Download Mutations of the PC2 Substrate Binding Pocket Alter Enzyme

THE JOURNAL OF BIOLOGICAL CHEMISTRY
© 2005 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 280, No. 36, Issue of September 9, pp. 31850 –31858, 2005
Printed in U.S.A.
Mutations of the PC2 Substrate Binding Pocket
Alter Enzyme Specificity*
Received for publication, May 20, 2005, and in revised form, July 6, 2005
Published, JBC Papers in Press, July 7, 2005, DOI 10.1074/jbc.M505567200
Magdalena M. Kacprzak‡, Manuel E. Than§, Luiz Juliano¶, Maria A. Juliano¶, Wolfram Bode§,
and Iris Lindberg‡储
From the ‡Department of Biochemistry and Molecular Biology, Louisiana State University Health Sciences Center,
New Orleans, Louisiana 70112, the §Department of Structural Research, Max-Planck-Institute for Biochemistry,
Am Klopferspitz 18, 82152 Martinsried, Germany, and the ¶Department of Biophysics, Escola Paulista de Medicina,
Rua Tres de Maio 100, 04044-020, Sao Paulo, Brazil
Prohormone convertase (PC)1 2 belongs to the family of
mammalian calcium-dependent serine proteases known as subtilisin/kexin-like proprotein convertases. The seven members of
this family, furin, PC2, PC1/3, PACE4, PC4, PC5/6, and PC7/8,
are the major endoproteolytic processing enzymes of the secretory pathway (1). PCs are multidomain proteins, all consisting
* This work was supported by Grant DA05084 from the National Institutes of Health (to I. L.), a grant from the Fundação de Amparo Pesquisa do Estado de São Paulo and Conselho Nacional de Desenvolvimento
Cientı́fico e Tecnológico (to L. J. and M. A. J.), and grants from the Deutsche
Forschungsgemeinschaft TH 862/1-1,3 (to M. E. T.). The costs of publication
of this article were defrayed in part by the payment of page charges. This
article must therefore be hereby marked “advertisement” in accordance with
18 U.S.C. Section 1734 solely to indicate this fact.
储 To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology, Louisiana State University, 1901 Perdido
St., New Orleans, LA 70112. Tel.: 504-568-4799; Fax: 504-568-2973;
E-mail: [email protected].
1
The abbreviations used are: PC, proprotein/prohormone convertase;
CT peptide, C-terminal peptide; D6R, hexa-D-arginine; D9R, nona-Darginine; PE, proenkephalin; pERTKR-MCA, pGlu-Arg-Thr-Lys-Arg-4methylcoumaryl-7-amide; BisTris, 2-[bis(2-hydroxyethyl)amino]-2(hydroxymethyl)propane-1,3-diol.
of a prodomain, believed to assist in correct folding of the
enzyme (2– 4), a catalytic domain with some homology to bacterial subtilisin, and a P- or homo-B domain that plays an
essential role in regulating stability and modulating activity
(4). The proprotein convertases are thought to be responsible
for the proteolytic maturation of precursors of a multitude of
proteins and peptides, such as polypeptide hormones, neuropeptides, proteases, blood coagulation factors, adhesion factors, and receptors (reviewed in Ref. 5). PCs process these
precursors at mono or dibasic residues. Furin, the best studied
convertase, most frequently cleaves following the consensus
sequence RX(R/K)R (5); PACE4 has a similar specificity (6).
Other PCs also accept Lys in the P1 position (for nomenclature
see Ref. 7) and are less dependent on the presence of basic
residues in the P4 position (8).
The prohormone convertases PC2 and PC1 are expressed
mainly in neuroendocrine tissues and act in the regulated
secretory pathway. These enzymes are involved in the proteolytic maturation of most if not all neuropeptide precursors,
including the opioid peptide precursors proopiomelanocortin,
proenkephalin, and prodynorphin (9). PC1 has been shown to
be responsible for the initial cleavages of these precursors, and
only the action of PC2 results in the major production of small
opioid-active peptides, presumably from PC1 cleavage product
intermediates (10). Little is known at this time as to the biochemical principles that underlie cleavage site preference for
this family of enzymes. The collective analysis of many sequences recognized by these neuroendocrine PCs has shown
that PC1 has a narrower spectrum of consensus sequences
than PC2. Sequences that contain Pro at P1⬘ or P2⬘, charged
residues at P2⬘, positively charged residues at P1⬘, P3⬘, or
negatively charged residues surrounding the cleavage site are
processed exclusively or preferably by PC2 (11).
To provide information on molecular contributions to enzyme
specificity, we have constructed and analyzed mutant forms of
PC2 which, based on the furin crystal structure (12) and computational modeling studies (13), contain PC1- or furin-specific
residues rather than PC2-specific residues positioned in or
near the substrate binding pocket.
EXPERIMENTAL PROCEDURES
PC2 Mutagenesis—The desired mutations of PC2 were introduced
into mouse PC2 cDNA (from pcDNA3-PC2 plasmid) using the QuikChange® site-directed mutagenesis kit (Stratagene) following the manufacturer’s protocol. The following sequences were used for mutated
primer design: S206R, 5⬘-CTGGTTCAAGACATGGAACTAGGTGTGC3⬘; S206K, 5⬘-GACTGGTTCAACAAGCATGGAACTAGGG3⬘; T271E, 5⬘CAAGCTGGGGCCCCAAGACTGGGAAGA-3⬘; T271N, 5⬘-GTGCAAGCTGGCCAAATGACAATGGGAAG-3⬘; D278E, 5⬘-GAAGACGGTTGAAGGGCCACGAGAGCTCACA-3⬘; RE281GR, 5⬘-GTTGATGGGCCAGAGCTCACACTCCAG-3⬘; AS322TN, 5⬘-GAGGACTCCCTGGTATGATGAG-
31850
This paper is available on line at http://www.jbc.org
Downloaded from www.jbc.org at Louisiana State University Health Sciences Center on August 22, 2006
By taking advantage of the recently published furin
structure, whose catalytic domain shares high homology with other proprotein convertases, we designed mutations in the catalytic domain of PC2, altering residues
Ser206, Thr271, Asp278, ArgGlu282, AlaSer323, Leu341,
Asn365, and Ser380, which are both conserved and specific to this convertase, and substituting residues specific to PC1 and/or furin. In order to investigate the
determinants of PC2 specificity, we have tested the mutated enzymes against a set of proenkephalin-derived
substrates, as well as substrates representing Arg, Ala,
Leu, Phe, and Glu positional scanning variants of a peptide B-derived substrate. We found that the exchange of
the Ser206 residue with Arg or Lys led to a total loss of
activity. Increased positive charge of the substrate generally resulted in an increased specificity constant. Most
intriguingly, the RE281GR mutation, corresponding to a
residue placed distantly in the S6 pocket, evoked the
largest changes in the specificity pattern. The D278E
and N356S mutations resulted in distinct alterations in
PC2 substrate preferences. However, when other residues that distinguish PC2 from other convertases were
substituted with PC1-like or furin-like equivalents,
there was no significant alteration of the PC2 specificity
pattern, suggesting that the overall structure of the substrate binding cleft rather than individual residues
specifies substrate binding.
