First insight into structure-activity relationships of selective meprin β inhibitors

Article abstract of DOI:10.1016/j.bmcl.2017.04.012

The astacin proteases meprin α and β are emerging drug targets for treatment of disorders such as kidney failure, fibrosis or inflammatory bowel disease. However, there are only few inhibitors of both proteases reported to date. Starting from NNGH as lead

Full text of DOI:10.1016/j.bmcl.2017.04.012

Accepted Manuscript
First insight into structure-activity relationships of selective Meprin β inhibitors
Daniel Ramsbeck, Antje Hamann, Dagmar Schlenzig, Stephan Schilling, Mirko
Buchholz
PII:
S0960-894X(17)30375-X
http://dx.doi.org/10.1016/j.bmcl.2017.04.012
BMCL 24865
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
15 February 2017
3 April 2017
3 April 2017
Please cite this article as: Ramsbeck, D., Hamann, A., Schlenzig, D., Schilling, S., Buchholz, M., First insight into
structure-activity relationships of selective Meprin β inhibitors, Bioorganic & Medicinal Chemistry Letters (2017),
doi: http://dx.doi.org/10.1016/j.bmcl.2017.04.012
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

First insight into structure-activity relationships of selective Meprin β
inhibitors
Daniel Ramsbecka, ∗ ,Antje Hamanna, Dagmar Schlenziga, Stephan Schillinga, Mirko
Buchholza,
∗
a Fraunhofer Institute for Cell Therapy and Immunology IZI, Department of Drug Design and
Target Validation MWT, Biocenter, Weinbergweg 22, 06120 Halle (Saale), Germany
* Corresponding author. Tel.: +49-345-13142826; fax: +49-345-13142801; e-mail: daniel.ramsbeck@izi.fraunhofer.de
∗ Corresponding author. Tel.: +49-345-13142825; fax: +49-345-13142801; e-mail: mirko.buchholz@izi.fraunhofer.de
Graphical Abstract
Leave this area blank for abstract info.
First insight into structure-activity
relationships of selective Meprin β inhibitors
Daniel Ramsbeck , Antje Hamann, Dagmar Schlenzig, Stephan Schilling, Mirko Buchholz
Keywords
Meprin β, hydroxamic acid, sulfonamides, astacins
Abstract
The astacin proteases meprin α and β are emerging drug targets for treatment of disorders such as
kidney failure, fibrosis or inflammatory bowel disease. However, there are only few inhibitors of both
proteases reported to date. Starting from NNGH as lead structure, a detailed elaboration of the
structure-activity relationship of meprin β inhibitors was performed, leading to compounds with
activities in the lower nanomolar range. Considering the preference of meprin β for acidic residues in
the P1' position, the compounds were optimized. Acidic modifications induced potent inhibition and
>100-fold selectivity over other structurally related metalloproteases such as MMP-2 or ADAM10.
∗ Corresponding author. Tel.: +49-345-13142826; fax: +49-345-13142801; e-mail: daniel.ramsbeck@izi.fraunhofer.de
∗ Corresponding author. Tel.: +49-345-13142825; fax: +49-345-13142801; e-mail: mirko.buchholz@izi.fraunhofer.de

Compelling evidence suggests the endoproteases meprin α and β as novel drug targets[1,2]. Due to their
procollagenase activity, meprins have been linked to disorders that involve collagen deposition, e.g.
skin or lung fibrosis and keloids[1,3]. Moreover, meprins are associated with nephritis, cancer and
inflammatory bowel disease[4,5] and in particular meprin β has also been shown to act as a β-secretase
independent from BACE1[6–8]. The meprins belong to the astacins, a subfamily of the metzincin
superfamily of proteolytic enzymes. Hence, they are structurally and functionally related to well-
characterized metalloproteases such as MMPs or ADAMs[9]. Meprin β exhibits an unique substrate
recognition pattern, preferring acidic amino acids in P1 to P2´-position[10]. This unusual substrate
specificity is caused by a clustering of arginine residues within the active site (figure 2b), which is not
observed with other metzincins and astacins, e.g. meprin α or BMP-1[11,12]. On our quest to decipher the
role of meprin activity in these disorders, we aimed here at the discovery of selective meprin
β inhibitors.
Figure 1: Inhibition constants of known meprin β inhibitors
Only few inhibitors of meprins have been reported to date (figure 1). Among those, the naturally
occurring compound Actinonin proved to be the most potent inhibitor of both, meprin α and β. Some
other inhibitors of MMPs or ADAMs also exhibited inhibitory activity on meprins, but displayed
much lower potency than Actinonin[13]. More recently, the compounds PPNDS and NF449 have been
reported as very potent and partially selective inhibitors of meprin β[14].

