Identification and characterization of LIM family genes in Tartary buckwheat: insights into functional divergence and abiotic stress responses

Abstract

Background

LIM domain proteins have been reported to participate in cytoskeleton dynamics, transcriptional regulation, and stress-related processes in plants. However, the LIM gene family has not been systematically characterized in Tartary buckwheat (Fagopyrum tataricum), and its potential roles in stress responses remain unclear.

Results

In the present study, 16 LIM genes were identified from the Tartary buckwheat genome and grouped into two subfamilies, CRP-like and DA1&DAR. Comparative analyses showed clear differences between the two subfamilies in gene structure, conserved motifs, and predicted protein architectures, implying functional diversification. Only one segmental duplication event (FtPLIM2a/FtPLIM2b) was detected, and the low Ka/Ks value indicated that this gene pair has undergone purifying selection. Synteny analysis showed that FtLIM genes displayed closer collinearity with dicot species, especially soybean, while no orthologous pairs were detected with the monocots included in this study. Most collinear gene pairs belonged to the CRP-like subfamily. Promoter analysis identified multiple cis-acting elements associated with growth and development, phytohormone responsiveness, and abiotic stress responses. Public transcriptome datasets revealed tissue-specificity expression patterns of FtLIM genes and distinct expression features between the two subfamilies. Several genes showed differential expression under salt, drought, and low-temperature conditions. Among them, FtLIM1a was consistently up-regulated across different stress treatments, while FtDAR4 was down-regulated under all three stress conditions. In the low-temperature dataset, FtWLIM2b showed altered expression under repeated cold treatment, indicating a potential association with cold adaptation. In addition, real-time fluorescence quantitative polymerase chain reaction analysis showed that most FtLIM genes responded to exogenous ABA and MeJA treatments.

Conclusions

This study provided a genome-wide characterization of the LIM gene family in Tartary buckwheat and identified several candidate genes potentially involved in abiotic stress responses. These results laid a foundation for future functional studies of FtLIM genes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-09184-4.

Background

Agricultural production worldwide is increasingly influenced by environmental constraints such as rising temperatures, soil salinity, and recurring drought events. These factors pose persistent challenges to maintaining yield stability and productivity [1]. In practice, crop yield depends not only on genetic potential but also on the capacity of plants to withstand adverse conditions, such as drought and salinity [2]. Consequently, there is growing interest in identifying gene families that participate in stress responses, as these genes provide important information for improving crop yield through molecular breeding. Among the protein families, the LIM domain proteins have attracted attention due to their biological functions. Previous studies suggest that LIM proteins are involved in both transcriptional regulation and cytoskeletal organization. LIM proteins can influence gene expression associated with developmental plasticity and environmental responses, and contribute to the organization and stability of actin filaments. This may enable plants to coordinate structural and transcriptional adjustments under stress conditions. Therefore, LIM proteins are increasingly considered potential regulators linking cellular architecture with adaptive responses, which may be relevant for improving stress resilience and yield stability in crops [3–5].

The LIM protein derives its name from the initials of three homeodomain proteins, namely LIN-11 from Caenorhabditis elegans [6], ISL-1 from rats [7], and MEC-3 from C. elegans [8]. LIM proteins possess a cysteine-rich zinc-binding domain, known as the LIM domain, which serves as a functional module for protein-protein interactions, and its consensus sequence is [C-X₂-C-X_16–23-H-X₂-C]-X₂-[C-X₂-C-X_16–21C-X_2–3-(C/D/H)] [3]. In plant species, LIM proteins can be classified into two distinct subfamilies according to the number of LIM domains, namely the CRP-like subfamily and the DA1&DAR subfamily, which contain two and one conserved LIM domains respectively [9]. The two LIM domains of the CRP-like subfamily LIM proteins are separated by an inter-LIM spacer. When compared with their animal orthologs, plant LIM proteins have a shorter C-terminal region and lack the glycine-rich domain. Plant CRP-like LIM domain proteins can be grouped into different categories. When classified by amino acid sequence similarity, they can be divided into four types: αLIM1, βLIM1, γLIM2, and δLIM2. Alternatively, when grouped according to their expression profiles, they are divided into two types: WLIMs (widely expressed across various tissues) and PLIMs (specifically expressed in pollen tubes) [10]. The LIM proteins of the DA1&DAR subfamily possess only one LIM domain, which may or may not be accompanied by the ubiquitin interaction motif (UIM) domain, along with a conserved C-terminal region [11].