PC2 Substrate Binding Pocket Mutants
31851
AGTTGCTC-3⬘; L340W, 5⬘-GCAGGACGCCTGATATGAG-3⬘; N356S,
5⬘-TCCACCTTCGTGGTAGAAGAGGTCC-3⬘; S380T, 5⬘-CTCTGAGACAACTGGGACTGCAGCT-3⬘.
Stable Transfection with PC2 cDNAs—CHOK1 cells were transfected
with pcDNA3s (Invitrogen) containing mutated PC2s using FuGENE 6
(Roche Applied Science), according to the manufacturer’s protocol and
overexpressed using the glutamine synthase-coupled method (14). The
7B2 protein, which is necessary for proper maturation and activity of
PC2 (15), was cotransfected into the cells in a pEE14 plasmid (Lonza)
containing the glutamine synthase gene and the coding sequence of rat
7B2 (pEE147B2). Clones were selected in minimum Eagle’s medium
lacking glutamine and supplemented with the nonessential amino acids, gentamycin, and sodium pyruvate (all from Invitrogen); 10% well
dialyzed fetal bovine serum (Irvine Scientific); 25 ␮M methionine sulfoximine (Sigma); and 1.2 mg/liter of 50% active G418 (Invitrogen; the
latter two compounds were added to achieve 7B2 and PC2 selection,
respectively). Cells were grown in roller bottles in minimum Eagle’s
medium, and conditioned medium was harvested daily in Opti-MEM
(Invitrogen), centrifuged, and frozen.
Purification of Recombinant PC2—Recombinant PC2s were partially
purified from 500 ml of conditioned media diluted 1:3 with buffer A (20
mM BisTris, 0.1% Brij, pH 6.5) on 1.3-ml UNO Q1 columns (7 ⫻ 35 mm;
Bio-Rad) using a gradient from 0 to 70% of buffer B (1 M NaCl, 20 mM
BisTris, 0.1% Brij, pH 6.5) in 90 min at flow rate of 1 ml/min.
Internally quenched (IQ) substrates were synthesized as described
previously (16, 17). These substrates have the overall structure of
Abz-peptidyl-Q-EDDnp, where Abz is o-aminobenzoic acid and EDDnp
is N-(2,4-dinitrophenyl)ethylenediamine. Cleavage of the peptide results in an increase in fluorescence because of separation of the quenching group from the fluorophore.
Enzyme Assays—Unless otherwise stated, PC2 assays were performed in 50 ␮l of 100 mM sodium acetate buffer, pH 5.0, containing 2
mM CaCl2, 0.1% Brij, 0.5 mM dithiothreitol, and a protease inhibitor
mixture as follows: 280 ␮M N␣-p-tosyl-L-phenylalanine chloromethyl
ketone (Sigma), 140 ␮M N␣-p-tosyl-L-lysine chloromethyl ketone hydrochloride (Sigma), 10 ␮M E-64 (Calbiochem), and 1 ␮M pepstatin (Calbiochem) in 96-well polypropylene plates. The substrates used were
either pERTKR-MCA at 200 ␮M or the internally quenched substrates
at 2.5 ␮M. For testing of the pH optima, a mixture of 100 mM BisTris,
100 mM sodium acetate brought to different pH values was used.
Determination of kcat/Km—The kcat/Km value was determined as
described previously (18). Briefly, PC2, wild type, or mutants (at nanomolar concentrations exhibiting a proteolytic activity against the standard fluorogenic pERTKR-MCA substrate of 1 fluorescence unit (5 pmol
of MCA/min)) were assayed in the PC2 assay buffer described above, in
the presence of 2.5 ␮M IQ substrate. Fluorescence (excitation, 305 nm;
emission, 405 nm) was measured using the Ascent program in a microtiter plate fluorometer every 3 min for a 3-h period. One-phase exponential decay calculations were made using GraphPad Prism, version
4.00. The k values were calculated from the equation Y ⫽ span ⫻ e⫺kt
⫹ plateau, where Y is the amplitude of the fluorescent change; k is the
apparent first-order rate constant; and “plateau” is the fluorescence at
the end point of the reaction, which was measured for each substrate by
tryptic digestion. The resultant apparent first-order rate constants
were divided by the molar concentration of enzyme (estimated by 7B2
titration and/or Western blotting; see below).
Active Site Titration and Inhibition by 7B2 CT Peptide—The concentration of active PC2 mutant in the enzyme preparations was determined by titration with the potent specific inhibitor of PC2, the 7B2 CT
peptide (19). Preactivated enzyme was incubated with varying nanomolar concentrations of 7B2 CT peptide in PC2 buffer for 40 min in
37 °C. After this incubation, pERTKR-MCA was added to a final concentration of 800 ␮M, which is four times higher than the highest Km
value of the mutants. The experiment was repeated twice with similar
results. Simultaneously, the Ki of CT peptide for different mutants was
calculated, using the equation Ki ⫽ [I]/(V/Vi ⫺ 1) ⫻ (Km/([S] ⫹ 1)).
Modeling—The (L)-RRRRRRDL inhibitor chain, which had been
modeled into furin (20), was similarly placed into the active site of PC2
(13) to guide the designation of the specificity subsites. Based on a
superposition with this model, the proenkephalin-derived substrate
PE4 was manually docked into the active site cleft of PC2, optimizing
all possible hydrogen bonds and noncovalent interactions between the
substrate and the enzyme. The intramolecular energy of PE4 was
minimized by MAIN (21) using the Engh and Huber parameters (22) for
bond length, bond angles, dihedral angles, and improper angles.
RESULTS
Analysis of the PC2 Substrate Binding Pocket—PC2 and
furin possess 37% identical amino acids; by taking only the
catalytic domain into consideration, this homology rises to 55%.
