Expression analysis of plant intracellular Ras-group related leucine-rich repeat proteins (PIRLs) in Arabidopsis thaliana

Arabidopsis thaliana contains a family of nine genes known as plant intracellular Ras-group related leucine-rich repeat (LRR) proteins (PIRLs). These are structurally similar to animals and fungal LRR proteins and play important roles in developmental pathways. However, to date, no detailed tissue-specific expression analysis of these PIRLs has been performed. Therefore, in this study, we generated promoter:GUS transgenic plants for the nine A. thaliana PIRL genes and identified their expression patterns in seedlings and floral organs at different developmental stages. Most PIRL members showed expression in the root apical region and in the vascular tissue of primary and lateral roots. Shoot apex-specific expression was recorded for PIRL1 and PIRL8. Furthermore, PIRL1, PIRL3, PIRL5, PIRL6, and PIRL7 showed distinct expression patterns in flowers, especially in pollen and anthers. In addition, co-expression network analysis identified cases where PIRLs were co-expressed with other genes known to have specific functions related to growth and development. Taken together, the tissue-specific expression patterns of PIRL genes improve our understanding of the functions of this gene family in plant growth and development.


Introduction
Leucine-rich repeat (LRR) proteins contain a characteristic LRR domain consisting of 18-29 amino acid repeats rich in leucine residues. These proteins are widespread in plants, animals, and bacteria [1,2]. In animals and lower eukaryotes, an intracellular LRR protein subfamily known as Ras-group LRR proteins with a variably long LRR domain are functionally involved in signal transduction [3][4][5][6][7][8]. Plants contain a small group of proteins that are structurally analogous to Ras-group LRRs, named plant intracellular Ras-group related LRR proteins (PIRLs) [4,9].
The Arabidopsis thaliana genome has nine PIRL genes that are categorized into three subfamilies based on their evolutionary relationship [4]. To date, the functions of five PIRLs have been identified. PIRL6 has been found to be essential for male and female gametogenesis [10], while PIRL1 and PIRL9 redundantly work during the differentiation of microspores into pollen [11]. In addition, PIRL2 and PIRL3 have been found to be involved in pollen morphogenesis [12]. However, to elucidate the physiological functions of the remaining PIRL genes, it is requisite to know their precise expression patterns.
In this study, we used a promoter:GUS assay to examine in detail the expression patterns of PIRL genes in A. thaliana. We found that PIRL genes were expressed in the root, root hairs, shoot apex, leaves, anther, pollen, and pollen tube. We also performed a phylogenetic analysis of A. thaliana PIRL proteins with homologous PIRL proteins from other plants and in silico analyses examining the co-expression of PIRL genes. The results of these experiments should improve our understanding of the function of PIRL genes during plant growth and development.

Plasmid construction
The promoter entry clones pDONR201-ProPIRL1 to pDONR201-ProPIRL9 were prepared with the primers listed in Supplementary  Table 2. The promoter:GUS binary constructs were prepared with pDD333 and pGWB3450 vectors according to the method described by Sultana et al. [15].

Generation of transgenic A. thaliana
Transformation of Agrobacterium tumefaciens C58C1 (pMP90) was performed using the freeze-thaw method as per Weigel and Glazebrook [16]. Transformation of A. thaliana (Col-0 accession) was performed using the floral inoculation method [17]. Harvested T0 seeds were kept at 4 • C for 3 days, after which they were grown on Murashige and Skoog (MS) agar medium containing kanamycin (30 mg/L) and Cefotax (100 mg/L) (Chugai Pharmaceutical Co., Tokyo, Japan) at 22 • C under continuous light. Fifteen-day-old selected plants (T1) were transferred to Jiffy-7 plant pots (Jiffy Products International BV, Zwijndrecht, Belgium) and were grown at 22 • C under a 16-hr light/8-hr dark daily cycle.

