No differences in the vegetative growth were observed for any of the six and nine lines selected carrying VviINO.1 and VviINO.2, respectively , even though some variability was observed in the size of silique in mature plants . Careful observation of ovule phenotypes both by stereo and optical microscope showed that both VviINO.1 and VviINO.2 coding regions were able to restore outer integument growth. However, differently to complementation with AthINO transgene, which leads to the high frequency of full complementation, transgenic plants complemented with grape INO CDS exhibited different ovule morphologies consistent inside lines. We have classified lines as “wild type-like” when normal ovule development was fully rescued or alternatively as “sup-like”, “weakino-like”, or “ino-like” when some atypical growth of outer integument or no growth at all was observed. To further support our scoring, some lines with representative morphologies were also analyzed by CRYO-SEM . Observation of transgene effects on ino-1 ovule morphologies are summarized in Table 1, and representative photos are shown in Fig. 4. Ovules of transgenic lines expressing VviINO.1 exhibited either wild-type growth of the outer integument or a “sup–like” phenotype with outer integument growing symmetrically on both sides of the ovuleprimordium. This demonstrates that VviINO.1 can restore the compromised outer integument growth during ovule development and, to some extent,40 litre pot also an asymmetric growth. Surprisingly, all transgenic lines expressing VviINO.2 rescued the outer integument growth too. In addition to two lines fully recovered to normal ovule morphology , six lines showed a partial “suplike” phenotype.

However, symmetric growth in these lines was much weaker compared to the sup mutant . Finally, one transgenic line showed a typical “weak-ino” phenotype . Ino-1 mutant is strongly affected in female fertility and homozygous plants produce approximately one to three seeds. Thus, the seed set allows additional evaluation of the complementation. In our hands, wild-type plants produced more than 50 seeds on average per silique. Transgenic lines presented a variable number of seeds per silique, but for all the seed set was significantly different compared to the ino-1 mutant. Only two lines carrying VviINO.1 showed a seed set comparable to that of wild type plants. Interestingly, both lines were previously classified as “wild type-like” according to their ovule morphology. Seed set in all other lines was significantly different to the wild type, even though also significantly different to ino-1, with an average seed number per silique proportional to the rescue of the wild-type ovule morphology scored by microscope . Finally, we have investigated the molecular basis of the different levels of complementation.As expected, no expression was found in the wild-type Arabidopsis, and lines transformed with VviINO.1 did not show any expression of VviINO.2. On the contrary, lines transformed with VviINO.2 produced both mRNAs, confirming that the VviINO.2 is likely an mRNA intermediate that can be successfully further processed in Arabidopsis to encode the complete functional protein, thus explaining the observed rescue in ovule morphology in these lines. Unfortunately, since we only conducted a relative quantification, the expression of VviINO.1 cannot be compared either to its endogenous expression level in grape nor to the expression of AthINO. Relative expression of both transgenes was variable in the different lines, and an inverted relationship between transgene expression and complementation level was observed, which was especially evident in lines carrying the VviINO.1 transgene. Altogether these results demonstrate that VviINO can restore outer integument growth in A. thaliana ino-1mutant as well as a partial asymmetric growth, therefore its function is conserved across the two species.

In this work, we have first of all identified the grapevine orthologue of the AthINO transcription factor, starting from the recent characterization of the grapevine YABBY gene family. We confirmed that, according to all recent grapevine genome annotations, VviINO is encoded by a unique gene located on top of chromosome 1 comprising six exons and five introns. Differing from what expected from current annotations, cDNA cloning resulted in the identification of two mRNAs, the second one retaining the intron four. Alternative splicing due to intron retention could have functional implications, especially when translation results in a mis-functional protein due to a frame shift mutation. However, we demonstrated that the VviINO.2 mRNA can be further spliced, even in a heterologous system, thus suggesting its presence is more likely related to an incomplete processing event or to an mRNA storing mechanism than working through translation in a mis-functional protein, as already reported in the literature for other plant genes. We confirmed a specific expression in flowers and fruit while no expression was found in other organs similarly as reported by other authors, in agreement with the specific role of INO in ovule development. In the past, detailed characterization of INO expression patterns in several species suggested wide conservation of INO function down to the early divergent Angiosperm clades. Despite that, more recent functional studies provided controversial results. While VIGs silencing of the NbINO orthologue in tobacco inhibited the growth of the outer cell layer of the integument, leading to a decrease in both integument extension and ovule curvature, the tomato SlINO coding region was not able to complement the Arabidopsis ino-1 mutant. With the final aim to deepen our knowledge of grapevine reproductive biology and especially ovule development to identify new targets for breeding purposes, we have enquired VviINO protein function conservation. By comparing protein sequences from 30 INO orthologues, we found that VviINO grouped in a clade with AthINO while both were more distantly related to SlINO. This supported the hypothesis that VviINO could complement the Arabidopsis ino-1 mutant phenotype, unlike SlINO.

