Total synthesis of human galanin-like peptide through an aspartic acid ligation

Article abstract of DOI:10.1021/ol402984r

Human galanin-like peptide (hGALP) is a newly discovered hypothalamic peptide that plays important roles in the regulation of food intake and energy balance. Here, we demonstrate that the aspartic acid ligation can be employed to achieve an efficient synt

Full text of DOI:10.1021/ol402984r

ORGANIC
LETTERS
2013
Vol. 15, No. 24
6128–6131
Total Synthesis of Human Galanin-Like
Peptide through an Aspartic Acid Ligation
Xiaoyang Guan, Matthew R. Drake, and Zhongping Tan*
Department of Chemistry and Biochemistry and BioFrontiers Institute, University of
Colorado, Boulder, Colorado 80303, United States
zhongping.tan@colorado.edu
Received October 16, 2013
ABSTRACT
Human galanin-like peptide (hGALP) is a newly discovered hypothalamic peptide that plays important roles in the regulation of food intake and
energy balance. Here, we demonstrate that the aspartic acid ligation can be employed to achieve an efficient synthesis of hGALP. The total
synthesis of hGALP enhances our ability to study its biology and facilitates the development of more stable analogues.
Chemical synthesis is a valuable tool for studying
the structure and function of naturally occurring proteins
as well as modified and engineered proteins because it
imparts a previously unattainable level of control over
protein composition.1 Encouraged by previous studies on
protein chemical synthesis, we initiated a program directed
toward the total chemical synthesis of a newly discovered
neuropeptide, human galanin-like peptide (hGALP).2
Recent findings have revealed that GALP plays important
roles in energy metabolism, thermoregulation, reproduc-
tion, as well as appetite control.3 When administered by
intracerebroventricular (i.c.v.) injection, GALP decreases
overall food intake and causes a decrease in body weight in
rats.4 Because of its important physiological functions,
there is a strong interest in developing GALP as a ther-
apeutic for the treatment of obesity.3 However, similar to
other peptide regulators, GALP has a very short half-life in
blood, which necessitates frequent injections or infusions.5
Therefore, the development of metabolically more stable
analogues of GALP is needed. Chemical synthesis could be
well suited for this purpose. As has been widely demon-
strated, once a general synthetic route to GALP is estab-
lished, its synthetic variants can be produced relatively
quickly. In principle, GALP analogues with more suitable
bioprofiles can be identified by screening the collection of
its synthetic variants.
(1) (a) Dawson, P. E.; Muir, T. W.; Clark-Lewis, I.; Kent, S. B.
Science 1994, 266, 776. (b) Erlanson, D. A.; Chytil, M.; Verdine, G. L.
Chem. Biol. 1996, 3, 981. (c) Muir, T. W.; Sondhi, D.; Cole, P. A. Proc.
Natl. Acad. Sci. U.S.A. 1998, 95, 6705. (d) Beligere, G. S.; Dawson, P. E.
Biopolymers 1999, 51, 363. (e) Marcaurelle, L. A.; Mizoue, L. S.; Wilken,
J.; Oldham, L.; Kent, S. B.; Handel, T. M.; Bertozzi, C. R. Chemistry
2001, 7, 1129. (f) Grogan, M. J.; Kaizuka, Y.; Conrad, R. M.; Groves,
J. T.; Bertozzi, C. R. J. Am. Chem. Soc. 2005, 127, 14383. (g) Yamamoto,
N.; Tanabe, Y.; Okamoto, R.; Dawson, P. E.; Kajihara, Y. J. Am. Chem.
Soc. 2008, 130, 501. (h) Piontek, C.; Ring, P.; Harjes, O.; Heinlein, C.;
€
€
ꢀ
Mezzato, S.; Lombana, N.; Pohner, C.; Puttner, M.; Varon Silva, D.;
Martin, A.; Schmid, F. X.; Unverzagt, C. Angew. Chem., Int. Ed. 2009,
48, 1936. (i) Kumar, K. S.; Bavikar, S. N.; Spasser, L.; Moyal, T.;
Ohayon, S.; Brik, A. Angew. Chem., Int. Ed. 2011, 50, 6137. (j) Shang, S.;
Tan, Z.; Danishefsky, S. J. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 5986.
