Genome-wide analysis of lectin receptor-like kinases family from potato (Solanum tuberosum L.)

Annotation of LecRLKs

The amino acid and nucleotide sequence data of Potato (Solanum tuberosum) v3.4, Arabidopsis thaliana Araport11, Rice (Oryza Sativa Japonica Group) IRGSP-1.0, Apple (Malus domestica) GDDH13 v1.1, and Tomato (Solanum lycopersicum) SL2.50 were downloaded from the Potato Genome Sequencing Consortium (PGSC) Public Data (http://solanaceae.plantbiology.msu.edu/pgsc_download.shtml), The Arabidopsis Information Resource (TAIR) (https://www.arabidopsis.org), the Ensembl Genomes (http://ensemblgenomes.org/), the genome database for Rosaceae (GDR) (https://www.rosaceae.org/), and the Solanaceae Genomics Network (https://solgenomics.net/), respectively. Proteins contained kinase domain (PF00069) or Pkinase_Tyr (PF07714) in above five species were following the method described by Zhu et al. (2018). After that, sequences contained a N-terminal domain B_lectin (PF01453), Lectin_legB (PF00139) or Lectin_C (PF00059) were considered as LecRLKs. All of the candidate proteins were determined by using online software Pfam (http://pfam.xfam.org/) and SMART (http://smart.embl-heidelberg.de/) (E-value < 10⁻⁵) to ensure the presence of N-terminal lectin domain, a transmembrane region and C-terminal kinase domain in each putative member.

Multiple sequence alignment, phylogenetic analysis, physical and chemical properties and subcellular localization classification

The sequences of all LecRLKs from six species were aligned using ClustalX 2.0 with default parameters (Larkin et al., 2007) and a phylogenetic tree was constructed using the Neighbor-Joining (NJ) method in MEGA 7 (Kumar, Stecher & Tamura, 2016). The parameter partial deletion and P-distance Model were used with 1,000 bootstrap replicates. The protein theoretical molecular weight (MV) and theoretical isoelectric point (pI) were predicted using ProtParam (https://web.expasy.org/protparam/) (Gasteiger et al., 2005). The subcellular localization of proteins was predicted using the WOLF PSORT (https://www.genscript.com/wolf-psort.html) and CELLO v2.5: subcellular Localization predictor (http://cello.life.nctu.edu.tw/).

Analysis of chromosomal distribution, gene duplication and Gene Ontology overview

The chromosomal positions of potato LecRLKs (StLecRLKs) were obtained from PGSC (http://solanaceae.plantbiology.msu.edu/pgsc_download.shtml) and the chromosomal distribution were visualized by using software MapDraw 2.1 (Liu & Meng, 2003). Tandem duplicated genes are defined as those genes that (1) are ≤l0 genes apart, (2) belong to the same phylogenetic group, and are (3) within 100 kb (Hofberger et al., 2015). The Whole-Genome Duplication (WGD)/segmental duplication was calculated by MicroSyn software (Cai et al., 2011). The projected GO annotation for the candidate StLecRLK genes was then analyzed for enrichment of GO terms using GOEast with default settings (Zheng & Wang, 2008). For cis-acting element analysis, genomic DNA sequences in the promoter region (−1,500 bp) were downloaded from PGSC and then scanned for in the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) (Lescot et al., 2002).

Expression patterns of StLecRLKs in response to biotic infection

To investigate the responses of StLecRLKs to biotic stress, expression data were obtained from the published studies, including potato infected with pathogenic bacteria Candidatus Liberibacter solanacearum (Cls) (Levy et al., 2017), Pectobacterium carotovorum subsp. brasiliense (Pcb) (Kwenda et al., 2016) and Ralstonia solanacearum (Rs) (Zuluaga et al., 2015), fungi Phytophthora infestans (Pi) (Yogendra & Kushalappa, 2016), and Fusarium sulphureum (Fs) (He et al., 2018). The expression pattern was extracted by the accession number of StLecRLKs. False Discovery Rate (FDR) < 0.05 and |log2 (fold change)| ≥ 1 were used as a threshold to judge the significance of the difference in gene expression. The heatmap was visualized by using software Multi experiment viewer software (MeV 4.9.0) (http://mev.tm4.org/).

