|
Structure, interaction and design studies
involving regulatory peptides and proteins |
|
Principal Investigator : Dinakar M Salunke Co-Investigator Project Associates/Assistants CSIR Associate CSIR Sr Research Fellow Ph D Students Collaborators The
structural aspects of molecular recognition and its applications in analyzing
the mechanisms associated with specific regulatory events and in rational
molecular design is the theme of research. The main objectives include i)
understanding the protein architecture, ii) analysis of the structural
principles of molecular recognition and mimicry, iii) Structural biology of
various regulatory events and iv) rational molecular design studies based on
the above. Conacanavalin
A (ConA) is an extensively studied model lectin. It has provided vital
insights regarding protein-carbohydrate recognition. It has also been shown to
bind to a variety of other ligands including certain peptides and porphyrin
derivatives. Different porphyrins bind to ConA with affinities comparable to
that of Mea-D-mannopyranoside. The crystal structure of
meso-tetrasulphanatophenylporphyrin complexed with ConA was analyzed in order
to define the binding site and mode of interactions of porphyrins with ConA
and correlate them with the corresponding interactions involving other ligands.
A pair of stacked porphyrin molecules stabilizes the crystal structure by
end-to-end cross-linking of ConA resulting in a three-dimensional network
similar to that observed on agglutination of cells. The porphyrin binds to
ConA predominantly through hydrogen bonds and water-mediated interactions
involving the sulphanatophenyl side group at the site overlapping with the
Mea-D-mannopyranoside binding site. Although the sulphanatophenyl side group
of the porphyrin does not show any topological resemblance with the
monosaccharide, it mimics hydrogen bonding interactions of ConA-sugar binding,
implying molecular mimicry between sugar and porphyrin. In order to define the invariant structural properties
associated with molecular recognition, a set of genetically distinct
monoclonal antibodies are being subjected to crystallographic analyses. The
model system used here constitutes the IgG antibodies PC283, PC282 and PC287
which have been raised against the peptide antigen PS1 (HQLDPAFGANSTNPD). The
peptide is derived from the large surface antigen of the Hepatitis B virus.
The immune response against this peptide antigen has been studied extensively.
It has been shown that the three antibodies bind to an epitope defined by the
same continuous stretch of amino acids-DPAF. The antibodies PC283, PC282 and
PC287 bind the DPAF epitope with comparable affinities. A detailed structural
analysis of the bound and unbound states of these antibodies was completed. The comparison of the variable
region sequences of both heavy and light chains of the three antibodies shows
that the three antibodies are genetically heterogeneous. For CDR sequences,
there are a total of 14 differences between PC283 and PC282, 8 differences
between PC282 and PC287 and 11 differences between PC287 and PC283. The
crystals of the Fab of PC283, PC282 and PC287 in complex with the antigen PS1
were obtained and they diffracted up to 2.9, 2.5 and 2.5 Å, respectively. The
crystals of the Fab of PC282 and PC287 were obtained in their native state
also and they diffracted up to 1.8 and 2.3 Å, respectively. The three complex
and the two native structures have been refined using all data to their
maximum resolutions. The crystallographic R factors for all the structures
were within 20% and the free R factors were within 26%. The structures of the
three complexes were compared in all aspects. There were significant
differences in the elbow angle and the VH:VL interface area in the three
antibodies. It has been postulated that differences in elbow angle and VH:VL
interface area reflect the differences in relative orientation of the VH and
VL domains. Changes in the relative orientation of the variable domains can
lead to changes in the topology of the antigen-binding site. When the variable
domains of the three antibodies were structurally aligned it was observed that
the CDR L3 shows a significant change in conformation in case of PC287 and
PC282 as compared to PC283. This is due to the insertion present in CDR L3 of
PC287 and PC282. A comparison of the rmsd values in the position of Ca atoms
shows that there is only slight variation in the main chain conformation of
the other five CDRs. The conformation of the peptide in
the three complexes was compared. For PC282 density was seen for the stretch
QLDPAFG and for PC287 the stretch HQLDPAFGA could be fitted to the observed
density, unlike in case of PC283 where electron density could be seen for all
fifteen residues. The residues Asp4P, Pro5P, Ala6P and Phe7P form a b-turn in
all the three complexes. It is appropriate to mention here that the peptide
shows a random conformation as shown by CD and NMR studies. Thus the main
chain conformation is very similar in case of these four residues. In
addition, the side chain of the residues Pro5P, Ala6P and Phe7P is also very
similar. However, the side chain conformation of Asp4P in case of PC283 is
very different from that in case of PC287 and PC282. The conformation of the
residues before Asp4P is different in case of PC283 than in case of PC282 and
PC287. A comparison of the interactions between the peptide and antibody
highlights the similarities and differences in the recognition of the peptide
by the three antibodies. There is a great deal of degeneracy in the location
and nature of the antibody residues, which interact with the three residues
Pro5P, Ala6P and Phe7P. This is true even for residues from CDR L3 which has a
different main chain conformation in PC283 than that seen in PC282 and PC287.
