Synthesis and evaluation of fluorescent Pam3Cys peptide conjugates

Article abstract of DOI:10.1016/j.bmcl.2016.05.094

Chirally pure R- and S-epimers of TLR2 ligand Pam₃CysSK₄were prepared and separately conjugated to an OVA model epitope, in which lysine was replaced by azidonorleucine. The azide function in the conjugate permitted labelling with di

Full text of DOI:10.1016/j.bmcl.2016.05.094

Bioorganic & Medicinal Chemistry Letters 26 (2016) 3641–3645
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and evaluation of fluorescent Pam3Cys peptide conjugates
Geoffroy P. P. Gential a, Nataschja I. Ho b, Fabrizio Chiodo a, Nico Meeuwenoord a, Ferry Ossendorp b,
a,
Herman S. Overkleeft a, Gijs A. van der Marel a, , Dmitri V. Filippov
⇑
⇑
a Bio-organic Synthesis, Leiden Institute of Chemistry, Leiden University, The Netherlands
b Department of Immunohematology and Blood Transfusion, Leiden University Medical Centre, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Chirally pure R- and S-epimers of TLR2 ligand Pam3CysSK4 were prepared and separately conjugated to an
OVA model epitope, in which lysine was replaced by azidonorleucine. The azide function in the conjugate
permitted labelling with different fluorophores by use of strain-promoted 3+2 cycloaddition. The R-epi-
mer of the labelled conjugates induced TLR2-dependent DC maturation, while S-epimer proved to be
inactive. Combining the lipophilicity of Pam3CysSK4 ligand with fluorophores influenced the solubility
of the resulting conjugates in an unpredictable way and only the conjugates labelled with Cy-5 were suit-
able for confocal fluorescence microscopy experiments. It was shown that both epimers of the Cy-5
labelled lipopeptides were internalized equally well, indicating TLR2-independent cellular uptake. The
presented results demonstrate the usefulness of strain-promoted azide-alkyne cycloaddition in the label-
ling of highly lipophilic lipopeptides without disturbing the in vitro activity of these conjugates with
respect to activation of TLR-2.
Received 6 April 2016
Revised 30 May 2016
Accepted 31 May 2016
Available online 1 June 2016
Keywords:
Lipopeptides
Toll-like receptor 2
Fluorescent labeling
Strain promoted cycloaddition
Ó 2016 Elsevier Ltd. All rights reserved.
Conjugated cancer vaccines have attracted much attention as a
promising lead for innovative therapeutic interventions.1–5 A par-
ticular flavour of conjugated vaccines, that has been extensively
investigated through the years, comprises a structurally defined
construct of a Toll-like receptor agonist covalently attached to a
synthetic peptide, that contains a T-cell epitope, either model or
tumour associated.6 It has been discovered that a conjugate of this
kind show improved T-cell priming and tumour protection when
compared to a mixture of the individual antigenic peptide and
Toll-like receptor agonist.7,8 The usefulness of such synthetic pep-
tide based conjugates in tumour vaccination has been demon-
strated as well. A commonly used agonist in these studies is a
lipopeptide known as Pam3CysSK4 that binds to TLR2/TLR1.9–11
This compound has been derived from the N-terminus of bacterial
lipoprotein of, among others, Escherichia coli.12 Notably, Pam3-
CysSK4 when applied as a component of a vaccine candidate either
covalently attached to a longer peptide sequence or simply
admixed with a peptide,7,10,13–16 is often present as a mixture of
R- and S-epimers at the glycerol moiety, while it is known that
the R-epimer is the biologically active one.11,17
glycerol moiety of the Pam3Cys residue, as judged by the level of
the antigen presentation by DC’s.17 In this paper we show that flu-
orescently labelled and chirally pure Pam3Cys-lipopeptides repre-
sent useful tools in the studies of antigen processing because
these constructs allow a visual evaluation of the antigen uptake
irrespective of the DC-maturation status. Towards this end conju-
gates 1–4 (Fig. 1) with the fluorescent label covalently attached
to the modified side chain of a lysine residue in the commonly used
model MHC-I epitope (SIINFEKL) have been synthesized. This
design of the labelled construct proved to be successful in our past
studies that involved the monitoring of the intracellular trafficking
of Pam3Cys-lipopeptides as mixtures of epimers at C-2 of the glyc-
erol moiety.8 To be able to vary the type of fluorophore more read-
ily a convergent approach based on copper free click chemistry has
been chosen in the present work.18–20 The DC-maturation capacity
of the constructs has been evaluated and the uptake of these was
studied using confocal microscopy.
