Novel highly potent, structurally simple γ-trifluoromethyl γ-sulfone hydroxamate inhibitor of stromelysin-1 (MMP-3)

Article abstract of DOI:10.1016/j.tetlet.2005.02.063

The γ-trifluoromethyl γ-sulfone hydroxamate 1 was synthesized both in racemic and enantiomerically pure forms by means of a thia-Michael reaction of p-methoxythiophenol on achiral and chiral 3,3,3-trifluorocrotonoyl Michael acceptors. The (R)-1 enantiomer

Full text of DOI:10.1016/j.tetlet.2005.02.063

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 2393–2396
Novel highly potent, structurally simple c-trifluoromethyl c-sulfone
hydroxamate inhibitor of stromelysin-1 (MMP-3)
Monica Sani,a Gabriele Candiani,b Franc¸oise Pecker,b
Luciana Malpezzia and Matteo Zandaa,*
aDipartimento di Chimica, Materiali ed Ingegneria Chimica ÔG. NattaÕ del Politecnico di Milano and CNR—Istituto di Chimica del
Riconoscimento Molecolare, Sezione ÔA. QuilicoÕ, via Mancinelli 7, I-20131 Milano, Italy
` ´
Inserm U581, avenue du Marechal de Lattre de Tassigny, F-94010 Creteil, France
b
Received 4 January2005; revised 9 February2005; accepted 11 February2005
Abstract—The c-trifluoromethyl c-sulfone hydroxamate 1 was synthesized both in racemic and enantiomerically pure forms by
means of a thia-Michael reaction of p-methoxythiophenol on achiral and chiral 3,3,3-trifluorocrotonoyl Michael acceptors. The
(R)-1 enantiomer was the most potent inhibitor of MMP-3 (stromelysin-1), showing an IC50 = 3.2 nM, as well as the most selective
with respect to MMP-9 (65-fold).
Ó 2005 Elsevier Ltd. All rights reserved.
Matrix metalloproteinases (MMPs) are a familyof zinc
metallo-endopeptidases secreted bycells, which are
responsible for much of the turnover of matrix
components.1
A common feature of manyserious diseases, such as
heart failure and cancer, is that MMPs have been re-
ported to playa key-role, in combination with their nat-
Figure 1.
ural tissue inhibitors (TIMPs). Progression and growth
of both pathologies, mainlyin the earlystages, is
favoured bythe proteolytic activityof several MMPs,
such as stromelysin-1 (MMP-3) and gelatinase-B
(MMP-9). For these reasons inhibition of MMPs is
activelystudied as a promising therapeutic target for
heart failure and cancer therapy.2
The R side chain was found to be critical not onlyfor
potency, but it could also dramatically influence the en-
zyme selectivity profile of the inhibition.
Within the frame of a research project aimed at a better
understanding and rationalization of the Ôfluorine-effectÕ
in peptidomimetic structures, particularlythose dis-
playing activity as protease inhibitors,5 we decided to
investigate the effect of a trifluoromethyl (Tfm) group,
positioned as R substituent of structures A, on the inhib-
itorypotencyof MMPs. It is in fact widelyaccepted that
the Tfm group is at the same time highlyhydrophobic
Some years ago Groneberg, Salvino, et al.3 and Freskos
et al.4 described as structurallyverysimple class of
hydroxamate inhibitors bearing an arylsulfone moiety
in c-position, such as A (Fig. 1), which showed nanomo-
lar inhibitorypotencyagainst MMP-2, 3 and 13.
6
and stericallydemanding. Moreover its high electron-
densitycould provide additional interactions within the
MMPs active sites, possiblyincluding hydrogen bond-
ing.7 This choice was encouraged bythe observation that
compounds A bearing large hydrophobic groups R (such
as alkyl, cycloalkyl and arylalkyl groups) showed low
nanomolar, and even subnanomolar affinityfor
Keywords: Trifluoromethyl; Matrix metalloproteinases; Inhibitors;
Asymmetric synthesis.