Plant materials, treatment and gene expressional assay

Potato cultivar Helan 15 (Favorita, Qingshenshu 2007001) were cultured in the greenhouse with 24 °C. After 5 weeks cultivation, leaf tissues were collected, sterilized with 75% ethanol and transferred into a petri dish with a diameter of 9 cm. To maintain humidity, a wet filter paper was placed in each petri dish. Pi strain “HB09-16-2 (race 1.2.3.4.5.6.7.9.10.11, collected from Hubei Province, China)” were friendly supplied by Prof. Tian Zhendong, Huazhong Agricultural University. The pathogen was cultured at 19 °C for 2 weeks. The conidia were collected and adjusted to a concentration as 7 × 10⁴ sporangia per milliliter. Inoculation was performed following the method described by He et al. (2018). For the control, the conidia suspension was replaced with the same volume of sterilized water. Inoculated and control tissues were collected in triplicate at 1 d, 4 d, 6 d, and 8 d after inoculation and quick-frozen in liquid nitrogen and stored at −80 °C.

The method of inoculation of Fs was performed as described by Li et al. (2009). Longshu No. 7 (Guoshenshu 2009006), without physical injuries or visible infection, was used as the material for the inoculation for 1 d, 2 d, 3 d and 4 d. Before treatment, tubers were surfaced-disinfected with 2% sodium hypochlorite for 3 min, and then rinsed with tap water and air-dried.

RNA isolation, reverse-transcription and qRT-PCR were performed following the method described by Zuo et al. (2018). Sequences of primers and actin genes could be found out in File S1. The relative expression levels of each gene were calculated based on the 2^−ΔΔCT method (Livak & Schmittgen, 2001).

The above samples at different treatment time were collected from three independent replicates. Each experimental unit consisted of three plants or tubers, three leaves per plant and each leaf or tuber with one inoculation site. The data were statistically analyzed using software Origin 8.0 and the significant differences were analyzed using the t-test (*P < 0.05; **P < 0.01).

LecRLKs annotation and classification

Based on domain screening, a total of 113 transcripts encoding LecRLKs were identified in potato, including 85 G-type, 26 L-type and 2 C-type members (File S2). For some transcripts, two or more transcripts had the same genome loci and corresponded with the same gene accession, but most of these showed the distinct sequences. Therefore, we determined the number of LecRLKs as the transcript’s numbers. With the same criteria, 80, 76, 139 and 167 LecRLKs were annotated in Tomato (Solanum lycopersicum), Arabidopsis, Rice (Oryza sativa) and Apple (Malus domestica), respectively. From the published study, we also re-identified the LecRLKs in different plant species, including Populus, Eucalyptus, Shrub, Soybean, Corn and Moss (Table 1; Yang et al., 2016). To ensure the uniformity of the obtained data, domains of all sequences were determined based on the same criterion as above. We excluded several proteins which did not contain typical domains of LecRLKs (File S2). Compared to Moss, the member of G-type and L-type LecRLKs was much larger in other plants, and the number of G-type LecRLKs was exceed that of L-type LecRLKs except for Shrub and Arabidopsis. There were only one or two C-type LecRLKs in all the detected plants. Over half of LecRLKs were classified as G-type except in Shrub, Arabidopsis, Corn and Moss. In potato, tomato, apple and Populus, G-type LecRLKs accounted for more than 70% of the LecRLKs, which suggested that the G-type LecRLKs were rapidly expanded in potato, tomato and Populus than in the other detected plants.

To further investigate the classification and evolution relationships of LecRLKs, phylogenetic tree of three independent types (G-type, L-type and C-type) were constructed separately based on the multiple sequence alignment of the full-length amino acid sequences from five plant species, including potato, tomato, apple, Arabidopsis and rice. For G-type LecRLKs, the members could be subdivided into five clades (G-I to G-V) (Fig. 1). Clades G-I, G-II, G-IV and G-V contained LecRLKs family members from all the tested plants, while G-III only contained members from rice. Potato G-type LecRLKs were discovered from four clades except G-III. There were 28, 3, 25 and 29 members in G-I, G-II, G-IV and G-V clade which accounted for 25.93%, 7.14%, 25% and 26.61% in each subgroup respectively.

L-type LecRLKs was subdivided into eight clades (Fig. 2). L-V had two Arabidopsis thaliana LecRLKs. L-VI contained members from four investigated plant species except rice. Other subgroups contained members from all five species. Potato LecRLKs located in six clades except L-I and L-V. C-type LecRLKs could be divided into four subgroups (Fig. S1), in which subgroup IV contained members from potato and tomato. For the other subgroups, each of these only contained members from one single plant species. Taken together, these results indicated the distinctly evolutional process of LecRLKs in plants. The division of each clade was supported by high bootstrap values.

