Abrahamson, Department of Laboratory Medicine, Division of Clinical Chemistry and Pharmacology, Lund University, University Hospital, SE-221 85 Lund, Sweden Fax: +46 46 130064 Tel: +46 4
Trang 1complex with papain allows identification of structural requirements for broad reactivity and specificity
determinants for target proteases
Izabela Redzynia1,*, Anna Ljunggren2,*, Anna Bujacz1, Magnus Abrahamson2, Mariusz Jaskolski3,4 and Grzegorz Bujacz1,4
1 Institute of Technical Biochemistry, Faculty of Biotechnology and Food Sciences, Technical University of Lodz, Poland
2 Department of Laboratory Medicine, Division of Clinical Chemistry and Pharmacology, Lund University, Sweden
3 Department of Crystallography, Faculty of Chemistry, A Mickiewicz University, Poznan, Poland
4 Center for Biocrystallographic Research, Institute of Bioorganic Chemistry, Polish Academy of Sciences, Poznan, Poland
Papain (EC 3.4.22.2) from the latex of the papaya fruit
(Carica papaya) was one of the first known proteolytic
enzymes, and its digestive properties were already
being utilized in the 19th century Detailed biochemical
studies in the 20th century peaked with efforts in the 1960s, defining the chemistry of the enzymatic mecha-nism, delineating the concept of specificity for protein substrate recognition [1–3], and with elucidation of the
Keywords
Chagas disease; cruzipain; cysteine
proteases; papain; protein inhibitors
Correspondence
G Bujacz, Institute of Technical
Biochemis-try, Faculty of Biotechnology and Food
Sciences, Technical University of Lodz, ul.
Stefanowskiego 4/10, 90-924 Lodz, Poland
Fax: +48 42 636 66 18
Tel: +48 42 631 34 31
E-mail: gdbujacz@p.lodz.pl
M Abrahamson, Department of Laboratory
Medicine, Division of Clinical Chemistry and
Pharmacology, Lund University, University
Hospital, SE-221 85 Lund, Sweden
Fax: +46 46 130064
Tel: +46 46 173445
E-mail: magnus.abrahamson@med.lu.se
*These authors contributed equally to this
paper
Database
Atomic coordinates, together with structure
factors, have been deposited in the Protein
Data Bank under the accession code 3E1Z
(Received 13 October 2008, revised 15
November 2008, accepted 1 December
2008)
doi:10.1111/j.1742-4658.2008.06824.x
A complex of chagasin, a protein inhibitor from Trypanosoma cruzi, and papain, a classic family C1 cysteine protease, has been crystallized Kinetic studies revealed that inactivation of papain by chagasin is very fast (kon= 1.5· 106m)1Æs)1), and results in the formation of a very tight, reversible complex (Ki= 36 pm), with similar or better rate and equilib-rium constants than those for cathepsins L and B The high-resolution crystal structure shows an inhibitory wedge comprising three loops, which forms a number of contacts responsible for the high-affinity binding Com-parison with the structure of papain in complex with human cystatin B reveals that, despite entirely different folding, the two inhibitors utilize very similar atomic interactions, leading to essentially identical affinities for the enzyme Comparisons of the chagasin–papain complex with high-resolution structures of chagasin in complexes with cathepsin L, cathepsin B and falci-pain allowed the creation of a consensus map of the structural features that are important for efficient inhibition of papain-like enzymes The compari-sons also revealed a number of unique interactions that can be used to design enzyme-specific inhibitors As papain exhibits high structural simi-larity to the catalytic domain of the T cruzi enzyme cruzipain, the present chagasin–papain complex provides a reliable model of chagasin–cruzipain interactions Such information, coupled with our identification of specifi-city-conferring interactions, should be important for the development of drugs for treatment of the devastating Chagas disease caused by this parasite
Trang 2crystal structure of the enzyme, one of the first protein
structures to be determined [4] Since then, papain has
been used as a model protein in many studies, and is
the founding member of the large C1 family of
papain-like cysteine proteases [5] Approximately 12
mammalian cysteine proteases are evolutionarily
clo-sely related to papain and hence belong to this family
(e.g cathepsins B, H, L, S and K) Enzymes from the
C1 family generally function in every cell as
compo-nents of the lysosomal degradation system,
participat-ing in the turnover of proteins, but, in addition, have
