Identification and characterization of the first pectin methylesterase gene discovered in the root lesion nematode Pratylenchus penetrans

Similar to other plant-parasitic nematodes, root lesion nematodes possess an array of enzymes that are involved in the degradation of the plant cell wall. Here we report the identification of a gene encoding a cell wall-degrading enzyme, pectin methylesterase PME (EC 3.1.1.11), in the root lesion nematode Pratylenchus penetrans. Both genomic and coding sequences of the gene were cloned for this species, that included the presence of four introns which eliminated a possible contamination from bacteria. Expression of the Pp-pme gene was localized in the esophageal glands of P. penetrans as determined by in situ hybridization. Temporal expression of Pp-pme in planta was validated at early time points of infection. The possible function and activity of the gene were assessed by transient expression of Pp-pme in plants of Nicotiana benthamiana plants via a Potato virus X-based vector. To our knowledge, this is the first report on identification and characterization of a PME gene within the phylum Nematoda.


Introduction
The plant cell wall plays an important role in various fundamental physiological processes of plant growth and development, such as maintaining the integrity of cellular content, morphogenesis, and cell signaling. In addition, the cell wall is the primary interface for most plantpathogen interactions, since it is the first physical barrier against invasion and infection [1,2]. The primary cell wall has an intricate structure composed of a complex network of cellulose microfibrils interconnected within a matrix of polysaccharides, including pectins, hemicelluloses, and glycoproteins [3,4]. Pectin, a highly abundant polysaccharide, is an important component of both primary and secondary cell walls, forming the main component of the middle lamella [5,6]. Modification of the pectin network is tightly regulated by the action of a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 mechanically disrupts the cell wall and through which CWDEs are secreted to facilitate penetration and migration of the nematode through host roots.
Pratylenchus penetrans (Cobb, 1917) Filipjev and Schuurmans Stekhoven, 1941 is regarded as one of the most destructive species of this genus because of its ability to parasitize a wide range of economically important host plants (e.g. alfalfa, corn, and potato), and its broad geographic distribution [29]. The core set of genes encoding CWDEs identified for P. penetrans so far comprises β-1,4-endoglucanases (GH5), pectate lyases (PL3), arabinogalactan endo-1,4-βgalactosidases (GH53), xylanases (GH30), and expansin-like genes [30]. As in other PPNs, the majority of these CWDEs are localized in the esophageal gland cells of P. penetrans, and are actively produced during the early time points of plant infection [31]. Interestingly, a refined data mining of the P. penetrans transcriptome [30] resulted in the identification of a transcript encoding a PME gene, here after named as Pp-pme. Here we report the identification and molecular characterization of this Pp-pme gene, which is, to the best of our knowledge, the first report of a PME encoding gene within the phylum Nematoda.

Nematode collection and extraction
Two different isolates of P. penetrans were used in this study: 1) NL 10p RH collected in Beltsville, MD, USA, and 2) A44L4 from potato (Solanum tuberosum L.) fields in Coimbra, Portugal. Both isolates were maintained and multiplied in vitro in roots of corn (Zea mays L. cv. 'Iochief') growing in Murashige and Skoog basal medium containing vitamins (MS) (Sigma-Aldrich, MO, USA), 3% (w/v) of sucrose (pH 5.8) and solidified with 1.5% (w/v) agar (Sigma-Aldrich, MO, USA). Nematodes were re-cultured every two months into new roots of corn and maintained in the dark at 28˚C. Nematodes were extracted from infected corn roots as described in Vieira et al. [30].