Modeling studies of PC2 and other convertases have recently
shown that the shape of the substrate binding cleft is also very
similar (13). Fig. 1 presents the alignment of the substrate
binding pocket fragments of three protein convertases, PC2,
PC1, and furin, showing that the majority of the amino acids
lining the substrate binding pocket surface is conserved. We
focused our interest on residues that (i) differ at least between
PC2 and PC1/furin and (ii) are PC2-specific, as defined by
conservation among all PC2s, both invertebrate and vertebrate. As indicated in Fig. 1, many residues fit both of these
criteria. These are Trp203, Ser206; the cluster QPFMT-(242–
246), Thr271, Asp278, Arg281, Asp309, Ala322, Leu341, Asn356, and
Ser380 (PC2 numbering, see Table I).
From the above residues, we selected those listed and described in Table I and Fig. 2. The deletion of a segment, including residue Trp203 (residues 201–206), has already been shown
to lead to PC2 inactivation (23), and residues 242–246 appear
to be responsible for binding of the 7B2 CT peptide (24). In
PC2, Asp309 serves as the oxyanion hole-forming residue for
the catalytic triad, whereas in every other member from the
subtilisin family this position is occupied by an Asn. However, this residue has been studied previously, revealing that
Asp is essential for PC2 efficiency and might participate in
7B2 binding (25); therefore, we omitted it from further
studies.
Downloaded from www.jbc.org at Louisiana State University Health Sciences Center on August 22, 2006
FIG. 1. Alignment of fragments of catalytic domains of PC2, PC1, and furin. The conservation between the different enzymes is encoded
as follows: dark gray-shaded, amino acids are conserved among all convertases; black-shaded, amino acids are conserved within the particular
species of convertase but different among the convertase family; unshaded and light-gray-shaded, residues are not conserved and conserved weakly
within the same species of convertases; *, residues exposed to the substrate binding surface; Œ, residues mutated in this work; ⌬, residues
previously investigated; 䡺, catalytic triad and oxyanion hole (Asp309).
31852
PC2 Substrate Binding Pocket Mutants
TABLE I
List of the PC2 mutants created in this study and their presumed locations in the PC2 substrate binding pocket
PC2 mutation
PC1
Furin
S206K/S206R (Ser95)a
Lys207
Arg193
2
T271N/T271E
(Thr160)
Asn271
Glu257
3
D278E (Asp167)
Glu278
Asp264
4
RE281GR (Arg170)
GlyArg282
AlaArg268
5
AS322TN (Ala211)
ThrAsp324
ThrAsn310
6
L341W (Leu230)
Trp342
Trp328
7
N356S (Asn245)
Gly/Ser357
Ser343
8
S380T (Ser269)
Thr379
Thr365
a
Description
Position
Ser206 is located at the northeastern edge of the
S1⬘ pocket (and in part the S3⬘ subsite),
creating in comparison to Arg193 of furin a
much larger S1⬘ pocket with a less positive
potential
Thr271 separates the S3 and S5 subsites and
may have a large effect on the P5 side chain
that can extend alongside residue 271. For
certain substrates, the main chain (Nterminal of P4) might also contact Thr271, as
modeled for PE4
Asp278 provides a carboxylate anchor for the
guanidyl moiety of P4-Arg. It might also have
an electrostatic effect on the P5 side chain
Arg281 is at the edge of the S6 subsite in the
PC2 model; we expected no direct effect on P4
or P5 residues. However, there might be a
general electrostatic effect
Ala322 can influence the S1 pocket. However, it
is deeply buried, and mutations here could
have many effects (i.e. destabilization of the
entire enzyme)
Leu341 is a surface-located residue at the far
edge of the S2⬘ subsite and might contact
only large P2⬘ side chains. The mutation to
Trp might especially affect the P2⬘-S2⬘
interaction
Asn356 is located on the distant prime side,
possibly making contact with the main and/or
side chains of the residues P4⬘ and beyond.
However, the exact location of these distant
substrate residues is uncertain
Ser380 contacts the main chain of P2⬘–3⬘. It may
make a direct H-bond to the main chain
S1⬘ (S3⬘)
S3 S5
S4 S5
S6 (S4)
(S1)
S2⬘
Distant
prime
side
(S2⬘ S3⬘)
The amino acid number in the PC2 model (13) is given in parentheses.
FIG. 2. Illustration of the substrate interaction with PC2. Solid surface representation of the extended active site cleft of PC2 (13) in
complex with modeled substrates. The surface has been colored according to the calculated negative (red, ⫺35 e/kT) and positive (blue, 35 e/kT)
electrostatic surface potential. The sites used for mutations in this study are indicated in black (except for AlaSer323, which is deeply buried
beneath the S1 pocket). Further prominent differences between PC2 and the other PCs have been labeled in green. A, the generic substrate
RRRRRR2DL (gray stick model) indicates the specificity subsites (P6 to P3⬘) of the enzyme. B, the proenkephalin-derived substrate PE4 shows
the proposed altered geometry N-terminal to P4 and the potential interaction between Arg-P3⬘ and the PC2-specific Asp202. This figure was made
with GRASP (32), MOLSCRIPT (33), and RASTER3D (34).
Downloaded from www.jbc.org at Louisiana State University Health Sciences Center on August 22, 2006
1
PC2 Substrate Binding Pocket Mutants
TABLE II
Sequences of internally quenched proenkephalin-derived substrates
Boldface indicates the variable residue of the peptides.
Sequence
BR3
BR4
BR5
BR6
BR7
BA3
BA4
BA5
BA6
BA7
PE1
PE2
PE3
PE4
PE5
PE6
PE7
PE8
PE9
PE10
PE11
BA1⬘
BA2⬘
BA3⬘
BA4⬘
BA5⬘
BE1⬘
BE2⬘
BE3⬘
BE4⬘
BE5⬘
BF1⬘
BF2⬘
BF3⬘
BF4⬘
BF5⬘
BL1⬘
BL2⬘
BL3⬘
BL4⬘
BL5⬘
Modeling of the PC2 Substrate Binding Pocket—Based on the
crystal structure of furin (12) and the homology models of the
entire family of PCs (13), we have analyzed in silico the potential binding modes of a generic polyarginine-based substrate
(RRRRRRDL) as well as the PC2-specific substrate PE4 (see
Table II and Fig. 3) toward the active site cleft of PC2 (Fig. 2).
