Genome-wide identification and expression analysis of the CBF transcription factor family in Lolium perenne under abiotic stress

ABSTRACT

C-repeat binding factor (CBF) subfamily genes encoding transcriptional activators are members of the AP2/ERF superfamily. CBFs play important roles in plant tolerance to abiotic stress. In this study, we identified and analyzed the structure, phylogeny, conserved motifs, and expression profiles of 12 CBFs of the grass species Lolium perenne cultured under abiotic stress. The identified LpCBFs were grouped into three phylogenetic clades according to their protein structures and motif organizations. LpCBF expression was differentially induced by cold, heat, water deficit, salinity, and abscisic acid, among which cold treatment induced LpCBF gene expression significantly. Furthermore, association network analysis indicated that different proteins, including certain stress-related proteins, potentially interact with LpCBFs. Altogether, these findings will enhance our understanding of LpCBFs protein structure and function in the regulation of L. perenne stress responses. Our results will provide valuable information for further functional research of LpCBF proteins in L. perenne stress resistance.

KEYWORDS:

• LpCBF
• perennial ryegrass
• expression level
• abiotic stress
• cold tolerance

Introduction

C-repeat binding factor (CBF) proteins comprise a subfamily of the AP2 transcription factor superfamily. It has been established that these proteins play important roles in facilitating resistance to abiotic stress, particularly low-temperature stress, thus enhancing cold resistance.¹ For example, it has been demonstrated that CBF proteins can bind specifically to dehydration-responsive element/C-repeat cis-acting elements in downstream gene promoters,² thereby regulating the transcription of downstream COR (cold-regulated) genes in response to low temperatures.

Previously, CBF subfamily genes have been isolated, identified, and further studied in the model plant Arabidopsis.³ Subsequently, CBF genes homologous to those in Arabidopsis were identified in numerous species, including Solanum Lycopersicum,⁴ Oryza sativa L.,⁵ Prunus persica,⁶ Populus hopeiensis,⁷ and Kandelia candel,⁸ thereby confirming that these are common genes among plants. Furthermore, it has been demonstrated that CBF family genes are induced by a range of different abiotic stress conditions, with the characteristics and levels of expression differing in each case. For example, expression of an LlCBF genes from Lepidium latifolium was found to be enhanced in response to high salinity, dehydration, and low-temperature treatments.⁹ Similarly, among the six CBF subfamily genes from P. persica (PpCBF1-6), the expression of PpCBF1, −5, and −6 was observed to be induced by exposure to low temperature, whereas that of the other three PpCBF genes remained relatively constant.¹⁰

As key transcription factors that respond to several abiotic stress factors, CBF genes can also activate the expression of other stress responsive genes,¹¹ as indicated by RNA-Seq analysis of Arabidopsis mutants, which revealed that 134 genes are regulated by CBF genes.¹² Furthermore, CBFs reportedly play an important role in the regulation of low-temperature responses in higher plants.¹³ For example, CBFs have been shown to upregulate the expression of COR genes that promote increased accumulation of proline and soluble sugars, thereby enhancing plant resistance to low temperature.¹⁴ Similarly, Phalaenopsis aphrodite PaCBF1 was cold-induced, promoting the upregulated expression of cold-regulated genes COR6.6 and RD29a, thereby protecting plants from cold damage.¹⁵ Furthermore, Lee found that the expression of eight CBF/DREB1 genes in Brassica rapa increased during cold (4°C) treatment.¹⁶ Consistently, Barrero-Gil showed that overexpression of AtCBF3 increased tolerance to freezing stress in Arabidopsis,¹⁷ with cold resistance being enhanced by an accumulation of CBF transcripts in response to cold treatment duration. Similarly, Zarka found that the transcript levels of AtCBF1, AtCBF2, and AtCBF3 were enhanced and reached relatively high levels at approximately 3 h after transferring Arabidopsis plants from 20°C to 4°C.¹⁸ Additionally, enhancement of the transcriptional levels of SlCBF1 and SlCBF2 has been observed in tomato after exposing plants to a temperature of 10°C,¹⁹ and ectopic expression of Lolium perenne LpCBF3 has been found to enhance the cold resistance of transgenic Arabidopsis plants.²⁰ However, the function of LpCBFs needs to be further studied to enable enhancement of the increased reversibility of L. perenne plants.

