Total synthesis of the triglycosyl phenolic glycolipid PGL-tb1 from mycobacterium tuberculosis

Article abstract of DOI:10.1002/anie.201206221

Complex: The synthesis of the glycolipid PGL-tb1 present in the outer membrane of hypervirulent strains of Mycobacterium tuberculosis has been accomplished for the first time by using a highly convergent strategy featuring a Sonogashira coupling to unite a phenolic trisaccharide with the phthiocerol. Efficient asymmetric Cu-catalyzed 1,4-additions to unsaturated thioesters and cyclic enones have been employed to introduce the methyl groups. Copyright

Full text of DOI:10.1002/anie.201206221

.
Angewandte
Communications
DOI: 10.1002/anie.201206221
Virulence Factors
Total Synthesis of the Triglycosyl Phenolic Glycolipid PGL-tb1 from
Mycobacterium tuberculosis**
Santiago Barroso, Riccardo Castelli, Marc P. Baggelaar, Danny Geerdink, Bjorn ter Horst,
Eva Casas-Arce, Herman S. Overkleeft, Gijsbert A. van der Marel, Jeroen D. C. Codꢀe,* and
Adriaan J. Minnaard*
Mycobacterium tuberculosis (M. tb) is one of the most
important pathogens. Despite the availability of antibiotics
and a vaccine (BCG), one third of the worldꢀs population is
infected with M. tb, causing 8 million casualties and
1.5 million deaths yearly.[1] Synergy with HIV and the
appearance of M. tb strains that are multi-drug resistant or
hypervirulent, poses further threats.[2]
The search for novel drugs and more effective vaccines
entered a new era with the publication of the genome
sequence of M. tb H37Rv.[3] Based on this sequence, genes
that code for enzymes involved in the critical steps of host–
pathogen interaction were identified. Many of these
enzymes are involved in the synthesis and transport of
Figure 1. PGL-tb1 (1).
complex lipids, in particular phthiocerol dimycocerosates
(DIM[4] or PDIM[5]) present in the outer layer of the M. tb cell
envelope. Furthermore, several M. tb strains synthesize
closely related phenolic glycolipids (PGL-tb1, Figure 1) in
which the phthiocerol is connected to a glycosylated phenol.
It has been shown that DIM/PDIMs are required for multi-
plication and persistence of M. tb in vivo.[6] Next to this, PGL-
tb1 (1, Figure 1) is suspected to be involved in hypervirulence
of specific M. tb strains.[7]
The interplay of M. tb with the human host is very
complex, with PGL-tb1 as one of the most unusual virulence
factors modulating its defense systems and causing disease.
Thus, there is a great need for antigens that permit to
distinguish between prior BCG vaccination and infection.
Recently, an enzyme-linked immunosorbent assay (ELISA)
based on PGL-tb1 has shown potential for the diagnosis of TB
in HIV-infected patients.[8] Furthermore, a lipidomics plat-
form has been established for chemotaxonomic analysis of
M. tb.[9]
Thus, access to pure, chemically synthesized PGL-tb1 (1)
has become crucial for reliable immunological studies.
Preluded by the first synthesis of phthiocerol dimycocerosate
PDIM A[10] we now report the first total synthesis of PGL-
tb1.[11]
The size and complexity of PGL-tb1 is impressive. The
parent phenylphthiocerol has four stereocenters and is
esterified to two molecules of mycocerosic acid (2), a long-
chain quadruple methyl-branched fatty acid. Through the
phenol terminus, the aglycon is linked to a linear trisacchar-
ide.
[*] Dr. S. Barroso, M. P. Baggelaar, D. Geerdink, Dr. B. ter Horst,
Dr. E. Casas-Arce, Prof. A. J. Minnaard
To get PGL-tb1 within reach, a strongly convergent
synthetic strategy had to be designed (see Scheme 1).
Esterification with mycocerosic acid was planned for a late
stage because of its precious nature. Instead of the glyco-
sylation of the phenol terminus with the trisaccharide
construct, we chose to connect the trisaccharide in the form
of a para-iodophenoxy-substituted glycan to a terminal
alkyne through Sonogashira coupling. This strategy avoids
a stereoselective glycosylation step with an activated trisac-
charide in a late stage of the synthesis. The alkyne function
should be fully reduced to the corresponding aliphatic
fragment together with the removal of the benzyl protecting
groups in the final step. An additional advantage is that the
bifunctional “spacer” of the required length, equipped with
a terminal alkyne, would be readily accessible from commer-
cially available alkynol 5 in a few steps.