The key step of the convergent synthesis of conjugates 1–4 in
which the fluorescent labels are appended to the peptide with
the aid of strain promoted [3+2] azide alkyne cycloaddition
(Scheme 2) required the availability of azide containing lipopep-
tides (29, 30) and dyes functionalized with a strained alkyne (15,
16, Scheme 1). The lipopeptides 29 and 30 were accessible via
standard Fmoc-based solid phase synthesis using chirally pure
Fmoc-Pam2Cys-OH building blocks prepared as described in Sup-
plementary information. The click-reaction we decided to apply
With the aid of non-labelled Pam3CysSK4 conjugates it has been
shown that R-epimer of Pam3Cys is indeed the one responsible for
dendritic cell (DC) maturation and the S-epimer is inactive while
the cellular uptake remained unaffected by the chirality of the
⇑
Corresponding authors.
http://dx.doi.org/10.1016/j.bmcl.2016.05.094
0960-894X/Ó 2016 Elsevier Ltd. All rights reserved.

3642
G. P. P. Gential et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3641–3645
Figure 1. Target labelled Pam3Cys-lipopeptides.
Scheme 1. Synthesis of the reactive dyes 15 and 16 functionalized with a strained alkyne. Reagents and conditions: (i) N2CH2C(O)OEt, Cu(C5H7O2)2, EtOAc, 78%, (ii) LiAlH4,
THF/Et2O, 91%, (iii) Br2, DCM, (iv) KOtBu, THF, 35%, (v) p-NO2PhOC(O)Cl, DCM, 59%, (vi) 1,8-diamino-3,6-dioxaoctane, NEt3, DMF, 76% (vii) Cat. H2SO4, AcOH, reflux, 30% (viii)
SuOH, DIC, DMF (ix) DiPEA, DMF.
prevents the use of a copper catalyst and requires the availability
of the bifunctional (1R,8S,9s)-bicyclo[6.1.0]non-4-yn-9-yl methyl
(BCN) linker 9 to which fluorescent labels of choice can be attached
via amide bond formation. The synthesis of BCN linker 9 (Scheme 1)
is based on the coupling of known BCN 4-nitrophenyl carbonate (8)
with 2,20-(ethylenedioxy)bis(ethylamine). The reported proce-
dure21 to BCN alcohol (6) commences with cyclopropanation of
Simmons–Smith type reaction to provide exo-5 (28%) and endo-5
isomers (58%). Although this step is arguably efficient, the cheaper
copper acetoacetonate was evaluated as catalyst, in order to facil-
itate the scaling up of the synthesis. By using ethyl acetate instead
of DCM to bring the reaction to higher temperature endo-5 and
exo-5 could be obtained in 18% and in 58% yield, respectively. A
notable difference with the rhodium catalysed reaction is the
appearance of exo-5 as a major isomer (lowest running spot on
1,5-cyclooctadiene through
a rhodium tetraacetate mediated

G. P. P. Gential et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3641–3645
3643
H-L
i
Boc tBu
SK4DEVSGLEQLESIINFEKL
tBu tBu Trt R1
23: R1=N3
H
24: R1=NHBoc
tBu tBu tBu Trt Trt
ii, iii
tBu
O
O
R1
Boc tBu
tBu tBu Trt
25: R, R1=N3
*
N
H
SK4DEVSGLEQLESIINFEKL
C15H31
O
O
S
26: S, R1=N3
27: R, R1=NHBoc
28: S, R1=NHBoc
O
NH2
tBu tBu tBu Trt Trt
iv, v
tBu
C15H31
29: R, R1=N3
30: S, R1=N3
31: R, R1=NH2
32: S, R1=NH2
O
O
R1
*
C15H31
O
O
S
O
N
H
SK4DEVSGLEQLESINFEKL-OH
O
NH
C15H31
C15H31
O
Fl
O
N
H
O
NH
H
N
O
O
Fl
Dry DMSO
ON
1:1
O
HN
O
O
N
1: R, Fl=TAMRA