Functionally, the CRP-like subfamily and the DA1&DAR subfamily exhibit significant divergence. CRP-like proteins have been extensively documented to directly bind and bundle F-actin, playing a crucial role in polarized pollen tube growth, cell elongation, and the maintenance of structural homeostasis under stress conditions [4]. Tobacco WLIM1 localizes in the nucleus and cytoplasm and binds F-actin with high affinity, thus defining a novel actin binding protein where the LIM domains themselves mediate F-actin binding and bundling. These properties serve as the basis for cytoskeletal remodeling during growth and stress [12]. In vitro, All LIM proteins of Arabidopsis thaliana are capable of bundling actin. However, the three pollen-specific PLIM proteins are inactivated under alkaline pH values, which differs from WLIM proteins, suggesting that this subfamily has specific sensitivity to the cytoplasmic conditions associated with the tip growth of pollen tubes [4, 13]. Genetic analysis further indicates that PLIM proteins are essential for pollen development and pollen tube elongation, which supports the role of CRP-like LIM proteins at the interface of mechanics, transport and signal transduction [14, 15]. Apart from cytoskeletal regulation, CRP-like members may also affect cell wall properties and stress resistance through the phenylpropanoid or lignin biosynthesis pathway. Tobacco NtLIM1 can bind to the Pal-box cis-acting element and activate reporters. Antisense inhibition of NtLIM1 results in a down-regulation of the transcription of key phenylpropanoid pathway genes, such as phenylalanine ammonialyase, hydroxycinnamate CoA ligase, cinnamyl alcohol dehydrogenase, and a reduction of lignin content by approximately 27%, implying a potential bridging role of CRP-like proteins in cell wall reinforcement, mechanical resistance, and stress adaptation [16]. Recent studies have demonstrated that the phosphorylation of WLIM2A plays a crucial role in plant immunity, suggesting that certain CRP-like members may serve as a bridge between stress signals and cytoskeleton remodeling [17]. Moreover, MaWLIM2 has been reported to act with MaZFP6 to directly repress MaRZF1 during the high-temperature response in banana, indicating that some LIM proteins may also participate in stress adaptation through transcriptional regulatory networks [18]. In contrast, DA1/DAR members regulate organ size and protein homeostasis through the ubiquitin system. Arabidopsis DA1 is a ubiquitin receptor containing a single LIM domain. The da1-1 mutant results in enlarged seeds and organs, indicating that DA1 plays a negative regulatory effect on restricting the cell proliferation period [19]. DA1 forms a critical axis with the E3 ubiquitin ligase DA2 and ENHANCER1 OF DA1 (EOD1)/BIG BROTHER to limit seed and organ size. DA1 physically interacts with DA2 and synergistically restricts cell proliferation in the maternal integuments of developing seeds [20]. Further studies have shown that DA1 regulates seed and organ size by affecting cell proliferation through modulating the stability of UBIQUITIN-SPECIFIC PROTEASE15 (UBP15) [21]. Members of the DA1/DAR subfamily, DAR1 and DAR2, share certain functional features with DA1. Previous studies have shown that these proteins are involved in regulating the duration of cell proliferation and the transition to endoreplication. In addition, they interact with multiple regulatory components, such as transcription factor TCP14 and TCP15 [22], and the deubiquitinating enzyme UBP12 and UBP13 [23]. Through these interactions, DA1/DAR proteins contribute to the control of leaf growth and developmental processes. Taken together, the two LIM subfamilies may operate in different but complementary ways: the CRP-like proteins are primarily linked to structural regulation, while the DA1/DAR subfamily proteins are more closely related to developmental control mediated by ubiquitin-dependent pathways, ultimately influencing organ size and growth patterns.

Tartary buckwheat (Fagopyrum tataricum) is a pseudocereal valued for both its nutritional properties and its ability to grow under relatively harsh conditions, such as poor soils and high-altitude environments. It is particularly known for its high flavonoid content that contributes to its health-promoting qualities [24]. In recent years, several gene families related to stress responses, including MYB transcription factors [25] and PYL kinases [26] gene families, have been investigated in Tartary buckwheat under diverse stress conditions. However, studies focusing on LIM domain proteins remain blank. The availability of a high-quality, chromosome-level genome assembly has made it possible to explore the LIM gene family in Tartary buckwheat [25]. In this study, a total of 16 LIM proteins were identified from the Tartary buckwheat genome, and the chromosomal distribution, phylogeny, conserved motifs, duplication patterns, and promoter cis-acting elements were systematically analyzed. In addition, expression patterns were examined across different tissues, and under various abiotic stress conditions. Co-expression relationships were also explored to provide clues about potential regulatory interactions. This study could provide a basis for future functional studies of LIM genes and contribute to the variety improvement and germplasm innovation of Tartary buckwheat.

Results

Identification and phylogenetic analysis of LIM proteins in Tartary buckwheat

In this study, a total of 16 LIM proteins were identified from the Tartary buckwheat genome (Table 1 and Table S1). All of these proteins contained one or two LIM domains. To explore the evolutionary relationships among FtLIM proteins, a phylogenetic tree was constructed, which included 40 LIM proteins from Tartary buckwheat (16), the model species Arabidopsis (14), and rice (10) (Fig. 1 and Table S2). These proteins could be initially classified into two subfamilies, the CRP-like subfamily and the DA1&DAR subfamily. The Tartary buckwheat LIM proteins belonging to the CRP-like subfamily were named according to their phylogenetic association, while those belonging to the DA1&DAR subfamily were named in their chromosomal order. In total, seven FtLIM proteins were clustered into the DA1&DAR subfamily, and the remaining nine proteins were clustered into the CRP-like subfamily. The CRP-like subfamily could be further divided into four subgroups, namely βLIM1, WLIM1, WLIM2 and PLIM2, which contained three, one, two and three FtLIM proteins, respectively.

Table 1.