been shown to participate in a number of specialized
functions, such as proteolytic cleavages activating
pro-hormones, regulation of antigen presentation, etc C1
family proteases are evolutionarily old, are found in
both prokaryotic and eukaryotic organisms, and in
many cases show activity that is indispensable for the
organism The unicellular parasite Trypanosoma cruzi
is an example of such an organism, in which the
papain-like enzyme, cruzipain, is essential for the
life-cycle of the parasite and also acts as a virulence factor
when the parasite infects its human host, causing the
devastating Chagas disease [6,7]
In a variety of species, from mammals, plants and
insects to simpler eukaryotes such as the filarial
parasites Onchocerca volvulus and Acanthocheilonema
viteae, C1 family cysteine proteases are in equilibrium
with protein inhibitors belonging to the cystatin
fam-ily, I25 [5,8–10] Most cystatins, such as human
cysta-tin B, are single-domain proteins of 100–120 residues
with a characteristic wedge-like epitope consisting of
the N-terminus and two b-hairpin loops, which blocks
the active site cleft of the target enzyme, thereby
inhib-iting the activity in a reversible manner [11,12]
Cysta-tins show high affinity for their target enzymes due to
a large binding area, with dissociation constants (Ki) in
the range 10)9–10)11m In extreme cases, such as the
human cystatin C–papain complex, Kivalues as low as
10)14mhave been reported [13]
Trypanosomatids, such as various Trypanosoma and
Leishmania species, produce inhibitors of their own
family C1 proteases [14] Chagasin, a tight-binding
inhibitor of cruzipain found in T cruzi [15], exhibits
no sequence similarity with cystatins (GenBank⁄ EMBL
[16] accession number AJ299433), despite its similar
size (110 residues) Molecular modeling studies
pre-dicted an immunoglobulin-like fold for chagasin [17],
which was essentially confirmed by subsequent NMR
[18] and crystallographic studies [19,20] Recently,
crys-tal structures of chagasin in complex with human
cath-epsins L and B [20,21], and additionally with falcipain
from the malaria parasite [22], have been determined
The complex structures demonstrate that the
enzyme-binding epitope of chagasin consists of three loops (L4, L2, L6) that together form a wedge-like enzyme-binding epitope
In this study, we present a high-quality crystal structure of chagasin in complex with papain, the model C1 family cysteine protease and one of only two enzymes in the family for which structural infor-mation for a cystatin complex is available [23,24] Based on the amino acid sequence and structure-based alignment, papain has been shown to be a close homolog of cruzipain [25] Our results confirm map-ping of the enzyme-binding epitope to the three loops,
as in chagasin complexes with mammalian enzymes, and illustrate the degree of structural adjustments as well as precise atomic contacts formed during enzyme binding Moreover, comparative analysis of several chagasin complexes has revealed a strikingly similar core structure involved in enzyme binding, which results in sub-nanomolar Ki values and rate constants for inactivation in the 105–106 m)1Æs)1 range in all cases Additionally, several contacts unique to the individual enzyme complexes could be identified, rais-ing the prospect of accurate structure-aided design of specific inhibitors of cruzipain and cathepsins Detailed knowledge of the structure and inhibition mode of chagasin should be valuable in guiding the development of drugs for the prevention and treat-ment of Chagas disease
Results
Function of chagasin as an inhibitor of papain Chagasin used in this study was expressed in Escheri-chia coli and purified to homogeneity as reported previously [20] The recombinant protein contains five extra N-terminal amino acid residues from the expres-sion construct, and has a mass of 12 440 Da as expected [20] The protein shows almost 100% activity
as a protease inhibitor based on titration of a papain solution with known activity, forms stoichiometric
1 : 1 complexes with cathepsin L or B, and is not cleaved by these proteases [20,21]
Kinetic parameters for the interaction of chagasin with papain at pH 6.0 were determined in a continuous-rate assay using the sensitive fluorogenic substcontinuous-rate car-boxybenzoyl-Phe-Arg-7-(4-methyl)coumarylamide, with
