Introduction

Abiotic stresses, including soil salinity, drought, high and low temperatures, have emerged as major limiting factors affecting crop yield and quality1. With climate change intensifying and extreme weather events becoming more frequent, it is anticipated that abiotic stresses will increasingly impact crop production, posing a threat to global food security2,3. In response to these challenges, plants have evolved a diverse array of molecular mechanisms to rapidly perceive and adapt to environmental changes.

Eukaryotes respond to and adapt to environmental changes by employing transcriptional and post-transcriptional regulatory mechanisms4. Alternative splicing (AS) is a process wherein potential splicing sites on precursor mRNA (pre-mRNA) transcripts may or may not be utilized following transcription of the master gene, resulting in the generation of diverse mature mRNA isoforms from an intron-containing gene. AS regulation enhances the diversity of the transcriptome and proteome expressed from the same set of pre-mRNA transcripts5,6. Recent data from the human transcriptome and translatome have confirmed that AS can increase protein diversity7,8. Similarly, AS contributes to the diversity of the transcriptome and proteome in the plant kingdom9,10. Particularly, plants employ more intensive AS regulation to produce protein isoforms adapted to environmental stress11,12. Therefore, AS regulation represents an important post-transcriptional strategy in plants' response to environmental stress13,14. The application of AS regulation holds promise for enhancing the stress tolerance of crops in the face of global climate change10,15.

The majority of post-transcriptional RNA metabolism regulation is mediated by various RNA-binding proteins (RBPs)16. GR-RBP proteins represent a subgroup of RBPs within the fourth subfamily of the glycine-rich protein (GRP) superfamily. They are characterized by glycine-rich regions, (Gly)n-X (where n is typically an odd number and X can be any amino acid, including glycine), primarily located at the C-terminus of the protein17,18. Based on the motifs present in these proteins, GR-RBPs can be further categorized into four subclasses: IVa, IVb, IVc, and IVd. Among these, both IVa and IVd contain RNA recognition motifs (RRMs) at the N-terminus, with the distinction that IVa has one RRM, while IVd has two. Conversely, IVb and IVc have one or more CCHC zinc fingers located at the C-terminus, with IVb featuring an RRM at the N-terminus and IVc featuring a cold shock domain (CSD) at the N-terminus18,19. However, the classification of GR-RBP proteins remains unclear and occasionally contentious. For example, AT2G21660 (AtGRP7) was initially categorized as a protein in the IVa subgroup20, it was recently reclassified as a member of IVb21. Similarly, OsGRP4 and OsGRP5 in rice, exhibiting strong homology with AtGRP2/7 in Arabidopsis were initially considered GR-RBP proteins19,22. Additionally, GRMZM2G001850 was classified as a GR-RBP protein in maize54.

The GR-RBPa protein is chaeaterized by a conserved RNA Recognition Motif (RRM) domain located at the N-terminal and a glycine-rich C-terminal. Both terminals of GR-RBPa proteins play important roles in their functionality. Studies have shown that variations in the length of the glycine-rich region of AT3G223830 can lead to functional differences in cold stress adaptation, while another study suggested that the difference may originate from the N-terminal RRM domain itself55,56. Our analysis revealed significant variation in the length of the C-terminal region among different GR-RBPa members (Fig. 2). Moreover, GR-RBPa proteins from the five plant species could be further classified into two branches based on the conserved motifs they possess, consistent with previous findings24. For instance, motif 8 is exclusively found in branch I, while motifs 4, 7, and 9 are unique to branch II. GR-RBPa proteins within the same branch of a plant species are believed to exhibit more conservative molecular functions than those between different branches. For example, AT2G21660 and AT4G39260 of branch I regulate their own pre-mRNAs, cross-regulate each other’s pre-mRNAs, and share several downstream pre-mRNAs33,57. In contrast, although both AT4G13850 (belonging to branch II) and AT2G21660 have been shown to enhance cold resistance in plants, AT4G13850 promotes the germination of Arabidopsis seeds under salt and drought stresses, while AT2G21660 has a negative effect on seed germination and seedling growth under salt or dehydration stress conditions27,55,58,59. In addition, overexpression of AT3G23830 in branch II did not confer cold tolerance but only enhanced salt and drought resistance in the transgenic plants, similar to AT4G1385056. Interestingly, our protein–protein interaction analysis revealed that GR-RBPa proteins in branch I shared more interacting proteins than those in branch II, suggesting more conserved molecular function of GR-RBPa proteins in branch I (Fig. 3, Supplemental Fig. 2). Multiple and diverse roles of GR-RBPa proteins could be attributed to the different motifs they possess, as protein–protein interactions and binding of specific nucleotides of proteins may be affected by the motifs they contain60. Therefore, further systematic comparisons are needed in the future to confirm the functions of specific motifs of GR RBPa proteins.

