Total synthesis of human hepcidin through regioselective disulfide-bond formation by using the safety-catch cysteine protecting group 4,4′-dimethylsulfinylbenzhydryl

Article abstract of DOI:10.1002/anie.201310103

A safety-catch cysteine protecting group, S-4,4′- dimethylsulfinylbenzhydryl (Msbh), was designed and developed to expand the capabilities of synthetic strategies for the regioselective formation of disulfide bonds in cysteine-rich peptides. The directed regioselective synthesis of human hepcidin, which contains four disulfide bonds, was undertaken and led to a high-resolution NMR structure under more physiologically relevant conditions than previously. Conversely, hepcidin synthesized with the formerly assigned vicinal disulfide-bond connectivity displayed significant conformational heterogeneity under similar conditions. The two synthetic forms of human hepcidin induced ferroportin internalization with apparent EC ₅₀ values of 2.0 (native fold, 1) and 4.4 nM (non-native fold, 2), with 2 undergoing isomerization to 1 in the presence of ferroportin expressing cells.

Full text of DOI:10.1002/anie.201310103

Angewandte
Chemie
DOI: 10.1002/anie.201310103
Controlled Peptide Folding
Hot Paper
Total Synthesis of Human Hepcidin through Regioselective Disulfide-
Bond Formation by using the Safety-Catch Cysteine Protecting Group
4,4’-Dimethylsulfinylbenzhydryl**
Zoltan Dekan, Mehdi Mobli, Michael W. Pennington, Eileen Fung, Elizabeta Nemeth, and
Paul F. Alewood*
Abstract: A safety-catch cysteine protecting group, S-4,4’-
dimethylsulfinylbenzhydryl (Msbh), was designed and devel-
oped to expand the capabilities of synthetic strategies for the
regioselective formation of disulfide bonds in cysteine-rich
peptides. The directed regioselective synthesis of human
hepcidin, which contains four disulfide bonds, was undertaken
and led to a high-resolution NMR structure under more
physiologically relevant conditions than previously. Con-
versely, hepcidin synthesized with the formerly assigned vicinal
disulfide-bond connectivity displayed significant conforma-
tional heterogeneity under similar conditions. The two syn-
thetic forms of human hepcidin induced ferroportin internal-
ization with apparent EC50 values of 2.0 (native fold, 1) and
4.4 nm (non-native fold, 2), with 2 undergoing isomerization to
1 in the presence of ferroportin expressing cells.
the increasing impact of genomics on the discovery of novel
disulfide-rich peptides, the assignment of their disulfide-bond
connectivity may be solely based on 2D NMR and conse-
quently is often plagued with ambiguity owing to the tight
packing of their cysteine frameworks.[3] In principle, unam-
biguous regioselective[4] chemical synthesis of each potential
regioisomer[5] that conforms to the NOE restraints obtained
by NMR would allow the connectivity to be confidently
imposed when the final structure is calculated.
The directed synthesis of up to three disulfide bonds in
a cysteine-rich peptide has been accomplished by using
various protecting-group schemes with either Boc or Fmoc
chemistry; most commonly with combinations of S-triphenyl-
methyl (Trt), S-acetamidomethyl (Acm), S-4-methylbenzyl
(Meb), S-4-methoxybenzyl (Mob), S-tertbutyl (tBu), or S-
tertbutylthio (StBu) groups (Boc = tert-butoxycarbonyl,
H
epcidin is a small peptide hormone that contains 25 amino
Fmoc = 9-fluorenylmethoxycarbonyl).[4a,6]
Whereas
the
acids and four disulfide bonds. It is involved in the regulation
of systemic iron homeostasis and in the pathogenesis of
several iron disorders[1] and acts by inducing the internal-
ization of the iron exporter ferroportin in vertebrates.[1]
Determination of the native connectivity of the four disulfide
bonds has proved to be a daunting task with early NMR
characterization suggesting a rare vicinal disulfide bond to be
present; this was recently revised and the native structure,
robustness of the Trt, Acm, and Meb or Mob protecting
groups is established, the removal of tBu groups often results
in the formation of side products and in low yields,[4a,7] and the
removal of StBu groups by reducing agents has been observed
to be sequence dependent.[6b,8] The only reported regioselec-
tive synthesis of a tetracystine peptide was also reliant on the
use of the tBu group.[9] In light of these shortfalls, we sought to
expand the scope of regioselective strategies through the
development of a novel safety-catch thiol protecting group
and to employ it strategically to synthesize the two reported
forms[2] of human hepcidin regioselectively.