Histochemical GUS staining
Plants were selected at different time points for histochemical staining, including 1-day-after-germination (DAG) seedlings, as well as plants at 4-DAG, 8-DAG, 16-DAG, and the flowers of 40-DAG plants. Staining was performed as described by Nakamura et al. [18] and plants were observed using a stereo microscope (SZX16, Olympus, Tokyo, Japan) equipped with a Nikon digital camera (DXM1200F) as well as an all-in-one microscope (BZ-710, Keyence, Osaka, Japan).

In silico expression and co-expression of PIRL genes
The expression value was obtained for each PIRL gene in several different tissues (cotyledon, hypocotyl, root, leaf, flower, sepal, petal, and pollen) using microarray data from the Arabidopsis eFP browser [19]. Expression values were then normalized using the log2 transformation. Finally, a heat map was generated by TBtools software [20]. Microarray data showing visual expression during different developmental stages was also retrieved from the Arabidopsis eFP browser [19]. The co-expression gene network of PIRLs and co-expressed genes were analyzed using the ATTED-II (https://atted.jp/) [21].

Phylogenetic analysis shows a relationship between A. thaliana PIRLs and PIRLs found in other species
Phylogenetic analysis of amino acid sequences from A. thaliana (PIRL), Brassica rapa (BrPIRL), Capsella rubella (CrPIRL), Oryza sativa (OsPIRL), Hibiscus syriacus (HsPIRL), Zea mays (ZmPIRL), and Ricinus communis (RcPIRL) showed that, the PIRLs from A. thaliana showed some similarity to those from other plants. Our tree revealed that the phylogenetic relationships between and among species were clustered into six groups (Fig. 1A). In group I, PIRL4 and PIRL5 formed a cluster with CrPIRL4 and CrPIRL5, with bootstrap values of 93 and 87, respectively. The highest number of clusters was found in group III, where PIRL6 formed a cluster with BrPIRL6, and PIRL7 and PIRL8 formed a cluster with CrPIRL7 and CrPIRL8, respectively. Group IV contained PIRL2 and PIRL3, which formed a cluster with CrPIRL2 and CrPIRL3. Finally, PIRL1 and PIRL9 formed a cluster with BrPIRL1 and CrPIRL9, and these were grouped together into Group V. Next, we predicted the number of LRR in each A. thaliana PIRL ( Fig. 1B

PIRL expression in different vegetative tissues in A. thaliana
We subjected T2 or T3 plants from ProPIRL:GUS transgenic lines to histochemical GUS staining at different developmental stages. As shown in Fig. 2, seedlings (1-, 4-, 8-, and 16-DAG) exhibited different expression patterns in their vegetative organs. For PIRL1, GUS expression was observed at the junction between the root and hypocotyl at 4-DAG, at the primary root tip and in the differentiated region of the primary and secondary roots at 8-DAG, and in the shoot apex at 16-DAG ( Fig. 2A). For PIRL2, we observed unique expression in root hair, and while lower expression was observed in the vascular tissues of the primary and secondary roots from 1-to 8-DAG, we also observed moderate expression in leaves and strong expression in stipules at 16-DAG (Fig. 2B). PIRL3, PIRL4, PIRL7, and PIRL9 showed two similar trends: for each of these PIRL genes, we observed GUS expression in root (1-to 8-DAG) but faint or no signal in aboveground organs at 16-DAG ( Fig. 2C, D, G and I). For PIRL5, we observed expression in the root tip and in the vascular tissue of the first and secondary roots at 4-and 8-DAG, but again almost no expression in aboveground organs at 16-DAG (Fig. 2E). PIRL8 showed GUS expression in roots (8-DAG), the shoot apex region, and young leaves (Fig. 2H). We did not observe considerable expression of PIRL6 in any vegetative organs (Fig. 2F). All the results of GUS staining experiments are summarized in Supplementary Table 4.

