Characterization of the ERP gene family in Arabidopsis thaliana

ABSTRACT

Plant genomes encode numerous proteins with obscure features (POFs) that lack recognized domains or motifs. However, there is little functional information for POFs even in Arabidopsis because biochemical, physiological, and genetic assay are required for the functional annotations of POFs. Here, we identified a small gene family, the endoplasmic reticulum-localized POF (ERP) family, in Arabidopsis. Phylogenetic analysis revealed that the number of ERP family members was conserved in the plant kingdom, suggesting strong selective pressure was imposed on ERP family during plant evolution. No recognizable domains were identified in the predicted ERP proteins, except for the N-terminal signal peptide. ERPs were found to be widely expressed during Arabidopsis development and showed endoplasmic reticulum localization. It was reported that ERP1 is an inositol-1,4,5-trisphosphate 5-phosphatase (5PTase), but ERP1 could not substitute for At5PTase12 in precocious pollen germination, indicating that ERP1 did not have the similar functions as At5PTase12 in inositol 1,4,5-trisphosphate [Ins(1,4,5)P3] metabolism. Further studies are needed to dissect the functions of ERP family proteins in Arabidopsis development.

Introduction

The advances in sequencing technologies have greatly expanded the sequence data. However, the functional annotation of genomes lags behind the rate of genome sequencing. For example, in the Arabidopsis genome, the functions of 30–34% of the predicted proteins are unclear or unknown, these proteins are, therefore, classified as proteins of unknown function (PUFs).¹ The PUFs are divided into two subcategories, PDFs, proteins containing a hidden Markov model (HMM)-recognized domain of unknown function, and POFs, proteins with obscure features that lack any recognized domains or motifs.² POFs have also been found to be more disordered in structure and tend to be shorter and more hydrophilic than PDFs.³ Approximately one-quarter of eukaryotic proteins are estimated to be POFs. Plant genomes contain large numbers of POFs and it has been suggested that they play diverse roles in plant development and stress adaption. For example, overexpression of two Arabidopsis-specific POFs (AT1G21520, AT1G50290) enhances Arabidopsis seedling tolerance to oxidative stress by involving in cellular repair and/or protection against oxidative stress.⁴ Arabidopsis Qua Quine Starch (QQS) is an orphan gene encoding a small POF protein. The QQS protein regulates carbon and nitrogen partitioning as well as susceptibility to pathogens and pests by interacting with Nuclear Factor Y-C4 in Arabidopsis.^5,6 Several cell wall-related studies have revealed that POFs are involved in cell wall formation.² However, there is little functional information on POFs even in Arabidopsis because no homologous genes have been characterized in other species. In this study, we characterized a POF gene family ERP in Arabidopsis. ERP proteins did not contain any recognized domains. We found that ERPs homologs were present in green plants and protozoans. In Arabidopsis, ERPs were widely expressed in various tissues and located in endoplasmic reticulum.

Results

ERP family in plant

In Arabidopsis, three gene loci, At5g54870, At4g27020, and At1g70160, encode three predicted proteins of unknown function that lack any HMM-recognized domains. Only these three homologous genes in Arabidopsis thaliana were identified by BLAST analysis and amino acid sequence alignment (Figure 1(a)). In the present study, we found that these three proteins were all localized to the endoplasmic reticulum (ER); and therefore, we named them the “ER-localized POFs” (ERPs). No recognizable domains were identified in the predicted ERP proteins, except a putative signal peptide sequence localized at their N termini (Figure 1(a)). To investigate whether ERPs were conserved in other organisms, we performed BLAST searches and identified ERP homologs in representative species. Our results revealed that the ERP homologs were present in green plants (Viridiplantae) and protozoans. No ERP homologs were identified from animals (Metazoa) (Figure 1(b), Supplementary Table S2). Phylogenetic analysis revealed that the ERP-like proteins were clustered into two major clades in spermatophytes (Figure 1(b)). ERP1 and ERP2 were located in clade I, and ERP3 in clade II. After analyzing the ERP copy numbers in green plant species, we found that the copy number of the ERP family in each species was conserved. Two to three family members were identified in each species except the green alga Klebsormidium nitens, in which only one ERP gene was identified (Figure 1(b)). Only three members were seen in three polyploid species (Brassica rapa, Glycine max, and Populus trichocarpa) (Figure 1(b)), suggesting that the ERP family was subjected to strong selective pressure during plant evolution.

Figure 1.

Alignments of the amino acid sequence of Arabidopsis ERPs and phylogenetic analysis of ERP and its homologs

(a) Alignments of the amino acid sequence of Arabidopsis ERPs. Black background indicates the same amino acid. Light blue background indicates the similar amino acid. The predicated signal peptides are underlined. (b) Phylogenetic tree of ERPs and its homologs from 36 species.

