A comprehensive analysis to speculate the role of trithorax genes in photoperiod-mediated tuberization in potato and comparison with storage root crops

Abstract

Background

Trithorax genes are crucial epigenetic regulators, driving gene activation through histone tail modifications and chromatin remodeling. These genes are fundamental in multifaceted aspects of plant development. However, our knowledge of trithorax genes is predominantly focused on model plant Arabidopsis, particularly in flowering. Considering the similarities between flowering and potato tuberization, we hypothesise the role of trithorax genes in stolon-to-tuber transition. To explore their origin, diversification, and molecular conservation, we conducted a comprehensive bioinformatic analysis of trithorax proteins in selective plant lineages with an emphasis on potato. Further, selected trithorax members were analyzed in five storage root crops to assess their potential role in storage organ development.

Results

We identified 268 trithorax family members across 9 species, including 2 algae, 2 bryophytes, and 5 angiosperms, categorised into three groups: SET domain proteins, chromatin remodelers, and associated proteins based on functional domains and motifs. These proteins were analyzed in potato and subsequently compared with other storage root crops. It was observed that the distribution of trithorax genes is universal among photosynthetic organisms from unicellular algae to highly evolved storage crops like potato, with the whole genome, segmental, and tandem duplications driving their expansion. The trithorax proteins cluster in unique monophyletic groups with conserved functional domains across species. The predicted physico-chemical properties, cis-regulatory elements (CREs), and transcription factors (TFs) reflected the involvement of trithorax genes in different developmental events, including tuberization in potato, stress-responsiveness, and flowering. Moreover, expression analysis confirmed the role of the identified genes in potato tuberization under photoperiod-inductive conditions. The study revealed the clustering of trithorax genes in potato into four groups of 7 slightly upregulated, 9 downregulated, and 8 highly upregulated genes during the stolon-to-swollen stolon transition stage. The functional domains and the abundance of CRE and TFs were found consistent in storage root crops as that of potato.

Conclusions

We show the existence of trithorax genes in photosynthetic organisms and demonstrate their potential role in photoperiod-mediated potato development. The gene regulatory mechanism appears to be conserved amongst potato and other storage root crops. This study may act as a prelude for further biotechnological interventions in potato and other storage crops.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-025-07548-w.

Background

Trithorax group proteins are crucial modulators of gene expression and are mainly responsible for maintaining genes in “on” state [1]. Their involvement in gene activation is intricate and encompasses multiple processes, including histone tail modifications, chromatin remodeling, and interactions with the transcriptional machinery. These proteins ensure that genes remain in an active state by modifying histones to promote a more accessible chromatin structure, facilitating the binding of TFs and RNA polymerase [2]. The significance of these proteins is emphasized by the prominent developmental phenotypes observed when their function is disrupted, highlighting their essential role in proper gene regulation during development [1].

Trithorax proteins, initially identified in Drosophila melanogaster, are pivotal transcriptional activators that maintain the expression of homeotic genes, ensuring normal body structure development in the appropriate locations [3]. The trithorax gene family has been well investigated in the model plant Arabidopsis thaliana, but research on other plant species is limited. These genes are significant for a range of biological processes, including development [4], stress responses [5], regeneration [6], immunity [7], and circadian regulation [8]. Most research focuses on the development of different plant organs, such as flowers [9], leaves [10], roots [11], stamens [12], and gynoecium [13], as well as the regulation of organ size [14], and developmental roles such as phase transitions [15].

Trithorax group proteins ARABIDOPSIS TRITHORAX1 (ATX1), SPLAYED (SYD), and BRAHMA (BRM) activate the floral homeotic genes APETALA3 (AP3) and AGAMOUS (AG) in A. thaliana to control floral organ identity [4, 16]. ATX1 additionally activates PISTILLATA (PI), APETALA1 (AP1), and AP2 [4]. In addition to influencing gynoecium patterning through its interaction with KANADI (KAN), ULTRAPETALA1 (ULT1) also initiates AG expression, which ends meristem activity in the centre of the flower [9, 13]. Premature flowering is inhibited by the interaction of ATX1, ARABIDOPSIS TRITHORAX RELATED3 (ATXR3), and ATXR7/SDG25 with FLOWERING LOCUS C (FLC), respectively, in the vegetative stage [15, 17, 18]. Furthermore, ATX1 and PICKLE (PKL) function to prevent FLOWERING LOCUS T (FT) suppression by the polycomb group, which stimulates flowering [19]. Early Heading Date 3 (Ehd3) and ATX1 work together in Oryza sativa to control flowering time [20]. OsWDR5a forms a COMPASS-like complex with OsTrx1/SET domain group protein 723 (SDG723) to promote flowering [21]. In woodland strawberries, ULT1 also regulates the development of flowers [22]. Further investigation of the relevant genes in other unexplored species is necessary to strengthen our understanding of the multifaceted role played by these genes in flowering and tuberization.

Two plant reproductive strategies, flowering and tuberization, share several endogenous factors and external cues [23]. Potato, the fourth most significant crop after cereals, is vital for relieving food insecurity and serves as a model for studying the molecular mechanisms of tuber formation. The tuberization process begins with the below-ground modified stem, known as a stolon, which undergoes various developmental stages to form a mature edible tuber. Studies on transgenic lines of key tuberization genes have revealed differential regulation of trithorax members. Trithorax genes such as StASH1-RELATED 2 (StASHR2) are upregulated, while StULT, StATXR2, and StATXR3 are downregulated in stolons of StBEL5 transgenic lines [24]. We have previously shown that StATXR3 and StASH1-RELATED3 (StASHR3) expression are elevated in the axillary meristems, while StATX4 is downregulated in stolons of transgenic lines overexpressing StMSI and StPOTH15, respectively [25, 26]. Moreover, several trithorax members, including StULT, StASH1-RELATED PROTEIN1 (StASHH1), StASHR2, and StASHR3, were targets of StE(Z)2 in short-day induced stolons [27].

A recent study on SET domain proteins in potato has largely focused on heat and drought stress, but has not explored their role in tuber development and was restricted to only a few trithorax group members [28]. Trithorax proteins, a subset of SET domain proteins, not only include SET domain proteins but also feature chromatin remodelers and associated proteins. Based on our previous findings, we believe that it is critical to comprehend the function of trithorax genes in tuber growth, considering the significance of potato as a major food crop and the crucial role of trithorax genes in plant development.

In this regard, our study aims to broaden the understanding of trithorax proteins by examining their presence across various plant species (algae, bryophytes, and angiosperms), evolutionary relationships, structural features, and potential functions in storage crops, including potato and five root crops. We conducted a comprehensive bioinformatics analysis of trithorax proteins and identified 268 members in 9 different species, and further analyzed these through phylogenetic analysis, domain and motif identification, conserved region analysis, sub-cellular localization, and determination of protein physico-chemical properties. To further explore the role of trithorax members in photoperiod-mediated tuberization in potato, we have analyzed the gene duplication, regulation, and expression profile in the stolon-to-tuber transition stages. Considering the nutritive values of storage root crops, the gene regulation of selective trithorax members was compared with five storage root crops. This study involving a broad description of the trithorax gene family across all the echelons of green plants provides important insights into their conservation and diversification. We speculate that the expression profile and molecular mechanisms of trithorax genes identified in potato may reflect similar regulatory functions in the development of other storage root crops. Our study would provide the foundation for further experimental research into the role of the trithorax gene family in tuber and storage organ development.