StLecRLKs classification, domain organization, physicochemical properties and gene ontology analysis

To further confirm the classification and evolutionary relationship of StLecRLKs, phylogenetic tree generated from the alignment result of the full-length amino acid sequences (Fig. 3; File S3). StLecRLKs could be divided into seven clades, including C-type, L-type, G-I, G-II, G-III, G-IV and G-V, which contained 2, 26, 28, 3, 8, 15 and 31 members, respectively.

Domain organization was also investigated for each subgroup of StLecRLKs (File S2). Unique N-terminal domain was discovered from both L-type (Lectin_legB) and C-type members (Lectin_C). However, complex domain composition was found from G-type LecRLKs. Besides B_lectin domain, multiple additional N-terminal domains were detected, such as S_locus_glycop (PF00954), PAN_2 (Pfam accession number was not detected) and EGF (SM000181). Additionally, two intracellular domains, DUF3660 (PF12398) and DUF3403 (PF11883) were observed from some members. Domain DUF3403 was in C-terminal of specific protein, but its function is still unknow. Above all, G-type StLecRLKs, being the largest group, showed a more complex evolutionary relationship and domain organization than the other types.

After determining the conserved domains of StLecRLKs, the physicochemical properties and subcellular localization of these proteins were also analyzed (File S4). The results showed that 113 StLecRLKs had an amino acid size ranging from 419 to 1,504 aa, a molecular weight ranging from 47.07 to 169.51 kDa, and pI value ranging from 5.02 to 9.26. The subcellular localization of 113 StLecRLKs were predicted by the online software CELLO and PSORT. 87 StLecRLKs were located in the plasma membrane, two in the cytoplasm, one was in the extracytoplasmic surface and three in both plasma membrane and cytoplasm. Moreover, there were 20 StLecRLKs with inconsistent prediction results from the two software. Further, 15 and 20 GO terms were distributed into molecular function and biological process (File S5). Within biological process, a larger number of family members involved in “metabolism process”, “cell recognition”, and “protein modification process”. For the molecular function categories, genes were mainly associated with “kinase activity”, “catalytic activity”, “transferase activity” and “binding”.

Chromosomal distribution, gene duplication

Tandem duplication mechanism represents a common role in the expansion of RLK family. To identify the gene clusters and to investigate such events, we scanned the gene ID and the corresponded protein ID of StLecRLKs based on PGSC database (File S2). For several accessions, unique gene accession was corresponded to two or more protein accessions. For instance, gene PGSC0003DMG400019237 encoded three proteins, including PGSC0003DMP400033418, PGSC0003DMP400033419 and PGSC0003DMP400033420. Finally, 88 of the 113 StLecRLKs was visualized on chromosome across 12 chromosomes (Fig. 4). There were 15, 10, 9, 20, 11 and 7 members scattered on Chr 02, 03, 04, 07, 09 and 10, respectively. Additionally, 15 tandem duplicated gene clusters (marked in red vertical lines) and five WGD/segmental duplicated gene pairs (marked in blue dotted lines) were found from StLecRLKs. The tandem duplicated gene clusters contained 47 StLecRLKsconsisting of 40 G-type (black) and 7 L-type (green). On the other hand, WGD/segmental duplicated gene pairs contained seven G-type and two L-type StLecRLKs. Above results indicated that tandem duplications played important roles in the amplification of the StLecRLK gene family.

Cis-elements involved in StLecRLKs

In order to further explore the functions of the candidate StLecRLKs, a 1,500 bp promoter region was extracted and PlantCARE was used to analyze the cis-elements related to stress challenge. A great number of cis-elements responsive to defense and stress, phytohormines (abscisic acid (ABA), methyl jasmonate (MeJA) and salicylic acid (SA)), biotic stressors (elicitor) and abiotic stressors (drought, low temperature and wounding) were identified in the promoter region of StLecRLKs (Fig. 5; File S6). Of these, the number of genes which had the cis-element involved in ABA-, MeJA- and SA-responsiveness were 57, 57 and 34, respectively. Most of the StLecRLKs belonged to the phylogenetic group G-I, G-V and L-type. Additionally, some StLecRLKs had elements responsive to abiotic stress, and several members possessed elements responsible for elicitor. The above results indicated that StLecRLKs had a potential role in response to signals from phytohormones and multiple stresses (Fig. 5).

