Discovery of selective hydroxamic acid inhibitors of tumor necrosis factor-α converting enzyme

Article abstract of DOI:10.1016/S0960-894X(01)00603-5

Modification of the P₁′ substituent of macrocyclic matrix metalloproteinase (MMP) inhibitors provided compounds that are selective for inhibition of tumor necrosis factor-α converting enzyme (TACE) over MMP-1 and MMP-2. Several analogues potent

Full text of DOI:10.1016/S0960-894X(01)00603-5

Bioorganic & Medicinal Chemistry Letters 11 (2001) 2907–2910
Discovery of Selective Hydroxamic Acid Inhibitors of Tumor
Necrosis Factor-ꢀ Converting Enzyme
James Holms,* Katherine Mast, Patrick Marcotte, Ildiko Elmore, Junling Li, Lori Pease,
Keith Glaser, Douglas Morgan, Michael Michaelides and Steven Davidsen
Cancer Research Area, Abbott Laboratories, Dept. 47J, Bldg. AP10, 100 Abbott Park Road, Abbott Park, IL 60064, USA
Received 23 July 2001; accepted 13 August 2001
0
Abstract—Modification of the P1 substituent of macrocyclic matrix metalloproteinase (MMP) inhibitors provided compounds that
are selective for inhibition of tumor necrosis factor-a converting enzyme (TACE) over MMP-1 and MMP-2. Several analogues
0
potently inhibited the release of TNF-a in a THP-1 cellular assay. Compounds containing a trimethoxyphenyl group in the P1
substituent demonstrated TACE selectivity across several series of hydroxamate-based inhibitors. # 2001 Elsevier Science Ltd. All
rights reserved.
Tumor necrosis factor-a (TNF-a) is a cytokine pro-
duced mainly by activated monocyte/macrophages and
released as a major mediator of inflammatory and
immune responses.1 The over-expression of TNF-a has
been implicated in diseases such as rheumatoid arthritis,
Crohn’s disease, septic shock, AIDS, insulin resistance,
cachexia, and cancer. Agents that block the action of
TNF-a may therefore be effective in the treatment of
these disease states. The approval of biologics such as
TNF antibodies (Remicade1) and TNF soluble recep-
tors (Enbrel1) for treatment of rheumatoid arthritis
and Crohn’s disease has validated the therapeutic effec-
tiveness of TNF-a inhibition. Recently, a small mole-
cule inhibitor of TNF-a processing through the selective
inhibition of TNF-a converting enzyme (TACE) has
entered clinical studies.2
were found to inhibit TNF-a production through inhi-
bition of TACE.4 With the discovery of these dual
MMP/TACE inhibitors, the objective of the current
study was to identify structural substitutions of known
MMP inhibitors which could differentiate these two
activities. This effort ultimately led to the discovery of
potent inhibitors that were selective for TACE relative
to two MMPs, fibroblast collagenase (MMP-1) and
gelatinase A (MMP-2).
The investigation of TACE inhibition began by explor-
ing the SAR of a series of macrocyclic hydroxamic acid
MMP inhibitors similar to those previously described in
reports from our laboratories5 and others.6 The syn-
thetic route to the macrocyclic hydroxamates is shown
in Scheme 1. Succinate 17 was alkylated with the silyl
protected 4-bromobutanol to afford the diastereomer 2
after chromatography. Epimerization with LDA gave
the desired diastereomer 3. After coupling with l-tyr-
osine-N-methyl amide, deprotection of the silyl alcohol,
and cyclization under Mitsunobu conditions, macro-
cycle 5 was obtained. Treatment with 9-BBN followed
by coupling with various aryl bromides via palladium-
catalyzed S0uzuki reactions afforded analogues with
modified P1 substituents. Hydrolysis of the tert-butyl
ester to the carboxylic acid followed by conversion to
the hydroxamic acid gave analogues 7a–h. Further
investigation of the SAR of the macrocyclic hydrox-
amates was performed on phenyl ketone analogues. The
synthetic strategy employed was similar to the method
to prepare the above N-methyl amide analogues except
TNF-a is produced by enzymatic cleavage of an Ala-Val
bond of a 26 kDa membrane-bound pro-TNF releasing
the 17 kDa mature protein. The major enzyme respon-
sible for this proteolytic cleavage is the sheddase TACE,
a zinc-containing metalloproteinase belonging to a dis-
integrin and metalloproteinase domain (ADAM) sub-
family of the metzincin family of enzymes.3 The active
site of TACE was found to be very similar to the active
site of another metzincin subfamily, the matrix metallo-
proteinases (MMPs), and several inhibitors of MMPs
*Corresponding author. Fax: +1-847-935-5165; e-mail: james.
holms@abbott.com
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00603-5

