Cleavage of peptide bonds bearing ionizable amino acids at P1 by serine proteases with hydrophobic S1 pocket

Article abstract of DOI:10.1016/j.bbrc.2010.08.078

Enzymatic hydrolysis of the synthetic substrate succinyl-Ala-Ala-Pro-Xxx-pNA (where Xxx=Leu, Asp or Lys) catalyzed by bovine chymotrypsin (CHYM) or Streptomyces griseus protease B (SGPB) has been studied at different pH values in the pH range 3-11. The pH

Full text of DOI:10.1016/j.bbrc.2010.08.078

Biochemical and Biophysical Research Communications 400 (2010) 507–510
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications
journal homepage: www.elsevier.com/locate/ybbrc
Cleavage of peptide bonds bearing ionizable amino acids at P1 by serine proteases
with hydrophobic S1 pocket
Mohammad A. Qasim a, , Jikui Song b,1, John L. Markley b, Michael Laskowski Jr. c,
⇑
a Department of Chemistry, Indiana University Purdue University Fort Wayne, 2101 E. Coliseum Blvd., Fort Wayne, IN 46805, USA
b Department of Biochemistry, University of Wisconsin—Madison, WI 53706, USA
c Department of Chemistry, Purdue University, West Lafayette, IN 47907, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 3 August 2010
Available online 26 August 2010
Enzymatic hydrolysis of the synthetic substrate succinyl-Ala-Ala-Pro-Xxx-pNA (where Xxx = Leu, Asp or
Lys) catalyzed by bovine chymotrypsin (CHYM) or Streptomyces griseus protease B (SGPB) has been stud-
ied at different pH values in the pH range 3–11. The pH optima for substrates having Leu, Asp, and Lys
have been found to be 7.5–8.0, 5.5–6.0, and ꢀ10, respectively. At the normally reported pH optimum
(pH 7–8) of CHYM and SGPB, the substrate with Leu at the reactive site is more than 25,000-fold more
reactive than that with Asp. However, when fully protonated, Asp is nearly as good a substrate as Leu.
The pK values of the side chains of Asp and Lys in the hydrophobic S1 pocket of CHYM and SGPB have
been calculated from pH-dependent hydrolysis data and have been found to be about 9 for Asp and
7.4 and 9.7 for Lys for CHYM and SGPB, respectively. The results presented in this communication suggest
a possible application of CHYM like enzymes in cleaving peptide bonds contributed by acidic amino acids
between pH 5 and 6.
Keywords:
Chymotrypsin,
Enzyme kinetic
pK shift
Serine protease
Standard Mechanism inhibitor
Streptomyces griseus protease B
Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction
enzyme [9,10]. This is not the case with substrates, which can read-
ily shift their scissile peptide bond (their reactive site) so as to un-
Serine proteases are ubiquitous and exist in a wide spectrum of
specificities [1,2] with the capability of hydrolyzing virtually any
peptide bond. Serine proteases that preferentially hydrolyze pep-
tide bonds contributed by hydrophobic (Ala, Val, Leu, Ile, Met,
Phe, Tyr, Trp), cationic (Lys, Arg), anionic (Asp, Glu), and polar
(Ser, Gln) side chains at P1 (Schechter–Berger nomenclature [3])
as well as proline at P1 are known to exist [1]. Of these specific clas-
ses of serine proteases, the ones that hydrolyze at hydrophobic and
cationic amino acids at P1 are the best characterized.
A great deal has been learned about the size, shape, and chem-
istry of the primary specificity pocket (S1 pocket) of serine prote-
ases by studying their interaction with Standard Mechanism
protein inhibitors [4–6]. The inhibitor–protease complex mimics
the substrate protease transition state complex, and, therefore,
the good correlation often found [7,8] between association equilib-
rium constant (Ka) values for a set of inhibitor variants and the cor-
responding kcat/Km values for substrates is not surprising. In
Standard Mechanism inhibitors the reactive site peptide bond
never shifts, even if the P1 residue is deleterious for the target
dergo cleavage according to the specificity of the serine protease.