2: S, Fl=TAMRA
3: R, Fl= Cy5
N
O
O
N
*
C15H31
O
O
S
O
N
H
SK4DEVSGLEQLESINFEKL-OH
4: S, Fl= Cy5
O
NH
C15H31
C15H31
O
O
*
C15H31
O
S
OH
NHFmoc
C15H31
O
R: 21
S: 22
O
Scheme 2. Convergent synthesis of labelled Pam3Cys-lipopeptides. Reagents and conditions: (i) SPPS Fmoc automated synthesis, (ii) 21 or 22, HCTU, DiPEA, NMP, (iii) 20%
piperidine, NMP, (iv) PamCl, Pyridine/DCM (v) 95% TFA, 2.5% TIS, 2.5% H2O. The designators R and S refer to the configuration of the chiral centre marked with an asterisk.
TLC). The rest of the synthesis was performed without any major
changes except that the exo-isomer was used to proceed with
the synthesis. Ester 5 was reduced using a LiAlH4 in a mixture of
Et2O and THF to give BCN alcohol 6. Subsequent bromination of
the double bond in 6 using Br2 followed by double elimination of
bromide from the crude dibromide intermediate generated alkyne
7 in 35% yield. Treatment of 7 with p-nitrophenyl chloroformate,
followed by addition of 2,20-(ethylenedioxy)bis(ethylamine) to
the resulting carbonate gave target bifunctional BCN linker 9 in
11% overall yield based on 1,5-cyclooctadiene.
With the availability of BCN linker 9 the fluorescent labels
TAMRA and Cy5 can be connected to the amine in bifunctional lin-
ker 9. In order to allow optimisation of the click reaction sufficient
quantities of the relatively stable TAMRA dye should be available.
Hence, using a slightly modified procedure from the literature22
TAMRA was prepared and coupled to BCN linker 9 on mmol scale
(Scheme 1). Sulfuric acid mediated condensation of dimethy-
laminophenol 10 with trimellitic anhydride 11 in acetic acid
instead of butyric acid proceeded smoothly to give 12 as a mixture
of regioisomers. Crude 12 was precipitated from diethyl ether and
the obtained partially purified compound was converted into
hydroxysuccinimide ester 13. Subsequently BCN linker (9) was
added to the reaction mixture to give fluorescent reagent 15. After
HPLC-purification TAMRA reagent 15 could be obtained as a single
isomer and high purity in a low overall yield. The corresponding
Cy5 reagent 16 was prepared according to the same procedure
using the commercially available hydroxysuccinimide ester of
Cy5 14 and the crude product was immediately used in projected
copper free labelling step.
Having all building blocks at our disposal the R- and S-Pam3-
CysSK4 peptide conjugates 31–32 were assembled by standard
solid-phase peptide synthesis SPPS using Fmoc-chemistry
(Scheme 2). Commercially available suitably protected amino acids
were applied while Fmoc-azidonorleucine was prepared based on a
published procedure.23 Automated SPPS was performed until the
azide containing peptide 23 was reached. The optically pure R-
and S-Pam3CysSK4 moieties were appended manually to immobi-
lized peptide fragment 23 using modified cysteine building blocks
21 (R) and 22 (S), respectively and HCTU as a coupling agent. This
manual coupling saved building blocks as only 1.2 equiv 21 and 22
overnight instead of the standard 5 equiv for 1 h could be used.