*
Corresponding author. Tel.: +39 02 23993084; fax: +39 02
23993080; e-mail: matteo.zanda@polimi.it
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.02.063

2394
M. Sani et al. / Tetrahedron Letters 46 (2005) 2393–2396
Scheme 1.
MMP-2, 3 and 13, and excellent selectivitywith respect
to MMP-1.3,4
carboxylic acid 7.9 Coupling with O-Bn hydroxylamine
delivered the O-Bn hydroxamate 8 in satisfactoryyield,
and the subsequent oxidation to the sulfone 9 took place
in nearly quantitative yield. Hydrogenolysis with the
PearlmanÕs catalyst afforded the target racemic hydrox-
amic acid 1 in good overall yields.10
In this communication we describe the synthesis of the
racemic target structure 1, as well as a route to both
its pure enantiomers, and the determination of their
inhibitoryactivitytowards MMP-3 and MMP-9.
Next, we turned our attention to the synthesis of (R) and
(S)-1. It was decided to explore the use of both an oxa-
zolidin-2-one11 10 (Scheme 3) and an oxazolidine-2-thi-
one12 11 as chiral auxiliaries, since theyare known to
feature a remarkablydifferent chemistryat the cleavage
stage (the oxazolidin-2-one is more chemicallyrobust
but more resistant to exocyclic cleavage).13
First attempts to synthesize racemic 1 were disappoint-
ing. Thia-Michael addition of thiol 3 (Scheme 1) to ethyl
3,3,3-trifluorocrotonate 2 occurred efficientlyproviding
4 in good yields.8 Then, oxidation to the sulfone 5 oc-
curred nearlyquantitatively. However, the a-Tfm centre
of 5 proved to be rather labile at basic pH, owing to the
strong acidityof the proton in a-position to the strongly
electronwithdrawing Tfm and sulfone moieties. There-
fore, attempts to perform an alkaline hydrolysis of the
ester function of 5 invariablyproduced intractable mix-
tures of by-products.
Acylation of lithiated 10 and 11 with 3,3,3-trifluoro-
crotonoyl chloride 12 occurred smoothlyaffording the
Michael acceptors 13 and 14, respectively.14 Thia-Michael
addition of thiophenol 3 delivered in quantitative yields
the oxazolidine-2-one 15 and the oxazolidine-2-thione
16, in both cases as nearlyequimolar mixtures of dia-
stereomers, which could be obtained in stereo- and che-
micallypure form byflash chromatography. No attempts
were made to improve the stereocontrol, as both the
epimeric forms were needed for biological evaluation.
The stereochemistryof ( S,S)-15 was assessed byX-ray
diffraction of suitable single crystals (Fig. 2).15
Further attempts to hydrolyze the sulfide 4 under the
same conditions gave rise to a retro-Michael reaction.
We therefore decided to use 3,3,3-trifluorocrotonic acid 6
(Scheme 2) as starting material, that would allow to skip
the ÔdifficultÕ hydrolysis step. The thia-Michael addition
of 3 to 6 occurred in reasonable yields, delivering the
Scheme 2.
Scheme 3.

M. Sani et al. / Tetrahedron Letters 46 (2005) 2393–2396
2395
The diastereomer (S,S)-15 was subjected to exocyclic
cleavage of the oxazolidin-2-one auxiliary, affording
the carboxylic acid (S)-7 (Scheme 4) in fair yields.16
Coupling with O-Bn hydroxylamine occurred efficiently,
and the resulting hydroxamate (S)-8 was quantitatively
oxidized to the sulfone (S)-9. The enantiomericallypure
hydroxamic acid (+)-(S)-1 was obtained in good yields
by hydrogenolysis. An identical synthetic sequence was
performed from (S,R)-15 to provide (ꢀ)-(R)-1.17
could be then converted into the final hydroxamate
(S)-1 through the identical sequence described above.