2908
J. Holms et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2907–2910
Scheme 1. Reagents: (a) LDA, THF, I(CH2)4OSi(CH3)2tBu, ꢀ78 ꢁC to rt; (b) LDA, THF, ꢀ78 ꢁC, MeOH quench; (c) 1-hydroxybenzotriazole
(HOBT), DMF, l-Tyr-NHMe HCl, 4-methylmorpholine (NMM), 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide (EDC), 0 ꢁC; (d) tetra-
.
butylammonium fluoride (TBAF), THF, 0 ꢁC; (e) Bu3P, 1,10-(azodicarbonyl)dipiperidine, THF/benzene; (f) 9-BBN, THF/DMF, PdCl2 (dppf),
Cs2CO3, RPhBr; (g) TFA, CH2Cl2, 0 ꢁC; (h) HOBT, DMF, 0 ꢁC, NMM, EDC, O-(t-butyldimethylsilyl)hydroxylamine.
Table 1. Macrocyclic amide hydroxamic acids
Table 2. Macrocyclic ketone hydroxamic acids
Compd
R
TACE
IC50 (nM) IC50 (nM) IC50 (nM) IC50 (nM)
MMP-1
MMP-2TNF-THP-1
Compd
R
TACE MMP-1 MMP-2TNF-THP-1
IC50 (nM) IC50 (nM) IC50 (nM) IC50 (nM)
8a
96
55
43
87
74
320
>10,000
>10,000
9200
11
3400
1000
230
>5000
3500
7a
7b
3.4
2.7
2.3
6.8
0.26
0.79
440
400
8b
8c
8d
8e
2900
7c
7d
7e
7f
30 5.2210
180
>5000
15,000
6.8
7.6
33
140
13
4100
140
300
440
220
>10,000
4000
2300
>10,000
28
110
2200
Selective inhibition of MMPs has been achieved for
0
several series of inhibitors by varying the size of the P1
substituent.9 Fibroblast collagenase has a shallow S1
0
7g
7h
6.0
12
0.23
pocket compared0 to gelatinase A and is generally intol-
erant of large P1 substituents. The X-ray structure of
the TACE catalytic domain with bound hydroxamate
suggests the enzyme could accommodate inhibitors
>1,000
96
0
containing large P1 substituents.10 Therefore, com-
pounds with substituted phenylpropyl groups were pre-
pared to explore the different S1 pockets of TACE,
0
MMP-1, and MMP-2and to identify P modifications
0
1
succinate 3 was coupled to the amino ketone derivative
of l-tyrosine.8 Conversion to the hydroxamic acids
afforded compounds 8a–e.
that provide TACE selectivity. The compounds were tes-
ted for TACE11 and MMP12 inhibition using fluoro-
metric assay systems. Several of these macrocyclic