Thus serine proteases with a hydrophobic S1 pocket, such as bovine
chymotrypsin (CHYM) and Streptomyces griseus protease B (SGPB),
have not been reported to hydrolyze peptide bonds in which the P1
residue is contributed by Asp or Glu. In contrast to substrates, Stan-
dard Mechanism inhibitors with a P1 Asp or Glu are able to bind,
albeit weakly, with CHYM or SGPB [4]. Our investigations on the
binding of serine proteases to variants of turkey ovomucoid third
domain (OMTKY3) containing all possible ionizable residues at P1
led to the finding of large shifts of the pKa values of these P1 resi-
dues when in the inhibitor–protease complex [11]. We ask a sim-
ilar question here for the hydrolysis of synthetic substrates bearing
Asp and Lys at the reactive site by two serine proteases that have a
neutral hydrophobic S1 pocket, CHYM and SGPB. The pH-depen-
dence of the hydrolysis of these substrates reveals interesting re-
sults that are discussed in this paper.
2. Materials and methods
TPCK treated bovine chymotrypsin was obtained from Sigma.
SGPB was purified from a commercial (Sigma) preparation of
Pronase as described earlier [12]. The purity of SGPB was confirmed
by ion exchange and size exclusion column chromatographies, while
its identity was confirmed by amino acid composition determination.
⇑
Corresponding author. Fax: +1 260 481 6070.
E-mail address: qasimm@ipfw.edu (M.A. Qasim).
Present address: Structural Biology Program, Memorial Sloan-Kettering Cancer
1
Center, New York, NY 10032, USA.
Deceased.
0006-291X/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.bbrc.2010.08.078

508
M.A. Qasim et al. / Biochemical and Biophysical Research Communications 400 (2010) 507–510
Synthetic protease substrates, such as Suc-Ala-Ala-Pro-Leu-pNA
(Suc-AAPL-pNA), Suc-Ala-Ala-Pro-Asp-pNA (Suc-AAPD-pNA), and
Suc-Ala-Ala-Pro-Lys-pNA (Suc-AAPK-pNA), were purchased from
BACHEM. Various buffers and solutions used with their pH ranges
in parentheses were: 0.1 M NaCl–NaOH (pH 9.5–11.0), 0.1 M
Tris–HCl (pH 7.0–9.0), 0.1 M Bis–Tris (pH 6.0–7.0), 0.1 M sodium
acetate–acetic acid (pH 4.0–5.5), 0.1 M glycine buffer (pH 3.0–3.5).
Stock solutions of 100 mM substrates were prepared in dimethyl
sulfoxide (DMSO).
Kinetics of enzymatic hydrolysis of substrates were performed
at different pH values and at 22 °C by following the change in
absorbance over a period of 120–300 s in the wavelength range
380–410 nm on a Hewlett–Packard HP8453 diode array spectro-
photometer. The kinetic data were automatically corrected for
any light scattering by continuously subtracting the absorbance
values in the wavelength range 650–700 nm during the kinetic
run. The rates of hydrolysis of substrates obtained as change in
absorbance per second were converted to moles of product (p-
nitroaniline) produced per second (Vo) by using a molar extinction
coefficient of 8800 Mꢁ1 cmꢁ1 [13] for p-nitroaniline. Km and kcat
values were calculated from the rates of hydrolysis of substrates
at different substrate concentrations using the Lineweaver–Burk
equation. The expression of the P1Q variant of OMTKY3 and the
procedure for the measurement of the association equilibrium con-
stant are described in an earlier publication [4].
neutral hydrophobic side chains (such as Ala, Phe, Val) in place
of Leu gave results similar to those shown in Fig. 1A.
The pH-dependence of the association equilibrium constant, Ka,
of inhibitors with non-ionizable residues at P1 has the same shape
as that for the pH-dependence of the hydrolysis of Suc-AAPL-pNA
(Fig. 1A). The pH-dependence of Ka of P1 variants of OMTKY3 with
different serine proteases has been extensively studied in our lab-
oratory at Purdue. As an example, the pH-dependence of Ka of P1Q
variant of OMTKY3 with CHYM and SGPB is shown here (Fig. 1B). The
strong correlation between free energy of association of inhibitors
(log Ka) and transition state free energies of substrates (log kcat/Km)
has been observed by us as well as by others [4,8,15,16].
