α-Biphenylsulfonylamino 2-methylpropyl phosphonates: Enantioselective synthesis and selective inhibition of MMPs

Article abstract of DOI:10.1016/j.bmc.2006.10.047

(R)-α-Biphenylsulfonylamino 2-methylpropyl phosphonates attain nM potency against several MMPs and are the most effective inhibitors based on phosphonate as zinc binding group. Since their preparation by direct N-acylation of expensive, enantiopure, α-aminophosphonic acids proceeds in low yields, we devised and evaluated a stereoselective and straightforward method of synthesis that avoids the unfavourable step of N-acylation. The key intermediate (R)-4-bromophenylsulfonylamino 2-methylpropyl phosphonate 9 was obtained by highly stereoselective addition of dibenzylphosphite to the enantiopure (S)-N-isobutylidene-p-bromobenzenesulfinamide 3, followed by oxidation with m-CPBA. Suzuki coupling of 9 with the desired arylboronic acids, gave the expected biphenylsulfonylamino derivatives in satisfactory yields. Liberation of the phosphonic group by hydrogenolysis led to the desired (R)-α-biphenylsulfonylamino 2-methylpropyl phosphonates 14a-i. Screening of the new compounds on MMP-1, -2, -3, -7, -8, -9, -13 and -14 showed IC₅₀ in the range of nM in most cases.

Full text of DOI:10.1016/j.bmc.2006.10.047

Bioorganic & Medicinal Chemistry 15 (2007) 791–799
a-Biphenylsulfonylamino 2-methylpropyl phosphonates:
Enantioselective synthesis and selective inhibition of MMPs
Alessandro Biasone,a Paolo Tortorella,b Cristina Campestre,a Mariangela Agamennone,a
Serena Preziuso,a Marika Chiappini,a Elisa Nuti,c Paolo Carelli,c Armando Rossello,c
Fernando Mazzad and Carlo Gallinaa,*
a
`
Dipartimento di Scienze del Farmaco, Universita ‘‘G. d’Annunzio’’, Via dei Vestini 31, 66013 Chieti, Italy
b
`
Dipartimento Farmaco-Chimico, Universita degli Studi di Bari, Via Orabona 4, 70125 Bari, Italy
`
`
Dipartimento di Chimica, Universita di L’Aquila, Via. Vetoio, 67010 L’Aquila, Italy
c
Dipartimento di Scienze Farmaceutiche, Universita degli Studi di Pisa, Via Bonanno 6, 56126 Pisa, Italy
d
Received 26 July 2006; revised 17 October 2006; accepted 23 October 2006
Available online 25 October 2006
Abstract—(R)-a-Biphenylsulfonylamino 2-methylpropyl phosphonates attain nM potency against several MMPs and are the most
effective inhibitors based on phosphonate as zinc binding group. Since their preparation by direct N-acylation of expensive, enan-
tiopure, a-aminophosphonic acids proceeds in low yields, we devised and evaluated a stereoselective and straightforward method of
synthesis that avoids the unfavourable step of N-acylation. The key intermediate (R)-4-bromophenylsulfonylamino 2-methylpropyl
phosphonate 9 was obtained by highly stereoselective addition of dibenzylphosphite to the enantiopure (S)-N-isobutylidene-p-
bromobenzenesulfinamide 3, followed by oxidation with m-CPBA. Suzuki coupling of 9 with the desired arylboronic acids, gave
the expected biphenylsulfonylamino derivatives in satisfactory yields. Liberation of the phosphonic group by hydrogenolysis
led to the desired (R)-a-biphenylsulfonylamino 2-methylpropyl phosphonates 14a–i. Screening of the new compounds on
MMP-1, -2, -3, -7, -8, -9, -13 and -14 showed IC50 in the range of nM in most cases.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
Hydroxamates7 are the most popular MPIs owing to the
exceptional strength of their ZBG, although metabolic
instability may compromise their oral bioavailability.
While hydroxamate optimization4 points to improve-
ment of selectivity and metabolic stability, other inhibi-
tors based on different zinc binding functions, such as
carboxylate, phosphonate and thiolate,8 are currently
investigated, and we also have been studying phospho-
nate MPIs for a long time.9
Matrix metalloproteinases (MMPs) are zinc endopepti-
dases that can degrade virtually all the constituents of
the extracellular matrix (ECM).1 While they intervene
in various physiological processes, chronic over-regula-
tion of MMP activity results in excessive degradation of
ECM components and has been implicated in numerous
pathological conditions.2 A great variety of synthetic,
low molecular weight, MMP inhibitors (MPIs) are there-
fore studied for the development of innovative chemo-
therapeutics in several fields where effective drugs are
not yet available,3–5 and some of them are presently in ad-
vanced clinical trials.4,6 Their structures include a peptide
or a peptidomimetic moiety that is preferably accommo-
dated in the S0 region of the active site, and a zinc binding
group (ZBG) able to coordinate the catalytic zinc ion.
Pursuing our programme in this subject, and stimulated
by basic contributions discussing SAR versus several
MMPs,10 we focused our attention on a-biphenylsulfo-
nylamino 2-methylpropyl phosphonates 1, analogues
of known carboxylate11 and hydroxamate12,13 MPIs.
Keywords: MMP phosphonate inhibitors; Enantiopure sulfinimine;
Dibenzyl phosphite stereoselective addition; Suzuki coupling.