Results

Trithorax proteins are distributed across all lineages of the plant kingdom

Thirty-two trithorax proteins from the model plant A. thaliana were selected based on the existing literature. Utilizing BLASTp against a range of plant databases, 268 trithorax proteins were identified and further confirmed using the NCBI CDD Pfam database. It was observed that these proteins are spread across all lineages of photosynthetic organisms, from unicellular algae to highly evolved angiosperms. Out of the 268 trithorax proteins, the number of these proteins in 2 algal (Chlamydomonas reinhardtii and Volvox carteri), 2 bryophyte (Physcomitrium patens and Marchantia polymorpha), and 5 angiosperm (A. thaliana, S. tuberosum, Solanum lycopersicum, O. sativa, and Zea mays) species is 44, 61, and 163, respectively. The increase in the number of proteins per species from algae (22 per species) to angiosperms (32.6 per species) depicts their diversification in flowering plants. This also suggests an increase in the complexity of trithorax proteins along the evolution in different lineages (Table S1). These results illustrate the wide distribution of trithorax proteins, indicating their potential functional diversity across various plant taxa.

Trithorax proteins cluster in unique monophyletic groups

To investigate the evolutionary relationship among trithorax proteins, multiple sequence alignment was conducted using 268 trithorax members, followed by the generation of the phylogenetic tree. The combined tree demonstrated the origin of trithorax proteins in algae and further diversification in angiosperms (Fig. S1). In addition, these proteins were found to form distinct monophyletic clusters, reflecting their separate evolutionary origin and divergence. To precisely explore the evolutionary relationship of these clusters and based on the topology of the tree and the depicted patterns, we segregated these proteins into three groups of SET domain proteins, chromatin remodelers, and other associated proteins. The structure of all three phylogenetic trees demonstrates the division of these proteins into three main groups (Group I, II, and III), with SET domain proteins, chromatin remodelers, and associated proteins organized into nine, eight, and five clades, respectively (Fig. 1).

Fig. 1.

Phylogenetic analysis of trithorax proteins from Solanum tuberosum and different plant species. A SET domain proteins B Chromatin remodelers C Associated proteins belonging to A. thaliana, S. tuberosum, Solanum lycopersicum, O. sativa, Zea mays, Physcomitrium patens, Marchantia polymorpha, Chlamydomonas reinhardtii, and Volvox carteri. SET domain proteins, Chromatin remodelers, and Associated proteins are divided into 3 groups each, with 9, 8, and 5 clades, respectively. The phylogenetic trees were constructed using MEGA11 with the Neighbor-Joining (NJ) method and 1000 bootstrap replicates. The outer circle of each tree represents the groups, while the inner circle highlights the clades within these groups. The label above the node represents node ages based on 1000 bootstrap

Analysing the phylogeny of SET domain-containing proteins, it was observed that these proteins form nine different clades clustering into three groups based on their evolutionary relationships. Group I contained 4 clades (clade A to D), with clade A comprising ASHH2 and ASHR1 of algae rooted at the base of the group, suggesting these proteins as the earliest forms of trithorax SET domain proteins in plants. While clade B contained ATXR2/5/6, clade C was formed of ASHR3 and ATXR7 of bryophytes and angiosperms. The ATX proteins of the different species, however, were found to cluster into a single monophyletic clade D, highlighting their common origin and subsequent diversification (Fig. 1A). Further, while group II comprises three clades, it represents the most recent members among SET domain-containing trithorax proteins. Except for the clade F, which majorly has ATXR1 proteins from algae, the clades E and G comprise ATXR1, ASHR1/2 proteins from algae, bryophytes, and angiosperms together (Data_Sheet_1). Within group III, clade H consists of ATXR3/4 members, while Clade I encompass ASHH proteins from all groups of plants studied. Overall, the topology of the tree depicts group I proteins as the ancient type and group II as the most recent ones (Fig. 1A).

Among chromatin remodelers, group I is rooted at the base of the tree representing the most ancient origin. Groups II and III appear to have emerged more recently, likely to have diverged at a similar time (Fig. 1B). Within these groups, chromatin remodelers are categorized into distinct clades based on their functional roles. Specifically, PKL is placed in clade A, SWITCH2 (SWI2) in clade B, SYD in clade C, BRM in clade D, TRAUCO (TRO) in clade F, and SWITCH1 (SWI1) in clade H, reflecting their functional similarities and evolutionary relationships. Moreover, BUSHY (BSH) proteins are distributed across two distinct clades: clade E, which encompasses BSH proteins from bryophytes and angiosperms, and clade G, which includes BSH proteins from algae (Fig. 1B).

Among the associated proteins, group I lies at the base of the phylogenetic tree, followed by groups II and III. Within this classification, WD-40 REPEAT PROTEIN5 (WDR5), RbBP5 LIKE (RbBP5-L), and REBELOTE (RBL) are located in clades A, B, and D, respectively, while ULT is present in both clades C and E (Fig. 1C).

Phylogenetic analysis revealed that each trithorax protein exhibits a distinct monophyletic origin, suggesting independent evolutionary diversification. Specifically, trithorax proteins from algae appear to have evolved earlier and independently, compared to those from bryophytes and angiosperms, highlighting their distinct evolutionary trajectories. Furthermore, trithorax proteins from monocots and dicots form separate clusters, reflecting their divergent evolutionary paths between these major plant lineages. Moreover, the trithorax proteins from potato, cluster closely with their counterparts in the model plant A. thaliana, suggesting that these proteins might have similar roles or conserved functions across these two plant species. The individual phylogenetic trees of algae, bryophytes, and angiosperms demonstrate consistent clustering patterns of trithorax proteins, as depicted in Fig. S2.

Functional domains and motifs of trithorax proteins are conserved across plant species

To elucidate the functional roles of trithorax proteins, we performed domain and motif analysis. This analysis identified three significant functional domains: the SET domain, domains associated with chromatin remodeling, and the WD domain or Noc2 domain. In terms of domain distribution, SET domain proteins, chromatin remodelers, and associated proteins contained 44, 48, and 11 domains, respectively, underscoring their diverse functional roles (Fig. S3) (Data_Sheet_2).

Trithorax SET domain proteins display significant structural diversity with domains such as SET, PHD, zinc finger (zf) types, FYRC/N, PWWP, and TPR (Fig. S3A). The SET domain’s position varies across clades: it is N-terminal in clade A, C-terminal in clades B, C, D, and central in clades E, F, G. Clade H features ATXR4 with a C-terminal SET domain and ATXR3 with a central SET domain. Clade A also includes zf-CW and zf-MYND domains for protein interactions, while clade B proteins like ATXR5/6 have the PHD domain for binding methylated H3, some members also containing the zf-MYND domain. Motif analysis shows SET motifs across all clades, PHD motifs in clades B, C, D, and the FYRN motif in clade I, respectively (Data_Sheet_3). Additionally, ATXR2 (Chlamydomonas reinhardtii, clade B), ASHR3 (Marchantia polymorpha, clade C), and ATXR3 (Solanum lycopersicum, clade H) each contain two SET domains and distinct PHD domains, while some of the clade D members also have two PHD domains (Fig. S3A).

Nearly all clades of chromatin remodelers feature domains essential for chromatin remodeling, such as SNF2-rel, DEAD box, CHDII_SANT-like, SNF5, and SPRY domains (Fig. S3B). The SNF2-rel domain, crucial for chromatin unwinding, is present in clades A, B, C, and D. Clade A also includes the CHDII-SANT-like domain, which is involved in chromatin remodeling, but is not present in other groups. Clades B, C, and D feature the Helicase_C domain, which separates double-stranded DNA. The Bromodomain, recognizing acetylated lysines on histone tails, is found in the BRM protein of clade D, while clade C and E contain the SnAC and the SNF5, and SPRY domains, respectively, the latter regulating histone H3 methylation. Motif analysis shows that clades A-D have various SNF2-rel motifs, with clade A additionally featuring P-loop containing nucleotide triphosphate hydrolase, PHD, and CHDII-SANT motifs. Clade B has CHDII-SANT motifs, clade D includes a P-loop, the SPRY motif is present in clade F, and clade H contains the DYAD protein motif, highlighting the distinct and specialized nature of chromatin remodeling domains and motifs across different clades (Fig. S3B).

