Selectivity profiling of DegP substrates and inhibitors

Article abstract of DOI:10.1016/j.bmc.2009.01.073

Protein quality control factors are involved in many key physiological processes and severe human diseases that are based on misfolding or amyloid formation. Prokaryotic representatives are often virulence factors of pathogenic bacteria. Therefore, protein quality control factors represent a novel class of drug targets. The bacterial serine protease DegP, belonging to the widely conserved family of HtrA proteases, exhibits unusual structural and functional plasticity that could be exploited by small molecule modulators. However, only one weak synthetic peptide substrate and no inhibitors are available to date. We report the identification of a potent heptameric pNA-substrate and chloromethyl ketone based inhibitors of DegP. In addition, specificity profiling resulted in the identification of one strong inhibitor and a potent substrate for subtilisin as well as a number of specific elastase substrates and inhibitors.

Full text of DOI:10.1016/j.bmc.2009.01.073

Bioorganic & Medicinal Chemistry 17 (2009) 2920–2924
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Selectivity profiling of DegP substrates and inhibitors
Patrick Hauske a,b, , Michael Meltzer b, , Christian Ottmann a, Tobias Krojer c,
a,
Tim Clausen c, Michael Ehrmann b,d, , Markus Kaiser
*
*
a Chemical Genomics Centre der Max-Planck-Gesellschaft, Otto-Hahn-Str. 15, D-44227 Dortmund, Germany
b Zentrum für Medizinische Biotechnologie, FB Biologie und Geographie, Universität Duisburg-Essen, D-45117 Essen, Germany
c Research Institute for Molecular Pathology (IMP), Dr. Bohrgasse 7, A-1030 Vienna, Austria
d Cardiff School of Biosciences, Cardiff University, Museum Avenue, Cardiff CF10 3US, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
Protein quality control factors are involved in many key physiological processes and severe human dis-
eases that are based on misfolding or amyloid formation. Prokaryotic representatives are often virulence
factors of pathogenic bacteria. Therefore, protein quality control factors represent a novel class of drug
targets. The bacterial serine protease DegP, belonging to the widely conserved family of HtrA proteases,
exhibits unusual structural and functional plasticity that could be exploited by small molecule modula-
tors. However, only one weak synthetic peptide substrate and no inhibitors are available to date. We
report the identification of a potent heptameric pNA-substrate and chloromethyl ketone based inhibitors
of DegP. In addition, specificity profiling resulted in the identification of one strong inhibitor and a potent
substrate for subtilisin as well as a number of specific elastase substrates and inhibitors.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 28 October 2008
Revised 21 January 2009
Accepted 25 January 2009
Available online 5 February 2009
Keywords:
HtrA protease
DegP
Allosteric regulation
Substrate
Inhibitor
1. Introduction
the resting5 and active states.6 Its proteolytic activity is regulated
in at least two complementary modes. First, DegP functions in a
DegP is a key protein quality control factor of the bacterial un-
folded protein response of the cell envelope and functions as an
essential virulence factor of several pathogenic bacteria.1,2 The un-
folded protein response is believed to counteract the protein fold-
ing stress caused by the host’s immune system attacking the
bacterial invader. Consequently, inhibition of DegP might repre-
sent an alternative strategy for the development of novel antibiot-
ics.3 To date, the rational design of small molecule inhibitors failed
due to the regulatory and mechanistic complexity as well as the
unusual plasticity of the active site, illustrating that further sys-
tematic studies with small molecules probing DegP are required.
DegP belongs to the HtrA proteins,4 an ubiquitously conserved
family of serine proteases and is best characterized in Escherichia
coli. The monomer consists of a protease domain resembling chy-
motrypsin and two C-terminal PDZ domains. PDZ domains are pro-
tein modules that mediate specific protein–protein interactions by
binding preferentially the C-terminal 3 or 4 residues of target pro-
teins. Recent studies revealed that DegP is a highly dynamic pro-
tein, assembling into distinct, multimeric particles that represent
temperature dependent manner.7 At low temperatures, it acts as
a molecular chaperone, while at elevated temperatures DegP
switches into a protease, exhibiting a profound proteolytic activity
at 37 °C. Second, DegP is allosterically regulated.8,9 Upon binding
the C-termini of hydrophobic peptides or proteins to the PDZ1 do-
main, the proteolytic activity of DegP is activated or amplified. This
mechanism involves dramatic changes in the oligomeric states that
is a change from the hexameric resting state into the proteolyti-
cally active 12 and 24-meric forms.6 Importantly, the ‘chemical’
allosteric regulation displays the higher priority, overriding the
temperature control. Recent structural studies revealed that DegP
is also able to perform the antagonistic activities of protein repair
and degradation in a single molecule. While in the large 12 and 24
mers, folded monomers of mislocalised outer membrane proteins
are protected from degradation, unfolded substrates are at the
same time efficiently degraded.6
In addition to its striking structural properties, DegP features an
unusual substrate binding and processing mode. In contrast to
classical serine proteases, DegP consists of an bipartite degradation
system, in which the PDZ1 and protease domains collaborate dur-
ing peptide cleavage.10 Substrates are first recognized at their C-
terminal end via binding to the PDZ1 domain, resulting in allosteric
activation of DegP. After fixation at PDZ1, the N-terminal substrate
segment then extends into an active site of a protease domain,
where it is preferentially cleaved after valine, alanine, threonine
* Corresponding authors. Fax: +49 201 183 3315 (M.E.); +49 231 9742 6479
(M.K.).
