PDF《SUPPLEMENTARY INFORMATION》

W W W.

N A T U R E . C O M / N A T U R E |

1 SUPPLEMENTARYINFORMATION doi:10.1038/nature10431 2? ? Supplementary Text 1. Expression of the StSP6A gene fully correlates with tuberisation induction. Expression analyses of the StSP6A FT-homologue showed that this gene is strongly up-regulated in leaves and stolons of short day induced plants (Fig. 1f). Accumulation of this transcript is observed as early as 2-4 days after transfer to short day inductive conditions (Fig. 1f) and thus precedes tuber formation. Interruption of the long nights with a

30 min night break inhibits tuberisation and prevents StSP6A accumulation both in the leaves and stolons. Moreover, high levels of the StSP6A mRNA are detected in leaf or stolon tissues of long day grown antisense phyB plants, with a defect in photoperiodic tuberisation control1 . This transcript is not detected in lines over-expressing the Arabidopsis CO protein (AtCOox) after

8 days cultivation in short days. These plants are strongly delayed in tuberisation2 and need longer than

30 inductive days to tuberize. Supplementary Text 2. Roles of the StSP6A and StSP3D genes in short day tuberization induction and in day length-independent floral promotion. To assess a role of StSP6A in tuberization control, we generated transgenic Andigena lines that over-expressed or were silenced for expression of this gene. Transgenic StSP6Aox lines tuberized under non-inductive long days (Fig 2a, b) and were induced to flower, although their flowering phenotype was less severe than that of Hd3a lines (Supplementary Fig. 3). A good correlation was observed between the tuberisation response of these plants and StSP6A transcript levels (Fig. 2a, b), consistent with a principal role of this gene in tuberization control. RNA profiling analyses of the stolons revealed up-regulated levels of expression of tuber-specific transcripts in long days (Supplementary Fig. 4), thus confirming induction of these plants. StSP6Aox scions grafted to wild-type stocks induced the stocks to tuberize (Fig. 3c, d), supporting movement of the StSP6A inductive signal through the graft junction, as seen for Hd3a plants. Transgenic StSP6A-RNAi (SP6Ai) lines, in opposite, exhibited strongly delayed tuberization in shot days (Fig. 2c, d). RNA profiling studies of the RNAi stolons showed that tuber-specific genes are not yet up-regulated after

6 inductive days (Supplementary Fig. 4). While non-transgenic plants tuberize after 6-8 days under short days, SP6Ai plants required on average more than

3 weeks prior to tuberize. However, when cultivated under high irradiance conditions, they flowered at the same time as the untransformed controls, evidencing that StSP6A repression does not affect floral transition. By RT-PCR studies we confirmed that expression of the StSP3D and StSP5G genes was not affected in these plants (Supplementary Fig. 5a, b), silencing therefore being specific to the StSP6A gene. SUPPLEMENTARY INFORMATION

2 | W W W. N A T U R E . C O M / N A T U R E RESEARCH 3? ? To test whether StSP3D is essential for floral induction, we down-regulated expression of this gene by using an RNAi construct. Silenced lines showed a late flowering response, with 20% of the less severely repressed lines succeeding to flower (line 32, Fig. 2e, f). Flowering was suppressed in the strongest silenced lines. Scanning micrographs confirmed that the silenced plants were in a vegetative stage or had just started floral transition (Fig. 2g, h), whereas fully open flowers were observed in the controls. These plants, however, tuberized in short days at the same time as the untransformed controls and produced an equivalent number of tubers (Supplementary Fig. 7), which demonstrates that expression of this gene is not required for normal tuber development. RT-PCR studies confirmed that StSP6A gene expression is not affected in these plants (Supplementary Fig. 5c), StSP3D playing a major role in flowering but not in day length tuberization control. Supplementary Text 3. Possible StSP6A-antagonistic function of the StSP5G and StSP5G-like FT-paralogs. Expression analyses of the two additional FT-like members, StSP5G and StSP5G-like (PGSC0003DMG400016180) showed that transcripts for these genes are elevated in leaves of long day grown plants (Figure 1e). The StSP5G transcript, in addition, accumulates to high levels in the stolons of short day induced plants, although expression in these organs is delayed with respect to the StSP6A gene. These expression profiles are suggestive of a negative role of these FT-like paralogs in day length control of tuberization. In line with such function, recent studies in sugar beet showed that in this long day vernalization-requiring species, flowering time is controlled by the interplay of an antagonistic pair of mutually exclusive FT-homologs that respectively promote or suppress flowering3 . BvFT2 is essential for flowering, while BvFT1 prevents bolting in non-vernalized plants. This gene is down-regulated in long days, in annual and vernalized-biennial plants, decreased levels of expression being required for BvFT2 expression and floral transition. Interestingly, BvFT1 carries three amino acid substitutions in an external loop formed by the fourth exon of the protein and both StSP5G and StSP5G-like homologs differ from the........