Figure 2: Schematic depiction of the assumed binding mode to meprin β (left); active site of meprin β (right)
S1
Zn2+
S2´
R146
R184
R238
S1´
a)
b)
Despite the higher potency of Actinonin, NNGH (N-Isobutyl-N-(4-methoxyphenylsulfonyl)glycyl
hydroxamic acid) [13] served as our lead structure for the development of meprin β inhibitors, due to the
easy chemical accessibility of this well-known sulfonamide MMP-inhibitor scaffold. Metzincin-
metalloprotease inhibitors with sulfonamide-scaffold bind to their target enzymes as depicted in figure
2a[15,16]. A similar binding mode of NNGH to the active site of meprin β was assumed. Thus, a ligand-
based compound optimization was performed, with modifications of the moieties targeting the
individual subsites of meprin β, i.e. arginine residues shaping S1, S1´or S2´.
The compounds have been synthesized using glycine or β-alanine derivatives as precursors (scheme
1). For the synthesis of the secondary sulfonamides 7a-w, either glycine or β-alanine methyl ester was
sulfamoylated with the respective sulfamoyl chloride, followed by conversion of the ester to the
hydroxamic acid by means of hydroxylamine under microwave irradiation. Some compounds were
synthesized on solid support using chlorotrityl-hydroxylamine resin, which was preloaded with Fmoc-
glycine. After Fmoc-removal under standard conditions, the amino group was sulfamoylated, followed
by cleavage from the resin using trifluoroacetic acid in dichloromethane. For the synthesis of the
tertiary sulfonamides 8a-m, sulfonamides 4 were treated with alcohols under Mitsunobu conditions or
were alkylated using alkyl or benzyl halides under microwave irradiation. The resulting glycine-
sulfonamide methyl esters 6 were transformed to the corresponding hydroxamic acids 8 using
hydroxylamine hydrochloride under microwave irradiation. Inhibitors carrying an N-(3-
carboxybenzyl) moiety (11a-g) were synthesized by reductive amination of 3 with tert-Butyl-3-
formylbenzoate, followed by sulfamoylation under microwave irradiation and subsequent
deprotection.

Scheme 1: General synthesis scheme of the sulfonamide hydroxamic acids
O
O
R2
N
O
O
O
R2
N
c or d
e (R=H3CO)
H
O
O
O
S O
R1
N
HO
HO
HO
R
S
O
R
S
O
N
H
n
R1
R1
R =
R =
H3CO-
-Trt-ONH-
R = H3CO-
4
5
6
8
a or b
e (R=H3CO-)
(R= -Trt-ONH)
O
f
H
N
O
NH2
R
N
H
S
O
n
n
R1
R =
R =
R =
H3CO-
-Trt-ONH-
Trt-ONH-
1
2
3
7
g
O
COOtBu
O
COOtBu
COOH
h
f
O
O
NH
N
N
R
R
S
O
N
H
S
O
R1
R1
9
10
11
:
Conditions
a) R2-SO2Cl, TEA,THF, rt; b) R1-SO2Cl, K2CO3, H2O/dioxane, rt; c) R2-OH, DIAD, PPh3, THF, 0°-rt; d) R2-Hal, K2CO3, MeCN, MW, 120°C; e) NH2OH*HCl,
KOH, MeOH, MW 80°Cf) TFA/DCM (1:1 v/v), TIS, rt; g) tert-butyl 3-formylbenzoate, NaBH4, MeOH, rt; h) R1-SO2Cl, THF, TEA, MW 100°C
In the first optimization step the sulfonamide residue was modified. This residue addresses the S1´
pocket of the metzincin proteases and is the major determinant of compound selectivity. Exploration
of the S1´-substituent was performed by introducing diverse aromatic secondary sulfonamide moieties
(table 1). However, the impact of most substituents on the activity was very low and only the
introduction of acidic functional groups (7q-s) led to an increased activity compared to NNGH. A very
similar inhibitory potency was also achieved with the respective β-alanine derivatives 7v-w. These
results nicely correspond to the preference of meprin β for acidic amino acids in P1´position of the
substrates, which is determined by ionic interactions with positively charged arginine R238 forming
the S1´-subpocket (figure 2). Because halophenols have been reported as carboxylic acid bioisosters[17],
a dichlorophenol was introduced (7s), which exhibited similar inhibitory potency as the carboxylic
acid derivatives. Interestingly, the introduction of an additional methylene group between the aromatic
ring and the carboxylic acid (7t) led to a loss of activity compared to 7q and r. This suggests a steric
hindrance within the S1´-pocket, ruling out an interaction of the carboxylic acid with the S1´-arginine
residue.
Table 1: Inhibition of meprin β by secondary sulfonamides
n
R1
R2
R3
R4
R5
IC50 (µM)
7a
1
1
1
1
1
1
1
H
H
H
OCH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
9.02 0.30
18.85 0.35
26.90 9.62
7.42 0.03
6.19 0.80
18.80 2.55
25.70 0.28
7b
7c
7d
7e
7f
OCH3
OCH3
H
OCH3
H
-OCH2CH2O-
H
F
H
H
OCH3
OPh
CH3
H
7g
OCH3