E-mail addresses: michael.ehrmann@uni-due.de (M. Ehrmann), markus.kaiser@
cgc.mpg.de (M. Kaiser).
Both authors contributed equally.
0968-0896/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2009.01.073

P. Hauske et al. / Bioorg. Med. Chem. 17 (2009) 2920–2924
2921
or isoleucine residues. This reaction produces a further binding site
for a PDZ domain that holds the product of the first reaction for a
second cycle of processing. This processive mechanism produces
products of rather defined sizes.10,11 These unusual and unique
peculiarities in combination with its interesting biological rele-
vance turn DegP into a highly interesting protease for probing its
function with small molecules.
We have recently reported a first para-nitroaniline (pNA)-based
substrate (1) for DegP,8 which was also used in later mechanistic
studies of substrate degradation.10 The sequence of the substrate
was derived from digestion experiments with native DegP sub-
strates. However, this substrate displays only weak affinities and
turnover rates, which can be explained by the substrate binding
mode, in which a C-terminal binding to the PDZ1 domain is re-
quired for efficient substrate hydrolysis. However, the availability
of a functional pNA-based substrate proves that besides the usual
bipartite also other substrate binding modes are possible that
could be exploited with small molecules. We therefore started a
comprehensive investigation to discover further pNA-substrates
‘violating’ the DegP bipartite binding mode. Furthermore, we pur-
sued the synthesis of a first series of irreversible DegP inhibitors
based on the chloromethylketone (CMK) reactive group for co-
crystallization and further biochemical experiments.
of HtrA proteases were carried out. The following proteins were
used for digestion experiments: alkaline phosphatase A (PhoA)
and Sigma-E factor negative regulatory protein (RseA) from
E. coli, porcine citrate synthase (cs) and human amyloid b-A4 pro-
tein (Ab). Subsequent analysis of cleavage products by LC–MS usu-
ally resulted in numerous fragments, most of which featured
alanine or valine as C-terminal residues corresponding to the P1
site. However, no further definite conserved sequence motifs could
be detected, confirming the low cleavage specificity of DegP. In
analogy to previous reports,10 the majority of identified fragments
had a length of 9–20 amino acids.
2.2. Chemical synthesis of potential DegP substrates
Due to their robust and broad applicability in biochemical as-
says, we decided to synthesize chromogenic substrates based on
the para-nitroaniline motif performing straight-forward synthesis
of pNA derivatives of the determined cleavage fragments. It con-
sisted of a combination of Fmoc-based solid phase peptide synthe-
sis and solution phase coupling, following essentially previously
described methodologies (Scheme 1 and Table 1).8,12
In brief, 2-chlorotrityl chloride resin was loaded with a standard
Fmoc-protected amino acid, using DIEA as a base in dry DCM. After
capping of the resin with DCM/MeOH/DIEA (17:2:1), solid phase
peptide synthesis with HBTU/HOBt as coupling reagents was per-
formed. For all couplings but the N-terminal amino acid for which
2. Results
a
an N -Boc-protected amino acid was used, standard Fmoc-amino
2.1. Identification of peptide sequences for the design of DegP
substrates and inhibitors
acid building blocks were employed. This led to fully protected
precursor peptides that were cleaved from the solid support by
addition of acetic acid/trifluoroethanol/DCM (2:2:6) under preser-
vation of the side residue protecting groups. In a solution phase
To identify suitable peptide sequences for the design of sub-
strates or inhibitors for E. coli DegP, complete digests of substrates
O
Rn
O
O
H
H
N
H
N
SPPS
a
N
O
N
H
Boc
Cl
O
Fmoc
O
R1
O
Rn+1
R1
n
Rn
O
R1
O
Rn
O
H
N
H
H
N
H
N
b
c, d
H2N
N
N
H
N
H
Boc
N
H
OH
Rn+1
O
R2
O
Rn+1
O
R1
NO2
n
n
Scheme 1. Combined solid phase and solution synthesis of chromogenic DegP substrates on 2-chlorotrityl chloride resin. Reagents and conditions: (a) (i) Fmoc-AA-OH
(1.2 equiv), DIEA (4 equiv), DCM, rt, 2 h; (ii) DCM/MeOH/DIEA (17:2:1), rt, 10 min; (b) acetic acid/trifluoroethanol/DCM (2:2:6); (c) H-Ala-pNA ꢀ HCl or H-Val-pNA ꢀ HCl
(2 equiv), EDC ꢀ HCl (1.5 equiv), HOBt (2 equiv), DIEA (3 equiv); rt, 16 h; (d) TFA/H2O/ethandithiol/triisopropylsilane (94:2.5:2.5:1), rt, 2 h.
