Guar (Cyamopsis tetragonoloba (L.) Taub.) is an economically important short-day legume, whose seeds are the main source of guar gum, a compound used as a thickening agent in the food industry and as a fracturing fluid during oil and gas extraction [1]. At high latitudes, e.g., in Russia, the main obstacle to its cultivation is the long photoperiod during summer which delays flowering and maturation of guar plants [2]. Knowledge of the genes involved in guar flowering pathways will facilitate the implementation of genome editing technologies in guar breeding programs to improve yield and adaptability.

Currently, another valuable crop, soybean, serves as a model plant for understanding photoperiodism in short-day legumes [3]. Different allele combinations of genes involved in photoperiod responses have facilitated the adaptation of soybean varieties to the diverse geographic regions and farming systems [4]. In soybean the photoperiod-specific expression of the florigen genes is directed by interaction between light perception and endogenous circadian rhythms. The florigen genes, sequentially, encode mobile signals that move from leaf and shoot apex to induce flowering [5].

At least 11 flowering and maturity loci were identified in soybean (E1-E10, J) [6, 7] that are involved in the molecular mechanisms of flowering of this short-day legume under long days condition. However, the most significant influence on flowering time in soybean is exerted by four maturity genes, designated E1, E2, E3 and E4 [4].

Two florigen orthologs of Arabidopsis FLOWER LOCUS T (FT) identified in soybean, GmFT2a and GmFT5a, collectively regulate flowering time, with GmFT5a, but not GmFT2a, being essential for soybean adaptation to high latitude regions [8]. Induction of GmFT2a and GmFT5a expression, in turn, is under the control of the two photoreceptors encoded by phytochrome A (PHYA) genes E3 and E4; which inhibit expression of these florigens under long-day (LD) conditions. This PHYA-dependent regulation of flowering is mediated by the E1 gene, a repressor of GmFT2a and GmFT5a [6]. The legume-specific E1 gene is considered to be a major locus associated with flowering time and maturity in soybean, and its malfunctioning alleles promote plant flowering under long-day conditions [9].

Besides PhyA genes (E3 and E4), there is another pathway that regulates GmFT expression. Arabidopsis GIGANTEA (GI) ortholog, GmGIa, was identified as a candidate gene for E2 locus that suppresses flowering of soybean under long photoperiod. The recessive nonsense e2 allele of this gene promotes early flowering through the up-regulation of GmFT2a expression, although has no visible effect on the expression of GmFT5a [10].

Various combinations of mutations at E1-E4 loci underlie the diversity of flowering and maturity time in modern soybean varieties, allowing them to adapt to the high latitudes with a long photoperiod. In general, mutations of E1-E4 genes reduce the photoperiod sensitivity and shorten the growth period of soybean cultivars [4]. Thus, the E loci, and especially the E1 gene, are becoming a major target for CRISPR-Cas editing in studies aimed at creating of early flowering germplasm of soybean suitable for high latitudes [7, 11]. Information about flowering genes gained from the well-studied model crop soybean can be used to understand the photoperiodic pathway in guar.

Recently, we have reported the results of comparative transcriptome and metabolome analysis of early- and late-flowering guar genotypes grown under long-day conditions [2]. For that analysis we generated de novo genome-guided Trinity transcriptome assembly employing the draft of guar genome from our previous study [12]. As a result, 102,539 transcripts of the Trinity genome-guided assembly were combined in 96,447 clusters of all revealed isoforms. When the 78,015 longest ones were blasted against the Arabidopsis thaliana genome, only 15% of those 78,015 guar transcripts were successfully annotated.

In the present study we made an effort to annotate these 96,447 guar cluster transcripts using blast against Glycine max genome to identify orthologs of soybean genes in guar genome, e.g., loci responsible for flowering and maturity. We also used a bioinformatics approach employing transcriptome assembly, draft genome [12] and the reference genome of guar (Cte V1.0, GCA_037177725.1) to characterize the E1-E4 guar orthologs, predicting the exon-intron structure of these flowering-related genes as promising targets for CRISPR-Cas editing.

In the present study, we provide the first accessible version of guar transcriptome assembly containing sequences of 96,447 clustered transcript isoforms detected in leaves of guar plants grown under the long day conditions. To date, several RNAseq studies have been reported for guar [33, 34, 35, 36, 37], however, no accessed Transcriptome Shotgun Assembly (TSA) was available for Cyamopsis tetragonoloba in the European Nucleotide Archive (https://www.ebi.ac.uk/ena) (27 July 2024). The developed sequenced resource will complement the recently published C. tetragonoloba reference genome assembly (CteV1.0) [38] and contribute to guar breeding programs via an application of biotechnological methods, e.g., CRISPR-Cas genome editing. As was shown with the example of several genes (Table 1), the detected transcripts perfectly match the sequence of the reference CteV1.0 guar genome. Thus, for most of the expressed genes, their physical position on the guar chromosomes can now be easily determined.