a sufficiently high inhibitor concentration for the binding reaction to be of pseudo-first order The kon value was determined to be 1.5· 106m)1Æs)1, very simi-lar to that determined for cathepsin L and higher than that for cathepsin B under the same conditions (Table 1) The equilibrium constant for dissociation
Trang 3(Ki) of the chagasin–papain complex was calculated
from the results of similar assays, under conditions
when steady-state enzyme rates could be determined
before and after addition of chagasin to a specific
con-centration The Kivalue for the papain–chagasin
com-plex, corrected for substrate competition in the assays,
was estimated as 36 pm, again similar to that of
cathep-sin L [20] and significantly lower than the values for
wild-type cathepsin B or for a cathepsin B variant with
an H110A substitution in the occluding loop, for which
the structure of its chagasin complex is known [21]
(Table 1)
Crystallization and structure determination
A complex between chagasin and papain was formed
by incubating the proteins in a 1.3 : 1 molar ratio for
approximately 4 h before setting up crystallization
drops Single crystals of the chagasin–papain complex
were obtained using Crystal Screen II in Hepes buffer
at pH 7.5 without further optimization The crystal
structure of the complex was solved to 1.86 A˚
resolu-tion by molecular replacement using the chagasin–
cathepsin L model (PDB code 2NQD) [20] as a probe
The initial atomic coordinates of the chagasin–papain
complex were obtained by rigid-body substitution of
cathepsin L by a papain model (PDB code 1KHQ)
[26] After least-squares refinement, the main-chain
traces of the chagasin and papain molecules were
visi-ble in 2Fo–Fc electron density maps without breaks at
the 1.7 r level, except for the N- and C-termini of the chagasin molecule All side chains, as well as both ter-minal segments, are clearly visible when the electron density maps are contoured at the 1.0 r level The GPLGS peptide introduced as an N-terminal extension
of the recombinant chagasin is totally disordered and not visible in the electron density maps In addition to
298 water molecules, the model includes 10 formate ions from the crystallization buffer The refinement statistics are presented in Table 2 The residues of the inhibitor are labeled without a chain designator The residues of the enzyme are marked ‘e’ When cystatin sequences are discussed in this paper, amino acid num-bering according to the chicken cystatin sequence is used, as in the original papain–cystatin B structure [23] To convert to human cystatin C numbering,
Table 1 Function of chagasin as an inhibitor of papain and other
C1 family enzymes Equilibrium constants for dissociation (K i ) of
chagasin–papain complexes were determined under steady-state
conditions at pH 6.0 as described in Experimental procedures
Cor-responding values for the papain-like cysteine proteases
cathep-sin L, cathepcathep-sin B and falcipain, with known inhibitor complex
structures [20–22], as well as for the papain complex with human
cystatin B [12], are included for comparison The K i values
pre-sented were corrected for substrate competition in the assays, as
described in Experimental procedures ND, not determined.
Enzyme
Ki(n M ) kon( M )1Æs)1)
Chagasin Cystatin B Chagasin
a Determined under slightly different assay conditions than in the
present study [22] b Determined for a recombinant variant of
chagasin with a 16 residue N-terminal extension [15].
Table 2 Data collection and structure refinement statistics Data collection
Temperature of measurements (K) 100
Resolution range (A ˚ ) 60.0–1.86 (1.93–1.86) a
Refinement Number of reflections in the working ⁄ test sets
31 568 ⁄ 1694
Number of atoms (protein ⁄ solvent ⁄ Zn ⁄ other) rms deviations from ideal
2561 ⁄ 298 ⁄ 1 ⁄ 30
Residues in Ramachandran plot (%)
a
Values in parentheses correspond to the last resolution shell.
b Rint= P
h
P
j | Ihj)<I h >| ⁄ P
h
P
j Ihj, where Ihj is the intensity of observation j of reflection h c Rpim= P
h (1 ⁄ n h )1)Pj |Ihj)<I h >| ⁄ P
h P
j <I hj > [42], calculated using SCALA [43] (from data processed using Denzo) d R = P
h | | Fo| )|F c | | ⁄ P
h |Fo| for all reflections, where Fo and Fc are observed and calculated structure factors, respectively R free is calculated analogously for the test reflections, randomly selected and excluded from the refinement e Ramachandran ‘favored’ region, as defined by MolProbity [50].