Table 1
Cleavage rate of pNA-based chromogenic DegP substrates and selectivity evaluation versus a panel of representative proteases
a
Substrate
Specific activity (nmol mgꢁ1 minꢁ1
)
DegP
Elastase
Subtilisin
Plasmin
Trypsin
Chymotrypsin
Thrombin
Kallikrein
SPMFKGV-pNA (1)
SDAEFRHDSGYEV-pNA (2)
STDGGV-pNA (3)
SGRVVPGYGHA-pNA (4)
SPLPEGV-pNA (5)
GLATGNVSTAELQDATPA-pNA (6)
KGKNSGSGATPV-pNA (7)
KGASVPGAGLV-pNA (8)
PMMGKASPV-pNA (9)
VFNTLPMMGKASPV-pNA (10)
PVFNTLPMMGKASPV-pNA (11)
SPAKGGEEPLPEGV-pNA (12)
Bz-R-pNA (13)
18
<1
<1
<1
<1
<1
<1
<1
1
32
18
<1
<1
473
7
8
5
3
2
<1
<1
2
<1
<1
<1
<1
<1
4
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
1
<1
<1
<1
<1
<1
26
5
<1
<1
8
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
3
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
24
<1
<1
2971
5
1521
34
<1
<1
<1
<1
8
56
49
24
74
6
96
89
69
<1
<1
4
3
<1
758
<1
<1
<1
<1
<1
1
DPMFKLV-pNA (14)
<1
a
Standard deviation was <15%.

2922
P. Hauske et al. / Bioorg. Med. Chem. 17 (2009) 2920–2924
coupling step, these peptides were coupled with commercially
available H-Ala-pNA ꢀ HCl or H-Val-pNA ꢀ HCl to yield fully pro-
tected chromogenic substrates which upon addition of a TFA-based
cleavage buffer were globally deprotected. Subsequent HPLC puri-
fication led to the desired chromogenic substrates 1-12 and 14.
Beside these synthesized chromogenic substrates, also the com-
mercial available Bz-Arg-pNA (13) was employed for later bio-
chemical testing.
the turnover rates by elastase and subtilisin were decreased (selec-
tivity of >100:1) (Table 1).
Interestingly, the panel wide evaluation of all synthesized pNA-
substrates allowed the identification of several elastase substrates
that differed in specificity and activity. For example, the potent
pNA-substrate 9 was rather specific (selectivity of >20:1). More-
over, substrate 10 also proved as a specific elastase substrate at
least when compared to the other 5 mammalian serine proteases
under investigation. pNA-substrate 5 was also specific when com-
paring elastase with the other 5 mammalian serine proteases
tested but the turnover rate was lower. In addition, peptides 4
and 6 were excellent substrates for subtilisin displaying rates of
2.3. Determination of cleavage rates of pNA-substrates and
selectivity profiling
The synthesized pNA peptides were evaluated for their poten-
tial to act as DegP substrates. To obtain initial evidence for their
selectivity, they were also tested in a panel of commercially avail-
able serine proteases, being elastase, subtilisin, plasmin, trypsin,
chymotrypsin, thrombin, and kallikrein.
about 3 and 1.5 l .
mol mgꢁ1 minꢁ1
2.4. Synthesis of chloromethyl ketone (CMK) based irreversible
DegP inhibitors
Following the optimization of assay conditions, all pNA-sub-
strates 1–13 were tested for cleavage efficiency at a concentration
of 0.5 mM (Table 1). As expected, most pNA-derivatives did not
show any biochemical activity, as predicted from the DegP bipar-
tite binding mode. However, three substrates, that is, 1, 10 and
11, with sufficient cleavage specificity were identified, of which
substrate 1 with a cleavage rate of 18 nmol mgꢁ1 minꢁ1 displayed
the best specificity versus the profiled proteases while substrate 10
showed the highest turnover with 32 nmol mgꢁ1 minꢁ1, albeight
with much lower specificity. As substrate 1 was the best short sub-
strate but was cleaved only with low rates, a derivative 14 with im-
proved features was synthesized by adapting its sequence to native
citrate synthase. In comparison with 1, DPMFKLV-pNA (14) had
two modifications. P7 was changed from Ser to Asp and P2 from
Gly to Leu. This modification increased DegP activity 26-fold at a
substrate concentration of 0.5 mM, the km decreased from 2.6 to
A series of CMK based inhibitors were synthesized some of
which might prove useful for biochemical and structural studies.