Characteristics of LIM genes in Tartary buckwheat

Gene name	Gene ID	Chromosome	CDS length (bp)	Protein length (aa)	Mw (kDa)	pI	Instability index	GRAVY
FtLIM1a	FtPinG0001249600.01.T01	Ft3	666	221	24.84	8.40	39.57	-0.777
FtLIM1b	FtPinG0001183300.01.T01	Ft3	609	202	22.34	8.31	24.29	-0.434
FtLIM1c	FtPinG0005852800.01.T01	Ft5	597	198	21.98	6.97	28.16	-0.604
FtPLIM2a	FtPinG0005304800.01.T01	Ft6	663	220	24.92	6.12	37.95	-0.490
FtPLIM2b	FtPinG0004144300.01.T01	Ft7	654	217	24.18	6.00	39.58	-0.625
FtPLIM2c	FtPinG0006465400.01.T01	Ft2	594	197	22.23	8.62	40.11	-0.393
FtWLIM1	FtPinG0003186400.01.T01	Ft1	564	187	20.86	9.10	34.34	-0.682
FtWLIM2a	FtPinG0009328900.01.T01	Ft4	630	209	22.98	9.09	59.91	-0.390
FtWLIM2b	FtPinG0002888600.01.T01	Ft7	627	208	22.79	8.85	45.66	-0.333
FtDAR1	FtPinG0008310200.01.T01	Ft2	1560	519	58.67	8.33	63.42	-0.610
FtDAR2	FtPinG0000716000.01.T01	Ft2	1530	509	57.69	5.90	52.18	-0.593
FtDAR3	FtPinG0005946600.01.T01	Ft3	1458	485	54.79	5.70	56.89	-0.540
FtDAR4	FtPinG0008649200.01.T01	Ft4	1227	408	46.37	7.97	60.15	-0.521
FtDAR5	FtPinG0002733600.01.T01	Ft7	1425	474	53.71	6.29	54.15	-0.480
FtDAR6	FtPinG0002149100.01.T01	Ft8	1404	467	52.79	5.38	54.44	-0.416
FtDAR7	FtPinG0008245700.01.T01	Ft8	1485	494	55.87	5.48	52.84	-0.534

CDS Coding sequence, bp base pair, aa amino acid, Mw Molecular weight, pI isoelectric point, GRAVY Grand average of hydropathicity

Fig. 1.

Phylogenetic tree of LIM proteins from Tartary buckwheat, Arabidopsis, and rice. The name of each LIM protein commences with a species acronym: Ft, Fagopyrum tataricum; At, Arabidopsis thaliana; Os, Oryza sativa

The physicochemical properties of FtLIM proteins belonging to the two subfamilies exhibited certain differences (Table 1). Regarding the FtLIM proteins of the CRP-like subfamily, their molecular weights (Mw) were comparatively small, ranging from 20.86 to 24.92 kDa, with an average of 23.01 kDa. The protein length was in the range of 188 to 222 amino acids (aa). The majority of the proteins were stable (except for FtPLIM2c, FtWLIM2a, and FtWLIM2b). In contrast, the Mw of FtLIM proteins in the DA1&DAR subfamily ranged from 46.37 to 58.67 kDa, with an average of 54.27 kDa. The protein length varied from 409 to 520 aa. All proteins within this subfamily were highly unstable, with an instability index ranging from 52.18 to 63.42. The isoelectric point (pI) of FtLIM proteins ranged from 5.38 (FtDAR6) to 9.1 (FtWLIM1), with an average of 7.28. The predicted grand average of hydropathicity (GRAVY) indicated that all FtLIM proteins were hydrophilic.

Analysis of the gene structure and motif composition of FtLIMs

A phylogenetic tree containing only FtLIM proteins was constructed (Fig. 2a), which exhibited similar topological structure to Fig. 1. Notably, the FtLIM proteins belonging to the CRP-like subfamily and the DA1&DAR subfamily differed greatly in terms of gene structure (Fig. 2b) and motif composition (Fig. 2c, d and Table S3). All CRP-like proteins possessed two LIM domains, four to six exons, and three to five introns. Among these, the introns of FtPLIM2c were relatively long. In contrast, the DA1&DAR subfamily proteins had only one LIM domain and one UIM domain (except FtDAR1 and FtDAR4). There were numerous exons and introns, with 11 or 12 exons and 10 or 11 introns. Regarding motif composition, Motif 1 was shared by all FtLIM proteins. Apart from Motif 1, the CRP-like subfamily proteins also possessed two unique motifs (Motif 2 and Motif 8). The motif composition of all proteins within the CRP-like subfamily was identical. However, it varied for the DA1&DAR subfamily proteins, which had a large number of motifs, including Motif 1, 3, 4, 5, 7 and 9. Most proteins had Motif 6 and Motif 10, while FtDAR4 lacked both, and FtDAR1 lacked Motif 10. The results indicated that the gene structure and motif composition of proteins belonging to the same subfamily were similar.

Fig. 2.

Phylogenetic tree, gene structure and motif composition analysis of FtLIMs. a Phylogenetic tree of the LIM proteins from Tartary buckwheat. b Exon-intron structure of FtLIM genes. Exons and domains are represented by right-angled boxes, while introns are represented by dark-gray lines. c Motif composition of the FtLIM proteins. Different motifs are represented by rounded rectangles in different colors. d Amino acid sequences of different motifs

Analysis of the secondary and tertiary structure of FtLIMs

The secondary and tertiary structures of FtLIM proteins were predicted, as shown in Fig. 3. The secondary structures were mainly composed of α-helix, β-turn, extended strand, and random coil. However, the proportions of these four components varied significantly between the two subfamilies (Fig. 3a). For the DA1&DAR subfamily proteins, the combined proportion of random coil and α-helix was nearly 90%. The extended strand and β-turn accounted for 9.4% and 3.7%, respectively. In contrast, the α-helix of the CRP-like subfamily proteins accounted for a relatively low proportion (21.9%), while the extended strand and β-turn accounted for 15.7% and 8.9%, respectively. The tertiary structure of proteins within the same subfamily or subgroup was generally similar, and the tertiary structure of DA1&DAR subfamily proteins was more intricate (Fig. 3b and Fig. S1), suggesting that FtLIM genes had undergone functional differentiation during the evolutionary process.