which has
a different disulfide-bond connectivity, was
obtained, albeit under nonphysiological conditions.[2] With
Safety-catch protecting groups are designed to be stable
under a desired set of conditions until made labile through
specific derivatization.[10] Alkylsulfinylbenzyl-type safety-
catch protecting groups have been studied in detail for the
majority of natural amino acid functional groups and
employed successfully in peptide synthesis.[11] In their oxi-
dized forms, they contain electron-withdrawing sulfoxide
groups and show high acid stability to trifluoroacetic acid
(TFA) and hydrogen fluoride (HF) while their reduced forms,
which contain electron-donating sulfide groups, are readily
removed by TFA.[11a] The 4,4’-dimethylsulfinylbenzhydryl
(Msbh) structure was selected for cysteine protection based
on its anticipated stability during peptide synthesis and
applicability for a one-step reductive acidolytic cleavage
with concomitant disulfide-bond formation in TFA. We
employed Na-Fmoc-S-Msbh cysteine for the regioselective
synthesis of two forms of hepcidin, specifically the native form
according to Jordan et al.[2b] (1) and the vicinal form originally
proposed by Hunter et al[2a] (2).
[*] Z. Dekan, Prof. P. F. Alewood
Institute for Molecular Bioscience, The University of Queensland
St Lucia, Queensland, 4072 (Australia)
E-mail: p.alewood@imb.uq.edu.au
Dr. M. Mobli
Centre for Advanced Imaging, The University of Queensland
St Lucia, Queensland, 4072 (Australia)
Dr. M. W. Pennington
Peptides International, Inc
Louisville, KY 40299 (USA)
Dr. E. Fung, Dr. E. Nemeth
David Geffen School of Medicine, University of California
Los Angeles, CA 90095 (USA)
[**] This work was supported by Discovery Grant DP1096866 from the
Australian Research Council (P.F.A.) and Future Fellowship
FT110100925 (M.M.). We thank Alun Jones for obtaining high-
resolution mass spectra, and Dr. Bo Qiao for generating and sharing
with us the inducible K8R ferroportin-GFP cell line.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201310103.
Angew. Chem. Int. Ed. 2014, 53, 2931 –2934
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2931

.
Angewandte
Communications
Initially, we prepared both Na-Fmoc- and Na-Boc-Cys-
(Msbh)-OH (see the Supporting Information, p 5) and
evaluated their broad suitability for peptide synthesis with
both Boc- and Fmoc- chemistry using oxytocin
(CYIQNCPLG; see the Supporting Information, p 12),
which contains a single disulfide bond. For the reductive
acidolytic cleavage of the Msbh group, we chose to employ
the NH4I/dimethylsulfide (DMS)/TFA system because its
compatibility has been demonstrated with a range of cysteine-
containing peptides.[12] Iodine or iodosulfonium ions pro-
duced during this reaction oxidize the liberated thiols to
disulfides[12b] (see the Supporting Information, p 12), a con-
venient byproduct since our aim is the formation of disulfide
bridges. While broadly applicable, a limitation of these
deprotection conditions for use in tryptophan-containing
peptides is discussed in the Supporting Information (p 34).
Next, Cys(Msbh) was applied with paired Cys(Trt), Cys-
(Acm), and Cys(Meb) to the regioselective synthesis of the
disulfide bonds of human hepcidin (1, Scheme 1A). Chain
assembly was performed on a Thr(tBu)-Wang resin by using
Fmoc chemistry with HBTU/DIEA activation for couplings
(DIEA = diisopropylethylamine). Cleavage of the protected
peptide from the resin with TFA/TIPS/H2O yielded Cys 1,8
(Meb)/Cys 3,6(Acm)/Cys 5,7(Msbh) hepcidin with free thiols
at Cys 2 and 4; these were subsequently oxidized with
aqueous DMSO to give 3. Subsequent oxidative deprotection
of the Cys 3,6(Acm) pair was accomplished with I2 in 0.1%
TFA/MeCN/H2O to give 4. The Meb groups of Cys 1 and 8
were removed with HF/p-cresol and the liberated thiols
subsequently oxidized with I2 in 0.1% TFA/MeCN/H2O to
give 5. Alternatively, Cys(Meb) could be substituted with the
more acid labile Cys(Mob) to circumvent the need for the HF
cleavage step, in which case trifluoromethanesulfonic acid
(TFMSA)/TFA/p-cresol was used to remove the Mob group
(see the Supporting Information, p 19). The Msbh groups
remained intact throughout the course of these treatments.