The interactions between the substrate side chains P4 through
P1⬘ and the corresponding specificity pockets/subsites of the
enzyme are well defined as follows: (i) by the main chain-main
chain interactions between the substrate and the alignment
segment of PC2 (Ser267–Pro270); (ii) by the location of the active
site residues; and (iii) by the corresponding subsites, which are
especially well defined for S4, S2, and S1. Although the overall
placement of the P4 side chain is obvious, the detailed shape of
the S4 pocket is somewhat unclear as PC2 significantly deviates in sequence and number of residues from furin and kexin
in the “canopy” creating the upper border of this pocket. Similar to furin, the guanidino group of Arg P4 should be anchored
by Asp278 (13). Presumably, the S4 pocket should also accept
hydrophobic substrate side chains, such as Phe. The PC2specific Arg281 residue might also influence to some degree the
electrostatic potential at the S4 pocket.
By extending the analysis of the substrate described above to
both sides, further subsites can be designated at the surface of
PC2 (however, with much less confidence). Nevertheless, for
most substrates the side chains P6 to P2⬘/P3⬘ should bind to (or
close to) the regions indicated in Fig. 2A. Especially for flexible
substrates (such as PE4, which contains two glycines N-terminal to Phe-P4), the main chain N-terminal to P4 might make a
strong H-bond to T271N (see Fig. 2B). At the prime side of the
scissile peptide bond of PE4, Val-P1⬘ can be positioned quite
accurately. The conformation and placement of the following
chain become increasingly less predictable, in particular due to
the highly flexible Gly in position P2⬘. In the conformation
shown, Arg-P3⬘ of PE4 might approximate the PC2-specific
residue Asp202.
Kinetic Properties of PC2 Mutants—The mutant enzyme
preparations were assayed with the standard fluorogenic substrate pERTKR-MCA, and kinetic parameters for this substrate were determined (Table III). The PC2 activity was significantly impaired only in the case of the mutants S206K,
S206R, RE281GR, and S380T. It is especially noteworthy that
substitution of Ser206 either by Arg or Lys led to an almost total
loss of activity.
The T271E and S380T as well as the L341W mutants had Km
values that were not significantly different from the wild type
enzyme. Slight increases in Km values were observed in the
AS322TN and N356S mutants. Most surprisingly, the D278N
mutant exhibited a Km value much higher than wild type,
whereas that of the RE281GR mutant was significantly lower.
The concentration of PC2 enzyme in each preparation was
determined by active site titration using the 7B2 CT peptide, a
potent PC2 inhibitor. Simultaneously, the Ki value for this
inhibitor was determined (Table III). Every mutant retained
the ability to bind the 7B2 CT peptide. The largest increase in
Ki value was observed for the mutant D278E, but this was still
in the nanomolar range (30 nM). It is interesting that this
mutation caused the largest decrease in 7B2 CT peptide binding, because the Asp278 residue is located on the southern rim
of the substrate binding cleft, opposite to the area described
previously to be involved in 7B2 CT peptide binding (the canopy region in Fig. 2) (24).
The mutations also did not alter the ability of PC2 to autoactivate. We incubated all PC2 mutants at pH 5, in the presence of 2 mM calcium, and collected samples every 10 min for 40
min. After 40 min, all of the mutated PC2s as well as the wild
type enzyme were correctly processed to the active 66-kDa PC2
forms. The rate of processing was similar for all PC2 preparations (data not shown).
The mutants were also assayed in the presence of two potent
furin and PC1 inhibitors, D9R and pro-SAAS CT peptide. Wild
type PC2 and the mutants retained ⬎90% activity in the presence of those inhibitors at concentrations of 1 mM and 10 ␮M,
respectively, indicating that single PC2 mutations with PC1- and
furin-like residues within and at the edge of the specificity pockets do not render PC2 sensitive to PC1 and furin inhibitors.
Calcium and pH Requirements of PC2 Mutated Enzymes—
The activities of all mutants were tested at different pH values
(Fig. 4). A slight shift of the pH optimum was observed for the
D278E mutant and to a lesser extent for the N356S, T271N,
and L341W mutants. In the case of the D278E mutant, the pH
optimum shifted to 5.3, i.e. closer to the pH optimum of PC1;
however, this residue surely cannot be decisive for this pH
optimum because furin, whose pH optimum is 7 (26), also
possesses an Asp in this location. Similarly, the introduction of
furin-type mutations did not alter the pH optima of the T271E
and AS322TN mutants. The calcium requirement remained
unchanged in the mutant enzymes, suggesting that these res-
Downloaded from www.jbc.org at Louisiana State University Health Sciences Center on August 22, 2006
Arginine and alanine scanning of
nonprime sites of PE1 substrate
VPEMRKR2YGGFM
VPEREKR2YGGFM
VPRMEKR2YGGFM
VREMEKR2YGGFM
RPEMEKR2YGGFM
VPEMAKR2YGGFM
VPEAEKR2YGGFM
VPAMEKR2YGGFM
VAEMEKR2YGGFM
APEMEKR2YGGFM
Proenkephalin-derived
VPEMEKR2YGGFM
YGGFLKR2FAESL
MDYQKR2YGGFL
YGGFMRR2VGRPE
FMRGLKR2SPQLE
EDSTSKR2YGGFM
YGGFMKK2DADEG
GEILAKR2YGGFM
YGGFMKK2MDELY
YGGFMKR2YGGFM
SHLLAKK2YGGFM
Alanine, glutamic acid, phenylalanine,
and leucine scanning of prime sites
of PE1 substrate
VPEMEKR2AGGFM
VPEMEKR2YAGFM
VPEMEKR2YGAFM
VPEMEKR2YGGAM
VPEMEKR2YGGFA
VPEMEKR2EGGFM
VPEMEKR2YEGFM
VPEMEKR2YGEFM
VPEMEKR2YGGEM
VPEMEKR2YGGFE
VPEMEKR2FGGFM
VPEMEKR2YFGFM
VPEMEKR2YGFFM
VPEMEKR2YGGFM
VPEMEKR2YGGFF
VPEMEKR2LGGFM
VPEMEKR2YLGFM
VPEMEKR2YGLFM
VPEMEKR2YGGLM
VPEMEKR2YGGFL
Name
31853
31854
PC2 Substrate Binding Pocket Mutants
FIG. 3. Structure of proenkephalin.
The presumed PC1 and PC2 cleavage sites
are indicated. Numbers correspond to the
numbers of substrates listed in Table I.
TABLE III
Catalytic features of wild type PC2 and mutants for the fluorogenic substrate pERTKR-MCA
Values represent averages from at least four independent experiments ⫾ S.E. WT indicates wild type.