Isolation of Pp-pme genomic DNA and cDNA sequences
During our previous analyses of the transcriptome assembly generated for P. penetrans [30], a partial transcript of 415 bp showing similarity to a bacterial pectin methylesterase gene (WP_090106049.1) was identified. In order to obtain the corresponding full-length coding sequence, BLAST searches were initially performed against the skimming genome assembly of P. penetrans [32]. Based on the assembly of the retrieved contigs, primers were designed to flank the in silico predicted fragment (S1 Table), as well as the putative full-length proteinencoding Pp-pme sequence, including partial sequences of both 5´and 3´regions (S1 Table). DNA extraction was performed from mixed stages (eggs, juveniles, females and males) of P. penetrans using the PureLink Genomic DNA Mini kit (Invitrogen, Carlsbad, CA, USA), following the manufacturer's instructions. Total RNA was extracted from mixed life stages of P. penetrans using the RNeasy Plant Mini Kit (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. RNA was then treated with RNase-Free DNase (Qiagen, Valencia, CA, USA) before reverse transcription. The quantity and quality of the extracted RNA was assessed with a ND-1000 Nanodrop spectrophotometer (NanoDrop). The first strand cDNA was synthesized using the iScript first-strand synthesis kit (Bio-Rad, Hercules, CA, USA) following the manufacturer's instructions. The putative Pp-pme genomic coding region of the isolate NL 10p RH was obtained by PCR amplification using 5 μl of the nematode DNA extract, 1x PCR buffer, 1 U Platinum Taq DNA polymerase (Invitrogen) and 0.2 μM of each primer in a 50 μl solution. A PTC-200 Peltier thermocycler was used for amplification using the following PCR conditions: one denaturation step at 94˚C for 2 min, followed by 30 cycles at 94˚C for 30 secs, 55˚C for 30 secs, 72˚C for 3 min, and a final extension step of 10 min at 72˚C. For amplification of the cDNA sequences of NL 10p RH and A44L4 isolates, 1 μl of cDNA was used which derived from each of the corresponding mixed-stage libraries, followed by similar PCR conditions described above, with an adjusted extended time (i.e. 1 minute per kb was used). The corresponding genomic and cDNA amplicons of P. penetrans were separated on a 1% agarose gel and stained with SYBRSAFE, and the corresponding bands were recovered from the gel using the MinElute Gel Extraction kit (Qiagen, Valencia, CA, USA). Both amplicons were then ligated to the TOPO TA Cloning kit (Invitrogen, Carlsbad, CA, USA) using the protocol provided by the manufacturer, and transformed into E. coli TOP10 competent cells (Invitrogen, Carlsbad, CA, USA). PCR colony positive clones were grown in 3 mL of LB overnight at 37˚C followed by plasmid DNA extraction (QIAprep Spin Miniprep kit, Qiagen, Valencia, CA, USA). DNA and cDNA clones were verified by automated sequencing using the universal M13 primers by Macrogen USA (MD, USA). Unless otherwise stated, all the following analyses were conducted with P. penetrans isolate NL 10p RH.

Sequence analysis
To identify introns in the genomic sequence, both cDNA and genomic sequences were aligned using MUSCLE [33]. Gene schematics were generated with the Exon-Intron Graphic maker available at WormWeb.org. The nucleotide and translated amino acid sequences were analyzed for similarity to other genes and proteins using BLAST analyses against the NCBI non-redundant nucleotide and protein databases (http://www.ncbi.nlm.nih.gov/), the nematode.net (http://www.nematode.net/), and NEMBASE4 database (http://www.nematodes.org/ nembase4/). In addition, protein sequence analyses were conducted using the following programs: SignalP 4.0 for prediction of protein signal peptide [34]; ProteParam for determination of the protein molecular mass and theoretical isoelectric point [35]; and CLC Main Workbench v. 8.0 software for protein secondary structure predictions. InterPro scan was performed using the public available software package (https://www.ebi.ac.uk/interpro/sequencesearch/iprscan). The overall GC content (%) and codon usage (GC1, GC2 and GC3 corresponding to the 1 st , 2 nd and 3 rd position in the codon triplet, respectively) were calculated using the on-line EMBOSS cusp package (http://www.bioinformatics.nl/cgi-bin/emboss/cusp). Modelling of the tertiary structure of the P. penetrans PME and the corresponding PME TOP hits of Bacteria/ Fungi/ Plants/Insects were performed using Phyre2 in the normal mode analysis [36] (http://www.sbg. bio.ic.ac.uk/phyre2). Structural classification of the translated protein followed the convention of SCOPe (Structural Classification of Proteins-extended, http://scop.berkeley.edu/) [37].
BLASTp search was done against the non-redundant (nr) protein database at NCBI, and sequence selection was based on the TOP hits of Bacteria/Archaea/Fungi/Plants/Insects, excluding sequences of the same species with 100% identity. The percentage of protein identity (identical residues in alignment positions) between the P. penetrans PME and those of the other organisms was determined by pairwise comparison. Multiple sequence alignment of PME was conducted with MAFFT algorithm [38] with default settings, trimmed with TRIMAL [39] and manually curated and edited for the selection of protein domain. The phylogenetic relationships were estimated using maximum likelihood (ML) analysis. The best model for protein evolution, determined with AMINOSAN [40] was the "Whelan and Goldman" (WAG) model with discrete gamma distribution (number of gamma categories = 8). The robustness of ML analysis was inferred using 1,000 bootstrap replicates.