Expression patterns of ERP family members in Arabidopsis

To investigate the expression patterns of ERPs in Arabidopsis, we performed qRT-PCR analysis with gene-specific primers using the cDNA templates from the different organs and tissues of Arabidopsis, including roots, stems, leaves, inflorescences, flowers, siliques, and seedlings. In these experiments, Tubulin2 was used as an internal control to normalize the RNA contents of different cDNA samples. Our results revealed that ERPs were constitutively expressed in vegetative and reproductive tissues with higher expression in leaves (Figure 2). Furthermore, the detailed expression patterns of ERPs in various tissues were detected by GUS histochemical staining in the transgenic plants expressing proERP1/2/3::GUS construct, in which the GUS gene uidA was driven by the over 2000-bp promoter of ERPs. Strong GUS signals were detected in seedling, the vascular tissue of the root tip, leaf, inflorescence, and the developing embryo (Figure 3). In addition, ERPs showed high expression in the sepal, petal, and pistil in the flower, while their expression were low in mature pollen, ovule, and the developing embryo except for ERP2, which showed higher expression in pollen, ovule, and the developing embryo (Figure 3).

Figure 2.

qRT-PCR analysis of the relative expression of ERPs in different organs of Arabidopsis.

Figure 3.

GUS staining analysis in transgenic Arabidopsis plants expressing proERPs::GUS.

(a1-l1) GUS staining in transgenic plants of proERP1::GUS. (a2-l2) GUS staining in transgenic plants of proERP2::GUS. (a3-l3) GUS staining in transgenic plants of proERP3::GUS. GUS staining are analyzed in different organs, including seedlings (a1, a2 and a3), root tips (b1, b2 and b3), leaves (c1, c2 and c3), inflorescences (d1, d2 and d3), flowers (e1, e2 and e3), sepals (f1, f2 and f3), petals (g1, g2 and g3), pistils (h1, h2 and h3), anthers (i1, i2 and i3), pollen grains (j1, j2 and j3), ovules (k1, k2 and k3), and young embryos (l1, l2 and l3). Bars, 500 µm in a, b, d, e, f, g, h, k, and 50 µm in c, i, j, l.

Subcellular localization of ERPs

Proteins targeted to the ER for subsequent transport through the secretory pathway generally contain signal peptides. The ERP proteins have predicted signal peptides at their N termini, indicating the ERP proteins might be ER-targeted. To test our hypothesis, we made the construct proERP1/2/3::ERP1/2/3-GFP, in which the fusion gene ERP1/2/3-GFP were driven by the ERP1/2/3 themselves promoters (more than 2000-bp). We obtained the proERP1/2/3::ERP1/2/3-GFP ER-rb plants by crossing between the ERP1/2/3-GFP transgenic plants and the transgenic-line ER-rb. ER-rb is a ER-marker in which a synthetic ER retention signal His-Asp-Glu-Leu was added at the C-terminus of mCherry and the signal peptide of AtWAK2, a wall-associated receptor kinase, was fused at the N-terminus of mCherry.⁷ In the ERP1/2/3-GFP ER-rb plants, the mCherry and GFP signals colocalized, indicating the ERPs are ER-localized proteins (Figure 4).

Figure 4.

Colocalization analysis of ERPs and ER in the Arabidopsis meristem zone of root expressing ERPs-GFP and ER-rb. Bars, 10 µm

ERP1 does not have 5PTase activity

At Araport11 (https://www.araport.org), ERP1 and ERP2 all were annotated as inositol-1,4,5-trisphosphate 5-phosphatase (5PTase),⁸ a key enzyme in the phosphatidylinositol (PI) signaling pathway.⁹ Arabidopsis genome encodes 15 5PTases.¹⁰ At5PTase12 is a pollen-expressed 5PTase, which functions in inhibiting precocious pollen germination in anther.⁹ As shown in Figure 3, ERP1 and ERP2 all were expressed in pollen. To test whether ERP1 played similar functions as At5PTase12 in Ins(1,4,5)P₃ metabolism, we performed the complementary analysis by introducing the coding sequence of ERP1 to At5PTase12 mutant (5pt12) under the control of the 2000-bp promoter sequence of At5PTase12, as described.¹¹ A total of 32 transgenic lines (proAt5PTase12::ERP1 5pt12) were obtained, and the line 15, 17, and 22 were chosen for pollen germination analysis in anthers. However, the precocious pollen germination phenotype of 5pt12 could not be rescued by ERP1 (Figure 5). These results indicate that ERP1 is not a functional substitute of At5PTase12 in Ins(1,4,5)P₃ metabolism. In addition, we conducted phylogenetic analysis between ERPs and 5PTases in Arabidopsis, and found that they were far from each other in evolution (Figure S1). Thus, ERPs are not 5PTase.