44.2% of transcripts from the TSA showed similarity to Glycine max genes, while only 15% of transcripts were successfully annotated earlier via blast against the Arabidopsis thaliana genome [2]. Although the majority of Arabidopsis flowering genes are represented in legume sequence databases, several gene families including MADS-box, CONSTANS, and FLOWERING LOCUS T families have been reported to have undergone significant changes in their composition [39]. As an example, compared with their roles in Arabidopsis, photoreceptors GmPHYA3 (E3) and GmPHYA2 (E4) have conserved and novel functions in soybean [40]. In contrast, many flowering gene orthologs in different legume genera and even tribes demonstrate a high level of synteny and are located in the conserved map positions, e.g., Pisum (Tribe: Vicieae) and Medicago (Tribe: Trifolieae) [39]. Considering that both Glycine max and Cyamopsis tetragonoloba belong to the same Indigoferoid/Millettioid clade of legumes, the high similarity of their genomes was expected and estimated as the highest (41.91%), followed by Phaseolus vulgaris, Cicer arietinum and Medicago truncatula [33].

We used the developed transcriptome assembly to discover in the guar genome orthologs of four soybean E loci that have the greatest impact on flowering and maturation of varieties of the short-day legume crop. A high level of identity was detected between peptide sequences encoding by orthologous genes E1 and CtE1 (80%), E2 and CtE2 (93%), E3 and CtE3 (83%), E4 and CtE4 (91%). Not only the sequences but also the intron-exon structure of the genes in soybean and guar were found to be similar, suggesting that the genetic pathways underlying basic flowering mechanisms are conserved between these two legume crops.

E1 has the largest effect on flowering in soybean, encoding a transcription factor that acts as a floral repressor. It was reported that E1 homologues in other legumes differ in their functions in flowering, e.g., in Phaseolus vulgaris (Indigoferoid/Millettioid clade) it inhibits flowering (similar to E1), but in Medicago truncatula (Hologalegina clade) it does not affect flowering [6, 31]. Zhang et al. [31] also reported that the ectopic expression of PvE1L, an E1 homolog from common bean (Phaseolus vulgaris), delayed the onset of flowering in soybean. By contrast, the ectopic expression of MtE1L, an E1 homolog from Medicago truncatula, did not affect the flowering of soybean. Thus, functional diversification of legume E1 family genes in the regulation of flowering may be associated with lineage specification and genomic duplication [31].

No different expression of E1 gene was reported for guar plants grown under short- and long day conditions, although two more B3 transcription factors were reported to be up-regulated in guar under LD, suggesting that they repressed flowering [41]. Since two rounds of whole genome duplication events are hypothesized for soybean, paralogues are described for E loci in soybean genome. Two paralogs of E1 gene, namely E1La (Glyma.04G156400.1) and E1Lb (Glyma.04G143300.1) were found in soybean to control the onset of flowering [9]. No similar sequences for both of the genes were detected in our guar transcriptome assembly. However, their absence could also be due to untimely sampling of plant tissue, coinciding with a very low level of expression of this circadian clock gene.

E2 has been identified as an orthologue of the Arabidopsis GIGANTEA (GI) gene [10]. For the E2 locus, three GIhomologs (GmGI1, GmGI2, and GmGI3) were discovered in the soybean genome, located on chromosomes 20, 9 and 10 respectively. Among them GmGI3, described earlier as E2 locus (Glyma10g36600.1) [10], is considered to be a more ancestral gene than GmGI1 (Glyma20g30980.1) and GmGI2 (Glyma09g07240.1) [42]. In that study alternative splicing was also reported for the 11th exon of GmGI1, resulting in two different versions—the longer GmGI1a and the shorter GmGI1$\beta$. Likewise, our analysis found two isoforms of the CtE2 gene (CtE2 i1 and CtE2 i2) that are perfectly aligned to each other, except for a deletion in the middle of CtE2 i2 isoform spanning 269 bp. The longer isoform CtE2 i1 was hypothesized to be an Intron Retaining Isoform (IRI), containing an intron between 11th and 12th exons. IRIs can be generated through the special alternative splicing mode in which introns, instead of being excised as usual, are retained in mature mRNAs [43]. It was reported that in most cases, IRIs are destroyed by nonsense-mediated decay (NMD), since retained introns often contain premature termination codons, although a certain proportion of IRIs escape from the NMD pathway. Alternatively, we may assume that among the 18 plants subjected to transcriptome profiling, a nucleotide polymorphism in the CtE2 gene was encountered, since some of these plants had a truncated version of the gene.