Analogously, the epimer (S,R)-16 was converted in com-
parable yields into (R)-8, and then into the enantiomer
(ꢀ)-(R)-1.
With racemic 1, as well as both the single enantiomers in
hand, we next addressed the inhibition tests on the cat-
alytic domains of MMP-3 and MMP-9.18
In order to improve the synthetic process, and particu-
larlythe auxiliarycleavage step, we next focused our
attention on the oxazolidine-2-thione intermediate
(S,S)-16 (Scheme 5). Satisfactorily, we found that the
hydroxamate (S)-8 could be obtained directly, in reason-
able yields, by reaction with O-Bn-hydroxylamine, that
The (R) enantiomer of 1 is the most potent compound
(Table 1), but it is worth noting that all of them are sin-
gle digit nanomolar inhibitors of MMP-3. Moderate
selectivitywas observed with respect to MMP-9, in par-
ticular (R)-1 that was ca. 65-fold selective. Interestingly,
both the pure enantiomers of 1 showed better inhibitory
potencythan the racemic compound. The reasons for
this surprising finding are presentlyunder investiga-
tion.19 It is apparent however that the inhibitorypo-
tencyof 1 is independent of its stereochemistry. This
could be the result of an easyinterchange of position
of the Tfm and sulfone moieties in two different enzyme
pockets, most likelyS 01 and S02.20 These findings also sug-
gest that judicious introduction of a Tfm group, and
possiblyof other fluoroalkly groups (such as CF 2Cl,
C2F5, etc.), onto the backbone of protease inhibitors
could represent a successful strategyin order to improve
and modulate the inhibitoryactivity.
We are currentlyinvestigating the snythesis and the
inhibitoryactivityof other fluoroalkyl analogues of
1
Table 1. Inhibition tests
Compound
IC50/MMP-3 (nM)
IC50/MMP-9 (nM)
(R)-1
(S)-1
Racemic 1
3.2
4.3
8.4
209.3
51.9
515.9
Figure 2. ORTEP view of 4-benzyl-3-[4,4,4-trifluoro-3-(4-methoxy-
phenylsulfanyl)-butyryl]-oxazolidin-2-one (S,S)-15.
Scheme 4.
Scheme 5.

2396
M. Sani et al. / Tetrahedron Letters 46 (2005) 2393–2396
CDCl3) d: ꢀ71.6 (d, 3F, J = 8.9 Hz); MS (DIS EI 70 eV)
m/z (%): 280 [M+] (15), 260 (20), 139 (100).
on a wider range of MMPs both in vitro and in vivo, in
order to have a more complete picture of the effect of
fluorine in this particular class of protease inhibitors.
10. To a stirred solution of 9 (0.41 mmol) in a MeOH/AcOEt
4:1 mixture (5 mL), a catalytic amount of Pd(OH)2/C was
added and the reaction mixture was kept vigorously
stirred under hydrogen atmosphere at rt for 1 h. The
palladium powder was filtered over Celite pad, washing
the dark solid with MeOH. The solvent was removed in
vacuo and the residue was purified byflash chromato-
Acknowledgements
We thank the European Commission (IHP Network
grant ÔFLUOR MMPIÕ HPRN-CT-2002-00181), MIUR
(Cofin 2004, Project ÔPolipeptidi Bioattivi e Nano-
strutturatiÕ), Politecnico di Milano and CNR for eco-
nomic support.
graphy(CHCl /MeOH, 95:5), affording 1 (74% yield): Rf:
3
1
0.36 (CHCl3/MeOH, 9:1); H NMR (400 MHz, CD3OD)
d: 7.86 (d, 2H, J = 8.9 Hz), 7.14 (d, 2H, J = 8.9 Hz), 4.62
(m, 1H), 3.89 (s, 3H), 2.92 (dd, 1H, J = 16.2 and 5.9 Hz),
2.63 (dd, 1H, J = 16.2 and 6.1 Hz); 13C NMR (400 MHz,
CDCl3, 305 K) d: 169.3, 168.9, 135.1, 132.9, 127.1 (q,
J = 279.1 Hz), 118.4, 66.7 (q, J = 28.1 Hz), 59.0, 31.1; 19F
NMR (235.3 MHz, CDCl3) d: ꢀ64.1 (d, 3F, J = 8.2 Hz);
MS (DIS EI 70 eV) m/z (%): 327 [M+] (3), 295 (12), 155
(100).