Examples reported in a patent14 and our own findings
(Ki in the nM range for MMP-2, MMP-3, MMP-8
*
Corresponding author. Tel.: +39 0871 355 5373; fax: +39 0871 355
5267; e-mail: cgallina@unich.it
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.10.047

792
A. Biasone et al. / Bioorg. Med. Chem. 15 (2007) 791–799
and aggrecanase) proved that the a-biphenylsulfonyla-
mino substitution gives rise to the most potent MPIs
based on phosphonate as ZBG. While the search for
new analogues was abandoned, due to the ample cover-
age of the patent, we turned our attention to a stereo-
selective method of synthesis and to the elucidation of
their mode of binding9 in the active site of MMP-8.
2. Chemistry
Addition of lithium dialkyl phosphites to enantiopure
and configurationally restricted sulfinimines occurs with
excellent diastereoselectivity and the resulting N-sulfi-
nyl-a-amino phosphonates can be readily converted into
the enriched (R)- and (S)-a-amino phosphonic acids.15
Further studies16 confirmed the effectiveness of this
method, clarified17 that preferred adducts are (SS, R)-
N-sulfinyl-a-amino phosphonates and brought out the
proposal18 of a unified asymmetric induction model to
explain the stereoselective outcome of the addition.
Biphenylsulfonylamino 2-methylpropyl phosphonates
were previously prepared14 by direct N-acylation of the
appropriate a-aminophosphonic acid (Scheme 1). Yields
were rather low, owing probably to the steric hindrance
around the amino group. Moreover, the R- and S-enan-
tiomers of the a-aminophosphonic acids, when available,
are expensive. We have therefore devised a new stereo-
selective and straightforward approach that avoids the
unfavourable step of N-acylation. The method, previ-
ously used to prepare the a-biphenylsulfonylamino
2-methylpropyl phosphonates (R) and (S)-14a,9 has been
further improved, in this paper, and evaluated for the
preparation of a series of the analogues 14b–i and 16a,
b, only one of which (14e) was previously reported.14
Adaptation of this excellent diastereoselective method
(Scheme 2) to prepare the biphenylsulfonylamino
2-methylpropyl phosphonates (R)- and (S)-14a for
cocrystallisation in complexes with MMP-89 was
previously outlined. Enantiopure (S)-p-bromobenzene-
sulfinamide 2, readily obtained from the (1R,2S,5R)-
(S)-p-bromobenzenesulfinate,19
converted into the (S)-p-bromobenzenesulfinimine 3
by treatment with i-butyraldehyde in the presence of
menthyl
can
be
Scheme 1.
Scheme 2.

A. Biasone et al. / Bioorg. Med. Chem. 15 (2007) 791–799
793
Table 1. Addition of lithium dialkyl phosphites to the enantiopure
sulfinimine (S)-3a
Table 2. Arylation of p-bromophenylsulfonylamino phosphonates
7–9a
Reagent
Product
Yield (%)
SS,R/SS,S
Ester
Boronic acid
Baseb
Product
Yield (%)
(EtO)2POLi
(MeO)2POLi
(BnO)2POLi
49
5
84
83
86
91/9b
90:10c
95.5/4.5
7
8
8
9
9
9
9
10a
10a
10a
10a
10a
10a
10a–i
A
A
B
A
B
C
C
11a
12a
12a
13a
13a
13a
13a–i
86
6
6
15
5
a All reactions were carried out for 3 h, in THF at ꢀ78 ꢁC.
b By integration of 31P NMR resonances in CDCl3 at 25.12 and
25.25 ppm.
c By integration of 31P NMR resonances in CDCl3 at 27.42 and
27.58 ppm.
45
73
43–73
a All reactions were carried out in toluene under reflux by using
Pd(PPh3)4 as catalyst (3 mol%) and boronic acids 10a–i (1.2 equiv).
b A, aqueous Na2CO3; B, anhydrous Na2CO3; C, anhydrous K3PO4.
5 equiv of Ti(OEt)4.20 Addition of lithium diethyl phos-
phite at ꢀ78 ꢁC to the N-isobutylidene-(S)-p-bromo-
benzenesulfinimine 3, according to the procedure of
Lefebvre,15 afforded a mixture of the two diastereoiso-
mers (SS,R)-4 and (SS,S)-4 in a 91:9 ratio (Table 1).
The major adduct (SS,R)-4 could be further enriched
by silica-gel chromatography to improve the enantio-
meric purity (99.8%) of the final phosphonate 14a. Oxi-
dation of the (SS,R)-N-p-bromobenzene sulfinamide 4
with m-CPBA gave the corresponding N-p-bromoben-
zene sulfonamide (R)-7 in essentially quantitative yield.
This intermediate was converted into the diethyl
a-biphenylsulfonylamino 2-methylpropyl phosphonates
(R)-11a by coupling with 4-methoxyphenylboronic acid
in the presence of Pd(PPh3)4 and aqueous sodium
carbonate, and finally converted into the free phospho-
nates 14a by acid hydrolysis.
12a (Table 2). When coupling was carried out in the
presence of anhydrous sodium carbonate, yield was only
slightly increased since sodium carbonate, in its behav-
iour as a base, releases a water molecule and dimethyl
phosphonates are particularly sensitive to base catalyzed
hydrolysis.