The grass species L. perenne adapts to a wide range of soils and is used extensively in animal husbandry as livestock forage, and as a garden ground cover plant.²¹ Accordingly, as a good type of lawn and forage grass, genome-wide analysis of L. perenne CBF genes is of particular importance with respect to breeding of varieties resistant to abiotic stress. In this study, we examined 12 LpCBF genes obtained from the NCBI database. On the basis of BLAST analysis of multiple sequences, these 12 CBF genes were clustered into three groups. Additionally, we searched for gene sequences homologous to CBF in other species that may have similar evolutionary histories and functions, and subjected these to phylogenetic analysis. CBF expression analyses were conducted on seedlings exposed to a range of abiotic stress conditions, namely high temperature, low temperature, water deficit, salinity, and abscisic acid (ABA) stress. Our findings enabled us to establish the basic characteristics of the CBF subfamily in L. perenne and have practical applications in molecular breeding programs. Furthermore, our data provide a useful reference for further studies on CBF mechanisms of action and have practical applications in molecular breeding.

Materials and methods

Database search for CBF protein sequences

Sequences of LpCBF subfamily genes were retrieved from the NCBI online (https://www.ncbi.nlm.nih.gov/) with Lolium perenne (taxid: 4522) database, and based on the findings of previous studies.²² LcCBF sequences were further authenticated based on the conserved domains using CDD (http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) and SMART (http://smart.embl-heidelberg.de/). Arabidopsis subfamily AtCBF genes were retrieved from the TAIR database (https://www.arabidopsis.org/) with reference to the study by Hu et al.²³ The protein sequences of CBFs from other gramineous species were retrieved from the NCBI database, and with reference to previous studies, including those on Oryza sativa,²⁴ Triticum aestivum,²⁵ Secale cereale,²⁶ and Hordeum vulgare.²⁷ The structural characteristics and isoelectric point (pI), molecular weight (MW), and grand average of hydropathicity (GRAVY) values of LpCBF proteins were determined using the ExPASy ProtParam tool (http://web.expasy.org/protparam/), while the subcellular localization of LpCBF genes was predicted using UniProt (https://www.uniprot.org/) and WoLF PSORT (https://wolfpsort.hgc.jp/) online tools.

Sequence alignment and phylogenetic analysis

In order to examine the evolutionary relationship among CBF proteins, multiple amino acid sequences were aligned using Clustal 2.1 software with default settings. An unrooted phylogenetic tree was constructed using MEGA7.0, based on the neighbor-joining (NJ) method and a Poisson substitution model with 1000 bootstrap replications was calculated.

Structural analysis of LpCBF proteins

The motifs of assessed LpCBF proteins were identified using MEME online software (https://meme-suite.org/meme/tools/meme) with advanced default settings. Details of the top 10 predicted motifs were obtained from the MEME suite. The conserved domains of LpCBF proteins were predicted using the SMART web server tool (http://smart.embl-heidelberg.de/), and the three-dimensional structure of LpCBF proteins was predicted using the SWISS-MODEL online server (https://swissmodel.expasy.org/) with advanced default settings. A three-dimensional template for modeling LpCBF proteins was determined based on the order of GMQE.

Plant growth conditions and stress treatments

For the purposes of the present study, the Yatsyn ecotype of L. perenne was used as an experimental material. After sterilization with 1% sodium hypochlorite, selected fully mature seeds were planted in pots containing perlite and thereafter incubated in a growth chamber at 24°C and 65% relative humidity under a 16 h/8 h (day/night) photoperiod. Thirty-day-old seedlings were used in stress treatments. The seedlings were placed in an incubator at 4°C and 42°C for the low- and high-temperature treatments, respectively. For PEG treatment, we removed the original culture solution and then watered the solution with 15% PEG. For the salt treatment, the original culture solution was removed and the solution was watered with 250 mM NaCl. In turn, the water was replaced with a 100 μΜ ABA solution for the ABA treatment. Leaf-blade tissue samples were collected for analysis from each of the five treatments after 0, 3, 6, 12, and 24 h of exposure to the various stress conditions. In each treatment, 20 plants and five plant leaves were collected from different seedlings per treatment.

