Silanediol-based inhibitor of thermolysin

Article abstract of DOI:10.1016/S0960-894X(02)00804-1

The first silanediol inhibitor of thermolysin is reported, prepared by analogy with the Grobelny/Bartlett phosphinate inhibitor. A Cbz group on nitrogen proved to be unstable to the triflic acid mediated silanediol deprotection and was replaced with a dihydrocinnamoyl group. The silanediol was prepared in high purity by hydrolysis of a difluorosilane intermediate and proved to be an effective inhibitor, differing from the phosphinate by a factor of 4 (K_i=41 nM).

Full text of DOI:10.1016/S0960-894X(02)00804-1

Bioorganic & Medicinal Chemistry Letters 12 (2002) 3625–3627
Silanediol-Based Inhibitor of Thermolysiny
Jaeseung Kim,a Athanasios Glekasa and Scott McN. Sieburtha,b,
*
aDepartment of Chemistry, State University of New York at Stony Brook, Stony Brook, NY 11794-3400, USA
bDepartment of Chemistry, Temple University, 1901 N. 13th St, Philadelphia, PA19122, USA
Received 7 August 2002; accepted 26 August 2002
Abstract—The first silanediol inhibitor of thermolysin is reported, prepared by analogy with the Grobelny/Bartlett phosphinate
inhibitor. A Cbz group on nitrogen proved to be unstable to the triflic acid mediated silanediol deprotection and was replaced with
a dihydrocinnamoyl group. The silanediol was prepared in high purity by hydrolysis of a difluorosilane intermediate and proved to
be an effective inhibitor, differing from the phosphinate by a factor of 4 (Ki=41 nM).
# 2002 Elsevier Science Ltd. All rights reserved.
Thermolysin is the archetype metalloprotease and has
been one of the most intensively studied examples.2 In
the accepted mechanism for peptide hydrolysis, the
active site zinc activates the scissile amide carbonyl,
delivers water, and stabilizes the resulting tetrahedral
intermediate.3
ating the stability and utility of these new structures.
Compound 1 would provide a nearly ideal comparison
with silicon-based inhibitor 2 because the center of the
tetrahedral transition state analogues of 1 and 2 are
both second row elements. We describe here chemistry
directed toward the synthesis of 2.
Several analogous metalloproteases are of significant
pharmaceutical importance,4 including angiotensin-
converting enzyme (ACE),5 and the matrix metallopro-
teases.6 Additional metalloprotease targets of note
include TNF-a converting enzyme,7,8 anthrax lethal
factor,9 and the botulinum toxin.10
A straightforward construction of 2 was expected to
involve a protected version of 4 (Fig. 2), which could be
generated from the commercially available chloromethyl-
trichlorosilane 5 and enantiomerically pure lithium
reagent 6. A diphenylsilane (4) was chosen as a silanediol
Silanediols are new transition state analogue inhibitors,
effective with both aspartic11 and metalloproteases.12,13
Silanediol 3 (Fig. 1) has recently been described as an
ACE inhibitor.12,13 As a vehicle for comparison and
understanding of protease inhibitors, thermolysin pro-
vides a useful benchmark. The Grobelny/Bartlett phos-
phinate 114,15 is
a low nanomolar inhibitor of
thermolysin and an attractive starting point for design
of a silanediol-based inhibitor. Silanediols are best
known for their self-condensation, forming silicone
polymers (siloxanes),16 but silanol polymerization can
be inhibited by steric effects. In comparison with 3 and
the silanediol HIV protease inhibitors,11 the steric
environment around the silanediol in 2 is less hindering,
providing a new level of silanediol exposure for evalu-
Figure 1. Phosphinate inhibitor of thermolysin 1, silanediol analogue
2, and ACE inhibitor 3.
*Corresponding author. Fax.: +1-631-632-882; e-mail: sieburth@
astro.temple.edu
ySee ref 1.
Figure 2. Synthesis of 2 via diphenylsilane 4, derived from commer-
cially available 5 and enantiomerically pure lithium reagent 6.
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00804-1