7h
7i
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH3
H
H
H
H
H
H
H
H
H
H
H
Cl
H
H
H
Cl
H
H
14.77 0.78
17.17 2.57
14.32 0.59
17.97 1.25
10.26 1.12
3.85 0.65
4.42 0.18
7.31 1.06
5.65 0.66
0.41 0.01
0.31 0.01
0.43 0.08
5.03 0.26
0.34 0.04
0.22 0.02
0.40 0.13
Cl
7j
H
F
H
H
7k
7l
F
CH3
H
CF3
H
Cl
Ph
H
7m
7n
7o
7p
7q
7r
7s
H
Ph(p-OCH3)
Ph( p-F)
Ph(p-Cl)
COOH
H
H
H
H
H
H
H
H
COOH
Cl
H
OH
H
7t
H
CH2COOH
COOH
H
H
7u
7v
7w
H
H
COOH
Cl
H
OH
H
Next, the impact of addressing the S2´-subpocket of meprin β was evaluated (table 2). For this
purpose, the 4-methoxysulfonamide moiety was kept constant and several alkyl- and arylalkyl
substituents were introduced. The inhibitory potency of the synthesized compounds (8a-i) was
virtually not affected by small alkyl or larger arylalkyl moieties. In all cases except 8h, the activity
was lower compared to NNGH. Therefore, attention was drawn again on acidic sidechains. The
activity could be improved by the introduction of an acetic acid, propionic acid and a p-carboxybenzyl
sidechain (8j-l), and was apparently dependent on the spacer length between the nitrogen and the
carboxylic acid. The introduction of an m-carboxybenzyl sidechain (8m) led to a boost of activity,
triggering an about 300-fold improved potency compared to NNGH. This suggests an improved
orientation of the carboxylic acid towards arginine R146 shaping the S2´-pocket (figure 2), enabling
the proper formation of an ionic interaction.
Table 2: Inhibition of meprin β by tertiary sulfonamides
R
IC50 (µM)
-CH2CH(CH3)2
-CH3
17.10 3.01
32.53 4.22
31.20 3.22
37.50 16.89
38.80 2.36
40.47 3.40
34.95 18.88
NNGH
8a
8b
-CH2CH3
8c
-(CH2)3CH3
-CH2CH2CH(CH3)2
-CH2Ph
8d
8e
8f
- CH2Ph(p-OCH3)