Gene expression analysis

Total RNA was extracted from the leaves of L. perenne seedlings using plant isolation kits (Sangon Biotech, Shanghai, China, Cat. #B518631). Complementary DNA (cDNA) was prepared from the isolated total RNA using a cDNA synthesis kit with random primers (Vazyme, Nanjing, China, Cat. #R312-02). The cDNA thus obtained was used as a template for quantitative real-time PCR (qPCR) analysis using a SYBR Green real-time PCR master mix (Vazyme, Nanjing, China, Cat. #Q711-02); qRT-PCR was conducted in a CFX96 real-time system (Bio-Rad, Hercules, CA, USA) following the manufacturer’s instructions. Primers are designed to exclude conservative sequences. The primers used for qRT-PCR amplification are listed in Table S1. L. perenne eIF4A (eukaryotic translation initiation factor 4 alpha) was used as a reference gene. Gene expression in response to different stress treatments was analyzed over time relative to that recorded at 0 h, with relative gene expression being quantified using the 2^−ΔΔCT method. PCR was performed in a 10 μL amplification volume consisting of 5 μL of 2X SYBR Green Mix, 0.2 μL of forward and reverse primer (10 μM), 2 μL of cDNA (50 ng/μL), and 2.6 μL of ddH₂O. The PCR program was as follows: 1 cycle of 95°C for 30 s, followed by 40 cycles of 95°C for 10 s, 60°C for 30 s, and 72°C for 30 s. The primers used are shown in Table S1, and the specificity of primer pairs was evaluated by dissociation-curve analysis. The resulting clusters were visualized with MeV software.

Prediction of interacting proteins

To identify candidate proteins that interact with LpCBFs, we analyzed the protein interaction network of LpCBF and used the STRING software (https://string-db.org/cgi/input.pl) to predict the interacting proteins, with Triticum aestivum as the background and advanced default settings. Functional enrichment in the network was determined based on local network clusters (STRING).

Statistical analysis

All the data in this study were analyzed using SPSS version 17.0 and the least significant difference (LSD) test. The means and standard errors were calculated, and p < .05 was considered statistically significant in different gene expressions.

Results

Identification and characterization of CBF genes in L. perenne

We identified 12 L. perenne CBF genes in the NCBI GenBank database and predicted their physicochemical characteristics (Table 1). The corresponding LpCBFs-encoded proteins had lengths ranging from 210 to 252 amino acid residues, molecular weights between 22.53 and 26.30 kDa, and isoelectric points ranging from 4.81 to 7.10. The overall average hydropathicity GRAVY values of all predicted LpCBF proteins were negative, ranging from −0.541 to −0.125, thereby implying their hydrophilic nature. The subcellular localization of the LpCBF proteins was examined using UniProt software and predicted using WoLF PSORT (Table S2). The detection of nuclear localization signal (NLS) sequences indicated that these proteins were localized in the cell nucleus.

Table 1. Protein physical and chemical properties of LpCBF gene family in Lolium perenne.

Gene name		Accession/locus ID	CDS (bp)			Protein length (aa)	MW (KD)	pI		GRAVY	Subcellular localization
LpCBF3a	ABK32847			714	237		25.52		4.91	−0.468	Nucleus
LpCBF3b	AAX57275			672	223		24.46		4.81	−0.437	Nucleus
LpCBF3c	ABK32848			714	237		25.55		4.84	−0.453	Nucleus
LpCBFVb	BAF36846			630	210		22.53		7.06	−0.145	Nucleus
LpCBFIVb	BAF36844			675	224		24.32		7.10	−0.456	Nucleus
LpCBFIVa	BAF36843			672	223		24.51		6.97	−0.541	Nucleus
LpCBFIIIc	BAF36842			744	247		25.76		5.68	−0.331	Nucleus
LpCBFIIIb	BAF36841			759	252		26.30		5.12	−0.327	Nucleus
LpCBFIIIa	BAF36840			726	242		25.24		5.23	−0.323	Nucleus
LpCBFII	BAF36839			711	236		25.21		5.27	−0.125	Nucleus
LpCBFIb	BAF36838			729	242		25.74		4.90	−0.303	Nucleus
LpCBFIa	BAF36837			714	237		24.99		4.84	−0.190	Nucleus

