In pursuit of carbohydrate-based HIV vaccines, part 2: The total synthesis of high-mannose-type gp120 fragments--evaluation of strategies directed to maximal convergence.

Full text of DOI:10.1002/anie.200353626

Communications
Glycopeptides (2)
synthesis of any Man9GlcNAc2 containing glycopeptides has
been reported.[8]
In Pursuit of Carbohydrate-Based HIV Vaccines,
In our route to the glycan portion of the glycopeptide, we
utilized, as proposed earlier, trisaccharide 2,[9] which already
encompasses the synthetically difficult b-mannosidic linkage,
as well as differentiated C3 and C6 access points (see
asterisks)for the subsequent introduction of the nonsym-
metrical mannose branching pattern.
From this point onward, two strategies for progression to
the octamannose motif presented themselves. One strategy
would start with two consecutive mannosylations of the 3-OH
and 6-OH groups of 2, employing mannoside donors 3 and 4,
respectively, to complete the first “mannose layer”. In turn,
the second “layer” of three mannose units would be
introduced by triple mannosylation of the pentasaccharide
triol acceptor with mannoside donor 3, providing the Man-6
octasaccharide. Saponification of the esters followed by the
introduction of another trimannose layer should provide the
desired Man-9 undecamer glycan (Scheme 1; “layered
approach”).
Alternatively, one would construct the “upper” penta-
mannose 5 and “lower” trimannose 6 building blocks
separately, followed by coupling them with the key trisac-
charide 2 at the “real” (C3)and the “virtual” (C6)acceptor
sites (see asterisks), thus reaching the undecamer 7b in a
highly convergent fashion (Scheme 1; “block approach”).
With the glycan matrix assembled, the next phases of the
program would involve global deprotection[10] followed by
amination at the anomeric site.[11] We initially envisioned that
a small (penta)peptide would be introduced by aspartyla-
tion.[12,13] Finally, native chemical ligation (NCL)would
complete the synthesis of 1 (Scheme 1), paving the way for
conjugation to a carrier immunogen en route to fashioning a
testable vaccine.[14,15]
The “layered” approach was explored first (Scheme 2).
Glycosylation of the 3-OH group of trisaccharide 2 with
ethylthiomannoside donor 3 under the Sinay¨ radical activa-
tion conditions[16] gave tetrasaccharide 8 bearing the benzyl-
idene group spanning C4 and C6. The acetal linkage was
opened in a reductive fashion to afford tetrasaccharide 9. The
primary hydroxy group of 9 was in turn mannosylated with
phenylthiomannoside 4. Saponification of the resulting pen-
tasaccharide 10 exposed the three required acceptor sites (see
asterisks). Trimannosylation of 11 delivered octasaccharide 12
in high yield (55%). This protocol (saponification followed by
trimannosylation)was repeated to synthesize the desired
protected undecasaccharide 7a.
Part 2: The Total Synthesis of High-Mannose-
Type gp120 Fragments—Evaluation of Strategies
Directed to Maximal Convergence**
Xudong Geng, Vadim Y. Dudkin, Mihirbaran Mandal,
and Samuel J. Danishefsky*
There are strong grounds to suppose that some selected
glycosylation patterns of the HIV viral protein gp120 can
themselves serve as epitopes for potent, broadly neutralizing
antibodies (e.g. 2g12).[1,2] The epitopes in question may
comprise several hybrid or high-mannose-type glycans at
particular asparagine loci (Asn295, 332, 339, 386, and 392).
The 2g12 antibody has been shown to recognize a cluster of
a1!2 linked mannose residues on the HIV surface. Another
argument in favor of the high-mannose-type glycan cluster
epitope was reported by Burton, Wilson, and co-workers.[3]
These workers described a structure of 2g12 cocrystallized
with the high-mannose-type reducing oligosaccharide Man9-
GlcNAc2. The crystal structure demonstrated that the anti-
body may bind up to four individual high-mannose glycans
simultaneously, thus favoring a very high affinity recognition.