The treatment of 5 with NH4I/DMS/TFA resulted in complete
removal of the Msbh groups of Cys 5 and 7 with simultaneous
disulfide-bond formation to give 1 as the single major product
without significant byproducts (Figure S2 in the Supporting
Information).
A vicinal disulfide form of hepcidin (2) was then
synthesized with the alternate connectivity of Cys 1–8, 2–7,
3–6, 4–5[2a] by using an analogous strategy as outlined in
Scheme 1B. Product 1 co-eluted by HPLC with a commercial
sample of human hepcidin (PI 4392s), which has the native
fold and was prepared by the reported procedure[13] (see the
Supporting Information, p 22), whereas product 2 eluted 14
seconds earlier (Figure 1).
Structural analysis of 1 by NMR spectroscopy revealed
that the cysteine residues in particular were substantially
broadened. This is a common observation in disulfide-rich
peptides because the disulfide bonds connect distal segments
of the peptide chain, which often undergo uncorrelated
motion.[14] This conformational plasticity around disulfide
bonds exacerbates an already difficult problem and as
anticipated, the disulfide-bond connectivity of 1 could not
be assigned from the 2D NMR experiments alone. However,
by utilizing the knowledge of the disulfide connectivity gained
from the regioselective chemical synthesis, a high-resolution
3D structure of 1 could be determined by using the
experimental NOE restraints and chemical-shift-based dihe-
dral-angle restraints.[15] The structure was found to be in good
agreement with the NMR and X-ray structures of native
Scheme 1. Regioselective synthesis of hepcidins 1[2b] (A) and 2[2a] (B) by using Trt, Acm, Meb, and Msbh protecting groups. a) 95% TFA/2.5%
TIPS/2.5% H2O, b) 30% DMSO/5% AcOH/H2O, c) I2, 0.1% TFA/50% MeCN/H2O, d) HF/p-cresol (9:1), e) NH4I (40 equiv), 1% DMS/TFA.
TIPS=triisopropylsilane, DMSO=dimethyl sulfoxide.
2932
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 2931 –2934

Angewandte
Chemie
spectra of the two compounds revealed little overlap, thus
indicating that none of the conformations of 2 are consistent
with the fold of 1 (Figure S4). Since a single major product
was obtained after each disulfide-forming step and 2 appears
as a single symmetrical peak by analytical HPLC, it is unlikely
that the multiple conformations seen by NMR are due to the
presence of undesired disulfide-bond isomers. Nevertheless,
to discount this possibility, a tryptic digest was performed.
Each of the products appeared as a single peak of correct
mass by analytical HPLC (see the Supporting Information,
p 27). Additionally, fractionation of 2 gave peaks of correct
mass that co-eluted on re-injection (see the Supporting
Information, p 27).
The hepcidin–ferroportin interaction is thought to involve
a thiol–disulfide interchange between ferroportin thiol C326
and the hepcidin disulfide framework.[16] We assessed the
bioactivity of the two regioselectively synthesized forms of
human hepcidin through the quantitation of ferroportin–GFP
degradation.[17] Forms 1 and 2, as well as native human
hepcidin (PI 4392s), stimulated the internalization and
degradation of ferroportin–GFP with comparable EC50
values of 2.0, 4.4, and 3.9 nm respectively (see the Supporting
Information, p 30). It was previously demonstrated that the
nine-residue N-terminal fragment of hepcidin retains some
bioactivity provided that there is at least one cysteine residue
present.[16] This fact, taken together with the equipotency of
1 and 2, may suggest that the native disulfide framework of
hepcidin is not essential for its bioactivity. Alternatively, it is
conceivable that the C326 thiol of ferroportin or other cell
surface thiols induce disulfide-bond rearrangement of 2 to the
native isomer 1. Indeed we found that catalytic amounts of
reduced glutathione are sufficient to effect the conversion of 2
into 1 (see the Supporting Information, p 31), thus suggesting
that the fold of 1 is thermodynamically favored over that of 2,
something that may contribute to the conformational insta-
bility of 2 observed by NMR. To investigate the possibility of
2 rearranging into 1 under the conditions of the biological
assay, 2 was incubated in the presence of cells expressing
a non-internalizing ferroportin mutant[18] and the supernatant
was analyzed by HPLC. Form 2 again rearranged into 1 (see
the Supporting Information, p 32), therefore it was not
possible to conclude whether 2 is indeed equipotent with
1 or whether the activity observed is due to 1 being generated
from 2.