The interactions shown by Leu3P and Asp4P are very different in case of PC283
as compared to that in PC282 and PC287. In case of PC283 Asp4P is oriented
towards the surface of the groove due to the formation of a salt bridge with
the side chain of Arg53H. In case of PC287 and PC282 this interaction is
absent as there is a Ser residue at the 53H position the side chain of, which
is too short for any such interactions to occur. As a result the Asp4P side
chain turns downwards into the groove and forms hydrogen bonding interactions
with the side chain of Ser91L (which is Thr in PC283). Consequently the Leu3P
also has to change its conformation in case of PC282 and PC287 because Asp4P
occupies the position it occupied in PC283. The changes in conformation of
Leu3P and Asp4P show that the flexible nature of the peptide allows it
compensate for changes in the paratope by changing its conformation in such a
way that the interactions with the CDR residues are optimized. When the bound and native structures of PC287 were
compared it was seen that the main chain conformation of all the CDRs were
similar. The side chain conformation of the residues from CDRs L1, L2, H1, H2
and H3 were also similar. In case of CDR L3, the comparison showed that the
residue Tyr94 flips outwards to facilitate peptide binding. In case of PC282
it was seen that the main chain as well as side chain conformation of the
residues from CDRs H1, H2 and L2 were similar. The CDR H3 undergoes a
significant outward movement on peptide binding. The CDRs L1 and L3 show a
slight inward movement on peptide binding. Thus, it appears that the two
antibodies PC282 and PC287 follow two different mechanisms to arrive at a
common conformation of peptide and CDRs in the bound state. In spite of
repeated attempts, PC283 Fab was not crystallized in its native state. It is
inferred that the PC283 CDRs are flexible and the peptide binding is required
for their stabilization. This flexibility would lead to a repertoire of
conformations for the CDRs in the unliganded state compared to the bound form. The ability of the peptide to move from a disordered
conformation to an ordered b-turn conformation in different antigen binding
sites could be one of the reasons for its well characterized immunodominant
properties. The similarities observed between the three complexes point that
the immune response appears to have evolved a common conformation of CDRs in
genetically distinct antibodies to bind to a flexible epitope in a similar
conformation with similar affinities. The differences observed in the native
structure lead to the conclusion that different mechanisms are followed in the
antibodies to converge to the same bound conformations. The crystal structure of an antibacterial protein of
immune origin, which was purified from tasar silkworm (Antheraea mylitta)
larvae after induction by E.coli infection, has been determined and refined.
The core structure of this protein is similar to c-type lysozymes and a-lactalbumins.
The catalytic residues are conserved with respect to the chicken lysozyme.
While the A.mylitta protein is functionally similar to chicken lysozyme unlike
human a-lactalbumin, it is significantly different in certain structural
features with respect to the other two proteins. Although physiological
origins of the tasar silkworm protein and chicken lysozyme are different, the
catalytic mechanism employed by them would probably be similar with subtle
differences in the specificity and level of activity. On the basis of the
structural comparisons between tasar silkworm protein, chicken lysozyme and
human a-lactalbumin it can be suggested that the conformational changes in a
protein are minimal during rapid evolution as compared to those in the normal
course of evolution. Publications Original peer-reviewed articles 1. Nair DT, Singh K, Sahu N, Rao KVS and Salunke DM (2000) Crystal structure of an antibody bound to an immunodominant peptide epitope: Novel features in peptide-antibody recognition. J Immunol 165:6949-6955. 2. Manivel V, Sahoo NC, Salunke DM and Rao KVS (2000) Maturation of an antibody response is governed by modulations in flexibility of the antigen-combining site. Immunity 13:611-620. 3. *Jain D, Kaur KJ, Sundaravadivel B and Salunke DM (2000) Structural and functional consequences of peptide-carbohydrate mimicry: crystal structure of a carbohydrate-mimicking peptide bound to concanavalin A. J Biol Chem 275:16098-16102 (*in press last year, since published). 4. Jain D, Kaur KJ, Goel M and Salunke DM (2000) Structural basis of functional mimicry between carbohydrate and peptide ligands of ConA. Biochem Biophys Res Commun 272:843-849. 5. Kaur KJ, Jain D, Goel M and Salunke DM (2001) Immunological implications of structural mimicry between a dodecapeptide and a carbohydrate moiety. Vaccine (in press). 6. Chavali GB, Vijayalakshmi C and Salunke DM (2001) Analysis of sequence signature defining functional specificity and structural stability in helix-loop-helix proteins. Proteins: Struct Func Genet 42:471-480. 7. Jain D, Kaur KJ and Salunke DM (2001) Plasticity in protein-peptide recognition: crystal structures of two different peptides bound to concanavalin A. Biophys J (in press). |