Synthesis of Mannose-6-Phosphate Analogues and their Utility as Angiogenesis Regulators

Full text of DOI:10.1002/cmdc.201100293

DOI: 10.1002/cmdc.201100293
Synthesis of Mannose-6-Phosphate Analogues and their Utility as
Angiogenesis Regulators
Vꢀronique Barragan-Montero,*[a] Azzam Awwad,[a] Stꢀphanie Combemale,[a] Pascal de Santa Barbara,[b]
Bernard Jover,[c] Jean-Pierre Molꢁs,*[d] and Jean-Louis Montero[a]
Although carbohydrates are the most abundant natural prod-
ucts, their use as therapeutic agents has been limited. Howev-
er, since carbohydrate binding proteins are involved in many
biological processes, including cellular communication,[1–2] the
prospects for carbohydrate-based drugs seem bright. Here, we
provide a synthetic route to bioactive mannose derivatives
that serve as both positive and negative effectors of angiogen-
esis, thereby laying the groundwork for future drug develop-
ment.
partner receptor.[7–9] It binds insulin-like growth factor-2 (IGF-2)
at the cell surface and internalizes this growth factor for degra-
dation inside lyzozomes. The functions of other proteins har-
boring M6P residues are also dependent on their interactions
with this receptor. For example, latent transforming growth
factor-b (TGF-b) is converted into active TGF-b upon M6P inter-
action with the receptor.[7] Leukemia inhibitory factor (LIF), a
growth factor, is cleared from the extracellular medium via re-
ceptor interaction with M6P, and subsequent internalization
prevents an excessive accumulation that is detrimental to
health especially during development.[10] Moreover, proliferin-
dependent angiogenesis also requires CI-M6PR.[11] The obvious
importance of M6P in biological processes, including angio-
genesis, prompted us to evaluate the activity of our new ana-
logues in blood vessel formation. As will be shown, the ana-
logues do indeed have substantial effect on angiogenesis.
The monosaccharides synthesized and examined here in-
clude a variety of substituents at the mannose 6-position:
azido (4), aminomethyl (6), carboxyl (7), malonate (8), sulfonate
(9), carboxymethyl (10), and phosphonate (3). The replacement
of the phosphate head group by other moieties, mostly bioi-
sosteres of phosphate, are intended to provide enzymatically
stable compounds that could be used as tools to better under-
stand the chemical factors involved in the modulation of an-
giogenic activities. In the past, the development of carbohy-
drates for therapeutic purposes has been considered problem-
atic due to the challenges in synthesizing carbohydrate
mimics; however, recent progress in this area has allowed us
to overcome these limitations. We now present a simple and
efficient synthesis of seven M6P analogues using a cyclic sul-
fate intermediate.
The current, limited applications of carbohydrates as thera-
peutics may, in part, be related to the high complexity of inter-
actions between carbohydrate and carbohydrate binding pro-
teins. Carbohydrate oligomers are themselves complex; for ex-
ample, four different monosaccharides can form 35560 distinct
tetrasaccharides—this large number reflects the multiple hy-
droxy attachment sites on each component sugar. Thus, a rela-
tively small polysaccharide has an enormous capacity to
encode biological information. When these polysaccharides are
conjugated to proteins, the complexity further increases. To
date, more than 80 carbohydrate binding proteins have been
identified, and their binding specificities have been described
(or are about to be).[3] Among these proteins, the lectin family
has been extensively studied and classified into subfamilies ac-
cording to their cellular location and their carbohydrate bind-
ing specificities.[4] For example, the P-type lectins recognize
mannose-6-phosphate (M6P), the motivation behind efforts in
the design and synthesis of new M6P analogues.
