Floral development is strictly controlled by complex regulatory networks to ensure successful reproduction in plants [44]. In natural conditions, the transition from vegetative to reproductive growth is irreversible so the correct tissue patterning can be guaranteed during the floral development [45]. Here we identified a recessive mutant of SLGT11 gene which has the phenotypes resembling some previously characterized mutants. Sl, a male sterile mutant was first reported in 1954 in which stamens degenerated and fused with the style [37]. This defect in floral meristem finally caused longitudinal cracks on the surface of the parthenocarpic fruit. Stamenless mutant was characterized as reduced and stunted stamens which led to the transformation of stamen into carpel structure [38, 39]. The similar defects in slf mutant suggest that SLGT11 could function in the same pathway as these previously reported genes.
However, slf also showed a reversion of floral development to vegetative organs, indicating that meristem termination in flowers became defective. This reversion phenotype is not necessarily associated with the defect of fused stamen and carpel, as evidenced by a number of previously characterized mutants including TAP3, TPIB and TM6. These tomato homeotic protein mutants cause severe floral organ patterning. In Arabidopsis, a weak allele ag-4, the carpel was transformed into sepal while stamens and carpels in a strong allele ag-6 were entirely transformed into petals and sepals [46]. Despite new flowers formed in whorl four of ag-2 flowers [47], no leaves were seen, indicating this defect represents only the aberrant termination of flower meristem. A direct homologous gene of AG in tomato is TAG1. In line with the conserved function of AG, tag1 showed a transformation of stamens into sepals [32]. However, the reversion of floral meristem back to vegetative development was neither reported in Arabidopsis ag mutant, nor in tomato tag1 mutant in nature condition. However, slf, we identified here exhibited an inhibited termination of flower meristem and the floral development was reversed into vegetative organs.
Interestingly the homologous gene of SLGT11 in Medicago truncatula seemed to only function as C type gene as the phenotype of the Medicago mutant caused the transformation of stamens and carpels into petals and stamens [40, 42]. This report has not described the reversal phenotype of floral organs, indicating that SLGT11 function may have experienced sub-functionalization in different species. In Arabidopsis, this reversal phenomenon was observed in the triple mutant of class E gene sep1/2/3 [10]. When the function of these three E genes were lost, all floral organs transformed into sepals and blades. Such phenomenon was even more pronounced in tomato. After SEPALLATA homologous gene TM29 was silenced in tomato, the development of inner three floral organs was affected, and the aberrant stems with flowers and leaves were initiated from the fruits [21].
Based on the defects in stamen and carpel, as well as the reversal phenotype in slf mutant, we hypothesize that SLGT11 regulates the stamen and carpel through AG, and also maintains the flower identity together with TM29. In our observation, we found the phenotype of slf was stronger than that of TM29 and TAG1. In addition, both TM29 and TAG1 were significantly down-regulated in slf mutant, indicating that SLGT11 could function in the upstream of TM29 and TAG1. The high temperature treatment enhanced slf phenotypes by increasing floral organ reversal rate, promoting petals to turn green and making sepals appear in the fourth whorl. These phenotypes indicated that the expression of class A genes could spread to the position of the fourth whorl, which is consistent with the phenotype of class C gene deletion. Our qRT-PCR also showed that TM29 was significantly down-regulated at high temperature, supporting the hypothesis that SLGT11 could regulate TAG1 and TM29.
SLGT11 expressed extensively in the early stage of flower, but its expression gradually became concentrated in stamens and the vascular bundles of the middle columns of the ovary. It was speculated that SLGT11 plays roles in the initiation of each whorl of flower organs, especially the initiation of stamens. It has been reported that the expression of BCE gene which affects stamen development overlaps with SLGT11 expression domain. Class B genes including TAP3, TM6 and TPI were all shown to have expression in stamen position [29, 30]. The C gene TAG1 is also mainly expressed in the stamens and carpel during tomato flower development [32]. Compared with the class B and C genes, the expression of TM29 in early stage was more extensive, including vascular bundles, But during the later stage of floral development, TM29 expression is mainly concentrated in stamens and carpels [21], which overlaps with the expression region of SLGT11, and the SLGT11 also express in vascular bundles, which could be the origin of the abnormal stem. Compared with the WT, the expression of TAP3, TPI, TPIB and TM29 in slf mutant was all down-regulated, suggesting SLGT11 could regulate the BCE gene expression to promote stamen development. Therefore, SLGT11 could be one of regulators in addition to the ABC model genes that regulate floral organ development.
In tomato, the regulation of stamen development seems to be more complicated than that in Arabidopsis. When the class B genes including AP3 or PI in Arabidopsis were mutated, the petals turned into sepals, and the stamens were missing or turned into carpels [26]. In tomato, there are four homologous class B genes: TAP3, TM6, TPI and TPIB. Despite similar phenotypes were observed when TAP3 and TPIB were mutated, mutations in TM6 or TPI only caused stamen transformation into carpel without affecting petal and carpel [30]. Since no homologous shift phenomenon was observed in slf mutant, it is difficult to fully explain SLGT11 function in floral development by including SLGT11 into any class of genes in the classical ABCE model. It seemed that SLGT11 had a wider range of functions. It is possible that SLGT11 not only regulates one type of ABCE genes, but also affects ABCE genes together.