Km for pERTKR-MCA
Vmax % of WT
Ki for 7B2 CT peptide
WT
S206K
S206R
T271N
T271E
D278E
RE281GR
AS322TN
L341W
N356S
S380T
61 ⫾ 7
100
⬍5
⬍5
96 ⫾ 32
120 ⫾ 46
150 ⫾ 30
39 ⫾ 15
98 ⫾ 43
139 ⫾ 32
102 ⫾ 40
33 ⫾ 17
4.7 ⫾ 1.3
55 ⫾ 2
84 ⫾ 18
278 ⫾ 26
24 ⫾ 1
133 ⫾ 16
88 ⫾ 20
181 ⫾ 20
108 ⫾ 21
FIG. 4. pH optima of PC2 mutants. The mutations introduced only
slightly shifted the pH optimum of PC2 activity; for details see text. wt,
wild type.
idues do not contribute to the much lower calcium requirement
of PC2 (27) (data not shown).
Specificity Constants of PC2 Mutants for PE-derived Substrates—A series of IQ substrates representing sequences surrounding dibasic cleavage sites within proenkephalin has been
tested previously for cleavage by PC2 (18) (for sequences see
Table II). The kcat/Km values of these substrates were determined for the PC2 mutants. As shown in Fig. 5, enzymes
bearing substrate binding pocket mutations did not exhibit
dramatically altered enzyme specificities against these substrates. The high number of variable residues in these substrates makes the detailed analysis of PC2 substrate interaction quite difficult, but some overall observations could be
derived.
The T271E mutant exhibited increased activity against the
substrates PE5 and PE11. In case of PE5, this might be due to
the ionic interactions that become possible between P5-Arg and
the new Glu. The Thr271 mutant, when mutated to the PC1-like
Asn (T271N), exhibited relatively increased affinity for the
PE4, PE2, and PE5 substrates and decreased affinity for PE8.
Thr271 is situated between the S3 and S5 subsites and may
have a large effect on P5 residue binding, which could extend
alongside residue 271. Unfortunately, the PE4 non-prime subsites are similar to many of the other substrates having the
Met-enkephalin motif “YGGFM”; therefore, this substrate does
14 ⫾ 1
6.0 ⫾ 3.4
29 ⫾ 6
16 ⫾ 11
15 ⫾ 2
11 ⫾ 2
13 ⫾ 2
5.5 ⫾ 0.8
not introduce any variable that could further contribute to
specificity. The only unique aspects of these substrates are as
follows: a dibasic pair at P1, P2 consisting of two Arg residues
in PE4; a Leu at P3, present in both PE2 and PE5 substrates;
and Ile in P5⬘ in the PE8 substrate.
Many changes in specificity constants were observed for the
D278E mutant, which is a carboxylate anchor for the P4 Arg
guanidino group of furin substrates and might also have an
effect on P5. We observed increased kcat/Km values of two
PE-derived substrates (PE9 and PE3) and decreased values for
at least five others (PE1, PE2, PE8, PE10, and PE11). The fact
that the specificity pattern of this mutant is so complex as
compared with wild type PC2 indicates that Asp278 might indeed be important for PC2 specificity, or rather for PC1 specificity, because this Glu is conserved only in PC1s.
The RE281GR mutant exhibited the greatest changes in
specificity constants against PE-derived substrates. In our PC2
model, this mutation is placed at the edge of the S6 pocket; for
this reason, we did not expect to observe a direct effect on
binding of P4 or P5 side chains. However, an overall electrostatic effect might be possible. The most significant change is a
great drop in affinity for the PE1 substrate. This would represent an expected characteristic of PC1, which cannot internally
cleave peptide B. One of the unique features of the PE1 peptide
is a Pro in the P6 site, but this is perhaps not decisive for PE1
substrate binding by this mutant, because substitution of Pro
by Ala (see below) does not affect binding. In addition to PE1,
several other substrates, such as PE8, PE10, and PE3, showed
relatively large but unexpected decreases in specificity. Of special interest is the increase in activity against those peptides
that represent the worst substrates for wild type PC2. The
specificity constant for the PE4 substrate (see also Fig. 2B),
poorly cleaved by wild type PC2, increased by 4-fold, and for
PE9 by 3-fold. It is very interesting that residues placed so far
from the cleavage site have such enormous influences on substrate preference. Apparently, small structural differences
even at distant sites might have large effects (either directly or
via more proximal pockets) on the detailed geometry by which
the scissile bond is presented to the active site serine. The
changes in specificity might also be caused by (possibly even
indirect) cross-talk between the mutated residues and the canopy region, which among the PCs is unique to PC2.
The S1 pocket mutant AS322TN resembles the wild type
enzyme with respect to specificity for PE-derived substrates.
Downloaded from www.jbc.org at Louisiana State University Health Sciences Center on August 22, 2006
PC2 variant
PC2 Substrate Binding Pocket Mutants
31855
From this amino acid pair, only Ala is predicted to make contact with the S1 subsite. Because of the deep burial within the
interior of the structure, any mutation at this site is more likely
to affect the overall folding and structure rather than showing
a specific effect for S1 binding. Correspondingly, this residue
does not participate significantly in substrate specificity, but
only generally weakens turnover and increases Km values (see
above).
One of the prime site mutations we created is L341W. This
mutant is expected to be placed at the edge of the S2⬘ subsite,
thus contacting only large P2⬘ side chains. Leu341 may also
influence the P4⬘ residue binding. Except for three peptides,
the general pattern of substrate specificity for proenkephalinderived substrates remained similar to that of wild type PC2.
Against this background, the distinct elevation of kcat/Km values for PE2, PE5 (both having Leu at P3), and PE11 peptides is
very clear. Again, this case shows that the primary interaction
between enzyme prime subsites and the substrate side chains
is not decisive for the enzyme-substrate interaction; the prime
site sequence of substrate PE11, “YGGFM,” is identical to PE1,
PE6, PE8, and PE10 peptides, but none of these substrates
exhibits better activity with this mutant than with the wild
type enzyme.
Another residue from the region thought to interact with the
prime sites of substrates is Asn365; we substituted this with
Ser, the amino acid occurring in other convertases. We hypothesized that the distant substrate chain (P4⬘ and beyond) contacts residue 356. Asn and Ser are both polar, neutral residues
but of very different shapes and sizes; this mutation caused a
general drop in PC2 activity. Particularly large decreases were
observed for the PE8 and PE1 substrates, the peptides that
represented the best wild type substrates but had nothing in
common with each other. The N356S mutant preferentially
cleaved PE3 and PE6, substrates uniquely possessing the P3
positions occupied by polar, noncharged residues Gln and Ser,
respectively.