P-type lectins encompass the 46 kD cation-dependent M6P
receptor (CD-M6PR), the 300 kD cation-independent M6P re-
ceptor (CI-M6PR), and proteins harboring M6P homology do-
mains.[5] One major cellular function of the receptors is to help
cargo M6P-containing proteins between various subcellular
compartments.[6] In addition, CI-M6PR is actually a large multi-
Current routes to M6P analogues exploit activating the C-6
position by a variety of methods, such as halogenation, sulfo-
nation, epoxidation and the use of phosphonium salts fol-
lowed by nucleophilic substitution.[12–19] Homologation of man-
nose has been achieved previously by oxidation of the primary
alcohol followed by a Wittig–Horner reaction on the resultant
aldehyde.[20] Recently, we introduced substitutions at the C-6
position of mannose via a Mitsunobu reaction.[21] These ap-
proaches can be efficient, but a need for a more versatile
method led us to access such compounds via the cyclic sulfate
method previously developed by us[22] to obtain, in just a few
steps with the potential for scale-up, a wide variety of C-6
mannose derivatives (Scheme 1).
[a] Dr. V. Barragan-Montero, A. Awwad, S. Combemale, Prof. J.-L. Montero
Institut des Biomolꢀcules Max Mousseron (IBMM)
UMR 5247 UM2-UM1-CNRS, ENSCM
8 rue de l’Ecole Normale, 34296 Montpellier cedex 5 (France)
E-mail: veronique.montero@univ-montp2.fr
[b] Dr. P. de Santa Barbara
INSERM U1046, UM1
371 avenue du doyen G. Giraud, 34295 Montpellier Cedex 5, (France)
[c] B. Jover
CNRS-FRE3400 IURC, UM1
641 avenue du doyen G. Giraud, 34093 Montpellier Cedex 5 (France)
[d] Dr. J.-P. Molꢁs
Regioselective nucleophilic displacement of cyclic sulfate 2
is the key step to obtain all the described mannose derivatives.
In the case of phosphonate 3, malonate 8, and sulfonate 9, the
corresponding nucleophiles were prepared with the aid of a
strong base such as n-butyllithium or sodium hydride. An
INSERM U1058, UM1, CHU Montpellier
15 avenue Charles Flahault, 34093 Montpellier Cedex 5 (France)
E-mail: jean-pierre.moles@inserm.fr
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201100293.
ChemMedChem 2011, 6, 1771 – 1774
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1771

MED
acidic ion-exchange resin (Amberlyst 15-H+) then led to the re-
moval of the monosulfate and isopropylidene groups. In the
case of nitrile 5, sodium cyanide was used in N,N-
dimethylformamide without adding a base in order to avoid a
premature hydrolysis of the functional group. Starting from ni-
trile 5, hydrolysis or hydrogenolysis led to carboxylate 7 or
aminomethyl derivative 6, respectively. Nucleophilic substitu-
tion of the cyclic sulfate is, therefore, a versatile method that
gives mannose derivatives in good to excellent (quantitative)
yields. Obtaining agents for biological evaluation in such a flex-
ible manner allowed us to unveil the best candidates for con-
trolling angiogenesis.