Associated proteins primarily feature the Noc2 domain and various WD40 domains. Clades A and B are characterized by multiple types of WD40 domains, while clade D is notable for the presence of the Noc2 domain. Though clades A, B, C, and E exhibit diverse WD40 motifs, while clade F includes the ultrapetala 1, and clade D contains the noc2p motif (Fig. S3C). Though this study identifies different domains and motifs characteristic to each group, wet lab experiments would be required to further validate the functional promiscuity of each of these domains.

To identify conserved functional regions, we conducted multiple sequence alignments and predicted the structures of the corresponding A. thaliana proteins using AlphaFold, achieving high-confidence predictions (Fig. S4). The conserved regions in each model of the specific clades are highlighted in green against the white theme (Fig. 2, S5). For SET domain proteins, conserved regions were primarily found in the SET domain across clades B-E and G-I, with clade G also showing conservation in the zf-MYND domain. Clade A exhibited lower sequence conservation due to outliers, and clade G lacked a representative A. thaliana member (Fig. 2A). In chromatin remodelers, conserved regions were identified within domains such as SNF2-rel, CHDII-SANT, and SNF5 in clades A, B, D, and E. Clade F had conservation in the SPRY domain, while clade H did not have a specific domain associated with its conserved region (Fig. 2B). For associated proteins, conserved regions were found in the WD40 domain for clades A and B, and the Noc2 domain for clade C. Notably, nearly all proteins in clade E showed high conservation, indicating least divergence within this clade (Fig. 2C).

Fig. 2.

Conserved protein region of respective clades depicted in 3D topology. The topology represents conserved regions highlighted in green, modelled on the A. thaliana member in each clade (A) SET domain proteins (B) Chromatin remodelers (C) Associated proteins. The structures of these representative proteins were predicted using the AlphaFold Protein Structure Database

Predicted subcellular localization, physico-chemical properties, and gene structure of trithorax proteins reflect their evolutionary adaptations

The sub-cellular localization of proteins is essential for their functional roles, and our analysis of SET domain proteins reveals that the majority of proteins in clades A, B, C, D, I, and clades E, G are predicted to be nuclear and cytoplasmic respectively, except ATXR1 proteins in clade F from V. carteri that localizes to the plastid. Clade H shows variation: most ATXR4 proteins are mitochondrial, whereas ATXR3 is nuclear. Generally, chromatin remodelers and associated proteins are nuclear with nuclear localization signals. Further, one ULT protein from S. lycopersicum and one from Z. mays are cytoplasmic, reflecting their placement in distinct clades (Data_Sheet_4).

The physico-chemical properties of proteins exhibit considerable variability, influenced by protein type and species origin. The amino acid counts vary dramatically from 144 amino acids (ULT2 of Z. mays) to 3,815 amino acids (SYD of Z. mays) (Data_Sheet_5). The theoretical isoelectric points (pI) of these proteins exhibit substantial variability from 4.46 (ASHR2 of S. tuberosum) to 9.71 (BSH of S. tuberosum). Stability indices provide insights into protein robustness, with the majority of trithorax proteins showing indices above 40, indicating their instability at room temperature. Interestingly, WDR5 proteins have instability indices below 40, highlighting their stability. Aliphatic indices, which reflect thermostability, vary widely among proteins, ranging from 45.72 (ULT2 of S. lycopersicum) to 126.44 (ULT2 of S. lycopersicum and BSH of V. carteri). Hydropathy profiles reveal that most of the trithorax proteins are hydrophilic, except ULT2 from S. lycopersicum and ATX3 from C. reinhardtii (Data_Sheet_5). Thus, each protein in this diverse set offers unique insights, shaped by its molecular structure, stability, and hydrophilicity, revealing the complex interplay between structure and function across different biological contexts.

Understanding exon-intron structures is crucial for exploring the evolution of gene families. Our analysis of the trithorax gene family reveals significant variations in exon-intron patterns, including differences in intron length and exon count (Fig. S6). Moreover, gene lengths vary markedly among species, with S. tuberosum displaying the longest intron-rich genes and V. carteri having the shortest. Remarkably, A. thaliana and S. tuberosum each have one intron-less gene (Fig. S6A-B), P. patens has five (Fig. S6C), and V. carteri has two (Fig. S6D). These structural variations offer valuable insights into the evolutionary dynamics and functional adaptations of the trithorax gene family across diverse plant species.

Gene duplication events promoted the expansion of the trithorax gene family in potato

Given the importance of potato as a major food crop, understanding the role of trithorax genes in tuber development is essential. Our comprehensive analysis of the potato genome identified 31 trithorax genes distributed across 9 of the 12 chromosomes, with 22 genes present in duplicated pairs (Fig. 3). The average number of trithorax genes present is 2.58 genes per chromosome, with the maximum number of trithorax genes on chromosomes 7 and 11, while none on chromosomes 4, 8, and 10. The gene arrangement on the chromosomes appears irregular and does not correlate with chromosomal length, nucleotide composition, GC skew, or overall gene density (Fig. 3). This irregular distribution suggests that the trithorax gene family expansion is due to both whole-genome, segmental, and tandem duplications. The higher number of duplicated genes is also evident from one recent and one ancient whole genome duplication in potato [29].

Fig. 3.

Circos plot of trithorax genes in S. tuberosum. The plot displays the distribution, chromosomal localization, gene collinearity, and genomic context of trithorax genes in S. tuberosum. The duplicated trithorax genes are highlighted in red, while the position of each gene ID represents its chromosomal localization. The inner circle represents the chromosome skeleton, followed by the GC ratio, GC skew, and the outer circle represents gene density in bar plots. Gene IDs used in this plot are as per the SpudDB database

To explore the evolutionary dynamics of these duplicated genes, we calculated the non-synonymous (Ka) to synonymous (Ks) substitution ratios. Most Ka/Ks ratios are below 1 (Table 1), indicating strong purifying selection pressure on trithorax genes in S. tuberosum. Divergence times for intraspecific duplicated trithorax genes were estimated to range from 3.03 to 93.95 million years ago (MYA). The gene pair Soltu.DM.11G010650.1/Soltu.DM.11G010620.1 (StBSH_1/StBSH_2) represents a recent duplication event, as evidenced by their proximity in the phylogenetic tree and the similarity in the domains and motifs of their proteins. In contrast, Soltu.DM.01G033850.1/Soltu.DM.07G019980.1 (StBRM/StSWI2) is identified as the most ancient duplicated gene pair, also evident from their placement in separate monophyletic clades and the distinct differences in their protein domains and motifs.

Table 1.

Divergence rate and time of duplicated trithorax genes in S. tuberosum

Gene pair	Ka	Ks	Ka/Ks	T(MYA)	Selection Pressure
Soltu.DM.01G035010.1/Soltu.DM.03G022600.2	0.38	1.62	0.23	47.08	Purifying
Soltu.DM.02G027040.1/Soltu.DM.09G019870.2	0.38	NA	NA	NA	Purifying
Soltu.DM.06G017500.1/Soltu.DM.11G010640.2	1.13	NA	NA	NA	Purifying
Soltu.DM.11G010650.1/Soltu.DM.11G010620.1	0.09	0.10	0.92	3.03	Purifying
Soltu.DM.05G012810.1/Soltu.DM.02G021490.1	0.65	2.53	0.26	73.65	Purifying
Soltu.DM.05G005310.1/Soltu.DM.03G010460.1	0.54	2.49	0.22	72.52	Purifying
Soltu.DM.12G000190.1/Soltu.DM.01G001770.2	0.15	0.47	0.31	13.68	Purifying
Soltu.DM.07G018390.1/Soltu.DM.03G010440.1	0.93	1.86	0.50	54.16	Purifying
Soltu.DM.07G001180.1/Soltu.DM.03G014760.1	0.96	NA	NA	NA	Purifying
Soltu.DM.01G033850.1/Soltu.DM.07G019980.1	0.86	3.23	0.27	93.95	Purifying
Soltu.DM.11G008270.1/Soltu.DM.03G022130.1	0.31	2.82	0.11	82.06	Purifying

T(MYA) = Ks/2r (where r = 1.72 × 10⁻⁸)

Na = High Sequence Divergence Value (pS > = 0.75)

Cis-regulatory elements predict the major role of potato trithorax genes in stress and hormonal responses

To explore the regulation of trithorax genes in potato, we identified 23 different CREs within the 2 kb promoter region of these genes and categorized them into three main groups: hormone-related, stress-related, growth and development-related elements (Data_Sheet_6). The overall distribution and abundance of CRE are given in Fig. 4.

