Identification of TALE Transcription Factor Family and Expression Patterns Related to Fruit Chloroplast Development in Tomato (Solanum lycopersicum L.)

The TALE gene family is an important transcription factor family that regulates meristem formation, organ morphogenesis, signal transduction, and fruit development. A total of 24 genes of the TALE family were identified and analyzed in tomato. The 24 SlTALE family members could be classified into five BELL subfamilies and four KNOX subfamilies. SlTALE genes were unevenly distributed on every tomato chromosome, lacked syntenic gene pairs, and had conserved structures but diverse regulatory functions. Promoter activity analysis showed that cis-elements responsive to light, phytohormone, developmental regulation, and environmental stress were enriched in the promoter of SlTALE genes, and the light response elements were the most abundant. An abundance of TF binding sites was also enriched in the promoter of SlTALE genes. Phenotype identification revealed that the green shoulder (GS) mutant fruits showed significantly enhanced chloroplast development and chlorophyll accumulation, and a significant increase of chlorophyll fluorescence parameters in the fruit shoulder region. Analysis of gene expression patterns indicated that six SlTALE genes were highly expressed in the GS fruit shoulder region, and four SlTALE genes were highly expressed in the parts with less-developed chloroplasts. The protein-protein interaction networks predicted interaction combinations among these SlTALE genes, especially between the BELL subfamilies and the KNOX subfamilies, indicating a complex regulatory network of these SlTALE genes in chloroplast development and green fruit shoulder formation. In conclusion, our result provides detailed knowledge of the SlTALE gene for functional research and the utilization of the TALE gene family in fruit quality improvement.


Introduction
Homeobox (Homologous alien box) genes, which produce the same phenotype when mutated, are a family of transcription factors (TFs) that are highly conserved in plants, animals, and fungi [1]. The first homeobox gene was identified in Drosophila, and KNOTTED1 (KN1 or ZmKN1), obtained from a functionally acquired mutant of maize, was the first homeobox gene isolated from plants [2,3]. Homeobox genes encode a typical DNA-binding domain of 60 amino acids, known as the homeodomain (HD), which forms three-helix regions; the first and second helices form a loop structure, and the second and third helices 2 of 17 form a helix-corner-helical structure [4]. Many homeobox genes have been subsequently identified from all major eukaryotic lineages [5]. Plant homeobox genes can be classified into 11 classes based on their structural characteristics and phylogeny: HD-ZIP (including HD-ZIP I to HD-ZIP IV), WOX, NDX, PHD, PLINC, LD, DDT, SAWADEE, PINTOX, KNOX, and BELL [6].
Unlike typical homeobox genes, the TALE (Three Amino acid Loop Extension) homeobox genes encode an atypical structure of 63 amino acids with three additional amino acid residues (P-Y-P) inserted between the first and the second helix [7][8][9]. The TALE family consists of five members, but only members of the KNOX (KNOTTED-like homeodomain) subfamilies and BELL (BEL1-like homeodomain) subfamilies have been observed in plants. KNOX proteins contain a MEINOX (from MEIS and KNOX) domain composed of KNOX1 and KNOX2 subdomains separated by a flexible linker, an ELK domain, and a TALE homeodomain. The MEINOX domain mediates the interaction between KNOX and BELL proteins [10,11]. In addition, KNATM, a novel TALE family member without a homeodomain, interacts with BELL proteins through the MEINOX domain and interacts with BP via the BPID (BP interacting domain) [11,12]. BELL proteins contain a TALE homeodomain and a MID (MEINOX interacting domain, or POX) composed of the SKY and BELL domains [13,14]. The interaction of BELL proteins with KNOX proteins targets the KNOX-BELL heterodimer to the nucleus [15,16]. The function of the ELK domain remains unclear, and it might be related to interactions between proteins. However, some studies have suggested that it contains a nuclear localization signal and might be involved in inhibiting the transcription of genes [14].
TALE family members play an important role in hormone regulation, signal transduction, meristem formation, organ morphogenesis, and the development of plants [17][18][19][20][21][22]. For example, ATH1 (Arabidopsis homeobox 1) interacts with STM (homeobox protein SHOOT MERISTEMLESS) and KNAT2 (Arabidopsis KNOX 2) and plays a role in the development of meristems and inflorescence tissue [23,24]. The interaction between BLH6 and KNAT7 affects the formation of the secondary cell wall [25]. TKN4 and TKN2 regulate chloroplast development and the distribution of chloroplasts in tomato fruit by targeting APRR2-like and GLK2 [22,26]. SlBEL11 is responsible for regulating chloroplast development, chlorophyll accumulation, and the chlorophyll quality of tomato fruit by modifying the expression of POR, CAB, and TKN2 [27]. SlBEL4 regulates chloroplast development and cell wall metabolism via the transcriptional regulation of CHLD, POR, TKN2, and PE [17]. Although TALE family members involved in the regulation of plant development have been well characterized, the functions of TALE genes in tomato fruits have been less studied by comparison because TALE family members have diverse functions and often show complex interactions with other genes.
To identify potential TALE members and understand the evolution and function of TALE genes in tomato, genome-wide identification and analysis of the TALE genes were conducted. The expression of TALE genes was evaluated between the stem end (fruit shoulder area) and the base (fruit navel area) of non-uniform ripening tomato fruit. The results of the systematic identification and analysis of TALE genes provide valuable insights into the regulatory mechanisms of TALE genes in chloroplast development and the formation of green fruit shoulders during tomato fruit development.