Angiogenesis is a complex phenomenon that leads to the
formation of new blood vessels from pre-existing ones.[23] This
process is crucial for development and plays a key role in vari-
ous normal and pathological states, including cancer and cardi-
ovascular diseases. Its regulation involves a tuned balance of
proangiogenic and antiangiogenic factors.[24] The relationship
between M6P and angiogenesis appears firmly established,
through both the transforming growth factor (TGF-b) pathway
activation and the proliferin (PLF) signal. TGF-b pathways or li-
gands are thought to have both pro- and antiangiogenic prop-
erties. Low TGF-b levels contribute to an angiogenic switch by
up-regulating angiogenic factors and proteinases. On the other
hand, high TGF-b levels inhibit endothelial cell growth, stimu-
late smooth muscle cell differentiation, and recruitment and
promote basement membrane reformation.[25] Regarding the
proliferin signal, M6P completely blocks PLF-induced angiogen-
esis both in cell culture and in vivo.[11] Finally, the receptor
itself can affect angiogenesis by clearing active plasminogen
through its soluble form, and this has been shown to block
tumor cell invasion in vitro, endothelial cell invasion in vivo,
and tumor growth in vivo.[26]
Scheme 1. Synthesis of mannose-6-phosphate (M6P) analogues. Reagents
and conditions: a) DMP, p-TsOH, acetone; b) H2O, 63%; c) Et3N, CH2Cl2, SOCl2,
08C then NaIO4, H2O, RuCl3, CH2Cl2/CH3CN (1:1), RT, 84%; d) CH3PO(OCH3)2,
1,1-diphenylethylene, nBuLi, HMPT, THF, À788C, quant; e) Amberlyst 15-H+,
MeOH/THF (1:1), quant; f) (CO2H)2CH, NaH, DMF, 30%; g) KOH, THF/H2O
(6:4), 95%; h) CH3SO3iPr, 1,1-diphenylethylene, nBuLi, HMPT, THF, quant;
i) TMSBr, CH2Cl2, 82%; j) NaN3, DMF, quant; k) NaCN, DMF, quant; l) H2O, H2,
Raney Ni, MeOH, 10%; m) H2O2, NaOH, then Amberlite IRC-50-H+, quant;
n) BrCH2CO2Et, 1,1-diphenylethylene, nBuLi, HMPT, THF, 39%; o) pyridine,
DMAP, POCl3, CH2Cl2, 08C, 75%. For abbreviations, see the Experimental
Section.
The M6P analogues presented in Scheme 1 were subjected
to angiogenic assays using two experimental models. The first
of these employed the rat aortic ring assay, an ex vivo angio-
genic model in which our ana-
logues were examined at 10À4
m
Table 1. Cytotoxicity, tumor growth and evaluation of angiogenic effects of M6P analogues.[a]
over 11 days for their ability to
stimulate or inhibit capillary
growth in rat aortic rings.[27–28]
Sunitinib (marketed by Pfizer as
Sutent and formerly known as
SU11248)[29] and endothelial cell
growth supplement (ECGS)[30]
were used as known negative
and positive stimuli, respectively.
The results are shown in Table 1.
Certain derivatives behave as in-
hibitors: MeM6P 11, malonate 8,
and phosphonate 3, while
others are activators: sulfonate 9
and azide 4. Still some deriva-
tives have only slight or no
effect on angiogenesis: carboxyl-
ate 7 and amine 6. Carboxylate
10 exhibited the same effect as
Compd
ARA[b,c]
no. sprouts
ARS[d]
[%]
Cell toxicity[e] [%]
10À4 10À6
m
B16 Tumor growth
vol[f,c] [mm3]
survival [%]
10À2
m
m
Sunitinib
11
9
8
10
6
PBS control
7
4
3
21Æ13
34Æ29
97Æ9
45Æ21
N.D.
58Æ33
88Æ21
71
49Æ2
78Æ5.5
110Æ3.5
80Æ5.5
113Æ4
110Æ2.5
80Æ10
100
123Æ8.5
109Æ4.9
99Æ3
107Æ2.5
101Æ3
101Æ2
103Æ4.5
104Æ1.5
120Æ23
100
N.D.
–
43Æ2
0.79Æ0.69
N.D.
75
–
69Æ3.5
70Æ0.5
73Æ0.5
81Æ2
119Æ3.5
110Æ1.9
101Æ13.5
100
N.D.
–
N.D.
–
N.D.
–
100
2.45Æ0.97
N.D.
50
–
123Æ7.5
125Æ3.5
130Æ3.5
172Æ8.5
119Æ2.5
84Æ9
110Æ2
90Æ5.5
86Æ4.5
N.D.
104Æ1.5
121Æ5
98Æ8.5
N.D.
115 Æ7
56
N.A.
1.5Æ0.5
2.92Æ1.5
N.D.
50
57
–
131Æ7.5
N.D.
ECGS
[a] N.A.: not applicable; N.D.: not done; [b] Aortic ring assay (ARA); [c] data represent the meanÆSD; data
were analyzed by one-way ANOVA or two-way ANOVA for repeated measures when required. Between-group
differences were determined with Student’s t-test. The level of significant difference was set for p <0.05.