References and notes
1. (a) Whittaker, M.; Floyd, C. D.; Brown, P.; Gearing, A. J.
H. Chem. Rev. 1999, 99, 2735–2776; (b) Bode, W.; Huber,
R. Biochim. Biophys. Acta 2000, 1477, 241–252; (c)
Giavazzi, R.; Taraboletti, G. Crit. Rev. Oncol. Hemat.
2001, 37, 53–60.
11. Evans, D. A.; Gage, J. R.; Leighton, J. L.; Kim, A. S. J.
Org. Chem. 1992, 57, 1961–1963, and references cited
therein.
2. For a critical overview of the status and perspectives of the
clinical use of MMPs: Coussens, L. M.; Fingleton, B.;
Matrisian, L. M. Science 2002, 295, 2387–2392.
3. (a) Groneberg, R. D.; Burns, C. J.; Morrissette, M. M.;
Ullrich, J. W.; Morris, R. L.; Darnbrough, S.; Djuric, S.
W.; Condon, S. M.; McGeehan, G. M.; Labaudiniere, R.;
Neuenschwander, K.; Scotese, A. C.; Kline, J. A. J. Med.
Chem. 1999, 42, 541–544; (b) Salvino, J. M.; Mathew, R.;
Kiesow, T.; Narensingh, R.; Mason, H. J.; Dodd, A.;
Groneberg, R.; Burns, C. J.; McGeehan, G.; Kline, J.;
Orton, E.; Tang, S.-Y.; Morrissette, M.; Labaudiniere, R.
Bioorg. Med. Chem. Lett. 2000, 10, 1637–1640.
12. (a) Crimmins, M. T.; King, B. W.; Tabet, E. A.;
Chaudhary, K. J. Org. Chem. 2001, 66, 894–902; (b)
Crimmins, M. T.; McDougall, P. J. Org. Lett. 2003, 5,
591–594, and references cited therein; (c) Palomo, C.;
´
Oiarbide, M.; Garcıa, J. M. Chem. Eur. J. 2002, 8, 37–44.
13. Sani, M.; Belotti, D.; Giavazzi, R.; Panzeri, W.; Volon-
terio, A.; Zanda, M. Tetrahedron Lett. 2004, 45, 1611–
1615, and references cited therein.
14. Under these conditions no intramolecular thia-Michael
addition byaction of the oxazolidinethione sulfur atom
was observed. For an overview of this process see:
´
Palomo, C.; Oiarbide, M.; Dias, F.; Lopez, R.; Linden,
4. Freskos, J. N.; Mischke, B. V.; DeCrescenzo, G. A.;
Heintz, R.; Getman, D. P.; Howard, S. C.; Kishore, N. N.;
McDonald, J. J.; Munie, G. E.; Rangwala, S.; Swearingen,
C. A.; Voliva, C.; Welsch, D. J. Bioorg. Med. Chem. Lett.
1999, 9, 943–948.
5. For a review see: Zanda, M. New J. Chem. 2004, 28, 1401–
1411.
6. Mikami, K.; Itoh, Y.; Yamanaka, M. Chem. Rev. 2004,
104, 1–16.
7. For a discussion of organic fluorine as hydron-bond
acceptor, see: (a) Carosati, E.; Sciabola, S.; Cruciani, G.
J. Med. Chem. 2004, 47, 5114–5125; (b) Dunitz, J. D.
ChemBioChem 2004, 5, 614–621.