Addition of lithium dibenzyl phosphite to the enantio-
pure (S)-N-isobutylidene-p-bromobenzene-sulfinimine 3
occurred in high yields and with an even higher dia-
stereoselectivity (Table 1). The major diastereoisomer
(SS,R)-6 could be easily purified by silica-gel chromatog-
raphy and highly enriched (99.8% de) by crystallisation
from EtOAc/n-hexane. After satisfactory oxidation with
m-CPBA, however, the dibenzyl phosphonate 9 showed
to be too sensitive to basic catalyzed hydrolysis in the
Suzuki coupling. In the presence of either aqueous or
anhydrous sodium carbonate, unsatisfactory recovery
of the expected biphenylsulfonylamino dibenzyl phos-
phonate 13a was obtained (Table 2). When anhydrous
potassium triphosphate was used in place of sodium car-
bonate, the dibenzylester hydrolysis was notably re-
pressed and recovery of the dibenzyl phosphonate was
satisfactory. The free a-biphenylsulfonylamino (R)-2-
methylpropyl phosphonic acid 14a was obtained by Pd
catalyzed hydrogenolysis and isolated as the cyclohexyl-
amine salt in practically quantitative yields. The enan-
tiomeric purity of the phosphonate (99.9% ee) was
monitored by HPLC analysis on a Chiralpack AD
column, after conversion into the corresponding methy-
lester by treatment with diazomethane and guarantees
that no racemization intervenes in the Suzuki coupling
and hydrogenolysis step. To confirm the usefulness of
this process, the a-biphenylsulfonylamino dibenzyl
phosphonates 13b–i were also prepared according to
Scheme 2 and successfully converted into the phospho-
nates 14b–i.
Since the Suzuki–Miyaura reaction21 is one of the most
efficient methods for creation of a C–C bond between
aromatics and a large range of the required arylboronic
acids are commercially available, we decided to evaluate
the effectiveness of this method for the preparation of a
small series (14a–i) of a-biphenylsulfonylamino (R)-2-
methylpropyl phosphonates (Scheme 2).
While diastereoselectivity of lithium diethyl phosphite
addition was satisfactory and yields were good, the rath-
er vigorous conditions (6 M HCl in H2O/acetic acid
under reflux) required for hydrolysis of the phosphonate
diethyl ester caused an extensive hydrolysis of the sulfo-
nylamino group, and an unsatisfactory recovery (50%)
of the expected phosphonate (R)-14a. With the aim to
further improve the overall yields of this method, addi-
tion of both lithium dimethyl phosphite or dibenzyl
phosphite in the first step of the process was tried since
the final dimethyl phosphonates require milder condi-
tions for their hydrolysis22 and dibenzyl phosphonates
undergo facile and quantitative hydrogenolysis.23
Addition of lithium dimethyl phosphite (Scheme 2) to
the enantiopure (S)-N-isobutylidene-p-bromobenzene-
sulfinimine 3 and the following oxidation of the adduct
5 to the dimethylphosphonate 8 occurred with yield and
stereoselectivity similar to that of the diethyl analogue
(Table 1). The Suzuki reaction in toluene under reflux,
in the presence of aqueous sodium carbonate, however,
caused extensive hydrolysis of the dimethyl to monom-
ethyl phosphonate, poor yields of the arylation reaction
and poor recovery of the expected biphenyl derivative
Since the Suzuki reaction also included coupling with
vinylboronic acids, the two arylsulfonylamino dibenzyl
phosphonates 15a and 15b (Scheme 3) were successfully
prepared, starting from two, commercially available,
2-arylvinylboronic acids. The expected a-arylsulfonyla-
mino phosphonic acids, where the two aromatics are
connected by means of an ethylenic spacer, would have
been potentially interesting as selective MPIs, in accor-
dance with previous studies on a carboxylate ana-
logue.10a Liberation of the phosphonic group by Pd

794
A. Biasone et al. / Bioorg. Med. Chem. 15 (2007) 791–799
Scheme 3.
catalyzed hydrogenolysis led, however, to the saturated
analogues 16a and 16b.
in vitro potencies were determined by fluorimetric assay.
To achieve a wide selectivity profile, in view of possible
clinical applications,6b the screening was extended to
MMP-1, -2, -3, -7, -8, -9, -13 and -14 (Table 3). The fol-
lowing results, in terms of potency and selectivity, may
be underlined.
3. Biological evaluation and structure–activity
relationships
Since the extension of the present method of synthesis to
a series of a-biphenylsulfonylamino 2-methylpropyl
phosphonates led to a small library of new MPIs, their
The 40-methoxybiphenylsulfonylamino-2-methylpropyl
phosphonate (R)-14a is 750- and 580-fold more potent
than the opposite enantiomer (S)-14a9 against MMP-2
Table 3. Inhibition data of 2-methylpropyl phosphonates 14a–i and 16a, b against MMPsa
Compound
Ar
MMP-1
5%b
MMP-2
1100
MMP-3
16%b
MMP-7
17%b
MMP-8
810
MMP-9
40%b
MMP-13
68%b
MMP-14
63%b
(S)-14a
(R)-14a
(R)-14b
(R)-14c
(R)-14d
(R)-14e
(R)-14f
(R)-14g
150
450
600
160
77
1.5
0.39
1.8
20
52
12
460
450
1.4
8.0
0.56
24
2.6
1.1
4.2
13
79
11
0.37
1.50
1.1
36
2000
1400
1700
150
130
32
150
28
59
8.7
2.3
53
0.81
0.39
2.4
21
5.3
3.3
63
47
98
45
4.1
11
590
1200
20,000
170
180
(R)-14h
(R)-14i
320
320
30
24
440
230
2900
1100
1.3
0.4
160
64
24
15
75
26
(R)-16a
(R)-16b
33%b
9%b
980
790
34%b
31%b
50,000
260
110
7100
680
820
480
6000
130,000
40,000
a IC50 (nM); errors in the range of 5–10% of the reported value (from three different assays).
b % Inhibition at 100 lM.

