device. CD: circular dichroism spectroscopy; DMF: N,N-
dimethylformamide; DMSO: dimethyl sulfoxide; DSS: 3-(tri-
methyl-silyl)propane-1-sulfonic acid; EDT: 1,2-ethanedithiol;
Fmoc: 9-fluorenylmethoxy-carbonyl. FTIC: fluorescein-5-isothi-
ocyanate; HEP: hydrophilic elastin peptides; HPLC: high-perform-
ance liquid chromatography; IBCF: isobutyl chloroformate; LSC:
laser scanning cytometer; NHDF: normal human dermal fibro-
blasts; NMM: N-methylmorpholine; NMR: nuclear magnetic reso-
nance; NOE: nuclear Overhauser enhancement; NOESY: NOE
spectroscopy; P (or Pro): proline; PBS: phosphate-buffered saline;
PBST: phosphate-buffered saline containing 0.05 % Tween-20; PI:
propidium iodide; Q (or Gln): glutamine; TFA: trifluoroacetic acid;
TFE: 2,2,2-trifluoroethanol; TOCSY: total correlation spec-
1
a white solid (30.0 g, 84 % yield). H NMR ([D6]DMSO): δ = 8.11
(d, 1 H; NH Val2), 7.40–7.23 (m, 6 H; H Ph and NH Val1), 5.01
(s, 2 H; Ph-CH2), 4.13 (m, 1 H; Hα Val), 3.96 (m, 1 H; Hα Val),
3.59 (s, 3 H; CH3), 2.11–1.83 (m, 2 H; Hβ Val), 0.92–0.80 ppm (m,
12 H; Hγ Val).
Z-Val-Val-OH: NaOH (1 n, 86.7 mL) was added with continuous
stirring to a solution of Z-Val-Val-OMe (30.0 g, 82.4 mmol) in ace-
tone (206.4 mL). After 3 h at room temperature, the organic solvent
was evaporated under reduced pressure and water was added. The
unchanged reagent was extracted with ethyl acetate, while the aque-
1648
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 1644–1651

Val-Val-Pro-Gln, a Bioactive Elastin Peptide
FULL PAPER
ous solution was cooled to 0 °C and neutralized by the dropwise
addition of 1 n HCl (86.7 mL). The crude material was extracted
with ethyl acetate and dried with anhydrous sodium sulfate. The
solution was then concentrated to dryness and the remaining resi-
due was crystallized from ethyl acetate/diethyl ether to give 26.7 g
maintained under nitrogen for an additional 15 min. The catalyst
was filtered and repeatedly washed with ethanol/80 % aqueous
AcOH (7:3). The solution obtained after filtration was evaporated
under vacuum, water was then added, and the solution was lyophi-
lized to remove any trace of unreacted products. The debenzyloxy-
carbonylated peptide was obtained in 60 % yield (113 mg). 1H
NMR ([D6]DMSO): δ = 8.62 (d, 1 H; NH Gln), 8.22 (d, 1 H; NH
1
of crystalline solid (93 % yield). H NMR ([D6]DMSO): δ = 12.54
(br. s, 1 H; OH), 8.01 (d, 1 H; NH Val2), 7.39–7.22 (m, 6 H; H Ph
and NH Val1), 5.00 (s, 2 H; Ph-CH2), 3.97 (m, 2 H; Hα Val), 2.10– Val), 7.43–7.12 (m, 16 H; H Ph and NHδ Gln), 4.37–4.26 (m, 2 H;
1.82 (m, 2 H; Hβ Val), 0.97–0.86 (m, 12 H; Hγ Val).
Hα Pro and Hα Gln), 4.12 (m, 1 H; Hα Val), 3.90 (m, 1 H; Hα Val)
3.71 (m, 2 H; Hδ Pro), 2.40–2.27 (m, 2 H; Hγ Gln), 2.10–1.66 (m,
8 H; Hβ Pro, Hγ Pro, Hβ Gln and Hβ Val), 0.83 ppm (m, 12 H; Hγ
Val).