Accordingly, a synthetic construct that is able to elicit a strong
immune response to a conserved cluster of gp120 high-
mannose glycans could potentially emerge as a valuable
candidate for incorporation into an HIV vaccine. In the
preceding paper,[4] we related a strategy for the construction
of a hybrid type gp120 glycopeptide construct.
Herein we describe the synthesis of gp120 fragments
comprising one of key asparagine sites (332)modified with a
fully synthetic high-mannose glycan. Although the nonaman-
nose section of the molecule was previously prepared and
tested in binding with cyanovirin-N,[5–7] no total chemical
[*] Dr. X. Geng, Dr. V. Y. Dudkin, Dr. M. Mandal, Prof. S. J. Danishefsky
Laboratory for Bioorganic Chemistry
Sloan–Kettering Institute for Cancer Research
1275 York Avenue, New York, NY 10021 (USA)
Fax: (+1)212-772-8691
E-mail: s-danishefsky@ski.mskcc.org
and
Department of Chemistry
Columbia University
Having demonstrated that the protected undecasaccha-
ride could be assembled by the “layered approach” in an
efficient manner (7 steps, 11% overall yield), we explored a
still more convergent “block approach” (Scheme 3). Penta-
saccharide block 5 was assembled efficiently through two
consecutive dimannosylation reactions starting from phenyl-
thiol mannoside 14 and chloromannose donor 15. The
Havemeyer Hall, 3000 Broadway, New York, NY 10027 (USA)
[**] This work was supported by the National Institutes of Health (CA-
28824). US Army Breast Cancer Foundation postdoctoral fellowship
supportis gratefully acknowledged by X.G. (BC022120) and V.D.
(BC020513). We thank Ms. Anna Dudkina and Ms. Sylvi Rusli (NMR
Core Facility, CA-02848) for mass spectrometric analyses, Dr.
George Sukenic for NMR spectroscopic analyses, and Dr. J. David
Warren for help in the preparation of starting materials. We also
thank Dr. Justin S. Miller, Dr. Tom Muir, Mr. Michael Hahn, and Mr.
Matthew Sekedat for their generous help in the preparation of the
peptides.
“lower” two trisaccharide “blocks” were joined by
a
MeOTf-mediated glycosylation to afford hexasaccharide 18
efficiently. Reduction of 18 released the primary hydroxy
group to give 19, which was then subjected to a 6 + 5
glycosylation with donor 5. Following the examination of
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
2562
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200353626
Angew. Chem. Int. Ed. 2004, 43, 2562 –2565

Angewandte
Chemie
Scheme 1. Synthetic strategy for the assembly of the gp120 glycopeptide fragments. TBS=tert-butyldimethylsilyl.
Scheme 2. Synthesis of Undecasaccharide 7a through the “layered” approach. a) 3, (BrC6H4)3NSbCl6, CH3CN, 4 h, 78%; b) BH3, Bu2BOTf, THF,
08C, 7 h, 90%; c) 4, (BrC6H4)3NSbCl6, CH3CN, 4 h, 74%; d) NaOMe, MeOH, 12 h, 91%; e) 3, (BrC6H4)3NSbCl6, CH3CN, 12 h, 55%; f) NaOMe,
MeOH, 12 h, 84%; g) 3, (BrC6H4)3NSbCl6, CH3CN, 12 h, 51%. Tf=trifluoromethanesulfonyl.
Angew. Chem. Int. Ed. 2004, 43, 2562 –2565
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2563

Communications
Scheme 3. Synthesis of undecasaccharide 7b through “block” approach. a) 15, AgOTf, DTBP, CH2Cl2, ꢀ108C!RT, 18 h; b) NaOMe, MeOH, 10 h,
50% for two steps; c) 15, AgOTf, DTBP, CH2Cl2, ꢀ108C to RT, 18 h; 87%; d) MeOTf, DTBP, CH2Cl2, ꢀ408C to RT, 12 h, 70%; e) BH3, Bu2BOTf,
THF, 08C, 7 h, 86%; f) 5, (BrC6H4)3NSbCl6, CH3CN, 10 h, 63% (85% based on recovered 19). DTBP=2,6-di-tert-butylpyridine.
several protocols for coupling, it was found
that the Sinay¨ radical activating conditions
worked best,[16] delivering the desired unde-
casaccharide 7b in 63% yield (85% based
on recovered acceptor 19). At least in this
endeavor, the ultimately convergent “block
approach” was indeed shown to be more
concise, bringing forward the protected
high-mannose 7b in 51% yield over three
steps (starting from 2).