Ensuing Fmoc deprotection with piperidine was followed by cou-
pling with 10 equiv of palmitoyl chloride. Finally, TFA mediated
removal of the side chain protecting groups and concomitant
cleavage from resin yielded the lipopeptides 29–30 in 12% and
11% overall yield, respectively. It is important to recognise that

3644
G. P. P. Gential et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3641–3645
Figure 2. Activation of dendritic cells. DCs were stimulated with titrated amounts of either R-Pam3Cys, S-Pam3Cys [in the labelled (1–4) or the unlabelled (31–32) form;
LPS (positive control; 1.25 g/ml) or peptide (SK4DEVSGLEQLESIINFEKL; negative control) for 48 h. Supernatants were harvested and analysed for IL-12 cytokine secretion by
ELISA. One representative from three independent experiments is shown.
lM],
l
lipopeptides 29–30 are poorly soluble in both aqueous and organic
solvents and pure DMSO is needed for further processing. The use
of DMSO brings along precautions as the oxidative power of DMSO
together with traces of aqueous acid may induce oxidation of the
thioethers24 in 29 and 30. We did not observe such an oxidation
when pure dry DMSO was use to dissolve the Pam3Cys-containing
conjugates. After purification by HPLC the azide containing Pam3-
CysSK4 peptide conjugates 29 (R) and 30 (S) were labelled with
TAMRA and Cy-5. The azide containing conjugate (29–30) was dis-
solved in dry DMSO and TAMRA reagent (15) was added in 1:1
ratio. After overnight stirring at room temperature, LCMS analysis
showed complete conversion of de starting peptides and the
untreated reaction mixture was used for purification by prepara-
tive RP HPLC, yielding the labelled lipopeptides 1–2 in 45% and
37% yield, respectively. Introduction of the Cy5-fluorophore with
crude reagent 16 using the same procedure, as described for the
100nM
5 0 0 0 0
4 0 0 0 0
3 0 0 0 0
2 0 0 0 0
1 0 0 0 0
0
50nM
25nM
32
P a m 3 C S K 4 31
3
4
c o n tro l
Figure 3. Ability of immunogenic lipopeptides in triggering human IL-8 production
via TLR-2. (a) HEK TLR-2 cells were incubated with compounds 31, 32, 3 and 4 (100-
25nM) or 100 ng/mL Pam3CysSK4 for 24 h. Error bars represent SD.
Figure 4. Uptake of Pam-conjugates by dendritic cells. DCs were incubated for 15 min with compounds 3 or 4 (1
lM). The uptake and localisation of the compounds were
analysed with confocal laser scanning microscopy with Leica system settings as described.25 The images are representative for multiple cells in at least 3 experiments.

G. P. P. Gential et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3641–3645
3645
TAMRA dye (15), gave after HPLC purification the labelled lipopep-
tides 3 (52% yield) and 4 (48% yield).
Immunological evaluation of labelled conjugates 1–4 started
with assessing murine DC-maturation upon exposure to the conju-
gates as well as relevant reference compounds.
Organisation for Scientific Research (NWO-ZonMw). We thank
Dr. Gerbrand J. van der Heden van Noort for his contribution in
the optimisation of the cylcopropanation reaction. F.C. acknowl-
edges financial support through an NWO-CW Veni grant number
722.014.008.
DCs were stimulated for 48 h with either the R-Pam3Cys or the
S-Pam3Cys and DC maturation was measured by IL-12 production
(Fig. 2). Cells treated with R-Pam3Cys containing lipopeptide (31)
produced significantly higher amounts of IL-12 compared to the
S-Pam3Cys based counterpart (32). Similar results were found with
the compounds labelled with Cy5 or TAMRA, showing intact
immunogenicity of the fluorophore-labelled conjugates 1 and 3.