CMK derivatives 15–20 were prepared employing a subset of se-
quences used for the development of chromogenic substrates (Ta-
ble 2). Compound 15 was derived from 4; 16 is an extended version
of 15; 17 is a derivative of 1; 18 of 11; 19 of 10 and 20 of 9. Syn-
thesis involved coupling of the fully side chain protected peptides
derived from solid phase synthesis with either H-Ala-CMK ꢀ HCl or
H-Val-CMK ꢀ HCl to yield fully protected CMK precursors which
upon protecting group cleavage and subsequent HPLC purification
yielded the potential DegP chloromethylketone inhibitors 15–20.
H-Ala-CMK and H-Val-CMK were derived from solution synthe-
sis, starting from commercially available Boc-Ala-OH and Boc-Val-
OH. Briefly, these amino acid derivatives were activated by the
mixed anhydride method using isobutyl chloroformate in THF; fol-
lowed by addition of freshly prepared diazomethane in diethyl
ether. Subsequent addition of hydrogen chloride in dioxane re-
sulted in the generation of the desired chloromethyl ketones as a
hydrogen chloride salt (Scheme 2). These intermediates were then
0.6 mM, while Vmax increased from 1.2 l lmol -
mol minꢁ1 to 8.7
minꢁ1 and a stronger cooperativity raising from 2.6 to 5.1 was ob-
served. In addition, the specificity of the substrate was increased as
Table 2
Remaining proteolytic activity after inhibition of DegP and selected proteases by 0.5 mM of chloromethyl ketone based inhibitors
a
Inhibitor
Specific activity (nmol mgꢁ1 minꢁ1
)
DegP (%)
Elastase (%)
Subtilisin (%)
Plasmin (%)
Trypsin (%)
Chymotrypsin (%)
Thrombin (%)
Kallikrein (%)
SGRVVPGYGHA-CMK (15)
IWNTLNSGRVVPGTGHA-CMK (16)
SPMFKGV-CMK (17)
PVFNTLPMMGKASPV-CMK (18)
VFNTLPMMGKASPV-CMK (19)
PMMGKASPV-CMK (20)
102
101
3
63
57
49
83
60
<1
<1
<1
<1
15
15
<1
143
155
5
62
74
81
62
75
71
285
410
354
591
287
348
69
78
92
48
22
53
131
98
96
101
102
122
135
111
101
122
88
108
a
Specific activity was determined using SPMFKGV-pNA (DegP), KGKNSGSGATPV-pNA (elastase), SGRVVPGYGHA-pNA (subtilisin), Bz-R-pNA (plasmin, trypsin, thrombin
and kallikrein), and GLATGNVSTAELQDATPA-pNA (chymotrypsin) as substrates (final concentration 0.5 mM). Standard deviation was <18%.
R1
R1
R1
c
a, b
Boc
Boc
OH
N
H
N2
HCl*
H2N
Cl
N
H
O
O
O
O
O
Rn
R1
H
N
NH2
d, e
N
H
N
H
Cl
Rn+1
O
R2
O
n
Scheme 2. Synthesis of chloromethyl ketone inhibitors (R1 = CH3 or (CH3)2CH). Reagents and conditions: (a) i-butylchloroformate (1.05 equiv), N-methylmorpholine
(1.05 equiv), THF, ꢁ20 °C to 50 °C, 20 min; (b) (i) diazomethane (3 equiv), Et2O, 0 °C, 2 h; (ii) HOAc (10 equiv), 0 °C, 0.5 h; (c) 4 M HCl in dioxane, dioxane, 0 °C, 2 h; (d) SPPS-
derived peptides (0.5 equiv), EDC ꢀ HCl (0.75 equiv), HOBt (1 equiv), DIEA (1.5 equiv), DCM, rt, o/n; (e) TFA/H2O/ethandithiol/triisopropylsilane (94:2.5:2.5:1), rt, 2 h.