Fig. 4.

The distribution of Cis-regulatory elements (CREs) in S. tuberosum trithorax promoter sequences. The distribution of CREs in 2 kb promoter sequences of trithorax genes was analyzed using the PlantCARE Query Search online tool. A total of 31 CREs were categorized into three groups of hormone-related, stress-related, and growth & development-related. The grey circles indicate the absence of CRE, while the coloured circles indicate its presence. The bar plot shows the abundance of each CRE across all promoter sequences, while the graph on the right illustrates the abundance of CREs within individual trithorax genes. The blue represents hormone-related, red represents stress-related, and green represents the growth & development-related CREs

Among three hormone-related CREs, Ethylene Response Elements (EREs) and Abscisic Acid Response Elements (ABREs) are the most prevalent, while Auxin Response Elements (AuxRR) are the least. In eight stress-responsive CREs, MYB and MYC elements are the most abundant, with Wound Responsive Element 3 (WRE3) being the least. For growth and development, LAMP and as-1 elements are the most frequent, whereas RY-elements and APETALA1 (AP1) were less abundant. Overall, this distribution indicates that these gene promoters are abundant in abscisic acid, ethylene, drought stress, and light-responsive CREs. Among SET domain proteins, Soltu.DM.03G022130.1 (StATXR6) features the highest number of CRE types, whereas Soltu.DM.07G004030.1 (StASHH1) has the fewest. Among chromatin remodelers, Soltu.DM.11G018160.4 (StSYD) possesses the most diverse CRE types, while Soltu.DM.07G019980.1 (StSWI2) has the least. Similarly, among associated proteins, Soltu.DM.02G021490.1 (StREBELOTE_2) shows the highest variety of CRE types, in contrast to Soltu.DM.07G019360.2 (StULT). The specific location of each CRE is given in Fig. S7. This analysis predicts that stress-responsive CREs are the most abundant on the promoter of the trithorax genes, followed by hormone-related. Therefore, it suggests the role and induction of these genes under different hormonal and stress environments besides in growth and development.

Transcription factors governing reproductive development are predicted to regulate trithorax genes in potato

TFs play a crucial role in gene regulation by binding to CREs. In our study, we identified TFs that interact with the promoters of trithorax genes in S. tuberosum using the A. thaliana database as a reference. Given the overlap in regulatory mechanisms between flowering and tuberization, we selected twenty-four TFs associated with the flowering pathway for detailed analysis. Our investigation revealed that a diverse array of TFs binds to the trithorax gene promoters (Fig. 5). These TFs (belonging to the homeobox genes, and Dof gene family) encompass those involved in regulating flowering time, floral organ development, light-mediated flowering responses, floral meristem identity, and photomorphogenesis processes (Data_Sheet_7).

Fig. 5.

Transcription Factors (TFs) binding to trithorax gene promoters in S. tuberosum. TFs binding to the promoter regions of trithorax genes were predicted using PlantPAN4.0, and the results were visualized using an UpSet plot. Circles connected by blue lines represent the presence and interrelation among TFs binding at the respective promoters of the trithorax genes. The graph on the left shows the conditional occurrence of TFs across the promoter regions of trithorax genes, and the graph on the right depicts the total occurrence of each TF. The bar plot at the top illustrates the abundance of different TFs on individual trithorax genes. The plot was created in the R studio environment using the package “ComplexUpset”

TFs which govern floral organ development, such as BIGPETAL (BPE), HECATEs (HECs), HOMEODOMAIN-LEUCINE ZIPPER PROTEIN 1 (HAT1), KANADI (KAN), LEAFY COTYLEDON2 (LEC2), SHOOTMERISTEMLESS (STM), SHI RELATED SEQUENCE (SRS), PERIANTHIA (PAN), SEPALLATA (SEP), RGA-LIKE2 (RGL2), AUXIN RESPONSE TRANSCRIPTION FACTOR 3 (ETT), and SQUAMOSA PROMOTER BINDING PROTEIN-LIKE (SPL) were found to interact with trithorax promoters.

MicroRNAs 156 and 172 play a significant role in the vegetative-to-reproductive phase transition in plants by regulating the expression of SPL and TARGET OF EAT2 (TOE2) [30, 31]. Trithorax genes are predicted as targets of these TFs. Other TFs such as CRY2-interacting bHLH 1 (CIB), SPL, AGAMOUS-LIKE (AGL), TEMPRANILLO 1 (TEM1), SUPPRESSOR OF OVEREXPRESSION OF CONSTANS1 (SOC1), SUPPRESSOR OF FRIGIDA 4 (SUF4), and FLC, which are critical for controlling flowering time, were also found to bind to these gene promoters. Additionally, TFs like CIB, CYCLING DOF FACTOR (CDF), TEM1, TOE2, and SOC1, which are pivotal in regulating key floral pathway genes such as CONSTANS (CO) and FT, were predicted to interact with gene promoters. The SUF4 and SHORT LIFE (SHL) TFs, which are known to recruit ASHH2 to FLC and recognize the H3K4me3 mark, respectively, also bind to trithorax gene promoters (Fig. 5).

Among the trithorax gene promoters, Soltu.DM.06G019650 (StPKL) exhibited the highest binding affinity (frequency and diversity of TF binding motifs on the promoter) for the flowering-related TFs, whereas Soltu.DM.05G005310 (StATXR1) showed the least (Data_Sheet_7). Notably, BPE was found to bind to fewer trithorax gene promoters, while the CDF binding site occurred the maximum number of times in all the promoters. Moreover, Soltu.DM.03G010440 (StATXR4), Soltu.DM.06G019650 (StPKL), and Soltu.DM.05G012810 (StREBELOTE) bind to the maximum (21) type of TFs, whereas Soltu.DM.05G005310 (StATXR1) binds to the minimum (11) type of TFs (Fig. 5).

Trithorax genes are involved in photoperiod-mediated stolon-to-tuber transition in potato

In potato, the below-ground modified stem (stolon) undergoes distinct developmental stages (stolon, swollen stolon, and mini-tuber) to form the edible tuber. To investigate the role of trithorax genes in tuber development, we examined their transcript abundance across various developmental stages under both tuber-inductive (in short-day (SD) condition) and non-inductive (in long-day (LD) condition) conditions in the photoperiod-sensitive wild type variety S. tuberosum subsp. andigena line 7540. Our analysis revealed significant differential expression of trithorax genes at different stages of potato development (Fig. S8).

Heat map analysis resulted in the clustering of genes into 11 groups: 4 major groups and 7 minor groups (Fig. 6). One group, consisting of nine genes, exhibited decreased expression post-induction throughout tuber development, including Soltu.DM.12G000190.1 (StASHH3_1), Soltu.DM.01G001770.2 (StASHH3_2), Soltu.DM.01G033850.1 (StBRM), and Soltu.DM.11G018160.4 (StSYD), suggesting their potential role as negative regulators of tuberization. Conversely, other major group showed a slight increase in expression levels during tuber development, such as Soltu.DM.02G027040.1 (StWDR5A), which exhibited a progressive increase throughout the developmental stages (Fig. 6). Another major group, containing three genes (Soltu.DM.11G008270.1 (StATXR5), Soltu.DM.09G026630.1 (StASHR2), and Soltu.DM.05G012810.1 (StREBELOTE_1)) and a minor group containing Soltu.DM.03G014760.1 (StSWI1) exhibited a marked increase in expression, particularly during the swollen stolon stage, indicating their possible involvement in radial stolon expansion. Three minor groups contained genes with elevated expression levels during tuber development stages (SD-induced stolon, swollen stolon, and mini tuber), including Soltu.DM.01G036130.1 (RbBP5-L), Soltu.DM.07G019360.2 (StULT), and Soltu.DM.05G005310.1 (StATXR1) displayed a gradual increase in expression higher than the previously described group. Though other genes clustered in groups, the Soltu.DM.07G018390.1 (StATXR3) specifically showed a significant increase in expression in the swollen stolon stage. Moreover, some of the genes (Soltu.DM.03G022600.2 (StATX4), Soltu.DM.11G018160.4 (StSYD), Soltu.DM.01G035010.1 (StATX3), Soltu.DM.11G008270.1 (StATXR5), Soltu.DM.01G036130.1 (StRbBP5-L), Soltu.DM.05G005310.1 (StATXR1), and Soltu.DM.07G018390.1 (StATXR3)) exhibited higher expression in SD induced stolons compared to uninduced ones, suggesting their potential role in photoperiod-mediated tuberization (Fig. 6).