Trang 4which is widely used, ‘2’ should be added to all residue
numbers, so that G9 in cystatin B corresponds to G11
in cystatin C [27]
A strong residual peak in the Fo–Fcelectron density
map, in close proximity to H72, H74, E23 and one of
the formate ions, was interpreted as a zinc cation This
interpretation is supported by the bond valence test
[28,29] and by the tetrahedral coordination of this
cation Although chagasin inhibition is not dependent
on any cofactors, this site at the surface of the
mole-cule may be of structural significance, as the same
his-tidine residues in the cathepsin B complex were found
to bind a phosphate ion [21]
The chagasin–papain interface
The papain chain in the present complex starts with
residue I1e, which is well defined in the electron
den-sity map The last residue, N212e, is also clearly visible
because the side chain is stabilized by hydrogen bonds
with D108e and I148e, and the C-terminal carboxylate
forms a salt bridge with R188e, the latter two
inter-actions involving a symmetry-related molecule The
enzyme used for crystallization was in an inactive
form, with the catalytic C25e residue protected by
carboxymethylation The blocking group is clearly
visible in the electron density maps
The overall conformation of the chagasin molecule
in the present complex is similar to that found for free
chagasin (PDB code 2NNR) [20] The C- and
N-termi-nal residues of chagasin are somewhat flexible, but the
contour level of 1 r for the 2Fo–Fc electron density
maps was sufficient for unambiguous modeling The
first visible residue at the N-terminus is S2, which is
anchored by a side-chain hydrogen bond to N64e from
a symmetry-related molecule The C-terminal N110
residue points to a water channel
In overall shape, the present complex is similar to the previously described complex structures of chaga-sin with cathepchaga-sins L and B [20,21], resembling an inverted mushroom, with the stalk formed by the cylindrical chagasin molecule and the cap by the glob-ular papain (Fig 1) The C25e-H159e-N175e catalytic triad of papain is located at the bottom of a long cleft running across the width of the molecule, dividing it into the L and R domains [30]
The binding region of chagasin formed by the loops L2 (N29–F34), L4 (P59–G68) and L6 (R91–S100) is docked very tightly to the papain molecule (Fig 2A) The main hydrogen bonds between chagasin and papain observed in the complex are listed in Table S1 All three loops are located in the catalytic groove, with the 310 tip of loop L2 inserted directly into the cata-lytic center Loops L4 and L6 embrace the enzyme molecule from both sides
The interactions of each loop have different charac-teristics Loop L6 forms three types of interactions with the enzyme: hydrogen bonds (R91), hydrophobic contacts (P92) and p interactions (W93), which ‘probe’ different elements of the catalytic apparatus First, W93 interacts with a cluster of aromatic residues (F141e, W177e, W181e) that serve to position the N175e element of the catalytic triad (C25e-H159e-N175e) through N-H p hydrogen bonds R91 assumes
a fully extended conformation reaching to the catalytic site of the enzyme and loop L2 of chagasin The R91 guanidinium group forms two hydrogen bonds with the carbonyl group of T32 in loop L2, which is located next to the active-site-blocking residue, T31 [20,21] The other segment of the guanidinium group of R91 forms a pair of hydrogen bonds with the oxygen atom
of the side-chain amide group of N18e It is interesting
to note that the equivalent position in cruzipain is occupied by an aspartate, making the interaction with
Fig 1 Stereoview of the chagasin–papain complex The chagasin molecule is colored green and papain is colored pink The surfaces of both proteins are marked correspondingly The view is along the catalytic cleft of papain and corresponds to the standard orientation used for cysteine proteases, with the L and R lobes on the left and right, respectively.
Trang 5R91 even stronger Finally, the guanidinium group of
R91 is also hydrogen-bonded to the carbonyl group of
G20e The third element of L6, P92, shapes the loop
for optimal interactions with the enzyme by forming
hydrophobic contacts with the side chain of L143e
(Fig 2B) In addition to the direct interactions of loop
L6 described above, there are also contacts mediated
by water molecules
The interactions of loop L4 with the enzyme are based on formation of an antiparallel intermolecular b-sheet Two residues from chagasin, G66 and L65, interact with the papain main-chain atoms N64e–G66e (Fig 2C) In addition, the side-chain carbonyl Od1 atom of N64e forms a water-mediated contact with the main-chain N atom of G68, and the main-chain nitrogen of G66 of chagasin forms a water-mediated
C
D
Fig 2 Interactions of chagasin and cystatin B with papain Color coding: chagasin (green)–papain (pink); cystatin B (brown)–papain (gray) (A) Stereoview of aligned molecules created by superposition of the Ca atoms of papain from the crystal structures of its complexes with cystatin B (PDB code 1STF) and chagasin (this work) The upper panel emphasizes the different angle of approach of the two inhibitors in the standard orientation of papain The lower panel, rotated by 90 (papain R domain at the front) emphasizes the similar shape of inhibitory elements (loops and the cystatin B N-terminus) (B) Zoom-in view of the interactions of papain with loop L6 of chagasin and loop L2 of cysta-tin B (C) Zoom-in view of the interactions of papain with loop L4 of chagasin and the N-terminal segment of cystacysta-tin B (D) Zoom-in view of the interactions of papain with loop L2 of chagasin and loop L1 of cystatin B.