Note: CDS, coding sequence length of LpCBF genes; MW, molecular weight of LpCBF proteins; pI, isoelectric point; GRAVY, grand average of hydropathicity.

Phylogenetic analysis

To examine the evolutionary relationships of CBFs found in L. perenne with those of other plant species, we constructed an unrooted phylogenetic tree comprising 70 CBF protein sequences from different species, namely, the model plant Arabidopsis, and the gramineous species, L. perenne, O. sativa, T. aestivum, H. vulgare, and S. cereale (Figure 1, Table S3). CBF proteins from the above species were divided into five groups. In addition, proteins LpCBFIIIc, LpCBFIIIb, LpCBFIIIa, LpCBFII, LpCBFIb, and LpCBFIa were found to be highly similar, clustering on the same branch, indicating that these six proteins may have maintained conserved structural fragments during the course of evolution. Further, LpCBF3a, LpCBF3b, and LpCBF3c apparently bear relatively close genetic relationships with HvCBF6 and ScCBFIIIa-6, and were clustered together with certain proteins from T. aestivum in common composition group II. LpCBFIVa and LpCBFIVb were clustered as an orthologous pair with HvCBF14 grouped at the secondary end of a clade, suggesting a close genetic relationship, whereas LpCBFVb was located on a different short branch of group III. The tree indicated that none of the assessed L. perenne CBFs clustered in either group IV or V, the former of which contained all assessed CBF genes from Arabidopsis (Figure 1). It should be noted that our assessment of genetic relationships among CBF genes was determined by the branch length of the phylogenetic tree, which was constructed based on sequence alignment.

Figure 1. Phylogenetic relationships among CBF proteins of different plants species. The relationships among CBF proteins were inferred from a phylogenetic tree constructed using the neighbor-joining method. The numbers shown at tree nodes indicate the percentage bootstrap replicate trees in which the associated taxa clustered together. Labels above nodes indicate the source species of the respective CBF proteins.

Multiple sequence alignments

The multiple sequence alignments of the LpCBF genes were determined based on the protein sequence, which enabled us to assign the 12 LpCBF genes into three groups (Figure 2a). Furthermore, LpCBF protein sequences were compared to assess amino acid homology, and possible duplication mechanisms were assessed using a matrix of protein sequence identities/positives and sequence coverage (Table S4). Predictably, we obtained high scores for identities and positives among members of the same group.

Figure 2. Different motifs analysis diagram in LpCBF proteins. (a) Different motifs were identified using the MEME search tool. Different motifs are indicated by different colors. Boxes represent conserved motifs among the proteins analyzed. (b) Amino acid chain-length and sequence information for Motif 1 to 10.

Further, we used the MEME program to identify common motifs in LpCBF proteins, the 10 most highly enriched of which are shown in Figure 2a. The location and number of motifs were found to be very similar among the members clustering on the same branch of the phylogenetic tree, such as LpCBF3a, LpCBF3b, and LpCBF3c, with LpCBF3a and LpCBF3c being located on the same short branch and characterized by almost identical motif patterns. Nine motifs were identified in CBFs located on the second largest branch, except for LpCBFII, which lacked motif 10. A similar pattern was detected for CBFs on the third branch. Among these motifs, further analysis of motif 1 revealed an AP2 domain (50 amino acid residues), a common feature of the CBF family. The sequence and structural characteristics of other motifs (motifs 2–10) are shown in (Figure 2b). Multiple sequence alignments of the CBF gene family revealed that similar motif structures were clustered together on a single branch, implying their functional similarity.