3626
J. Kim et al. / Bioorg. Med. Chem. Lett. 12 (2002) 3625–3627
Scheme 1. Synthesis of silanediol 14.
precursor based on the successful use of this strategy for
preparation of 3. The phenyl–silicon bond is stable to a
wide range of chemical transformations, yet hydrolysis to
a silanediol was readily achieved within the tripeptide
analogue structure 3 under strongly acidic conditions.13
what appeared to be adequate precedent for the stability
of a benzyl carbamates to treatment with triflic acid,23 a
survey of conditions for triflic acid-mediated hydrolysis
of the phenyl–silicon bonds in 12a convinced us that it
could not be accomplished without loss of this Cbz
group. We therefore turned our attention to 12b, sub-
stituting dihydrocinnamoyl for benzyloxycarbonyl.
To prepare 2, chloromethyltrichlorosilane (5) was coupled
with two equivalents of phenylmagnesium chloride, fol-
lowed by treatment of the resulting product with aqueous
HF17 (Scheme 1). The resulting fluorosilane (7) was iso-
lated by distillation in 65% overall yield. Fluorosilanes
were shown by Eaborn to be moisture stable reagents that
undergo efficient coupling with organometallics.17
Inspection of the crystal structure of inhibitor 1 bound to
the thermolysin active site24 shows that the Cbz benzyl
group makes little contact with the enzyme, and there-
fore substitution of phenethyl 14 for benzyloxy 2 would
be expected to have little, if any, effect on the binding.
Lithium reagent 6 was prepared from the corresponding
iodide 17 (Scheme 2) by metal-halogen exchange,18,19
which in turn was prepared using Evans’ asymmetric
alkylation chemistry.20
Treatment of 12b with an excess of triflic acid in meth-
ylene chloride at 0 ꢀC, followed by addition of ammo-
nium hydroxide, led to the removal of the tert-butyl
ester and both phenyl groups on silicon. The addition of
ammonium hydroxide serves to neutralize the excess
triflic acid and to ensure that any bonds between the
amides and the silicon that may have formed during the
silicon-phenyl bond cleavage had been hydrolyzed.11,12
Addition of 6 to fluorosilane 7 gave silane 8 in 74%
yield (Scheme 1). Gabriel synthesis was used to convert
the chloride of 8 to a protected amine. Treatment of 8
with potassium phthalimide in DMF, was followed by
three equivalents of boron tribromide to cleave the
benzyl ether, forming alcohol 9. Notably, the potential
reaction of the phenyl-silicon bond with this Lewis acid
was not observed.21,22 Oxidation of the alcohol to the
acid 10 was performed in two stages, TPAP followed by
sodium chlorite. Coupling of this acid with the tert-
butyl ester of l-leucine gave the amide as a single dia-
stereomer, confirming that epimerization of the stereo-
genic center of 10 had not occurred.
The silanediol product from this hydrolysis could be iso-
lated directly, but was most cleanly generated by conver-
sion of the product to a crystalline difluorosilane 13 that
precipitated when the silanediol was stirred with aqueous
HF. It is likely that the structure of 13 involves a hyper-
valent silicon.25,26 When 13 was taken up in water and
treated with three equivalents of sodium hydroxide,17 the
diastereotopic fluorines on silicon (19F NMR in acetone-
d6, d ꢁ119 and ꢁ124 ppm, J=20Hz,) rapidly hydrolyzed
to give silanediol 14 that was spectroscopically pure.
Hydrazinolysis removed the phthalimide group to give
amine 11 in high yield and without epimerization. This
amine was coupled with benzyl chloroformate. Despite
Inhibition of thermolysin (Sigma) by 14 was evaluated
using the substrate N-[3-(2-furyl)acryoyl]glycyl-l-leucin-
amide (FAGLA, Sigma) following the method of
Feder.27 Silanediol 14 showed competitive binding and
a Ki=41 nM. This value is only slightly higher than that
reported for phosphinate 1, with a Ki=11 nM.15
The inhibition of this metalloprotease by a neutral spe-
cies not known for its ability to coordinate metals (the
silanediol) could be viewed as surprising. Eliminating
Scheme 2. Preparation of enantiomerically pure iodide 17.