Identification and Sequence Analysis of SlTALE Family Members
A total of 52 and 28 candidate TALE members were identified from the tomato genome by HMMER search and BLASTP, respectively. Next, the two identification results were combined and taken together. The results were further verified by PfamScan and SMART; a total of 24 SlTALE members were confirmed in tomato (Table 1). Fourteen SlTALE members belonged to the BELL subfamilies and possessed a POX domain and homeodomain, and the remaining 10 members belonged to the KNOX subfamilies. In the KNOX subfamily, six members contained KNOX1, KNOX2, ELK, and HOX domains, but the KNOX2 domain of SlTALE2 was located at the same position as the first ELK. SlTALE12 and SlTALE18 lacked a KNOX1 domain and ELK domain, respectively, and SlTALE7 and SlTALE13 lacked both ELK and HOX domains (Table S1).
The 24 SlTALE genes were predicted to encode proteins ranging from 169 to 726 amino acids in length. The molecular weights ranged from 19.5 to 79.2 kDa, and isoelectric points ranged from 4.38 to 8.91. The instability index ranged from 42.9 to 65.2, and the aliphatic index varied from 60.7 to 86.4, suggesting that SlTALEs were all unstable and fat-soluble proteins. The GRAVY value was between −0.995 and −0.385, suggesting that SlTALEs were all hydrophilic proteins. The signal peptide prediction showed that all SlTALE proteins were non-secreted proteins. Subcellular localization prediction suggested that all SlTALE proteins were distributed in the nucleus (Table 1).

Phylogenetic Analysis of the TALE Family
A phylogenetic tree of TALE family genes from various plant species was constructed to characterize the evolutionary relationships among these genes, elucidate their biological functions, and clarify the similarities in the SlTALE proteins among tomato and other plants ( Figure 1, Table S2).

Analysis of Gene Structures, Chromosomal Locations, and Conserved Motifs
SlTALE genes were mapped to the tomato genome to study their distribution. SlTALE genes were randomly and non-randomly distributed across all 12 chromosomes; there were 2-4 members on all chromosomes except 3, 5, 7, 9, 10, and 12, which had only 1 member (Figure 2A). Both chromosomes 4 and 8 had the highest (16.66%) number of genes compared with the other chromosomes. Furthermore, most of the SlTALE genes Figure 1. Phylogenetic tree of the TALE family. Phylogenetic relationships among 203 TALE proteins in Solanum lycopersicum, Solanum tuberosum, Arabidopsis thaliana, Hordeum vulgare, Nicotiana benthamiana, Oryza rufipogon, and Vitis vinifera. The phylogenetic tree was constructed using MEGA6 with a neighbor-joining method and 1000 bootstrap replicates. Solanum lycopersicum, Solanum tuberosum, Arabidopsis thaliana, Hordeum vulgare, Nicotiana benthamiana, Oryza rufipogon, and Vitis vinifera were marked with a red star, red rectangle, white red star, white green circle, green check, white red triangle, and white green triangle, respectively.

Analysis of Gene Structures, Chromosomal Locations, and Conserved Motifs
SlTALE genes were mapped to the tomato genome to study their distribution. SlTALE genes were randomly and non-randomly distributed across all 12 chromosomes; there were 2-4 members on all chromosomes except 3, 5, 7, 9, 10, and 12, which had only 1 member (Figure 2A). Both chromosomes 4 and 8 had the highest (16.66%) number of genes compared with the other chromosomes. Furthermore, most of the SlTALE genes were located at both ends of tomato chromosomes, especially at the long arm. The syntenic relationship analysis showed that there were no collinear pairs of SlTALE genes, but tandem duplications were observed between SlTALE10 and SlTALE11 as well as between SlTALE22 and SlTALE23 ( Figure 2A). ranged from 4 to 7 and from 3 to 6, respectively ( Figure 2D). The gene structure and distribution of conserved motifs were similar among members of the same subfamily. For example, four KNOX-II members (SlTALE15, SlTALE16, SlTALE18, and SlTALE24) contained five to six exons and four identical motifs, and three BELL-IV members (SlTALE1, SlTALE14, and SlTALE22) contained five exons and seven identical motifs ( Figure 2C). The above results indicated that the SlTALE family members were conserved, and the differences in gene structure or protein-conserved motifs might be related to their various biological functions.  Structural analysis showed that the number of exons and introns of the SlTALE genes ranged from 4 to 7 and from 3 to 6, respectively ( Figure 2D). The gene structure and distribution of conserved motifs were similar among members of the same subfamily. For example, four KNOX-II members (SlTALE15, SlTALE16, SlTALE18, and SlTALE24) contained five to six exons and four identical motifs, and three BELL-IV members (SlTALE1, SlTALE14, and SlTALE22) contained five exons and seven identical motifs ( Figure 2C). The above results indicated that the SlTALE family members were conserved, and the differences in gene structure or protein-conserved motifs might be related to their various biological functions.