(k) Wilkinson, B. L.; Stone, R. S.; Capicciotti, C. J.; Thaysen-Andersen,
M.; Matthews, J. M.; Packer, N. H.; Ben, R. N.; Payne, R. J. Angew.
Chem., Int. Ed. 2012, 51, 3606. (l) Wang, P.; Dong, S.; Brailsford, J. A.;
Iyer, K.; Townsend, S. D.; Zhang, Q.; Hendrickson, R. C.; Shieh, J.;
Moore, M. A.; Danishefsky, S. J. Angew. Chem., Int. Ed. 2012, 51, 11576.
(m) Zhan, C.; Varney, K.; Yuan, W.; Zhao, L.; Lu, W. J. Am. Chem. Soc.
2012, 134, 6855.
The synthesis of hGALP (1) may be achieved in several
ways. Detailed analysis of its sequence suggested to us the
possibility of disconnecting it into two fragments (1ꢀ31) 3
(3) Shioda, S.; Kageyama, H.; Takenoya, F.; Shiba, K. Int. J. Obes.
2011, 35, 619.
(4) Kumano, S.; Matsumoto, H.; Takatsu, Y.; Noguchi, J.; Kitada,
C.; Ohtaki, T. Endocrinology 2003, 144, 2634.
(5) Nonaka, N.; Farr, S. A.; Kageyama, H.; Shioda, S.; Banks, W. A.
J. Pharmacol. Exp. Ther. 2008, 325, 513.
(2) Ohtaki, T.; Kumano, S.; Ishibashi, Y.; Ogi, K.; Matsui, H.;
Harada, M.; Kitada, C.; Kurokawa, T.; Onda, H.; Fujino, M. J. Biol.
Chem. 1999, 274, 37041.
r
10.1021/ol402984r
Published on Web 11/22/2013
2013 American Chemical Society

Scheme 1. Retrosynthetic Analysis for the Synthesis of hGALP
Scheme 2. Synthesis of the Ligation Precursor of Aspartic Acida
and (32ꢀ60) 4, each containing about 30 amino acids. In
this way, it may be possible to effectively prevent synthetic
problems that may arise from aspartimide formation.6
Human GALP contains four Asp residues (Scheme 1,
enlarged amino acids). By strategically disconnecting
hGALP at the Gln-Asp site, two aspartic acids, Asp30
and Asp32, are the firstand lastbuilding block being added
to the growing peptide chain during Fmoc-based solid-
phase peptide synthesis (SPPS). They are either not next to
a R-carboxamide or not exposed to strong bases and are
thus stable during synthesis. Aspartimide formation at
Asp43 and 49 can be reduced by using Fmoc-Asp(OMpe)-
OH and Fmoc-Asp(OtBu)-(Dmb)Gly-OH (Mpe = β-3-
methylpent-3-yl, Dmb =2,4-dimethoxybenzyl).6b,c
We began our synthesis by preparing β-thioaspartic acid
5. As shown in Scheme 1, the Asp surrogate 5 has two func-
tional groups, a sulfhydryl group and a carboxyl group,
attached to its β-carbon atom. There are two possible
diastereomers of protected thioaspartic acids: 5 and its
β-epimer.7 Recently, Payne and co-workers reported a na-
tive chemical ligation at aspartic acid. Their results showed
that there is no noticeable difference in the ligation rate
between the two epimers.7d
this hypothesis, we first prepared the thiobenzamidoma-
lonic ester 7.9 The lithium dianion of 7 was formed at 0 °C
in THF using LiHMDS (2.0 equiv). Happily, quenching
the dianion with I2 followed by an acidic workup gave
the trans thiazoline 8 in 38% yield over two steps.8 No cis-
isomer was detected. In addition to the desired product,
about 30% 1,3-thiazole compound was also produced
in this reaction. Selective hydrolysis of thiazoline 8 using
2.0 N HCl afforded the N-benzoyl derivative 9 in 54%
yield.8 The free thiol in 9 was protected with a trityl group.