PIRLs are expressed in various reproductive tissues in A. thaliana
We subjected the flowers of 40-DAG transgenic plants to GUS staining. For PIRL1, GUS was weakly expressed in petals and filaments, but was strongly expressed in anthers and pollen (Fig. 3A). Expression was also detected in the pollen tube (Fig. 3A). For PIRL2, expression was strong in flower buds but was lower in sepals, petals, filaments, and stigmas of open flowers (Fig. 3B). We observed distinct expression of PIRL3 in anthers and pollen, but not in the pollen tube (Fig. 3C). Next, for PIRL4, we did not find any expression in any floral organs whatsoever (Fig. 3D). Similarly, PIRL5 was detected very weakly and only in anthers (Fig. 3E). However, clear GUS expression was detected in both  I, II, III, IV, V,  and VI are denoted by blue, fuchsia, purple,   pollen and pollen tubes for both PIRL6 and PIRL7 (Fig. 3F and G). GUS activity was not observed in any floral organ for either PIRL8 or PIRL9 ( Fig. 3H and I). All GUS staining results are summarized in Supplementary Table 4.

In silico expression and co-expression analysis of PIRLs
We generated a heat map that used microarray datasets of PIRL genes sourced from the Arabidopsis eFP browser (Fig. 4A). This heat map analysis showed that PIRL6 was specifically expressed in pollen  PIRLs are shown on the right side, and tissues are indicated at the bottom. Red indicates a higher level of gene expression, while green indicates lower expression. (B) Co-expression networks were generated by the ATTED-II using PIRL as query genes. Red, yellow, blue, sky-blue, and green circles indicate groups of genes involved in different metabolic pathways. (Fig. 4A), that PIRL1 and PIRL3 were expressed strongly in pollen and moderately in vegetative organs (Fig. 4A), and that PIRL5 was expressed in pollen and roots (Fig. 4A). This analysis also showed that both PIRL4 and PIRL9 were expressed in a variety of organs, although PIRL4 was not expressed in pollen (Fig. 4A). Interestingly, weak expression in root tissue was observed for PIRL8 (Fig. 4A). Moreover, a microarray study using visual expression of PIRL genes during developmental stages revealed that PIRL1 and PIRL3 were significantly expressed in floral organs, specifically in mature pollen rather than vegetative organs ( Supplementary Fig. S1). According to the microarray expression developmental map, PIRL4 and PIRL9 were expressed in a number of regions, while no expression was found for PIRL4 in pollen (Supplementary Fig. S1). PIRL5 and PIRL6 showed vigorous expression in mature pollen, whereas root-specific expression was also observed for PIRL4 and PIRL8, respectively. The microarray datasets contained no expression values for PIRL2 and PIRL7. Further, to identify the functional relationship between PIRLs and other genes, we generated a coexpression network using the ATTED-II (ath-m.c9-0 platform) (Fig. 4B). Of the nine PIRL genes, only PIRL1 and PIRL3 were found to co-express with each other. Furthermore, PIRL1 expression was correlated with the expression of VQ-motif containing protein 9 (VQ9, At1g78310), which is a class of transcription factor [22], and zinc finger nuclease 2 (ZFN2, AT2G32930) (Fig. 4B). We also found that PIRL3 co-expressed strongly with a member of the Type One Protein Phosphatase family known as TOPP8 (AT5G27840), as well as IBR5 (AT2G04550) and an F-box domain-containing protein (At3g17710) (Fig. 4B). Expression of PIRL4 was found to strongly correlate with expression of TMK1 (AT1G66150) (Fig. 4B), while PIRL5 was weakly co-expressed with SK32 (AT4G00720) and CYP721A1 (AT1G75130) (Fig. 4B). PIRL6 was co-expressed with GLOX1 (AT1G67290) and PAB3 (AT1G22760) (Fig. 4B), two genes whose expressions are known to have been restricted to the anther and tapetum [23]. PIRL8 showed co-expression with IMMUNE ASSOCIATED NUCLEOTIDE BINDING 4 (IAN4, At1g33900) and a member of the Alpha Expansion gene family (EXPA17, AT4G01630) (Fig. 4B). The expression of EXPA17 is root-specific and it is vital for lateral root formation [24]. Finally, PIRL9 was found to co-express with UDP-DEPENDENT GLYCOSYL-TRANSFERASE 76B1 (UGT76B1, AT3G11340) and At4g02940 (Fig. 4B). Co-expression data was not available for PIRL2 and PIRL7.