Fig. 6.

Real-time gene expression of 29 trithorax genes in the stolon-to-tuber transition stages of S. tuberosum subsp. andigena line 7540. Heat map showing the relative gene expression in terms of fold change with respect to LD stolon. The colour within rows depicts normalized (log) average fold change values from LD stolon to SD stolon, swollen stolon and mini tuber of tuber developmental stages. The analysis was performed using three biological replicates. The plot was created in R studio using the “pheatmap” package

Based on the above findings, we selected trithorax members showing significant differential expression in potato development for interaction analysis. Our findings predicted that many of these trithorax proteins interact amongst themselves and also with polycomb proteins involved in tuberization (Fig. 7). Furthermore, we determined the tertiary structure of the Trithorax proteins in potato, providing insights into their functional mechanisms during tuber formation (Table S2) (Fig. S9, S10).

Fig. 7.

The schematic illustrates the putative interactions between trithorax proteins and tuberization pathway proteins, alongside the selective TFs that bind to trithorax gene promoters during stolon-to-tuber transitions. The trithorax genes are depicted at the stage of their peak expression. Solid pink and green lines represent protein-protein interactions, while dashed arrows indicate the targets of TFs

Trithorax proteins may potentially be involved in storage organ development in root crops

Given the fact that the development of storage organs shares several molecular factors with that of potato tuberization, trithorax proteins may also play a conserved function in the development of these storage crops. Hence, in this study, we identified 170 protein orthologs of the A. thaliana trithorax proteins in various storage root crops. These comprise 28 proteins in Daucus carota (carrot), 35 in Manihot esculenta (cassava), 31 in Beta vulgaris (sugar beet), 40 in Ipomoea trifida (sweet potato), and 36 proteins in Raphanus sativus (radish), respectively (Table S3).

A phylogenetic tree was generated to examine the evolutionary links between the identified proteins and the trithorax members of S. tuberosum and (A) thaliana (Fig. S11). The proteins were segregated into three different groups (groups I–III), having 9 clades. Group I and III were branched at the base, while group II may have evolved later. Four clades were identified within group I, with clade A including the PKL, SYD, SWI2, and BRM orthologs from all the species. As anticipated, the functional domain SNF2-rel_dom was highly conserved among these proteins (Fig. S12). The clade B contains WDR5A/5B and RbPB5-like orthologs with conserved repeats of WD40 domains, the clade C comprises SET domain and AWS domain containing ASHH1/2/3/4 orthologs, while clade D encompasses ULT1/2 and TRO orthologs. Group II is composed of proteins containing the SET domain, separated into two clades. Clade E contains all the REBELOTE proteins, whereas clade F contains ATXR1/2/4. Interestingly, the ASHR2 appear in two groups: ASHR2 of R. sativus, D. carota, S. tuberosum, and I. trifida in clade C of group I, whereas ASHR2 of (B) vulgaris, M. esculenta, and A. thaliana in clade F of group II (Data_Sheet_8). Group III is further subdivided into three clades. While the PHD, PHD_2, and PWWP domains containing ATX1/2/3/4/5 proteins cluster in clade G, all of the ATXR3 proteins and BSH, ASHR1/3, ATXR5/6 group in clade H, and I, respectively (Fig. S11).

The orthologs of trithorax proteins, which were differentially expressed during the stolon-to-tuber transition in S. tuberosum subsp. andigena line 7540, as mentioned previously, were selected to explore the similarities in the regulation of these genes (Fig. S13). The comparative analysis of the gene promoters with differentially expressed genes of S. tuberosum revealed a high degree of similarity in the CRE distribution and abundance (Fig. 8A). In general, the stress-responsive CREs have the highest abundance, followed by phytohormone-related and development-related CREs. It is interesting to note that while the D. carota ortholog of SYD had equal amounts of CREs associated with development, hormones, and stress, the R. sativus SWI1 ortholog possessed a 100% abundance of stress-related CREs (Fig. 8A). The presence of these CREs suggests that these genes may be involved in phytohormone-signalling, stress response, and in regulating developmental pathways.

Fig. 8.

Identified CREs and flowering-related TFs found in the promoter of trithorax gene orthologs in storage root crops. A The distribution of CREs in 2 kb promoter sequences of trithorax gene orthologs in underground storage crops was analyzed using the PlantCARE Query Search online tool. A total of 33 CREs were categorized into three groups: hormone-related, stress-related, and growth & development-related. The bar plot with blue colour, red colour, and green colour indicates the abundance of stress-related, development-related, and hormone-related CREs across the promoter sequences, respectively. B Circular heat map showing the number of binding sites of flowering-related TFs in the promoter regions of trithorax orthologs in storage root crops. TFs were predicted using PlantPAN4.0. The colour within the cells represents the number of binding sites of the respective TF

We also examined the possible interaction of TFs involved in the flowering pathway with the promoters of these genes for their regulation. TFs such as, CDF3, SPL9, AGL6, STM, TEM1, and CIB1, which are involved in flowering time, floral transition, and floral organ development, had the highest number of binding sites in the promoters of trithorax orthologs in all storage root crops and this is consistent with potato (Fig. 8B). Possible interactions between flowering-related TFs with the trithorax genes of storage root crops also indicates that these genes may have a role in storage organ development because of the similarities with potato tuberization and flowering pathways in A. thaliana.

Discussion

The evolution of trithorax proteins in plants

Trithorax genes are essential epigenetic regulators and play a crucial role in plant development [1, 4]. These genes are majorly studied in A. thaliana and relatively less explored in other plants. Previous studies have investigated ATX1 and ATX2 in microalgae and compared them with orthologs in D. melanogaster, Mus musculus, and A. thaliana [32]. Additionally, RegA protein, a member of the same family (ULT proteins), has been studied in V. carteri for its role in re-differentiating somatic cells into gonidia cells [33]. Moreover, some research has also explored trithorax protein functions in other plants, such as O. sativa [20], Fragaria vesca [22], and Pinus radiata [34]. In this study, we have carried out a comprehensive analysis to explore the trithorax proteins in the larger context of plant evolution, diversification, functionalities, and gene regulation, with a special emphasis on photoperiod-mediated tuberization in potato. Further, a comparative analysis of selective trithorax proteins has been conducted in five storage root crops.