J. Holms et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2907–2910
Table 3. Non-succinate hydroxamates containing the trimethoxyphenyl group
2909
Compd
TACE
IC50 (nM)
MMP-1
IC50 (nM)
MMP-2TNF-THP-1
IC50 (nM) IC50 (nM)
9 (Prinomastat)
7.9
5.7
0.048
8600
10
11
12
1.5
26
>10,000
>10,000
21
3400
430
550
Not tested
41,000
180
6.5
hydroxamates failed to demonstrate a trend toward
TACE selectivity and, as shown in Table 1, potent
inhibition for both TACE and MMP-2was observed
(7a–d). However, an approximate 10- to 20-fold selec-
tivity for gelatinase A versus fibroblast collagenase was
observed as 0 expected for these compounds (7a–d).
When the P1 substituent was modified to contain the
trimethoxyphenyl (TMP) group (7e) and the 3,5-bis(tri-
fluoromethyl)phenyl group (7f), potent TACE activity
was achieved and gelatinase A activity was substantially
diminished providing compounds with overall TACE
selectivity. The data suggest that substituents at the 3
In order to further investigate the capacity of sub-
0
stituted aryl P1 groups to impart TACE selectivity, the
trimethoxyphenyl group was incorporated onto several
series of non-succinate hydroxamic acids and evaluated
for TACE inhibition, TACE versus MMP selectivity,
and cellular inhibition of TNF-a production.
As shown in Table 3, several TMP-containing non-suc-
cinate hydroxamates were synthesized and evaluated.
The penicillamine derived sulfonamides14 (9 and 10), the
a-methyl sulfide15 (11), and the a-methyl sulfone15 (12)
were prepared according to known procedures.
0
and 5 phenyl positions are necessary and that bulky P1
groups alone are not enough to decrease gelatinase A
activity since the analogue with the hindered 1-naphthyl
group (7g) does not provide TACE selectivity.
The reference compound prinomastat (9) demonstrated
very potent inhibitory activity against MMP-2as well as
potent activity against MMP-1 and TACE. As observed
in the macrocycle examples, the TMP-containing ana-
logue (10) exhibited a decrease in the MMP-2inhibitory
potency. This 400-fold decrease, however, was not suf-
ficient to provide significant TACE selectivity even with
the slight increase in TACE potency because of the
intrinsic selectivity for MMP-2observed with this series.
Other non-succinate TMP-containing compounds (11–
12) were found to be potent inhibitors of TACE while
differing in the degree of selectivity over the MMPs.
Compound 11 was moderately potent against TACE
with an IC50 value of 180 nM but was also less than 20-
fold selective vs MMP-2. Compound 12 showed good
selectivity over the MMPs (66-fold vs MMP-2and
>1,500-fold versus MMP-1) and good potency against
TACE with an IC50 value of 6.5 nM but failed to
potently inhibit the release of TNF-a in the cellular
assay.
The macrocyclic hydroxamates 7a–h were also eval-
uated for their ability to block TNF-a production in a
THP-1 cellular assay.13 The majority of these com-
pounds were active in the whole cell assay with com-
pound 7e, containing the trimethoxyphenyl group,
exhibiting an IC50 value of 140 nM.
Since it had been previously disclosed that ketones
0
0
could replace the P2 –P3 amide bond of succinyl
hydroxamate MMP inhibitors,8 additional SAR of the
macrocyclic hydroxamates was conducted on phenyl
ketone analogues to determine whether the TACE
selective aryl substitutions would have the same effect in
another series. As shown in Table 2, potent inhibition
and selectivity for TACE was observed with comp0ounds
containing various substituted phenylpropyl P1 sub-
stituents (8b–e), compared to0 the reference compound
8a containing a tolylpropyl P1 group. The TMP ketone
analogue 8b demonstrated TACE selectivity compar-
able to the corresponding TMP-containing N-methyl
amide 7e. However, in contrast to the macrocyclic amide
hydroxamates, the phenyl ketone analogues were found
to be only weakly active in the TNF cellular assay.
0
In conclusion, several substituted propylphenyl P1
groups were discovered that provide selectivity for inhi-
bition of TACE over MMP-1 and MMP-2. Differences
0
0
in the S1 pockets may explain why these P1 groups
cause a decrease in the inhibitory activity of the MMPs
while maintaining potent inhibitory activity of TACE.