Pp-pme expression at different nematode developmental stages
RNA of different nematode developmental stages [eggs, juveniles (J2 to J4), males, and females] was obtained from 150 nematodes and then isolated using the same kit and conditions mentioned above. Expression of the Pp-pme at different developmental stages of P. penetrans was analyzed by semi-quantitative RT-PCR using a pair of primers that amplify a Pp-pme fragment of 221 bp (S1 Table). The P. penetrans 18S rDNA gene was used as the reference gene (S1 Table). PCR products were separated on a 1% agarose gel and stained with SYBR Safe.

In situ hybridization of Pp-pme transcripts
To assess the localization of Pp-pme transcripts, whole mount in situ hybridization was performed using all stages of P. penetrans NL 10p RH following the protocol of de Boer et al. [41]. For localization of the Pp-pme transcripts, the same primers described above for RT-PCR analyses were used, while a fragment of 241 bp of the Pp-eng-1 gene was used as positive control in this experiment (S1 Table). The resulting PCR products were used as a template for generation of sense and antisense DIG-labeled Pp-pme probes using a DIG-nucleotide labeling kit (Roche, Indianapolis, IN, USA). Hybridized probes within the nematode tissues were detected using anti-DIG antibody conjugated to alkaline phosphatase and its substrate. Nematode sections were then observed using a Nikon Eclipse 5i light microscope.

Pp-pme transcript expression in host roots after nematode infection
To validate the expression of Pp-pme during its interaction with plants, in vitro assays were conducted using roots of alfalfa, corn, potato, and soybean and challenged with P. penetrans as described by Vieira et al. [42]. Total RNA was then extracted from a pool of infected roots at 1, 3, 7 or 11 days after nematode infection (DAI) as mentioned previously. Semi-quantitative RT-PCR analyses were performed as described above. Reference genes included an actin gene for corn and potato, the Ubiquitin-3 (Ubi3) for soybean, and the NP_001237047 gene for alfalfa (S1 Table).

Expression of Pp-pme in planta
For expression of the Pp-pme in planta, the full-length coding sequence was amplified from a nematode cDNA library (S1 Table). RT-PCR products were cloned into the pCR TOPO II vector (ThermoFisher Scientific, MA, USA), digested with EcoRV restriction enzyme (restriction site was incorporated into both PCR primers), gel-purified and sub-cloned into the EcoRV-linearized PVX-based vector pP2C2S (obtained from D. Baulcombe, Sainsbury Laboratories, Norwich, England) [43]. The integrity of all clones was verified by automated sequencing. pP2C2S plasmids were linearized with SpeI, and capped transcripts were generated from cDNA clones using Ambion's T7 mMessage Machine kit (ThermoFisher Scientific, MA, USA). The transcripts were mechanically inoculated onto fully expanded leaves of three N. benthamiana plants. Transcripts were also produced from pP2C2S plasmids without inserts ("empty" PVX vector) and inoculated onto three plants to serve as controls representing a wild-type PVX infection (PVX-WT). Three more plants were buffer-inoculated to serve as negative ("healthy") controls. Inoculated plants were monitored daily for symptoms. The inoculation experiments that included all sets of plants of each variant was repeated three times. At 14 days after inoculation leaves and roots were photographed and collected, snap-frozen in liquid nitrogen and stored at -80˚C until RNA extraction. cDNA was synthesized using the same protocol mentioned above. Semi-quantitative RT-PCR analyses were performed for transcript detection of Pp-pme or PVX wild-type fragments from infected N. benthamiana leaves and roots (S1 Table).