Fig. 3.

Prediction of the secondary and tertiary structures of the FtLIM proteins. a Pie chart of the secondary structure composition of the CRP-like (top) and DA1&DAR (bottom) subfamily FtLIM proteins. b Representative tertiary structure diagrams of the FtLIM proteins in different subgroups

Chromosomal distribution and synteny analysis of LIM genes in Tartary buckwheat

The 16 LIM genes identified in this study were distributed across eight chromosomes of Tartary buckwheat (Fig. S2). Among them, Ft2, Ft3, and Ft7 each harbored three genes, Ft4 and Ft8 contained two genes, and the remaining chromosomes contained only one gene. The distribution of CRP-like subfamily genes on the chromosomes was uniform, while DA1&DAR subfamily genes were concentrated on chromosomes Ft2, Ft3, Ft4, Ft7, and Ft8.

Gene duplication plays an important role in the amplification process of gene families. In the FtLIM gene family, only one segmental duplication event (FtPLIM2a and FtPLIM2b) was identified, and no tandem duplication event was detected (Fig. 4a). The estimated Ka/Ks ratio (ω) of the FtPLIM2a/FtPLIM2b gene pairs was 0.11 (95% confidence interval (CI): 0.0445–0.1728), which was significantly lower than 1 (likelihood ratio test (LRT), P = 2.60 × 10^− 25), indicating that this duplicated gene pair had mainly evolved under purifying selection. Furthermore, we analyzed the collinearity of the LIM gene family between Tartary buckwheat and eight other species (four dicots and four monocots) (Fig. 4b, Fig. S3 and Table S4). The collinearity analysis revealed that there were no homologous gene pairs in the LIM gene family between Tartary buckwheat and four monocotyledon plants. The number of homologous gene pairs between Tartary buckwheat and Glycine max was the highest, with eight pairs, followed by Arabidopsis thaliana (three pairs), Solanum lycopersicum (two pairs), and Beta vulgaris (two pairs). It was worth noting that homologous gene pairs mainly encompass genes in the CRP-like subfamily, particularly the βLIM1 subgroup (FtLIM1a, FtLIM1b, and FtLIM1c), and another gene, FtPLIM2a, from PLIM2 subgroup. The DA1&DAR subfamily involved only one gene, FtDAR3, which formed homologous gene pairs with G. max and B. vulgaris, respectively. These results suggested that the evolution of Tartary buckwheat LIM gene family was relatively conservative, and CRP-like genes may play important roles in this process.

Fig. 4.

Gene duplication and synteny analysis of the FtLIM genes. a Gene duplication analysis, with red line indicating segmental duplication event. b Synteny analysis of the LIM genes between Tartary buckwheat and other species. Ft, Fagopyrum tataricum; At, Arabidopsis thaliana; Os, Oryza sativa; Gm, Glycine max; Sl, Solanum lycopersicum; Bv, Beta vulgaris

Analysis of the cis-acting elements and expression patterns of LIM genes in Tartary buckwheat

Abundant cis-acting elements in the promoter region of Tartary buckwheat LIM gene members were detected. A total of 2,356 cis-elements were identified, which were mainly associated with development, stress-response, and hormone-response (Fig. 5 and Table S5). Therefore, the tissue expression characteristics (Fig. 6a and Table S6) of the Tartary buckwheat LIM genes and their response to different abiotic stresses (Fig. 6b and Table S7) were investigated based on publicly available transcriptome datasets for salt, drought, and low-temperature treatments.

Fig. 5.

Distribution of cis-acting elements in the promoter region of the FtLIM genes. a Phylogenetic tree of the LIM proteins from Tartary buckwheat. b Distribution of cis-acting elements within the 2 kb promoter region of the FtLIM genes. Different element types are indicated by distinct colors as presented in the legend

Fig. 6.

Heatmap of FtLIM genes expression profiles across different tissues and under various abiotic stress conditions. a Expression patterns of FtLIM genes in different tissues and at different stages of grain development based on publicly available transcriptome datasets. The color scale represented expression level, with lighter colors indicating lower expression and redder colors indicating higher expression. b Expression patterns of FtLIM genes under salinity, drought, and low-temperature stress based on publicly available RNA-seq datasets. Expression values were normalized as log₂ (FPKM + 1) for visualization. In the heatmap, blue indicated relatively low expression, the central light-gray color corresponded to a normalized value of 0, and red indicated relatively high expression. The colored bar above the heatmap denoted stress type: blue, salinity; red, drought; yellow, low temperature. “Control” and “Treatment” represented the control and salt-treated samples, respectively; “PEG_0h”, “PEG_1h”, “PEG_3h”, and “PEG_6h” represented drought treatment time points; “Control”, “Memory”, and “Shock” represented the low-temperature treatment groups. The asterisk (*) denoted a significant difference compared with the corresponding control at P < 0.05