[d] Angiogenic relative surface (ARS) was determined using a chorioallantoic membrane (CAM) assay; data rep-
resent the meanÆSD versus the PBS control. [e] Data represent the meanÆSD versus the control. [f] Tumor
volume was determined on day 19. Animal experiments complied with the European and French laws and
with the guiding principles for experimental procedures as set forth in the declaration of Helsinki.
1772
www.chemmedchem.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1771 – 1774

MeM6P (11) itself, with the number and the length of sprouts
being equivalent.
that the azide and carboxylate derivatives, both of which dis-
played lower activities compared with the other test com-
pounds, are not bioisosteric analogues of M6P. It is also note-
worthy that the results of our two assays do not correspond
exactly. During our study, we analyzed two different angiogen-
ic processes, sprouting using aortic ring assay and intussuscep-
tion (splitting) angiogenesis with CAM assays.[32] These two
processes can be differently modulated by M6P analogues.
This is not unusual in the angiogenesis arena and explains why
multiple assays are often needed to portray the efficacy of a
potential drug.
The second biological assay used to evaluate our M6P deriv-
atives was the avian chorioallantoic membrane (CAM) assay.[31]
Control- or M6P-analogue-treated membranes were deposited
on nascent CAM at embryonic day 7 and grown for 4 days
in ovo at 388C. The same positive and negative controls were
included as before. Quantification of the angiogenic response
was carried out by measuring the area of neovascularization
on each particular membrane. These experiments demonstrate
divergent activities of the synthesized compounds (Figure 1).
Finally, compounds showing the most potent an-
giogenic properties, namely MeM6P (11) for inhibi-
tion and phosphonate 3 for activation, as well as the
azido 4 (moderate activity), were tested in a B16 mel-
anoma tumor growth model,[33] and their cytotoxicity
was also evaluated in primary human endothelial cell
cultures. Indeed, the antiangiogenic properties of
some compounds could be the result of specific cell
toxicity.[34] At the three concentrations tested, slight
or no effect on the cell number in primary human en-
dothelial cell cultures was observed after 48 h expo-
sure (Table 1). Mice were injected with B16F1 cells
(day 0). The mice were then divided into groups and
were treated (i.p.) with 300 mgkgÀ1 of test com-
pound three times a week, starting from day 0.
MeM6P (11) showed 79% tumor growth inhibition
and 75% survival at day 19. Azide 4 is also an inhibi-
tor of tumor growth but to a lesser extent (50%). No
effect was observed for phosphonate 3 (Table 1).
In conclusion, we observed that, of the M6P ana-
logues evaluated, some display proangiogenic activi-
ties, while others display antiangiogenic activities.
These latter analogues were tested in a melanoma
B16 tumor growth model, and at least two of them
have been shown to inhibit tumor growth. We have
clearly demonstrated that M6P and its analogues
assist in the control of neoangiogenesis, and these
compounds can be considered as leads for the devel-
opment of a novel class of therapeutics.
We have presented an efficient method for synthe-
Figure 1. Chorioallantoic membrane (CAM) assays performed with mannose-6-phosphate
(M6P) analogues 3, 4, 6–11, the antiangiogenesis inhibitor sunitinib and the angiogenesis
activator, endothelial cell growth supplement (ECGS). Also shown is the phosphate-buf-
fered saline (PBS) control experiment for comparison. The values given represent the
extent of angiogenesis, where the PBS control is defined as 100%.