With the protected oligosaccharide in
hand, we proceeded to the next phase,
global deprotection (Scheme 4). Deacety-
lation, desilylation, and reduction (dissolv-
ing metal)of 7 afforded free glycan 20.[10]
The latter was advanced to glycosylamine
21 through Kochetkov amination.[11] This
compound was first coupled with gp120331–
335
pentapeptide segment 22 (bearing Asp
with the protected cysteine thiol and the N-
terminal amino groups). Following removal
of the Fmoc group and reduction of the
Scheme 4. Synthesis of gp120 glycosylated fragments 25. a) NaOMe/MeOH, 12 h, 96%;
disulfide, gp120331–335 pentapeptide–high-
b) TBAF, HAc, THF, 08C, 1 h, 98%; c) Na/NH3(L), ꢀ788C, 2 h; d) Ac2O, NaHCO3(sat.
mannose glycan conjugate 25 was in hand.
Our initial expectation was to utilize
native chemical ligation to complete the
assembly of the gp120316–335 peptide–high-
mannose glycan conjugate. However, NCL
with 24 (thiol on Cys protected)or 25 (free
thiol on Cys)failed to deliver the desired
gp120316–335 peptide–high-mannose glycan
aqueous solution), 87% for two steps; e) NH4HCO3(sat. aqueous solution), 2 d, 408C;
f) 22, HATU, DIPEA, DMSO, 7 h; g) piperidine, DMF, 15 min, 24% from 20;
h) HSCH2CH2SO3Na, TCEP, 3 days, 60%. Fmoc=9-fluorenylmethyloxycarbonyl. TBAF=
tetra-n-butylammonium fluoride; HATU=O-(7-azabenzotriazol-1-yl)-N,N,N’,N’-tetramethy-
luronium hexafluorophosphate; DIPEA=diisopropylethylamine; DMSO=dimethyl sulfox-
ide; DMF=N,N-dimethylformamide; TCEP=tris(2-carboxyethyl)phosphane hydrochloride.
conjugate after several attempts. We recall that in the
preceding manuscript,[4] a similar breakdown was noted with
the same polypeptide elements and a related complex glycan.
Together, these cases underscore unexpected limitations in
the applicability of NCL in such highly ornate settings.
Fortunately, as reported earlier,[4] direct coupling with a
gp120316–335 eicosapeptide fragment 27 (bearing Asp at 332
with protected Lys and the N-terminus amino groups)was
feasible (Scheme 5). The desired conjugate gp120316–335 pep-
tide–high-mannose glycan, 1, was isolated following asparty-
2564
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 2562 –2565

Angewandte
Chemie
Scheme 5. Synthesis of gp120 glycosylated fragments 1. a) 27, HATU, DIPEA, DMSO, 7 h; b) N2H4, piperidine, DMF, 15 min, 16% from 20.
ivDde=4,4-dimethyl-2,6-dioxocyclohex-1-ylidine-3-methylbutyl.
K. H. Roux, J. W. Kelly, P. M. Rudd, R. A. Dwek, H. Katinger,
D. R. Burton, I. A. Wilson, Science 2003, 300, 2065 – 2071.
[4] See preceding manuscript in this issue: M. Mandal, V. Y.
Dudkin, X. Geng, S. J. Danishefsky, Angew. Chem. 2004, 116,
2611 – 2615; Angew. Chem. Int. Ed. 2004, 43, 2557 – 2561.
[5] L. G. Barrientos, J. M. Louis, D. M. Ratner, P. H. Seeberger,
A. M. Gronenborn, J. Mol. Biol. 2003, 325, 211 – 223.
[6] I. Botos, B. R. O'Keefe, S. R. Shenoy, L. K. Cartner, D. M.
Ratner, P. H. Seeberger, M. R. Boyd, A. Wlodawer, J. Biol.