Fmoc-Pro-Gln(Trt)-OH: IBCF (3.4 mL, 25.7 mmol) was added at
–5 °C to a solution of Fmoc-Pro-OH (8.7 g, 25.7 mmol) and NMM
(2.8 mL, 25.7 mmol) in chloroform (100.0 mL). The temperature
was maintained at –5 °C for 1 min, and Cl–H2+-Gln(Trt)-OH·H2O
(10.0 g, 25.7 mmol) and NMM (2.8 mL, 25.7 mmol) were then
added. The mixture was continuously stirred at room temperature
for 26 h. The organic solvent was removed under reduced pressure
and the oily residue was dissolved in ethyl acetate and washed with
5 % sodium hydrogencarbonate, water, 5 % citric acid and water.
The solution was dried with anhydrous sodium sulfate and concen-
trated to dryness, and the oil obtained was crystallized from ethyl
acetate/diethyl ether to give the crystalline dipeptide (13.3 g, 73 %
H2-Val-Val-Pro-Gln-OH: H2-Val-Val-Pro-Gln(Trt)-OH (96 mg,
0.14 mmol) was dissolved in water (0.58 mL). The resulting solu-
tion was cooled in an ice bath, and EDT (0.24 mL) and TFA
(9.7 mL) were added slowly.[24] The reaction mixture was stirred at
0 °C for few minutes and then at room temperature for 4 h. Follow-
ing this incubation period, approximately 1 mL of dichloromethane
was added and the solvent was partially evaporated under reduced
pressure. Cold diethyl ether was then added to give a white solid
1
(50 mg, 81 % yield). H NMR ([D6]DMSO): δ = 8.49 (d, 1 H; NH
1
Gln), 8.39 (d, 1 H; NH Val), 7.53 (s, 1 H; NHδ Gln), 6.82 (s, 1 H;
NHδ Gln), 4.38–4.27 (m, 2 H; Hα Pro and Hα Gln), 4.15 (m, 1 H;
Hα Val), 3.99 (m, 1 H; Hα Val) 3.69 (m, 2 H; Hδ Pro), 2.40–2.30
(m, 2 H; Hγ Gln), 2.22–1.97 (m, 8 H; Hβ Pro, Hγ Pro, Hβ Gln and
Hβ Val), 0.91 ppm (m, 12 H; Hγ Val). Purification by HPLC and
electrospray-mass spectrometry of the peptide are reported in Fig-
ure 8.
yield). H NMR ([D6]DMSO): δ = 12.58 (br. s, 1 H; OH), 8.50 (d,
1 H; NH Gln), 7.91–7.0 (m, 24 H; H Ph and NHδ Gln), 4.42–3.88
(m, 6 H; Hα Pro, Hα Gln, Hδ P and CH2), 2.29 (br. t, 2 H; Hγ Gln),
1.98–1.62 (m, 6 H; Hβ Pro, Hγ Pro and Hβ Gln), 1.07 ppm (t, 1 H;
CH).
H-Pro-Gln(Trt)-OH: Piperidine (102.6 mL, 1.0 mol) was added to
a stirred solution of Fmoc-Pro-Gln(Trt)-OH (12.8 g, 18.1 mmol) in
N,N-dimethylformamide (200.0 mL). The reaction was monitored
spectrophotometrically at λ = 301 nm for the absorbance of the
free Fmoc group. After 2 h and 30 min at room temperature, the
organic solvent was evaporated under reduced pressure and the oily
residue was crystallized from methanol/diethyl ether to give 3.9 g
1
of crystalline solid (45 % yield). H NMR ([D6]DMSO): δ = 12.57
(br. s, 1 H; OH), 8.49(d, 1 H; NH Gln), 7.88–6.7 (m, 16 H; H Ph
and NHδ Gln), 4.4–3.86 (m, 4 H; Hα Pro, Hα Gln, Hδ Pro), 2.3 (br.
t, 2 H; Hγ Q), 2.0–1.71 (m, 6 H; Hβ Pro, Hγ Pro and Hβ Gln).
Z-Val-Val-Pro-Gln(Trt)-OH: IBCF (0.8 mL, 6.5 mmol) was added
at –5 °C to a solution of Z-Val-Val-OH (2.3 g, 6.5 mmol) and
NMM (0.7 mL, 6.5 mmol) in N,N-dimethylformamide (150.0 mL).