In general, the expression levels of the DA1&DAR subfamily genes were higher than those of genes in the CRP-like subfamily. Notably, four members (FtPLIM2a, FtPLIM2b, FtPLIM2c, and FtWLIM2a) of the CRP-like subfamily were not expressed in any investigated tissues of Tartary buckwheat. The expression of Tartary buckwheat LIM genes exhibited tissue specificity. For example, the expression of FtLIM1a and FtDAR5 in the roots was relatively high; FtWLIM1 and FtDAR1 were highly expressed in the stems; FtDAR2 was highly expressed in the leaves; FtLIM1b, FtLIM1c, and FtDAR6 were specifically expressed in the flowers; FtDAR4 was highly expressed in the stems and leaves, while FtWLIM2b was highly expressed in the leaves and grains at the S1 stage; FtDAR3 and FtDAR7 were highly expressed in all tissues, and their expression levels gradually decreased as the grains developed.

A total of seven genes showed significant changes in expression under salt stress. Among them, except for FtLIM1a, which was up-regulated, all other genes were down-regulated. Notably, the expression level of FtLIM1a was also up-regulated under both drought and low-temperature stress. In contrast, FtDAR4 was significantly down-regulated under all stress conditions. FtDAR7 was sensitive to low temperature, and its expression was significantly down-regulated, while FtDAR3 was up-regulated at low temperature. Gene expression usually undergoes changes at different stress durations. For example, the expression level of FtWLIM2b first decreased and then increased under continuous drought stress treatment. The expression level of FtDAR7 also showed phased changes. Additionally, the results indicated that the expression level of FtWLIM2b was significantly down-regulated when suddenly exposed to low-temperature stress (shock). However, after the plants developed memory through repeated low-temperature training, the expression level did not change significantly under low-temperature stress.

It was worth mentioning that cis-acting elements associated with abscisic acid (ABA) and methyl jasmonate (MeJA) were present in almost all LIM gene members. Therefore, exogenous ABA and MeJA treatments were conducted on Tartary buckwheat seedlings to verify whether the relevant genes would respond. The results indicated that the majority of LIM genes were sensitive to ABA and MeJA treatments. The expression levels of nine genes were statistically analyzed, and except for FtDAR2, the other genes showed significant responses to at least one hormone (Fig. 7 and Table S8). Regarding ABA treatment, there were significant changes in the expression levels of six genes, with three genes significantly up-regulated (FtWLIM1, FtDAR4, and FtDAR7), and three genes significantly down-regulated (FtLIM1a, FtWLIM2b, and FtDAR5). In contrast, under MeJA treatment, except for FtDAR2, the expression levels of the other eight genes showed significant changes, with five genes significantly down-regulated (FtLIM1a, FtWLIM2b, FtDAR3, FtDAR5, and FtDAR7), and three genes significantly up-regulated (FtWLIM1, FtDAR1, and FtDAR4).

Fig. 7.

Expression profiles of FtLIM genes under different exogenous hormones treatments. a–i represent the changes in the expression levels of different genes following exogenous hormone treatments. ABA, abscisic acid; MeJA, methyl jasmonate. The * and ** denote significance at the 0.05 and 0.01 levels, respectively

GO enrichment and interactive network analysis of FtLIMs

The GO enrichment analysis showed that Tartary buckwheat LIM genes could be significantly enriched into 14 GO terms, which belong to the biological process, cellular component, and molecular function categories (Fig. 8 and Table S9). Specifically, the GO terms enriched in the biological process were the most numerous, with 9 GO terms, including the developmental process (GO:0032502), the reproductive process (GO:0022414), and others. There were four GO terms enriched in the cellular component. Only one GO term, binding (GO:0005488), was enriched into the molecular function, yet the number of genes was the largest, with 14 genes.

Fig. 8.

GO enrichment analysis of FtLIM genes

Interactive network analysis revealed that FtWLIM1, FtWLIM2a, and FtWLIM2b might interact with FtPinG0008499500.01, a protein of unknown function. The proteins FtDAR2, FtDAR3, FtDAR5, FtDAR6, and FtDAR7 of the DA1&DAR subfamily may interact with the four proteins FtPinG0000998800.01, FtPinG0001692500.01, FtPinG0001692700.01, and FtPinG0006380500.01 (Fig. 9). According to the genome annotation of Tartary buckwheat, FtPinG0000998800.01, FtPinG0001692700.01, and FtPinG0006380500.01 all encode cytochrome c oxidase (complex IV), which was a key enzyme in mitochondrial energy metabolism. The results suggested that LIM genes were widely involved in the important physiological metabolism of Tartary buckwheat.

Fig. 9.

Prediction of protein–protein interactive networks of FtLIM proteins

Discussion

LIM domain proteins are widely conserved in plants and have been reported to participate in multiple biological processes, including cytoskeleton organization, gene regulation, and development. Recently, their potential involvement in abiotic stress responses has also attracted increasing attention [5]. In this study, 16 LIM genes were identified in Tartary buckwheat and grouped into two subfamilies, CRP-like and DA1&DAR (Fig. 1). Obvious differences were observed between two subfamilies with respect to gene structure (Fig. 2b), conserved motifs (Fig. 2c), and predicted protein architectures (Fig. 3 and Fig. S1). Similar patterns had been described in other plant species such as Arabidopsis, where CRP-like proteins were mainly associated with actin-binding, while DA1&DAR proteins were related to the ubiquitin pathways controlling organ growth [4, 27]. When compared with other dicot species, the number of LIM genes in Tartary buckwheat was similar to that in Solanum lycopersicum (15 genes) [11], Phaseolus vulgaris (17 genes) [28] and Lactuca sativa (14 genes) [29]. However, the LIM gene number in Tartary buckwheat was lower than that reported in certain leguminous crops, such as Medicago sativa, where gene family expansion of this gene family had been observed [30]. This difference may reflect distinct evolutionary progresses among species, and in Tartary buckwheat the LIM gene family seemed to have remained relatively stable without obvious expansion.