Figure 5.

ERP1 does not complement the precocious germination of 5pt12 pollen

(a) No pollen germination in the wild-type anther. (b) Precocious pollen germination in the 5pt12 anther. (c-e) Precocious pollen germination in the anthers of the transgenic lines 15 (c), 17 (d) and 22 (e) of proAt5PTase12::ERP1 5pt12. (f) Statistical analysis of pollen germination ratio in the anthers of the wild type, 5pt12, and the transgenic lines 15, 17, and 22 of proAt5PTase12::ERP1 5pt12. Data are presented as means ± SD of three biological replicates. n > 300. n.s., not significant (two-tailed Student t-test). Bars, 50 µm.

Discussion

It is estimated that there are over 4000 entries in the Pfam database describing domains of unknown function.¹² In eukaryotes, approximately 30–34% of the predicted proteins encode PUFs that have been classified into PDFs and POFs depending on the presence or absence of defined motifs or domains.^13,14 Elucidating the functions of PUFs is essential for a complete understanding of the complexities of an organism’s growth and development as well as its interactions with the biotic and abiotic environment. Compared with the PDFs, it is more challenging to characterize the functions of POFs due to the absence of any recognized domains or motifs in their sequences.² In this study, we identified the Arabidopsis ERP family, a typical POF family. ERP family proteins do not contain any defined motifs or domains except an N-terminal signal peptide. About 60% of the POFs identified in the 10 predicted proteomes are species-specific.³ It was found that POFs are predicted to contain higher percentages of disordered structures, supporting a hypothesis that POFs are implicated in species- or phylogenetically specific regulatory and signaling networks.³ In this study, we performed BLAST analysis against many representative genome-sequenced species using the Arabidopsis ERP proteins, and did not find any homologs of ERP in the animal kingdom. Our results revealed that ERPs are specific for plants, protozoans, and bacteria. Studying the functions of POFs is still technically challenging due to their lack of conserved domains, functional annotation can only be obtained by biochemical, physiological, and genetic tests. The Araport website (https://www.araport.org/) annotated ERP1 and ERP2 as inositol-1,4,5-trisphosphate 5-phosphatase despite the fact that neither protein contains the 5PTase conserved domain, the catalytic domain of inositol polyphosphate 5-phosphatases. There are 15 5PTases in Arabidopsis that function in plant growth and development, and phytohormone signaling.¹⁵ 5PTase11 is a cytosol- and plasma membrane-localized protein.¹⁶ In contrast, in this study, GFP-labeled fusion protein analysis revealed that the ERP family is localized in the ER; these results are consistent with the predicted signal peptide sequences at their N termini. Thus, the ERP family might play roles in protein processing and modification in the ER. 5PTase12 is preferentially expressed in leaves and mature pollen grains, and the 5PTase12 knockout mutation results in precocious pollen germination within anthers.¹¹ We found that the precocious germination phenotype of the 5PTase12 mutation could not be recovered by ERP1. This suggests that ERP1 is not functionally homologous to 5PTase12 in Arabidopsis, and the biochemical functions of ERPs need to be investigated.

Materials and methods

Plant materials and growth conditions

The Arabidopsis Columbia-0 (Col-0) ecotype was used as the wild type. All transgenic plants were created using Col-0. The 5pt12 mutant was obtained from the Arabidopsis Biological Resources Center (ABRC). Seeds were first disinfected with 70% ethanol for 5 minutes, then incubated in 2.5–2.6% sodium hypochlorite solution containing 0.5% Tween 20 for 10 minutes, and washed eight times in sterile distilled water. Seeds were germinated on 1/2 Murashige and Skoog (MS) plates containing 1% agar. After 10 days, the plants were transplanted to soil and grown at 22°C with a 16-h light and 8-h dark photoperiod.

Aniline blue staining of pollen

Dehiscent anthers from wild-type, 5pt12, and 5pt12 plants with the complementary expression of ERP1 were fixed in 1:3 acetic acid: ethanol solution for 2 h, cleared three times with distilled water, softened overnight in 8 M NaOH solution at room temperature, and then washed three times with distilled water. The anthers were then stained with 0.1% aniline blue solution (0.1% aniline blue in 0.1 M K₂HPO₄-KOH buffer, pH 11) for at least 2 h in the dark. Stained anthers were observed in ultraviolet light with an Olympus BX51 fluorescence microscope (Olympus, Tokyo, Japan).