Protein identification
Symptomatic leaves from two individual plants infected with PVX-Pp-pme recombinant virus vector and two plants infected with empty vector (PVX-WT) were sent to Bioproximity LLC (Virginia, USA) for protein extraction and identification services to reveal the top most abundant protein groups expressed in the infected plants. All procedures were performed according to the company protocols and protein assays were carried out on a Thermo Q-Exactive HF-X Orbitrap mass spectrometer (https://www.bioproximity.com/protein-identification). The obtained Ultra Performance Liquid Chromatography-Tandem Mass Spectrometer (UPLC-MS/MS) peptide datasets (mzML format for proteomics mass spectrometric data) were exported as Mascot generic format (.mgf) file using PRIDE Inspector software [44]. Peptide-Shaker version 1.16.36 [45] was used for protein identification and validation. Protein identification was performed against a concatenated target/decoy version of N. tabacum cv. TN90 proteome (ID UP000084051) [46] retrieved from the UniProtKB (Universal Protein Resource Knowledgebase-http://www.uniprot.org/), with 73,605 proteins complemented with Pp-PME protein sequence (3 Jan, 2019). Decoy sequences were generated by reversing the target sequences in SearchGUI 3.3.11 [47]. Identifications settings were: Trypsin with a maximum of 1 missed cleavages; 10.0 ppm for precursor m/z tolerance (MS1) and 0.5 Da for fragment m/z tolerance MS2; fixed modifications: carbamidomethylation of C (+57.021464Da), and variable modifications: deamination of N (+0.984016 Da), deamination of Q (+0.984016 Da), oxidation of M (+15.994915 Da), pyrolidone from E (-18.010565 Da) and pyrolidone from Q (-17.026549 Da). Peptides and protein inferences were identified from the spectrum of identification results. Protein Inference was considered with 100% confidence when identification was based on 2 or more validated peptides. Peptide Spectrum Matches (PSMs), peptides and proteins were validated at 1.0% False Discovery Rate (FDR) estimated using the decoy hit distribution. UPLC-MS/MS data with respective protein inference were deposited at the proteomics data repository PRIDE Archive (http://www.ebi.ac.uk/pride/archive/) with the dataset identifier PXD012419.

Pp-pme genomic and cDNA coding sequences
Data mining of the transcriptome assembly of P. penetrans revealed a transcript of 415 bp with high similarity to a pectin methylesterase gene from a bacterium Chitinophaga sp. CF118 (WP_090106049.1, BLAST e-value of 1,31e-111). To validate whether this transcript originated from a prokaryote or an eukaryote, BLAST search analyses were initially performed using the skimming genome assembly of P. penetrans [32], which confirmed the presence of a PME-coding gene in P. penetrans. The cloned genomic DNA amplicon revealed a transcribed genomic sequence of 2,501 bp, while the cloned cDNA amplicon contained a putative open reading frame of 987 bp for the isolate NL 10p RH (Fig 1A). The exon/intron boundaries of the genomic sequence were determined by aligning both genomic and cDNA sequences (Fig 1B and S1  Fig). The genomic sequence revealed the presence of four introns (368, 377, 493 and 282 bp long), all following the canonical GT/AG splicing site (Fig 1B). The coding sequence has an overall GC content of 56.36%, with a GC1 of 52.25%, GC2 of 45.95%, and GC3 of 70.87% (S2 Table). Using the same set of primers, the full-length cDNA sequence of a P. penetrans A44L4 isolate collected in Portugal was cloned, thus confirming the presence of a PME gene in a distant geographic isolate.