The temperature was maintained at –5 °C for 1 min, and H-Pro-
Gln(Trt)-OH (3.1 g, 6.5 mmol) and NMM (0.7 mL, 6.5 mmol) were
then added. The mixture was stirred at room temperature for 24 h.
The organic solvent was removed under reduced pressure and the
oily residue was dissolved in ethyl acetate and washed with 5 %
sodium hydrogencarbonate, water, 5 % citric acid and water. The
solution was dried with anhydrous sodium sulfate and concentrated
to dryness, and the oil obtained was crystallized from ethyl acetate/
diethyl ether to give a white solid (3.8 g, 72 % yield). 1H NMR
([D6]DMSO): δ = 8.61 (d, 1 H; NH Gln), 8.12 (d, 1 H; NH Val),
7.80 (d, 1 H; NH Val), 7.40–7.11 (m, 21 H; H Ph and NHδ Gln),
5.01 (s, 2 H; CH2), 4.35–4.25 (m, 2 H; Hα Pro and Hα Gln), 4.10
(m, 1 H; Hα Val), 3.90 (m, 1 H; Hα Val) 3.70 (m, 1 H; Hδ Pro),
3.55 (m, 1 H; Hδ Pro), 2.39–2.25 (m, 2 H; Hγ Gln), 2.0–1.65 (m, 8
H; Hβ Pro, Hγ Pro, Hβ Gln and Hβ Val), 0.82 (m, 12 H; Hγ Val).
Figure 8. a) HPLC chromatogram of purified Val-Val-Pro-Gln.
b) Electrospray-mass spectrometry of Val-Val-Pro-Gln.
H2-Val-Val-Pro-Gln(Trt)-OH: Palladium/charcoal (Pd/C, 5 %, Circular Dichroism Spectroscopy: CD spectra of 0.1 mgml–1 solu-
49.5 mg) was suspended in 7.6 mL of ethanol/80 % aqueous AcOH
(7:3) and the system was vigorously stirred under hydrogen for
tions of peptides in water or trifluoroethanol were recorded on a
JASCO J600 CD spectropolarimeter in a cell with a 1 mm optical
path length with use of a HAAKE waterbath to control the tem-
perature. Usually, 16 scans were acquired in the 190–250 nm range
at temperatures of 273 K and 289 K, by taking points every 0.2 nm,
with a 20 nm min–1 scan rate, an integration time of 2 s, and a 1 nm
30 min. Then
a
suspension of Z-Val-Val-Pro-Gln(Trt)-OH
(210.0 mg, 0.27 mmol) in a minimal amount of the same solvent
mixture was added. The reaction mixture was monitored until the
evolution of CO2 had stopped (about 3 h), and the mixture was
Eur. J. Org. Chem. 2005, 1644–1651
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1649

A. M. Tamburro et al.
FULL PAPER
bandwidth. Buffer solution was recorded under the same condi-
tions and subtracted from the spectra before noise reduction.
Smooth noise reduction was applied. The data are expressed as
molar ellipticity [θ]M in deg cm2 dmol–1.
were harvested by fixation in 70 % ethanol at –20 °C.[27] The fixed
cells were incubated with murine monoclonal anti-BrdU (2
μg mL–1) overnight at 4 °C. Excess anti-BrdU was removed by
washing the coverslips in PBS containing 0.05 % Tween-20 (Sigma,
St. Louis, MO) (PBST). The cells were then incubated with Alexa-
488 conjugated goat anti-mouse secondary antibody (Molecular
Probes, Eugene, OR) for 4 hrs at 22 °C, washed with PBST and
counterstained for 10 min with propidium iodide (PI, 5 μg mL–1)
(Molecular Probes). The coverslips were protected with Permafluor
(Fisher Scientific, Tustin, CA), placed onto microscope slides and
dried flat in the dark, overnight.