Stratingh Institute for Chemistry
University of Groningen
Nijenborgh 7, 9747 AG, Groningen (The Netherlands)
E-mail: a.j.minnaard@rug.nl
Homepage: http://www.rug.nl/fmns-research/stratingh
Dr. R. Castelli, Prof. H. S. Overkleeft, Prof. G. A. van der Marel,
Dr. J. D. C. Codꢀe
Leiden Institute of Chemistry, Gorlaeus Laboratories
Leiden University
P.O. Box 9502, Leiden (The Netherlands)
E-mail: jcodee@chem.leidenuniv.nl
Homepage: http://biosyn.lic.leidenuniv.nl/
[**] Financial support from The Netherlands Organization for Scientific
Research (NWO-CW) is acknowledged. S. B. thanks the Generalitat
Valenciana and the Universitat de Valꢁncia (Spain) for a postoctoral
grant (VALi+D program).
The right-hand side of phenylphthiocerol was planned to
be prepared by using our asymmetric conjugate addition–
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201206221.
11774
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11774 –11777

Angewandte
Chemie
introduction of the third stereocenter.
Thus, b-ketoester 16 was treated with
(R)-[{RuCl(tol-binap)}2(m-Cl)3][NH2Me2]
(1 mol%;
tol-binap = 2,2’-bis(di-p-tolyl-
phosphanyl)-1,1’-binaphthyl) and H2 at
20 bar.[18] Gratifyingly, this gave hydroxy-
ester 17 in good yield and more than 99%
de. To finalize the building block, 17 was
transformed into its Weinreb amide 18 by
treatment with AlMe3 and N,O-dimethyl-
hydroxylamine, which gave higher yields
than using the corresponding lithium
amide.
Scheme 1. Retrosynthetic analysis of 1.
alkylation strategy (see below). The advantage of this
approach compared to an asymmetric aldol reaction is that
it avoids extensive functional group manipulation to extend
this fragment to the required length. The subsequent trans-
formation of this building block to the corresponding b-keto
ester, followed by RuII-catalyzed asymmetric hydrogenation,
would provide the right-hand side of phenylphthiocerol,
which is suitable to connect to the alkyne iodide. After
coupling both fragments, the b-hydroxy ketone thus obtained
should then be converted into the 1,3-anti diol.
The target trisaccharide should be accessible in two
distinct glycosylation events between p-iodophenol rhamno-
side 8, rhamnosyl thioglycoside 9, and fucosyl thioglycoside
10. To achieve a-selective glycosylations, a benzoate ester was
chosen as a temporary protecting group for rhamnoside donor
9, while acetate esters were installed on fucosyl donor 10 to
tune its reactivity[12] and to be finally replaced by methyl
ethers, as present in the final product. Although the trisac-
charide portion had been prepared before,[13] we devised
a different approach that centered around the p-iodophenol
moiety, relying on its pivotal role both for the Sonogashira
cross-coupling reaction with alkyne 4 and as anomeric
protection for the reducing end of the growing trisaccharide.
The construction of the two glycosidic bonds profited from
the N-thiophenyl-e-caprolactam/Tf2O combination to acti-
vate the thioglycoside donors.[14] This recently developed
method was chosen for its orthogonality to the allyl ether
protection in the 3-O position of the central rhamnose unit;
which is readily installed by means of a Bu2SnO-mediated
regioselective allylation.[15,16]
Scheme 2. Reagents and conditions: a) Cu(OTf)2 (0.5 mol%), (S,R,R)-
phosphoramidite (1 mol%), Me2Zn, toluene, À258C. Then EtI, HMPA,
08C, 83%, >20:1 trans/cis, 95% ee (for trans); b) m-CPBA, CH2Cl2,
reflux, 60%; c) K2CO3, MeOH, RT; d) NaH, MeI, DMF, RT, 92%;
e) DIBAL-H, Et2O, À848C, 86%; f) ethyl diazoacetate, NbCl5
(5 mol%), CH2Cl2, 86%; g) (R)-[{RuCl(tol-binap)}2(m-Cl)3][NH2Me2]
(1 mol%), 20 bar H2, EtOH, RT, 76%, de>99.5%; h) AlMe3, MeNH-
(OMe)·HCl, THF, 73%. HMPA=hexamethyl phosphoramide, m-
CPBA=m-chloroperbenzoic acid, DIBAL-H=diisobutylaluminum hy-
dride.