2910
J. Holms et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2907–2910
TACE selectivity was demonstrated across several dif-
ferent series of hydroxamate-based inhibitors0 with
compounds containing a TMP group in the P1 sub-
stituent.
Hardman, K.; Nelson, D.; Copeland, R.; Jaffee, B.; Decicco,
C. J. Med. Chem. 1998, 41, 1745.
7. Intermediate 1 was prepared according to the procedure in
ref 5, substituting pent-4-enoic acid chloride for 4-methyl-
valeric acid chloride.
8. Sheppard, G.; Florjancic, A.; Giesler, J.; Xu, L.; Guo, Y.;
Davidsen, S.; Marcotte, P.; Elmore, I.; Albert, D.; Magoc, T.;
Bouska, J.; Goodfellow, C.; Morgan, D.; Summers, J. Bioorg.
Med. Chem. Lett. 1998, 8, 3251.
9. White, A.; Bocan, T.; Boxer, P.; Peterson, J.; Schrier, D.
Curr. Pharm. Des. 1997, 3, 45.
References and Notes
1. Bemelmans, M.; van Tits, L.; Buurman, W. Crit. Rev.
Immunol. 1996, 16, 1.
2. Decicco, C. Abstracts of Papers, 221st National Meeting of
the American Chemical Society, San Diego, CA, April 1–5,
2001; American Chemical Society: Washington, DC, 2001;
MEDI 189.
3. Black, R.; Rauch, C.; Kozlosky, C.; Peschon, J.; Slack, J.;
Wolfson, M.; Castner, B.; Stocking, K.; Reddy, P.; Srinivasan,
S.; Nelson, N.; Bolani, N.; Schooley, K.; Gerhart, M.; Davis,
R.; Fitzner, J.; Johnson, R.; Paxton, R.; March, C.; Cerretti,
D. Nature 1997, 385, 729.
4. (a) McGeehan, G.; Becherer, J.; Bast, R.; Boyer, C.;
Champion, B.; Connolly, K.; Conway, J.; Furdon, P.; Karp,
S.; Kidao, S.; McElroy, A.; Nichols, J.; Pryzwansky, K.;
Schoenen, F.; Sekut, L.; Truesdale, A.; Verghese, M.; Warner,
J.; Ways, J. Nature 1994, 370, 558. (b) Gearing, A.; Beckett,
M.; Christodoulou, M.; Churchill, M.; Clements, J.; David-
son, A.; Drummond, A.; Galloway, W.; Gilbert, R.; Gordon,
J.; Leber, T.; Mangan, M.; Miller, K.; Nayee, P.; Owens, K.;
Patel, S.; Thomas, W.; Wells, G.; Wood, L.; Wooley, K. Nat-
ure 1994, 370, 555.
10. Maskos, K.; Fernandez-Catalan, C.; Huber, R.; Bour-
enkov, G.; Bartunik, H.; Ellestad, G.; Reddy, P.; Wolfson, M.;
Rauch, C.; Castner, B.; Davis, R.; Clarke, H.; Petersen, M.;
Fitzner, J.; Cerretti, D.; March, C.; Paxton, R.; Black, R.;
Bode, W. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 3408.
11. A soluble form of TACE, lacking the carboxyterminal
membrane binding domain, was produced from a baculovirus
expression system. The secreted, active enzyme was partially
purified using Q-Sepharose and hydroxyapatite chromato-
graphy. TACE inhibition was assessed in a microtiter plate
format by a protocol similar to that used for MMP assays,
using a TACE specific fluorogenic substrate, Acetyl-Gly-
Glu(EDANS)-Leu-Ala-Gln-Ala-Val-Arg-Ser-Ser Lys(DAB-
CYL)-Gly-NH2, in 10 mM Tris–HCl, 1 mM MgCl2, pH 7.4.
12. Marcotte, P.; Davidsen, S. In Current Protocols in Phar-
macology; Enna, S., Williams, M., Ferkany, J., Kenakin, T.,
Porsolt, R., Sullivan, J., Eds.; John Wiley and Sons: New
York, 2001; p 3.7.1.
13. Glaser, K.; Pease, L.; Li, J.; Morgan, D. Biochem. Phar-
macol. 1999, 57, 291.
14. Zook, S.; Dagnino, R.; Deason, M.; Bender, S.; Melnick,
M. PCT Patent Application WO 97/20824.
15. Warpehoski, M.; Harper, D. PCT Patent Application WO
98/13340.
5. Steinman, D.; Curtin, M.; Garland, R.; Davidsen, S.; Hey-
man, R.; Holms, J.; Albert, D.; Magoc, T.; Nagy, I.; Marcotte,
P.; Li, J.; Morgan, D.; Hutchins, C.; Summers, J. Bioorg. Med.
Chem. Lett. 1998, 8, 2087.
6. Xue, C.; He, X.; Roderick, J.; DeGrado, W.; Cherney, R.;