The two serine proteases used here have a hydrophobic S1 pock-
et and therefore have a preference for amino acids with hydropho-
bic side chains at P1. Thus OMTKY3 variants with Asp, Glu, Lys or
Arg at P1 bind weakly to these serine proteases [4]. We performed
pH-dependent measurements of association equilibrium constants
of such inhibitors with different serine proteases and found that
when a charged side chain is present at P1, it undergoes large pK
shifts in the direction that makes it neutral in the S1 pocket. Thus
Asp and Glu undergo large increases in their pKa values whereas
Lys undergoes a decrease in its pKa value [11,15].
The pH-dependence of kcat/Km determined for Suc-AAPD-pNA
and Suc-AAPK-pNA with CHYM and SGPB (Fig. 2) shows a large
shift in the optimum pH with both enzymes. When Asp is present
at the scissile peptide bond, the optimum shifts to pH 5.0–5.5,
whereas when Lys is present, the pH optimum is 10 or higher.
These shifts are consistent with the binding of Asp and Lys in their
neutral state. Suc-AAPD-pNA is an extremely poor substrate for
CHYM and SGPB at pH 8 but becomes a moderately good substrate
(ꢀ100-fold better) at pH 5.5. The pH-dependence of Asp and Lys
substrates can be used to determine the pKa values of these two
residues when they are inserted in the S1 pocket of the enzyme
(designated as pKc) in the transition state complex. We have used
a procedure we initially described for inhibitors [11] to determine
the pKc values of Asp and Lys in the substrates. In this procedure an
assumption is made that the difference in pH-dependence of log k-
cat/Km of two similar substrates, one with an ionizable side chain at
P1 and the other with a non-ionizable side chain at P1, can solely be
explained by the pKa value of ionizable group in free-state (pKf)
and in bound state (pKc). Free and bound states refer, respectively,
to the substrate that is not bound to a protease and to the substrate
that is bound to a protease. We used Suc-AAPL-pNA with Leu at P1
as our representative non-ionizable P1 residue. As an example,
plots obtained with CHYM for Suc-AAPD-pNA and Suc-AAPK-pNA
are shown in Fig. 3. In these plots, the difference in log kcat/Km val-
ues between Suc-AAPL-pNA and Suc-AAPD-pNA (or Suc-AAPK-
pNA) designated as log R is plotted against pH. The pKf values for
3. Results and discussion
Hydrolysis of Suc-AAPL-pNA by CHYM and SGPB was studied at
different pH values between pH 4.0 and pH 10.0 (Fig. 1A). CHYM
and many other serine proteases show maximum enzymatic activ-
ity between pH 7 and 8. The decline in enzyme activity below pH 7
is well known and has been attributed to protonation of catalytic
histidine 57 (chymotrypsinogen numbering). The decrease in
activity above pH 8 can be attributed to the deprotonation of the
a-amino group of Ile16 and the disruption of charge–charge inter-
action between Ile16 and the side chain carboxyl group of Asp194
[14]. The salt bridge between Ile16 and Asp194 is characteristic of
CHYM like enzymes but is either absent in other serine proteases
or is substituted by a different type of salt bridge. For example,
SGPB, a bacterial serine protease, has a salt bridge involving
Asp194 and the side chain of Arg138 [14]. Because the arginine
side chain has a very high pKa value (>12), the interaction between
Asp194 and Arg138 in SGPB is unlikely to be disrupted significantly
below pH 11. This is consistent with the observation that the pH-
dependence of kcat/Km for Suc-AAPL-pNA with SGPB shows a flat
region between pH 8 and pH 10 (Fig. 1A). Substrates with other
6
5
4
3
2
1
0
10
8
6
4
2
3
5
7
9
11
2
4
6
8
10
12
pH
pH
Fig. 1. (A) log kcat/Km plot for Suc-AAPL-pNA as a function of pH for CHYM (N) and for SGPB (j). (B) log Ka plot for P1Q variant of OMTKY3 for CHYM (N) and SGPB (d).