Synthesis of the required alkyne iodide 22 started with
commercially available alkynol 5 (Scheme 3). A Zipper
reaction converted 5 into terminal alkyne 19 (61% yield),
which was TMS-protected and transformed into iodide 22 in
two steps.[19]
The construction of the aglycone started with a copper/
phosphoramidite-catalyzed asymmetric conjugate addition of
Me2Zn to 6 (Scheme 2). The resulting enolate was alkylated
in situ with EtI to afford 11 in high yield and excellent
stereoselectivity. A subsequent Baeyer–Villiger oxidation of
11 gave the desired lactone 12 in high regioselectivity though
with a moderate 60% yield. Efforts to improve this yield were
unsuccessful. Lactone 12 was subjected to methanolysis and
subsequent O-methylation of the resulting hydroxyester.
DIBAL-H reduction afforded the desired aldehyde 15.
Efficient transformation of 15 into b-ketoester 16 was
achieved by treatment with ethyl diazoacetate employing
NbCl5 as catalyst.[17] This transformation set the stage for the
Scheme 3. Reagents and conditions: a) NaH, 1,3-diaminopropane,
708C, 61%; b) nBuLi, TMSCl, THF, À408C to RT; c) p-TsCl, pyridine,
CHCl3, RT, 87%; d) NaI, acetone, RT, 86%; e) 22, tBuLi, Et2O, À848C,
2 h, then amide 18, 81%; f) NH4BH(OAc)3, MeCN, AcOH, THF,
À258C, 76%; g) MeOH, K2CO3, RT, quant.; TMS=trimethylsilyl,
p-TsCl=p-toluenesulfonyl chloride.
Angew. Chem. Int. Ed. 2012, 51, 11774 –11777
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11775

.
Angewandte
Communications
The addition of 22, as its corresponding Li reagent, to 18
turned out to be not trivial.[20] Initially, a stepwise procedure
with initial deprotonation of the hydroxy group in 18 by
strong bases suffered from various problems, depending on
the base. Therefore it was finally decided to use an excess of
lithiated 22 both for the deprotonation and the addition.
Extensive optimization showed that using 22 and tBuLi in
a 1:1.6 ratio for the lithiation prevented the formation of an
inseparable tert-butyl ketone and gave the maximum yield of
the desired hydroxyketone 23, a satisfying 81%. Anti-1,3-diol
24 was obtained stereoselectively in good yields using
NH4BH(OAc)3 according to the Evans protocol.[21] Removal
of the TMS group led quantitatively to the required terminal
alkyne 4.
With the required building blocks 2, 7, and 4 in hand, we
could enter the convergent part of the synthesis, depicted in
Scheme 5. A key step, Sonogashira reaction between the
alkyne 4 and iodophenyl trisaccharide 7, gave after optimi-
zation the desired 31 in 51% yield, along with 35% of
recovered starting material. Esterification of the diol moiety
in 31 with mycocerosic acid 2 was accomplished using EDC
and DMAP, to afford diester 32 in a 72% yield.
Synthesis of trisaccharide 7 (Scheme 4) commenced with
connecting rhamnosides 8 and 9 by exposure to an equimolar
amount of N-thiophenyl-e-caprolactam and Tf2O in a TTBP-
buffered solution, without the requirement of preactivation.
Scheme 4. Reagents and conditions: a) 9, N-thiophenyl-e-caprolactam,
TTBP, Tf2O, CH2Cl2, À258C to 208C, 4 h, 71%; b) MeONa, MeOH,
208C, 16 h; c) BnBr, NaH, DMF, 08C to 208C, 2 h, 87% two steps;
d) PdCl2 (5 mol%), MeOH, 208C, 16 h, 82%; e) 10, N-thiophenyl-e-
caprolactam, TTBP, Tf2O, CH2Cl2, À258C to 208C, 4 h, 52%;
f) MeONa, MeOH, 208C, 16 h; g) MeI, NaH, DMF, 08C to 208C, 16 h,
85% two steps. TTBP=2,4,6-tri-tert-butyl pyrimidine,
Tf2O=trifluoromethanesulfonic anhydride.
Scheme 5. Reagents and conditions: a) [PdCl2(PPh3)2] (5 mol%), PPh3
(5 mol%), CuI (10 mol%), Et3N, 408C, 51% (36% 4 recovered);
b) EDC·HCl, DMAP, CDCl3, 72%; c) Pd/C, 1 bar H2, EtOAc, EtOH, RT,
86%. EDC·HCl=N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide hy-
drochloride, DMAP=4-(dimethylamino)pyridine, TMS=trimethylsilyl.