Characterization and sequence analysis of the predicted Pp-PME protein
In silico translation of the Pp-pme full-length cDNA revealed a protein sequence of 328 amino acids with a predicted molecular weight of 35.815 kDA and a pI of 5.87. The protein was predicted to have an N-terminal signal peptide of 16 amino acids as determined by SignalP v. 4.0, with a cleavage site located between positions 16 and 17 (VRG-QQ), and no predicted transmembrane domain. Pfam domain search and InterPro scan confirmed this protein as a member of the CE8 family (pfam01095; InterPro IPR000070). PMEs can also be distinguished on the basis of specific signature patterns in their amino acid sequences [48]. Based on the multiple sequence alignment of Pp-PME with PMEs of other organisms, the following five conserved sequence segments typical of PMEs were identified: 64_GxYxE, 146_QAVAL, 168_QDTLY, 195_DxIFG and 251_LGRPR (Fig 2).
A three-dimensional model of Pp-PME of P. penetrans (Fig 3) was predicted using the PME of Erwinia chrysanthemi as template (NCBI TaxId: 556), which was the top hit species obtained by a BLASTp similarity search against the Protein Data Bank. There was 41% identity and 293 residues were covered with 100% confidence. This structure displayed a single-stranded, righthanded beta-helix fold (SCOPe: 51125) composed predominantly of beta-strands (class All beta proteins, SCOP 48124). One alpha-helix was predicted at the N-terminal end of the PME (14 amino acids) and another alpha-helix (5 amino acids) was predicted and at the C-terminal end. The latter was followed by a beta-strand (5 amino acids) and a short alpha-helix (4 amino acids). For the other organisms, the model of each corresponding PME was predicted with 100% confidence (Fig 3).

Phylogenetic analysis of Pp-PME
Although PMEs occur naturally in both prokaryotes and eukaryotes, prior to this work there was no record of a PME within the phylum Nematoda. BLAST searches were performed using the Pp-PME protein sequence as a query against different public databases. The top 15 BLAST hits, consisting of PMEs of bacteria origin only, are presented in Table 1. The amino acid sequence of Pp-PME was found to be most closely related to PMEs of different bacteria of the phylum Bacteriodetes (E-values ranging from 1e-111 to 1e-105), showing 56.86% identity with Chitinophaga sp. CF118, 53.65% with Mucilaginibacter sp. PPCGB 2223, and 53.54% with Flexibacter sp. ATCC35208. When BLASTp searches were performed against other organisms for which PMEs have been previously identified (i.e. archaea, fungi, insects, and plants) (S3 Table), the highest homology of the Pp-PME was found with the PME of the soil arthropod Folsomia candida (Arthropoda: Collembola: Isotomidae) (51.58% protein identity; Evalue = 7e-96), and with the PME of Haloterrigena salina (Halobacteriaceae: Euryarchaeota: Archaea) (46.18% protein identity; E-value = 4e-83). BLAST results against insects and plants showed a lower identity (up to 43.17% and 40.62%, respectively) against the nematode Pp-PME (S3 Table). Despite exhaustive BLAST searches against the large number of sequences and genomes available for other RLNs or PPNs, no homologues were found in any available nematode sequence datasets. The analysis of codon usage of Pp-PME and the corresponding top hit of each group of organisms (S4 Table) indicates that Pp-PME exhibits codon usage that is more similar to other eukaryotic species, than to prokaryotes.
The predicted pectinesterase domain sequence (pfam01095, position 23-307 aa, Evalue = 1.8e-68) of Pp-PME was then aligned with the corresponding domain sequences of the top hit PME proteins of 53 eukaryotic species, including those of fungi, insects, and plants, as well as 29 prokaryotic species distributed among bacteria and Archaea (S3 Table). In order to infer the phylogenetic relationship of the Pp-PME and other organisms, ML analyses were performed (Fig 4). These analyses revealed a clear separation of the PMEs of plant origin with the remaining PMEs (i.e. archaea, bacteria, fungi, insects, and P. penetrans), and highly supported by our bootstrap results (>98%). The PMEs of the remaining taxa could be separated into three main phylogenetic clades, which is also highly supported by the bootstrap analyses. Interestingly, the PMEs of the insects of the family Curculionidae (Coleoptera) clustered together The five conserved sequence segments typical of PME proteins are numbered according to their positions in the corresponding Pp-PME predicted protein (64_GxYxE, 146_QAVAL, 168_QDTL, 195_DxIFG, and 251_LGRPW). Conserved residues among species are indicated by dark blue shading and dots, whereas similar residues are represented in light blue using a threshold for shading of 50% similarity. The accession numbers corresponding to each species are presented in S3 Table. https://doi.org/10.1371/journal.pone.0212540.g002 with bacterial PMEs of the Proteobacteria phylum, while the PMEs of the whitefly B. tabaci (family Aleyrodidae, Hemiptera) grouped with a monophyletic clade formed by fungi only. As denoted by the protein alignment described above, Pp-PME clustered with bacterial sequences of the phylum Bacteriodetes (80% bootstrap value), including the soil collembolan F. candida (family Isotomidae, Collembola), although with a lower bootstrap value. All remaining PME sequences of bacteria (Phylum Firmicutes) and Archaea formed well-separated sub-clusters.