Results of cross-species complementation demonstrate that the VviINO can indeed successfully functionally complement the Arabidopsis ino-1 mutant phenotype when expressed from the AthINO promoter. We analyzed the ovules of 6 independent T3 transgenic lines expressing the VviINO.1 cloned CDS and 9 independent T3 transgenic lines expressing the VviINO.2 alternative CDS which, as previously mentioned, can be further spliced in Arabidopsis to encode a functional VviINO protein identical to that encoded by the VviINO.1 transcript. Careful microscopic observations showed that in all lines the outer integument growth and some ovule curvature was rescued. Seed set evaluation also confirmed a significantly different behavior in the transgenic lines compared to the background ino-1 line. However, differently from Arabidopsis lines complemented with the AthINO, we observed a high number of transgenic lines displaying a “sup-like” phenotype, with also outer integument growth from the adaxial side of the ovule primordium associated with a still partially reduced seed set. This was reminiscent of behaviors previously observed in domain swap experiments. Replacement of the C-terminus of AthINO with AthCRC resulted in a significant proportion of transgenic lines that contained ovules with a “sup-like” phenotype,collection drainage supporting the involvement of this region in the repressive action of SUP. Our results strongly resembled these results, suggesting that, although VviINO can effectively complement the outer integument growth promotive effects in the ino-1 mutant, it was less responsive than the endogenous AthINO to the SUP inhibition. Interestingly, alignment of VviINO and AthINO protein sequences highlighted low conservation of the C-terminal portion in the grapevine protein , which could explain the less effective SUP suppression and the consequent “sup-like” phenotype. Furthermore, we speculated that the consequent abaxalization of the adaxial domain would be likely more pronounced in lines with higher transgene expression and that this could likely explain the observed variability in phenotypes. Accordingly, we found a correlation between compromised ovule asymmetric growth and the expression levels of VviINO.1 and VviINO.2 transcripts, which further supports our data interpretation . These data improve our understanding of grapevine ovule development, with potential implications also for table grape breeding. The major source of seedlessness currently exploited in cultivated grapevine has been recently characterized, being due to an amino acid substitution in the VviAGL11 gene controlling seed coatdevelopment and lignification, downstream of ovule development. Accordingly, no difference in the VviINO.1 and VviINO.2 transcripts expression was found in seeded and seedless grape varieties . However, alternative sources with potential implications for breeding purposes could also exist. The description of seedless normal size fruit production in Ts mutant of A. squamosa lacking the orthologous INO gene, beside the demonstrated functional conservation in ovule development of VviINO , supports the VviINO gene as a candidate for grapevine seedlessness. However, despite findings in Annona crop, in Arabidopsis ino-1, as previously mentioned, “fruit” development appears as compromised. The different effect of ino defects in ovule development and on seeds and fruit production in Annona and Arabidopsis species has been the subject of further characterizations. In these studies, the Arabidopsis ino-1 pollen tubes grow through the transmitting tract but were never observed inside the micropyle, and the majority of ovules fail to form embryo sacs. In contrast, in the A. squamosa Ts mutant, pollen tube growth was more normal often targeting the micropyle, and most of the ovules contained fully developed, but sometimes degenerating, embryo sacs.

Authors have speculated that the outer integument role in pollen tube guidance and embryo sac development could have been a recent acquisition in Arabidopsis. They suggested that the absence of an essential role of the outer integument in pollen tube guidance in Annona could be related to its endostomal type of micropyle, with the outer integument not fully covering the inner integument and participating in the micropyle. Furthermore, concerning embryo sac development, a higher sensitivity to changes in integument development could be due to a much thinner fraction of tissue around the female germline in the tenuinucellate Arabidopsis compared to the crassinucellate Annona. Interestingly, grapevine also shows an endostomal micropyle and several cell layers surrounding the embryo sac. Only comparative studies in grapevine plants knocked-out for the now confirmed grapevine functional VviINO will conclusively allow to define implications of the defect in outer integument development for pollen tube guidance and embryo sac development and thus on seed and fruit production, eventually validating the utility of VviINO for table grape breeding. Interestingly, early genetic studies on grape seedlessness reported that seed coat hardness and endosperm/embryo development were behaving as separate sub-traits, confirming the existence of alternative contributions. Moreover, more recently QTLs studies have revealed the contribution of a region located on top of chromosome 1 including VviINO gene to the total seeds fresh weight per berry trait, further supporting VviINO as a candidate for grapevine seedlessness. An association to seedlessness in this genomic part of Chr1 was also confirmed by resequencing of seedless and seeded varieties, even though no associated SNP located in this gene were found in the studied panel. In conclusion, sequence comparison and the rescue of the outer integument growth in all Arabidopsis ino-1 lines expressing the VviINO protein from the AthINO promoter demonstrate that VviINO is the AthINO orthologue and that it plays the same function in promoting outer integument growth during ovule development. The high number of transgenic lines displaying a “sup-like” phenotype found in our cross-species complementation suggests a reduced sensitivity of VviINO compared to AthINO to the Arabidopsis SUP-mediated repression of expression in the adaxial side of the ovule primordia. Therefore, the mechanism involved in the tight control of INO spatial expression for proper ovule asymmetric growth could have partially diverged in the grapevine. Now that the functional involvement of VviINO in outer integument growth during grape ovule development has been demonstrated, functional studies in grape can further elucidate the mechanism for the asymmetric growth and the impacts on fruit and seed formation and their potential implications for table grape breeding purposes.Almond [Prunus dulcis D.A. Webb, syn. P. amygdalus Batsch] is the most economically important temperate tree nut crop worldwide. Due to increasing demand, production areas are expanding into warm and cold climatic regions of both hemispheres. Almond world production is led by the USA , Australia , and Spain. The origin of almond within the Amygdalus subgenus, including cultivated almond and its wild relatives such as P. fenzliana Fritsh, P. bucharica Fedtsch, P. kuramica Kitam., and P. triloba Lindl took place ~5.88 million years ago. Almond originated in the arid mountainous regions of Central Asia, where it was first cultivated around 5000 years ago and then moved to the Mediterranean region and later to California and thesouthern hemisphere.