Floral development is strictly controlled by complex regulatory networks to ensure successful reproduction in plants [44]. In natural conditions, the transition from vegetative to reproductive growth is irreversible so the correct tissue patterning can be guaranteed during the floral development [45]. Here we identified a recessive mutant of SLGT11 gene which has the phenotypes resembling some previously characterized mutants. Sl, a male sterile mutant was first reported in 1954 in which stamens degenerated and fused with the style [37]. This defect in floral meristem finally caused longitudinal cracks on the surface of the parthenocarpic fruit. Stamenless mutant was characterized as reduced and stunted stamens which led to the transformation of stamen into carpel structure [38, 39]. The similar defects in slf mutant suggest that SLGT11 could function in the same pathway as these previously reported genes.
However, slf also showed a reversion of floral development to vegetative organs, indicating that meristem termination in flowers became defective. This reversion phenotype is not necessarily associated with the defect of fused stamen and carpel, as evidenced by a number of previously characterized mutants including TAP3, TPIB and TM6. These tomato homeotic protein mutants cause severe floral organ patterning. In Arabidopsis, a weak allele ag-4, the carpel was transformed into sepal while stamens and carpels in a strong allele ag-6 were entirely transformed into petals and sepals [46]. Despite new flowers formed in whorl four of ag-2 flowers [47], no leaves were seen, indicating this defect represents only the aberrant termination of flower meristem. A direct homologous gene of AG in tomato is TAG1. In line with the conserved function of AG, tag1 showed a transformation of stamens into sepals [32]. However, the reversion of floral meristem back to vegetative development was neither reported in Arabidopsis ag mutant, nor in tomato tag1 mutant in nature condition. However, slf, we identified here exhibited an inhibited termination of flower meristem and the floral development was reversed into vegetative organs.
Interestingly the homologous gene of SLGT11 in Medicago truncatula seemed to only function as C type gene as the phenotype of the Medicago mutant caused the transformation of stamens and carpels into petals and stamens [40, 42]. This report has not described the reversal phenotype of floral organs, indicating that SLGT11 function may have experienced sub-functionalization in different species. In Arabidopsis, this reversal phenomenon was observed in the triple mutant of class E gene sep1/2/3 [10]. When the function of these three E genes were lost, all floral organs transformed into sepals and blades. Such phenomenon was even more pronounced in tomato. After SEPALLATA homologous gene TM29 was silenced in tomato, the development of inner three floral organs was affected, and the aberrant stems with flowers and leaves were initiated from the fruits [21].
Based on the defects in stamen and carpel, as well as the reversal phenotype in slf mutant, we hypothesize that SLGT11 regulates the stamen and carpel through AG, and also maintains the flower identity together with TM29. In our observation, we found the phenotype of slf was stronger than that of TM29 and TAG1. In addition, both TM29 and TAG1 were significantly down-regulated in slf mutant, indicating that SLGT11 could function in the upstream of TM29 and TAG1. The high temperature treatment enhanced slf phenotypes by increasing floral organ reversal rate, promoting petals to turn green and making sepals appear in the fourth whorl. These phenotypes indicated that the expression of class A genes could spread to the position of the fourth whorl, which is consistent with the phenotype of class C gene deletion. Our qRT-PCR also showed that TM29 was significantly down-regulated at high temperature, supporting the hypothesis that SLGT11 could regulate TAG1 and TM29.
SLGT11 expressed extensively in the early stage of flower, but its expression gradually became concentrated in stamens and the vascular bundles of the middle columns of the ovary. It was speculated that SLGT11 plays roles in the initiation of each whorl of flower organs, especially the initiation of stamens. It has been reported that the expression of BCE gene which affects stamen development overlaps with SLGT11 expression domain. Class B genes including TAP3, TM6 and TPI were all shown to have expression in stamen position [29, 30]. The C gene TAG1 is also mainly expressed in the stamens and carpel during tomato flower development [32]. Compared with the class B and C genes, the expression of TM29 in early stage was more extensive, including vascular bundles, But during the later stage of floral development, TM29 expression is mainly concentrated in stamens and carpels [21], which overlaps with the expression region of SLGT11, and the SLGT11 also express in vascular bundles, which could be the origin of the abnormal stem. Compared with the WT, the expression of TAP3, TPI, TPIB and TM29 in slf mutant was all down-regulated, suggesting SLGT11 could regulate the BCE gene expression to promote stamen development. Therefore, SLGT11 could be one of regulators in addition to the ABC model genes that regulate floral organ development.
In tomato, the regulation of stamen development seems to be more complicated than that in Arabidopsis. When the class B genes including AP3 or PI in Arabidopsis were mutated, the petals turned into sepals, and the stamens were missing or turned into carpels [26]. In tomato, there are four homologous class B genes: TAP3, TM6, TPI and TPIB. Despite similar phenotypes were observed when TAP3 and TPIB were mutated, mutations in TM6 or TPI only caused stamen transformation into carpel without affecting petal and carpel [30]. Since no homologous shift phenomenon was observed in slf mutant, it is difficult to fully explain SLGT11 function in floral development by including SLGT11 into any class of genes in the classical ABCE model. It seemed that SLGT11 had a wider range of functions. It is possible that SLGT11 not only regulates one type of ABCE genes, but also affects ABCE genes together.