Discussion
LRR proteins are found in many plant species, where they perform important biological functions linked to growth and development. For instance, many proteins from the LRR-receptor-like kinases (RLKs) and related-receptor-like proteins (RLPs) families play crucial roles in plant growth and development and in defense-related pathways [25][26][27]. In A. thaliana, nine PIRLs have been identified, some of which have been previously examined for functional roles in developmental signaling. In the present study, we used promoter:GUS assays and in silico approaches to analyze the amino acid sequences and detailed expression patterns of nine PIRLs found in A. thaliana. Phylogenetic analysis revealed the relationships of the A. thaliana PIRL proteins with other plants PIRLs and resolved all sequences into six groups (Fig. 1A). Each A. thaliana PIRL was found to be close in sequence to one of the PIRLs found in C. rubella (CrPIRL) and others from B. rapa (BrPIRL). A. thaliana PIRL1 and PIRL9 belong to Group V, a group that also contains PIRLs from C. rubella and B. rapa. A. thaliana PIRL4 and PIRL5 belong to a clade that includes PIRL4 and PIRL5 from both C. rubella and B. rapa, designated here as Group I. This group also contained PIRL4 and PIRL5 from O. sativa, Z. mays, H. syriacus, and R. communis. PIRL6, PIRL7, and PIRL8 from A. thaliana, C. rubella and B. rapa all belong to Group III, a group that also includes PIRL1 from O. sativa and Z. mays. PIRL2 and PIRL3 from A. thaliana, C. rubella, and B. rapa belong to the same clade, designated here as Group IV. Interestingly, Groups IV and V did not include PIRLs found in O. sativa and Z. mays. Group VI contains PIRLs only from rice and maize. Focusing on A. thaliana and rice, there is no OsIRL corresponding to PIRL1, PIRL9, PIRL2 and PIRL3, and there is no PIRL corresponding to OsIRL2 and OsIRL3. PIRL1 and PIRL9 have been reported to be essential for pollen development [11]. Although the function of OsIRL2 and OsIRL3 is not yet known, they may be redundantly involved in pollen development like PIRL1 and PIRL9 because they are most similar to PIRL1 and PIRL9 among OsIRLs, and their expression was shown in stamens [9]. OsIRL2 and OsIRL3 may have diverged and acquired additional functions in rice. For example, expression of OsIRL2 and OsIRL3 is also shown in endosperm [9]. Since the osirl3 mutant exhibited a normal phenotype, analysis of osirl2;osirl3 double mutant will be a clue to clarity their function. Analysis of the domain organizations by Uniprot showed that the number of LRRs present in A. thaliana PIRLs is 10 or 11 ( Fig. 1B and Supplementary Table 3). The number of LRRs found in PIRL1, PIRL2, PIRL3, PIRL6, PIRL7, PIRL8, and PIRL9 was one more than reported by Forsthoefel et al. [4]. This difference may be due to differences in the length of the first LRR between the previous prediction and the Uniprot prediction used in this study.
Our expression experiments revealed that PIRL1, PIRL3, and PIRL6 were preferentially expressed in pollen (Fig. 4A). Microarray developmental map, RNA-seq [19,28] and promoter:GUS assays ( Fig. 3 and Supplementary Fig. S1) also indicated strong expression of PIRL1, PIRL3, and PIRL6 in pollen. These results are consistent with reports that have demonstrated that PIRL1 and PIRL9 are vital for the differentiation of microspores into pollen [11]. In addition, recent studies have also shown that pirl2;pirl3 double mutant generated a wide range of abnormal pollen morphologies [12]. Moreover, our promoter:GUS assay of PIRL6 showed expression also in the ovule (Fig. 3F), a finding that is consistent with a previous result that showed that PIRL6 was essential for both male and female gametogenesis processes in A. thaliana [10]. Forsthoefel et al. [10] reported abundant functional transcript in flowers and a small amount of non-functional transcript by alternative splicing in leaf and root. These and our promoter:GUS results indicated transcriptional and post-transcriptional regulation of PIRL6. Next, our co-expression analysis showed that the expression of PIRL1 and PIRL3 are correlated with each other and that PIRL1 expression is also associated with the expression of VQ9 and ZFN2 (Fig. 4B). The essential role of VQ proteins in