Circadian rhythm and light signals have been reported to be involved in several physiological processes, including flowering, various stress responses and metabolism [26, 30, 31, 38]. However, this is the first report suggesting a role for circadian rhythm and flowering signals in the regulation network of male sterility. In general, these genes promote the transition from vegetative growth to reproductive growth under SD conditions in SD plants and under LD conditions in LD plants . Although the male sterility transition in PGMS rice is known to be regulated by LD conditions, the mechanism of regulation remains unknown. Global gene expression profiling under LD conditions revealed a suppressive trend and categorised into the circadian rhythm and flowering pathways. Both pathways regulate the process in a coordinated manner. Further studies revealed that the diurnal expression profiles of OsPRR1, OsPRR37, OsGI, OsLHY, OsDof and Hd1 under LD conditions were reprogrammed.
In this study, OsPRR1, OsPRR37 and OsGI exhibited similar expression patterns in the leaf under LD conditions, suggesting that these three genes may function together. OsLHY and OsDof also exhibit similar differentially expressed patterns, suggesting that they may function coordinately. In Arabidopsis, TOC1/PRR1, LHY and GI, key component factors in the central clock, comprise interlocked transcriptional feedback loops to regulate plant circadian rhythms . In PGMS rice, OsPRR1, OsLHY and OsGI were not only involved in the circadian rhythm, but were also involved in the male sterility transition in coordination with OsPRR37 and OsDof. The LD signals maybe integrated and transmitted to the downstream genes through rice central clock in leaves. The day length signal may have been transmitted to the inflorescence via Hd1 and OsPRR37, ultimately leading to pollen abortion.
In the rice flowering pathway, OsGI is upstream of Hd1 and is positively correlated with Hd1 expression. Hd1 is activated by OsGI under LD or SD conditions. Hd1 activates Hd3a/RFT1 expression under SD conditions to promote flowering. However, Hd1 suppresses Hd3a/RFT1 expression under LD conditions to inhibit flowering . In PGMS rice, Hd1 expression may also have a positive correlation with OsGI under LD conditions in Nongken 58S, as its expression was activated after OsGI expression (Figure 3 C, D). This implies that Hd1 may function downstream of OsGI in the male sterility transition in PGMS rice. In this case, Hd3a/RFT1 may not be the downstream regulator of Hd1 since there were no rhythm differences between Nongken 58S and Nongken 58. Hd1 was activated by OsGI under LD conditions and suppressed the unknown factors. This may influence downstream gene expression and finally affect male fertility.
In leaves, the diurnal expression patterns of OsPRR1, OsPRR37, OsGI, OsLHY, OsDof and Hd1 showed significant differences between Nongken 58S and Nongken 58. In the inflorescence, only OsPRR37 and Hd1 exhibited different expression patterns between Nongken 58S and Nongken 58 under both LD and SD. Therefore, it is speculated that the clock genes in leaves may function as sensors for day length. These genes may receive day length signals and integrate and transmit the signals into the inflorescence through a series of unknown pathways. Both OsPRR37 and Hd1 may be the effectors in the inflorescence, and they are likely involved in regulation of male sterility directly.
Previous studies have indicated that phytochromes and cryptochromes are involved in the male sterility transition. However, there were no diurnal expression differences in OsPhyA, OsPhyB, OsCry1a, OsCry1b, OsCry2 and OsCry3 under LD conditions between Nongken 58S and Nongken 58 (data not shown). Using an ELISA assay, Wang, et al. found that phyA content in Nongken 58S leaves was higher than in Nongken 58 under the identical day length treatment . The differences in phytochromes and cryptochromes may be at the protein level and needs further investigation.