Analysis of Cis-Elements and TF Binding Sites
To characterize the functions and regulatory mechanisms of SlTALE family genes, the cis-elements and TF binding sites of their promoter regions were analyzed ( Figure 3, Table S3). A total of 41 different cis-elements identified in the promoter of SlTALE family genes were detected in response to 16 signal factors: light, phytohormone (abscisic acid, auxin, gibberellin, and zein), developmental regulation (circadian control, endosperm, flavonoid biosynthesis, endosperm, and meristem), and environmental stress (anaerobic, anoxic, defense and stress, drought, low temperature, methyl jasmonate, and salicylic acid). A total of 21 cis-elements involved in light responsiveness regulation were detected, and they accounted for 53.4% of all cis-elements ( Figure 3A,B, Table S3). According to the TF binding site analysis, a total of 1045 transcription factor binding sites belonging to 33 different TFs were identified in the promoters of SlTALE family genes. ( Figure 3C, Table S3). In plants, these TFs regulate various metabolic pathways, suggesting that the regulatory functions of SlTALE family genes are diverse.

Analysis of Cis-Elements and TF Binding Sites
To characterize the functions and regulatory mechanisms of SlTALE family genes, the cis-elements and TF binding sites of their promoter regions were analyzed (Figure 3, Table S3). A total of 41 different cis-elements identified in the promoter of SlTALE family genes were detected in response to 16 signal factors: light, phytohormone (abscisic acid, auxin, gibberellin, and zein), developmental regulation (circadian control, endosperm, flavonoid biosynthesis, endosperm, and meristem), and environmental stress (anaerobic, anoxic, defense and stress, drought, low temperature, methyl jasmonate, and salicylic acid). A total of 21 cis-elements involved in light responsiveness regulation were detected, and they accounted for 53.4% of all cis-elements ( Figure 3A,B, Table S3). According to the TF binding site analysis, a total of 1045 transcription factor binding sites belonging to 33 different TFs were identified in the promoters of SlTALE family genes. ( Figure 3C, Table S3). In plants, these TFs regulate various metabolic pathways, suggesting that the regulatory functions of SlTALE family genes are diverse.

Interaction Network of the SlTALE Proteins
Clarifying the functional relationships among SlTALE proteins is important for characterizing their potential functions and regulatory pathways. A protein-protein interaction network among SlTALE proteins was predicted using STRING software (Figure 4). Most of the SlTALE proteins could interact with multiple other members except SlTALE2 and SlTALE10, as no interactions of SlTALE2 and SlTALE10 with other members were

Interaction Network of the SlTALE Proteins
Clarifying the functional relationships among SlTALE proteins is important for characterizing their potential functions and regulatory pathways. A protein-protein interaction network among SlTALE proteins was predicted using STRING software (Figure 4). Most of the SlTALE proteins could interact with multiple other members except SlTALE2 and SlTALE10, as no interactions of SlTALE2 and SlTALE10 with other members were observed. SlTALE11 and BELL-III subfamily members (SlTALE4, SlTALE6, SlTALE9, and SlTALE19) were central to the interaction network and interacted strongly with most of the KNOX subfamily members, including KNOX-IV (SlTALE5 and SlTALE8), 3 KNOX-II members (SlTALE15, SlTALE16, and SlTALE24), and 1 KNOX-III member (SlTALE12). In addition, BELL-II subfamily members (SlTALE20, SlTALE21, and SlTALE23) might interact with KNOX-IV subfamily members (SlTALE5 and SlTALE8) as well as 1 KNOX-III member (SlTALE12). BELL-IV subfamily members (SlTALE1, SlTALE14, and SlTALE22) might interact with SlTALE18 and SlTALE18. Moreover, the KNOX-I member SlTALE7 might interact with BELL-II (SlTALE20, SlTALE21, and SlTALE23), BELL-III (SlTALE4, SlTALE6, SlTALE9, and SlTALE19), and KNOX-II (SlTALE15, SlTALE16, SlTALE18, and SlTALE24). Functional interactions, especially gene co-occurrence and co-expression, were mainly observed between the BELL subfamilies and the KNOX subfamilies, indicating that these genes might be synergistically involved in certain regulatory pathways.