Conserved domain analysis and three-dimensional model prediction

Multiple sequence analysis showed that the LpCBF protein contained an AP2 domain, a PKKPAGR motif (PKKPAGRxKFxETRHP) and a conserved sequence of DSAWR (Figure S1). Using the SMART program to investigate the conserved domains within LpCBF protein sequences (Figure 3), a very similar distribution of conserved domains among all LpCBF families was revealed, as they were all found to contain an AP2 DNA-binding domain that has been shown to play an important role in biotic responses.

Figure 3. Analysis of LpCBF-protein conserved domains. The conserved domains of LpCBF proteins were identified using SMART software, and the different domains are indicated by different colors.

The higher-level structure of plant proteins is related to their biological function and activity. In this study, we examined the three-dimensional structure of LpCBF proteins based on predictive modeling. Accordingly, we found that the LpCBF family members are highly similar with respect to α-helix and β-sheet configurations, with only certain small differences in random coils being identified (Figure 4). LpCBF sequence identity was found to range from 46.03% to 52.46%, and the GMQE and QMEAN trends were also small (Table S5). The template used to generate the three-dimensional model of LpCBF proteins was measured relative to the structure of the complex of GCC-box binding domain by solution nuclear magnetic resonance spectroscopy. Therefore, the three-dimensional structure of LpCBFs may play an important role in transcriptional regulation.

Figure 4. The three-dimensional structures predicted using the Swiss-MODEL server was shown for each of the analyzed proteins in the LpCBF-protein subfamily. Sequence identities >30% between the template and targets were obtained. Template model selection was based on the model with the highest GMQE score.

LpCBF gene expression in response to abiotic stress

To identify the functional role of CBFs in L. perenne, we used qRT-PCR to analyze gene expression in response to the exposure to different abiotic stress factors including cold, heat, salinity, PEG 6000, and ABA stress. Our results showed that the 12 CBF genes under study responded to different extents to the abiotic stress conditions tested (Figure 5a, Table S7). Most of the 12 genes were upregulated across the stress treatments, with all 12 CBFs being upregulated in response to cold (4°C). We observed that when the cold stress duration was extended, the expression of the 12 CBF genes initially increased and subsequently decreased with time, generally peaking after 6 h under stress. The only exception in this regard was LpCBF3a, the expression of which peaked at 12 h. The expression of 10 of the CBF genes was also upregulated under 42°C and water deficit treatments, whereas that of the remaining two (LpCBF3c and LpCBFIVb) was down-regulated under the same conditions. Meanwhile, the expression of nine CBFs was markedly upregulated under salinity stress, whereas that of LpCBF3c, LpCBFIIIa, and LpCBFIIIb was slightly upregulated when exposed to high salinity. Ten CBFs also showed upregulated expression in response to ABA stress, whereas the expression of the remaining two (LpCBFIIIb and LpCBFIVb) was down-regulated (Figure 5a).

Figure 5. Temporal expression of 12 LpCBF genes in response to diverse stress factors. Heat map of the 12 LpCBF genes in seedlings subjected to 4°C, 42°C, salinity, water deficit, and ABA stress treatments over a 12-h time course (a). Temporal expression of LpCBF3a (b), LpCBFIa (c), LpCBFIIIc (d) and LpCBFVb (e) under 4°C, 42°C, salinity, PEG 6000 and ABA stress treatments.

The expression of four genes, namely, LpCBF3a, LpCBFIa, LpCBFIIIc, and LpCBFVb, was found to be markedly upregulated in response to all five stress treatments, particularly in seedlings exposed to 4°C; furthermore, among these genes, the expression of LpCBFIa and LpCBFIVb was markedly higher under the 4°C treatment than under any other stress condition (Figure 5b-e). Additionally, we compared the expression of the 12 CBFs in seedlings that had been exposed to 4°C for 6 and 12 h. In this case, we found that, at both 6 and 12 h, the expression of LpCBFIa was highest and that the expression of LpCBF3b, LpCBFIb, and LpCBFVb was higher than that of other genes at 6 h, whereas the expression of LpCBF3b, LpCBFIb, LpCBFII, LpCBFIIIa, and LpCBFVb was higher than that of other genes when exposed to 4°C for 12 h (Figure S2). Furthermore, we established that the response of these genes to cold stress was treatment duration-dependent. These findings indicated that LpCBFs may potentially play an important role in cold resistance in L. perenne, particularly, LpCBF3a, LpCBFIa, LpCBFIIIc, and LpCBFVb.