J. Kim et al. / Bioorg. Med. Chem. Lett. 12 (2002) 3625–3627
2. Matthews, B. W. Acc. Chem. Res. 1988, 21, 333.
3627
the negative charge associated with phosphinate 1, a
group that binds to the positively charged active site
zinc, might be expected to result in a lower affinity for
the enzyme. The effect of the acidity and related attri-
butes of the phosphorus has received considerable
attention, but all of the examples are negatively
charged.14,15,28 Notably, charged species carry solvation
that must be displaced during a binding event.
3. Antonczak, S.; Monard, G.; Lopez, M. R.; Rivail, J. L. J.
Mol. Model 2000, 6, 527.
4. Leung, D.; Abbenante, G.; Fairlie, D. P. J. Med. Chem.
2000, 43, 305.
5. Topol, E. Lancet 1999, 354, 1797.
6. Foda, H. D.; Zucker, S. Drug Discov. Today 2001, 6, 478.
7. Letavic, M. A.; Axt, M. Z.; Barberia, J. T.; Carty, T. J.;
Danley, D. E.; Geoghegan, K. F.; Halim, N. S.; Hoth, L. R.;
Kamath, A. V.; Laird, E. R.; Lopresti-Morrow, L. L.;
McClure, K. F.; Mitchell, P. G.; Natarajan, V.; Noe, M. C.;
Pandit, J.; Reeves, L.; Schulte, G. K.; Snow, S. L.; Sweeney,
F. J.; Tan, D. H.; Yu, C. H. Bioorg. Med. Chem. Lett. 2002,
12, 1387.
8. Rabinowitz, M. H.; Andrews, R. C.; Becherer, J. D.; Bick-
ett, D. M.; Bubacz, D. G.; Conway, J. G.; Cowan, D. J.; Gaul,
M.; Glennon, K.; Lambert, M. H.; Leesnitzer, M. A.;
McDougald, D. L.; Moss, M. L.; Musso, D. L.; Rizzolio,
M. C. J. Med. Chem. 2001, 44, 4252.
9. Mock, M.; Fouet, A. Annu. Rev. Microbiol. 2001, 55, 647.
10. Johnson, E. A. Annu. Rev. Microbiol. 1999, 53, 551.
11. Chen, C.-A.; Sieburth, S.McN.; Glekas, A.; Hewitt, G. W.;
Trainor, G. L.; Erickson-Viitanen, S.; Garber, S. S.; Cordova,
B.; Jeffrey, S.; Klabe, R. M. Chem. Biol. 2001, 8, 1161.
12. Sieburth, S.McN.; Nittoli, T.; Mutahi, A. M.; Guo, L.
Angew. Chem., Int. Ed., Engl. 1998, 37, 812.
The structure of 14 differs from that of 1 in two ways. The
replacement of the Cbz group in 1 by the dihy-
drocinnamoyl group replaces an oxygen with a methylene
group. This exchange was anticipated to have a minimal
effect on the binding, based on inspection of the crystal
structures of inhibitors bound to the thermolysin active
site.24 The substitution of silicon for phosphorus, an
exchange of two second row elements, eliminates the
charge on the transition state analogue. In addition,
there are intrinsic differences in bond angles and bond
lengths for the central silicon and phosphorous units.29
Conclusions
13. Mutahi, M. w.; Nittoli, T.; Guo, L.; Sieburth, S.McN. J.
Am. Chem. Soc. 2002, 124, 7363.
14. Grobelny, D.; Goli, U. B.; Galardy, R. E. Biochemistry
1989, 28, 4948.
Silanediol 14 is an effective inhibitor of thermolysin,
with a Ki similar to that of the Grobelny/Bartlett phos-
phinate 1.
15. Morgan, B. P.; Scholtz, J. M.; Ballinger, M. D.; Zipkin,
I. D.; Bartlett, P. A. J. Am. Chem. Soc. 1991, 113, 297.
16. Rochow, E. G. Silicon and Silicones; Springer-Verlag:
New York, 1987.
17. Eaborn, C. J. Chem. Soc. 1952, 2846.
18. Bailey, W. F.; Punzalan, E. R. J. Org. Chem. 1990, 55,
5404.
19. Negishi, E.; Swanson, D. R.; Rousset, C. J. J. Org. Chem.
1990, 55, 5406.
20. Evans, D. A.; Urpi, F.; Somers, T. C.; Clark, J. S.; Bilo-
deau, M. T. J. Am. Chem. Soc. 1990, 112, 8215.
21. Haubold, W.; Herdtle, J.; Gollinger, W.; Einholz, W. J.
Organomet. Chem. 1986, 315, 1.
Silanediols, as potentially useful structural units, suffer
from a reputation for irreversible condensation to form
siloxane polymers. Some condensation of 14 may have
been observed under the acidic and basic conditions used
for hydrolysis of the phenyl–silicon bonds. Nevertheless,
the transient conversion of crude silanediol to a difluoro-
silane and subsequent hydrolysis results in a very pure
silanediol product. A similar hydrolysis of difluoro-
silanes to yield silanediols was reported by Eaborn in
195217 and a recent example was reported by Organ et
al.30 The resulting silanediol 14 can be precipitated as
stable monomeric species, or left in aqueous solution for
weeks (at least) without decomposition. Alternative
chemistry to prepare 2, the precise silanediol analogue of 1
carrying the Cbz group, is currently under study.
22. Kaufmann, D. Chem. Ber. 1987, 120, 853.
23. Earle, M. J.; Fairhurst, R. A.; Heaney, H.; Papageorgiou,
G. Synlett 1990, 621.
24. Tronrud, D. E.; Holden, H. M.; Matthews, B. W. Science
(Washington, D.C.) 1987, 235, 571.
25. Struchkov, Y. T.; Ovchinnikov, Y. E.; Shipov, A. G.;
Baukov, Y. I. Russ. Chem. Bull. 1995, 44, 1705.
26. Bassindale, A. R.; Borbaruah, M.; Glynn, S. J.; Parker,
D. J.; Taylor, P. G. J. Organomet. Chem. 2000, 606, 125.
27. Feder, J.; Brougham, L. R.; Wildi, B. S. Biochemistry
1974, 13, 1186.
Acknowledgements
This work was supported by a grant from the National
Institutes of Health.
28. Christianson, D. W.; Lipscomb, W. N. J. Am. Chem. Soc.
1988, 110, 5560.
29. For a comparison of carbon, phosphorus and silicon
structures as they relate to protease inhibitor structures, see ref
13.
References and Notes
1. Kim, J.; Sieburth, S. McN. Proceedings of 222nd ACS
National Meeting, Chicago, IL, 2001.
30. Organ, M. G.; Buon, C.; Decicco, C. P.; Combs, A. P.
Org. Lett. 2002, 4, 2683.