Fig. 8: Seesaw working model of TCP transcription factors in the control of ovule identity.

a Ovule development in wild-type under 22°C. The activity of AG-SEPs-STK/SHPs was predominant due to the repressive effects of TCPs and BEL1 on the activity of AG-SEPs. b Ovule development in wild-type under 28 °C. Although the degradation of BEL1 was induced by HT, TCPs and the remaining BEL1 were sufficient to repress the activity of AG-SEP quartets. The activity of AG-SEPs-STK/SHPs quartets was still predominant. c Ovule development in tcpDUO under 22 °C. Without the repressive effects of TCPs on AG-SEP quartets, the relatively stable BEL1 protein only slightly inhibited the activity of AG-SEP quartets, causing a low proportion of carpeloid structures to be derived from ovule conversion in tcpDUO. d Ovule development in tcpDUO under 28 °C. Without the repressive effects of TCPs on AG-SEP quartets, degradation of BEL1 was not sufficient to inhibit the activity of AG-SEP quartets, causing the conversion of all ovules into carpeloid structures in tcpDUO. e Ovule development in the tcpDUO bel1 tredecuple mutant. Without the repressive effects of both TCPs and BEL1 on AG-SEP quartets, all ovules in tcpDUO bel1 were converted into carpeloid structures under 22 °C. f Ovule development in the tcpDUO stk tredecuple mutant. Without the repressive effects of TCPs on AG-SEP quartets, the relatively stable BEL1 protein under 22 °C was barely sufficient to inhibit the activity of AG-SEP quartets. However, the activity of AG-SEPs-SHPs was low without STK function, causing the relative predominance of AG-SEP quartets and carpeloid structures in the tredecuple mutant tcpDUO stk.