Prediction of interacting proteins

To gain a further insight into the function of LpCBF proteins, we sought to predict the proteins with which LpCBFs might interact. To this end, we generated protein association networks for LpCBF3a, LpCBFIa, LpCBFIIIc, and LpCBFVb (Figure 6a-d). A number of different proteins were predicted to interact with these four LpCBF proteins, thereby suggesting a regulatory function. Somewhat surprisingly, different models were predicted for LpCBFIa and LpCBFIIIc, although the associated proteins were the same, thereby indicating that they play similar functional roles (Figure 6b, c). Two of the predicted proteins, NAC2b and LEA (late embryogenesis abundant), are known to be associated with stress. NAC transcription factors have been shown to be involved in plant tolerance to abiotic stress,²⁸ a member of the LEA protein family, which is abundant in seeds and pollen, and plays a regulatory role in tolerance against multiple abiotic stress conditions.²⁹ We also predicted LEA proteins that interact with LpCBF proteins in the network based on functional enrichments (Table S6). Overall, these findings indicate that LpCBF proteins play regulatory roles in response to adverse environmental stimuli, although the underlying mechanisms may differ.

Figure 6. Predicted protein interaction networks of LpCBF3a (a), LpCBFIa (b), LpCBFIIIc (c), and LpCBFVb (d). Protein interaction networks were generated using STRING online with Triticum aestivum as the background. Edges represent protein–protein associations, nodes represent proteins, and red nodes represent query proteins. Black, green, blue, light sky-blue, and purple lines represent co-expression, text-mining, gene co-occurrence, protein homology, and experimental determination, respectively.

Discussion

CBF gene families have been identified in a number of plant species,²³ and sequences have also been analyzed in rye (Secale cereal),³⁰ wheat (Triticum aestivum),³¹ cotton (Gossypium hirsutum),³² and barley (Hordeum Vulgare).³³ In this study, we identified 12 CBF genes in the grass species Lolium perenne, which were classified into three groups (Table 1). Phylogenetic analysis revealed that among the species under study, LpCBF sequences are more closely related to those of T. aestivum, S. cereale, and H. vulgare than to those of O. sativa (Figure 1). Although the sequences tend to be highly conserved and are predicted to have a similar function, it has been found that different CBF family genes within the same species can differ with respect to structure and function, despite having similar sequences. Genome-wide identification and analysis are often used to study gene families;³⁴ further, bioinformatics-based control of protein function is a practical method in gene research.³⁵ To gain further understanding of the properties of LpCBF proteins, we performed conserved domain and three-dimensional modeling analyses, the results of which revealed that LpCBF family members typically have conserved domains and protein structures (Figure 3 and 4). Furthermore, our comparative genome sequence analysis using the MEME program indicated that LpCBF members within the same clade are characterized by the same motifs (Figure 2). The diversity of CBF sequences in terms of phylogenetic relationships, structural and functional characteristics, and regulatory levels contribute to the differentiation in the regulatory properties of these proteins.³⁶