8g
8h
8i
-CH2Ph(3,4OCH2O)
-CH2CH2Ph
1.93 0.21
42.90 4.00
43.87 2.08
5.98 0.37
2.55 0.48
0.98 0.05
0.06 0.001
-CH2-(2-Norbornyl)
-CH2COOH
8j
8k
8l
-CH2CH2COOH
-CH2Ph(p-COOH)
-CH2Ph(m-COOH)
8m
Since the introduction of the m-carboxybenzyl side chain led to a significant increase of activity, this
moiety was kept constant and the SAR of the S1´-substituent was re-evaluated. Caused by the
additional S2´-interaction of the acidic side chain, all compounds (11a-e, table 3) exhibit increased
activities compared to the respective secondary sulfonamides. However, only the introduction of a
benzodioxolane led to an IC -value in the lower nanomolar range, similar as observed with 8m before.
50
Interestingly, in contrast to the secondary sulfonamide derivatives 7q and r, a significant difference in
the activity of the p- and m-carboxyaryl-sulfonamide was observed. For the respective tertiary
sulfonamides 11f and g a gain of activity was just observed by introduction of the m-carboxylic acid
(11g). This could be driven by a slightly shifted binding mode due to the S2´-interaction, leading to a
different positioning within the S1´-subpocket that rules out the proper interaction of arginine R238
with the p-carboxylic acid of 11f.
Table 3: Inhibition of meprin β by m-carboxybenzyl substituted sulfonamides
R1
R2
IC50 (µM)
-OCH2O-
0.27 0.02
0.65 0.06
0.70 0.08
1.08 0.45
1.06 0.06
11a
11b
11c
11d
11e
11f
F
H
OCH3
OCF3
Ph
H
H
F
H
COOH 0.32 0.02
0.06 0.01
COOH
H
11g
Numerous hydroxamate-based MMP inhibitors have been tested in model systems and clinical trials
during the past years. However, most of them have been withdrawn due to toxicity issues. The reason
of most of those effects are interactions not only with the respective target enzyme, but also with off-
targets, i.e. other metalloproteases[18]. Thus, the selectivity of novel metalloprotease inhibitors should

be treated with special regards. To verify the selectivity profile of the most potent novel meprin β
inhibitors (8m and 11g), the inhibitory activity towards selected metzincin metalloproteases, i.e.
meprin α, MMP2, MMP9, MMP13, ADAM10 and ADAM17, has been evaluated.
Table 4: Inhibition data of selected metalloproteases
8m
IC50 (µM)
11g
IC50 (nM)
Protease
0.06 0.001
3.29 0.41
0.06 0.01
27.35 3.04
n.d.*
Meprin β
Meprin α
MMP2
0.002 0.0001
0.05 0.001
MMP9
n.d.*
MMP13
ADAM10
ADAM17
0.0003 0.000003
11.70 1.13
n.d.*
n.d.*
1.24 0.01
n.d.*
*n.d. (not determined) – no inhibition @ 200 µM inhibitor detected
To our surprise, 8m is not only a potent inhibitor of meprin β, but also a very potent inhibitor of
MMP2, 9 and 13. This might be due to the p-methoxyarylsulfonamide targeting the S1´-subpocket. As
mentioned above, this residue determines the compound selectivity within the metzincin-superfamily
and in particular the p-methoxyarylsulfonamide can be found in several non-specific metalloprotease
inhibitors, e.g. NNGH[19]. Nevertheless, this compound exhibits a good selectivity for the inhibition of
meprin β over meprin α by a factor of 60. The introduction of the m-carboxyphenylsulfonamide (11g)
significantly influenced the selectivity profile of the inhibitor. While the activity against meprin β was
not affected, the activity against meprin α dropped by a factor 7 compared to 8m, leading to a 500-fold
selectivity for the inhibiton of meprin β over α. Additionally, the activity against the other
metalloproteases was almost completely abolished. Thus, the acidic functionality targeting the S1´-
pocket is not only required for high potency against meprin β, but also leads to an excellent selectivity
over other metzincins, that have been investigated so far. Thus, this enhanced selectivity renders 11g
as a promising lead compound for further development of selective meprin β inhibitors.
In summary, a first insight into SAR of meprin β inhibitors could be gained, yielding compounds with
inhibitory activity in the lower nanomolar range. Furthermore, selectivity could be achieved against
the isoform meprin α and also against selected structurally related metalloproteases of the metzincin-
superfamily. However, additional studies, e.g. docking experiments or co-crystallization with meprin
β, should give a better insight into the binding mode of those inhibitors enabling further optimization
of the compounds.
Acknowlegments