To corroborate the TLR-2 dependent activation of DC’s by the
fluorescent conjugates the compounds were next assessed using
HEK-cells transfected with TLR2. The level of IL-8 produced in
the assay reflects the capacity of the conjugates to activate the
receptor. The results (Fig. 3) show the ability of compounds 31
and 3 to trigger human TLR-2. Compound 31 showed a behaviour
similar to the natural TLR-2 ligand Pam3CysSK4 while compound
3 showed a lower ability in triggering TLR-2 especially at lower
concentration (25 nM). Compounds 32 and 4 showed no ability
in triggering human TLR-2. To control the receptor specificity of
immunogenic lipopeptides for TLR-2, HEK cells expressing TLR-4
were stimulated with compounds 31, 32, 3 and 4 (Fig. 1S, Support-
ing information). None of the compounds were able to trigger
human TLR-4 showing not only the high specificity of the immuno-
genic lipopeptides for TLR-2 but also the absence of any inadver-
tent LPS contamination in the samples of the TLR-2 activating
conjugates of this study (3, 4, 31, 32).
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2016.05.
094.
References and notes
1. Chang, C. H.; Varamini, P.; Giddam, A. K.; Mansfeld, F. M.; D’Occhio, M. J.; Toth,
I. ChemMedChem 2015, 10, 901.
2. Zom, G. G. P.; Khan, S.; Filippov, D. V.; Ossendorp, F. TLR Ligand-Peptide
Conjugate Vaccines: Toward Clinical Application In Advances in Immunology;
Austen, K. F., Honjo, T., Melchers, F., Uhr, J. W., Unanue, E. R., Melief, C. J. M.,
Eds.; Synthetic Vaccines, 2012; Vol. 114, p 177.
3. Heit, A.; Schmitz, F.; O’Keeffe, M.; Staib, C.; Busch, D. H.; Wagner, H.; Huster, K.
M. J. Immunol. 2005, 174, 4373.
4. Maurer, T.; Heit, A.; Hochrein, H.; Ampenberger, F.; O’Keeffe, M.; Bauer, S.;
Lipford, G. B.; Vabulas, R. M.; Wagner, H. Eur. J. Immunol. 2002, 32, 2356.
5. Tighe, H.; Takabayashi, K.; Schwartz, D.; Van Nest, G.; Tuck, S.; Eiden, J. J.;
Kagey-Sobotka, A.; Creticos, P. S.; Lichtenstein, L. M.; Spiegelberg, H. L.; Raz, E. J.
Allergy Clin. Immunol. 2000, 106, 124.
6. Jackson, D. C.; Lau, Y. F.; Le, T.; Suhrbier, A.; Deliyannis, G.; Cheers, C.; Smith, C.;
Zeng, W. G.; Brown, L. E. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 15440.
7. Zom, G. G.; Khan, S.; Britten, C. M.; Sommandas, V.; Camps, M. G. M.; Loof, N.
M.; Budden, C. F.; Meeuwenoord, N. J.; Filippov, D. V.; van der Marel, G. A.;
Overkleeft, H. S.; Melief, C. J. M.; Ossendorp, F. Cancer Immunol. Res. 2014, 2,
756.
8. Khan, S.; Bijker, M. S.; Weterings, J. J.; Tanke, H. J.; Adema, G. J.; van Hall, T.;
Drijfhout, J. W.; Melief, C. J. M.; Overkleeft, H. S.; van der Marel, G. A.; Filippov,
D. V.; van der Burg, S. H.; Ossendorp, F. J. Biol. Chem. 2007, 282, 21145.
9. Jin, M. S.; Kim, S. E.; Heo, J. Y.; Lee, M. E.; Kim, H. M.; Paik, S. G.; Lee, H. Y.; Lee, J.
O. Cell 2007, 130, 1071.
The uptake of 3 and the 4 was measured with confocal micro-
scopy. After 15 min, both compounds were efficiently internalized
by murine DCs (shown in red and overlay with DC) and accumu-
lated in hot spots surrounding the nucleus (Fig. 4). Similar as we
have previously reported, no differences in localisation or uptake
were observed.8,17
10. Spohn, R.; Buwitt-Beckmann, U.; Brock, R.; Jung, G.; Ulmer, A. J.; Wiesmuller, K.
H. Vaccine 2004, 22, 2494.
11. Metzger, J.; Jung, G.; Bessler, W. G.; Hoffmann, P.; Strecker, M.; Lieberknecht,
A.; Schmidt, U. J. Med. Chem. 1991, 34, 1969.
