Several lines of evidence from this study suggest that MdANT1
function as transcription factors in apple. A motif of basic residues (KKKR) is essential for the nuclear localization of ANT, as replacement of two lysine residues within this motif resulted in a loss of nuclear localization in Arabidopsis [36
]. In MdANT1 and MdANT2, a major part of this element is conserved (TKKR), strongly suggesting that these ANTs are targeted to the nucleus, consistent with their proposed roles as transcription factors. The Arabidopsis ANT binds to the DNA at a consensus site of 16 bases through two AP2 domains and a conserved linker region [24
]. MdANT1 and MdANT2 shared greater than 88% sequence identity with the Arabidopsis ANT within these regions. All of the 19 residues identified as essential for the DNA binding activity of the Arabidopsis ANT [25
] are conserved in the apple ANTs suggesting that they may bind to similar DNA elements, further supporting their role as transcription factors. Domains within the amino-terminal region are also essential for the transcriptional activation properties of the Arabidopsis ANT [36
]. Although the apple ANTs display limited conservation of residues with that of the Arabidopsis ANT in this region, it should be noted that other plant ANTs also display significant sequence divergence within this region, indicating that distinct, species-specific features may be required for the transcriptional activation properties of the ANTs.
MdANT1 and MdANT2 are expressed in regions associated with fruit growth and development [a) ovary and floral-tube tissues before bloom; b) core and cortex tissues during early fruit growth]. MdANT1 and MdANT2 display high expression before bloom in the ovary as well as the floral-tube regions, strongly suggesting their association with cell production-mediated growth of the ovary and floral-tube tissues before bloom. Expression of MdANT1 and MdANT2 declines within these tissues during the period of temporary cessation of growth and quiescence in cell production (around full bloom). Subsequently, the expression of the ANTs increases sharply within the cortex tissue while little change in their expression is observed within the core tissue between bloom and 10 DAFB, coincident with the resumption of growth and re-initiation of cell production in the cortex during early fruit development. The sharp increase in expression at 10 DAFB within the cortex is likely triggered by pollination and/or fertilization and may mediate fruit set. MdANT1 and MdANT2 expression is high during the cell production-mediated phase of early fruit growth and subsequently declines greatly during exit from this phase. This pattern of expression is conserved under conditions of different carbohydrate availability and across genotypes differing in their fruit growth potential. Together, the data presented here indicate that the expression of MdANT1 and MdANT2 is consistently and closely associated with cell production during fruit growth in apple. Therefore, it is proposed that ANTs are important components of a developmental program that controls the extent of cell production and thereby regulates fruit growth in apple.
Cell production and fruit growth are limited by carbohydrate availability in many plant species [4
]. Consistent with previous studies, increase in carbohydrate availability through manual thinning during early fruit development in GS enhanced fruit growth and final fruit size. This was primarily achieved through sustained cell production in the fruit cortex during early fruit growth and a higher relative cell production rate, especially towards the later stages of the cell production phase. These data indicate that carbohydrate limitation due to increased competition among sinks decreases the extent of cell production in the fruit cortex. Under conditions of higher carbohydrate availability, the expression of MdANT1
was several-fold higher than that under carbohydrate limitation. Additionally, MdANT2
was up-regulated (>3-fold at 25 DAFB compared to 11 DAFB) in response to an increase in carbohydrate availability in GS. These data suggest that an increase in carbohydrate availability enhances the expression of the ANT
genes, especially MdANT2
, thereby increasing the competence of the fruit cortex cells for cell production. Hence, it may be proposed that the ANTs
, particularly MdANT2
, mediate the effects of carbohydrate availability on cell production and fruit growth. Increase in competence for cell production may be achieved either through an increase in the proportion of fruit cortex cells undergoing proliferation or through an increase in the capacity of individual cortex cells for division. Increase in carbohydrate availability also led to a minor increase in cell area during later stages of fruit growth in GS, inconsistent with previously reported results in the apple cultivar, ‘Empire’ [4
], but consistent with results in tomato fruit [39
]. It is likely that an increase in sink strength as a result of higher fruit cortex cell number in thinned fruit may subsequently aid in increasing the extent of cell expansion.