A. Biasone et al. / Bioorg. Med. Chem. 15 (2007) 791–799
795
and MMP-8, respectively. Preference of MMP-8 for the
R-enantiomer 14a has been discussed in detail on the ba-
sis of the crystal complexes of (R)-14a and (S)-14a with
MMP-8.9 The even larger potency of the R-relative to
the S-enantiomer, observed for all the other tested
MMPs, can be related to the strong similarities in the ac-
tive sites of this family of enzymes. Similar results (Table
4) were found for the examples previously reported [(R)-
14e, (S)-14e and (R)-14l] in the aforementioned patent.14
against MMP-2 and MMP-8. The IC50 values (2–
20 nM) against MMP-2 of the 40-substituted (R)-a-
biphenylsulfonylamino phosphonates (R)-14a–f are very
similar to those of previously studied carboxylate
analogues13 (S)-17a,d,f,m (Table 4). The (R)-a-
biphenylsulfonylamino hydroxamate (R)-18m (Table
4), in addition, is more potent than the carboxylate ana-
logue (S)-17m, and closely approaches potencies of
phosphonate analogues versus the tested enzymes.
All the (R)-a-biphenylsulfonylamino 2-methylpropyl
phosphonates 14a–i confirm to be excellent MPIs. They
exhibit IC50 values in the nM range against all the tested
enzymes and some of them attain sub-nM potency
While carboxylates retain nM potency only versus
MMP-2, -3 and -13 and exhibit micromolar potency ver-
sus MMP-1, -7 and -9, the majority of corresponding
phosphonates 14a–f retain nM potency also versus
Table 4. Inhibition of selected examples of phosphonates 14 and analogous carboxylates (17) and hydroxamates (18) against MMPsa
O
O
S
N
H
ZBG
Ar
Ar
Compound
ZBG
MMP-1
150
MMP-2
1.5
MMP-3
52
MMP-7
460
MMP-8
1.4
MMP-9
8.0
MMP-13
2.6
MMP-14
79
(R)-14a
PO3H2
(S)-17ab
(R)-14d
(S)-17db
(R)-14e
(R)-14ed
(S)-14ed
(S)-17eb
(R)-14ld
(R)-14f
COOH
PO3H2
COOH
PO3H2
PO3H2
PO3H2
COOH
PO3H2
PO3H2
COOH
COOH
CONHOH
1500
160
3
8
7200
1400
4800
1700
ND
NDc
1.1
2200
59
6
ND
32
20
39
8.7
ND
ND
11
ND
2.3
2
150
13
43
5.3
ND
ND
48
ND
3.3
11
8
4200
77
10
ND
0.81
1
64,000
21
ND
47
28
ND
ND
6500
ND
98
6
ND
ND
16,000
ND
4.1
ND
ND
ND
ND
11
10,000
ND
1000
ND
2
9
7500
ND
5
N
45
3
150
0.39
ND
NT
ND
(S)-17fb
(S)-17mb
(R)-18mb
2200
6000
110
4500
7200
140
3900
7900
18
ND
NT
ND
4
7
Br
1
5
2
a IC50 (nM).
b Ref. 13.
c Not determined.
d Ref. 14.

796
A. Biasone et al. / Bioorg. Med. Chem. 15 (2007) 791–799
MMP-1, -8 and -9, and decrease their potency to micro-
molar level only versus MMP-7. As previously observed
for carboxylates,13 the decrease in potency of phospho-
nates against MMP-1 and -7 can be attributed to the
restriction of the S10 pocket of these enzymes, where
an arginine or a tyrosine residue defining the bottom
of the pocket hinders accommodation of the distal ring
of the biphenyl moiety of the inhibitor.24,25 Flexibility of
the protein in this area in some cases allows an induced
fit, upon ligand binding, that moves the occluding resi-
due and favours accommodation of large P10
groups.26,27 Phosphonates 14a–i retain nM potency ver-
sus MMP-9 and are some 1000-fold more potent (Table
4) than the corresponding carboxylates.20 This large dif-
ference in selectivity between carboxylate and phospho-
nate analogues versus MMP-9 was unexpected. It can
probably be explained by assuming that only phospho-
nate, rather than carboxylate, can be coordinated as
ZBG in such a way as to allow proper orientation of
the biphenyl moiety for easy accommodation in the nar-
row S10 pocket25 of MMP-9.
for all the observed cases. When a methyl group in 30 is
involved a sub-nM IC50 against MMP-8 is retained and
selectivity versus the remaining MMPs is notably in-
creased. Binding affinities of the two a-arylsulfonylami-
no phosphonates 16a and 16b, where the phenyl rings
are connected by means of a CH2–CH2 spacer, are gen-
erally strongly decreased owing to the high flexibility
and conformational mobility of the P10 hydrophobic
group. IC50 in the range of nM are however retained
versus MMP-2 and -8.
4. Conclusion
The efficient stereoselective synthesis of the key interme-
diate (SS,R)-6 allows easy access into a series of various-
ly substituted (R)-a-biphenylsulfonylamino 2-methyl-
propyl phosphonates through the Suzuki coupling with
arylboronic acids, followed by Pd catalyzed hydrogenol-
ysis. Facile diastereomeric enrichment of the key inter-
mediate by column chromatography and crystallisation
guarantees very high levels of enantiomeric purity of
the final products.
Replacement of MeO by EtO in 40 position improves
selectivity towards MMP-1 and increases potency
against most of the tested MMPs, attaining IC50 in the
range 0.37–1.1 nM for MMP-2, -8, -9 and -13. When
R = i-Pr, a small decrease in potency against these
MMPs is observed, but selectivity against MMP-1,
MMP-7 and MMP-14 is improved. Docking studies
clearly show that 40-ethoxy group increases hydrophobic
interactions with Leu193 side chain (MMP-8 number-
ing), a conserved residue in all the MMPs of the present
study, except MMP-1 and -7. Further increase of hydro-
phobic interactions (i-PrO in place of EtO group) causes
an even greater increase of steric hindrance (Fig. 1) and
results in a decrease of the potency.
