CYSTINE-KNOT PEPTIDES AND BBX MICROPROTEINS AS CONTROLLING FACTORS OF FLOWER AND FRUIT DEVELOPMENT

Cystine-knot peptides are members of the large family of Cysteine-rich proteins with a dimension typically less than 50 amino acids in their mature form, characterized by the presence of six conserved cysteine (Cys) residues linked by three disulfide bonds in a knotted arrangement (Rees & Lipscomb, 1980). Peptides containing the knot motif are widespread in various species such as fungi, insects, mollusks, and mammals, where they mainly play a defense role against microorganisms, acting as toxins, or as signals in cell defense pathways (Craik et al., 2001; Iyer & Acharya, 2011; Schwarz, 2017; Vitt et al., 2001). In plants, cystine-knot peptides (also referred to as cysteine-knot miniproteins) are often involved in resistance to pathogens with the function of protease inhibitors, namely metallocarboxypeptidases and serine proteases (Daly & Craik, 2011; Molesini et al., 2017). A class of cystine-knot protease inhibitors specific to the Solanaceae family was described for the first time in the 1980s (Hass & Hermodson, 1981a; Rees & Lipscomb, 1980). This class includes the tomato cystine-knot miniproteins 1 and 2 (TCMP-1 and TCMP-2), of 37 and 44 amino acids, respectively. TCMP-1 and TCMP-2 display a sequential expression pattern, which is highly modulated during flower and fruit development. TCMP-1 is expressed at a very high level in flower buds before anthesis, then its expression decreases rapidly after anthesis and increases again during fruit development (Cavallini et al., 2011). TCMP-1 mRNA level is very low in leaves, although its expression is induced by wounding and elicitors of responses to biotic stress (Díez-Díaz et al., 2004; Martineau et al., 1991). The expression of TCMP-2 is low in flower buds before anthesis, and gradually increases after fertilization, reaching a maximum in green and ripe fruits, whereas it is apparently absent in leaves, roots, and stems (Cavallini et al., 2011; Pear et al., 1989; Treggiari et al., 2015). Indeed, the TCMP-2 promoter (also named 2A11; X13743; [Pear et al., 1989]) has been successfully used to improve qualitative trait in tomato fruit (Davuluri et al., 2005). Although the biological activity of metallocarboxypeptidase inhibitors supports a role for Solanaceous cystine-knot proteins in plant defense, it has recently been demonstrated that tomato TCMPs are implicated in fruit development (Molesini et al., 2018). In the paper by Molesini et al. (2018), tomato plants transformed with a chimeric gene containing the entire TCMP-1 coding region under the control of the TCMP-2 promoter (pTCMP-2::TCMP-1) showed a marked increase in the expression of TCMP-2 before anthesis, associated with anticipated fruit production. This evidence suggests that TCMPs are regulators of fruiting time and that the maintenance of a correct TCMP-1/TCMP-2 ratio is required for proper initial fruit growth. The mode of action of TCMPs remains largely unexplored also due to the absence of homologous genes in other model species, including Arabidopsis thaliana. In several cases, Cysteine-rich peptides act as signaling molecules in plant development, by interfering with receptors or modifying the activity of multimeric complexes (de Coninck & de Smet, 2016; Tavormina et al., 2015). The general aim of this thesis was a further investigation of the functional role played by TCMP-2 during the first phases of reproductive development. Specific aims were: 1) phenotypic and molecular analyses of pTCMP-2::TCMP-1 and 35S::TCMP-2 plants, during the transition from vegetative growth to reproductive development; 2) identification of TCMP-2 interacting partners by Yeast Two-Hybrid (Y2H) screening using a tomato cDNA library; 3) functional study of one of these interactors (i.e., a B-BOX motif-containing protein) in both Arabidopsis and tomato. A detailed analysis of pTCMP-2::TCMP-1 plants during the transition from the vegetative to the reproductive stage showed an anticipated termination of the sympodial units linked with an induced expression of the florigen gene SINGLE FLOWER TRUSS (SFT), which is the main inducer of flowering. Moreover, MicroTom plants over-expressing TCMP-2 under the control of the CaMV35S constitutive promoter exhibited a reduction of the primary shoot length, very often accompanied by a decreased number of leaves before the first inflorescence, which are indicators of early flowering. The Y2H screen permitted the identification of 47 potential interactors of TCMP-2. Among them, we focused on the B-Box domain-containing protein 16 (SlBBX16). The interaction between TCMP-2 and SlBBX16 was validated in vivo in plant cells by bimolecular fluorescence complementation (BIFC) analysis. We demonstrated that TCMP-2 is also able to interact with SlBBX17, which is the closest tomato homolog of SlBBX16, and with the SlBBX16 Arabidopsis homolog, AtBBX31. A recent study showed the involvement of AtBBX30 and AtBBX31 microproteins (also referred to as miP1b and miP1a) in a multiprotein complex which regulates flowering time in Arabidopsis (Graeff et al., 2016). These proteins interact with the flowering regulator CONSTANS (AtBBX1) and additionally engage in a larger protein complex involving the co-repressor protein TOPLESS through a specific amino acid motif (PFVFL). These interactions suppress the CO-mediated induction of FT expression, causing the severe late flowering phenotype observed in plants over-expressing AtBBX30/31. The implication of AtBBX30/31 in flowering control may indicate that the TCMP-2-SlBBX16/SlBBX17 interaction could play a role in the regulation of flowering. To test this hypothesis, we ectopically overexpressed TCMP-2, SlBBX16 and SlBBX17 in Arabidopsis. The overexpression of TCMP-2 led to an anticipated flowering phenotype linked to an increased FT mRNA level, whereas the overexpression of SlBBX16 and SlBBX17 in Arabidopsis WT and AtBBX30/31 KO mutant resulted in a slight delay in flowering time, suggesting that tomato BBXs were unable to fully phenocopy AtBBX30/31 overexpression. One of the possible reasons for the weak phenotype displayed by A. thaliana over-expressing SlBBX16 and SlBBX17 could be attributed to their inability to interact with AtCO and AtTPL, since the interaction between AtBBX30/31 with AtCO and AtTPL is due to the PFVLF motif (Graeff et al., 2016), which is absent in SlBBX16 and SlBBX17. Indeed, the Y2H experiments revealed no interaction between the tomato BBXs and AtCO and AtTPL, even when using a mutated version of SlBBXs containing the PFVLF motif. The functional study in tomato was conducted on SlBBX17, which presents a peculiar expression pattern in the floral organs. MicroTom plants over-expressing SlBBX17 showed a number of leaves at the first inflorescence similar to that of WT plants, but a delay in the time to reach the flower anthesis stage and a reduced number of ripe fruits. To investigate whether in tomato TCMP-2 and SlBBX17 may participate in a multiprotein complex similar to that described in Arabidopsis (AtBBX30/31-CO-TPL), we carried out ad hoc Y2H analyses to test the interaction between TCMP-2 and SlCO1 and SlTPL1 and the interaction between SlBBX17 and SlCO1 and SlTPL1. Our data indicate that neither TCMP-2 nor SlBBX17 can directly bind to SlCO1 and SlTPL1. Further investigation of the role in flowering and fruiting of SlBBX16, the homologue of SlBBX17, may provide additional insight into the function of BBXs microproteins in tomato.