Characterization of Fruit Phenotype and Chloroplast Development of the GS Mutant and WT
Unlike uniformly pigmented WT tomato fruit (TTD302A), the GS mutant was characterized by fruit with various pigments and a dark green fruit shoulder ( Figure 5A). The shoulder of GS fruit accumulated significantly large amounts of chlorophyll compared with the WT (both in the shoulder and base) ( Figure 5D). The photosynthetic performance of WT and GS fruits was assessed by determining the chlorophyll fluorescence parameters ( Figure 5B,E). Chlorophyll fluorescence parameters, including minimal fluorescence (Fo), maximal fluorescence (Fm), and variable fluorescence (Fv), were significantly higher in

Characterization of Fruit Phenotype and Chloroplast Development of the GS Mutant and WT
Unlike uniformly pigmented WT tomato fruit (TTD302A), the GS mutant was characterized by fruit with various pigments and a dark green fruit shoulder ( Figure 5A). The shoulder of GS fruit accumulated significantly large amounts of chlorophyll compared with the WT (both in the shoulder and base) ( Figure 5D). The photosynthetic performance of WT and GS fruits was assessed by determining the chlorophyll fluorescence parameters ( Figure 5B,E). Chlorophyll fluorescence parameters, including minimal fluorescence (Fo), maximal fluorescence (Fm), and variable fluorescence (Fv), were significantly higher in GS mutant fruits than in WT fruits ( Figure 5E). In addition, chloroplast development was accelerated and chloroplast number and size, the number of grana, and thylakoid density were higher in the shoulder of GS fruit ( Figure 5C). These results suggested that the green color formation of the shoulder in GS mutant fruit was promoted by chloroplast development and chlorophyll biosynthesis and accumulation.

Conserved and Divergent Patterns of SlTALE Genes Expression during Chloroplast Development and Green Shoulder Formation in Tomato Fruit
The expression of genes is often related to their function. To gain insight into the characteristics and functions of SlTALE genes in tomato fruit, the expression patterns of genes were analyzed based on RNA-seq data (Table S4). Most SlTALE genes were expressed during different fruit developmental stages in the GS mutant and WT, except for eight genes (SlTALE4, SlTALE7, SlTALE8, SlTALE10, SlTALE12, SlTALE13, SlTALE17, and SlTALE23) (Figure 6). Among the 16 expressed genes, 5 genes (SlTALE1, SlTALE14, SlTALE15, SlTALE18, and SlTALE22) had high expression levels but nonsignificant differences between the fruit shoulder of GS and WT. SlTALE2 and SlTALE5 exhibited similar expression patterns with significantly higher expression in the shoulder of GS compared with that of WT. The expression of SlTALE16 was higher in GS than in WT at 28 dpa and P. SlTALE24 was expressed exclusively in the fruit shoulder of WT before 28 dpa and was highly expressed in both GS and WT after 28 dpa. The expression profiles of SlTALE20 and SlTALE21 were similar with higher expression in GS than in WT. SlTALE6, SlTALE9, and SlTALE19 (all of which are in the same subfamily) were more highly expressed in WT than in GS. SlTALE3 and SlTALE11 were more highly expressed in GS than in WT.

Conserved and Divergent Patterns of SlTALE Genes Expression during Chloroplast Development and Green Shoulder Formation in Tomato Fruit
The expression of genes is often related to their function. To gain insight into the characteristics and functions of SlTALE genes in tomato fruit, the expression patterns of genes were analyzed based on RNA-seq data (Table S4). Most SlTALE genes were expressed during different fruit developmental stages in the GS mutant and WT, except for eight genes (SlTALE4, SlTALE7, SlTALE8, SlTALE10, SlTALE12, SlTALE13, SlTALE17, and SlTALE23) ( Figure 6). Among the 16 expressed genes, 5 genes (SlTALE1, SlTALE14, SlTALE15, SlTALE18, and SlTALE22) had high expression levels but nonsignificant differences between the fruit shoulder of GS and WT. SlTALE2 and SlTALE5 exhibited similar expression patterns with significantly higher expression in the shoulder of GS compared with that of WT. The expression of SlTALE16 was higher in GS than in WT at 28 dpa and P. SlTALE24 was expressed exclusively in the fruit shoulder of WT before 28 dpa and was highly expressed in both GS and WT after 28 dpa. The expression profiles of SlTALE20 and SlTALE21 were similar with higher expression in GS than in WT. SlTALE6, SlTALE9, and SlTALE19 (all of which are in the same subfamily) were more highly expressed in WT than in GS. SlTALE3 and SlTALE11 were more highly expressed in GS than in WT. had similar expression profiles in fruits from different tomato materials ( Figure 6A,B). BELL-IV subfamily genes (SlTALE1, SlTALE14, and SlTALE22) and KNOX-II (SlTALE15, SlTALE16, SlTALE18, and SlTALE24) exhibited similar expression patterns in different tomato materials and different parts of the fruit, indicating that the functions of these genes were conserved but not necessarily involved in the regulation of chloroplast development in tomato fruit ( Figure 6A,B).