M.A. Qasim et al. / Biochemical and Biophysical Research Communications 400 (2010) 507–510
509
2.5
2
5
4
3
2
1
1.5
1
0.5
0
3
5
7
9
11
-0.5
2
4
6
8
10
12
pH
pH
Fig. 2. pH-dependence of log kcat/Km for Suc-AAPD-pNA (N) and Suc-AAPK-pNA (j) for SGPB (A) and CHYM (B).
5
4
3
2
1
0
complex with CHYM, Lys is a stronger acid (pKc = 7.44) than Asp
(pKc = 8.92). The pK shifts of the Asp side chain were nearly identi-
cal for CHYM and SGPB (pKc ꢁ pKf =
DpK = 5.0 0.1) and represent
one of the largest shifts reported for acidic amino acid side chains
in proteins [11,18]. The pK shifts for the Lys side chain were quite
different for CHYM and SGPB (see Table 1). The probable cause of
this difference is the flexibility of the lysine side chain and its abil-
ity to adopt different conformations in the S1 pocket of proteases.
In an earlier study, we found that the pK shift for the P1 lysine of
BPTI is much smaller than that for the P1 Lys variant of OMTKY3
in complex with CHYM [15]. The molecular explanation emerged
from comparison of the X-ray crystallographic structures of the
complexes with CHYM of BPTI [19,20] and P1 Lys–OMTKY3 (1hja,
unpublished). In the P1 Lys-OMTKY3–CHYM complex, the P1 Lys
side chain is inserted deep into S1 pocket of CHYM, whereas in
the BPTI–CHYM complex, the P1 Lys side chain goes into the S1
3
5
7
9
11
pH
pocket but bends back in such a manner that the e-amino group
Fig. 3. pH-dependence of log R for Suc-AAPD-pNA (N) and Suc-AAPK-pNA (j) for
CHYM. For details, see text and Eq. (1).
is able to make a couple of hydrogen bonds with the backbone oxy-
gens of Ser217 of the enzyme and Pro13 (P3) of the inhibitor. The dif-
ference in the microenvironment of Lys in the BPTI–CHYM
complex and in the P1 Lys–OMTKY3–CHYM complex accounts for
the difference in the pKc value of these two lysine side chains. Thus
it is likely that a similar mechanism accounts for the difference in
the pKc values of Lys in Suc-AAPK-pNA in its transition state com-
plexes with CHYM and SGPB.
Suc-AAPD-pNA and Suc-AAPK-pNA were determined by NMR
spectroscopy [17]. The plot is fitted to the equation we developed
earlier [11] for the pH-dependence of association equilibrium con-
stants for the interaction of P1 variants of OMTKY3 with serine
proteases:
log R ¼ log R0 þ logð1 þ 10ðpH—pK ÞÞ ꢁ logð1 þ 10ðpH—pK ÞÞ
ð1Þ
c
Why chymotrypsin-like enzymes are not reported to hydrolyze
peptide bonds contributed by Asp and Glu is obviously due to ex-
tremely poor activity of these enzymes towards such peptides in
the pH range 7–8. For example, at pH 8, kcat/Km for the hydrolysis
of Suc-AAPD-pNA by CHYM was found to be 0.74 Mꢁ1 sꢁ1 com-
pared to 25,000 Mꢁ1 sꢁ1 for Suc-AAPL-pNA. This is because in order
bind to the S1 pocket of CHYM, the Asp side chain must become
protonated—energetically an expensive process at pH above 7.
However, at low pH values (pH 4), kcat/Km for Suc-AAPD-pNA ap-
proaches that of Suc-AAPL-pNA (10.7 Mꢁ1 sꢁ1 vs 28.2 Mꢁ1 sꢁ1).
This is also indicated by a near zero log R value at low pH values
(log R value of zero means that kcat/Km for Suc-AAPD-pNA and
Suc-AAPL-pNA are equal).
f
In this equation, R0 is the ratio of kcat/Km values for an ionizable
residue to a non-ionizable residue at a pH where the ionizable res-
idue is completely protonated and pKf and pKc are the pKa values of
the ionizable group in the substrate and enzyme–substrate transi-
tion state, respectively.