P. Hauske et al. / Bioorg. Med. Chem. 17 (2009) 2920–2924
2923
coupled with the fully protected peptides derived from solid phase
synthesis.
Future directions of research will include to exploit substrate 14
as an inhibitor scaffold using either chlormethyl ketone or boronic
acid derivatization. This inhibitor will allow us to address the open
question of how substrates bind to HtrA proteases using X-ray crys-
tallography. Such information is expected to be highly relevant for
inhibitor optimization, which is required for practical application.
2.5. Evaluation of the CMK inhibitors
The synthesized CMK derivatives were evaluated in two steps.
First, the residual proteolytic activity of several serine proteases
after pre-incubation with a 0.5 mM solution of CMKs 15–20 was
measured (Table 2). Of all serine proteases employed during this
pre-screening, only three proteases, namely elastase, subtilisin
and DegP were completely inhibited by some of the CMKs. Elastase
was inhibited by 17–20, subtilisin by 17 and DegP by 17. Inhibitor
17 was chosen for further characterisation. A titration experiment
was carried out using elastase, subtilisin and DegP revealing IC50 of
4. Experimental
4.1. General
All solvents and reagents were purchased from Fluka, Aldrich or
Acros and were generally of peptide synthesis grade or p.a. quality
and were used without further purification. Dry solvents were ob-
tained from Aldrich while Fmoc-amino acids were purchased from
IRIS biotech. H-Ala-pNA ꢀ HCl and H-Val-pNA ꢀ HCl were obtained
from Bachem while 2-chlorotrityl chloride resin was purchased
from Novabiochem.
1 nM for subtilisin, 75 lM for DegP and 220 lM for elastase.
3. Discussion
3.1. Substrates
LC-ESI-MS analysis was performed on a LCQ Advantage Max
system from Thermo Finnigan, coupled with the Agilent 1200
series analytical HPLC. If not stated explicitly, a standard gradient
of 10% acetonitrile in water (0.1% formic acid) to 100%
acetonitrile (0.1% formic acid) on a RP-C18 column in 10 min was
used for all analytical measurements. In certain cases, an alterna-
tive gradient (named gradient 2) of 5% acetonitrile in water (0.1%
formic acid) to 100% acetonitrile (0.1% formic acid) was employed.
Preparative HPLC purifications were performed on a Varian Pro-
This study identified a short but surprisingly potent DegP sub-
strate, DPMFKLV-pNA (14), increasing the specific activity of DegP
versus the similar substrate 1 by a factor of 26. This unexpected find-
ing suggests that a heptapeptide has the potential to convert DegP
from its resting hexameric state into the active and higher oligo-
meric state. The availability of a short pNA-substrate is expected
to facilitate the setup of a convenient enzyme assay for future high
throughput screens that could be instrumental in identifying small
molecule inhibitors of DegP. Such molecules could serve as leads
for a novel class of antibiotics targeting a key protein quality control
factor that was previously implicated in bacterial pathogenesis.3
Neutrophil elastase has an optimal pH of around 8 and the en-
zyme prefers small hydrophobic residues including Val, Ala and
Leu at P1 position. At P2, Pro is preferred, while substrates with
charged residues such as Lys, Arg and Asp are less well cleaved.
Our data correlate well with these observations as elastase cleaves
most substrates tested and the best ones have a Pro residue at P2.
As expected, not all substrates carrying a small hydrophobic resi-
due at P1 and a Pro at P2 are equally well processed as elastase
has several binding sites for substrate.13
Subtilisin has a broad substrate specificity with a preference for
aromatic or large nonpolar P1 and P4 substrate residues, even
though substrate binding determinants extend over nine amino
acids from P6 to P30.14 Interestingly, the two pNA-substrates with
an Ala at P1 were cleaved several hundred fold better compared
to the other pNA-substrates that had a Val residue at P1.
Chymotrypsin exhibits low specificity with a preference for Phe,
Tyr, Trp, Met or Leu at P1. This protease cleaves 4 of the 13 pNA-
substrates tested with GLATGNVSTAELQDATPA-pNA (6) being the
best substrate. Since kallikrein and thrombin prefer an Arg and
plasmin and trypsin a Lys or Arg at P1 position, respectively, the
synthesized pNA-substrates were not cleaved by these proteases.
Star HPLC, employing a 5
l C18 reverse phase HPLC column from
Macherey-Nagel and a gradient of 5% acetontrile in water (0.1%
TFA) to 60% acetonitrile in water (0.1% TFA) in 60 min.
The serine protease DegP was prepared following previously
published procedures.7 Chymotrypsin, trypsin, elastase from por-
cine pancreas, kallikrein, plasmin, subtilisin, and thrombin were
purchased from Sigma.