Comparison of apple genotypes differing in their growth potential further supports the proposed roles of the ANTs
in regulating cell production. Although, it is possible that some of the differences observed between the two genotypes are due to environmental effects, the overall patterns of fruit growth and gene expression reported here were consistent with that observed in other studies during different years (data not shown). The initial cell number and the duration of the cell production phase were similar in ‘Gala’ and GS (around 198 and 184 GDD after bloom, respectively), indicating that the higher final cell number within the fruit cortex of GS in comparison to that in ‘Gala’ was due to differences in the pattern of progression in cell production during early fruit development. GS fruit cortex cells displayed a more gradual increase in cell number after bloom in comparison to those of ‘Gala’ which displayed a short-lived early burst in cell production. In fact, the RCPR in GS reached the maxima around 54 GDD after that in ‘Gala’. Subsequently, the rate of cell production in GS was higher than that in ‘Gala’, especially between 73 and 184 GDD after bloom. Expression of MdANT1
in the two genotypes matched their respective patterns of cell production. Expression of these genes in GS was sustained at higher levels for a longer duration while in ‘Gala’, the expression of these genes displayed an initial rapid burst followed by a rapid decline. The expression of both these genes was higher in GS than in ‘Gala’ during the final stages of the cell production phase (around 129 and 184 DAFB). Sustained competence for cell production as a result of this pattern of expression of the ANTs
may allow for enhanced cell production and a higher final cell number in GS. Final cell number is often an important determinant of variation in fruit size across genotypes [5
]. Differences in the pattern of expression of the ANT
genes during early fruit growth may affect the final cell number and thereby final fruit size across genotypes. Similarly, differences in the pattern of expression of FW2.2
are thought to determine fruit size differences across tomato genotypes [44
]. Hence, it is likely that MdANT1
also function as regulators of fruit size in apple.
Expression of the apple ANT
genes was correlated with that of several positive regulators of the cell cycle, including B-type CDKs, A- and B-type cyclins, and MdDEL1
during different stages of flower and early fruit development. During the period of exit from cell production (around 15-25 DAFB in ‘Gala’), the expression of several cell cycle genes positively associated with cell production declined, while that of genes negatively associated with cell production increased [1
]. These changes in the expression of the cell cycle genes coincide with the decline in the expression of MdANT1
observed in this study. In fact, the expression patterns of the ANT
genes during fruit growth display high similarity with those of the core cell cycle genes involved primarily in the regulation of the G2-M phases of the cell cycle. Co-expression of these genes suggests coordinated regulation and their involvement in a common biological process [45
]. Considering that the ANT
genes may function as transcription factors, it is possible that MdANT1
regulate the expression of the core cell cycle genes and thereby coordinate cell production during fruit growth. In Arabidopsis, increased cell production as a result of the over-expression of ANT
was associated with the prolonged expression of D3-type cyclins [19
]. Identification of the genes targeted for direct regulation by the ANTs is essential to test this hypothesis.
The general similarities in the expression patterns of MdANT1
suggest overlapping roles for these genes in regulating flower and fruit development. In Arabidopsis, expression of four PLT
genes (members of the AP2 sub-family) in overlapping as well as specific regions of the root allows for PLT concentration-dependent regulation of root growth and development [32
]. Similarly in apple, the combined activity of MdANT1 and MdANT2 may have an additive effect on cell production and fruit growth. However, certain key differences between MdANT1
were also noted. The expression of these genes in the core tissue differed slightly during early fruit development. MdANT1 and MdANT2 also differed within the AP2-repeats and linker region in three residues (A354-S352; T365-A363; S388-F386, MdANT1-MdANT2, respectively). If the DNA binding characteristics are affected by the above residues, MdANT1 and MdANT2 may regulate different pools of downstream target genes. Together, the above data suggest that MdANT1
may also have distinct roles in regulating fruit growth and development. Functional characterization of MdANT1
and the identification of their downstream targets in vivo
are essential to determine their specific roles in regulating fruit growth.
All of the AIL
genes studied here contained the characteristic AP2-repeats and the conserved linker region suggesting that they function as transcriptional regulators. These genes displayed elevated expression during flower development and a sharp decline in expression during early fruit development, suggesting that they may be primarily involved in regulating flower growth and development in apple. In Arabidopsis, many of the AIL
genes are involved in regulating floral organ growth and development [21
]. MdAIL4 and MdAIL5 share significant amino acid identity with AtAIL5 and AtAIL6 respectively, genes which have been previously reported to regulate organ growth [21
]. Further characterization of the tissue-specific patterns of expression and the functional characterization of the AIL
genes is essential to determine their specific roles in apple.