Fitting the data according to the relation given above allowed us
to calculate the pKc values for the ionizable residues, Asp and Lys,
in the transition state complex of the substrates with the enzymes
(Table 1). It is interesting to note that in the transition state
Table 1
pKc values of Asp and Lys side chains in the S1 pockets of CHYM and SGPB.
The pKc values reported here are very similar to the one found
for P1 Asp and Lys variants of OMTKY3 in complex with SGPB,
CHYM, porcine pancreatic elastase and subtilisin Carlsberg
[11,15] (Qasim and Laskowski, unpublished). Thus, the strong cor-
relation reported between association equilibrium constants in
inhibitors and kcat/Km values in substrates at pH 8.3 [4,8,15,16]
for P1 variants extends over the whole ionizable range of pH 3–11.
The results presented here clearly show that the pH optimum of
a neutral serine protease is dependent on the nature of the side
chain at the scissile peptide bond. The large preferential increase
in the catalytic activity of neutral serine proteases for acidic amino
Substrates
pK values
CHYM
SGPB
Suc-AAPD-pNA
pKfa = 3.84
pKc
8.92
5.08
8.72
4.88
D
pK
Suc-AAPK-pNA
pKfa = 10.51
pKc
7.44b
3.07
9.71
0.80
D
pK
a
pKf values were determined by NMR as reported earlier [17].
Qasim et al. [15] determined pKc using Suc-AAPA-pNA as reference and found it
to be 7.35. In this study, Suc-AAPL-pNA was used as reference.
b

510
M.A. Qasim et al. / Biochemical and Biophysical Research Communications 400 (2010) 507–510
[8] M. Laskowski Jr., M.A. Qasim, What can the structures of enzyme–inhibitor
complexes tell us about the structures of enzyme substrate complexes?,
Biochim Biophys. Acta 1477 (2000) 324–337.
acid side chains at pH 5.0–5.5 may have a physiological role. For
example, granulocytes are rich in chymotrypsin-like serine prote-
ases and have a pH between 5.0 and 6.0 [21]—a pH range well sui-
ted for cleavage at peptide bonds contributed by Asp and Glu.
Regardless of any physiological role of the action of serine prote-
ases on acidic amino acids at low pH, the results presented here
suggest that such proteases can be used for cleavage at acidic ami-
no acids by performing the cleavage at an appropriate low pH.
[9] W. Ardelt, M. Laskowski Jr., Turkey ovomucoid third domain inhibits eight
different serine proteinases of varied specificity on the same . . . Leu18
Glu19 . . . reactive site, Biochemistry 24 (1985) 5313–5320.
-
[10] M. Laskowski Jr., M.A. Qasim, Z.-P. Yi, Additivity-based prediction of
equilibrium constants for some protein–protein associations, Curr. Opin.
Struct. Biol. 13 (2003) 130–139.
[11] M.A. Qasim, M.R. Ranjbar, R. Wynn, S. Anderson, M. Laskowski Jr., Ionizable P1
residues in serine proteinase inhibitors undergo large pK shift on complex
formation, J. Biol. Chem. 270 (1995) 27419–27422.
[12] M.A. Qasim, Streptogrisin B, in: A.J. Barrett, N.D. Rawlings, J.F. Woessner (Eds.),
Handbook of Proteolytic Enzymes, second ed., Elsevier Academic Press, 2004,
pp. 1458–1462.
Acknowledgments
[13] A.J. Barrett, Cathepsin G, Methods Enzymol. 80 (Pt. C) (1981) 561–565.
[14] A.R. Khan, M.N.G. James, Molecular mechanisms for the conversion of
zymogens to active proteolytic enzymes, Protein Sci. 7 (1998) 815–836.
[15] M.A. Qasim, S.M. Lu, J.H. Ding, K.S. Bateman, M.N.G. James, S. Anderson, J.K.
Song, J.L. Markley, P.J. Ganz, C.W. Saunders, M. Laskowski Jr., Thermodynamic
criterion for the conformation of P1 residues of substrates and of inhibitors in
complexes with serine proteinases, Biochemistry 38 (1999) 7142–7150.