Finally, the S380T mutation resides in a residue that makes
contact with the main chain proximal parts of the P2⬘ residue
but also with the main chain of P2⬘–P3⬘; it exhibited somewhat
generally lowered reactivity against the standard substrate.
Interestingly, the specificity of this mutant did not differ substantially from that of the wild type enzyme, indicating that
the introduction of the C␥ methyl group does not seem to
change the hydroxyl position and hence is not decisive for
substrate discrimination. The only visible change is a relatively
lower kcat/Km value for the PE2 substrate.
Specificity Constants of PC2 Mutants for Ala and Arg Scanning of the P3 to P7 Subsites—To learn more about PC2substrate interactions, we performed a series of assays using
derivatives of the PE1 substrate (this peptide contains the
internal cleavage site within peptide B, which is known to be
specific for PC2; see Fig. 3). The first set of substrates consisted
of Ala (BAn, “B” from peptide B-derived, where “A” indicates
Ala, and “n” is the position in the peptide) and Arg (BRn)
scanning of the P3 to P7 sites. Fig. 6, A and B, shows the
relationship of kcat/Km values of these substrates with respect
to the original substrate PE1.
In the Ala-scanning substrate series (BA3 to BA7) Ala in
position P4 appeared to be less well tolerated than the naturally occurring Met. The larger Met may form an important
interaction at the S4 subsite that is not available for the small
Ala. Moreover, this observation is in agreement with our previous data, where peptides containing Met in the P4 position
exhibited elevated inhibitory potency for PC2 (28). The exception seems to be the RE281GR mutation, which exhibited the
largest overall change in specificity for PE substrates (see
above).
Stronger specificity constant changes were observed in the
case of Arg substitutions (BR3–BR7). Arg seemed to be preferred at P4 and P5 for all enzyme variations. This is underscored by the D278E mutant results; this mutant exhibited
slightly increased activity for BR5 and decreased activity for
BR4 versus wild type PC2. One essential H-bonding partner for
the P4 arginine side chain of BR4 is missing in this mutant,
whereas little effect is observed for Arg at P5 (BR5). For the
T271E mutant, in addition to alterations at P4 and P5 residues,
we also observed large increases for substrates having an Arg
in the P3 position. This emphasizes that the P3 residue is also
affected by an interaction with the Thr271 residue. When the
same residue is substituted with Asn (as in the T271N mutant),
this effect is not seen. Correspondingly, the Arg substitutions
at P4 and P5 are not preferred as much as seen for the T271E
mutation. Additionally, in the RE281GR mutant, we have observed a positive effect of Arg substitution on P6 and P7 sites.
Downloaded from www.jbc.org at Louisiana State University Health Sciences Center on August 22, 2006
FIG. 5. Specificity constants kcat/Km of proenkephalin-derived substrates for PC2 wild type (wt) and mutant enzymes. For sequences
of the peptides also see Table II. The bars represent mean values from two or more experiments; in all cases standard errors were ⬍10%.
31856
PC2 Substrate Binding Pocket Mutants
This also agrees well with the modeling studies. Although in
the wild type enzyme Arg281 seems to face the substrate binding pocket and the Glu282 side chain is likely to be directed
away from the pocket, in this mutant this positively charged
Arg in the S6 and S7 vicinity is lacking; this most likely
improves the acceptance of the Arg-scanning peptides BR6 and
BR7.
Specificity Constants of PC2 Mutants for Prime Site (P1⬘ to P5⬘)
Substrates—In order to study the contribution of amino acids
placed in the prime sites of the substrate binding pocket, we used
a series of peptide substrate scans with four different kinds of
amino acids: alanine (BA1⬘–BA5⬘), glutamic acid (BE1⬘–BE5⬘),
phenylalanine (BF1⬘–BF5⬘), and leucine (BL1⬘–BL5⬘) scanning of
the peptide B-derived substrate PE1, which represents a specific
and favored PC2 substrate (Fig. 6, C–F). The most significant
finding was that substrates having Ala in position P1⬘ are much
better cleaved by all PC2s; small neutral amino acids are indeed
frequently found at this site within naturally occurring substrates (29). In the case of the wild type enzyme, the kcat/Km
values of BA1⬘ substrate (i.e. Ala in P1⬘ position) were elevated
almost 3-fold compared with the original substrate. Intriguingly,
the largest increase (about 10 times more than wild type) was
observed for the D278E mutant, which does not possess any
mutations in the residues forming the specificity subsites on the
prime side of the scissile bond.
Glu substitutions of prime sites did not affect PE1 substrate
binding for all mutants. In the PC2 wild type enzyme as well as
in most mutants this substitution resulted in only small favorable changes when placed at P5⬘ and P4⬘, i.e. when substituting
Phe and Met.
Concerning the Phe-scanning substrates BF1⬘ to BF5⬘, Phe
appeared to be better tolerated in the P1⬘ position than Glu or
the original Tyr from the PE1 substrate but was less well
accepted than Ala. This finding supports bioinformatics data
Downloaded from www.jbc.org at Louisiana State University Health Sciences Center on August 22, 2006
FIG. 6. Relative specificity constants of PC2 wild type (wt) and mutants for peptide B-derived substrates. A, Ala; B, Arg positional
scanning of P3–P7 sites; C, Ala; D, Glu; E, Phe; and F, Leu scanning of P1⬘–P5⬘ positions of the peptide B-derived IQ substrate. Enzymatic reactions
were performed as described under “Experimental Procedures.” Values represent the ratio of each peptide to the original peptide B-derived
substrate-PE1 (see Fig. 5). For sequences of the peptides see Table II.
PC2 Substrate Binding Pocket Mutants
indicating that aromatic residues are more tolerable in the P1⬘
position than are aliphatic residues (29). The presence of Phe
also appeared to be favorable in the P5⬘ position for the wild
type enzyme as well as for the mutants. Conversely, Gly at S2⬘
and S3⬘ sites was much better accepted than Phe.
Leu substitutions at prime subsites did not result in large
changes in substrate affinity. This substitution was, however,
less favorable for the N356S mutant at all sites. Most interestingly, the replacement of a PC2-specific prime site-interacting
residue to Ser, a residue common in other convertases, resulted
in the acquisition of PC1-like specificity, the loss of cleavage of
substrates with aliphatic residues at the P1⬘ site (29).