Expression pattern profile of Pp-pme at different stages of nematode development
The expression pattern of Pp-pme was determined by semi-quantitative RT-PCR for different nematode developmental stages [eggs, juveniles (J2-J4), adult males and females] (Fig 5A). Pppme expression was only detected for nematode motile stages, whereas no expression was detected from the eggs. Amplification of 18S rDNA gene fragment was used as positive control in this experiment. Identification of the gene-specific transcripts within the nematode tissues can provide insights into putative functions of the Pp-pme. Therefore, we performed whole mount in situ hybridization for detection of transcripts within the nematode tissues. As determined by the antisense probe, the Pp-pme transcripts accumulated within the esophageal glands of the nematode (Fig 5B), while no signal was detected using the sense probe of the Pp-pme as control (Fig 5C). As a positive control, in situ hybridization was performed with the antisense probe designed for a CWDE (Pp-eng-1) that specifically localizes within the esophageal glands ( Fig  5D). The localization of the Pp-pme within the glands and the presence of a N-terminal signal peptide of the deduced Pp-PME protein suggest its secretion by the nematode into the host.

Pp-pme transcript expression in host roots after nematode infection
The expression of the Pp-pme was analyzed during interaction with different host plants (e.g. roots of alfalfa, corn, potato and soybean) by semi-quantitative RT-PCR at early time points after nematode infection (Fig 6). In each respective nematode-plant interaction the expression of the Pp-pme could be detected, showing that this gene is actively transcribed during the nematode's interaction with the host plants.