Expression Analysis of SlTALE Genes in Tomato Fruit
To further investigate the SlTALE genes involved in chloroplast development and green shoulder formation in tomato fruit, the RNA-seq results of 16 expressed genes were validated via RT-qPCR (Figure 7 and Table S6). These genes can be divided into three categories according to their expression levels in the shoulder and base of tomato fruit. Eight of the sixteen analyzed genes, SlTALE2, SlTALE3, SlTALE5, SlTALE6, SlTALE11, SlTALE19, SlTALE20, and SlTALE24, showed high expression in the GS fruit shoulder region (Gs) in more developed chloroplasts. Four analyzed genes, SlTALE9, SlTALE18, SlTALE21, and SlTALE22, showed high expression in the parts with less-developed chloroplasts such as Gb, Ws, and Wb. Four genes, SlTALE1, SlTALE14, SlTALE15, and SlTALE16, showed no difference in expression among all parts of the fruit (Figure 7). In general, the expression patterns of these 16 genes in tomato fruits were consistent with the RNA-seq data. RNA-seq data of "Ailsa Craig" tomato (Table S5) were used to investigate the expression profile of the SlTALE family genes in different tomato materials. SlTALE family genes had similar expression profiles in fruits from different tomato materials ( Figure 6A,B). BELL-IV subfamily genes (SlTALE1, SlTALE14, and SlTALE22) and KNOX-II (SlTALE15, SlTALE16, SlTALE18, and SlTALE24) exhibited similar expression patterns in different tomato materials and different parts of the fruit, indicating that the functions of these genes were conserved but not necessarily involved in the regulation of chloroplast development in tomato fruit ( Figure 6A,B).

Expression Analysis of SlTALE Genes in Tomato Fruit
To further investigate the SlTALE genes involved in chloroplast development and green shoulder formation in tomato fruit, the RNA-seq results of 16 expressed genes were validated via RT-qPCR (Figure 7 and Table S6). These genes can be divided into three categories according to their expression levels in the shoulder and base of tomato fruit. Eight of the sixteen analyzed genes, SlTALE2, SlTALE3, SlTALE5, SlTALE6, SlTALE11, SlTALE19, SlTALE20, and SlTALE24, showed high expression in the GS fruit shoulder region (Gs) in more developed chloroplasts. Four analyzed genes, SlTALE9, SlTALE18, SlTALE21, and SlTALE22, showed high expression in the parts with less-developed chloroplasts such as Gb, Ws, and Wb. Four genes, SlTALE1, SlTALE14, SlTALE15, and SlTALE16, showed no difference in expression among all parts of the fruit (Figure 7). In general, the expression patterns of these 16 genes in tomato fruits were consistent with the RNA-seq data.

Discussion
TALE family genes play a crucial role in plant development and are widely distributed in the plant kingdom, but the number of these genes varies greatly among different plant species [33][34][35][36]. Tomato is one of the most widely grown vegetable crops, and it is an important model plant for studying fruit development [37]. Previous studies have shown that several TALE family genes regulate tomato fruit development, including fruit cell wall degradation, chloroplast development, pigment metabolism, fruit coloration and quality [17,22,27]. However, the number of TALE genes, their biological features, and regulatory functions in tomato fruit development remain unknown. In this study, 24 SlTALE family members were identified in tomato. Their biological features, structure, phylogenetic relationships, and expression regulation were characterized during tomato fruit development.
Gene organization, also known as exon-intron structure, plays a key role in the evolution of multigene families and can provide additional information for phylogenetic analysis [38,39]. Our results revealed a significant correlation between phylogeny and gene organization among SlTALEs, especially within the same groups and subfamilies ( Figure 2). In addition, intron number is usually associated with the sensitivity of gene transcriptional regulation [40]. The number and length of introns are negatively correlated Figure 7. Confirmation of the expression patterns of SlTALE genes in fruit by qRT-PCR. Sixteen differentially expressed or highly expressed genes were used, as well as the pericarp of the tomato fruit at 28 dpa. The values shown are the means ± standard deviation of three replicates. Different letters among columns denote significant differences among the treatments at p < 0.05.