Trang 6contact with the main-chain carbonyl group of D158e
from papain
Compared to the very strong and extended
interac-tions of loops L4 and L6, the interacinterac-tions of loop L2
are very limited A repulsive contact is seen between
the carbonyl O atom of T31 and the Nd1 atom of the
imidazole ring of the catalytic H159e residue A much
longer, attractive contact exists between the same T31
carbonyl and the Ne1 atom of W177e (Table S1) The
hydroxyl group of T31 interacts with the main-chain
carbonyl of D158e (Fig 2D) The four additional
atoms of the carboxymethyl modification of the
cata-lytic C25e residue are easily accommodated at the
inhibitor–enzyme interface The oxygen atoms of
the carboxymethyl block form contacts with both the
enzyme (main-chain N of C25e and side chain of
Q19e) and the inhibitor (OH group of T31)
The inhibition mode of chagasin
The best-studied group of cysteine protease inhibitors
are the cystatins, which are small proteins with a
molecular mass of 11–14 kDa [27] The structure of papain in complex with cystatin B (PDB code 1STF) [23] offers an excellent opportunity for comparison of the mode of interaction of the two very different inhib-itors with the same target enzyme
Although chagasin and cystatin B show essentially identical affinity for papain (Table 1), superposition of the two complexes based on Ca alignment [31] of the enzyme portions shows a completely different fold for the two inhibitors (Fig 2A) The characteristic b-sheet grip around a long a-helix, characteristic of cystatins, contrasts with the all-b structure of chagasin However, despite their different overall fold, the epitope presented by both inhibitors to the enzyme is arranged similarly The L4–L2–L6 wedge of chagasin overlaps with a similar wedge of cystatin B formed by the N-terminal segment and two b-hairpin loops, L1 and L2 (Fig 2A–D) This similarity does not extend beyond the active site, and, in fact, the two molecules approach the enzyme from a different angle We have defined the angle of approach, s (Table 3), as the dihedral angle between two planes, one (a) dividing the
Table 3 Comparison of various enzyme complexes of chagasin The superpositions of Ca atoms were calculated using ALIGN [31] for the entire complex (c), for the enzyme molecule only (e), and for the chagasin molecule only (ch) Each superposition is characterized by the root mean square (rms) deviation in A ˚ and the number of aligned Ca atoms (in parentheses) For comparison, superpositions with the crystallo-graphic models of cruzipain and free chagasin (molecules A and B) are also included Where appropriate, a number in square brackets shows the level of sequence identity (%) between the compared enzymes The last two rows characterize the chagasin complexes by giving the contact area (in A˚2 ) calculated using Areaimol [32] and by specifying the angle of inhibitor approach s (in degrees) relative to the enzyme framework (see definition in the text).
Chagasin–cathepsin B
Chagasin–cathepsin L Chagasin–papain Chagasin–falcipain
[27.8%]
0.81 (190) [43.4%]
1.07 (191) [35.8%]
1.10 (192) [37.7%] Chagasin
Chagasin–cathepsin B
e 0.54 (233)
ch 0.46 (100)
c 1.25 (301)
e 1.28 (198)
ch 0.43 (102)
c 1.15 (287)
e 1.32 (192)
ch 0.43 (96)
c 1.15 (294)
e 1.30 (193)
ch 0.37 (101)
e 1.31 (196)
ch 0.55 (107) [28.2%]
c 1.10 (278)
e 1.42 (191)
ch 0.42 (101) [29.7%]
c 1.19 (293)
e 1.35 (190)
ch 0.52 (107) [24.1%]
e 0.79 (188)
ch 0.61 (101) [40.6%]
c 1.19 (300)
e 1.00 (189)
ch 0.44 (105) [35.9%]
e 1.18 (187)
ch 0.34 (93) [37.7%]
Trang 7enzyme into the R and L lobes along the catalytic
groove, defined by the Ca atoms of three papain
resi-dues, I40e, Y67e and W177e (or their equivalents in
other enzymes), and the other (b) created by three Ca
atoms defining the inhibitor framework and passing
along the inhibitory wedge In the case of chagasin,
plane (b) is triangulated by the tips of the peripheral
loops L4, L6 and the C-terminus, or specifically by the
Ca atoms of G66, R91 and A109 The corresponding
Ca atoms of cystatin B are located in residues G9 (in
the N-terminal binding segment, according to chicken
cystatin numbering [23]), D68 (a loop from the
oppo-site pole) and L102 (loop L2) The s angle in the
chagasin–papain complex is 5.8, indicating that the
chagasin molecule is slanting towards the R domain
The angle of approach of cystatin B is)12.7, and the
inhibitor molecule is inclined towards the L domain of
the enzyme The difference in the angles of approach
between chagasin and cystatin B is 18.5 It is also of
note that the sequential epitope of cystatins