sizing M6P analogues. This route can be used to de-
velop additional carbohydrate analogues modified at
the C-6 position, thereby allowing access to a large
variety of original carbohydrate mimics. We also in-
vestigated the function of these monocarbohydrates
during angiogenic processes, showing for the first
As before, some M6P derivatives were identified as CAM in-
hibitors: sulfonate 9, malonate 8, amine 6 and MeM6P 11,
while other derivatives behaved as CAM activators: phospho-
nate 3, azide 4 and carboxylate 7. In the inhibitor group, com-
pared to the control, we observed 43% of neovascular vessels
for phosphate 11, 69% for sulfonate 9, 81% for amine 6, 70%
for malonate 8, and 73% for carboxylate 10. In the activator
group, compared to the control, an increase in neovascular
vessels of 123% was observed for azide 7, 125% for carboxyl-
ate 4, and 130% for phosphonate 3. It is worth pointing out
time that monocarbohydrates possess angiogenic activities via
the M6PR with no apparent toxicity. These results open the
possibility for developing angiogenesis regulator carbohy-
drates as anticancer agents (inhibitors) or for the treatment of
cardiovascular disease (activators). It is clear that our prelimina-
ry results are a promising start in the use of carbohydrates as
angiogenic regulators. The in vitro, in ovo, and in vivo assays
have divided the M6P analogues tested into angiogenesis acti-
vators and inhibitors. The mechanism of action of M6P ana-
logues as angiogenesis regulators has still to be elucidated,
ChemMedChem 2011, 6, 1771 – 1774
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1773

MED
[8] J. X. Kang, Y. Li, A. Leaf, Proc. Natl. Acad. Sci. USA 1997, 94, 13671–
13676.
[9] R. J. Wood, M. D. Hulett, J. Biol. Chem. 2008, 283, 4165–4176.
[10] F. Blanchard, L. Duplomb, S. Raher, P. Vusio, B. Hoflack, Y. Jacques, A.
Godard, J. Biol. Chem. 1999, 274, 24685–24693.
but it seems obvious that the M6P receptor is involved. Investi-
gation into the mechanism of M6P analogue action is under-
way now that a family of regulators is in hand.
[11] O. Volpert, D. Jackson, N. Bouk, D. I. H. Linzer, Endocrinology 1996, 137,
3871–3876.
Experimental Section
[12] R. S. Tipson, Adv. Carbohydr. Chem. 1953, 8, 107–215.
[13] D. B. Berkowitz, D. Bhuniya, G. Peris, Tetrahedron Lett. 1999, 40, 1869–
1872 and references cited therein.
[14] Q. Shen, D. G. Sloss, D. B. Berkowitz, Synth. Commun. 1994, 24, 1519–
1530 and references therein.
Supporting Information: Experimental protocols for the synthesis
and biological evaluation of the analogues described here can
be found on the WWW under http://dx.doi.org/10.1002/
cmdc.201100293.
[15] R. Nouguier, C. R. Acad. Sci. Ser. IIc 2000, 3, 373–377.
[16] A. Jeanjean, M. Garcia, A. Leydet, J.-L. Montero, A. Morere, Bioorg. Med.
Chem. 2006, 14, 3575–3582.
[17] S. Takahashi, H. Kuzuhara, M. Nakajimab, Tetrahedron 2001, 57, 6915–
6926.
[18] N. A. Hughes, N. D. Todhunter, Carbohydr. Res. 2000, 326, 81–87.
[19] B. Castro in Organic Reactions, Volume 29, 1st ed., (Ed.: W. G. Dauben),
John Wiley & Sons, New York, 1983, pp. 1–162.
[20] V. Barragan, F. M. Menger, K. Caran, C. Vidil, A. Morere, J.-L. Montero,
Chem. Commun. 2001, 85–86.
Abbreviations: para-Toluenesulfonic acid (p-TsOH); 2,2-dimethoxy-
propane (DMP); 4-dimethylaminopyridine (DMAP); N,N-dimethyl-
formamide (DMF); hexamethylphosphorous triamide (HMPT); tetra-
hydrofuran (THF); trimethylsilyl bromide (TMSBr).