4.2. Synthesis of chromogenic substrates 1-12 and 14
Synthesis of chromogenic substrates 1–12 and 14 was per-
formed on 2-chlorotrityl resin (0.4 g, 0.56 mmol, 100–200 mesh,
1% DVB, maximal loading 1.4 mmol gꢁ1) as a solid support. Loading
of the first amino acid was performed with the corresponding ami-
no acid (1.2 equiv) and DIEA (4 equiv) in dry DCM for 2 h, followed
by capping with DCM/MeOH/DIEA (17:2:1, 5 mL). Initial loading
efficiency as determined by UV absorption was between
0.50 mmol gꢁ1 and 0.90 mmol gꢁ1. All reactions on solid support
were done in syringe reactors in batches of 0.2 mmol and the
amount of resin was adjusted according to the initial loading effi-
ciency. Fmoc cleavage during peptide synthesis was achieved by
addition of 20% piperidine in DMF for 20 min and was repeated
to drive the reaction to completeness. Further amino acid cou-
plings were performed with Fmoc-amino acids (4 equiv), HOBt
(4 equiv), HBTU (3.5 equiv) and DIEA (4 equiv) in DMF and a cou-
pling time of 2 h. Coupling efficiency of each step was verified by
a Kaiser test or in case of Fmoc-Pro-OH with a chloranil test. For
3.2. Inhibitors
a
the last coupling step, N -Boc-protected amino acids were used.
The CMK derivative 17 was identified inhibiting DegP and
should prove useful for future structural studies as to date there
is no crystal structure of any HtrA1 protease with bound substrate
or peptide inhibitor. Moreover, 4 of the 6 CMK derivatives inhibited
elastase at high concentrations. In addition, the screen identified
17 as a potent inhibitor of subtilisin (IC50 of 1 nM). Therefore, this
limited search for peptide based inhibitors yielded useful tools for
further experimental studies. Unexpectedly, most CMK derivatives
such as 18 activated trypsin by up to a factor of 6. Even though
similar findings were reported previously, the underlying mecha-
nism is not yet understood.15
Resin cleavage was achieved with AcOH/TFE/DCM (2:2:6, 5 mL)
to yield fully protected peptides which were used for solution
phase couplings without any further purification.
Finally, the protected peptides were coupled with H-Val-
pNA ꢀ HCl (2 equiv) or H-Ala-pNA ꢀ HCl (2 equiv), respectively,
and EDC ꢀ HCl (1.5 equiv), HOBt (2 equiv) and DIEA (3 equiv) in
DCM (15 mL). Some drops of DMF were added to facilitate dissolu-
tion of reagents if necessary. The resulting mixture was stirred over
night at room temperature, evaporated to dryness and the residue
was taken up with DCM (150 mL). The organic phase was extracted
with brine (50 mL), dried over Na2SO4 and evaporated to dryness.

2924
P. Hauske et al. / Bioorg. Med. Chem. 17 (2009) 2920–2924
Cleavage of remaining protecting groups was achieved by addition
of a mixture of TFA/H2O/EDT/TIS (94:2.5:2.5:1, 6 mL) for 2 h, fol-
lowed by precipitation of crude product by addition of cold dieth-
ylether. HPLC purification on a C18 reverse phase column with a
gradient of 5–10% acetonitrile in water (0.1% TFA) to 60% acetoni-
trile in water (0.1% TFA) in 60 min yielded the final chromogenic
substrates.
SPMFKGV-pNA (1): Yield = 130 mg (73%). LC–MS (ESI): tR
5.93 min, m/z 885.93 [M+H]+, 1770.33 [2M+H]+, 885.43 calcd for
C41H60N10O10S+.
(1–12 and 14) using instead of H-AA-pNA ꢀ HCl the freshly synthe-
sized H-AA-CMK ꢀ HCl building block.
SGRVVPGYGHA-CMK (15): Yield = 84 mg (37%). LC–MS (ESI): tR
8.09 min (gradient 2), m/z 1131.60 [M+H]+, 566.75 [M+2H]2+
,
þ
1131.55 calcd for C49H75ClN16O13
.
IWNTLNSGRVVPGTGHA-CMK (16): Yield = 94 mg (25%). LC–MS
(ESI): tR 4.78 min, m/z 1875.07 [M+H]+, 937.73 [M+2H]2+, 1872.93
þ
calcd for C84H126ClN25O22
.
SPMFKGV-CMK (17): Yield = 46 mg (29%). LC–MS (ESI): tR
5.38 min, m/z 797.40 [M+H]+, 1594.33 [2M+H]+, 797.38 calcd for
C36H57ClN8O8S+.