corre-sponds to a non-sequential binding site of chagasin
The contact area [32] is similar for both complexes,
and is 853 and 922 A˚2 for the cystatin B–papain and
chagasin–papain complexes, respectively
The three crucial residues of loop L6 of chagasin
(R91, P92 and W93) correspond to L102, P103 and
H104, respectively, in the cystatin B molecule
(Fig 3A) It is noteworthy that the pattern
Pro–aro-matic residue is conserved in chagasin-like inhibitors
and in cystatins (where it is predominantly PW),
despite the lack of overall sequence similarity The
role of the proline residue appears to be to maintain
the specific shape of the loop The aromatic residue,
on the other hand, interacts with the aromatic
clus-ter of the enzyme (Fig 2B) The residue preceding
the Pro–aromatic motif, which is invariably an
argi-nine in chagasin-type inhibitors of protozoan origin,
is replaced by an aliphatic residue in cystatins
(Fig 3A) This difference may be an important
ele-ment regulating the enzyme specificity of these two
groups of inhibitors The R91 residue of chagasin
provides direct communication between loops L6 and
L2, and also interacts with the crucial D18e⁄ N18e
residues of cruzipain⁄ papain The role of the L102
residue of cystatin B is different, and supports
inter-action with the aromatic cluster of the enzyme
(Fig 2B) An additional interaction between loops
L2 and L6 of chagasin is provided by the carbonyl
group of the main chain of M90 and the nitrogen
atom of A35 A similar stabilizing contact between
cystatin B loops L2 and L1 is formed by the
main-chain carbonyl of Q101 and the peptide nitrogen
atom of T58
The interaction of loop L4 of chagasin with papain
is based on formation of an intermolecular b-sheet (Fig 2C) There is an analogous interaction between the N-terminus of cystatin B and papain G9 from the N-terminal cystatin B segment and G66 from loop L4
of chagasin provide a degree of flexibility, thus allow-ing optimal interactions between the two main chains The same role is played by G65e of papain The short antiparallel b-sheet interaction is formed by only one residue, G66e, of papain with L65 or S66 of chagasin
or cystatin B, respectively This antiparallel interaction
is supported by a water molecule linking the N atom
of G66 of chagasin and the main-chain O atom of D158e of papain In the cystatin B complex, an equiv-alent carbonyl is involved in a water-mediated interac-tion with the N atom of A10
The L2 loop of chagasin and the corresponding loop L1 of cystatin B interact with the catalytic center of papain (Fig 2D) Our structural alignment (Fig 3A) shows that loop L2 of chagasin is one residue longer, and thus T31 has no equivalent in loop L1 of cysta-tin B Loop L2 of chagasin not only interacts with loop L6 but also with loop L4, by forming a hydrogen bond between the side-chain amide of N29 and main-chain carbonyl of G66 A similar interaction is observed in cystatin B, where the side-chain amide of Q53 forms a hydrogen bond with the main-chain car-bonyl of G9, stabilizing the interaction between loop L1 and the N-terminus Although the conformation of these two loops is somewhat different, in both cases they have the same, substrate-like, polarity There is a surprisingly small number of specific interactions with the catalytic residues of the enzyme for both chagasin loop L2 and cystatin loop L1, which explains why chagasin (and also cystatins) can bind with high affin-ity to cysteine proteases with the catalytic -SH group protected by a carboxymethyl group The repulsive interactions between the chagasin loop L2 and the cat-alytic site of papain, described above, are reproduced
in the cystatin complex
Discussion
Comparison of the existing structures of chagasin complexes with cysteine proteases
In addition to the chagasin–papain complex presented
in this paper, four additional crystal structures of chagasin complexes with other cysteine proteases are available in the Protein Data Bank The target enzymes for chagasin in these complexes are cath-epsin L (PDB code 2NQD) [20], cathcath-epsin B in two crystal forms (PDB codes 3CBJ and 3CBK) [21] and
Trang 8falcipain (PDB code 2OUL) [22] These structural data
form an excellent platform for comparison of the
inter-actions between chagasin and the targeted proteolytic
enzymes The residues from the catalytic cleft of
vari-ous cysteine proteases that interact with chagasin are
structurally aligned (Fig 3B) The inhibitor binds
papain and cathepsin L with essentially the same, very
high, affinity (Ki approximately 0.03 nm); the affinity