DISCUSSION
the convertase family. This fact taken together with our experimental data suggests that the Arg281 residue indeed might be
important for substrate binding. In the PC2 model (13), the
side chain of Arg281 is pointing upward, where it could (together with Glu282) have an effect on substrate binding to
distant non-prime subsites, especially S6. However, the exact
conformation of this area is difficult to predict because of multiple surface-located amino acid exchanges. By comparing ArgGlu to GluArg, there could also be an overall electrostatic
effect, possibly altering to some degree the acceptance of P4
side chains. However, the strong effects we have seen when
mutating ArgGlu282 indicate that these residues might be especially important to PC2. By considering the major alterations
between PC2 and the other PCs in the canopy region that forms
the northern ridge of the S4 pocket, and the observed pronounced effects on the P4 specificity of the ArgGlu282 mutations, one might speculate regarding potential cross-talk between these two regions. However, an experimental structure
is necessary to understand the observed effects in detail.
Complex changes in the specificity pattern were also observed for the N356S and D278E mutants. These mutants
showed severely altered patterns of specificity that were difficult to interpret based on a simple comparison of substrate
composition and the modeled PC2 structure. The diversity of
changes may be dependent on general conformational changes
in the substrate binding pocket, rather than on simple residueresidue interactions. The overall increase in Km for D278E
determined for a substrate that has P4 Arg (the standard
substrate for overall kinetics) is understandable, because
Asp278 is one of the key residues (carboxylate anchor) that
tether the Arg-P4 guanidino group at this position. The longer
side chain of Glu278 cannot interact with the guanidino group
in an identical manner. Also, the Arg scanning data show less
preference for Arg at P4 in this mutant. For the PE-derived
substrates (none of which has an Arg at P4), the differences
between this mutant and the wild type enzyme were not large,
as expected.
The remaining mutations did not seriously alter PC2 activity, indicating that these mutated residues are not crucial for
substrate binding and cleavage. The mutants T271E, T271N,
AS322TN, L341W, and S380T retained a generally unchanged
order of substrate preference; those that were better substrates
for wild type PC2 remained better substrates, etc. However in
the mutants listed above, on a generally unaltered background,
we observed a few changes of specificity constants for individual substrates. Some of these differences appear to have a
biochemical rationale, but other changes are more difficult to
understand. For example, the increased kcat/Km value of the
T271E mutant for PE5 is most likely caused by the favorable
interaction between the P5 Arg and Glu271, not present in wild
type PC2. However, much speculation is required to explain
other effects, such as the increase in the kcat/Km value of
L341W for PE2 and PE5. These data indicate that the Thr271,
AlaSer323, Leu341, and Ser380 residues, although conserved and
unique to PC2, are most likely not crucial for PC2 specificity.
This might be understandable in the case of the AS322TN,
L341W mutants because of their location; the former is deeply
buried at the bottom of the S1 pocket and does not contact
residues other than P1, and it affects the overall activity rather
than specificity, whereas the latter residue lies on the distant
edge of the S2⬘ subsite and might only affect the binding of
certain large P2⬘ residues. The S380T mutation resulted only
in small effects. This mutation may reduce the conformational
space that is available to the hydroxyl group in comparison to
Ser380. The Thr271 residue in turn protrudes in close proximity
to the P5 residue and may certainly influence binding to that
Downloaded from www.jbc.org at Louisiana State University Health Sciences Center on August 22, 2006
The goal of this work was to gain information as to which
residues residing within the PC2 substrate binding pocket contribute to the broad specificity of PC2 as compared with PC1.
Based on the furin crystal structure (12), and PC2 modeling
(13), we have mutated several catalytic domain residues that
are presumably placed in the PC2 substrate binding cleft and
are contained within all PC2s but not in furins or PC1s. Inspection of the various sequences within GenBankTM showed
that few PC2 residues met these conditions; we selected eight
sites for mutation to PC1/furin residues (listed in Table I and
marked on Figs. 1 and 2). Although ultimately the overall
specificity of PC2 will depend on multiple substrate-enzyme
interactions as well as on interactions of the specificity subsites
with each other, we have attempted here to assign specificity
effects triggered by single amino acid changes within the
pocket.
Most unexpectedly, the mutation of Ser206 appeared to be
essential for PC2 activity in general. Both the S206R and the
S206K mutant totally lost activity; however, their secretion was
not impaired. This result is interesting because the PC2 enzyme
is the only member of the convertase family that possesses Ser in
this position; all other remaining convertases contain basic residues (most have Lys or Arg but PC7 has His). This result indicates that this residue must play an especially important role in
PC2-specific catalysis. Previous studies have also shown that the
region around residue Ser206 is important for PC2 activity; the
deletion of residues 201–206 described by Zhu et al. (23) decreased PC2 activity to 7% of the wild type enzyme. However,
because of the low activity of this enzyme, we were not able to test
its interaction with our substrate series. From our modeling
studies, we had expected large differences in specificity that
should be explainable in terms of a larger S1⬘ pocket (because of
the missing Arg). The presence of Ser206 should lead to a better
acceptance of hydrophobic residues, in contrast to furin and PC1.
Also, the S1⬘ subsite becomes less positively/more negatively
charged. This nicely agrees with the observation that ␣-MSH/
CLIP junction in proopiomelanocortin (PVGKKRR2RPV) is exclusively cleaved by PC2 (30). In addition, the strongly reduced
activity of the S206K and S206R PC2 mutants shows that these
mutations must affect more than the immediate S1⬘ subsite, as
one would otherwise expect a “normal” PC1- and/or furin-like
overall activity.
The second largest effect we observed was for the RE281GR
mutation. The ArgGlu282 residues, being placed in a distant
non-prime site area, seem to have an important influence on
substrate binding; in addition to the alterations in specificity,
this mutant exhibited a significantly lowered Km value for the
standard fluorogenic substrate pERTKR-MCA. The Arg281 residue (which, in contrast to Glu282, has direct contact with the
substrate binding pocket) is conserved and is unique for PC2.