Transient expression of Pp-pme in Nicotiana benthamiana
Plant-parasitic nematodes are able to secrete effector proteins directly into the apoplasm or cytoplasm of the plant cell. CWDEs are often secreted by PPNs to the apoplasm during invasion of the plant roots [49]. These CWDEs are thought to be involved in cell wall disassembly thus facilitating the penetration and migration process of the nematode into the roots. This is indicative that the Pp-PME could be also secreted into the apoplast by the nematode and be involved in the invasion of the roots. To direct the nematode PME into the secretory pathway of the host cell, and to assess putative Pp-pme functional role, the full-length coding sequence of the Pp-pme gene, including the signal peptide, was transiently expressed in N. benthamiana using a PVX-based vector. Phenotypic differences between symptoms generated by the recombinant virus containing Pp-pme and by the empty PVX vector are shown in Fig 7. Systemic expression of the Pp-pme gene via the PVX vector in N. benthamiana resulted in more pronounced chlorosis, vein clearing and yellowing of the leaves starting 10-14 days after inoculation (Fig 7B). In some cases, lesion-like symptoms were observed on the leaves of PVX-Pp-pme-infected plants (Fig 7C). Plants expressing empty PVX transcripts displayed characteristic mosaic-like symptoms (Fig 7A). In addition, all plants inoculated with PVX-Pp-pme exhibited distinct lesion-like symptoms in different areas of the stem, branches (Fig 7E and  7F) and main root (Fig 7H), which were absent in plants infected with empty PVX vector ( Fig  7D and 7G). Semi-quantitative RT-PCR conducted confirmed expression of the Pp-pme gene in plants inoculated with PVX-Pp-pme (Fig 7J), while a fragment of PVX plasmid confirmed the expression of PVX transcripts in inoculated plants (Fig 7I).
To detect the possible accumulation of the Pp-PME in the leaves of N. benthamiana plants, UPLC-MS/MS analyses were performed with proteins extracted from the leaves of PVX-Pppme and PVX-WT plants. The peptides identified by mass spectrometry were matched to a set of 73,605 proteins sequences of N. tabacum, and complemented with the predicted Pp-PME protein sequence (S6A and S6B Table). In both PVX-Pp-pme plants, seven to nine peptides were identified that were 100% identical to Pp-PME (Table 2 and S3 Fig). No Pp-PME-related peptides were retrieved from two PVX-WT plants (S6C and S6D Table).