for cathepsin B is approximately one order of
magni-tude lower, and that for falcipain is yet another order
of magnitude lower, although still in the nanomolar range (Table 1)
The contact surface area for chagasin–cysteine pro-tease complexes varies between 922 and 1373 A˚2 (Table 3), and does not directly correlate with the effi-ciency of inhibition The extra contact area found in both crystal forms of the chagasin–cathepsin B com-plex is created by the additional and unique occluding
A
B
Fig 3 Structure-based sequence alignment
of the interacting residues of cysteine prote-ases and their inhibitors (A) Alignment of structurally equivalent residues forming the enzyme-binding epitopes of chagasin-like (L4, L2 and L6) and cystatin-like inhibitors (N-terminus, L1 and L2) The following protein sequences have been used: inhibitors, Trypanosoma cruzi (GenBank accession number AJ299433), Trypanosoma brucei (AJ548777), Leishmania mexicana (AJ548776), Leishmania major (AJ548878), Entamoeba histolytica (AJ634054), Pseudomonas aeruginosa (AAG04167) [53], Gallus gallus cystatin (J05077), Homo sapiens cystatin B (BC010532), H sapiens cystatin C (BC110305); proteases, Carica papaya papain (M15203), H sapiens cathep-sin L (X12451), H sapiens cathepcathep-sin B (BC010240), Plasmodium falciparum falci-pain (AAF97809), T cruzi cruzifalci-pain (X54414) (B) Alignment of structurally equivalent resi-dues from the catalytic groove of various cysteine proteases, based on the crystal structures of their complexes with chagasin, except for cruzipain, for which a complex with a small-molecule inhibitor (PDB code 1ME3) is used The residues crucial for interactions with chagasin are color-coded
as yellow (catalytic triad), red (aromatic cluster), green (residues forming hydrogen bonds) and blue (hydrophobic contacts).
Trang 9loop of this enzyme The inhibition of cathepsin B by
chagasin is relatively weak, which may be due to the
fact that some of the binding energy has to be invested
in pushing the occluding loop out of the catalytic cleft
The angle of approach, calculated in the way described
above, has the lowest value for the tetragonal form of
the chagasin–cathepsin B complex and the highest for
the chagasin–cathepsin L complex (Table 3) The
dif-ference of 9 between these complexes may be
corre-lated with the variation of the rate of binding and
affinity for chagasin of these enzymes On the other
hand, in the two crystal forms of the
chagasin–cathep-sin B complex, the difference is 5, showing that there
is some degree of variability in inhibitor–enzyme
dock-ing, resulting either from inherent freedom of
move-ment or adaptability to environmove-mental factors, such as
crystal packing interactions
The rms deviations for the four enzyme-bound
chagasin molecules are 0.34–0.61 A˚, a range that is
similar to that for comparisons of the two crystal
structures of native chagasin (0.35–0.55 A˚) These results show that the chagasin molecule has a rigid conformation and does not change upon complex for-mation This contradicts the conclusions drawn from
an NMR study that predicted a high level of flexibil-ity of the chagasin molecule [18] A superposition of all the chagasin molecules from the complex and native structures is shown in Fig 4A A different conformation is only visible for a few N-terminal resi-dues Additionally, a small difference between native and complexed chagasin is observed at loop L4, which is rich in Gly residues, where a conformational change is responsible for adjustment of the inhibitor
to the enzyme in the b-sheet-forming motif The C-terminus has a relatively stable conformation, although the last two residues protrude from the pro-tein surface The C-terminal end of chagasin is a good marker of the variable angle of approach of the inhibitor relative to the enzyme, as illustrated in Fig 4B
L4
A
B
L2
L6
L4
L2
L6
Fig 4 Stereoview of aligned chagasin
com-plexes (A) Superposition of the Ca atoms of
the chagasin molecules from the complexes
of papain, cathepsins L and B, and falcipain
with native chagasin (B) Alignment of all
above chagasin complexes based on
superposition of the Ca atoms of the
enzyme components Color code: chagasin
(green)–papain (pink) (this work), chagasin
(gold)–cathepsin L (dark blue) (PDB code
2NQD), chagasin (orange)–cathepsin B
(mid-green) (monoclinic form, 3CBJ), chagasin
(yellow)–cathepsin B (lime green) (tetragonal
form, 3CBK), chagasin (light blue)–falcipain
(gray) (2OUL) The additional two molecules
of native chagasin (2NNR) are colored
dark-green (chain A) and purple (chain B).