The functions of some CBF transcription factor genes have been determined based on annotation analysis that has established that CBF genes play roles in the response of plants to different abiotic stress factors.³⁷ Moreover, examination of the different expression patterns of orthologous CBF genes under different stress conditions suggests that these genes may be involved in different regulatory pathways.³⁸ Similarly, our analysis of the expression profiles of LpCBFs reported herein revealed differential expression in response to low/high temperature, water deficit, salinity, and ABA treatments (Figure 5). We found that the expression of all 12 LpCBFs studied was upregulated in seedlings exposed to cold (4°C), initially increasing and subsequently decreasing when the stress period was extended (Figure 5). These observations are consistent with those reported by Han,³⁹ who demonstrated that cold treatment (4°C) induced the expression of LpCBF3. Similarly, our findings are consistent with those obtained for CBF family members in other plants. For example, among the CBF homologs LeCBF1–3 in tomato (Lycopersicon esculentum), only the LeCBF1 gene was found to be cold-inducible.⁴⁰ Similarly, in Arabidopsis, AtCBF1, AtCBF2, and AtCBF3 were shown to be induced in response to low temperature,¹¹ and in papaya (Carica papaya), the expression of CBF1 and CBF2 was observed to increase under low temperature.⁴¹ Furthermore, the findings of Zhao regarding low temperature-induced sensitivity and resistance of double mutants indicated that AtCBF2 plays a more important role than either AtCBF1 or AtCBF3 in freezing tolerance, whereas in Malus sieversii observed different expression patterns for four CBF homolog genes in response to a period of cold treatment.^42,43 Altogether, these findings provide important evidence indicating that LpCBFs play an important role in the resistance to low-temperature stress.

Previous studies have indicated that in addition to low temperature, the expression of CBFs is similarly induced by other stress factors. For example, Jiang found that the expression of CBF homologs was induced by exposure to low temperature, salinity, and water deficit stress,⁴⁴ while B yun observed that DaCBF7 was induced by water deficit, cold, and salinity.⁴⁵ In turn, Li demonstrated that the MtCBF4 gene from Medicago truncatula plays an important role in responses to water deficit and salinity stress conditions.⁴⁶ Consistently, herein, we found that the expression of most of the 12 LpCBFs studied was also induced under salinity and water deficit (Figure 5a). Studies have also revealed that ABA treatment induces the expression of CBF homologs, thereby indicating that an ABA-inducible signaling pathway is involved in the regulation of cold-regulated genes via the CRT promoter element.⁴⁷ Consistent with this assumption, the exogenous application of ABA has been shown to enhance the expression of VvCBF2, VvCBF3, VvCBF4, and VvCBF6,⁴⁸ and Peng observed that, whereas the BgCBF1 gene showed limited induction in response to salinity and water deficit stress, it exhibited relatively high upregulated expression following ABA treatment.⁴⁹ In this regard, it has been reported that in the regulation of OST1 (open stomata 1), a Ser/Thr protein kinase in the ABA signaling pathway interacts with ICE1, an upstream regulator involved in the CBF pathway.⁵⁰ Here, we found that the expression of 10 of the 12 assessed LpCBFs was up-regulated in response to ABA treatment, whereas that of the remaining two LpCBFs was down-regulated (Figure 5a). These observations indicate a complex and intimate interaction between the ABA and the CBF pathways.

One of the primary objectives of this study was to determine whether candidate LpCBFs play a role in plant adaptation. On the basis of our analysis of expression profiles, we identified four candidate genes (LpCBF3a, LpCBFIa, LpCBFIIIc, and LpCBFVb) that were highly expressed in response to exposure to multiple stress factors (Figure 5b-e). Notably, these four candidate genes were induced in seedlings subjected to low temperature (4°C), water deficit, and salinity stress; therefore, we speculate that they may play regulatory roles in the stress responses of L. perenne, which warrants further examination (Figure 5b-e). These results are also consistent with those reported previously, indicating that the SpCBF1 gene from Solanum pinnatisectum, which is induced by chilling stress, plays an important role in chilling stress.⁵¹ CBFs have been shown to participate in cold-induced activation of the transcriptional response,⁵² and the CBF subfamily has been found to be dependent on a low-temperature signal chilling pathway.⁵³ In addition, CBFs have been demonstrated to affect plant growth and development,⁵⁴ while PpCBF1 has been shown to be involved in regulation of the bud endodormancy process in pears.³⁵ Furthermore, CBFs have been established to play regulatory roles in the tolerance to metal stress, such as that induced by cadmium and molybdenum treatment.⁵⁵ Altogether, these findings imply that CBF members play distinct roles in the response to abiotic stress, whereby, it would be desirable to further investigate the functions of the LpCBF genes identified in the present study.

