Carotenoid biosynthesis in the rice endosperm
A scientific question that has accompanied all versions of
GR has been to explain howb-carotene and xanthophylls
are formed even though neither lycopene cyclase nor a
carotene hydroxylase [13] are transformed (i.e. the
biochemical pathway proceeds beyond the end-point
expected by the enzymatic activity of the expressed
transgenes). Published evidence pointed to the occurrence
of a feedback regulatory phenomenon in transgenic tomato
[14] capable of markedly modifying the expression of
several endogenous carotenoid biosynthetic genes. The
bacterial CRTI transgene in particular was suspected to be
the cause of this phenomenon because this desaturase
differs from the plant-type desaturases, not only structurally but also in the nature of the lycopene produced:
all-trans in the bacterial system but specifically
poly-ciswhen produced by plant-type carotene desaturases
(Figure 2) [15]. In plants, this requires an additional
enzyme, the recently identified carotenecis-trans-isomerase (CRTISO) [16,17],toenablecyclizationandthe
production of all-transb-carotene. The artificial accumulation of all-trans lycopene in daffodil flowers by the
application of the cyclase inhibitor CPTA [2-(p-chlorophenylthio)triethylammonium chloride] led to altered
expression levels of carotenogenic mRNAs and seemingly
underscored the relevance of such a feedback mechanism
[18]. However, the use of Real-Time PCR to quantify the
levels of rice-encoded carotenogenic mRNAs in the
endosperm showed no indication of such a regulatory
principle. Wild-type rice endosperm was shown to express
mRNAs for all the relevant carotenoid biosynthetic genes,
with the exception of PSY, which was effectively absent
[19]. The occurrence ofb-LCY-mRNA explains why it is not
needed as a transgene and the hydroxylase transcripts
explain the formation of xanthophylls in GR. However, the
presence of the phytoene desaturase (PDS) andz-carotene
desaturase (ZDS) transcripts conflicts with earlier data
that showed that expression of daffodil PSY as the sole
transgene in rice endosperm yielded only phytoene but not
colored carotenoids. However, the detection of PDS and
ZDS mRNAs does not necessarily imply the presence of
sufficient functional protein. Supporting this suggestion,
the overexpression of PDS and ZDS along with the daffodil
PSY (Figure 1, pPlantCar) resulted in b-carotene and
xanthophyll accumulation[19]. This suggests that the
cofactor requirements of PDS, including the presence of
active redox pathways[20], are met in the endosperm. The
introduction of the plant carotene desaturation pathway by
transformation also demonstrated the existence of
endogenous CRTISO activity because all-transb-carotene
was formed[19]. Because of the relevance of PSY for
the reaction velocity of the overall pathway, the relative
effectiveness of the two desaturation systems cannot be
estimated to date.
It is interesting to note that although no statistically
significant alterations in specific mRNA levels were
detected upon constitutive expression of CRTI, a decrease
in lutein in favor of b-carotene-derived xanthophylls was
observed in leaves of rice andArabidopsis[19]. To avoid
such effects that might influence the photosynthetic
performance, the preferred option is to express CRTI
under endosperm-specific promoter control, under which
conditions this effect does not occur