The whole-genome sequences for I. batatas, I. trifida, and I. triloba were obtained through Ipomoea Genome Hub (https://ipomoea-genome.org/) (viewed: 6 March 2023) and Sweetpotato Genomics Resource (http://sweetpotato.plantbiology.msu.edu/) (viewed: 6 March 2023). Three different screening techniques (Dai et al., 2022) were used at once to ensure that all members of GLR family were accurately identified.

The molecular weight, hydrophilicity, instability index, and theoretical isoelectric point for GLRs were computed through the Expert Protein Analysis System (ExPASy, https://www.expasy.org/) (viewed: 12 March 2023). Protein Subcellular Localization Prediction (PSORT, https://wolfpsort.hgc.jp/) (viewed: 12 March 2023) was employed to predict the subcellular localizations for GLRs.

IbGLRs, ItfGLRs, and ItbGLRs were separately mapped to the respective chromosomes for I. batatas, I. trifida, and I. triloba by Ipomoea Genome Hub (https://ipomoea-genome.org/) (viewed: 14 March 2023) and Sweetpotato Genomics Resource (http://sweetpotato.plantbiology.msu.edu/) (viewed: 14 March 2023). The Toolkit for Biologists integrating various biological data handling tools (TBtools) program (South China Agricultural University, Guangzhou, China) was used for visualization.

MEGA ClustalW 7.0 was used to phylogenetically analyze the amino acid sequences for GLRs of Arabidopsis, I. batatas, I. trifida, and I. triloba (Kumar et al., 2016). Bootstrapping was carried out with 1000 replicates, and Interactive Tree of Life (iTOL) software (http://itol.embl.de/) (viewed: 8 June 2023) was used to construct a phylogenetic tree.

To perform an in-depth study of GLRs’ conserved motifs, the Multiple Expectation Maximizations for Motif Elicitation (MEME) program (https://meme-suite.org/meme/) (viewed: 15 March 2023) was used. The maximum number of motifs that could be used has been set at five.

A database of Plant Cis-Acting Regulatory Element (PlantCARE, (http://bioinformatics.psb.ugent.be/webtools/plantcare/html) (viewed: 16 March 2023) predicted cis-elements within 1500 bp promoter region for GLRs (Lescot et al., 2002). Gene Structure Display Server (GSDS) 2.0 developed by Peking University, Beijing, China (http://gsds.gao-lab.org/) (viewed: 16 March 2023) and the TBtools software developed by South China Agricultural University, Guangzhou, China was used to obtain and visualize the exon-intron structures for GLRs, respectively.

Depending upon Arabidopsis orthologous proteins, the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING, https://www.string-db.org/) (viewed: 18 March 2023) predicted GLR protein interaction network. The Cytoscape version 3.2 was used to construct the network map (Kohl et al., 2011).

Transcriptome analysis and real-time quantitative PCR (qRT-PCR) was carried out at duration points (0 h, 36 h, 72 h, 120 h, and 10 d) following root rot induction using the underground stem of resistant variety Jishuzi203 and susceptible variety Jishuzi563. These plants were provided by the Institute of Cereal and Oil Crops, Hebei Academy of Agriculture and Forestry Sciences (Shijiazhuang, China) and planted in the natural root rot disease nursery in Xiong'an New Area, China. qRT-PCR was conducted on a CFX Opus 384 Real-Time PCR system (Bio-Rad, USA) by the ChamQ Universal SYBR qPCR Master Mix (Vazyme, China). The relative expression level of the target gene was presented as fold change compared with the internal control using the 2 ^{- Δct} method, and data were analyzed with Duncan’s multiple range test (p < 0.05). Three biological replications were performed for each test. A gene of I. batatas ADP-ribosylation factor (ARF, GenBank JX177359) was used as an internal control. The specific primers used for the qRT-PCR analysis were listed in Supplementary Table S2 . Based on related investigations (Zhang et al., 2017; Dong et al., 2019; Zhu et al., 2019) ribonucleic nucleic acid sequencing (RNA-seq) data for IbGLRs were collected. RNA-seq data for ItfGLRs and ItbGLRs were collected through Sweetpotato Genomics Resource (http://sweetpotato.plantbiology.msu.edu/) (viewed: 22 March 2023). The fragments per kilobase of exon per million mapped fragments (FPKM) method was used to calculate the GLRs expression levels. Tbtools was used to build the heat maps.