Full stereochemical control to the desired a-configuration
was ensured by the 2-O-benzoate ester in 9. The ester was
then replaced by a benzyl group in two steps. The allyl ether
moiety in disaccharide 27 was then selectively excised by
catalytic PdCl2 in MeOH.[22] The obtained disaccharide
acceptor 28 was welded together with fucose donor 10
under similar conditions as for the previous glycosylation,
yielding trisaccharide 29 in moderate yield and complete a-
selectivity. After catalytic transesterification, the hydroxy
groups of crude 30 were finally methylated by treatment with
an excess of sodium hydride and methyl iodide to furnish 7 in
good yield.
We earlier reported a stereoselective preparation of
mycocerosic acid in 12% yield over 15 steps, relying on the
copper-catalyzed conjugate addition of Grignard reagents to
a,b-unsaturated thioesters.[23] However, the iterative protocol
as optimized for the synthesis of mycolipenic acid was used,
by employing DIBALH for the reduction of thioesters and
Horner–Wadsworth–Emmons instead of Wittig olefina-
tions.[24]
We were very pleased to see that the total synthesis of
1 could be concluded efficiently by full reduction of the triple
bond and hydrogenolysis of the benzyl ethers with H2 and Pd/
C. Full analysis of the resulting compound by NMR spectros-
copy, MS, and optical rotation showed 1 to be identical with
PGL-tb1 isolated from M. tb.[11]
In summary, the first total synthesis of PGL-tb1 has been
accomplished. For mycocerosic acid and the aglycon phthio-
cerol, Cu-catalyzed asymmetric conjugate additions of
MeMgBr and Me2Zn were used as the key approach,
respectively. The synthesis of the iodophenol-functionalized
trisaccharide profited from the fully stereoselective glycosy-
lation with rhamnose and fucose thioglycosides using N-
thiophenyl-e-caprolactam and Tf2O. Connection of the build-
ing blocks by Sonogashira cross-coupling, esterification of the
11776 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11774 –11777

Angewandte
Chemie
[11] For the first isolation and characterization of this compound see:
mycocerosic acid units, and one-step deprotection completed
the synthesis. The so-prepared PGL-tb1 is currently applied in
a recently established lipidomics platform of M. tb and its
availability is expected to advance the development of
efficient diagnostic tools for the detection of hypervirulent
strains of M. tb and contribute to the understanding of the
role of PGL-tb1 in virulence.
a) M. Daffꢁ, C. Lacave, M.-A. Lanꢁelle, G. Lanꢁelle, Eur. J.
Biochem. 1987, 167, 155 – 160; b) M. Daffꢁ, P. Servin, Eur. J.
Biochem. 1989, 185, 157 – 162.
[12] D. Deprꢁ, A. Dꢂffels, L. G. Green, R. Lenz, S. V. Ley, C.-H.
Wong, Chem. Eur. J. 1999, 5, 3326 – 3340.
[13] a) G. H. Veeneman, S. H. Van Leeuwen, H. Zuurmond, J. H.
van Boom, J. Carbohydr. Chem. 1990, 9, 783 – 796; b) T. Fuji-
wara, Agric. Biol. Chem. 1991, 55, 2123 – 2128.
[14] a) S. G. Durꢃn, T. Polat, C. H. Wong, Org. Lett. 2004, 6, 839 –
841; b) E. J. Grayson, G. J. L. Bernardes, J. M. Chalker, O.
Boutureira, J. R. Koeppe, B. G. Davis, Angew. Chem. 2011, 123,
4213 – 4218; Angew. Chem. Int. Ed. 2011, 50, 4127 – 4132.
[15] A detailed synthesis of the monosaccharide building blocks is
reported in the Supporting Information.
[16] F. Guibꢁ, Tetrahedron 1997, 53, 13509 – 13556.
[17] J. S. Yadav, B. V. Subba Reddy, B. Eeshwaraiah, P. N. Reddy,
Tetrahedron 2005, 61, 875 – 878.
[18] a) R. Kramer, R. Brꢂckner, Angew. Chem. 2007, 119, 6657 –
6661; Angew. Chem. Int. Ed. 2007, 46, 6537 – 6541; b) R.
Noyori, T. Ikeda, T. Ohkuma, M. Widhalm, M. Kitamura, H.
Takaya, S. Akutagawa, N. Sayo, T. Saito, T. Taketomi, H.
Kumobayashi, J. Am. Chem. Soc. 1989, 111, 9134 – 9135; c) K.
Mashima, K. H. Kusano, N. Sato, Y. Matsumura, K. Nozaki, H.
Kumobayashi, N. Sayo, Y. Hori, T. Ishizaki, S. Akutagawa, H.
Takaya, J. Org. Chem. 1994, 59, 3064 – 3076.
