Is bicarbonate in Photosystem II the equivalent of the glutamate ligand to the iron atom in bacterial reaction centers?

Photosystem II of oxygen-evolving organisms exhibits a
bicarbonate-reversible formate effect on electron transfer between
the primary and secondary acceptor quinones, QA and QB. This effect
is absent in the otherwise similar electron acceptor complex of
purple bacteria, e.g. Rhodobacter sphaeroides. This distinction has
led to the suggestion that the iron atom of the acceptor quinone
complex in PS II might lack the fifth and sixth ligands provided in
the bacterial reaction center (RC) by a glutamate residue at
position 234 of the M-subunit in Rb. sphaeroides,RCs (M232 in Rps.
viridis). By site-directed mutagenesis we have altered GluM234 in
RCs from Rb. sphaeroides, replacing it with valine, glutamine and
glycine to form mutants M234EV, M234EQ and M234EG, respectively.
These mutants grew competently under phototrophic conditions and
were tested for the formate-bicarbonate effect. In chromatophores
there were no detectable differences between wild type (Wt) and
mutant M234EV with respect to cytochrome b-561 reduction following
a flash, and no effect of bicarbonate depletion (by incubation with
formate). In isolated RCs, several electron transfer activities
were essentially unchanged in Wt and M234EV, M234EQ and M234EG
mutants, and no formate-bicarbonate effect was observed on: (a) the
fast or slow phases of recovery of the oxidized primary donor (P+)
in the absence of exogenous donor, i.e., the recombination of P+QA−
or P+QB−, respectively; (b) the kinetics of electron transfer from
QA− to QB; or (c) the flash dependent oscillations of semiquinone
formation in the presence of donor to P+ (QB turnover). The absence
of a formate-bicarbonate effect in these mutants suggests that
GluM234 is not responsible for the absence of the
formate-bicarbonate effect in Wt bacterial RCs, or at least that
other factors must be taken into account. The mutant RCs were also
examined for the fast primary electron transfer along the active
(A-)branch of the pigment chain, leading to reduction of QA. The
kinetics were resolved to reveal the reduction of the monomer
bacteriochlorophyll (τ = 3.5 ps), followed by reduction of the
bacteriopheophytin (τ = 0.9 ps). Both steps were essentially
unaltered from the wild type. However, the rate of reduction of QA
was slowed by a factor of 2 (τ = 410 ± 30 and 47 ± 30 ps for M234EQ
and M234EV, respectively, compared to 220 ps in the wild type). EPR
studies of the isolated RCs showed a characteristic g = 1.82 signal
for the QA semiquinone coupled to the iron atom, which was
indistinguishable from the wild type. It is concluded that GluM234
is not essential to the normal functioning of the acceptor quinone
complex in bacterial RCs and that the role of bicarbonate in PS II
is distinct from the role of this residue in bacterial RCs.