Nuclear Magnetic Resonance Spectroscopy: All 1H NMR experi-
ments were performed on a Varian Unity INOVA 500 MHz spec-
trometer fitted with a 5 mm triple-resonance probe and z-axial gra-
dients. The purified peptide was dissolved in 700 μL of TFE-d3/
H2O (80:20), containing 0.1 mm of 3-(trimethylsilyl)propane-1-sul-
fonic acid (DSS) as internal reference standard at 0 ppm; 1.5 mm
1
peptide solutions were used. One-dimensional H spectra were ac-
quired with 32 K datapoints with a sweepwidth of 6000 Hz. The
residual HDO signal was suppressed by presaturation during the
relaxation period. TOCSY[14] spectra were collected using 256 t1
increments and a spectral width of 6000 Hz in both dimensions.
Relaxation delays were set to 2.5 s and spinlock (MLEV-17) mixing
time was 80 ms. Shifted sine bell squared weighting and zero filling
to 2 K×2 K were applied before Fourier transformation. NOESY
data were collected in essentially the same manner as for TOCSY
data, with mixing times ranging from 200 to 400 ms. The residual
HDO signal was suppressed by presaturation during relaxation
time for TOCSY experiments, while for NOESY experiments
Watergate pulse sequence were performed prior to acquisition.[25]
Amide proton temperature coefficients were measured from 1D 1H
NMR spectra recorded in 5 K increments from 278 K to 318 K.
Spectra were processed and analysed by VNMR Ver. 6.1C software
(Varian, Palo Alto, CA).
FITC-labelled BrdU incorporation was quantified in replicate sam-
ples by use of a laser scanning cytometer (LSC) (CompuCyte,
Cambridge MA)[27] The LSC is a microscope-based flow cytometer.
Green fluorescent intensity reflected a quantitative measure of
cellular activation as increased BrdU incorporation into S phase
DNA in treated cells relative to control cells. To confirm the gating
parameters, cells in selected gates were visually inspected through
the CCD camera attached to the microscope. Optimized protocol
and display settings were stored as configuration files and utilized
for all samples within a given experimental set. An average of 5
000 500 cells were scanned per treatment or condition.
Acknowledgments
This work was funded in part by grants from Connective Tissue
Imagineering, Visalia, California, from the European Community
(QLK6–2001–00332), and from the MIUR (COFIN 2002).
Structure Calculation: Experimentally measured NOE intensities
were converted into proton–proton distance constraints classified
into three ranges: 1.8–2.7 Å, 1.8–3.3 Å and 1.8–5.0 Å, correspond-
ing to strong, medium and weak NOEs, respectively. Pseudoatoms
were introduced when no stereospecific assignment was deter-
mined, and interproton distances were corrected accordingly.[26]
The two coupling constant of V2 and Q4, extracted from the 1D
1H NMR spectrum, were used as angle constraints. The structures
were calculated with the aid of the CYANA 1.0 program,[17] by
use of a standard simulated annealing procedure. From an initial
ensemble of 50 structures the best 10, in terms of target function
values and residual distance restraint violations, were chosen to
represent the conformations of the peptide. The resulting structures
were analysed with the MOLMOL graphics program,[19] which was
also used to produce the molecular model.
[1] L. B. Sandberg, Int. Rev. Connect. Tissue Res. 1976, 7, 159–
210.
[2] L. M. Green, P. J. Roos, L. B. Sandberg, Matrix Biol., in press.
[3] W. R. Gray, L. B. Sandberg, J. A. Foster, Nature 1973, 246,
461–466.
[4] J. Thomas, D. F. Elsden, S. M. Partridge, Nature 1963, 200,
651–652.
[5] W. Hornebeck, C. Lafuma, C. R. Séances Soc. Biol. Fil. 1991,
185, 127–134.
[6] Z. Werb, J. R. Chin, Ann. N. Y. Acad. Sci. 1998, 857, 127–134.
[7] L. S. Price, P. J. Roos, V. P. Shively, L. B. Sandberg, Matrix
1993, 13, 307–311.
[8] C. Spezzacatena, T. Perri, V. Guantieri, L. B. Sandberg, T. F.
Mitts, A. M. Tamburro. Eur. J. Org. Chem. 2002, 95–103.