Discussion
TALE family genes play a crucial role in plant development and are widely distributed in the plant kingdom, but the number of these genes varies greatly among different plant species [33][34][35][36]. Tomato is one of the most widely grown vegetable crops, and it is an important model plant for studying fruit development [37]. Previous studies have shown that several TALE family genes regulate tomato fruit development, including fruit cell wall degradation, chloroplast development, pigment metabolism, fruit coloration and quality [17,22,27]. However, the number of TALE genes, their biological features, and regulatory functions in tomato fruit development remain unknown. In this study, 24 SlTALE family members were identified in tomato. Their biological features, structure, phylogenetic relationships, and expression regulation were characterized during tomato fruit development.
Gene organization, also known as exon-intron structure, plays a key role in the evolution of multigene families and can provide additional information for phylogenetic analysis [38,39]. Our results revealed a significant correlation between phylogeny and gene organization among SlTALEs, especially within the same groups and subfamilies (Figure 2). In addition, intron number is usually associated with the sensitivity of gene transcriptional regulation [40]. The number and length of introns are negatively correlated with the ability of plants to adapt to diverse developmental processes and environmental conditions [41,42]. The diverse gene structures and intron phases of SlTALEs in different subfamilies might be related to their distinct biological functions. There was also some similarity observed within groups and subfamilies, as these members shared similar domains and possessed a similar motif distribution (Figure 2). The consistency in gene organization, domains, motifs, and phylogenetic relationships demonstrates that these genes were correctly classified.
Gene duplications and tandem duplication of chromosomal regions affect the formation of gene families [43]. Chromosomal distribution and syntenic relationship analysis of SlTALE were performed to clarify the origin of the SlTALE family. The SlTALE genes were unevenly distributed on every tomato chromosome, and more SlTALE genes were located at the long arm of the chromosome, which facilitates the exchange of DNA fragments and gene recombination [44]. No collinear pairs of SlTALE genes were detected, and two tandem duplications were observed ( Figure 2); this indicated that SlTALE family members might originate from different ancestors with high evolutionary diversity. These properties may explain the functional differences among SlTALE genes.
Promoter activity plays an important role in the regulation of gene functions [45]. Analysis results of the cis-elements and TF binding sites of the promoter regions revealed that SlTALE genes were found in response to various signal factors, such as light, phytohormone, developmental regulation, and environmental stress, and also are regulated by a variety of transcription factors (Figure 3 and Table S3). These properties might also account for the diversity of the regulatory functions of SlTALE family members and indicate that SlTALE family genes play an important regulatory role in different biological processes of plant development, especially those related to light regulation (e.g., photosynthesis, plastid development, and pigment metabolism) [8,12].
The expression patterns of genes are related to their potential functions. We analyzed the expression profiles of the 24 SlTALE genes. Some SlTALE genes showed specific expression patterns between the shoulder and base of tomato fruit during fruit development (chloroplast development and green shoulder formation). Six SlTALE genes (SlTALE2, SlTALE3, SlTALE5, SlTALE11, SlTALE20, and SlTALE21) were highly expressed in the GS fruit shoulder region (Gs), and four SlTALE genes (SlTALE6, SlTALE9, SlTALE19, and SlTALE24) were highly expressed in the parts with less-developed chloroplasts, such as Gb, Ws, and Wb ( Figure 6). These SlTALE genes, which show a latitudinal expression gradient, might regulate the establishment of a chloroplast developmental gradient between the shoulder and base of tomato fruit. SlTALE2 and SlTALE5 (i.e., SlTKN4 and SlTKN2) have been identified to play a role in regulating chloroplast development and green shoulder formation in tomato fruit [22]. However, SlTALE22, known as SlBEL11, was not differentially expressed between the shoulder and base of tomato fruit, and it has been reported to be involved in the regulation of chloroplast development and chlorophyll accumulation in tomato fruit [27].
Complex interactions between these SlTALEs were detected based on protein-protein interaction networks. SlTALE5 and SlTALE24 could interact with SlTALE6, SlTALE9, SlTALE11, and SlTALE19, and SlTALE5 could interact with SlTALE20 and SlTALE21. This suggested a possible interaction among these SlTALE genes in tomato fruit development, which is consistent with the frequent interactions observed between BELL and KNOX members. The predicted interaction networks provide new insights into how the regulation of TALEs affects the growth and development of tomato fruit, including chloroplast development, chlorophyll accumulation, and the formation of green fruit shoulder.
Immature green fruits are capable of photosynthesis, and up to 20% of the total fruit carbohydrate comes from the photosynthesis of fruit chloroplasts in tomato [46]. During fruit maturation, fruit chloroplasts develop into chromoplasts to synthesize and accumulate nutrients and flavor compounds [47]. It is an effective way to improve fruit quality by increasing the abundance and function of chloroplasts. In this study, GS tomato fruits with dark green shoulders had significantly enhanced chloroplast development and chlorophyll accumulation, resulting in a significant increase in the level of photosynthesis-related parameters (chlorophyll fluorescence parameters) ( Figure 5). As mentioned previously, four SlTALE genes had been reported to affect the development of tomato chloroplasts and fruit quality [17,27,28]. As positive regulators, SlTKN2 and SlTKN4 are promising candidates to target for improving fruit phytonutrients. Negative regulators, such as SlBEL4 and SlBEL11, can also be the ideal targets to be precisely edited by CRISPR/Cas9 to create fruits with improved phytonutrients. Notably, the TALE family is a pivotal transcription factor and involves many processes of plant growth and development, including some core processes such as floral meristem formation, organ morphogenesis, and fruit development [8,48]. When SlTALEs are applied to tomato breeding programs, they may have negative pleiotropic effects that affect other aspects of plant growth. That is a strategy to specifically target SlTALEs in fruits by using fruit-specific promoters or RNAi technology to avoid the negative or pleiotropic effects on plant growth. This study suggested several candidate genes for further analysis to improve fruit quality by regulating tomato fruit chloroplast development.