RNA extraction and quantitative real-time PCR (qRT-PCR)

For analyzing the expression patterns of ERP1, ERP2, and ERP3, RNA was extracted from different tissues of Col-0 using the Trizol reagent (CWBIO, Beijing, China, CW0581M). These tissues included roots, stems, leaves, inflorescences, flowers, siliques, and 7-day-old seedlings. Two micrograms of total RNA were used for reverse-transcription to synthesize cDNA using the RevertAid first-strand cDNA synthesis kit (TIANGEN, Beijing, China, KR116). qRT-PCR was performed in optical 96-well plates using a Chromo4 real-time PCR system (Roche, Basel, Switzerland). The housekeeping gene Tubulin2 was used as an internal control to normalize the transcript levels in each cDNA sample. All qRT-PCR experiments contained three biological replicates and three technical replicates. The primers used are listed in Supplementary Table S1.

GUS histochemical staining

GUS (β-glucuronidase) histochemical staining assays were performed as previously described.¹⁷ Harvested fresh tissues from the transgenic lines expressing proERP1::GUS, proERP2::GUS, and proERP3::GUS were fixed in 90% acetone on ice for 10 min and were then transferred to GUS staining solution [50 mM Na₂HPO₄ (pH 7.2), 0.5 mM K₃ Fe(CN)₆, 0.5 mM K₄Fe(CN)₆, 0.1% Triton X-100, and 2 mM X-Gluc (Sigma-Aldrich, 70036-M MSDS)], vacuum infiltrated for 10 minutes, and incubated overnight at 37°C. Finally, staining was stopped by replacing the staining solution with 70% ethanol solution. Stained specimens were photographed using an Olympus BX-51 microscope equipped with an Olympus DP71 digital camera or an Olympus SZX-16 stereomicroscope equipped with an Olympus DP72 digital camera.

Phylogenetic analysis

Full-length orthologs of ERP1, ERP2, and ERP3 from different organisms were retrieved from the NCBI database (https://www.ncbi.nlm.nih.gov//Protein) as described by Sun et al.¹⁸ Sequences were aligned by using ClustalX2.1 with default parameters. The phylogenetic tree was constructed by the maximum likelihood algorithm using MEGA 6.0 with a bootstrap of 1000 replicates.

Plasmid construction and plant transformation

For the complementation construct of the 5pt12 mutant, including the 2074-bp promoter of 5PTase12, the coding region and the 3ʹ untranslated region (UTR) of ERP1, was amplified by PCR from the Arabidopsis genome and cloned into pROKII to generate the pro5PTase12::ERP1 expression vector. This vector was transformed into 5pt12 using the Agrobacterium-mediated floral dip method. The primers used are listed in Supplementary Table S1.

The proERP1::GUS reporter construct was created as follows: the 2123-bp promoter of ERP1 was amplified by PCR from Arabidopsis genomic DNA using the primers ERP1-GUS-F and ERP1-GUS-R. The PCR product was cloned into pENTR/D-TOPO vector and subsequently recombined into pMDC163 vector to generate the proERP1::GUS expression vector. This vector was transformed into Col-0 using the Agrobacterium-mediated floral dip method. The construction method for the proERP2::GUS expression vector and the proERP3::GUS expression vector was the same as the proERP1::GUS expression vector. These three vectors were transformed into Col-0 using the Agrobacterium-mediated floral dip method, respectively. The primers used are listed in Supplementary Table S1.

For subcellular localization, the ERP1 genomic sequence containing a 2123-bp promoter and 2453-bp coding regions without the stop codon was PCR-amplified from Arabidopsis genomic DNA with the primers of ERP1-GFP-F and ERP1-GFP-R. The PCR product was recombined into pROKII-GFP to generate the proERP1::ERP1-GFP expression vector. The ERP2 genomic sequence containing a 2171-bp promoter and 2775-bp coding region without the stop codon was PCR-amplified. The resulting PCR product was cloned into pENTR/D-TOPO vector and subsequently recombined into pMDC107 vector to generate the proERP2::ERP2-GFP expression vector. The ERP3 genomic sequence containing a 2202-bp promoter and 2184-bp coding region without the stop codon was PCR-amplified. The PCR product was recombined into pROKII-GFP to generate the proERP3::ERP3-GFP expression vector. The above vectors were then transformed into Agrobacterium GV3101 for transformation into Col-0. The primers used are listed in Supplementary Table S1.

Confocal microscopy

The proERP1::ERP1-GFP, proERP2::ERP2-GFP, and proERP3::ERP3-GFP transgenic lines were crossed with ER-rb plants, respectively. T2 seedlings were grown vertically on 1/2 Murashige and Skoog (MS) plates containing 1% agar without selection. Roots of 4–5 d-old seedlings were observed, and fluorescent images were collected on a Zeiss confocal laser scanning microscope (Zeiss, Jena, Germany) with a 40× oil objective under 488 nm and 561 nm laser line.

Funding Statement

This study has been supported by National Natural Science Foundation of China (31770349).

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website.