GR-RBP proteins are known to play crucial roles in regulating plant growth and responding to stress17,31. Our analysis revealed that the promoters and introns of GR-RBPa genes are enriched with various cis-elements, including those involved in light signaling, hormone signaling, growth and development, and stress responses (Fig. 5). This indicates that GR-RBPa genes are subject to regulatory control at the expression level in response to various internal and external cues. Previous studies have shown that GR-RBP genes are highly induced by heat stress in sorghum61 and Pinellia ternate62. On the other hand, the expression level of LpGRP1 transcript in ryegrass was significantly induced with prolonged exposure to cold stress63. Additionally, the transcriptional levels of PpGRPs in Physcomitrella patents fluctuated during prolonged cold stress50. In our study, to ease and ensure the comparability between individual investigations, we use the same hydroponics-based experimental conditions when cultivate different plants. However, it is important to note that this approach may introduce extra stresses, such as oxygen and soil microbial deficiency, to the plants64. It's worth mentioning that we did not calculate whether there are differences and how significant they are in the gene transcription and splicing patterns of GR-RBP genes between plants cultivated in soils and those grown in water, as our focus was primarily on the effects generated from salt, drought and temperature here.

When we subjected the five plant species to salt, drought and temperature stresses, we observed significant and intricate changes in the expression of GR-RBPa genes compared to their expression levels in the original culture solution. For example, most GR-RBPa genes from maize and tomato showed transcriptional induction in response to all types of abiotic stress employed. Conversely, GR-RBPa genes in rice tended to be down-regulated under salt and cold stresses. Additionally, GR-RBPa genes in branch I were generally downregulated by all kinds of stress treatments, while those in branch II were more frequently induced by stress. The expression of some GR-RBPa genes exhibited fluctuations in response to different types of abiotic stress or even different durations of a particular stress. Moreover, homologous GR-RBPa genes from the same plant species could be regulated in opposite directions by the same stress. However, transcriptional regulations of GR-RBPa genes did not exhibit clear divergence between monocots and dicots. In conclusion, transcriptional levels of these GR-RBPa genes are not conserved among the plant species we tested. Therefore, it is essential to investigate in future studies whether transcriptional regulation of the GR-RBPa genes is controlled by coordinated modulations between different cis-elements. Alternatively, there could be other unidentified cis-elements that significantly affect transcriptional levels.

Accumulating studies have demonstrated that pre-mRNA alternative splicing (AS) regulation plays essential roles in plant growth and stress response because alterations in the relative abundances of different transcript isoforms can affect the abundance and diversity of protein products65,66,67,68. In our study, the results of gene structure analysis showed a clear distinction in the number of introns between GR-RBPa genes in branch I and branch II. The genes in branch I have fewer introns (only 1 or 2) than those in branch II (Fig. 4), suggesting a potential co-evolution of gene structure and protein function among GR-RBPa proteins in plants. Surprisingly, RT-PCR results demonstrated that all of the GR-RBPa genes in branch I tested were found to be alternatively spliced and subsequently generated multiple mRNA isoforms, while nearly all GR-RBPa genes in branch II were only constitutively spliced, even though they contained more introns (Fig. 9). In addition, AS patterns changed significantly among GR-RBPa genes in branch I, for various types of stress or different durations of the same stress treatment could induce different AS variations in pre-mRNAs of these genes. Moreover, AS regulations of homologous GR-RBPa genes changed significantly among different plant species, and no clear divergence of AS regulation model was found between monocot and dicot species. These findings highlight the complexity and variability of AS regulations in coding genes of the RNA binding proteins (RBPs), which have also been recently reported in another family, serine/arginine-rich (SR) protein45,69.

The reason for the differential response of different branches of GR-RBPa genes to identical stress at both transcription and splicing levels remains unclear. It remains uncertain whether or not the evolution in gene structure results in specific regulation of the master genes themselves, and subsequently affecting downstream genes at the genome-wide level. Studies in different plants have shown that AS of pre-mRNA of GR-RBPa proteins can alter the splicing patterns of downstream genes57,70. In our experiments, there were also significant changes in the abundances of different AS transcripts of GR-RBPa proteins when the plants were subjected to environmental stresses. In-silico analysis has also indicated that potential truncated proteins may be generated through AS of AT2g21660 gene (Fig. 9c–f). Existing evidence supports that AS-induced protein variants have different functions in response to environmental changes. For instance, HAB1 protein variants generated through pre-mRNA AS regulation play opposite roles in ABA signaling71,72. Additionally, two protein variants of OsZIP1 generated through pre-mRNA AS regulate development of rice plant under light and dark conditions, respectively73. However, further functional studies are required to confirm how different GR-RBPa alternatively spliced variants affect the stress response and adaptation in plants.

In conclusion, the intricate and diverse expression regulation of GR-RBPa genes across different plant species, at both transcriptional and pre-mRNA splicing levels, prompts questions about the contribution of these regulatory mechanisms to plant development and stress adaptation. Despite sharing relatively conserved functional domains, the impact of their diverse expression patterns on growth and development of the plants remains unclear. Understanding the biological significance of the variations in each pre-mRNA isoform could be crucial for future application of this gene family in breeding crops for enhanced stress tolerance.