Our analysis shows that all green lineages, from primitive unicellular algae to intricate multicellular angiosperms possess trithorax proteins, and this conservation across diverse plant lineages indicates their crucial role in regulation of gene expression throughout plant evolution and diversification. It is well reported that from unicellular algae to angiosperms, key activities of trithorax proteins such as histone tail methylation [15, 32], chromatin remodeling [35], and interactions with transcription machinery are conserved [33, 36]. This is also evident from our analysis, as the functional domains associated with these processes are conserved across all the trithorax proteins, indicating their sustained molecular function across all plant lineages (Fig. S3). Moreover, we found an increase in trithorax gene copy number from early land plants to advanced angiosperms (Table S1), which can be attributed probably to the greater complexity of angiosperm body plans and functions [37]. Further, proteins from algae and bryophytes (arising from the base of the tree) cluster apart from angiosperms, whilst proteins from monocots and dicots form unique clusters that illustrate multiple evolutionary routes (Fig. 1). This evolutionary divergence has also been reported earlier for SET domain proteins, which diversified through extensive duplications before the monocot-dicot split, illustrating their ancient and intricate evolution [38]. Though trithorax proteins from different lineages show multiple evolutionary routes, a close clustering pattern in the trithorax proteins of S. tuberosum and A. thaliana suggests their functional similarity among dicots besides a high degree of conserved nature in these two species. However, further experimental research will be necessary to reaffirm these speculated functional similarities.

Potential role of trithorax genes in potato tuber development

Before elucidating the probable role of trithorax genes in potato tuberization, it is pertinent to discuss how these genes diversified in potato. One of the most significant factors driving gene diversification is the process of gene duplication. Gene duplication plays an important role in the speciation and adaptation of plants into varied environments [39]. Similarly, gene duplication is considered responsible for the increased number of SET domain genes in plants, when compared with other organisms such as D. melanogaster, M. musculus and Saccharomyces cerevisiae [38]. We observed that gene duplication has played a significant role in the proliferation of trithorax genes in the potato (Fig. 3). This observation is consistent with a previous report, where it was shown that tandem duplications alone occurred in 23% of the SET domain genes [28], however, the study was restricted to the analysis of SET domain proteins only.

Amongst the targets of StBEL5, a key tuberization marker gene in potato, several were noted to be trithorax proteins, indicating their potential role in tuberization. Previously, we have reported the involvement of trithorax proteins in tuber development [25–27]. Exploring how trithorax genes could influence photoperiod-mediated tuberization will provide valuable insights into their regulatory roles and enhance our understanding of potato development.

Interestingly, two crucial microRNAs, miR156 and miR172, have been reported to play pivotal roles in the transition from the vegetative to the reproductive phase of plant development [40]. During tuberization, the former accumulates in below-ground stolon, and supporting vegetative growth in aerial parts under non-inductive conditions by targeting and cleaving StSPL TFs [41], while the latter enhances tuberization through the PHYB-miR172-BEL5 pathway [42], targets and cleaves AGL and TOE2 [43]. It is pertinent to mention here that StSPL genes are also regulated by photoperiod and gibberellins (GA) [44]. We predicted that SPL TFs bind to the promoters of trithorax genes (Fig. 5), suggesting a link between these genes and miR156, indicating their role in tuber development. We showed that miR156 contains an H3K4me3 activation mark in SD-induced stolon further validating the role of trithorax genes in tuberization [27]. Moreover, AGL TFs, which are targets of miR172, bind to the promoters of trithorax genes, suggesting their potential role in developmental regulation. It is also reported that the miR172-TOE-FT module regulates flowering in A. thaliana [45]. Keeping in view the similarities between flowering and potato tuberization, investigating the TOE2 TF led us to identify 21 trithorax genes as their putative targets in potato. It would be interesting to experimentally validate these putative targets in potato.

Besides these, another key gene involved in tuberization is StCDF1, which promotes tuber formation by inhibiting StCONSTANS-LIKE1 (StCOL1) activity. StCDF1 stability is regulated by the StGI-StFKF-StCDF1 module, and its expression is controlled by StCCA1 [46]. We noted that, in potato, CDFs and CCA1 bind to the promoters of all trithorax genes, suggesting potential involvement in tuber development. Among all trithorax family members, StPKL demonstrated the highest affinity for SPL, CDF, and CCA1. StPKL expression increases upon SD induction and then gradually decreases. This pattern suggests that StCDF and StSPL contribute to the upregulation of StPKL during tuber initiation, consistent with their known roles as positive regulators of tuberization in potato. Moreover, StCO and StSELF-PRUNING 6 A (StSP6A) are crucial in the tuberization process, with StCO acting as a negative regulator and StSP6A as a positive regulator [47]. In A. thaliana, CIB1 interacts with CO to enhance FT (the ortholog of StSP6A) transcription, promoting flowering [48]. Our study found that CIB1 binds to the promoters of 25 trithorax genes in potato, suggesting its potential role in tuberization. Additionally, the StSTM transcription factor, involved in heat stress-induced tuber sprouting in potato [49], binds to the promoters of all identified trithorax genes. In Arabidopsis, EFS (ASHH2) is recruited by SUF4 to FLC clade genes to facilitate flowering [50], and SUF4 binds to the promoters of 14 potato trithorax genes (Fig. 5). Overall, our analysis reveals that trithorax genes may play a crucial role in tuberization.

Functional significance of trithorax genes in potato tuberization

Trithorax proteins play a key role in reproductive organ development and flowering. Earlier, we have shown that key tuberization genes such as StBEL5, StSP6A, StPOTH1, miR156/172 and several others harbour the H3K4me3 activation mark, which is embarked by trithorax proteins [27]. In order to assess the role of trithorax genes in photoperiod-mediated stolon-to-tuber transition stages, expression analysis of all trithorax members has been performed. It was found that many of the trithorax genes involved in flowering were differentially expressed in the tuber developmental stages (Fig. 6). It is reported in A. thaliana that SYD is essential for the proper expression of floral homeotic genes, overcoming polycomb repression to regulate flowering [16]. Moreover, WDR5B, in conjunction with TRO and RbBP5, forms a COMPASS-like complex that controls the floral transition [51]. High transcript abundance of StATX4, StSYD, and StWDR5B in SD stolon suggests their possible role in photoperiod-mediated tuberization. This observation is consistent with the fact that StATX4 was downregulated in SD-induced stolon tissues of StPOTH15 overexpression transgenic lines, where StPOTH15 is known to repress tuberization [26]. Moreover, another trithorax member ULT1, together with KANADI is known to regulate genes involved in gynoecium and stamen development in Arabidopsis [13]. Several genes, including StATXR1, StRbBP5-L, StWDR5A, and StULT, showed a gradual increase in expression throughout tuber development, with StATXR1 exhibiting the highest expression, suggesting its role as positive regulators of tuberization, particularly at later stages of development. Consistently, StULT was found to be downregulated in the SD-induced stolon tissue of StPOTH15 overexpression transgenic lines and stolon tissues of StBEL5 transgenic lines [24, 26].

Previous studies have also shown differential expression of trithorax genes, with StATXR3 and StASHR3 being upregulated in axillary meristems of StMSI-1 overexpression lines, suggesting its involvement in aerial tuber development [25]. StATXR3 also harbors H3K27me3 mark in SD-induced below-ground stolons in StE(Z)2 overexpression lines, in which belowground tuber formation was reduced [27]. StATXR3 expression increases upon SD induction and peaks at the swollen stolon stage. Together, these findings suggest StATXR3 is a key trithorax member involved in potato tuberization, and could be considered for further studies. StASHR2 was upregulated, while StATXR2, StATXR3, and StULT were downregulated in stolon tissues of StBEL5 transgenic lines [24]. Moreover, this study also identifies interactions between various trithorax members and genes involved in tuberization, suggesting that these interactions may regulate potato tuber development. This is also evident from the fact that trithorax proteins typically function in complexes with other trithorax members and transcription factors to modulate their activity. For example, ULT1 and ATX1 collaborate with the polycomb member EMBRYONIC FLOWER1 (EMF1) to prevent untimely seed germination while they antagonize EMF1 to regulate flowering [52, 53]. In addition, trithorax proteins are part of the COMPASS complex, which controls flowering time in O. sativa and A. thaliana [21, 51]. Based on the above, it is presumable to speculate the involvement of trithorax genes in photoperiod-mediated tuberization.