37 GLRs for sweet potato (named Ib) and 32 for each of two diploid relatives (named Itf and Itb) were identified through a combination of three methods used by Dai et al. (2022). The sequences from I. batatas were used for the analysis of the physicochemical features of GLRs ( Table 1 ). The coding sequence length of IbGLRs varied from 1977 bp (IbGLR7) to 7326 bp (IbGLR36). The molecular weights of IbGLRs ranged from 73.627 to 270.105 kDa, the amino acid lengths extended from 658 to 2441 amino acids, and their isoelectric points varied from 5.49 (IbGLR32) to 9.31 (IbGLR30). Most IbGLRs were stable, only IbGLR2/10/11/13/14/15/21 and IbGLR30 were unstable (instability index > 40). 17 hydrophobic proteins (the grand average of hydropathy [GRAVY] score > 0) and 20 hydrophilic proteins (GRAVY score < 0) were identified for the IbGLR family. Subcellular localization prediction assessment demonstrated that except for IbGLR1 (located on chloroplast) and IbGLR12 (located on chloroplast and nucleus), the remaining IbGLRs were located on the plasma membrane.

Based on the physical locations of genes in the I. batatas, I. trifida, and I. triloba genomes, the chromosomal positions of GLRs in these crops were mapped. In I. batatas, IbGLRs were distributed unevenly on nine chromosomes. LG13 had the most IbGLR genes, with 11, followed by LG7 with seven. LG3, LG9, and LG12 had four IbGLRs. There were less than four IbGLRs on LG1, LG5, LG8, and LG15 ( Figure 1A ). In I. trifida, except for ItfGLR31 and ItfGLR32 located on Chr00 (not shown in Figure 1B ), the other ItfGLRs were located on the same chromosomes (Chr02, Chr03, Chr05, Chr06, Chr07, Chr10, Chr11, Chr12, and Chr14) compared to those in I. triloba. Chr02 had the most ItfGLR genes, with 12, followed by Chr14 with four. There were less than four ItfGLRs on Chr03, Chr05, Chr06, Chr07, Chr10, Chr11, and Chr12 ( Figure 1B ). In I. triloba, Chr02 had the most ItbGLR genes, with 11, followed by Chr03 with six and Chr14 with five. There were less than four ItbGLRs on Chr05, Chr06 Chr07, Chr10, Chr11, and Chr12 ( Figure 1C ). Chr06, Chr10, Chr11, and Chr12 had the same numbers of GLRs in I. trifida and I. triloba. These findings imply that two diploid relatives differed in the proportion and distribution of GLRs on chromosomes in comparison to sweet potato.

Phylogenetic trees were constructed to elucidate the evolutionary interactions between 121 different GLRs from Arabidopsis (20), I. batatas (37), I. trifida (32), and I. triloba (32) ( Figure 2 ). The GLRs for different species were categorized according to evolutionary distances into different subgroups, including five (Groups I to V) of I. batatas, four (Groups I to III, and V) of I. trifida, five (Groups I to V) of I. triloba, and three (Groups I to III) of Arabidopsis. The specific distributions for GLRs were as follows (total: I. batatas, I. trifida, I. triloba, Arabidopsis): Group I (37: 11, 12, 10, 4), Group II (27: 6, 6, 6, 9), Group III (47: 16, 10, 14, 7), Group IV (2: 1, 0, 1, 0), and Group V (8: 3, 4, 1, 0) ( Figure 2 ; Supplementary Table S1 ). All AtGLRs were found to have homologous proteins in I. batatas, I. trifida, and I. triloba, but IbGLR18/19/27/28, ItfGLR10/11/23/32, and ItbGLR10/24 showed no homology with Arabidopsis GLRs. Except for Groups II and IV, the numbers and types of GLRs distributed in the I. batatas subgroups were different from those in I. trifida, I. triloba, and Arabidopsis.