Evaluation of the inhibiting activities of (R)- and (S)-14a
against MMP-1, -2, -3, -7, -8, -9, -13 and -14 showed
IC50 in the range of nM in most cases only for the
R-enantiomer. The most powerful analogues studied in
this paper contain an alkoxy substitution in 40 and at-
tain IC50 in the range 0.37–1.1 nM for MMP-2, -8, -9
and -13. These examples, in addition to some 30 substi-
tuted analogues, show that appropriate structural varia-
tions on the biphenyl distal ring can strongly affect
potency and selectivity.
5. Experimental
5.1. Chemistry
The shift of an F or Cl substituent from the 40 to the 30
position generally causes a decrease in binding affinities
Melting points were determined on a Buchi B-540 appa-
¨
ratus and are uncorrected. Optical rotations were mea-
sured with a Perkin-Elmer 241 digital polarimeter at
20 ꢁC; concentrations are expressed as g/100 mL. HPLC
analyses were performed with a Perkin-Elmer HPLC
chromatograph equipped with a Rheodyne 7725i model
injector, a 785A model UV/Vis detector, a series 200
model pump and NCI 900 model interface. 1H, 13C
and 31P NMR spectra were recorded on a Varian
VXR 300 spectrometer, operating at 300, 75 and
121 MHz, respectively. Chemical shifts are expressed
in d (ppm) values relative to an internal standard
(TMS for proton and carbon; H3PO4 for phosphorus);
coupling constants (J) are given in Hz. 13C and 31P
NMR spectra are fully proton decoupled. Elemental
microanalyses of C, H and N were performed on a Carlo
Erba mod. 1106 analyzer and were within 0.4% of the
calculated values.
5.1.1. (S)-N-(2-Methylpropylidene)-p-bromobenzenesul-
finimine (3). Isobutyraldehyde (2.19 mL, 24.0 mmol)
and titanium(IV) ethoxide (29.8 mL, 120 mmol) were
added to a solution of the sulfinamide19 2 (5.29 g,
Figure 1.

A. Biasone et al. / Bioorg. Med. Chem. 15 (2007) 791–799
797
20
D
1
24.0 mmol) in CH2Cl2 (370 mL), under N2. The reaction
mixture was refluxed for 3 h and quenched at 0 ꢁC by
addition of H2O (270 mL). After filtration through a
short pad of Celite, the organic phase was secured and
the aqueous solution was further extracted with CH2Cl2
(2· 100 mL). The pooled organic phases were dried
(Na2SO4) and concentrated to leave the crude product
317 (6.02 g; 91%) as a yellow oil, that was employed
without further purification.
½aꢁ ꢀ 11:8ꢂ (c 1.0 CHCl3); H NMR (CDCl3) d 0.93–
0.98 (m, 6H), 2.11–2.23 (m, 1H), 3.73 (ddd, 1H,
J = 21.0, 9.9 and 3.3 Hz), 4.63–4.88 (m, 4H), 5.11 (dd,
1H J = 10.2 and 4.8 Hz), 7.18–7.22 (m, 4H), 7.31–7.33
(m, 6H), 7.46–7.49 (m, 2H), 7.65–7.68 (m, 2H); 13C
NMR (CDCl3) d 17.8 (d, J = 2.8 Hz), 20.6 (d, J =
13.1 Hz), 29.7 (d, J = 4.8 Hz), 56.2 (d, J = 151.4 Hz),
68.1 (d, J = 7.1 Hz); 68.3 (d, J = 7.1 Hz), 127.6, 128.4,
128.8, 128.8, 128.9, 132.3, 135.8 (d, J = 5.7 Hz), 135.9
(J = 5.4 Hz), 140.6. Enantiomeric excess 99.8 deter-
mined by integration of the HPLC chromatograms: Chi-
ralpack AD column (4.6·250 mm); iso-propanol/n-
hexane (1:1) containing 0.025% TFA, at a flow rate of
0.5 mL/min; peak detection at 280 nm; retention times
16.5 and 22.3 min for (R)- and (S)-9, respectively.
5.1.2. Dibenzyl (SS,R)-N-(p-bromobenzenesulfinyl)-2-
amino-2-methylpropylphosphonate [(SS,R)-6]. To a solu-
tion of dibenzyl phosphite (12.6 g, 48.0 mmol) in THF
(360 mL), contained in a 500 mL two-necked round-bot-
tommed flask equipped with a magnetic stirring bar and
rubber septum, a solution (1.0 M in THF) of lithium
bis(trimethylsilyl)amide (48 mL, 48.0 mmol) was slowly
added under nitrogen, at ꢀ78 ꢁC. After stirring for
30 min, the resulting mixture was transferred via cannula
into a 1 L two-necked round-bottommed flask equipped
with a magnetic stirring bar, rubber septum and nitro-
gen inlet, containing a solution of sulfinimine 3 (6.02 g,
22.0 mmol) in THF (230 mL), cooled at ꢀ78 ꢁC. After
3 h at ꢀ78 ꢁC, the reaction mixture was quenched with
saturated NH4Cl (250 mL) solution, the organic layer
was secured and the aqueous phase was further extract-
ed with EtOAc (3· 150 mL). The pooled organic phases
were dried, concentrated, and fractionated by silica gel
chromatography (3% iso-propanol in CHCl3), to give a
95.5/4.5 mixture of (SS,R)-6 and (SS,S)-6 (10.15 g,
86%). Relative abundances of the two diastereoisomers
were determined by integration of the HPLC chromato-
grams: Luna C18 column (4.6 · 250 mm); 75:25 MeOH/
H2O containing 1% TFA at a flow rate of 1 mL/min;
peak detection at 280 nm; retention times 9.9 and
12.0 min, respectively.