Expression patterns of StLecRLKs in response to biotic stresses

Based on the published transcriptomic datasets, the expressional patterns of StLecRLKs were investigated during potato infected with phytopathogenic bacteria (Levy et al., 2017; Kwenda et al., 2016; Zuluaga et al., 2015) and fungi (Yogendra & Kushalappa, 2016; He et al., 2018). For the bacteria, expression changes of StLecRLKs were extracted after potato inoculated with Cls, Pcb and Rs (Fig. 6). A total of 55 StLecRLKs were differently expressed, including 15 members in G-I, 2 in G-III, 10 in G-IV, 17 in G-V and 11 in L-type, respectively. Furthermore, 28 StLecRLKs were detected to be up-regulated under at least one bacterial infection.

Additionally, expression patterns of StLecRLKs were extracted from published transcriptomic data of potato tuber (Gao et al., 2013) and leaf (Yogendra & Kushalappa, 2016) inoculated with Pi (Fig. 7). A total of 26 members, including nine members in G-I (32.14%), two in G-III (25.0%), two in G-IV (13.33%), five in G-V (16.13%) and eight in L-type (30.77%) were differentially expressed (Fig. 7). PGSC0003DMP400030282 showed the same expressional pattern among resistant tuber and leaves. However, PGSC0003DMP400032146 and PGSC0003DMP400055136 in G-I and PGSC0003DMP400045382 in G-IV were detected to be down-regulated under bacteria attack while they were up-regulated under fungi infection. PGSC0003DMP400067047 in G-IV was down-regulated under both bacteria and fungi infection (Figs. 6 and 7). Combined with the above results, a great number of G-I, G-V and L-type StLecRLKs were differentially expressed after potato tissue inoculated with both phytopathogenic fungi and bacteria, which indicated the potential roles of these members in potato disease resistance.

Expression patterns of StLecRLKs during potato infected with Fs and Pi

To further validate the regulating roles of the candidate StLecRLKs to fungal disease resistance, the expression analysis of StLecRLKs during potato inoculated with Fs (Fig. 8) and Pi (Fig. 9) were detected. Two or three members in each phylogenetic group were selected, which were strongly affected under the biotic stress treatment (Figs. 6 and 7). Finally, 16 differently expressed members were selected to perform qRT-PCT analysis (File S7). During Fs infection, 10 of the 16 detected StLecRLKs were differentially expressed at least one time point (Fig. 8; File S7), and they were detected to be up-regulated after treatments. The expression of PGSC0003DMP400067047 was gradually upregulated during the experiment, and the expression level was 26.59 times that of the control group at 4 d treatment. When the leaves were inoculated with Pi, 12 representative genes were differentially expressed (Fig. 9; File S7). PGSC0003DMP400041968, PGSC0003DMP400049549 and PGSC0003DMP400055136 were significantly induced at the later stage (T4), and the expressional levels were 12.33, 430.80 and 34.08 times that of the control group. PGSC0003DMP400030282, PGSC0003DMP400001977 and PGSC0003DMP400045383 were significantly induced when the leaves were treated for 6 d (T3).

Evolution and expansion of StLecRLKs

Besides the differences in genome size, expansion rate is one of the most important contributors to the copy numbers of plant LecRLKs (Liu et al., 2018). Compared to Moss, total numbers of LecRLKs were drastically expanded in higher plants (Table 1), which mainly resulted from the expansion of the G-type LecRLKs members, followed by L-type. In potato, tomato, apple and Populus, about three-quarter of LecRLKs were identified as G-type members, although L-type LecRLKs were more expanded than G-type members in several species such as Shrub, Arabidopsis and Corn (Table 1; Vaid, Pandey & Tuteja, 2012). The results showed that different plant species had different numbers of LecRLKs, which was possibly related with the different growth condition and growth behaviors, or maybe with the plant life cycles and the mode of reproduction. Yang et al. (2016) also speculated that the difference in the number of G-type and L-type LecRLKs between Arabidopsis and Populus may be related with the fact that Arabidopsis is a self-fertile plant whereas Populus is an obligate outcrossing plant. In our study, the number of LecRLK in Arabidopsis and rice is different from the previous researchers, which may be due to the perfection of genome sequence or the difference of identification standards (Table 1).