The MEME website was used to analyze the sequence motifs of 37 IbGLRs, 32 ItfGLRs, and 32 ItbGLRs, and motif 1 to 5 were identified as five conserved motifs ( Figure 3A ). All GLR sequences were used to produce the sequences for the five most conserved motifs, shared between sweet potato and two diploid relatives ( Supplementary Figure S1 ). Despite being substantially similar, the GLRs for I. batata, I. trifida together with I. triloba could have different numbers and types of conserved domains. For example, ItfGLR1 contained motifs 2 to 5, ItbGLR1 contained motifs 1 to 5, and IbGLR31 contained motifs 1 to 5. ItfGLR4 contained motifs 1 to 5, ItbGLR4 contained motifs 1 to 5, and IbGLR34 contained motifs 1 to 3 and 5. ItfGLR8 comprised motifs 1 to 5, ItbGLR8 contained motifs 3 to 5, and IbGLR30 contained motifs 1 to 5 ( Figure 3A ). Additionally, the receptor domains (atrial natriuretic factor [ANF]_receptor and periplasmic_binding_protein) and four transmembrane helix domains (Lig_chan domains) of GLRs are closely related to plant functions. Most GLRs of I. batata, I. trifida, and I. triloba (30 IbGLRs, 28 ItfGLRs, and 29 ItbGLRs) contained ANF_receptor and Lig_chan domains. However, ItbGLR31 only contained a Lig_chan domain. IbGLR7/18/23/28, ItfGLR32, and ItbGLR24 contained periplasmic_binding_protein_type1 and Lig_chan domains, and IbGLR6/22/24, ItfGLR18/19/24, and ItbGLR25 contained ANF_receptor, periplasmic_binding_protein_type2, and Lig_chan domains ( Figure 3B ).

Exon-intron architectures were examined in order to gain a better understanding of structural variation among the GLRs ( Figure 3C ). The number of exons in the GLR genes varied from 3 to 18. In detail, GLRs of Group I contained 4 to 18 exons, GLRs of Group II contained 3 to 18 exons, GLRs of Group III contained 6 to 17 exons, GLRs of Group IV contained 5 to 6 exons, and GLRs of Group V carried 5 to 17 exons ( Figure 3C ). Additionally, exon-intron architecture for several homologous GLRs in I. batatas was identified as possibly different from the counterparts in I. trifida and I. triloba. Such as, IbGLR36 carried 18 exons, but ItfGLR7 and ItbGLR7, the corresponding homologous genes of IbGLR36 contained 10 and 11 exons in Group I, IbGLR30 contained 18 exons, but ItfGLR8 and ItbGLR8 both contained five exons in Group II, and IbGLR13 carried nine exons, but ItfGLR14 and ItbGLR14 contained 12 and six exons in Group III ( Figure 3C ). These findings suggest that sweet potato genome potentially underwent a lineage-specific differentiation event involving GLR gene family members.

A cis-element, such as the sequence upstream of GLRs, could play a significant role in plant development and stress responses. Hence, we used a 1500 bp DNA sequence upstream of IbGLRs for cis-element analysis in I. batatas. According to the predicted functions, such components were separated into five groups (core, hormone, developmental, light, and abiotic/biotic) ( Figure 4 ). The CAAT-box and TATA-box core promoter elements were present in all 37 IbGLRs. There were 19 to 47 CAAT-box and 17 to 129 TATA-box core promoter elements in 37 IbGLRs. In addition to IbGLR35, other IbGLRs were found to have several hormone elements, such as the P-box for gibberellic acid (GA)-responsive elements, TCA for SA-responding elements, AuxRR-core together with TGA-element for indole-3-acetic acid (IAA)-responsive elements, CGTCA-motif and TGACG-motif for MeJA-responsive elements/ABRE for ABA-responsive elements ( Figure 4 ).