Figure 1. Analytical HPLC traces of 1 (a), commercial human hepcidin
sample (Peptides International, PI 4392s; b), co-injection of 1 and PI
4392s (c), 2 (d), and co-injection of 2 and PI 4392s (e).
human hepcidin obtained by Jordan et al.;[2b] the core regions
(residues 7–23) overlaid with rmsd values of 0.90 ꢀ and
0.74 ꢀ, respectively (Figure 2). Previous NMR-based struc-
tural studies have been forced to extreme temperatures
(263 K or 325 K) to alleviate the difficulties associated with
the exchange processes proximal to the disulfide bonds;
however, we demonstrate herein that if the connectivity is
known a priori, a high-resolution structure can be obtained
even at ambient temperature (298 K). Indeed this structure is
in better agreement with the crystal structure than the
previous NMR structure at 325 K.[2b] Interestingly, the NMR
spectra of 2 indicates the presence of multiple conformations.
This non-native fold appears to be highly unstable and even at
323 K, the conformational exchange could not be averaged on
the NMR time scale, thus making structure calculation
unfeasible. Comparison of the 1H–15N and 1H–13C HSQC
In conclusion, the complete regioselective synthesis of two
disulfide-bond isomers of hepcidin was achieved by using the
newly developed Msbh thiol protecting group. We were able
to rule out 2 as a possible structure of native hepcidin and
provide further support for 1 as the native molecule with the
correct S–S connectivity. Moreover, Msbh thiol protection
expands the general capacity for the regioselective synthesis
of peptides with multiple disulfide bonds by employing either
Fmoc or Boc chemistry and complements the existing set of
sulfoxide-type safety-catch protecting groups for multidimen-
sional protection in peptide synthesis.
Figure 2. Superposition of the NMR structure of 1 obtained at 298 K
(green) with that of native hepcidin at 325 K (pink, PDB ID 2KEF) and
the X-ray structure of native hepcidin (blue, PDB ID 3H0T). Disulfide
bonds are shown in yellow.
Received: November 21, 2013
Revised: December 30, 2013
Published online: February 7, 2014
Angew. Chem. Int. Ed. 2014, 53, 2931 –2934
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2933

.
Angewandte
Communications
[8] B. Denis, E. Trifilieff, J. Pept. Sci. 2000, 6, 372 – 377.
[9] A. Cuthbertson, B. Indrevoll, Org. Lett. 2003, 5, 2955 – 2957.
[10] G. W. Kenner, J. R. McDermott, R. C. Sheppard, J. Chem. Soc.
Chem. Commun. 1971, 636 – 637.
[11] a) J. M. Samanen, E. Brandeis, J. Org. Chem. 1988, 53, 561 – 569;
b) M. Pꢅtek, M. Lebl, Tetrahedron Lett. 1991, 32, 3891 – 3894;
c) T. Kimura, T. Fukui, S. Tanaka, K. Akaji, Y. Kiso, Chem.
Pharm. Bull. 1997, 45, 18 – 26; d) S. Thennarasu, C. F. Liu,
Tetrahedron Lett. 2010, 51, 3218 – 3220.
[12] a) E. Nicolꢅs, M. Vilaseca, E. Giralt, Tetrahedron 1995, 51, 5701 –
5710; b) M. Vilaseca, E. Nicolꢅs, F. Capdevila, E. Giralt,
Tetrahedron 1998, 54, 15273 – 15286; c) A. Brust, A. E. Tickle,
J. Pept. Sci. 2007, 13, 133 – 141; d) A. Brust, C. I. A. Wang, N. L.
Daly, J. Kennerly, M. Sadeghi, M. J. Christie, R. J. Lewis, M.