Summarising, using strain-promoted [3+2] cycloaddition
a
12. Braun, V. Biochim. Biophys. Acta 1975, 415, 335.
small set of fluorescent Pam3Cys-based lipopeptides (1–4) has
been successfully synthesized and compared to known immuno-
genic compounds (LPS, 31, 32). The R- not S-epimer of Pam3Cys
in the prepared fluorescent lipopeptides triggered DCs maturation
in TLR-2-dependent manner and at approximately the same level
as their unlabelled analogues. However, the poor aqueous solubil-
ity of the conjugates containing TAMRA (1 and 2) precluded the use
of those for microscopy studies. This reminds that the attaining
sufficient solubility remains a major challenge in the synthesis of
Pam3Cys-based constructs labelled with fluorophores. Neverthe-
less, conjugate 3 (R-epimer) and conjugate 4 (S-epimer), both
labelled with Cy-5, could be successfully used for confocal micro-
scopy and were taken up by dendritic cells to the same extent. This
result corroborates previous findings that suggested a TLR-inde-
pendent uptake of the peptides conjugated to a TLR-ligand.
13. Moyle, P. M.; Dai, W.; Zhang, Y. K.; Batzloff, M. R.; Good, M. F.; Toth, I.
Bioconjugate Chem. 2014, 25, 965.
14. Shakya, N.; Sane, S. A.; Vishwakarma, P.; Gupta, S. Exp. Parasitol. 2012, 131, 377.
15. Caproni, E.; Tritto, E.; Cortese, M.; Muzzi, A.; Mosca, F.; Monaci, E.; Baudner, B.;
Seubert, A.; De Gregorio, E. J. Immunol. 2012, 188, 3088.
16. Welters, M. J. P.; Bijker, M. S.; van den Eeden, S. J. F.; Franken, K. L. M. C.; Melief,
C. J. M.; Offringa, R.; van der Burg, S. H. Vaccine 2007, 25, 1379.
17. Khan, S.; Weterings, J. J.; Britten, C. M.; de Jong, A. R.; Graafland, D.; Melief, C. J.
M.; van der Burg, S. H.; van der Marel, G.; Overkleeft, H. S.; Filippov, D. V.;
Ossendorp, F. Mol. Immunol. 2009, 46, 1084.
18. Zheng, M. M.; Zheng, L.; Zhang, P. Y.; Li, J. B.; Zhang, Y. Molecules 2015, 20, 3190.
19. Debets, M. F.; van Hest, J. C. M.; Rutjes, F. P. J. T. Org. Biomol. Chem. 2013, 11,
6439.
20. Sletten, E. M.; Bertozzi, C. R. Acc. Chem. Res. 2011, 44, 666.
21. Dommerholt, J.; Schmidt, S.; Temming, R.; Hendriks, L. J. A.; Rutjes, F. P. J. T.;
van Hest, J. C. M.; Lefeber, D. J.; Friedl, P.; van Delft, F. L. Angew. Chem., Int. Ed.
2010, 9422.
22. Uddin, M. J.; Marnett, L. J. Org. Lett. 2008, 10, 4799.
23. Sminia, T. J.; Pedersen, D. S. Synlett 2012, 2643.
24. Tam, J. P.; Wu, C. R.; Liu, W.; Zhang, J. W. J. Am. Chem. Soc. 1991, 113, 6657.
25. Streng-Ouwehand, I.; Ho, N. I.; Litjens, M.; Kalay, H.; Boks, M. A.; Cornelissen, L.
A.; Kaur Singh, S.; Saeland, E.; Garcia-Vallejo, J. J.; Ossendorp, F. A.; Unger, W.
W.; van Kooyk, Y. Elife 2006, 5.
Acknowledgements
This work is part of the research program TOP with project
number 91211011, which is financed in part by the Netherlands