5.1.4. Dimethyl (R)-N-(p-bromobenzenesulfonyl)-2-ami-
no-2-methylpropylphosphonate (8). Dimethyl phosphite
(0.46 mL, 5.0 mmol) and sulfinimine
3
(685 mg,
2.5 mmol) were reacted according to the procedure
reported for preparation of (SS,R)-6 to give a crude mix-
ture of (SS,R)-5 and (SS,S)-5 stereoisomers (846 mg,
83%; 90:10 by integration of 31P NMR resonances in
CDCl3 at 27.42 and 27.58 ppm). After enrichment by sil-
ica gel chromatography (3% iso-propanol in CHCl3), the
(SS,R)-5 stereoisomer was oxidized according to the pro-
cedure employed for preparation of 9, to give 614 mg of
the pure compound 8 as a white solid: mp 98.0–99.0 ꢁC;
20
½aꢁ ꢀ 11:8 (c 1.0 CHCl3); 1H NMR (CDCl3) d 0.94 (d,
D
3H, J = 6.3 Hz), 0.96 (d, 3H, J = 6.9 Hz), 2.14–2.22 (m,
1H), 3.45 (d, 3H, J = 7.5 Hz), 3.50 (d, 3H,
J = 7.5 Hz),3.57 (ddd, 1H, J = 19.2, 9.9 and 2.0 Hz),
5.14–5.19 (m, 1H), 7.64 (d, 2H, J = 8.1 Hz), 7.74 (d,
2H, J = 8.1 Hz); 13C NMR (CDCl3)
d 17.6 (d,
J = 2.8 Hz), 20.4 (d, J = 13.2 Hz), 29.8 (d, J = 4.8 Hz),
56.7 (d, J = 152.0 Hz), 127.8, 128.4, 132.8, 141.6.
Crystallisation from EtOAc/n-hexane of the 95.5/4.5
mixture afforded 6.4 g (54%) of the prevailing (SS,R)-6
diastereoisomer as white crystals: mp 88.0–88.5 ꢁC;
5.1.5. Synthesis of phosphonate dibenzylesters 13a–i and
15a, b (General Procedure A). A solution of 9 (1.0 mmol)
and the appropriate boronic acid (1.2 mmol) in toluene
(10 mL) was treated with tetrakis(triphenylphos-
phine)palladium(0) (22 mg) and K3PO4 (2.0 mmol), un-
der nitrogen. After being refluxed for 18–48 h (TLC
monitoring for disappearance of 9), the reaction mixture
was diluted with EtOAc (15 mL), added with 1 M HCl
(15 mL) under stirring and filtered through a short pad
of Celite. The organic layer was separated and further
washed with NaHCO3 saturated solution (20 mL) and
brine, dried and concentrated in vacuo. The resulting
yellow oil was purified by chromatography on silica
gel (40% EtOAc/n-n-hexane) and crystallisation.
20
1
½aꢁ þ 21:2ꢂ (c 1.0 CHCl3); H NMR (CDCl3) d 0.98
D
(d, 3H, J = 6.9 Hz), 1.07 (dd, 3H, J = 6.9 and 1.2 Hz),
2.21–2.33 (m, 1H), 3.46 (ddd, 1H, J = 20.1, 9.6 and
2.7 Hz), 4.58 (t, 1H, J = 9.6 Hz), 4.82–5.12 (m, 4H),
7.27–7.36 (m, 10H), 7.54–7.56 (m, 4H); 13C NMR
(CDCl3) d 18.9 (d, J = 6.0 Hz), 19.8 (d, J = 7.4 Hz),
28.3, 53.4 (d, J = 152.0 Hz), 68.9 (d, J = 6.9 Hz), 69.1
(d, J = 6.5 Hz), 128.1, 128.4, 128.8 (d, J = 3.0 Hz),
128.9 (d, J = 3.0 Hz), 131.6, 132.6, 133.2, 135.6 (d,
J = 4.3 Hz), 135.6 (d, J = 5.7 Hz), 138.1, 143.6. Relative
abundance (HPLC) of (SS,R)-6 relative to (SS,S)-6: 99.8/
0.2.
5.1.6. Hydrogenolysis of phosphonate dibenzylesters 13a–
i and 15a, b (General Procedure B). A solution of the
phosphonate dibenzylester (1.0 mmol) in 1:1 CH2Cl2/
MeOH (40 mL) was stirred under hydrogen, in the pres-
ence of 5% Pd/C (100 mg), for 24 h. The reacting mix-
ture was filtered and evaporated under reduced
pressure to give the crude phosphonic acid as a white
solid that was dissolved in CH3OH and added with a
small excess of cyclohexylamine. Evaporation of the sol-
vent and recrystallisation from CH3OH/Et2O gave the
pure cyclohexylamine salts 14a–i and 16a, b.
5.1.3. Dibenzyl (R)-N-(p-bromobenzenesulfonyl)-2-ami-
no-2-methylpropylphosphonate (9). To a solution of
(SS,R)-6 (6.4 g, 11.9 mmol) in CH2Cl2 (160 mL), m-
CPBA (4.1 g, 23.9 mmol) was added portionwise, at
0 ꢁC, under stirring. After 0.5 h, the reaction mixture
was sequentially washed with saturated solutions of
Na2S2O3 (2· 200 mL), NaHCO3 (2· 200 mL) and brine.