Discussion
In this work, we report the identification, characterization, and phylogenetic status of a putative pectin methylesterase gene PME that encodes a cell wall-degrading enzyme in the RLN, P. penetrans. Several previous studies have shown that PPNs possess a set of CWDEs to degrade the plant cell wall and thus facilitate their invasion and migration through the plant host tissues. The diversity and number of CWDEs endogenous to each PPN do not only reflect the complex nature of the plant cell wall and its structural components, but also indicate an elaborate parasitic strategy and adaptation of these obligatory plant-pathogens [50]. Although Although bacteria (e.g. Wolbachia) have been reported from within different parts of the P. penetrans body [51], the features of the Pp-pme gene exclude the possibility of a prokaryotic contamination. The gene structure of this new nematode CWDE encompasses typical eukaryotic features with four characteristic spliceosomal introns (GC/AG splice sites). The presence of a putative signal peptide in the corresponding Pp-PME sequence, coupled with its transcript expression in the esophageal glands of the nematode, highlights its potential importance in the parasitic process of P. penetrans. This is in agreement with our previous observations of other CWDEs found in this particular species [31], which were localized in the esophageal glands of the nematodes and potentially secreted into the host tissues.
Despite all PMEs being classified under the same family, their mode of action is likely different, depending on the pH and ionic environment, substrate specificity, and origin [17]. In plants, PMEs are encoded by a large family of genes, which emphasizes their functional diversity within the plant tissues [9]. In bacterial and fungal pathogens, PMEs play a critical role in their virulence [52], since their secretion has been related to enzymatic degradation of the pectin polysaccharides of the plant cell walls. P. penetrans has a wide range of host plants and is capable to parasitize on both mono-and dicotyledonous plants. The components of the cell wall in monocots and dicots vary extensively, thus requiring some plasticity in the substrate specificity of the enzymes secreted by the nematode. Our results indicate that Pp-pme is expressed during the early infection time points in different host plants. This is in line with the expression of other CWDEs identified in P. penetrans, such as pectin lyases and different glycoside hydrolase (GH) gene families [31]. Degradation of the plant cell wall pectin network requires the synergistic action of different pectolytic enzymes, such as polygalacturonases (GH28), PLs, and PMEs. PME catalyze the de-methylesterification of pectin, a major component of plant cell wall, where their activity is finely tuned through endogenous inhibitors. Plant-derived PMEs can contribute toward immunity against pathogens [53], while pathogenic PME may function to promote disease [18,19]. The activity of PMEs can improve cell wall accessibility for other CWDEs and consequently accelerate cell wall degradation [9,10,11]. The correctness of the discovered Pp-pme genetic sequence was confirmed by its transient expression in planta, where it translated into a PME-like protein. Therefore, it is plausible that the synchronized expression of the nematode PME with other CWDEs during infection could induce changes in the properties of the plant cell wall, contributing to penetration and migratory activity of the nematode. Speculatively, the symptoms observed on PVX-Pp-pme infected plants could be potentially related to the activation of the plant defense reactions against this effector-like protein. Although, further experiments are needed to confirm this. To date, PMEs are restricted to plants, bacteria, and fungi and are exclusively found in a small number of insect species belonging to the family Curculionidae [23,26,54], the whitefly and the springtail [24,25]. The predicted pectinesterase amino acid domain of the Pp-PME showed significant similarity to different bacterial PMEs. Phylogenetic analyses performed in this work, indicates that this gene has most likely been acquired by horizontal gene transfer (HGT) into P. penetrans. Horizontal gene transfer events of other CWDEs from ancestral microbial donors into sedentary and migratory PPNs of the Tylenchida have been extensively reported [50,55]. One common characteristic of these CWDE genes, which was also observed for the genomic Pp-pme sequence, is the presence of long introns [56,57], in contrast to the average intron size reported for the available nematode genomes [58]. The functions of the transferred gene products are often linked to particular steps of nematode-plant interaction, particularly to the parasitism of PPNs. As mentioned above, Wolbachia endosymbionts have been found in different tissues of P. penetrans [51], which could be seen as a potential route for HGT of the transmissible genomic fragments to the nematode. However, the symbiotic bacteria described so far from nematodes, including the most recently characterized Wolbachia strain (wPpe) isolated from P. penetrans [46], are not known to encode any of the CWDEs that have been potentially acquired by HGT [50].
The PME sequences of the bacterial origin used in this study are highly heterogeneous and were placed by the phylogenetic analysis into different branches of the phylogenetic tree correlating with different niches and taxa. While PME sequences of flying insects of the family Curculionidae strongly correlate with Proteobacteria, the whitefly B. tabaci clustered together with PME sequences of the fungal origin. The diversity of the genes encoding PME in the different groups of insects, support the idea that these PMEs have potentially been acquired multiple times independently by their receiver taxa [23]. In many cases, genes encoding CWDEs in PPNs seemed to be most closely related to sequences found in bacteria that inhabit the soil [55]. The predominant putative ancestral donors found so far are related to bacteria of the Phyla Proteobacteria, Actinobacteria, Firmicutes, and Bacteriodetes [55,59]. In line with these findings, P. penetrans PME is most similar to PMEs of bacterial sequences of the phyla Bacteriodetes (Chitinophagaceae, Chitinophaga), followed by the PME sequences of the soil collembolan F. candida (Collembola: Isotomidae). Species belonging to the Bacteriodetes are specialized in the degradation of complex organic matter [60,61], particularly the family Chitinophagaceae that showed a high ability for degradation of plant-derived carbohydrates (e.g. cellulose and chitin) [62,63]. Remarkably, this group of bacteria has been recently reported to be a part of the rhizosphere microbiome associated with the presence of P. penetrans [63] and also found in the microbiome of F. candida [64,65]. Similarly, the genome of F. candida, one of the most abundant arthropods in soil, showed an extensive number of carbohydrate-active enzymes potentially acquired by the HGT from soil bacteria. These enzymes enable F. candida to scavenge decaying plant matter as a food source [25]. As RLN are composed of a vast number of species with all of them occurring in the soil, it will be interesting to investigate whether they also carry a PME-coding gene.
Supporting information S1 Table. List of primers used in this study.  Table. List of PME sequences, and respective species, used for phylogenetic analysis of Pp-PME. These sequences of each particular taxa correspond to the top BLAST hits against P. penetrans PME sequence. (XLSX) S4 Table. Codon usage of PMEs of representative species of different taxa. The preferred codons are highlighted in red. (XLSX) S5 Table. Similarity matrix among the different PME sequences obtained for different groups of organisms including the PME sequence of Pratylenchus penetrans. (XLSX) S6 Table. List of peptides and corresponding protein identification generated through mass spectrometry analyses.