mation [53]. Once the reaction reverted, C678–690can be transformed back to C684–696in the dark, which
can be transformed again to C678–690by new illumination and so on. Therefore C678–690and C684–696
form a cycle. In vitro experiments have shown that the C678–690to C684–696transformation required
NADPH [42]. Consequently, when the cycle is turning, NADPH is consumed. It is probably oxidized
at each light-triggered C684–696→C678–690conversion [54]. We showed that this cycle is involved in
the photoprotection of newly formed Chlide against photooxidation [55,56]: when Chlide is in the
C678–688conformation, it is readily photodestroyed, whereas in the C684–696form, this is not the case.
This is in line with the action spectrum of the oxygen uptake by Chlide in shortly illuminated plastids
established by Redlinger and McDaniel [57]. It should be emphasized that a photoprotection mecha-
nism is needed at this stage because carotenoids, although present in the etiolated leaves [28,34,58], do
not protect the newly formed Chlide against photodestruction [59]. It should be noted here that unpro-
tected Chl(ide) is very reactive with oxygen when illuminated and generates activated oxygen species,
which are able to destroy cellular and subcellular structures (reviewed in Ref. 60). The photoprotection
mechanism is specifically NADPH dependent [55]. Because in the two spectral forms of Chlide-LPOR
aggregates involved in the cycle, i.e., C678-688and C684-696, the Chlide is still bound to the enzyme, we
can conclude that LPOR is involved in the transformation of C684-696to C676-688. The involvement of
LPOR in this process is further supported by the increase in Chlide photoprotection in Arabidopsis
overexpressing LPOR [61]. When the aggregates are dissociated, Chl(ide) is partially released from
LPOR and is esterified. Both events occur during the Shibata shift (reviewed in Ref. 32). These events
are summarized in Figure 4B.
During the Shibata shift, photoactive Pchlide is regenerated. This process has not yet been exten-
sively studied. It was shown that an aggregate similar to photoactive Pchlide, but containing NADP(i.e.,
P642-649) instead of NADPH, is formed very rapidly after the photoreduction [62]. Experiments using
inhibitors of protein synthesis have shown that LPOR is partly reused to regenerate photoactive Pchlide
[63,64]. On the other hand, full regeneration requires protein synthesis [65]. Although it is established that
at least enzymes involved in the -ALA synthesis are involved, the exact number of proteins synthesized
de novo remains undetermined.
C. Chlorophyll Formation in Partially Green Leaves and in Fully
Green LeavesNonilluminated leaves contain far less Pchlide than fully mature green leaves contain Chl (reviewed in
Ref. 66). Therefore, Chl should be produced during greening. In Secs. III.A and III.B the arguments in fa-
vor of the involvement of aggregates of LPOR-Pchlide a-NADPH complexes in Chlide aformation dur-
ing the first illumination were presented. Although some evidence suggests that the same types of aggre-
gates are involved in Chlide aformation during leaf greening and also in green leaves [27,67,68], no firm
proof was given in these works. Therefore it was crucial to determine whether Chl is formed during green-
ing and in green leaves according to the set of reactions illustrated in Figure 4. If so—i.e., similar spec-
tral forms of photoactive Pchlide are used to synthesize Chl during greening—a steady-state amount of
photoactive complexes should be detected when the plants are illuminated with nonsaturating light. This
was demonstrated by in situ fluorescence [69,70] and by absorbance measurements [7,71]. The amount
of photoactive Pchlide detected during greening is directly related to the light intensity used for cultiva-
tion [71]. The photoactive Pchlide involved in greening has a 77 K emission maximum slightly shifted to
the blue (653 nm) [69,70]. Under a saturating flash, the pool of photoactive Pchlide, not photoreduced by
the light used to drive greening, is transformed to C678–690, which is in turn transformed to C684–696and
subsequently to C672–682. The duration of these shifts, similar in nature to those observed in etiolated
leaves, is dramatically accelerated when compared with the etiolated material [69]. Therefore, it can be
concluded that the cycle presented in Figure 4B also describes the reactions leading to Chl production in
greening and green leaves.
This conclusion is in sharp contradiction to the view expressed by Lebedev and Timko [72], who pro-
posed the existence of a cycle similar to Figure 4A. It also apparently runs against the measurements of
the variations of LPOR messenger RNA (mRNA) and LPOR amounts during greening. In fact, both dra-
matically decrease during the first hours of greening (Figure 5) (reviewed in Ref. 73). This last contra-
diction vanished when it was found that most of the angiosperms contain two LPORs, denoted LPORA
and LPORB [74–76] (reviewed in Refs. 9 and 14). Exceptions have been found in cyanobacteria [77],
270 SCHOEFS