Plant Materials
The tomato (Solanum lycopersicum L.) inbred line TTD302A and mutant line GS were used in this study. The mutant line GS (in which immature fruit possess a green shoulder) derived from ethyl methyl sulfonate (EMS) mutagenesis of inbred line TTD302A (in which immature fruit lack a green shoulder). All plant materials were grown in a glasshouse under a 16-h light/8-h dark photoperiod, with 60-75% relative humidity and ambient temperature of 20 • C (night) to 30 • C(day), at Northwest A&F University, Yangling, Shaanxi. Flowers were tagged at anthesis, and fruit development was recorded as days post-anthesis (dpa). The different stages of fruit development included mature green (MG) at 42 dpa, which was characterized by bright green coloration without obvious color change; the first sign of color change was observed at the breaker (Br) stage at approximately 45 dpa; the Br+5 stage occurred after 5 days of the Br stage and/or the pink ripe (P) stage. All plant samples were collected at the same time each day. The samples were directly frozen in liquid nitrogen and stored at −80 • C for further use.

Identification of SlTALE Family Members
To identify the TALE family members of tomato, previously reported Arabidopsis and tomato TALE proteins sequences were used as queries to search against the tomato protein database with the BLASTP program (https://blast.ncbi.nlm.nih.gov, accessed on 3 November 2021) using an E-value below 10 −3 and an identity of 50% as the threshold. In addition, the Hidden Markov Model constructed by the HMMER 3.0 program was used to search the tomato protein sequences to identify TALE family members. Redundant sequences obtained from the HMMER and BLAST results were removed following the alignment of the predicted SlTALE members. The predicted SlTALE members were then aligned for comparison and removal of redundancy. The aligned sequences were used as candidate SlTALE family sequences. All candidate sequences were confirmed to be SlTALE proteins using PfamScan (https://www.ebi.ac.uk/Tools/pfa/pfamscan/, accessed on 15 November 2021) and SMART (http://smart.embl-heidelberg.de/, accessed on 15 November 2021).

Gene Structure, Conserved Motif, and Phylogenetic Analysis
Conserved motifs of SlTALE family proteins were identified by MEME software (http://meme-suite.org/, accessed on 7 December 2021), with motif width of 6-50 and maximum motif number of 12. The gene structures and chromosomal locations of SlTALE family genes were characterized by comparing the coding and genomic sequences with TBtools [50]. Gene duplication information was obtained by aligning all SlTALE family genes.
Multiple sequence alignments of candidate TALE family proteins from Solanum lycopersicum, Solanum tuberosum, Arabidopsis thaliana, Hordeum vulgare, Nicotiana benthamiana, Oryza rufipogon, and Vitis vinifera were performed using MAFFT. The phylogenetic tree was constructed with MEGA6 software using the neighbor-joining (NJ) method and 1000 bootstrap replicates, and the output was visualized using the online software tool EvolView (http://www.evolgenius.info/, accessed on 7 December 2021) and FigTree (v1.4.4) software. Finally, a figure displaying the phylogenetic tree, conserved motifs, and gene structure was built using TBtools.

Cis-elements, TF Binding Sites, and Protein Interaction Network
To analyze the cis-elements and TF binding sites of the promoters, the 2000-bp region upstream from the start codon was obtained from the tomato genome sequence by a Perl script (Perl S1) and TBtools. Cis-elements were identified by PlantCARE (http: //bioinformatics.psb.ugent.be/webtools/plantcare/html/, accessed on 29 November 2021). TF binding sites were searched by Plant Transcriptional Regulatory Map (http:// plantregmap.gao-lab.org/binding_site_prediction.php, accessed on 5 December 2021) [51,52]. The protein-protein interaction network of the SlTALE family was analyzed by String (https://string-db.org/, accessed on 7 December 2021) [53].

