The increase of spikelet number in the panicles of modern super rice has made the architecture compact, as the extra spikelets are accommodated mostly on secondary branches than on primary branches. However, the grain yield did not improve because of poor grain filling, which was more visible in the basal spikelets than apical spikelets. The objective of this study was to examine the effect of the compactness and positional difference of spikelets in the panicle on grain filling by comparing the activity and genetic expression of starch synthesising enzymes in the developing kernels of lax-(Upahar and CR3856-45-11-2-7-2-5 (CR-45)) and compact-(Mahalaxmi and CR3856-29-14-2-1-1-1 (CR-29)) panicle cultivars. Upahar and Mahalaxmi are genetically related, whereas CR-45 and CR-29 are recombinant inbred lines. The grain carbohydrate concentration and activity of sucrose synthase (SUS) enzyme were estimated during the active period of grain filling. Further, expression of isoforms of SUS, ADP glucose pyrophosphorylase (APL and APS for large and small units respectively) and starch synthase (SS and GBSS for soluble and granule bound starch synthases respectively) were also assayed through PCR studies. The genotype approach used revealed grain SUS activity and starch concentration high and sugar concentration low in the lax-compared with compact-panicle cultivars and in the apical spikelets compared with basal ones. The margin of variation between apical and basal spikelets was higher in the compact-than the lax-panicle cultivars. Genetic expression of most of the isoforms of the enzymes was higher in the lax-than the compact-panicle cultivars as seen in RT-PCR studies. A quantitative appraisal of transcript levels of isoforms in the qRT-PCR identified greater expression of SUS3 in the basal spikelets of Upahar than that in Mahalaxmi and in CR-45 over CR-29, most prominently during the active period of grain filling. We conclude that proximal location as well as increased density of spikelets on panicles affected SUS3 expression in the basal spikelets. The metabolic dominance of a spikelet in rice panicle is dependent on the expression of the genes for different isoforms of starch synthesising enzymes, but the expression of SUS3 could be more specific than the others. SUS3 expression is most active during grain filling of the lax-panicle cultivars, but its dominance is reduced significantly in the kernels of the compact-panicle cultivars.

Rice being a staple crop for human, its production is required to be increased significantly, particularly keeping in view the expected world’s population of 9.6 billion by the year 2050. In this context, although the rice breeding programs have been successful in increasing the number of spikelets per panicle, the basal spikelets remain poorly filled, undermining the yield potential. The present study also found the grain filling to bear negative correlation with the panicle grain density. The poorly filled basal spikelets of the compact-panicle cultivars showed a lower endosperm cell division rate and ploidy status of the endosperm nuclei coupled with no significant greater expression of CYCB;1 and CYCH;1 compared with the apical spikelets, unlike that observed in the lax-panicle cultivars, which might have prevented them from overcoming apical dominance. Significantly greater expression of CYCB2;2 in the basal spikelets than in the apical spikelets might also have prevented the former to enter into endoreduplication. Furthermore, expression studies of KRPs in the caryopses revealed that a higher expression of KRP;1 and KRP;4 in the basal spikelets than in the apical spikelets of the compact-panicle cultivars could also be detrimental to grain filling in the former, as KRPs form complex primarily with CDKA-CYCD that promotes S-phase activity and G1/S transition, and thus inhibits endosperm cell division. The study indicates that targeted manipulation of expression of CYCB1;1, CYCB2;2, CYCH1;1, KRP;1 and KRP4 in the basal spikelets of the compact-panicle cultivars may significantly improve their yield performance.

No abstract

No abstract