Acknowledgements
V.B.M. is grateful to the Mission de la Recherche et de la Techno-
logie (MRT), France for scholarships to S.C. and A.A., and thanks
Dr. Frꢀdꢀric Geniet (Laboratoire de physique thꢀorique et astro-
particules, Universitꢀ Montpellier 2, France) for helpful discus-
sions. J.P.M. thanks Anne Tesniere (Inserm U1058, Universitꢀ
Montpellier 1, France) for technical help and the Sociꢀtꢀ de Re-
cherche en Dermatologie for its financial support. P.d.S.B. is sup-
ported by the French Agence Nationale pour la Recherche (ANR-
07-JCJC-0112).
[21] C. Clavel, V. Barragan-Montero, J.-L. Montero, Tetrahedron Lett. 2004, 45,
7465–7467.
[22] N. A. Khanjin, J.-L. Montero, Tetrahedron Lett. 2002, 43, 4017–4020.
[23] J. Folkman, N. Engl. J. Med. 1971, 285, 1182–1186.
[24] W. Risau, Nature 1997, 386, 671–674.
[25] P. Carmeliet, Nat. Med. 2003, 9, 653–660.
[26] V. Leksa, R. Loewe, B. Binder, H. B. Schiller, P. Eckerstorfer, F. Forster, A.
Soler-Cardona, G. Ondrovicovꢅ, E. Kutejovꢅ, E. Steinhuber, J. Breuss, J.
Drach, P. Petzelbauer, B. R. Binder, H. Stockinger, Circ. Res. 2011, 108,
676–685.
[27] R. F. Nicosia, A. Ottinetti, Lab. Invest. 1990, 63, 115–122.
[28] R. F. Nicosia, Y. J. Lin, D. Hazelton, X. Qian, Am. J. Pathol. 1997, 151,
1379–1386.
Keywords: angiogenesis · carbohydrates · chemotherapy ·
glycobiology · mannose-6-phosphates
[29] K. J. Gotink, H. M. Verheul, Angiogenesis 2010, 13, 1–14.
[30] J. Folkman, C. Haudenschild, Nature 1980, 288, 551–556.
[31] J. Hasan, S. D. Schnyder, M. Bibby, J. A. Double, R. Bicknel, G. C. Jayson,
Angiogenesis 2004, 7, 1–16.
[32] C. A. Staton, M. W. R. Reed, N. J. Brown, Int. J. Exp. Pathol. 2009, 90,
195–221.
[33] K. D. Johnstone, T. Karoli, L. Liu, K. Dredge, E. Copeman, C. P. Li, K. Davis,
E. Hammond, I. Bytheway, E. Kostewicz, F. C. Chiu, D. M. Shackleford,
S. A. Charman, W. N. Charman, J. Harenberg, T. J. Gonda, V. Ferro, J. Med.
Chem. 2010, 53, 1686–1699.
[34] C. Cavallini, M. Trettene, M. Degan, P. Delva, B. Molesini, P. Minuz, T. Pan-
dolfini, Br. J. Pharmacol. 2011, 162, 1261–1273.
[1] N. Sharon, Biochem. Soc. Trans. 2008, 36, 1457–1460.
[2] H. Rꢃdiger, H.-C. Siebert, D. Solꢄs, J. Jimꢀnez-Barbero, A. Romero, C. W.
von der Lieth, T. Diaz-MauriÇo, H. J. Gabius, Curr. Med. Chem. 2000, 7,
389–416.
[3] B. Ernst, J. L. Magnani, Nat. Rev. Drug Discovery 2009, 8, 661–677.
[4] D. C. Kilpatrick, Biochim. Biophys. Acta Gen. Subjects 2002, 1572, 187–
197.
[5] K. Von Figura, A. Hasilik, Annu. Rev. Biochem. 1986, 55, 167–193.
[6] P. Ghosh, N. M. Dahms, S. Kornfeld, Nat. Rev. Mol. Cell Biol. 2003, 4, 202–
212.
[7] S. Godꢅr, V. Horejsi, U. H. Weidle, B. R. Binder, C. Hansmann, H. Stocking-
er, Eur. J. Immunol. 1999, 29, 1004–1013.
Received: June 15, 2011
Published online on July 26, 2011
1774
www.chemmedchem.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1771 – 1774