SDAEFRHDSGYEV-pNA (2): Yield = 138 mg (42%). LC–MS (ESI):
tR 5.43 min, m/z 1632.00 [M+H]+, 817.00 [M+2H]2+, 1631.67 calcd
PVFNTLPMMGKASPV-CMK (18): Yield = 54 mg (17%). LC–MS
for C70H94N20O26
.
(ESI): tR 6.21 min, m/z 1620.63 [M+H]+, 811.25 [M+2H]2+, 1620.81
þ
þ
STDGGV-pNA (3): Yield = 106 mg (82%). LC–MS (ESI): tR
calcd for C73H118ClN17O18S2
.
5.57 min, m/z 655.53 [M+H]+, 1309.60 [2M+H]+, 655.27 calcd for
VFNTLPMMGKASPV-CMK (19): Yield = 61 mg (20%). LC–MS
C26H38N8O12
.
(ESI): tR 6.21 min, m/z 1523.57 [M+H]+, 762.66 [M+2H]2+, 1523.75
þ
þ
SGRVVPGYGHA-pNA (4): Yield = 146 mg (60%). LC–MS (ESI): tR
calcd for C68H111ClN16O17S2
.
4.73 min, m/z 1220.00 [M+H]+, 611.00 [M+2H]2+, 1219.60 calcd
PMMGKASPV-CMK (20): Yield = 63 mg (34%). LC–MS (ESI): tR
for C54H78N18O15
.
5.47 min, m/z 949.36 [M+H]+, 475.46 [M+2H]2+, 949.44 calcd for
þ
þ
SPLPEGV-pNA (5): Yield = 107 mg (66%). LC–MS (ESI): tR
C40H69ClN10O10S2
.
5.90 min, m/z 818.60 [M+H]+, 1635.73 [2M+H]+, 818.41 calcd for
þ
C37H55N9O12
.
4.4. Enzyme assay conditions
GLATGNVSTAELQDATPA-pNA (6): Yield = 167 mg (45%). LC–MS
(ESI): tR 5.79 min, m/z 1836.87 [M+H]+, 919.00 [M+2H]2+, 1835.88
After optimizing assay conditions, the following buffers were
used for measuring the proteolytic activities: DegP: 50 mM
NaH2PO4, pH 8.0; chymotrypsin and trypsin: 80 mM Tris, 20 mM
CaCl2, pH 8.0; elastase: 100 mM Tris, pH 8.0; kallikrein: 20 mM
Tris, 100 mM NaCl, pH 7.8; plasmin: 50 mM NaH2PO4, 150 mM
NaCl, pH 8.5; subtilisin: 50 mM K2HPO4, pH 7.5; thrombin:
50 mM NH4HCO3, pH 8.0.
þ
calcd for C77H122N22O30
.
KGKNSGSGATPV-pNA (7): Yield = 97 mg (40%). LC–MS (ESI): tR
4.89 min, m/z 1222.87 [M+H]+, 612.07 [M+2H]2+, 1222.62 calcd
þ
for C51H83N17O18
.
KGASVPGAGLV-pNA (8): Yield = 103 mg (48%). LC–MS (ESI): tR
5.75 min, m/z 1075.87 [M+H]+, 1075.59 calcd for C48H78N14O14
.
þ
PMMGKASPV-pNA (9): Yield = 76 mg (37%). LC–MS (ESI): tR
5.59 min, m/z 1037.41 [M+H]+, 519.42 [M+2H]2+, 1037.49 calcd
4.5. Determination of cleavage efficiency of chromogenic
substrates 1–14
þ
for C45H72N12O12S2
.
VFNTLPMMGKASPV-pNA (10): Yield = 87 mg (27%). LC–MS
(ESI): tR 6.06 min, m/z 1611.80 [M+H]+, 1633.93 [M+Na]+, 806.53
For determination of the cleavage rate, 5
lg of the corresponding
[M+2H]2+, 1611.80 calcd for C73H114N18O19S2
.
protease was dissolved in 100 L buffer. Chromogenic substrates
l
þ
PVFNTLPMMGKASPV-pNA (11): Yield = 104 mg (30%). LC–MS
(1–14) were added to a final concentration of 0.5 mM and substrate
cleavage was monitored at 405 nm for 60 min. The determined spe-
cific activities (sa) are derived from at least three independent
(ESI): tR 6.19 min, m/z 1708.60 [M+H]+, 855.33 [M+2H]2+, 1708.86
þ
calcd for C78H121N19O20S2
.
SPAKGGEEPLPEGV-pNA (12): Yield = 160 mg (54%). LC–MS
duplicate measurements and calculated from sa =
(m ꢀ ꢀ t) ( OD405: change of absorption at k = 405 nm over
60 min; V: final volume of reaction, m: amount of protease (mg),
: molar extinction coefficient of para-nitroaniline).