Trang 10Considering the stability of the chagasin structure,
the similarity or dissimilarity of its complexes with
various enzymes may be regarded as the result of
two factors: (a) the overall similarity of the
enzy-matic part, and (b) the variability of the angle of
approach of the inhibitor relative to the catalytic
cleft of the enzyme The latter factor may reflect not
so much the geometry of the catalytic site itself,
which is highly conserved, but rather the general
shape of the peripheral regions surrounding the
active site of the enzymes, which may guide the
inhibitor molecule during its docking The data in
Table 3 show that the Ca traces of cathepsin L,
papain, falcipain and cruzipain have rms deviations
in the range 0.79–1.18 A˚ Much higher deviations are
observed for cathepsin B (1.28–1.42 A˚), in agreement
with the view that it is the most unique member of
this group of enzymes
Although the complexes include a variety of enzyme
sequences and differ in the angle of inhibitor
approach, they have a relatively similar shape; the rms
deviations for the entire complexes range from 0.87 A˚
for the chagasin–papain⁄ chagasin–falcipain pair, to
1.55 A˚ for the superposition of chagasin complexes
with cathepsins L and B
Core structural elements explaining the broad
inhibition profile of chagasin
Chagasin displays a broadly-reactive inhibition profile,
and inhibits all the investigated C1 family proteases
This efficient binding is achieved despite some
differ-ences in the architecture of the active site clefts of the
enzymes, which are especially evident for cathepsin B
[21] What are the principal elements utilized by
chaga-sin that enable it to become such a broadly-reactive
inhibitor? Correct identification of these core elements
would be useful for guiding the rational design of
efficient cysteine protease inhibitors
In all the presented structures, the inhibitory loops
creating the enzyme-binding epitope have the same
architecture except for loop L6 in the C-terminal
frag-ment from the chagasin–papain complex, which adopts
a slightly different conformation in comparison with
the other structures The different shape of this loop is
caused by formation of a hydrogen bond between the
side chain of D99 and the main-chain N atom of S21e
(Fig 2B)
Residue R91 of loop L6 forms important hydrogen
bonds in all chagasin complexes, both with the N⁄ D
residue at position 18e (papain numbering) and the
G⁄ K residue at position 20e (Fig 3B), and is an
important core elements explaining the broad
inhibi-tion profile of chagasin The crucial aromatic W⁄ F residue at position 93 of chagasin is conserved as
W⁄ H in cystatins This residue interacts with the aromatic cluster that is present in all cysteine prote-ases as an extension of the catalytic triad The pro-line residue at position 92 in chagasin, which is responsible for the shape of the L6 loop, is also con-served in cystatins
The shape of loop L4 is very similar in all com-plex structures Conserved interactions formed by loop L4 include those of residues L64–A67, which participate in both the antiparallel intermolecular b-sheet and hydrophobic contacts with Y67e and P68e (Fig 3B)
Although papain, cathepsin L and cruzipain show moderate sequence identity (36–43%), the residues responsible for the interaction with chagasin in the cat-alytic groove are conserved Chagasin thus utilizes a few conserved residues in the active site cleft of C1 family enzymes to become a broadly-reactive inhibitor with quite similar affinity for all these enzymes From
a biological perspective, it appears that these residues
in C1 family enzymes have been conserved to allow binding by chagasin- or cystatin-type inhibitors, result-ing in a means by which the organism can regulate cysteine protease activity as required
Enzyme-specific interactions of chagasin Chagasin also utilizes some enzyme-specific interac-tions, explaining why it binds more tightly to papain, cathepsin L and cruzipain than to cathepsin B and falcipain Identification of these interactions is now possible based on structural and functional data Detailed comparison of the various complexes reveals
a few contacts, all positioned close to the consensus elements of the L4–L2–L6 loops, that are unique to each of the papain, cathepsin L and cathepsin B complexes (Fig 3B) For papain, this is the contact
of S21e, for cathepsin L the contacts of Y72e and E141e, and for cathepsin B the contacts of E194e and D224e with chagasin residues boxed in Fig 3B The role of the conserved residues indicated by our structural data is consistent with published mutagene-sis studies [33] The identified characteristic interac-tions appear promising for use when designing specific inhibitors to a particular enzyme A struc-ture-based sequence alignment of the residues from the catalytic groove of cysteine proteases that inter-act with chagasin is shown in Fig 3B These residues are also preserved in cruzipain, which justifies our suggestion that these interactions are also maintained
in a chagasin–cruzipain complex