[19] a) C. A. Brown, A. Yamashita, J. Am. Chem. Soc. 1975, 97, 891 –
892; b) S. R. Abrams, Can. J. Chem. 1984, 62, 1333 – 1334;
c) F. M. Menger, X. Y. Chen, S. Brocchini, H. P. Hopkins, D.
Hamilton, J. Am. Chem. Soc. 1993, 115, 6600 – 6608.
[20] C. Roche, O. Labeeuw, M. Haddad, T. Ayad, J.-P. Genet, V.
Ratovelomanana-Vidal, P. Phansavath, Eur. J. Org. Chem. 2009,
3977 – 3986.
Received: August 3, 2012
Published online: October 19, 2012
Keywords: asymmetric synthesis · carbohydrates · glycolipids ·
.
mycobacterium tuberculosis · total synthesis
[1] a) C. Dye, Lancet 2006, 367, 938 – 940; b) Global tuberculosis
control: WHO report 2011, World Health Organization, link:
http://www.who.int/tb/publications/global_report/en/index.html,
2011.
[2] D. Donoghue, M. Spigelman, C. L. Greenblatt, G. Lev-Maor,
G. K. Bar-Gal, C. Matheson, K. Vernon, A. G. Nerlich, A. R.
Zink, Lancet Infect. Dis. 2004, 4, 584 – 592.
[3] S. T. Cole, R. Brosch, J. Parkhill, T. Garnier, C. Churcher et al.,
Nature 1998, 393, 537 – 544.
[4] L. R. Camacho, D. Ensergueix, E. Perez, B. Gicquel, C. Guilhot,
Mol. Microbiol. 1999, 34, 257 – 267.
[5] J. S. Cox, B. Chen, M. McNeil, W. R. Jacobs, Jr., Nature 1999,
402, 79 – 83.
[6] C. Rousseau, N. Winter, E. Pivert, Y. Bordat, O. Neyrolles, P.
Avꢁ, M. Huerre, B. Gicquel, M. Jackson, Cell. Microbiol. 2004, 6,
277 – 287.
[7] a) M. B. Reed, P. Domenech, C. Manca, H. Su, A. K. Barczak,
B. N. Kreiswirth, G. Kaplan, C. E. Barry III, Nature 2004, 431,
84 – 87; b) D. Sinsimer, G. Huet, C. Manca, L. Tsenova, M.-S.
Kool, N. Kurepina, B. Kana, B. Mathema, S. A. E. Marras, B. N.
Kreiswirth, C. Guilhot, G. Kaplan, Infect. Immun. 2008, 76,
3027 – 3036.
[8] N. Simonney, P. Chavanet, C. Perronne, M. Leportier, F. Revol,
J.-L. Herrmann, P. H. Lagrange, Tuberculosis 2007, 87, 109 – 122.
[9] E. Layre, L. Sweet, S. Hong, C. A. Madigan, D. Desjardins, D. C.
Young, T.-Y. Cheng, J. W. Annand, K. Kim, I. C. Shamputa, M. J.
McConnell, C. A. Debono, S. M. Behar, A. J. Minnaard, M.
Murray, C. E. Barry, I. Matsunaga, D. B. Moody, Chem. Biol.
2011, 18, 1537 – 1549.
[21] a) D. A. Evans, K. T. Chapman, Tetrahedron Lett. 1986, 21,
5939 – 5942; b) D. A. Evans, K. T. Chapman, E. M. Carreira, J.
Am. Chem. Soc. 1988, 110, 3560 – 3578.
[22] a) J. Sun, X. Han, B. Yu, Synlett 2005, 437 – 440; b) N. Minakawa,
Y. Kato, K. Uetake, D. Kaga, A. Matsuda, Tetrahedron 2003, 59,
1699 – 1702; c) T. Ogawa, H. Yamamoto, Agric. Biol. Chem.
1985, 49, 475 – 482.
[23] B. ter Horst, B. L. Feringa, A. J. Minnaard, Chem. Commun.
2007, 489 – 491.
[24] a) B. ter Horst, B. L. Feringa, A. J. Minnaard, Org. Lett. 2007, 9,
3013 – 3015; b) B. ter Horst, J. van Wermeskerken, B. L. Feringa,
A. J. Minnaard, Eur. J. Org. Chem. 2010, 38 – 41.
[10] E. Casas-Arce, B. ter Horst, B. L. Feringa, A. J. Minnaard,
Chem. Eur. J. 2008, 14, 4157 – 4159.
Angew. Chem. Int. Ed. 2012, 51, 11774 –11777
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11777