For development-related elements, the CAT-box related to meristem formation (found in IbGLR1/4/8/10/11/18/21/26/27/30/36), O2-site, a zein metabolism-regulatory element (observed in IbGLR1/5/10/11/12/20/21/28/30/31/35), I-box responding to light (found in IbGLR1/6/8/10/15/17/21/23/26/35), and GCN4_motif participating in regulating seed-specific expression (found in IbGLR7/10/11/12/15/18/34/37) were abundant in promoters of IbGLRs. The promoter regions of IbGLR2/16/19/22/32/33 did not contain any development-related elements ( Figure 4 ). Similarly, the promoters of 37 IbGLRs contained a number of light-responsive elements. IbGLRs were found to be rich in BOX4, G-boxes, GT1-motifs, and TCT-motifs. The ATC-motif was only present in IbGLR3, and 4cl-CMA2b only in IbGLR17 ( Figure 4 ).

Moreover, several abiotic elements, the MYC, MYB, and DRE core responded to drought stress, LTR, MBS, and W box responded to salt stress and biotic elements, the WUN, WRE3, and W box motifs were found in most IbGLRs ( Figure 4 ). This indicated that IbGLRs may have different roles under different stress conditions. The DRE core was only found in the promoter region of IbGLR18. The results demonstrate that IbGLRs take part in regulating the growth and development of plants, hormone crosstalk, and adaptation to abiotic and biotic stress in sweet potato.

An IbGLR interaction network was built through orthologous proteins from Arabidopsis to examine possible regulatory networking for IbGLRs ( Figure 5 ). The prediction of protein interactions suggested that some IbGLRs (IbGLR2/5/7/12/18/20/25) could interact with other IbGLRs. Moreover, IbGLR7, IbGLR17, and IbGLR20 were determined to interact with CNGC18, which was reported to regulate germination of pollen grains and the growth of pollen tubes (Chang et al., 2007; Cao et al., 2016; Gu et al., 2017). IbGLR17, IbGLR20, and IbGLR25 were determined to interact with GF14, which participates in transport, growth, metabolism, and the drought stress reactions (Li et al., 2016; Hajibarat et al., 2022). IbGLR2, IbGLR7, and IbGLR17 were determined to interact with TPK1, which participates in potassium transportation, fruit ripening, and quality formation (Lu, 2011; Latza et al., 2013; Wang et al., 2018). IbGLR2/3/12/20 were determined to interact with TPC1, which mediates Ca²⁺ release (Moccia et al., 2021; Navarro-Retamal et al., 2021; Pottosin and Dobrovinskaya, 2022), and IbGLR2/3/5/6/7/12/17/18/20 and IbGLR25 were determined to interact with AGB1, which regulates salt stress tolerance, fungal and bacterial immunity, and plant growth (Takahashi et al., 2018; Yu and Assmann, 2018; Zhang et al., 2020; Qi et al., 2021; Afrin et al., 2022). These findings indicate that IbGLRs have a distinct function in mediating biotic and abiotic stress and influence plant development in sweet potatoes.

Overall, I. trifida and I. triloba had the same number of GLRs (32 ItfGLRs and 32 ItbGLRs) identified, however they were less numerous than those (37 IbGLRs) in I. batatas. This is consistent with the fact that I. trifida and I. triloba are diploid and have a close genetic relationship. The evolution and differentiation of chromosomes were revealed by genomic alignment (Mukherjee et al., 2018). I. batatas, I. trifida, and I. triloba differed in terms of the distribution and proportion of GLRs in individual chromosomes. GLRs were located on nine chromosomes for I. batatas and I. triloba, however, GLRs were located on ten chromosomes for I. trifida. If different members of the same family are located within the same or adjacent intergenic regions, they have tandem repeat relationships (Freeling, 2009). Based on this standard, eight, six, and four tandem repeats were identified in I. batatas, I. trifida, and I. triloba, respectively ( Figure 1 ). This indicates that tandem repeats might be one of the reasons for the amplification of GLR genes in I. batatas, I. trifida, and I. triloba.