Trithorax genes in storage root crops

Although storage organ crops are a major part of the global staple diet, the molecular pathways regulating the development of these storage organs are poorly studied, except for potato. Studies have shown similarities in the pathways governing storage organ development in these crops. For example, TFs such as KNOTTED1-like (KNOX) contribute to the formation of storage organs in potato, and sweet potato and cassava [54–56], MADS-box genes regulate potato tuberization [57], and storage root development in sweet potato [58]. Likewise, StBEL5 (an important tuberization activator) and BEL1-like proteins, have been shown to promote root growth in potato and sweet potato, respectively [59, 60]. Moreover, other genes, such as the PHOSPHATIDYL ETHANOLAMINE-BINDING PROTEIN (PEBP) family, share a common developmental pathway and are mostly involved in the formation of storage organs [61]. However, other than the PEBP family, it’s unclear if the genetic pathways found in crops with different storage organs are all comparable or restricted to closely related species only. In this study, trithorax gene orthologs in storage root crops (D. carota, M. esculenta, B. vulgaris, I. trifida, and R. sativus) were found to have conserved functional domains and motifs similar to those observed in potato (Fig. S12). Persistent CRE abundance and TF binding amongst these storage root crops and potato, suggest a conserved role of trithorax proteins in storage organ development (Fig. 8). Particularly, TFs which had the maximum number of binding sites amongst the twenty-four flowering-related TFs were CDF3, SPL9, AGL6, STM, TEM1, and CIB1, which are known to be involved in the regulation of flowering time, vegetative-to-reproductive transition, and floral organ development [62–67], which is common to potato. These findings suggest that trithorax gene regulation in storage root crops is likely comparable to that of potato. This study reveals a consistent pattern of trithorax gene regulation across various storage crops and suggests that similar gene regulatory mechanisms might govern both stem and root storage organ development. However, additional experimental research is required to provide concrete evidence and validate this hypothesis.

Trithorax group members have not been extensively characterized in major crop plants. This study provides a foundational framework for future research by offering comprehensive insights into their potential roles, utilizing bioinformatics analyses and expression profiling. These findings open new avenues for understanding the functional significance of Trithorax members in crop development and stress responses.

Conclusion

In this study, we identified a total of 268 trithorax family members across a diverse range of plant species, including S. tuberosum, S. lycopersicum, O. sativa, Z. mays, P. patens, M. polymorpha, C. reinhardtii, and V. carteri. Our analysis revealed that trithorax proteins cluster into monophyletic groups, with distinct classifications among SET domain proteins, chromatin remodelers, and associated proteins. Interestingly, our phylogenetic analysis indicated that trithorax proteins in algae evolved earlier than in bryophytes and angiosperms. Within angiosperms, monocots and dicots are clustered into separate monophyletic groups. Moreover, the study demonstrated that functional domains and motifs, predicted sub-cellular localization, and physico-chemical properties of these proteins are largely conserved across the species analyzed. Additionally, gene structure analysis provided insights into the evolutionary divergence of trithorax genes. In potato, we observed a significant increase in trithorax gene copy number attributed to the gene duplication. To explore the regulation of these genes in potato, analysis of cis-regulatory elements (CREs) in their promoters identified an abundance of stress and hormonal-related elements. Several TFs associated with key developmental processes, such as flowering and tuberization were found to bind to trithorax gene promoters, suggesting a role for trithorax genes in the tuberization pathway. Further, gene expression analysis during the photoperiod-mediated stolon-to-tuber transition stages showed differential expression of many identified trithorax genes, amongst which StATXR3, StPKL, and StATXR1 were upregulated upon SD induction and in the later stages. Moreover, the promoter of these genes showed high affinity for important TFs like CDF, CCA1, SPL, and AGL, suggesting their probable role in tuberization. These genes can be further studied to understand their specific role in tuber development. A comparison of selective trithorax members in storage crops, suggested a common gene regulatory mechanism between the tuber development in potato and other root crops. This study offers novel insights into the evolution, conservation, diversification, and expression profiling of trithorax members in potato and unfolds their potential role in other storage crops, laying the foundation for further research.

Materials and methods

Identification of trithorax gene family members

The NLM-NIH NCBI Gene database (https://www.ncbi.nlm.nih.gov/gene) was used to retrieve the genes of trithorax members in the model plant A. thaliana, and the sequences were further confirmed in the Phytozome database (A. thaliana, TAIR10, version 13.0) (https://phytozome-next.jgi.doe.gov/info/Athaliana_TAIR10). A BLAST search was performed in the Phytozome v13.0 with A. thaliana trithorax proteins as a query against the proteome target of S. tuberosum (Potato, v6.1) with the expect (E) threshold value of −1 and comparison matrix BLOSUM62 (https://phytozome-next.jgi.doe.gov/blast-search). The retrieved trithorax protein sequences of S. tuberosum were further verified in the Spud DB Potato Genomics Research database (http://spuddb.uga.edu/integrated_searches.shtml). Further, the same (A) thaliana trithorax protein sequences were used as query, and BLASTp was performed against other storage root crops, such as D. carota subsp. sativus (Carrot, v3.0), M. esculenta (Cassava, v8.1), (B) vulgaris subsp. Vulgaris (Sugar Beet, EL10.2_2), I. trifida (Sweet Potato, v3), R. sativus (Radish, ASM80110v3) besides other non-vascular and vascular plants like the S. lycopersicum (Tomato, ITAG5.0), O. sativa (Rice, v7.0), Z. mays (Maize, PHJ40 v1.2), P. patens (v6.1), M. polymorpha (v3.1), (C) reinhardtii (CC-4532 v6.1) and V. carteri (v2.1) in Phytozome v13.0, Sweetpotato Genomics Resource database (http://sweetpotato.plantbiology.msu.edu/blast.shtml) and NCBI (www.ncbi.nlm.nih.gov) where ever applicable to find trithorax orthologs in these species respectively. To ensure that all the retrieved sequences are indeed trithorax members, these were further validated by checking the presence of the respective protein functional domain (SET domain- pfam00856, SNF2-rel_dom- pfam00176 and WD domain- pfam16529) using the NCBI Conserved Domain Database- CDD (searching against Pfam v35.0 with the threshold value of 0.01) [68]. Finally, a total of 268 trithorax protein sequences, including 152 SET domain proteins, 66 Chromatin remodelers, and 50 Associated proteins were selected for further analysis from 2 Algal, 2 Bryophyte, and 5 Angiosperm (2 Monocots and 3 Dicots) species. Additionally, to gain insights into the trithorax-mediated response, these 268 selected sequences were further compared with 168 trithorax members in 5 storage root crops, as mentioned previously.

Multiple sequence alignments and phylogenetic analysis

To look into the evolutionary relationship among the trithorax proteins, the selected 268 sequences were further categorized into three groups of SET domain proteins, Chromatin remodelers, and Associated proteins. The sequences were subjected to multiple sequence alignment (MSA) using the MUSCLE alignment tool [69], with the UPGMA cluster method and 16 max iterations on MEGA (version 11) software [70]. The phylogenetic trees were generated on MEGA (version 11) software using the Neighbor-joining (NJ) method with a Poisson substitution model created from 1000 bootstrap replicates [71]. A separate phylogenetic tree was generated among the trithorax protein orthologs of selected storage root crops (I. trifida, M. esculenta, B. vulgaris, R. sativus, and D. carota) as well as the protein sequences from S. tuberosum and A. thaliana following the same parameters as mentioned previously. The tree was visualized in iTOL (https://itol.embl.de/) [72], and the final version was generated in FigTree (version 1.4.4) for improved representation and analysis.