[9] L. B. Sandberg, P. J. Roos, T. F. Mitts, USA patent n°6,069,129,
2000.
[10] N. Sewald, H. D. Jakubke, Peptides: Chemistry and Biology,
2002, Wiley-VCH, Weinheim.
FTIR Spectroscopy: Infrared spectra at a resolution of 2 cm–1 were
obtained in TFE solution with a Jasco FTIR spectrometer with a
0.025 cm cell with CaF2 windows. The sample concentration was
50 mg mL–1. The solvent spectrum obtained under identical condi-
tions was subtracted from the sample spectrum. The amide I region
of the spectra was decomposed into component bands by use of
the software Grams/32® (Version 4.11 Level II Galactic Industries
Corporation). The ratio between Lorentzian and Gaussian func-
tions was fixed to the 2:8 ratio.
[11] M. Lisowky, G. Pietrzynski, B. Rzeszotarska, 1993, 42, 466–
74.
[12] L. Biondi, F. Filira, M. Gobbo, R. Rocchi, J. Pept. Sci. 2002,
8, 80–92.
[13] A. Perczel, M. Hollósi, P. Sàndor, G. D. Fasman, Int. J. Peptide
Protein Res. 1993, 41, 223–236.
HEP-Induced Cellular Activity (Bromodeoxyuridine Incorporation
Assay): Normal human dermal fibroblasts were purchased from
Clonetics (BioWhittaker, Inc., Walkersville, MD) and maintained
according to the specifications provided. The cells were growth ex-
panded and samples were stored frozen in liquid nitrogen; cells
were not utilized in experiments past passage 5. NHDFs were es-
tablished in 6-well plates with 25 mm round glass coverslips and
treated with 3 mm BrdU added simultaneously with 10 μg mL–1 of
a total mixture of hydrolysed elastin peptides (HEP), 10 μg mL–1
Val-Val-Pro-Gln, PBS (negative control) or BrdU only (basal level
of incorporation). The treated cells were incubated for 3 hours and
[14] K. Wüthrich, NMR of Proteins and Nucleic Acids, 1986, Wiley,
New York.
[15] A. Bax, D. G. Davis, J. Magn. Reson., Ser. B 1985, 65, 355–
366.
[16] J. Jeener, B. H. Meier, P. Bachmann, R. R. Ernst, J. Chem.
Phys. 1979, 71, 4546–4553.
[17] P. Güntert, M. Mumenthaler, K. Wüthrich, J. Mol. Biol. 1997,
273, 283–298.
[18] M. Karplus, J. Chem. Phys. 1959, 30, 11–15.
[19] R. Koradi, M. Billeter, K. Wüthrich, J. Mol. Graph. 1996, 14,
51–55.
1650
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 1644–1651

Val-Val-Pro-Gln, a Bioactive Elastin Peptide
FULL PAPER
[20] Ch. Pulla Rao, R. Nagaraj, C. N. R. Rao, P. Balaram, Biochem-
istry 1980, 19, 425–431.
[21] E. Vass, S. Holly, Zs. Majer, J. Samu, I. Laczkο´, M. Hollο´si, J.
Mol. Struct. 1997, 408/409, 47–56.
[22] M. Hollosi, Z. Majer, A. Z. Ronai, A. Magyar, K. Medzih-
radszky, S. Holly, A. Perczel, G. D. Fasman, Biopolymers 1994,
34, 177–85.
[25] M. Piotto, V. Saudek, V. Sklenar, J. Biomol. NMR 1992, 2,
661–665.
[26] K. Wüthrich, M. Billeter, W. Braun, J. Mol. Biol. 1983, 169,
949–961.
[27] L. M. Green, D. K. Murray, A. M. Bant, G. Kazarians, M. F.
Moyers, G. A. Nelson, D. T. Tran, Radiat. Res. 2001, 155, 32–
42.
[23] S. Cai, B. R. Singh, Biochemistry 2004, 43, 2541–2549.
[24] P. Sieber, B. Riniker, Tetrahedron Lett. 1991, 32, 739–742.
Received: July 21, 2004
Eur. J. Org. Chem. 2005, 1644–1651
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1651