This research is based on the GLRs for I. batatas, I. trifida, and I. triloba, which were segregated into five groups (Groups I to V), four groups (Groups I to III and V), and five groups (Groups I to V), respectively. The fourth and fifth subgroups consisted of 10 GLRs that were absent in Arabidopsis. The type and number of GLRs in various subgroups for sweet potato and two diploid relatives are different from those in Arabidopsis. These findings imply that the genomes may have experienced lineage-specific differentiation for GLR gene family.

Introns are important components of eukaryotic protein-coding genes, eliminated during messenger RNA precursor molecule splicing. They are characterized by their clear organization and abundance in eukaryotic genes (Rogozin et al., 2005; Mukherjee et al., 2018). Because of the presence of introns, gene expression in eukaryotic cells is much more complex than that in prokaryotic cells. Under stressful conditions, introns assume a key role in regulating cell growth (Morgan et al., 2019). Herein, in comparison with I. trifida and I. triloba, I. batatas had distinct exon-intron patterns for some homologous GLRs ( Figure 3C ). For example, IbGLR36, which was expressed in fibrous roots, contained 17 introns, whereas its homologous genes ItfGLR7 and ItbGLR7 contained nine and ten introns, respectively, and were expressed in leaves. Moreover, IbGLR30, which was expressed in the stem, contained 17 introns, whereas ItfGLR8 and ItbGLR8 contained four and five introns, respectively, and were expressed in flowers ( Figures 3C , 6 ). The corresponding differences in exon–intron structure between sweet potato and its two diploid relatives might lead to different functions for GLRs in various aspects of plant growth and development (Pang et al., 2018; Ma et al., 2019; Ma J et al., 2020).

Plant GLRs are actively involved in hormone biosynthesis and signal transduction concerning the regulation of plant growth and responses to stress (Weiland et al., 2015). PpGLR1 is involved in ABA-mediated growth regulation in Physcomitrium patens (Wang et al., 2022). Meanwhile, RsGluR could serve as a defensive mechanism against pathogen infection by triggering MeJA biosynthesis (Kang et al., 2006). In this work, the promoters of the IbGLRs contained 11 hormone-responsive elements, 12 developmental elements, 19 light-responsive elements, and 21 abiotic/biotic-response elements. In addition to IbGLR35, other IbGLRs comprised at least one hormone element ( Figure 4 ). IbGLR3 only contained the P-box of the GA-responsive elements, whereas its homologous gene ItbGLR18 was determined to be induced by ABA. IbGLR4 was found to contain the TGACG-motif and CGTCA-motif for MeJA-responsive elements, TCA for SA-responsive elements, and the AuxRR-core together with TGA-element for IAA-responsive elements, whereas its homologous gene, ItbGLR31, was determined to be induced by GA and IAA. IbGLR21 contained ABRE for ABA-responsive elements, the TGACG-motif together with CGTCA-motif for MeJA-responsive elements, and TCA for SA-responsive elements, whereas its homologous genes, ItfGLR21 and ItbGLR22, were both determined to be induced by ABA and GA. IbGLR30 carried the TGACG-motif and CGTCA-motif for MeJA-responsive elements, whereas its homologous gene, ItbGLR8, was determined to be induced by ABA ( Figures 4 , 11 ). Such results demonstrate GLRs involvement in the interplay of several hormones and the participation of homologous GLR genes in numerous hormone pathways in sweet potato and two diploid relatives. Further research is needed to fully understand how GLRs control hormonal crosstalk.