Other PCs have uncharged small residues in this position (with
the exception of His in PC7), and there is no homology across
31857
31858
PC2 Substrate Binding Pocket Mutants
REFERENCES
1. Steiner, D. F. (1998) Curr. Opin. Chem. Biol. 2, 31–39
2. Creemers, J. W. M., Siezen, R. J., Roebroek, A. J. M., Ayoubi, T. A. Y.,
Huylebroek, D., and Van de Ven, W. J. (1993) J. Biol. Chem. 268,
21826 –21834
3. Lusson, J., Benjannet, S., Hamelin, J., Savaria, D., and Chretien, M. (1997)
Biochem. J. 326, 737–744
4. Zhou, A., Martin, S., Lipkind, G., LaMendola, J., and Steiner, D. F. (1998)
J. Biol. Chem. 273, 11107–11114
5. Nakayama, K. (1997) Biochem. J. 327, 625– 635
6. Rehemtulla, A., Barr, P. J., Rhodes, C. J., and Kaufman, R. J. (1993) Biochemistry 32, 11586 –11590
7. Schechter, I., and Berger, A. (1967) Biochem. Biophys. Res. Commun. 27,
157–162
8. Duckert, P., Brunak, S., and Blom, N. (2004) Protein Eng. Des. Sel. 17,
107–112
9. Seidah, N. G., and Chretien, M. (1994) Methods Enzymol. 244, 171–188
10. Muller, L., and Lindberg, I. (1999) Prog. Nucleic Acids Res. 63, 69 –108
11. Day, R., Lazure, C., Basak, A., Boudreault, A., Limperis, P., Dong, W., and
Lindberg, I. (1998) J. Biol. Chem. 273, 829 – 836
12. Henrich, S., Cameron, A., Bourenkov, G. P., Kiefersauer, R., Huber, R., Lindberg, I., Bode, W., and Than, M. E. (2003) Nat. Struct. Biol. 10, 520 –526
13. Henrich, S., Lindberg, I., Bode, W., and Than, M. E. (2005) J. Mol. Biol. 345,
211–227
14. Bebbington, C. R., and Hentshel, C. G. C. (1987) in DNA Cloning (Glover,
D. R., ed) Vol. 3, pp. 163–180, Academic Press, San Diego
15. Muller, L., Zhu, X., and Lindberg, I. (1997) J. Cell Biol. 139, 625– 638
16. Hirata, Y., Cezari, M. H. S., Nakaie, C. R., Boschcov, P., Ito, A. S., Juliano,
M. A., and Juliano, L. (1994) Lett. Pept. Sci. 1, 299 –308
17. Nery, E. D., Juliano, M. A., Meldal, M., Svendsen, I., Scharfstein, J., Walmsley,
A., and Juliano, L. (1997) Biochem. J. 323, 427– 433
18. Johanning, K., Juliano, M. A., Juliano, L., Lazure, C., Lamango, N. S., Steiner,
D. F., and Lindberg, I. (1998) J. Biol. Chem. 273, 22672–22680
19. Lindberg, I., van den Hurk, W. H., Bui, C., and Batie, C. J. (1995) Biochemistry
34, 5486 –5493
20. Kacprzak, M. M., Peinado, J. R., Than, M. E., Appel, J., Henrich, S., Lipkind,
G., Houghten, R. A., Bode, W., and Lindberg, I. (2004) J. Biol. Chem. 279,
36788 –36794
21. Turk, D. (1992) Weiterentwicklung eines Programmes fuer Molekuelgrafik und
Seine Andwendung auf Verschiendene Protein-Strukturaufklaerungen.
Ph.D. thesis, Technische Universitaet, Muenchen, Germany
22. Engh, R. A., and Huber, R. (1991) Acta Crystallogr. Sect. A 43, 392– 400
23. Zhu, X., Muller, L., Mains, R. E., and Lindberg, I. (1998) J. Biol. Chem. 273,
1158 –1164
24. Apletalina, E. V., Muller, L., and Lindberg, I. (2000) J. Biol. Chem. 275,
14667–14677
25. Benjannet, S., Lusson, J., Savaria, D., Chretien, M., and Seidah, N. G. (1995)
FEBS Lett. 362, 151–155
26. Hatsuzawa, K., Nagahama, M., Takahashi, K., Takada, K., Murakami, K., and
Nakayama, K. (1992) J. Biol. Chem. 267, 16094 –16099
27. Lamango, N. S., Zhu, X., and Lindberg, I. (1996) Arch. Biochem. Biophys. 330,
238 –250
28. Apletalina, E., Appel, J., Lamango, N. S., Houghten, R. A., and Lindberg, I.
(1998) J. Biol. Chem. 273, 26589 –26595
29. Cameron, A., Apletalina, E. V., and Lindberg, I. (2001) Enzymes 22, 291–331
30. Dickinson, C. J., Sawada, M., Guo, Y. J., Finniss, S., and Yamada, T. (1995)
J. Clin. Investig. 96, 1425–1431
31. Dey, A., Lipkind, G. M., Rouille, Y., Norrbom, C., Stein, J., Zhang, C., Carroll,
R., and Steiner, D. F. (2004) Endocrinology 146, 713–727
32. Nicholls, A., Bharadwaj, R., and Honig, B. (1993) Biophys. J. 64, 166 –170
33. Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, 946 –950
34. Merritt, E. A., and Bacon, D. J. (1997) Methods Enzymol. 277, 505–524
Downloaded from www.jbc.org at Louisiana State University Health Sciences Center on August 22, 2006
subsite, but by some electrostatic interaction rather than
through spatial hindrance.
In agreement with earlier observations (8, 29), PC2 appeared
to favor small residues at the P1⬘ site. This preference is,
however, not unique to this convertase. Substrates with Ala at
P1⬘ of the peptide B-derived substrate showed a strong increase
in specificity. Large P1⬘ residues were better accepted when
aromatic (such as Phe or Tyr rather than Leu). PE5, the only
substrate having an additional Arg in a non-prime position
(besides P1 and P2), was generally well accepted by all PC2
variants. Similarly, Arg enhanced the specificity constant
when placed in every non-prime position of the peptide Bderived substrate. This brings us to the conclusion that perhaps the overall charge of the substrate is important for effective binding to the quite acidic (13) PC2 substrate binding cleft,
as previously observed for furin (20). However, there must be
important differences between the substrate binding pockets of
furin and PC2, because the positively charged peptides D6R
and D9R do not inhibit PC2 (20). PC2 has only 11 negative
charges in the direct vicinity of the cleft, the lowest number of
all the PCs; by comparison, furin has 16. It is interesting to
speculate that the lowered negative charge of the pocket contributes to the wider specificity of PC2 compared with other
convertase family members.
A detailed molecular analysis of the substrate binding pocket
of PC2 is fundamental to the rational design of PC2 inhibitors;
because PC2 is the major enzyme involved in the synthesis of
glucagon (31), these inhibitors could be useful in the treatment
of diabetes. Moreover, a better understanding of convertase
specificity could lead to the availability of “restriction proteases” as a biotechnology tool, for example for the accurate
production of bioactive peptides from recombinant precursors.