Page 123 - 82_02

The new levels of redox regulation of S-adenosylmethionine synthesis
In fact, this subunit followed a similar pathway to produce
MATa2 dimers, and only in a latter step association of
MATß takes place (46). Data regarding folding pathways
in vivo remain very limited, although E. coli MAT was
identified several years ago among the substrates for the
chaperonin GroEL (68).
All this information regarding folding and structure
could be of help to understand the impact of the different
MAT1A mutations detected in patients with isolated
persistent hypermethioninemia, and in fact we performed
this exercise in 2011 (69). Most of these patients are
asymptomatic, although demyelination and other central
nervous system manifestations have been described in
several cases (70,71). The number of patients is still small,
but newborn screenings in search for CBS defects are
aiding in the identification of hypermethioninemias due to
MAT1A mutations. Most mutations follow a recessive
autosomal inheritance trait, although a dominant pattern
has been described for the transition leading to the R264H
protein mutant (72). The presence of early stop codons
leading to shorter MATa1 subunits has been reported, as
well as the possibility of longer chains due to mutations in
the stop codon or splice-donor sites (70,71,73). Analysis of
some of these mutant proteins has been carried out upon
overexpression in E. coli and/or COS cells and their effects
on activity evaluated. The early stop codons identified to
date lead to proteins lacking from part (386X) or most
(350X) of the C-terminal domain, to others in which even
information for a complete domain is absent (92X and
185X). Additional early stop codons (351X) result not
only in the loss of half of the C-terminal domain, but also
in a change of sequence that affects a portion of the central advanced stage of the disease, where hepatitis was already
and the remaining C-terminal domains. In contrast, longer
subunits resulting from the stop codon mutation (X396Y-
464X) present with the whole wild type sequence, and
hence the information for the whole N- and central
domains plus and extended C-terminal domain. Longer
monomers due to splice-donor mutations on exon III, will
result in proteins whose sequence is drastically altered, and
hence, their folding and structure. If not eliminated by
proteolysis, the shortest MATa1 subunits are expected to
result in a complete change in the folding pattern,
precluding association. However, those mutants in which
the sequences for the N- and central domains remain
intact, and even part of the C-terminal domain is
preserved, may achieve an intermediate state that allows
association (69).
This is the case for mutations leading to MATa1
subunits of approximately 350 residues, which were the
first identified in patients with demyelination. In fact,
Hazelwood et al. proposed that this type of subunits might
associate with MATa2 monomers, leading to an inactive
hetero-oligomer (74). Although plausible, given the high
level of sequence identity between MATa1 and MATa2
subunits, the availability of new structural data showed
@Real Academia Nacional de Farmacia. Spain
that there are small differences in the orientation of the
monomers. Moreover, these mutants ~350 residues long
lack the C-terminal domain sequence involved the in
monomer-monomer interface, including the ß-sheet that
participates in dimer association. These divergences
together will impose a rearrangement to avoid exposure of
amino acids that are hidden in the wild type MATa1, and
hence reducing the probability of association with MATa2
(69). Once again, additional experiments are needed for
verification of the folding pathways followed by the
mutants, their putative ability to hetero-oligomerize with
wild type subunits, etc.
4.4. Oxidative stress induced by copper accumulation
impairs SAM synthesis in early stages of Wilson disease
Several models of liver disease that induce oxidative
stress by different mechanisms have been explored along
the years regarding the alterations induced in SAM
synthesis. Just to name some, carbon tetrachloride
intoxications (75,76), treatments with buthionine
sulfoximine (BSO) (42), Long Evans Cinnamon (LEC)
rats (20-22), etc. all concur with a reduction in MAT
activity and SAM levels. Now that the role of cysteines in
MAT I and MAT III is known this is no longer a surprise,
but since most of these models were explored earlier there
was a need to revisit certain aspects. Among them, the case
of the Wilson disease model provided by LEC rats
remained underexplored, probably because of the rare
condition of this pathology. The available data referred
only to activity parameters and were determined in an
present. Therefore, we decided to explore early
pathological stages to get insight into the copper effects on
SAM synthesis. For this purpose, we examined 9-week old
LEC rats, which already presented with copper deposits in
the liver, but no signs of hepatitis (20).
Hepatic copper accumulation at this early step led to
changes in expression that were unexpected for MAT
genes (20). The Mat1a/Mat2a expression switch was
already detected in LEC rats at this age, but Mat2b
expression was dramatically reduced, hence following an
opposite pattern to that classically described in liver
disease. These effects on Mat2b expression were prevented
by neocuproine, a chelator of Cu (I) (Figure 4). Reductions
in SAH concentrations were larger than effects on SAM
levels, probably due to the effects of copper on SAHH
(47). Activity measurements carried out under conditions
that allowed distinction between the three MAT
isoenzymes indicated a decrease in MAT III activity,
together with an increase in the combined activity of MAT
I and MAT II. These effects on activity correlated with an
enhancement in the MAT I content together with a
reduction in that of MAT III, according to analytical gel
filtration chromatography (AGFC) and dot blot detection
of MATa1 (20).
237