To decrease the radical-mediated addition of the thiol to
the aromatic carbons, this reaction was carried out in the
absence of base.10 After the trityl protection of the SH
group, the benzoyl group was replaced by a Boc group
undertheoptimized, nonepimerizingconditions.11 Finally,
the methyl ester deprotection was accomplished under
mild conditions using trimethyltin hydroxide to give the
desired aspartic acid surrogate 5 in 97% yield.12 This new
route is four steps longer than the method reported by
Payne et al.7d
With the thioaspartic acid in hand, we went on to
evaluate the feasibility and versatility of the aspartic acid
ligation using a model ligation test system, which has been
widely used in the mechanistic study of native chemical
ligation (Table 1).13 This study is similar to that of
Thompson et al. The only difference is that this study used
different model peptides and a different diastereomer of
the chiral thioaspartic acid. First, we tested the feasibility
of β-thioaspartic acid mediated ligations with peptide
To develop anefficient synthesis of β-thioaspartic acid5,
we attempted the direct thiylation of the protected aspartic
acid(Scheme 2). Previous work on stereoselective synthesis
of trans-oxazoline via iodination of dianion marginally
supports the feasibility of sucha synthetic strategy.8 To test
(6) (a) Tam, J. P.; Riemen, M.; Merrifield, R. B. Pept. Res. 1988, 1, 6.
(b) Karlstrom, A.; Unden, A. Tetrahedron Lett. 1996, 37, 4343. (c)
Cardona, V.; Eberle, I.; Barthelemy, S.; Beythien, J.; Doerner, B.;
Schneeberger, P.; Keyte, J.; White, P. D. Intl. J. Peptide. Res. Therapeut.
2008, 14, 285.
(9) Hauptmann, H.; Berl, H. J. Am. Chem. Soc. 1955, 77, 704.
(10) Ashby, E. C.; Park, W. S.; Goel, A. B.; Su, W. Y. J. Org. Chem.
1985, 50, 5184.
(11) Burk, M. J.; Allen, J. G. J. Org. Chem. 1997, 62, 7054.
(12) Nicolaou, K. C.; Estrada, A. A.; Zak, M.; Lee, S. H.; Safina,
B. S. Angew. Chem., Int. Ed. 2005, 44, 1378.
(7) (a) Chen, J.; Wan, Q.; Yuan, Y.; Zhu, J.; Danishefsky, S. J.
Angew. Chem., Int. Ed. 2008, 47, 8521. (b) Harpaz, Z.; Siman, P.; Kumar,
K. S.; Brik, A. ChemBioChem 2010, 11, 1232. (c) Tan, Z.; Shang, S.;
Danishefsky, S. J. Angew. Chem., Int. Ed. 2010, 49, 9500. (d) Thompson,
R. E.; Chan, B.; Radom, L.; Jolliffe, K. A.; Payne, R. J. Angew. Chem.,
Int. Ed. 2013, 52, 9723.
(8) Cardillo, G.; Gentilucci, L.; Tolomelli, A.; Tomasini, C. Synlett
1999, 1999, 1127.
(13) Hackeng, T. M.; Griffin, J. H.; Dawson, P. E. Proc. Natl. Acad.
Sci. U.S.A. 1999, 96, 10068.
Org. Lett., Vol. 15, No. 24, 2013
6129

22 and 24 were changed to prevent the overlap of the peaks
of 23and 25and that of MPAA during HPLC purification.
Similar to the ligations of isocysteine and γ-mercaptoglu-
tamine, the rate and efficiency of aspartic acid ligation was
not found to be markedly different from Cys-mediated
reaction, whichindicatesthatthepresenceof aneighboring
carboxyl group does not significantly affect the ability of
the β-thiol to catalyze native chemical ligations.7d,14
Table 1. Substrate Scope of Aspartic Acid Ligationa
Figure 1. (A) LCꢀMS trace of the authentic tripeptides. (B)
LCꢀMS trace of the desulfurization products. Reagents and
conditions: (a) 6 M Gn HCl, 300 mM Na2HPO4, 200 mM
3
MPAA, 20 mM TCEP, pH 7.9. (b) TCEP, VA-044, t-BuSH,
1.5 h.