We thank Mercedes Scharfe (IZI-MWT) and Dr. Andrea Porzel (Leibniz Institute of Plant
Biochemistry, Halle) for their excellent technical support. We gratefully acknowledge the funding of
this project by the federal state of Saxony-Anhalt, Germany.
Supplementary data
Supplementary data (experimental procedures and spectroscopic characterization data of the
compounds) associated with this article can be found, in the online version.
References
[1]
[2]
[3]
C. Broder, C. Becker-Pauly, Biochem. J. 450 (2013) 253–264.
J. Prox, P. Arnold, C. Becker-Pauly, Matrix Biol. 44-46 (2015) 7–13.
C. Broder, P. Arnold, S. Vadon-Le Goff, M.A. Konerding, K. Bahr, S. Muller, C.M. Overall,
J.S. Bond, T. Koudelka, A. Tholey, Hulmes, D. J. S., C. Moali, C. Becker-Pauly, Proc. Natl.
Acad. Sci. U.S.A. 110 (2013) 14219–14224.
[4]
[5]
[6]
[7]
S. Banerjee, G. Jin, S.G. Bradley, G.L. Matters, R.D. Gailey, J.M. Crisman, J.S. Bond, Am. J.
Physiol. Gastrointest. Liver Physiol. 300 (2011) G273-82.
V. Biasin, M. Wygrecka, L.M. Marsh, C. Becker-Pauly, L. Brcic, B. Ghanim, W. Klepetko, A.
Olschewski, G. Kwapiszewska, Sci Rep 7 (2017) 39969.
J. Bien, T. Jefferson, M. Causevic, T. Jumpertz, L. Munter, G. Multhaup, S. Weggen, C.
Becker-Pauly, C.U. Pietrzik, J. Biol. Chem. 287 (2012) 33304–33313.
C. Schönherr, J. Bien, S. Isbert, R. Wichert, J. Prox, H. Altmeppen, S. Kumar, J. Walter, S.F.
Lichtenthaler, S. Weggen, M. Glatzel, C. Becker-Pauly, C.U. Pietrzik, Mol Neurodegener 11
(2016) 19.
[8]
C. Becker-Pauly, C.U. Pietrzik, Front Mol Neurosci 9 (2017) 146.
[9]
E.E. Sterchi, W. Stöcker, J.S. Bond, Mol. Aspects Med. 29 (2008) 309–328.
[10]
C. Becker-Pauly, O. Barre, O. Schilling, auf dem Keller, U., A. Ohler, C. Broder, A. Schutte,
R. Kappelhoff, W. Stöcker, C.M. Overall, Mol. Cell Proteomics 10 (2011) M111.009233.
[11]
[12]
J.P. Villa, G.P. Bertenshaw, J.S. Bond, J. Biol. Chem. 278 (2003) 42545–42550.
J.L. Arolas, C. Broder, T. Jefferson, T. Guevara, E.E. Sterchi, W. Bode, W. Stöcker, C.
Becker-Pauly, F.X. Gomis-Ruth, Proc. Natl. Acad. Sci. U.S.A. 109 (2012) 16131–16136.
[13]
M.-N. Kruse, C. Becker, D. Lottaz, D. Köhler, I. Yiallouros, H.-W. Krell, E.E. Sterchi, W.
Stöcker, Biochem J. 378 (2004) 383–389.

[14]
[15]
F. Madoux, C. Tredup, T.P. Spicer, L. Scampavia, P.S. Chase, P.S. Hodder, G.B. Fields, C.
Becker-Pauly, D. Minond, Biopolymers 102 (2014) 396–406.
S. Hanessian, S. Bouzbouz, A. Boudon, G.C. Tucker, D. Peyroulan, Bioorg. Med. Chem. Lett.
9 (1999) 1691–1696.
[16]
[17]
P. Jain, C. Saravanan, S.K. Singh, Eur J Med Chem 60 (2013) 89–100.
J. Qiu, S.H. Stevenson, M.J. O'Beirn, R.B. Silverman, J. Med. Chem. 42 (1999) 329–332.[18]
R.E. Vandenbroucke, C. Libert, Nat Rev Drug Discov 13 (2014) 904–927.
[19]
S.P. Gupta, Chem. Rev. 115 (2015) 6427–6490.