E3 (GmPHYA3) and E4 (GmPHYA2) repress soybean flowering by responding to high and low red:far-red ratios in the light spectrum under LDs [44, 45]. The PHYA genes provide light perception and, therefore, stand in the initial point of the molecular network regulating flowering in soybean under LD conditions. Consequently, dysfunctional recessive alleles for the E3 and E4 loci are the most common mechanism underlying reduced photoperiod sensitivity in soybean [6]. Thus, the predicted intron-exon structure of their homologues in the guar genome offers clues for their targeted mutagenesis, e.g., using CRISPR-Cas method.

Although guar is an emerging economically important legume with a wide range of industrial applications [46], the productivity of this high-valued industrial crop is poor due to unsatisfactory availability of early maturing and high-yielding varieties [47]. The possibilities of marker-assisted selection of guar are still limited. Priti et al. [48] first developed a linkage map using microsatellite markers to genotype F3–F4 segregating populations from an interspecific cross between C. tetragonoloba cv HG563 and C. serrata. In this interspecific cross, C. serrata was highly resistant to bacterial leaf blight (BLB), and HG563 was a cultivar with high gum content. The linkage map was employed in order to identify quantitative trait loci (QTLs) for BLB resistance and gum content. Additionally, a genome-wide association study (GWAS) was conducted for guar, resulting in the discovery of six single nucleotide polymorphisms (SNPs) linked to the variation of an important agronomic trait—the percentage of pods matured to the harvest date under long light day conditions [49]. With very limited reports about linkage mapping and identification of quantitative trait loci associated with agrobiological traits in guar, the reverse genetics approach may facilitate marker-assisted breeding of guar.

Recent advances in generating mutants of E1-E4 genes using CRISPR-Cas technology have shown great potential to improve the performance of soybean varieties under different growing conditions [7, 40, 50]. Homozygous E1 mutants developed with CRISPR-Cas have shown significantly reduced onset of flowering time (by 20 days) under LD conditions [7]. CRISPR-Cas-induced e2 single mutant showed an early flowering phenotype, along with a distinctly increased yield performance relative to wild type in long photoperiod [50]. A core set of phytochrome mutants (including GmPHYA2/E4 and GmPHYA3/E3), with early-flowering phenotypes were obtained with CRISPR-Cas to characterize their functional mechanisms in regulating flowering of soybean under LD conditions [40]. Considering the established high level of homology between the E1-E4 genes of soybean and guar, one can expect to obtain similar CRISPR-Cas-engineered guar mutants with a shortened flowering onset, which will allow it to be cultivated in high latitudes.

In summary, our study provides the first accessible version of guar transcriptome assembly containing sequences of 96,447 clustered transcript isoforms detected in leaves of guar plants grown under the long day conditions. The developed transcriptome assembly was annotated against the Glycine max genome and used as a tool to identify and characterize in the guar genome orthologs of the E1-E4 soybean maturity genes. A high level of identity was detected between peptide sequences encoding by orthologous genes E1 and CtE1 (80%), E2 and CtE2 (93%), E3 and CtE3 (83%), E4 and CtE4 (91%). Moreover, not only the sequences but also the intron-exon structure of the genes in soybean and guar were found to be similar, suggesting that the genetic pathways underlying basic flowering mechanisms are conserved between these two legume crops. Consequently, the predicted intron-exon structure of the CtE1-CtE4 guar genes may offer clues for their targeted mutagenesis, e.g., using CRISPR-Cas method.

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Guar transcriptome assembly contig set is available from the European Nucleotide Archive database (Project: PRJEB78478, TSA Sequence Set: HCEE01000000.1).

Conceptualization, EP and LCD; methodology, LCD and AZ; software, LCD and AZ; validation, LCD and AZ; formal analysis, EP; investigation, LCD and AZ; data curation, AZ; writing—original draft preparation, LCD, AZ and EP. All authors have read and agreed to the published version of the manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Not applicable.

The research was funded by the Russian Science Foundation (project no. 24-26-00073).

The authors declare no conflicts of interest.

Supplementary Table 1: Guar transcriptome assembly contig set containing sequences of 96,447 clustered transcript isoforms detected in leaves of guar plants grown under the long day conditions; Supplementary Table 2: Annotation of the guar transcriptome assembly using blastn and blastx against the Glycine max genome. Supplementary Table 3: Evaluation of gene expression differences between early and late flowering guar plants. Supplementary Table 4: Exon-intron structure predictions of guar CtE1-CtE4 genes homologues to soybean E1-E4 maturity loci.

Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.31083/FBS26548.