Chlorophyll Content, Chlorophyll Fluorescence Parameters, and Chloroplast Structure
The chlorophyll content was determined spectrophotometrically following a previously described method [54]. Fruit samples were collected at 28 dpa, and the chlorophyll content was measured using the stem end (fruit shoulder area) and base (fruit navel area) parts of the fruits.
After the fruit at 28 dpa was conditioned in the dark for 30 min, a Multispectral Fluorescence Imaging System (Fc800 FluorCam, Photon Systems Instruments, Drásov, Czech Republic) was used for chlorophyll fluorescence dynamic imaging and measurements of chlorophyll fluorescence parameters.
The pericarp was isolated and fixed in 0.1 M cacodylate buffer consisting of 4% glutaraldehyde at 4 • C for 24 h, postfixed in 1% osmium tetroxide, and dehydrated in a gradient ethanol series. After embedding the sample in Spurr resin, the ultrathin sections of samples were obtained using the Leica UC7 ultramicrotome (Leica, Solms, Germany) on 200 mesh copper grids, and then stained with uranyl acetate and lead citrate. The Hitachi HT7700 transmission electron microscope (Hitachi Ltd., Tokyo, Japan) was used to observe the ultrastructures of plastids.

Transcriptome Analysis and Expression Analysis of SlTALE Family Genes
The pericarp of the shoulder region from WT and GS tomato fruits was taken for RNAseq analysis at five critical periods (Table S4), including 7, 14, 21, and 28 dpa as well as at P (Pink ripening period, or Br+5). Three biological replicates were collected from each tissue type and three technical replicates within each biological replicate. Library construction was performed using the Illumina TruseqTM RNA Sample Prep Kit, and RNA-seq was performed on the Illumina NovaSeq 6000 sequencing platform (Illumina, Shanghai, China) by Shanghai Majorbio Bio-pharm Technology Co., Ltd The data were analyzed using the Majorbio Cloud Platform (www.majorbio.com, accessed on 18 September 2021). In addition, RNA-seq data (Table S5) for "Ailsa Craig" tomato (which has a dark green shoulder) were downloaded from the TOMExpress website (http://tomexpress.toulouse.inra.fr, accessed on 28 September 2021) at different developmental stages (10 dpa, 20 dpa, MG, Br, and Br+5) between the shoulder and the base of the fruit, to investigate the expression profile of SlTALE family genes in different tomato materials.
A total of 16 differentially expressed or highly expressed genes were selected for relative expression analysis to validate the RNA-seq results. For quantitative RT-PCR, total RNA was extracted using an RNA Extraction Kit (Omega Bio-Tek, Shanghai, China), followed by an RNA integrity assay with 1% agarose gels; the quality and concentration of RNA were assessed using a Nanodrop 2000 spectrophotometer (Thermo Scientific, Waltham, MA, USA). First-strand cDNA was reverse-transcribed using an Evo M-MLV Mix Kit with gDNA Clean for qPCR (Accurate Biotechnology, Changsha, China). qRT-PCR was performed using gene-specific primers with the SYBR ® Green Premix Pro Taq HS qPCR Kit (Accurate Biotechnology, Changsha, China) in a QuantStudio ® 5 Real-Time PCR System (Life Technologies, Carlsbad, CA, USA). Three biological replicates were collected from each tissue type and three technical replicates within each biological replicate. The SlActin (Solyc01g104770) was used as the reference gene. The relative expression levels of the 16 SlTALE genes were calculated according to the 2 −∆∆Ct method. Primers used for qRT-PCR are shown in Table S7.

Conclusions
In this study, a total of 24 SlTALE genes were identified. The physicochemical characteristics, subcellular localization, phylogenetic analysis, gene structure, motifs, promoter activity, and expression profiles were investigated with diverse bioinformatics methods. Phylogenetic analysis showed that 24 SlTALE members could be classified into 5 BELL subfamilies and 4 KNOX subfamilies, and the members of the same subfamily had similar gene structures and motifs. The cis-elements responsive to light, phytohormone, developmental regulation, and environmental stress were enriched in the promoter of SlTALE genes, and the light response elements were the most abundant. An abundance of TF binding sites was also enriched in the promoter of SlTALE genes. An analysis of gene expression profiles indicated that six SlTALE genes (SlTALE2, SlTALE3, SlTALE5, SlTALE11, SlTALE20, and SlTALE21) were highly expressed in the GS fruit shoulder region (Gs), and four SlTALE genes (SlTALE6, SlTALE9, SlTALE19, and SlTALE24) were highly expressed in the parts with less-developed chloroplasts, such as Gb, Ws, and Wb. The protein-protein interaction networks predicted that a complex regulatory network controls the expression of these SlTALE genes during chloroplast development and green fruit shoulder formation. Overall, our findings provide new insight into SlTALE genes in tomato and their role in fruit chloroplast development, and suggested candidate genes for future research aimed at improving fruit quality by regulating tomato fruit chloroplast development.