[16] H. Czapinska, J. Otlewski, Structural and energetic determinants of the S1-site
specificity in serine proteases, Eur. J. Biochem. 260 (1999) 571–595.
[17] J. Song, M. Laskowski Jr., M.A. Qasim, J.L. Markley, NMR determination of pKa
values for Asp, Glu, His, and Lys mutants at each variable contiguous enzyme–
inhibitor contact position of the turkey ovomucoid third domain, Biochemistry
42 (2003) 2847–2856.
[18] V.L. Laurents, B.M.P. Huyghues-Despointes, M. Bruix, R.L. Thurlkill, D. Schell, S.
Newsom, G.R. Grimsley, K.L. Shaw, S. Trevino, M. Rico, J.M. Birggs, J.M.
Antosiewicz, J.M. Scholtz, C.N. Pace, Charge–charge interactions are key
determinants of the pK values of ionizable groups in ribonuclease Sa
(pI = 3.5) and a basic variant (pI = 10.2), J. Mol. Biol. 325 (2003) 1077–1092.
[19] C. Capasso, M. Rizzi, E. Menegatti, P. Ascenzi, M. Bolognesi, Crystal structure of
Financial support was provided by an NIH Grant GM 10831 to
M.L. M.A.Q. was a recipient of a summer Faculty Research Grant
from IPFW. We thank Dr. Ronald S. Friedman for careful reading
of the manuscript and for suggesting many changes.
References
[1] E. Di Cera, Serine proteases, IUBMB Life 61 (2009) 510–515.
[2] L. Hedstrom, Serine protease mechanism and specificity, Chem. Rev. 102
(2002) 4501–4523.
[3] I. Schechter, A. Berger, On the size of active site in proteases: I. Papain,
Biochem. Biophys. Res. Commun. 27 (1967) 157–162.
[4] W. Lu, I. Apostol, M.A. Qasim, N. Warne, R. Wynn, W.-L. Zhang, S. Anderson, Y.-
W. Chiang, E. Ogin, I. Rothberg, K. Ryan, M. Laskowski Jr., Binding of amino acid
side chains to S1 cavities of serine proteinases, J. Mol. Biol. 266 (1997) 441–
461.
[5] K.S. Bateman, S. Anderson, W. Lu, M.A. Qasim, M. Laskowski Jr., M.N.G. James,
Deleterious effects of b-branched residues in the S1 specificity pocket of
Streptomyces griseus proteinase B: crystal structures of the turkey ovomucoid
the bovine a-chymotrypsin:Kunitz inhibitor complex: an example of multiple
protein:protein recognition sites, J. Mol. Recognit. 10 (1997) 26–35.
[20] A.J. Scheidig, T.R. Hynes, L.A. Pelletier, J.A. Wells, A.A. Kossiakoff, Crystal
structure of bovine chymotrypsin and trypsin complexed to the inhibitor
domain of Alzheimer’s amyloid b-protein precursor (APPI) and basic
pancreatic trypsin inhibitor (BPTI): engineering of inhibitors with altered
specificities, Protein Sci. 6 (1997) 1806–1824.
third domain variants Ile18I Val18I Thr18I and Ser18I in complex with
, , ,
Streptomyces griseus proteinase B, Protein Sci. 9 (2000) 83–94.
[6] H. Czapinska, R. Helland, A.O. Smalås, J. Otlewski, Crystal structures of five
bovine chymotrypsin complexes with P1 BPTI variants, J. Mol. Biol. 344 (2004)
1005–1020.
[21] M.B. De Young, E.F. Nemeth, A. Scarpa, Measurement of the internal pH of mast
cell granules using microvolumetric fluorescence and isotopic techniques,
Arch. Biochem. Biophys. 254 (1987) 222–233.
[7] M. Laskowski Jr., M.A. Qasim, S.M. Lu, Interaction of standard mechanism,
canonical protein inhibitors with serine proteinases, in: C. Kleanthous (Ed.),
Protein–Protein Recognition, Oxford University Press, 2000, pp. 228–279.