D
OD405 ꢀ V/
(ESI): tR 5.61 min, m/z 1487.67 [M+H]+, 744.47 [M+2H]2+, 1486.72
e
D
þ
calcd for C65H99N17O23
.
DPMFKLV-pNA (14): Yield = 158 mg (82%). LC–MS (ESI): tR
6.76 min, m/z 969.80 [M+H]+, 1938.73 [2M+H]+, 969.49 calcd for
C46H68N10O11S+.
e
References and notes
1. Hasenbein, S.; Merdanovic, M.; Ehrmann, M. Genes Dev. 2007, 21, 6.
2. (a) Lipinska, B.; Fayet, O.; Baird, L.; Georgopoulos, C. J. Bacteriol. 1989, 171,
1574; (b) Strauch, K. L.; Johnson, K.; Beckwith, J. J. Bacteriol. 1989, 171,
2689.
3. Raivio, T. L. Mol. Microbiol. 2005, 56, 1119.
4. (a) Clausen, T.; Southan, C.; Ehrmann, M. Mol. Cell 2002, 10, 443; (b) Pallen, M.
J.; Wren, B. W. Mol. Microbiol. 1997, 26, 209.
5. Krojer, T.; Garrido-Franco, M.; Huber, R.; Ehrmann, M.; Clausen, T. Nature 2002,
416, 455.
6. Krojer, T.; Sawa, J.; Schäfer, E.; Saibil, H. R.; Ehrmann, M.; Clausen, T. Nature
2008, 453, 885.
4.3. Synthesis of chloromethyl ketone inhibitors 15–20
The valine chloromethyl ketone building block was freshly pre-
pared, adapting a previously published method.16 Boc-Val-OH
(15.2 g, 70 mmol) was activated with iso-butylchloroformate
(9.53 mL, 73.5 mmol, 1.05 equiv) and N-methlymorpholine
(8.08 mL, 73.5 mmol, 1.05 equiv) in dry THF (100 mL) at ꢁ20 °C un-
der an argon atmosphere for 20 min and was then allowed to
warm up to 5 °C. Addition of freshly prepared diazomethane
(8.83 g, 200 mmol, 3 equiv) in diethylether (200 mL), followed by
careful addition of acetic acid to destroy excess diazomethane after
5 h yielded after evaporation of the solvents Boc-Val-CH2N2 as a
yellow oil. This intermediate was then redissolved in dry dioxane
(50 mL) and cooled to 0 °C. 4 M HCl in dioxane (2 mL) was slowly
added and the resulting mixture was stirred for 2 h. Addition of
cold diethylether led to precipitation of H-Val-CMK ꢀ HCl (5.27 g,
28.3 mmol, 40% yield) which was used without further purification
in the next coupling step.
7. Spiess, C.; Beil, A.; Ehrmann, M. Cell 1999, 97, 339.
8. Meltzer, M.; Hasenbein, S.; Hauske, P.; Kucz, N.; Merdanovic, M.; Grau, S.; Beil,
A.; Jones, D.; Krojer, T.; Clausen, T.; Ehrmann, M.; Kaiser, M. Angew. Chem., Int.
Ed. 2008, 47, 1332.
9. Hauske, P.; Ottmann, C.; Meltzer, M.; Ehrmann, M.; Kaiser, M. Chembiochem
2008, 9, 2920.
10. Krojer, T.; Pangerl, K.; Kurt, J.; Sawa, J.; Stingl, C.; Mechtler, K.; Huber, R.;
Ehrmann, M.; Clausen, T. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 7702.
11. Kolmar, H.; Waller, P. R. H.; Sauer, R. T. J. Bacteriol. 1996, 178, 5925.
12. Tonello, F.; Ascenzi, P.; Montecucco, C. J. Biol. Chem. 2003, 278, 40075.
13. Korkmaz, B.; Moreau, T.; Gauthier, F. Biochimie 2008, 90, 227.
14. Perona, J. J.; Craik, C. S. Protein Sci. 1995, 4, 337.
15. Radau, G.; Rauh, D. Bioorg. Med. Chem. Lett. 2000, 10, 779.
16. Kato, D.; Boatright, K. M.; Berger, A. B.; Nazif, T.; Blum, G.; Ryan, C.; Chehade, K.
A. H.; Salvesen, G. S.; Bogyo, M. Nat. Chem. Biol. 2005, 1, 33.
The chloromethly ketone peptides (15–20) were then generated
in an analogues manner to the synthesis of the peptide substrates