Phylogenetic analysis revealed the distinct evolutionary process in plant species (Figs. 1 and 2; File S1). The rice specific clade was found both in G-type (G-III) and L-type (L-I) LecRLKs. Therefore, we suggested the distinct expanded characteristics between G-type and L-type LecRLKs in plant species. Compare to L-type LecRLKs in potato, more G-type members were found, indicating the crucial roles of G-type LecRLKs in development and environmental response.

Gene duplications and functional analysis

Duplication can lead to functional divergence of genes, which is a universal phenomenon in plants (Liu et al., 2012). LecRLKs duplication were mainly resulted from tandem duplication and segmental duplication/WGDs (Yang et al., 2016; Hofberger et al., 2015; Shumayla Sharma et al., 2016). In potato, 40 G-type StLecRLKs were involved in tandem duplication but only seven members in WGD/segmental duplication (Fig. 4). Similarly, there were more tandem duplicated L-type StLecRLKs than that in segmental duplication/WGD (Fig. 4). Hence, the tandem duplication might be the main reason for expansion of both G-type and L-type StLecRLKs. Compared to RLKs involved in development, stress responsive RLKs showed a higher number of tandemly duplicated genes. Based on the above conclusion, we deduced that StLecRLKs also took important roles in potato responses to biotic and abiotic stress. Furtherly, we used GO method to analyze the molecular function of these genes. The results showed that the enriched GO terms were kinase activity, catalytic activity, transferase activity and binding (phosphate binding and carbohydrate binding) (Fig. 5). Since chitin and β-1,3 glucans are components of the cell walls of some higher fungi, some plant lectins sharing chitin-binding domains can recognize the signal then naturally these substrates will be hydrolyzed and fungal growth will be disrupted (Sahai & Manocha, 1993). This type of lectin could inhibit the growth of several phytopathogenic and saprophytic chitin containing fungi in vitro (Broekaert et al., 1989). Moreover, the cis-element prediction furtherly demonstrated that StLecRLKs have the potential role against multiple stresses (Fig. 5; File S6). Previous study confirmed LecRLKs participated in the regulation of abiotic stress (Vaid, Pandey & Srivastava, 2015; Singh & Zimmerli, 2013; Xin et al., 2009) and biotic stress (Chen et al., 2006; Wang et al., 2015).

Complex domain architecture and organization of G-type StLecRLKs

Protein domains are some complex regions of a protein’s structure, which often perform some specific function (Buljan & Bateman, 2009). Architecture and composition of domain of G-type StLecRLKs were observed (File S2). In addition to B_lectin domain, most G-type members contained an S_locus_glycoprotein, EGF and PAN_2 or PAN-AP domain, in which PAN domains are involved in interactions of protein–protein and protein–carbohydrate (Naithani et al., 2007). Furthermore, some G-type StLecRLKs also contained intracellular domain DUF3660 or DUF3403. Our result was consistent with the findings in Populus although the function of the two domains was largely limited until now (Yang et al., 2016). Nevertheless, the complex domain architecture suggested the diverse functions of G-type StLecRLKs.

StLecRLKs in response to biotic stresses

Increased investigations reported the crucial role of both G-type and L-type LecRLKs in plant immunity (Zhao et al., 2018; Wang et al., 2015, 2014). Over-expression of LecRK-V (L-type) significantly enhanced the resistance of Yangmai158 to powdery mildew (Wang et al., 2018). Similar studies have been done in rice and tomatoes. Pid2 (G-type LecRLK in rice) and NbLecRK (L-type LecRLK in tomato) were shown to increase the resistance to fungal pathogen and Phytophthora, respectively (Chen et al., 2006; Wang et al., 2015). In this study, we found a large number of StLecRLKs were differentially expressed during potato infected both bacterial and fungal pathogens (Figs. 6 and 7). Most of these were distributed in clades G-I, G-V and L-type. Genes which contributed to plant resistance to a wide of pathogens could be classified as broad-spectrum resistant genes (Dangl & Jones, 2011; Xiao et al., 2004). In Populus, the root-expressed PtLecRLKs maybe have the potential function for perceiving signals from soil microbes (Yang et al., 2016). Intriguingly, the expression of PGSC0003DMP400030282 was induced under signals from both phytopathogenic bacteria and fungi, which suggested that the crucial roles in broad-spectrum disease resistance of potato, while PGSC0003DMP400055136 and PGSC0003DMP400067047 showed distinct expression, indicating a different role in response to different pathogen signal (Figs 7–9).