CONICET | Buscador de Institutos y Recursos Humanos

Accounting of a previous harvest?The mechanism proposed by Agustí et al. to explain alternate bearing in Citrus, also involves an FLC orthologue, CcMADS19, whose first intron is also a target of epigenetic changes, they show?Mauro Hoijemberg1,2 and Pablo D. Cerdán 1,2*1Fundación Instituto Leloir, IIBBA-CONICET, Avenida Patricias Argentinas 435, 1405-Buenos Aires, Argentina.2 Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Ciudad Universitaria, Buenos Aires, Argentina*Corresponding author e-mail: pcerdan@leloir.org.ar (Pablo D. Cerdán)Word count: 1230Number of figures: 1Monocarpic plants set fruit once and then senesce. On the contrary, polycarpic plants set fruit multiple times during their life cycle. In each flowering season, some vegetative meristems persist, allowing vegetative growth to resume in the following growing season. In fruit trees, it has long been known that a season of high harvest, the ON-season is followed by a season of low productivity, the OFF-season, a phenomenon known as ?alternate bearing? (Tukey, 1922). Alternate bearing is a widespread behavior in trees and has been proposed as a balancing mechanism between vegetative growth and the high demand of resources required for reproduction (Goldschmidt, 2013). Alternate bearing is a problem of high economic importance, and has been observed in important crop trees like apple, pear, mango, avocado, graprefruit, olive, pecan, pistachio, plum, apricot among others. It may be set by certain occurrences, like an early frost or water stress, for instance, that can eliminate flowering buds at once and synchronize all trees in an orchard, or even in a region, leading to highly oscillating harvests (Monselise & Goldschmidt, 2011). The physiological events leading to alternate bearing have remained elusive. Carbohydrate metabolism and hormonal balances have been implicated (Goldschmidt, 1999). It has been proposed that fruits bearing on trees exert a negative effect on flower meristem production (Monselise & Goldschmidt, 2011). Removal of young fruits from ON-trees (defruiting) allows flowering. Analyses performed in Citrus buds of defruited trees (DEF-trees) showed that these buds rapidly acquired characteristics of OFF-tree buds in opposition to ON-tree buds. These changes included altered ABA metabolism, ABA reduction and the induction of auxin transport genes (Shalom et al., 2014). In this issue of New Phytologist, Agustí et al. (pp. 000?000) have proposed a molecular mechanism to explain how Citrus branches ?remember? the previous bearing of fruits to repress flowering in the OFF-season.But then, how do plants track previous fruiting?Memory is a concept not usually associated with plants, but evidence shows that plants keep track of previous events and ?remember? these events. The flowering response to vernalization is one of these examples (Song et al., 2013). Plants are vernalized when exposed to low temperatures for several weeks. This exposure serves as evidence of the past winter, allowing plants to become more sensitive to the longer days of spring that induce flowering. The underlying mechanisms have been described in Arabidopsis, where a mechanism of cellular memory is in place. The MADS box transcription factor FLOWERING LOCUS (FLC) inhibits flowering, by repressing flowering promoters like the florigen, FLOWERING LOCUS T (FT). In annual accessions of Arabidopsis that germinate in autumn, FLC expression is high, but it is repressed during cold exposures. The peculiarity of FLC regulation is that once cold is sufficiently extended, the repression becomes epigenetically stabilized by trimethylation in Lysine 27 of Histone 3 (H3K27me3) in the FLC locus. Therefore, when warmer and longer days finally arrive, FLC expression is sufficiently low to allow the induction of FT by longer days, promoting flowering in spring. The epigenetic repression of FLC is reset in the next generation, so plants can respond again to vernalization clues (Whittaker & Dean, 2017). In Arabis alpina, an Arabidopsis-related perennial, PERPETUAL FLOWERING1 (PEP1), an FLC orthologue, is also repressed by winter cold, allowing the conversion of some vegetative meristems to inflorescence meristems but, contrary to FLC, PEP1 expression is reactivated in spring to ensure some meristems remain vegetative and are hence available for the next growing season, leading to perenniality (Wang et al., 2009). Interestingly, sequence variation in the regulatory regions within the first intron of FLC orthologues in the Brassicaceae correlates with divergence of the annual and perennial expression patterns (Kiefer et al., 2017).The mechanism proposed by Agustí et al. to explain alternate bearing in Citrus, also involves an FLC orthologue, CcMADS19, whose first intron is also a target of epigenetic changes, they show. Agustí et al. followed throughout a year the expression of CcMADS19 and the Citrus FT orthologue CiFT2 in leaves of ON and OFF trees of ?Moncada? mandarin. At the time of floral transition, the expression of CiFT2 increased in OFF but not in ON trees, in the young leaves formed in the previous spring. Interestingly, the expression of CcMADS19 followed an opposite pattern, it was higher in ON than in OFF trees at the end of winter, just the moment when floral transition is established in OFF trees (Fig. 1). Since flowering can be promoted by defruiting when fruits are still developing (Martinez-Fuentes et al., 2010), Agustí et al. showed that by defruiting, not only flowering was restored, but also the expression of CiFT2 and CcMADS19 in the leaves of DEF trees mimicked those of OFF trees. Further, CcMADS19 repressed the expression of CiFT2 in transient expression experiments carried out in leaves from OFF Citrus trees at the end of winter, suggesting CcMADS19 may act in a similar way to FLC. Therefore they decided to test if epigenetic changes could be involved in the activation of CcMADS19. First, methylation patterns in the first intron of CcMADS19 differed between OFF and ON trees and, more interestingly, patterns in DEF trees were similar to those of OFF trees. In addition, inhibitors of DNA methyltransferases led to higher CcMADS19 expression in ON trees. When they tested histone modifications, they found an enrichment of Histone 3 methylated in Lysine 4 (H3K4me3) in the promoter of ON trees compared to OFF and DEF trees. The enrichment in H3K4me3 accounts for the higher expression of CcMADS19 in the leaves of ON trees with respect to OFF and DEF trees. They also observed an increase in the expression of genes of the TRITHORAX group, whose role is to establish the H3K4me3 marks. Agustí et al. propose that fruits inhibit flowering by epigenetically activating CcMADS19, an event that would not occur if the plants are defruited.These results explained the inhibition of flowering by fruits, but not the ability of ON trees to flower in the next (OFF) season. In this regard, they found that before spring, buds displayed higher levels of H3K27me3 in the promoter of CcMADS19 in ON trees compared to OFF trees. These results suggest that before spring, the CcMADS19 locus is reprogramed in ON trees so the new emerging shoots are sensitive to the flowering signals of the next OFF season.The model proposed by Agustí et al. explains how fruits inhibit flowering in alternate bearing Citrus trees and it opens the way to further studies. It would be interesting to know if similar mechanisms occur in other species. The changes in gene expression of the TRITHORAX group could indicate that changes in H3K4me3 may be massive in branches bearing fruits. Therefore it would be also worth studying if the changes in H3K4me3 are widespread in the genome and if they are linked to sink-source relationships established during fruit development.It would also be important to further test the model by genetic approaches. This is not easy in fruit trees with long generation times. However, considering that genome editing and plant transformation is feasible (Zhang et al., 2017) and grafting techniques are already developed for fruit trees, including Citrus species, it should be possible to combine new genotypes on the same rootstocks. All these would allow the comparison of new genotypes in several aspects such as gene expression profiles, the reprograming of buds and, eventually, the effects of fruits on flowering.