The organic phase was dried and concentrated in vacuo
to give 6.39 g (97%) of the pure phosphonate 9 as a
white solid: mp 97.5–98.2 ꢁC (EtOAc/n-hexane);

798
A. Biasone et al. / Bioorg. Med. Chem. 15 (2007) 791–799
5.2. MMP inhibition assays28
The sulfonamide ligand 14a was used to define the active
site box and to set up grid calculation. The prepared
crystal structures were used for the docking campaign
with GLIDE that was successful in reproducing the
(R)-14a crystal structure binding mode. All ligands of
the present study were manually built in Maestro31,
exploiting the Build function and minimized with
MMFFs force field as implemented in the MacroModel
software package,32 using a distance-dependent dielec-
tric constant of 4.0, with Polak-Ribier conjugate gradi-
ent as minimizer and a threshold value of 0.01 kJ/
Recombinant human progelatinase A (pro-MMP-2) and
B (pro-MMP-9) from transfected mouse myeloma cells
were supplied by Prof Gillian Murphy (Department of
Oncology, University of Cambridge, UK); pro-MMP-
1, pro-MMP-3, pro-MMP-7, pro-MMP-8, pro-MMP-
13 and pro-MMP-14 were purchased from Calbiochem.
Proenzymes were activated immediately prior to use
with p-aminophenylmercuric acetate (APMA 2 mM for
1 h at 37 ꢁC for MMP-2, MMP-1 and MMP-8; 1 mM
for 1 h at 37 ꢁC for MMP-9 and MMP-7; 1 mM for
0.5 h at 37 ꢁC for MMP-13). Pro-MMP-14 and pro-
MMP-3 were activated with trypsin 5 lg/mL for
15 min at 37 ꢁC followed by soybean trypsin inhibitor
(SBTI) 23 lg/mL for pro-MMP-14 and 62 lg/mL for
pro-MMP-3.
˚
(A mol) as the convergence criterion. Minimized struc-
tures were the input for docking calculations into
each enzyme. GLIDE was used with Extra Precision
docking protocol default settings, enlarging the energy
window filtering higher level poses to 5 kcal/mol. The
top 10 docked poses for each docked structure were
analyzed.
For assay measurements, the inhibitor stock solutions
(DMSO, 100 mM) were further diluted, at seven differ-
ent concentrations (0.01 nM–300 lM) for each MMP
in the fluorimetric assay buffer (FAB: Tris 50 mM, pH
7.5, NaCl 150 mM, CaCl2 10 mM, Brij 35 0.05% and
DMSO 1%). Activated enzyme (final concentration
2.9 nM for MMP-2, 2.7 nM for MMP-9, 2.4 nM for
MMP-7, 1 nM for MMP-14 and MMP-3, 1.5 nM
for MMP-8, 0.66 nM for MMP-13 and 0.20 nM for
MMP-1) and inhibitor solutions were incubated in the
assay buffer for 4 h at 25 ꢁC. After the addition of
200 lM solution of the fluorogenic substrate Mca-Arg-
Pro-Lys-Pro-Val-Glu-Nva-Trp-Arg-Lys(Dnp)-NH2
(Sigma) for MMP-3 and Mca-Lys-Pro-Leu-Gly-Leu-
Dpa-Ala-Arg-NH2 (Bachem) for all the other enzymes
in DMSO (final concentration 2 lM), the hydrolysis
was monitored every 15 s for 20 min recording the
increase in fluorescence (kex = 325 nm, kem = 400 nm)
using a Molecular Device Spectramax Gemini XS plate
reader. The assays were performed in triplicate in a total
volume of 200 lL per well in 96-well microtitre plates
(Corning, black, NBS). Control wells lack inhibitor.
The MMP inhibition activity was expressed in relative
fluorescent units (RFU). Percent of inhibition was calcu-
lated from control reactions without the inhibitor. IC50
was determined using the formula: Vi/Vo = 1/(1 + [I]/
IC50), where Vi is the initial velocity of substrate cleav-
age in the presence of the inhibitor at concentration [I]
and Vo is the initial velocity in the absence of the inhib-
itor. Results were analyzed using SoftMax Pro software
and GraFit software.29
Acknowledgments
Financial support by Italian MIUR is gratefully
acknowledged. We thank Dr. Roberta Micello for preli-
minary experiments.
Supplementary data
1H and 13C NMR spectral data and physical constants
for compounds 13a–i, 14a–i, 15a, b and 16a, b are pro-
vided. Supplementary data associated with this article
can be found, in the online version, at doi:10.1016/
j.bmc.2006.10.047.
References and notes
1. Whittaker, M.; Floyd, C. D.; Brown, P.; Gearing, J. H.
Chem. Rev. 1999, 99, 2735.
2. Matrix Metalloproteinases and TIMPs; Woessner, J. P.,
Nagase, H., Eds.; Oxford University Press: New York,
NY, USA, 2000.
3. Puerta, D. T.; Cohen, S. M. Curr. Top. Med. Chem. 2004,
4, 1551.
4. Matter, H.; Schudok, M. Curr. Opin. Drug Discovery
2004, 7, 513.
5. Skiles, J. W.; Gonnella, N. C.; Jeng, A. Y. Curr. Med.
Chem. 2004, 11, 2911.
5.3. Molecular modelling
6. (a) Abbenante, G.; Fairlie, D. P. Med. Chem. 2005, 1, 71;
(b) Overall, C. M.; Kleifeld, O. Br. J. Cancer 2006, 94,
1355.