vegetative plant growth, differentiation, and seed development has been previously described [22]. The relative expression of VQ9 has also been reported in roots by Hu et al. [29], a finding that is consistent with the observed expression of PIRL1 in roots. Taken together, these findings suggest a contribution of PIRL1 to root development in addition to its role in pollen formation. PIRL1 and PIL9 are closely related gene pairs, and pirl1;pirl9 double knockout plants (pirl1 − /− , pirl9 − /− ) were not established due to the developmental defect of pirl1;pirl9 microspore (pirl1 -, pirl9 -) into viable pollen [11]. Analysis of double mutants using weaker alleles may reveal their function in the development of vegetative organs, including roots. Next, we also identified TOPP8 and IBR5 as genes that were co-expressed with PIRL3. TOPP8 is also known as ATUNIS2 (AUN2) and plays a vital role in pollen germination and pollen tube growth [30]. Pollen-and root-specific expression of AUN2 has also been reported by Franck et al. [30], who used a microarray analysis that showed a pattern of expression resembling that of PIRL3, further suggesting their functional similarity. We also found that PIRL6 expression was strongly correlated with the expression of genes coding for glyoxal-oxidase-related protein (GLOX1) and PAB3. GLOX1 was reported to be important for pollen development; Phan et al. [23] found GLOX1 promoter-driven GUS expression within anthers of floral buds as well as strong GUS activity in mature and released pollen grains. Belostotsky [31] also described tapetum and pollen-specific expression of PAB3 after performing promoter:GUS assays in A. thaliana. These results support the possibility that the function of PIRL6 in pollen development is related to GLOX1 and PAB3. Although no expression values were available in the database for PIRL7, our promoter:GUS assay revealed clear expression of PIRL7 both in pollen and in the pollen tube (Fig. 3G), suggesting that PIRL7 plays a role in pollen development. The promoter:GUS assay of PIRL2 showed unambiguous staining of filaments (Fig. 3B). Therefore, the requirement of PIRL2 for pollen development [12] may be exerted in the filament. We obtained root-specific promoter:GUS expression for PIRL4 (Fig. 2D). As shown in Fig. 4, co-expression analysis revealed that PIRL4 was co-expressed with LRR-RLK TMK1 [32], which has been found to be expressed in roots by RT-PCR and promoter:GUS assays [33]. Dai et al. [33] also reported that the root lengths of tmk1; tmk4 double mutants were approximately one-third of that of the wild type. The biological functions of PIRL4 remain unknown; however, co-expression of PIRL4 and TMK1 in roots suggests that PIRL4 may be involved in the growth and development of roots. Furthermore, we also observed promoter:GUS expression of PIRL5, PIRL8, and PIRL9 in the root apical region, as well as in the primary and secondary roots (Fig. 2E, H, and I). According to the heat map and microarray developmental map, PIRL5, PIRL8, and PIRL9 are also known to be expressed in roots (Fig. 4A and Supplementary Fig. S1). In addition, we found that PIRL8 showed co-expression with EXPA17; a previous study by Lee and Kim [24], who used the promoter:GUS of EXPA17 and recorded its expression in lateral root and lateral root primordia. Therefore, they suggested that EXPA17 may be required to promote lateral root formation during auxin response. Together, these findings signify that PIRL8 may have root-specific functions.

Conclusion
In this study, we investigated the detailed expression patterns of nine A. thaliana PIRL genes at different developmental stages. Promoter:GUS assay showed that most PIRLs were expressed in roots and in the vasculature of the primary and lateral roots. In addition, we observed that some PIRLs were strongly expressed in floral organs, suggesting their possible biological function to regulate the growth and development of these organs. Moreover, co-expression network analysis indicated that functionally similar PIRLs are expressed simultaneously. Therefore, these results will be helpful for future tissue-specific expression analyses of other LRR proteins and may provide useful information to improve our understanding of their biological function.