Gene duplication is generally considered an important mechanism contributing to the expansion of gene families [31]. However, in Tartary buckwheat, evidence for such expansion appears limited, as only one segmental duplication event (FtPLIM2a/FtPLIM2b) was identified (Fig. 4a). The calculated Ka/Ks value was relatively low, which implied that FtLIM genes may have retained conserved functions. Synteny analysis showed that the LIM genes in Tartary buckwheat share more homologous relationships with dicot species than with monocots (Fig. 4b and Fig. S3). The result was consistent with the findings of previous research [32, 33]. This phenomenon may reflect the divergence that occurred following the separation of monocots and dicots during evolution. In addition, most of collinear gene pairs belonged to the CRP-like subfamily, while only one gene from the DA1&DAR subfamily was involved (Table S4). The result suggested that CRP-like genes remained relatively stable during evolution, while DA1&DAR members might have undergone functional diversification.

Expression profiling revealed distinct subfamily divergence and tissue specificity among FtLIM genes. In general, DA1&DAR genes exhibited higher expression levels compared to CRP-like genes, while several CRP-like members showed extremely low or undetectable expression in the examined tissues and transcriptome datasets. Rather than indicating complete lack of expression, this pattern may suggest transcriptional specialization, conditional activation under developmental stages or environmental conditions not sampled in the present study, or transcript abundance below the detection threshold of the available RNA-seq data [34]. For example, TaLIM2 had been reported to be specifically expressed in fertile anthers in wheat [9], whereas the Tartary buckwheat tissue expression dataset used in this study did not include anther samples. It is also possible that some of these genes may represent pseudogenes or genes undergoing functional degeneration, although this possibility cannot be determined from the current data alone. Tissue-specific expression patterns indicated that FtLIM genes may participate in diverse developmental processes, including the development of roots, stems, and flowers (Fig. 6a). Similar tissue-preferential expression patterns had been reported in Arabidopsis, where LIM proteins were involved in cytoskeletal organization and organ-specific development [4]. Interestingly, some FtLIM genes, such as FtDAR3 and FtDAR7, showed broad expression across tissues and gradual down-regulation during grain development, implying potential roles in early developmental stages.

The promoter analysis indicated that the FtLIM genes contain abundant cis-acting elements related to development, hormone signaling, and stress responses (Fig. 5), suggesting that these genes may participate in complex regulatory networks. In particular, elements responsive to ABA and MeJA were widely distributed across the promoters, implying their potential involvement in hormone-mediated stress signaling pathways. Similar regulatory complexity had been reported in other plant species, where LIM genes were transcriptionally regulated in response to environmental stimuli and developmental cues [9, 10, 29]. Therefore, investigating their expression patterns under abiotic stress conditions may provide useful clues to the possible roles of FtLIM genes in stress adaptation. Among all members, FtLIM1a showed consistently up-regulated under salt, drought, and low-temperature stress in the analyzed transcriptome datasets (Fig. 6b), suggesting that it may be associated with responses to multiple abiotic stresses. FtLIM1a was homologous to Arabidopsis WLIM1 (Fig. 1), which had been implicated in actin cytoskeleton organization [4]. Based on this homology, FtLIM1a may contribute to the maintaining of cell structure and stability under stress conditions, although this possibility requires further experimental validation. In contrast, FtDAR4 was consistently down-regulated under the examined stress treatments (Fig. 6b), which may reflect a potential growth-defense trade-off mechanism [35], given that DA1&DAR family proteins were known to regulate organ size and growth [27]. In addition, the distinct responses of FtDAR3 and FtDAR7 under low-temperature treatment (Fig. 6b) may indicate functional divergence within the DA1&DAR subfamily. Interestingly, FtWLIM2b showed different expression patterns under cold-shock and repeated low-temperature treatments (Fig. 6b), suggesting that it may be associated with responses to repeated cold exposure or cold acclimation. However, this inference is based only on transcriptome data, and its precise role requires further validation through controlled cold-priming experiments and functional analyses. Stress memory had been recognized as an important adaptive mechanism that enhanced plant tolerance to recurring environmental stresses and represented a promising target for crop improvement [36]. Based on its expression pattern, FtWLIM2b may serve as a candidate gene for future functional studies in Tartary buckwheat. However, the stress-responsive expression patterns discussed above were inferred mainly from publicly available transcriptome datasets and were not independently validated by real-time fluorescence quantitative polymerase chain reaction (qRT-PCR) or other experimental approaches in this study. Thus, these findings should be regarded as preliminary, and further experimental verification will be necessary to confirm the expression profiles and biological functions of representative FtLIM genes under different stress conditions. Hormone treatment experiments showed that most FtLIM genes responded significantly to ABA and MeJA, particularly to MeJA (Fig. 7), which may suggest an involvement of jasmonate signaling in modulating FtLIM gene expression. Jasmonate pathways had been reported to participate in plant stress responses, especially in coordinating rapid transcriptional adjustments under adverse conditions [37].