GLR plays an indispensable role in the response of plants to different biological stresses (He et al., 2016). Liu and colleagues discovered a point mutation for GhGLR4.8 exon that can increase the resistance of upland cotton to Fusarium wilt, while knocking out the GhGLR4.8 lead the defense ability of cotton cell wall against G. hirsutum Fov race 7 to weaken (Liu et al., 2021). AtGLR3.3 also has a valuable function in the defense response induced by Pseudomonas syringae pv tomato DC3000. An AtGLR3.3 mutant was shown to exhibit high sensitivity to DC3000, and the response of defense genes was reduced in mutant lines induced by pathogenic bacteria (Li et al., 2013). Sweet potato root rot infection usually begins at the underground stem and tip or middle of fibrous roots. Severely diseased plants do not produce tuberous roots, whereas slightly diseased plants produce tuberous roots with diseased spots (Ma Z et al., 2020).

Transcriptome and qRT-PCR analysis showed that IbGLR18 expression was induced in Jishuzi203 (root rot-resistant variety) and repressed in Jishuzi563 (root rot-sensitive variety) after root rot infection. In addition, the IbGLR18 expression in fibrous roots was higher in comparison to that in other tissues (leaf, stem, and tuberous root). Furthermore, IbGLR18 expression was also higher in the fibrous root (diameter of root 1 mm) than that in tuberous roots (diameter of root 1/3/5/10 cm) in the five developmental stages of the Xushu22 root. Therefore, we speculate that IbGLR18 might be involved in resistance to root rot and the formation of fibrous roots in sweet potato, but its function should be verified in the future studies.

Usually, tuberous roots are the main harvesting tissues for hexaploid sweet potato. I. trifida and I. triloba are not capable of forming tuberous roots (Wu et al., 2018). The expression for IbGLR29 in tuberous roots was higher than that in other tissues (leaf, stem, and fibrous roots) of sweet potato, and its homolog genes ItfGLR9 and ItbGLR9 were highly expressed in the leaf. Further, in the five developmental stages of Xushu22 roots, the expression of IbGLR29 in tuberous roots (root diameter of 1/3/5/10 cm) was high in comparison to that in fibrous roots (diameter of approximately 1 mm). IbGLR29 may therefore be involved in the generation of tuberous roots in sweet potato.

GLRs play a crucial role in mediating plants responses to abiotic environmental stress (He et al., 2016). The degree of expression of ZmGLR2.3/3.1 demonstrated a significant increase after drought stress (Zhou et al., 2021). In this work, MYB and MYC were found to respond to drought stress, MBS to salt stress, and LTR to cryogenic stress, and these were identified in the IbGLR19 promoter, which was rapidly expressed after drought stress and reached its expression maximum at the early stage (1 h). Moreover, its expression was upregulated by NaCl treatment in ND98 (salt-tolerant line) and downregulated in Lizixiang (salt-sensitive variety, Figures 4 , 9 ). ItfGLR23, the homolog gene of IbGLR19, was induced by both drought and salt treatments in I. trifida ( Figure 10 ). These results suggest that IbGLR19 may contribute to abiotic stress. Additionally, I. trifida and I. triloba may be utilized for searching and identifying functional genes, particularly those that provide tolerance or resistance to biotic and abiotic stresses, which may have been lost during the domestication of cultivated sweet potato (Cao et al., 2016). In I. trifida and I. triloba, several genes (ItfGLR8 and its homologous gene ItbGLR8, ItfGLR17 and its homologous gene ItbGLR20, ItfGLR23 and its homologous gene ItbGLR24, and ItfGLR28 and its homologous gene ItbGLR29) exhibited the same expression patterns and were induced under 10/4°C (day/night) ( Figure 12 ). ItfGLR3 and its homologous gene ItbGLR3 were induced by drought and salt treatments ( Figure 10 ). In summary, the GLRs induced in I. trifida and I. triloba could serve as candidate genes for enhancing the abiotic stress resistance of sweet potato.