Page 122 - 82_02

constituted the interaction interface between monomers in
the dimer. The two active sites of each dimer located at
this interface, opposite one to another, and their structure showed that interconversion between MAT I and MAT III
required residues of both subunits for catalysis (59,60).
This organization therefore explained the need for
association exhibited by MATa1 subunits, and the fact
that dimers were the minimum active isoenzymes.
Production of the crystals in the presence of substrates or
certain analogues also allowed the identification of key
residues for catalysis (59,60). Precisely, F251 was
identified as key for methionine binding (59), whereas the
role in ATP binding of a P-loop previously identified by
photoaffinity labeling was confirmed (60,61). The next
level of association involved the central domains of each
monomer, which provided the few residues that allowed
tetramerization, being the resulting interaction pattern
much more limited then that previously described for E.
coli MAT (62). How these interactions contribute to the
changes in affinity and Vmax shown between MAT I and
MAT III is still unknown and requires further
investigation.
4.2. Interconversion between MAT I and III isoenzymes is
blocked by an intrasubunit disulfide bond
The next question we addressed concerned the
interconversion between MAT I and MAT III, and
specifically what blocks this exchange, which is not
observed in the liver purified isoenzymes. The sequence of
MATa1 includes 10 cysteine residues and N-
ethylmaleimide modification of just one of them was
responsible for a large loss of enzymatic activity.
Moreover, modification of an additional sulfhydryl group
led to inactive dimers (63). Analysis of mutants generated
in all the cysteines confirmed the role of these residues in calculate that approximately 50% of the global
activity, and demonstrated that those comprised between
C35 and C105 were involved in the control of the MAT
III/I isoenzyme ratio (64). Latter on, C121 was identified
as the target for inactivation by nitric oxide and
hydroxylation (44,45,65), a role explained by its location
in the loop of access to the active site (59,60). However,
our initial studies using chemical modification also
revealed that in the isoenzymes purified from rat liver two
cysteine residues remained elusive, and that only after
reduction they became accessible to N-ethylmaleimide
modification (63,66). These data suggested the presence of
a disulfide bond within MAT I and MAT III, which was
identified as an intrasubunit bond involving C35 and C61
in the isoenzymes isolated form rat liver (66) (Figure 3).
The crystal structure of MAT I, obtained under reducing
conditions, showed that these residues were well oriented
and at a distance short enough to form such an intrasubunit
bond (59).
The role of the disulfide was further explored taking
advantage of the available refolding procedure and
cysteine mutants prepared on MATa1 (62). For this
purpose, refolding was carried out under reduced (with
DTT) or mild oxidative conditions (with GSH/GSSG
mixtures). Analysis of the resulting refolded isoenzymes
236
María Ángeles Pajares
could be produced only when the proteins were fully
reduced (62). Moreover, mass spectrometry evaluation of
the refolded proteins revealed that this interconversion was
blocked by the presence of the C35-C61 disulfide bond in
the isoenzymes resulting from refolding under mild
oxidative conditions (62). Once more, the use of the
available structural data allowed us to propose that the role
of this intrasubunit disulfide relies in its ability to stabilize
the ß-sheet of the central domain involved in dimer-dimer
interactions. Therefore, production of the disulfide at
different steps during folding would result in stable
tetramers or dimers.
4.3. Unfolding of MAT I and MAT III takes place through
an intermediate state
Our two-step refolding protocol suggested the
requirement of an intermediate state that must be well
populated to obtain a correct association pattern of the
MATa1 monomers into dimers, in order to avoid
aggregation (57). Unfolding studies carried out on MAT
III using urea as denaturant confirmed this hypothesis, and
demonstrated a three-state mechanism that is reversible
(58). The intermediate identified by gel filtration
chromatography and sedimentation velocity was an
inactive monomer with 70% of the native secondary
structure, according to circular dichroism results.
Moreover, the instability exhibited by the dimer in the
presence of the denaturant was due to its dissociation,
probably by weakening of the interface interactions
between MATa1 monomers. In fact, we were able to
stabilization energy of MAT III was due to subunit
association, whereas only 25% derived from the
interactions stabilizing the intermediate state. MAT I
unfolding was also studied taking advantage of the
differences in fluorescence intensity exhibited by tetramers
and dimers upon 8-anilinonaphtalene-1-sulfonic acid
(ANS) binding. The results obtained suggested that around
65% of the global free energy derives from dimer
dissociation, thus supporting a lower stability for dimer-
dimer interactions (62). On the other hand, thermal
denaturation of MAT I and MAT III was shown to be
irreversible, regardless of the techniques used to follow the
process. Only one transition was observed for both
tetramer and dimer denaturation, hence suggesting that
stability of MAT I is highly dependent on that of the dimer
(67). Using two-dimensional infrared spectroscopy and the
structural data it was possible to ascribe the earliest
changes to the most exposed elements, a-helices and ß-
turns.
The high level of sequence and structural conservation
among a-subunits suggested no large differences in
folding pathways for MATa1 and MATa2 monomers,
although association steps may diverge due to the
incorporation of MATß into MAT II hetero-oligomers.
However, this initial hypothesis was demonstrated to be
partially wrong when refolding of MATa2 was performed.
@Real Academia Nacional de Farmacia. Spain