8. For a recent example of thia-Michael reaction with
trifluorocrotonic esters, see: Karimova, N. M.; Glazkov,
A. A.; Ignatenko, A. V.; Kolomiets, A. F. Russ. Chem.
Bull. Int. Ed. 2003, 52, 1621–1622.
9. To a stirred solution of 6 (5.25 mmol) in DCM (10 mL)
thiol 3 (3.5 mmol) and Et3N (8.75 mmol) were added.
After stirring for 24 h the reaction mixture was washed
with a 1 M solution of HCl and then with a 5% aqueous
solution of NaHCO3. The aqueous solution was acidified
with a 1 M solution of HCl until pH 1–2 was reached and
then extracted with AcOEt. The organic layer was dried
over anhydrous Na2SO4 and the solvent was removed
A. Angew. Chem., Int. Ed. 2004, 53, 3307–3310.
15. The absolute configuration of (S,S)-15 was confirmed on
the basis of the Flack parameter (x = 0.0329 (0.0284) for
the selected configuration and x = 1.0281 (0.0380) for the
opposite one): Flack, H. D. Acta Crystallogr. 1983, A39,
876–881, Full data (excluding structure factors) of the
crystal structure have been deposited with Cambridge
Crystallographic Data Centre as supplementary publica-
tion no. 263271. Copies of the data can be obtained, free
of charge, on application to CCDC, 12 Union Road, Cam-
bridge, CB2 1EZ, UK [fax: +44(0)-1223-336033 or e-mail:
deposit@ccdc.cam.ac.uk].
16. Although the cleavage reaction of 15 was conducted under
strictlycontrolled conditions, it is likelythat competitive
oxidation of the sulfur atom byaction of hdyrogen
peroxide is responsible for the rather modest yields.
17. The following values were recorded bypolarimetric
analysis: (R)-1: ½aŠ
23
D
ꢀ16.9 (c 0.88, EtOH); (S)-1:
23
D
½aŠ þ16:1 (c 1.1, EtOH).
18. The catalytic domains of MMP-3 and MMP-9 enzymes
were produced in E. coli, transfected with cDNAs corre-
sponding to the respective human sequences. Proteins
were purified byaffinitychromatographyand the inhib-
itorypotencies of racemic 1 and single enantiomers were
assayed with synthetic, general MMP fluorescent substrate
(Mca-PLGLDpaAR, Tebu-bio) using a FL600 Avantec
fluorimeter. For MMP-3 see: (a) Ye, Q. Z.; Johnson, L.
L.; Hupe, D. J.; Baragi, V. Biochemistry 1992, 31, 11231–
11235; For MMP-9 see: (b) Ye, Q. Z.; Johnson, L. L.; Yu,
A. E.; Hupe, D. Biochemistry 1995, 34, 4702–4708.
19. The tests were reproduciblyrepeated several times on
chemicallypure compounds, therefore this result is
unlikelyto be an artifact.
in vacuo, affording pure 7 (64% yield): Rf: 0.40 (CHCl3/
23
D
MeOH, 9:1); ½aŠ (R enantiomer) ꢀ5.6 (c 0.78, CHCl3);
23
D
½aŠ (S enantiomer) +5.1 (c 1.1, CHCl3); 1H NMR
(400 MHz, CDCl3) d: 10.7–9.20 (verybr s, 1H), 7.53 (d,
2H, J = 8.5 Hz), 6.88 (d, 2H, J = 8.5 Hz), 3.81 (s, 3H), 3.77
(m, 1H), 2.94 (dd, 1H, J = 17.0 and 3.8 Hz), 2.69 (dd, 1H,
J = 17.0 and 10.6 Hz); 13C NMR (400 MHz, CDCl3) d:
175.4, 160.8, 137.1, 126.8 (q, J = 276.3 Hz), 121.6, 114.8,
55.3, 48.5 (q, J = 29.7 Hz), 34.1; 19F NMR (235.3 MHz,
20. We thank one of the referees for this suggestion.