Chem. 2002, 277, 34336 – 34342.
lation and deprotection in 16% yield over 3 steps. Compound
1 was purified by reverse-phase HPLC and 1H NMR spectro-
scopic analysis and MS data[17] are consistent with the desired
structure as a homogeneous entity. Full characterizations are
provided in the Supporting Information.
In summary, we have reported the first chemical synthesis
of gp120 glycopeptide fragments (high-mannose-type conju-
gate gp120316–335 1 and gp120331–335 25). The glycan was
assembled through two efficient methods, that is, a “layered
approach” and a “block approach”, and then conjugated with
gp120 peptide segments through direct aspartylation. In
combination with the preceding manuscript,[4] our total
synthesis program provides direct access to mimics of the
epitope of broadly neutralizing antibody 2g12, that is, high-
mannose and hybrid-type gp120 glycopeptide fragments.
With the organic synthesis phase complete, the project has
entered the immunogen conjugation phase, en route to a
thorough evaluation of the immunological issues discussed
earlier.[4]
[7] D. M. Ratner, O. J. Plante, P. H. Seeberger, Eur. J. Org. Chem.
2002, 826 – 833.
[8] I. Matsuo, M. Wada, S. Manabe, Y. Yamaguchi, K. Otake, K.
Kato, Y. Ito, J. Am. Chem. Soc. 2003, 125, 3402 – 3403.
[9] V. Y. Dudkin, J. S. Miller, S. J. Danishefsky, Tetrahedron Lett.
2003, 44, 1791 – 1793. The stereochemistry of glycosidic linkages
1
in 7 was confirmed by a combination of HMQC and JC,H
measurements. For details, see Supporting Information.
[10] U. Iserloh, V. Dudkin, Z. G. Wang, S. J. Danishefsky, Tetrahe-
dron Lett. 2002, 43, 7027 – 7030.
[11] L. M. Likhosherstov, O. S. Novikova, V. A. Derevitskaja, N. K.
Kochetkov, Carbohydr. Res. 1986, 146, C1-C5.
[12] S. T. Cohen-Anisfeld, P. T. Lansbury, J. Am. Chem. Soc. 1993,
115, 10531 – 10537.
Received: December 29, 2003 [Z53626]
[13] J. S. Miller, V. Y. Dudkin, G. J. Lyon, T. W. Muir, S. J. Danishef-
sky, Angew. Chem. 2003, 115, 447 – 450; Angew. Chem. Int. Ed.
2003, 42, 431 – 435.
[14] T. J. Tolbert, C.-H. Wong, J. Am. Chem. Soc. 2000, 122, 5421 –
5428.
Keywords: antigens · glycoconjugates · glycopeptides ·
mannosylation · total synthesis
.
[15] P. E. Dawson, T. W. Muir, I. Clark-Lewis, S. B. H. Kent, Science
1994, 266, 776 – 779.
[1] R. W. Sanders, M. Venturi, L. Schiffner, R. Kalyanaraman, H.
Katinger, K. O. Lloyd, P. D. Kwong, J. P. Moore, J. Virol. 2002,
76, 7293 – 7305.
[2] C. N. Scanlan, R. Pantophlet, M. R. Wormald, E. Ollmann Sa-
phire, R. Stanfield, I. A. Wilson, H. Katinger, R. A. Dwek, P. M.
Rudd, D. R. Burton, J. Virol. 2002, 76, 7306 – 7321.
[16] Y. M. Zhang, J. M. Mallet, P. Sinay, Carbohydr. Res. 1992, 236,
73 – 88.
[17] LRMS (ESI)( m/z): calcd for C164H278N35O80S2 [M+3H]3+
1360.6, found: 1360.7; calcd for C164H279N35O80S2 [M+4H]4+
1020.7, found: 1020.6.
:
:
[3] D. A. Calarese, C. N. Scanlan, M. B. Zwick, S. Deechongkit, Y.
Mimura, R. Kunert, P. Zhu, M. R. Wormald, R. L. Stanfield,
Angew. Chem. Int. Ed. 2004, 43, 2562 –2565
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2565