a Reagents and conditions: 6 M Gn HCl, 300 mM Na2HPO4,
200 mM MPAA, 20 mM TCEP, pH 7.9. P1 = peptide fragments,
3
Having examined the reactivity of β-thioaspartic acid in
ligation reactions, we next investigated the efficiency of the
metal-free desulfurization (MFD) in converting thioaspar-
tic acid into the natural aspartic acid. Happily, we found
that the adjacent carbonyl group does not have a destruc-
tive effect on the desulfurization reaction. Under the orig-
inal conditions, we were able to obtain all the desired
peptide products in 4 h. The yield ranged from 50% for
the desulfurization of 25 to 77% for that of 15.15
We also confirmed that the stereochemical integrity of
the R-carbon in the thioaspartic acid was preserved during
all reaction steps. The level of racemization of aspartic
acid was investigated based on the LCꢀMS analysis of a
tripeptide, H-Gly-Asp-Phe-NH2. We first prepared two
authentic tripeptide samples, H-Gly-Asp-Phe-NH2 26 and
H-Gly-DAsp-Phe-NH2 27.16 Consistent with previous
reports, these two diastereomers can be well separated
Gn HCl = guanidine hydrochloride, TCEP = tris(2-carboxyethyl)-
3
phosphine, MPAA = 4-mercaptophenylacetic acid.
thioesters bearing C-terminal Gly (entry 1, Table 1). The
reaction proceeded rapidly and was nearly completed in
less than 10 min. After 2 h, the reaction was quenched with
sodium 2-mercaptoethanesulfonate (MESNA). A subse-
quent HPLC purification gave the desired peptide 15 in an
isolated yield of 85%. As expected, two other favorable
thioesters 16 and 18, which had C-terminal Phe and Leu
residues, also underwent rapid ligation with the peptide 13
to generate the corresponding products 17 and 19 (entries 2
and 3). To further probe the scope of the aspartic acid
ligation, peptide thioesters presenting β-branched C-term-
inal residues, including Thr, Val, and Ile, were also tested.
Interestingly, the ligation reaction was not slowed by the
β-hydroxyl group. It was also completed in 2 h to afford
the desired peptide product in 84% isolated yield. As
expected, the ligation reactions proceeded much less effec-
tively with the Val and Ile peptide thioesters, 22 and 24. As
shown in Table 1, these reactions were nearly completed in
10 h to afford the ligation products 23 and 25 in relatively
low yields (77% and 69%, respectively). The sequences of
(14) (a) Johnson, E. C.; Kent, S. B. J. Am. Chem. Soc. 2006, 128, 6640.
(b) Dose, C.; Seitz, O. Org. Biomol. Chem. 2004, 2, 59. (c) Siman, P.;
Karthikeyan, S. V.; Brik, A. Org. Lett. 2012, 14, 1520.
(15) (a) Wan, Q.; Danishefsky, S. J. Angew. Chem., Int. Ed. 2007, 46,
9248. (b) Haase, C.; Rohde, H.; Seitz, O. Angew. Chem., Int. Ed. 2008,
47, 6807. (c) Metanis, N.; Keinan, E.; Dawson, P. E. Angew. Chem., Int.
Ed. 2010, 49, 7049.
(16) Han, Y.; Albericio, F.; Barany, G. J. Org. Chem. 1997, 62, 4307.
6130
Org. Lett., Vol. 15, No. 24, 2013

by LCꢀMS, exhibiting a difference in retention time of
>0.5 min (Figure 1). The chirality of the thioaspartic acid
can therefore be determined by comparing the desulfuriza-
tion products of the tripeptide precursor 30 with 26 and 27.
The experiment was performed as shown in Figure 1B.
After native chemical ligation of 28 and 29, the roughly
purified30was subjected tothe standardMFD conditions.
Only a trace amount of 27 (<1%) was detected.