Identification of conserved functional protein domains, motifs, and gene structure analysis

The conserved domains in the trithorax proteins were identified using the Pfam (v35.0) resource base through the NCBI Conserved Domain Database (CDD) (https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi) with the expect threshold value of 0.01 [68], and the data was downloaded as.hitdata with full description. The MEME suit (Multiple Expectation maximizations for Motif Elicitation) version 5.5.6 online tool (https://meme-suite.org/meme/tools/meme) was used to identify 15 conserved motifs of length ranging from 6 to 250 residues [73]. The Bio Sequence Structure Illustrator (Advanced Gene Structure View) tool of TBtools-II was used for the schematic representation of both domains and motifs with respect to their evolutionary relatedness [74]. The functional description of each detected conserved domain and motif was obtained from the InterPro-EMBL-EBI database using Pfam sequence search (https://www.ebi.ac.uk/interpro/entry/pfam/). Moreover, the gene structure analysis was performed using the annotation files (.gff3) of the selected genomes in the Bio Sequence Structure Illustrator tool of TBtools-II [74].

Protein tertiary structure prediction

The tertiary structure, topology, and transmembrane helices of the trithorax proteins were predicted using the Phyre2 web portal (http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgi? id=index), run in expert mode following the batch processing [75]. The conserved regions in each clade of the phylogenetic tree were visualized in the AlphaFold Protein Structure Database (https://alphafold.ebi.ac.uk/) using the respective A. thaliana trithorax protein.pdb structure files [76]. The conserved regions were highlighted with green colour against a white theme.

Protein physico-chemical properties, sub-cellular localization, and protein-protein interaction

The physico-chemical properties, such as the number of amino acid residues, molecular weight, theoretical pI, instability index, aliphatic index, and grand average of hydropathicity of trithorax protein members were calculated using the Expasy ProtParam (https://web.expasy.org/protparam/) online tool [77], through TbTools-II [74]. In addition, the sub-cellular localization of trithorax proteins was predicted using DeepLoc (version 2.0) (https://services.healthtech.dtu.dk/services/DeepLoc-2.0/) [78], SignalP (version 6.0) (https://services.healthtech.dtu.dk/services/SignalP-5.0/) [79], and TargetP (version 2.0) (https://services.healthtech.dtu.dk/services/TargetP-2.0/) [80] online tools, however, the predictions from the DeepLoc–2.0 were finally considered in this study.

The protein sequences of genes involved in the tuberization pathway were retrieved from the Phytozome v13.0 database. The trithorax and tuberization pathway proteins of S. tuberosum were used as a query on the STRING (version 12.0) online tool (https://string-db.org/cgi/input? sessionId=br0W4AhKp4lx&input_page_active_form=multiple_sequences) to find out the protein-protein interactors among these.

Cis-Regulatory Element (CRE) analysis and transcription factor prediction

The 2 kb promoter regions of trithorax genes in S. tuberosum and other storage crops were obtained from Phytozome (v13.0), Sweetpotato Genomics Resource database (http://sweetpotato.plantbiology.msu.edu/blast.shtml), and NCBI (www.ncbi.nlm.nih.gov). The PlantCARE Query Search online tool (https://bioinformatics.psb.ugent.be/webtools/plantcare/html/) was used to identify the putative cis-regulatory elements (CREs) on these promoters [81]. The identified CREs were divided into three groups (hormone-related, stress-related, and growth & development-related) manually. The presence and absence of different types of CREs were further visualized using the HeatMap illustrator of TbTools-II [74].

The putative transcription factors involved in flowering and tuberization, binding to the promoter sequences were retrieved using the promoter Analysis software (PlantPAN4.0 http://plantpan.itps.ncku.edu.tw/plantpan4/promoter_multiple.php) against the transcription factor database of the model plant A. thaliana [82]. The transcription factors and their characteristics were confirmed using TF/TFBS (http://plantpan.itps.ncku.edu.tw/plantpan4/TF_TFBS_search.php) and Gene Search (http://plantpan.itps.ncku.edu.tw/plantpan4/gene_search_setting.php) tool of PlantPAN4.0 [82].

Chromosomal localization and gene duplication analysis

The chromosomal location and genome-wide distribution of trithorax genes in potato were obtained from the genome annotation files (.gff3) of S. tuberosum (v6.1) from Phytozome. The collinear trithorax genes in the potato genome were identified using the MCScanX software of TbTools-II [74, 83]. The GC skew, which indicates strand-specific guanine and cytosine overrepresentation, the chromosomal localization, and genome‐wide gene density was obtained from the genome assembly and annotation files using “Fasta Window Stat” and “Gene Density Profile,” functions of TbTools-II [74]. The final plot was generated as an advanced Circos plot showing chromosomal skeleton, GC ratio, GC skew, and gene density represented as circular rings from inside to outside [84]. The duplicated trithorax gene pairs were used for the calculation of non-synonymous substitution per non-synonymous site (Ka), synonymous substitution per synonymous site (Ks), and Ka/Ks ratio by the “Ka/Ks calculator” function of TbTools-II [74]. The divergence time in million years ago (MYA) of each duplicated trithorax gene pair was calculated by the formula T = Ks/2r, where r indicates the divergence rate. The value of the divergence rate was implicated as 6.5 × 10 − 9 in angiosperms [85].

Expression analysis of the identified trithorax genes in solanum tuberosum

Plant material and growth conditions

The photoperiod-sensitive wild type cultivar S. tuberosum subsp. andigena line 7540 which tuberizes under short-day conditions, was utilized for this study. Wild-type plants were propagated in vitro in Murashige and Skoog basal medium utilizing nodal cuttings as explants [86]. The cultures were maintained in a growth incubator (Percival Scientific) at 22 °C with light intensity of 300 µmol m⁻² s⁻¹ under long-day conditions (16-hour light/8-hour dark). Twenty-eight-day-old plants (36 plants) were transferred to soil and maintained under the same conditions for eight weeks. Subsequently, eighteen plants were maintained in long-day conditions, while the remaining eighteen plants were subjected to short-day conditions (8-hour light/16-hour dark) for seven days. The experiment was performed with three biological replicates with 6 plants per replicate. Stolon-to-tuber transition stages were harvested post-short-day induction and stored at −80 ℃ till further use.

RNA isolation and quantitative real-time PCR analysis

RNAiso Plus reagent (TaKaRa Bio, Japan) was used to isolate total RNA in accordance with the manufacturer’s instructions. cDNA synthesis was performed from 2 µg of the total RNA using M-MLV Reverse Transcriptase (Promega, USA), and oligo(dT) primers. Quantitative reverse-transcription PCR (RT-qPCR) reactions were carried out on a CFX96 Real-Time System (Bio-Rad, USA) with gene-specific primers (Table S4) and TB Green^® Premix Ex TaqTM II (TaKaRa Bio, Japan) by incubating at 95 °C for 30 s, followed by 40 cycles at 95 °C for 5 s, 56 °C for 15 s and 72 °C for 20 s. Data was normalized using the eIF3E housekeeping gene whose expression is shown to remain constant throughout all the tissue types and growth conditions [87]. The data was calculated as 2–ΔΔCT values with respect to long-day stolon’s and expressed as relative fold-change [88]. The experiment was performed with three bio-replicates in triplicate.

Abbreviations

CREs

Cis-regulatory elements

TFs

Transcription factors

NJ

Neighbor-joining

ERE

Ethylene responsive elements

ABRE

Abscisic acid responsive elements

AuxRR

Auxin responsive elements

WRE

Wound responsive elements

SD

Short-day

LD

Long-day

Authors’ contributions

Original idea was conceived by AKB. SS has designed the complete data analysis plan and necessary experiments, grown the plants to obtain stolon samples. SS and MPM extracted RNA samples and completed the qRT-PCR work. SS has performed the comprehensive data analysis for potato and other extant land plants. MPM performed data analysis for five storage root crops. MIK has analyzed all the data output in this project and provided necessary inputs. SS, MPM, MIK have written the manuscript. AKB obtained project funds, provided resources and supervised the entire project. AKB edited the final manuscript, and all authors read and approved it.

Funding

This Research was supported by DBT, Govt. of India, TATA Innovation Grant to AKB, Grant No: HRD-16012/7/2021-AFS-DBT.

Data availability

The datasets generated and analyzed during the current study are available as supplementary materials.

Declarations

Ethics approval and consent to participate

Not applicable.

Competing interests

The authors declare no competing interests.