Protein interaction analysis indicated that several members of the DA1&DAR group may have predicted associations with cytochrome c oxidase (complex IV) subunits (Fig. 9), a key component of the mitochondrial electron transport chain [38]. The result implied that FtLIM proteins may be linked to mitochondrial function and energy metabolism. However, these associations were inferred solely from orthology-based and in silico analyses and do not provide direct evidence of physical interaction. It is well known that abiotic stress can disturb redox balance and energy metabolism in plant cells [39], the potential association between DA1&DAR proteins and mitochondrial components may point to a possible connection between growth regulation and metabolic adjustment under stress conditions. However, this possibility remains speculative and requires further experimental validation.

Although LIM gene families had been characterized in several plant species, this study provided the first comprehensive analysis of the LIM family in Tartary buckwheat. More importantly, by integrating gene structure, promoter composition, stress-responsive expression patterns, hormone responses, and predicted protein associations, this work identified candidate FtLIM genes potentially involved in abiotic stress adaptation and provided a foundation for future functional studies in Tartary buckwheat.

Conclusions

In the present study, a total of 16 LIM genes were identified within the Tartary buckwheat genome and subsequently classified into the CRP-like and DA1&DAR subfamilies, which exhibited distinct structural and evolutionary divergence. Limited gene duplication and strong purifying selection indicated a high degree of conservation within the Tartary buckwheat LIM gene family. Expression and promoter analyses revealed that the FtLIM genes were involved in tissue development and responded to multiple abiotic stresses and hormone signals. Several genes exhibited consistent and dynamic stress responsiveness. Overall, this study provided important candidate genes and a foundation for improving abiotic stress tolerance in molecular breeding of Tartary buckwheat.

Materials and methods

Identification of the LIM gene family in Tartary buckwheat

The genome sequences and annotation files (V2) of Tartary buckwheat were retrieved from the MBKBASE database (https://mbkbase.org/Pinku1/). Based on the Hidden Markov Model (HMM) PF00412 acquired from InterPro (https://www.ebi.ac.uk/interpro/), potential Tartary Buckwheat LIM proteins were searched by using the HMMER (v3.3.2) software with an E-value of less than 0.01. Meanwhile, the LIM protein sequences of Arabidopsis and rice were obtained from TAIR (https://www.arabidopsis.org/) and iTAK (http://itak.feilab.net/cgi-bin/itak/index.cgi) respectively to query the potential LIM proteins in the Tartary Buckwheat genome by using the BLASTp program with an E-value of less than 1.0 × 10^− 5. Subsequently, all the screened putative LIM protein sequences were submitted to the InterPro, CDD (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi), and SMART (https://smart.embl-heidelberg.de/) databases to verify the conserved LIM domains. The physicochemical properties, pI, Mw, instability index, and GRAVY of FtLIM proteins were computed using the ProtParam tool (https://web.expasy.org/protparam/).

Phylogenetic analysis of FtLIMs

A total of 40 LIM proteins from Arabidopsis (14 LIMs), rice (10 LIMs), and Tartary buckwheat (16 LIMs) were used to construct a phylogenetic tree. Firstly, multiple sequence alignment was performed using the MUSCLE program, and then an evolutionary tree was constructed using the maximum likelihood method with the MEGA-X software (v10.0.4). The JTT + G model was selected, as it was determined to be the best-fit model through calculation. The bootstrap value was set at 1000. The phylogenetic tree was visualized by the iTOL online tool (https://itol.embl.de/). Tartary buckwheat LIM proteins were named according to the chromosome location information and the phylogenetic relationship with LIM proteins from Arabidopsis and rice.

Gene structure analysis and protein structure prediction

Leveraging the genome sequence and general feature format (GFF) files of Tartary buckwheat retrieved from the MBKBASE database, the exon and intron structures of FtLIM genes were analyzed and visualized using TBtools software (v2.086). The motif composition of the amino acid sequences of FtLIMs was analyzed by the MEME online tool (v5.5.5; https://meme-suite.org/meme/tools/meme) and visualized with TBtools. All parameters were set to their default values, with the exception that the maximum number of motifs was set to 10. The promoter sequences (2 kb upstream of CDS) of FtLIM genes were extracted and submitted to PlantCARE (https://bioinformatics.psb.ugent.be/webtools/plantcare/html/) for cis-acting elements analysis. Additionally, in order to investigate the structural differences of FtLIMs proteins among different subgroups, the amino acid sequences of FtLIM proteins were submitted to SOPMA (https://npsa-prabi.ibcp.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_sopma.html) and SWISS-MODEL (https://swissmodel.expasy.org/) for the prediction of protein secondary and tertiary structure, respectively.

Chromosomal distribution and synteny analysis of FtLIMs

The location information of FtLIMs was retrieved from the Tartary buckwheat genome GFF file and visualized with TBtools. Multiple collinear scanning toolkits (MCScanX) were employed to examine whether gene duplication events occurred among FtLIM genes. The genome files of Arabidopsis thaliana, Glycine max, Solanum lycopersicum, Beta vulgaris, Oryza sativa, Zea mays, Sorghum bicolor, and Setaria italica were downloaded from the Ensembl Plants database, and the collinearity relationships between the Tartary Buckwheat LIM family and these eight species were analyzed using the One-Step Mcscanx plugin in TBtools. Ka, Ks, and Ka/Ks values were estimated using codeml in the PAML package based on codon alignments. The significance of deviation from neutrality (Ka/Ks = 1) was assessed using LRT method by comparing a model with free ω to a null model with ω fixed at 1. The 95% CI of Ka/Ks was estimated by bootstrap resampling of codon sites.