The effectiveness of aspartic acid ligation in the elimina-
tion of the problem of aspartimide formation was also
investigated. The study was performed through the syn-
thesis of peptide H-Leu-Tyr-Arg-Ala-Gly-Asp-Gly-Tyr-
Leu-Ala-Ala-OH (31), which is highly prone to asparti-
mide formation. Under standard SPPS conditions, the
synthesis using Fmoc-Asp(OtBu)-OH resulted in an ex-
tensive formation of inseparable aspartimide and related
side products (>50%). Although the use of Fmoc-Asp-
(OMpe)-OH reduced the levels of aspartimide formation,
the side reactions still occurred to the extent of approxi-
mately 10% in such a short peptide. As expected, the as-
partic acid ligation provided sufficient protection against
aspartimide formation. No detectable aspartimide pro-
ducts were observed during peptide synthesis, ligation,
and desulfurization.
Having established the feasibility of Asp ligation for the
preparation of aspartimide-prone peptide sequences, we
set out to apply this strategy to the total synthesis of
hGALP (1). Following the synthetic plan shown in
Scheme 1, we first prepared the two hGALP fragments 3
and 4. Similar to the synthesis of the model peptides,
the peptide thioester 3 was prepared from the fully pro-
tected peptide fragment using the method described by
Danishefsky and co-workers.7c The β-thioaspartic acid-
containing peptide fragment 4 was synthesized by using 5,
Fmoc-Asp(OMpe)-OH, and Fmoc-Asp(OtBu)-(Dmb)-
Gly-OH as building blocks during solid-phase synthesis.
These two fragments were purified using HPLC. As ex-
pected, native chemical ligation between 3 and 4 proceeded
Figure 2. (A) LCꢀMS spectrum of the synthetic hGALP. [M þ
7H]7þ m/z = 929.0548 Da. (B) CD spectra of the synthetic
hGALP in 5 mM phosphate buffer (pH 7) containing 100 mM
NaF.
minimum at about 199 nm and a shape consistent with a
random coil conformation (Figure 2B).
In summary, the study described here highlights the
development of aspartic acid ligation for the synthesis of
an attractive target, hGALP. To achieve this goal, we
developed the synthesis of a useful native chemical ligation
precursor, β-thioapartic acid. For the first time, we demon-
strated the feasibility of direct thiylation for the enantio-
selective introduction of a thiol group to the β-carbon of
amino acids. In addition, our studies revealed that the
attachment of a carboxyl group on the adjacent carbon
does not significantly change the reactivity of the thiol
group in the NCL reaction. More importantly, we vali-
dated the efficacy of the Asp ligation in suppressing aspar-
timide formation. Finally, using the Asp-based NCL and
MFD, we were able to synthesize hGALP with a high yield
(44%) and purity. Our study further demonstrates the
power of NCL/MFD in protein chemical synthesis and
paves the way for future development of new hGALP
analogues.
rapidly under the standard conditions (6 M Gn HCl,
3
300 mM Na2HPO4, 200 mM MPAA, 20 mM TCEP,
pH 7.9) at room temperature. The reaction was completed
in 2 h to afford the thiol-substituted hGALP 2 in a 56%
isolated yield.
Acknowledgment. We are grateful to Professor Tarek
Sammakia of the University of Colorado Boulder for
useful discussions. We thank the University of Colorado
Boulder and the Butcher Seed Grant for financial support.
We also thank Dr. William Old and Dr. Nan Wang of
the University of Colorado Boulder for mass spectral
assistance.
Having established the feasibility of Asp ligation for the
preparation of 2, we next attempted the MFD of 2. As
expected, the desulfurization reaction proceeded rapidly
and cleanly [3 M Gn HCl, 100 mM NaH2PO4, 250 mM
3
TCEP, 40 mM glutathione, UV (365 nm), pH 6.5] toafford
hGALP (1) in 80% isolated yield (Figure 2A).15b The
LCꢀMS spectrum of the synthetic hGALP is identical to
that of the commercial one, while its purity is higher than
the reference sample. This result, together with the peptide
sequencing with tandem MS/MS, confirmed the identity
of synthetic product (Supporting Information). The CD
spectrum of the synthetic hGALP also closely resembles
that of the commercially available molecule, with a
Supporting Information Available. Experimental pro-
cedures; spectroscopic and analytical data. This material
is available free of charge via the Internet at http://pubs.
acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 24, 2013
6131