CBFs have previously been characterized as transcriptional factors that contribute to transcriptional regulation in response to abiotic stress;⁵⁶ furthermore, cold response-related genes implicated in the response to low temperature have been identified and the underlying mechanisms characterized.⁵⁷ Here, we used the aforementioned candidate genes as a basis to predict LpCBF protein association networks (Figure 6 and Table S6). Our data revealed that these CBF proteins potentially interact with NAC transcription factors and LEA proteins, as well as other currently uncharacterized proteins which, accordingly, warrant further study to enable a functional characterization of their roles in L. perenne. Similarly, further in-depth research is required to elucidate the mechanisms underlying transcriptional regulatory mechanisms of LpCBFs. Low-temperature environments induce CBF genes via associated signal transduction pathways, thereby initiating the subsequent regulation of target genes that mediate the response to low-temperature stress.⁵⁸ For example, it has been demonstrated that MdBBX37 interacts with MdICE1 to enhance the transcriptional activity of MdICE1 in the MdCBF1 pathway, which in turn initiates the regulation of cold stress tolerance-signaling.⁵⁹ Furthermore, CBFs have been demonstrated to enhance cold adaptability by regulating the expression of COR genes.⁶⁰

Abiotic stress limits plant growth and development. To counter the resulting adverse effects, plants have evolved sophisticated adaptive capacities to deal with different types of stress.⁶¹ Furthermore, from the perspective of the application of CBFs to molecular breeding, ectopic expression of CBF subfamily genes has been shown to enhance plant tolerance to abiotic stress. For example, the ectopic expression of HvCBF7 and HvCBF9a in Arabidopsis seemingly enhances salinity tolerance,⁶² while the ectopic expression of pear PpCBF1 confers increased cold hardiness.⁶³ Similarly, the ectopic expression of AtCBF3 in Solanum tuberosum enhanced the tolerance of transformed plants to high temperature (40°C) stress.⁶⁴ We anticipate that further research on the L. perenne CBF genes identified herein will contribute to enhancing environmental adaptation of this grass, and that a combination of modern molecular biology technology and traditional breeding methods will improve the efficiency with which this objective is achieved.

Conclusion

We identified 12 LpCBF genes in the grass species Lolium perenne and performed a comprehensive characterization of these genes with respect to protein features, phylogenetic relationships, motifs, and structural domains. Furthermore, our analysis of the expression profiles of LpCBF genes in seedlings exposed to different stress conditions indicated that these genes potentially play a variety of roles in response to cold, heat, salinity, water deficit, and ABA stress. In addition, on the basis of association network analysis, we predicted the identity of putative interacting proteins presumably regulated by LpCBFs. We believe the findings reported herein will provide valuable information for further functional research on the role of LpCBF proteins with respect to plant responses to stress and perspective applications in breeding for stress resistance. Finally, our observations provide a sound theoretical basis for functional studies of the LpCBF gene family and contribute to developing a potential strategy for further breeding of L. perenne.

Abbreviations

CBF	=	C-repeat binding factor
AP2/ERF	=	APETALA2/Ethylene responsive factor
DRE/CRT	=	dehydration responsive element/C-repeat
COR	=	Cold regulated
ABA	=	Abscisic acid
pI	=	Isoelectric point
MW	=	Molecular weight
GRAVY	=	Grand average hydropathicity
NLS	=	Nuclear localization signal
cDNA	=	Complementary DNA
PEG	=	Polyethylene glycol
eIF4A	=	Eukaryotic translation initiation factor 4 alpha
LEA	=	Late embryogenesis abundant
NAC	=	NAM、ATAF1/2、CUC2.

Author contributions

NS and WD wrote the main manuscript text. NS and WD contributed to the conception of the study. WD, CB, GH, and LY performed the experiments and the data analyses. All authors have read and agreed to the published version of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All data generated or analyzed during this study were included in this published article and the addition-al files.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15592324.2022.2086733

Additional information

Funding

This research was funded by the Sichuan Science and Technology Program (2020YJ0413) and the Fundamental Research Funds of China West Normal University (19E052). The funding bodies played no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.