7. Muri, E. M. F.; Nieto, M. J.; Sindelar, R. D.; Williamson,
J. S. Curr. Med. Chem. 2002, 9, 1631.
All calculations were performed on a notebook ASUS
A6770KLH, equipped with an AMD Turion MT32 pro-
cessor and a RedHat Enterprise Desktop 4 OS. Crystal
structures of MMP-2 (PDB code 1qib), MMP-3 (PDB
code 1caq) and MMP-8 (PDB code 1z) were used as
the basis of the docking experiments. The GLIDE30 pro-
tein preparation protocol was used to remove all crystal-
lographic water molecules and add hydrogen atoms
using the predicted protonation states of the amino acid
residues at pH 7.4. The crystal structures were then min-
imized to optimize hydrogen atoms position.
8. Breuer, E.; Frant, J.; Reich, R. Expert Opin. Ther. Patents
2005, 15, 253.
9. Pochetti, G.; Gavuzzo, E.; Campestre, C.; Agamennone,
M.; Tortorella, P.; Consalvi, V.; Gallina, C.; Hiller, O.;
Tschesche, H.; Tucker, P. A.; Mazza, F. J. Med. Chem.
2006, 49, 923, and references cited therein.
10. (a) Matter, H.; Schwab, W.; Barbier, D.; Billen, G.;
Haase, B.; Neises, B.; Schudok, M.; Thorwart, W.;

A. Biasone et al. / Bioorg. Med. Chem. 15 (2007) 791–799
799
Schreuder, H.; Brachvogel, V.; Lonze, P.; Weithmann, K.
U. J. Med. Chem. 1999, 42, 1908; (b) Matter, H.; Schwab,
W. J. Med. Chem. 1999, 42, 4506.
22. Christov, V. C.; Aladinova, V. M.; Prodanov, B. Phos-
phorus, Sulfur Silicon Relat. Elem. 1999, 155, 67.
23. Page, P. C. B.; McKenzie, M. J.; Gallagher, J. A. J. Org.
Chem. 2001, 66, 3704.
24. Pavlovsky, A.; Ye, Q. Z.; Ortwine, D.; Humblet, C.;
Purchase, C.; White, A.; Roth, B.; Johnson, L.; Hupe, D.;
Williams, M.; Dhanaraj, V.; Blundell, T. Protein Sci. 1999,
8, 1455.
25. Dhanaraj, V.; Ye, Q. Z.; Johnson, L. L.; Hupe, D. J.;
Ortwine, D. F.; Dunbar, J. B., Jr.; Rubin, J. R.;
Pavlovsky, A.; Humblet, C.; Blundell, T. L. Structure
1996, 4, 375.
26. Lovejoy, B.; Welch, A. R.; Carr, S.; Luong, C.; Broka, C.;
Hendricks, R. T.; Campbell, J. A.; Walker, K.; Martin, R.;
Van Wart, H.; Browner, M. Struct. Biol. 1999, 6, 217.
27. Moy, F. J.; Chanda, P. K.; Chen, J.; Cosmi, S.; Edris, W.;
Levin, J. I.; Rush, T. S.; Wilhelm, J.; Powers, R. J. Am.
Chem. Soc. 2002, 124, 12659.
28. (a) Murphy, G.; Crabbe, T. Methods Enzymol. 1995, 248,
470; (b) Tuccinardi, T.; Martinelli, A.; Nuti, E.; Carelli, P.;
Balzano, F.; Uccello-Barretta, G.; Murphy, G.; Rossello,
A. Bioorg. Med. Chem. 2006, 14, 4260.
29. (a) SoftMax Pro 4.7.1 by Molecular Device; (b) GraFit
version 4 by Erithecus Software.
30. GLIDE; Schro¨dinger, LLC: Portland, OR 97204.
31. MAESTRO; Schro¨dinger, LLC: Portland, OR 97204.
32. MACROMODEL, version 7.2; Schro¨dinger, LLC: Port-
land, OR 97204.
11. Watanabe, F.; Tsuzuki, H.; Ohtani, M. (Shionogi & Co.
Ltd. Japan), 1999, WO 9727174.
12. Robinson, R. P. (Pfizer Products Inc. USA), 1999, EP
895988 A1.
13. O’Brien, P. M.; Ortwine, D. F.; Pavlovsky, A. G.; Picard,
J. A.; Sliskovic, D. R.; Roth, B. D.; Dyer, R. D.; Johnson,
L. L.; Man, C. F.; Hallak, H. J. Med. Chem. 2000, 43,
156.
14. Schudok, M.; Schwab, W.; Zoller, G.; Bartnik, E.;
Buttner, F.; Weithmann, K. U. (Aventis Pharma Deu-
tschland Gmbh), 2001, US 6,235,727 B1, and 2002, US
6,500,811 B2.
15. Lefebvre, I. M.; Evans, S. A., Jr. J. Org. Chem. 1997, 62,
7532.
16. Mikolajczyk, M.; Lyzwa, P.; Drabowicz, J. Tetrahedron:
Asymmetry 1997, 8, 3991.
17. Lefebvre, I. M.; Evans, S. A. Phosphorus, Sulfur Silicon
Relat. Elem. 1999, 397.
18. Davis, F. A.; Lee, S.; Yan, H.; Titus, D. D. Org. Lett.
2001, 3, 1757.
19. Capozzi, M. A.; Cardellicchio, C.; Naso, F.; Spina, G.;
Tortorella, P. J. Org. Chem. 2001, 66, 5933.
20. Davis, F. A.; Zhang, Y.; Andemichael, Y.; Fang, T.;
Fanelli, D. L.; Zhang, H. J. Org. Chem. 1999, 64, 1403.
21. Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457.