Plant materials and hormone treatments

The Tartary buckwheat variety Xinong-9940 used in the experiment was provided by the Minor Grain Crops Research Centre of Northwest A&F University. Tartary buckwheat seedlings were cultivated in the greenhouse of Northwest A&F University at a temperature of 25℃. When the seedlings reached 21 days of growth, they were subjected to treatments with a 100 µM solution of ABA and MeJA by foliar spray, respectively, while water treatment served as a control. The second leaves (from top to bottom) of the seedlings under different treatments were harvested after 6 h [40], immediately frozen with liquid nitrogen, and stored at -80℃.

Investigation of expression patterns of FtLIM genes

The tissue expression characteristics of FtLIM genes and their expression profiles under salt, drought, and low-temperature stresses were analyzed using publicly available transcriptome datasets, whereas the responses to exogenous ABA and MeJA treatments were experimentally validated by qRT-PCR in this study.

Transcriptome sequencing datasets PRJNA854877 [41] and GSE111937 [42] were retrieved from the National Center for Biotechnology Information (NCBI) to investigate the expression patterns of FtLIM genes in different tissues and at different grain developmental stages, respectively. In addition, three publicly available stress-related RNA-seq datasets were used to analyze the expression responses of FtLIM genes to abiotic stresses in Tartary buckwheat. For salt stress, PRJNA528524 [43] was generated from 10-day-old seedlings of Tartary buckwheat cultivar ‘Yu Qiao 2’ treated with 100 mM NaCl. Seedlings were grown in soil under a 16 h light/8 h dark photoperiod at 25°C with approximately 50% relative humidity, and shoot tissues without roots were sampled after 4 days of treatment. Untreated seedlings grown under the same conditions were used as controls. For low-temperature stress, GSE138546 [44] included three treatments in Tartary buckwheat cultivar ‘Dingku 1’: control plants grown under normal conditions, cold-stock plants directly exposed to 0°C for 6 h, and cold-memory plants pretreated at 4°C for 6 h followed by 18 h recovery at room temperature for four cycles before a final exposure to 0°C for 6 h. For drought stress, CRA003335 [45] was derived from seedlings of Tartary buckwheat cultivar ‘Jinqiao No. 2’, in which drought stress was simulated by 20% PEG-6000 treatment. Samples were collected at 0, 1, 3, and 6 h after treatment, with the 0 h samples used as controls. These experimental details were summarized from the corresponding original publications. The calculated Fragments Per Kilobase of exon model per Million mapped fragments (FPKM) value was used to represent the expression level of genes in the sample and was normalized to a log₂ (FPKM + 1) value for visualization.

The qRT-PCR was used to detect the expression changes of FtLIMs following the application of exogenous hormones. Referring to the method described by Li et al. [32], RNA was extracted from leaves subjected to different treatments, and then the first-strand cDNA was synthesized and amplified as a template. FtH3 was selected as the internal reference gene, and the relative expression levels of FtLIM genes were calculated using the 2^−ΔΔCT method. The information of all primer sequences was listed in Table S10.

GO enrichment and prediction of protein interactive networks

GO enrichment analysis of FtLIMs was performed using the OmicShare online tool (https://www.omicshare.com/tools). The interaction network of FtLIM proteins was predicted by the STRING web server (https://cn.string-db.org/) based on the homologous relationship with Arabidopsis and visualized using Cytoscape software (v3.6.1).

Statistical analysis

The Student’s t-test was carried out using R software (version 4.4.1) to investigate whether there were significant alterations in the expression of the FtLIM genes subsequent to different stress treatments and exogenous hormone stimulation. Additionally, Origin 2026 software was utilized for image plotting.

Abbreviations

aa

amino acids

ABA

Abscisic acid

At

Arabidopsis thaliana

bp

base pair

Bv

Beta vulgaris

CDS

Coding sequence

CI

Confidence interval

FPKM

Fragments Per Kilobase of exon model per Million mapped fragments

Ft

Fagopyrum tataricum

GFF

General feature format

Gm

Glycine max

GRAVY

Grand average of hydropathicity

HMM

Hidden Markov Model

LRT

Likelihood ratio test

MeJA

Methyl jasmonate

Mw

Molecular weights

NCBI

National Center for Biotechnology Information

Os

Oryza sativa

pI

isoelectric point

qRT-PCR

Real-time fluorescence quantitative polymerase chain reaction

Sl

Solanum lycopersicum

Authors’ contributions

JL planned and designed the research, analyzed data, and wrote the original manuscript. YZ and YY performed the bioinformatics analysis. YZ prepared the figures. CW and HY performed the experiments. SL and BF reviewed and edited the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the Scientific Research Program Funded by Education Department of Shaanxi Provincial Government (25JK0744), Natural Science Basic Research Program of Shaanxi Province (2024JC-YBQN-0220), and Doctoral Support Project of Yan’an University (YDBK2022-82).

Data availability

All data generated or analyzed during this study are included in this published article and its supplementary information files.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.