Mobli, P. F. Alewood, Angew. Chem. 2013, 125, 12242 – 12245;
Angew. Chem. Int. Ed. 2013, 52, 12020 – 12023.
[13] J. Zhang, S. Diamond, T. Arvedson, B. J. Sasu, L. P. Miranda,
Biopolymers 2010, 94, 257 – 264.
[14] N. J. Saez, M. Mobli, M. Bieri, I. R. Chassagnon, A. K. Malde, R.
Gamsjaeger, A. E. Mark, P. R. Gooley, L. D. Rash, G. F. King,
Mol. Pharmacol. 2011, 80, 796 – 808.
[15] Y. Shen, F. Delaglio, G. Cornilescu, A. Bax, J. Biomol. NMR
2009, 44, 213 – 223.
[16] G. C. Preza, P. Ruchala, R. Pinon, E. Ramos, B. Qiao, M. A.
Peralta, S. Sharma, A. Waring, T. Ganz, E. Nemeth, J. Clin.
Invest. 2011, 121, 4880 – 4888.
Keywords: disulfide bonds · hepcidin · peptides ·
protecting groups · regioselective folding
.
[1] E. Nemeth, T. Ganz, Annu. Rev. Nutr. 2006, 26, 323 – 342.
[2] a) H. N. Hunter, D. B. Fulton, T. Ganz, H. J. Vogel, J. Biol.
Chem. 2002, 277, 37597 – 37603; b) J. B. Jordan, L. Poppe, M.
Haniu, T. Arvedson, R. Syed, V. Li, H. Kohno, H. Kim, P. D.
Schnier, T. S. Harvey, L. P. Miranda, J. Cheetham, B. J. Sasu, J.
Biol. Chem. 2009, 284, 24155 – 24167.
[3] M. Mobli, G. F. King, Toxicon 2010, 56, 849 – 854.
[4] a) C. Boulꢁgue, H. J. Musiol, V. Prasad, L. Moroder, Chim. Oggi
2006, 24, 24 – 36; b) L. Moroder, D. Besse, H. J. Musiol, S.
Rudolph-Bçhner, F. Siedler, Biopolymers 1996, 40, 207 – 234;
c) K. Adermann, K. Barlos in Oxidative Folding of Peptides and
Proteins (Eds.: L. Moroder, J. Buchner), The Royal Society of
Chemistry, London, 2009, pp. 297 – 317.
[5] H. Liu, M. A. Boudreau, J. Zheng, R. M. Whittal, P. Austin, C. D.
Roskelley, M. Roberge, R. J. Andersen, J. C. Vederas, J. Am.
Chem. Soc. 2010, 132, 1486 – 1487.
[6] a) M. Gꢂngora-Benꢃtez, J. Tulla-Puche, M. Paradꢃs-Bas, O.
Werbitzky, M. Giraud, F. Albericio, Biopolymers 2011, 96, 69 –
80; b) R. Eliasen, T. L. Andresen, K. W. Conde-Frieboes,
Peptides 2012, 37, 144 – 149; c) H. Gali, G. L. Sieckman, T. J.
Hoffman, N. K. Owen, D. G. Mazuru, L. R. Forte, W. A. Volkert,
Bioconjugate Chem. 2002, 13, 224 – 231; d) C. Kellenberger, H.
Hietter, B. Luu, Pept. Res. 1995, 8, 321 – 327.
[7] a) E. Klꢄver, S. Schulz-Maronde, S. Scheid, B. Meyer, W. G.
Forssmann, K. Adermann, Biochemistry 2005, 44, 9804 – 9816;
b) A. Schulz, E. Klꢄver, S. Schulz-Maronde, K. Adermann,
Biopolymers 2005, 80, 34 – 49; c) I. Szabꢂ, G. Schlosser, F.
[17] E. Nemeth, G. C. Preza, C. L. Jung, J. Kaplan, A. J. Waring, T.
Ganz, Blood 2006, 107, 328 – 333.
[18] B. Qiao, P. Sugianto, E. Fung, A. del-Castillo-Rueda, M. J.
Moran-Jimenez, T. Ganz, E. Nemeth, Cell Metab. 2012, 15, 918 –
924.
˝
Hudecz, G. Mezo, Biopolymers 2007, 88, 20 – 28.
2934 www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 2931 –2934