StPIP1, a PAMP-induced peptide in potato, elicits plant defenses and is associated with disease symptom severity in a compatible interaction with Potato virus Y

Abstract

The role of small secreted peptides in plant defense responses to viruses has seldom been investigated. Here, we report a role for potato (Solanum tuberosum) PIP1, a gene predicted to encode a member of the pathogen-associated molecular pattern (PAMP)-induced peptide (PIP) family, in the response of potato to Potato virus Y (PVY) infection. We show that exogenous application of synthetic StPIP1 to potato leaves and nodes increased the production of reactive oxygen species and the expression of plant defense-related genes, revealing that StPIP1 triggers early defense responses. In support of this hypothesis, transgenic potato plants that constitutively overexpress StPIP1 had higher levels of leaf callose deposition and, based on measurements of viral RNA titers, were less susceptible to infection by a compatible PVY strain. Interestingly, systemic infection of StPIP1-overexpressing lines with PVY resulted in clear rugose mosaic symptoms that were absent or very mild in infected non-transgenic plants. A transcriptomics analysis revealed that marker genes associated with both pattern-triggered immunity and effector-triggered immunity were induced in infected StPIP1 overexpressors but not in non-transgenic plants. Together, our results reveal a role for StPIP1 in eliciting plant defense responses and in regulating plant antiviral immunity.

Pathogen-associated molecular patterns (PAMPs), pattern-triggered immunity, peptides, potato, symptoms, viruses

Introduction

Plants have evolved a multi-layer immune system to allow them to deal with the threat of pathogens such as bacteria, fungi, and viruses (Jones and Dangl, 2006; Wang et al., 2019). A first layer of defense is provided by cell surface receptors called pattern recognition receptors (PRRs) that detect conserved features of pathogens termed pathogen-associated molecular patterns (PAMPs) in the extracellular space (Boutrot and Zipfel, 2017). Upon detection of PAMPs by PRRs, plant cells initiate immune responses including release of Ca²⁺ ions, reactive oxygen species (ROS), increased expression of pathogen response genes, and deposition of callose at the site of infection, ultimately producing increased resistance to pathogens termed pattern-triggered immunity (PTI) (Jones and Dangl, 2006). Because viruses are obligate intracellular parasites, the extent to which PTI is involved in plant defense against viruses has been understated. Recent studies have revealed that virus components can act as PAMPs and trigger PTI-like responses through PRR co-receptors such as SERK1 and NIK1 (Niehl et al., 2016; Zvereva et al., 2016; Gouveia et al., 2017).

To overcome PTI, pathogens have evolved proteins called effectors that act to disable the plant innate immunity and allow them entry into the cell or greater access to host resources. Plants, in turn, have evolved intracellular receptors, often called R proteins, which detect effectors and initiate effector-triggered immunity (ETI) (Jones and Dangl, 2006). Detection of effectors by R proteins triggers an intense immune response, sometimes resulting in programmed cell death, a reaction known as the hypersensitive response (HR) (Valkonen et al., 2017).

Plants have a diverse array of small endogenous peptides, also known as phytocytokines, which are released from pathogen-challenged cells (Gust et al., 2017). These small secreted peptides (SSPs), such as the plant elicitor peptides (PEPs) and PAMP-induced peptides (PIPs), trigger or modulate PTI-like immune responses in neighboring cells, priming them to defend against an oncoming infection. The PIPs in Arabidopsis, like other SSPs, are produced as precursor polypeptides (pre-propeptides) with N-terminal signal sequences recognized by the secretion pathway. After entering the secretion pathway, the N-terminal signal sequence is removed, and the resulting propeptide further processed into small (~15–25 amino acids) mature peptides. Ultimately, the fully mature peptides are released into the extracellular space where they are perceived, like PAMPs, by PRR-like receptors on neighboring cells (Hou et al., 2014; Matsubayashi, 2018). In Arabidopsis, AtPIP1 and AtPIP2 are expressed in response to PAMPs such as flagellin and chitin, and trigger PTI-like immune responses, including ROS and defense gene expression, in perceiving cells (Hou et al., 2014). AtPIP3 was shown to modulate plant immunity by regulating crosstalk between salicylic acid and jasmonic acid signaling pathways (Najafi et al., 2020).

Potato virus Y (PVY) is the type member of the largest group of RNA plant viruses, the Potyviridae (Wylie et al., 2017). Its host range is broad, infecting most solanaceous species including potato, tomato, peppers, and tobacco, in addition to other plant groups. PVY is listed as one of the top 10 plant viruses in terms of scientific and economic importance (Scholthof et al., 2011). PVY is a single-stranded, positive-sense RNA virus with a genome of 9.7 kb, and exists as a large number of strains, variants, recombinants, and isolates. The most commonly found strains in growers’ fields are the necrotic strain N and recombinants between the N and O strains, N-Wilga, NTN, and N:O. The O strain occurrence has been declining to low levels in recent years (Karasev and Gray, 2013; Funke et al., 2017).

The cultivated potato, Solanum tuberosum L., is the fourth most cultivated staple food crop worldwide (FAOSTAT, 2017). Because of its high level of heterozygosity, potato is propagated vegetatively by using tubers as seeds to ensure genetic identity of the progeny. To maintain low levels of pathogens in potato seed production, seed lots are regularly inspected. Due to its prevalence, PVY is currently the number one reason for rejection of seed lots (Karasev and Gray, 2013). The symptoms caused by PVY infection in potato vary depending on the viral strain and host cultivar. Common symptoms include foliar mosaic, rugose mosaic, leaf wrinkling, and various necrotic lesions (Lacomme and Jacquot, 2017).

In a previous study, we profiled PVY-induced changes in the transcriptome of potato cultivar Premier Russet (PR), and identified PGSC0003DMG400014879, as the most significantly differentially expressed gene (DEG) in an incompatible interaction with PVY^O (Goyer et al., 2015). This prompted us to investigate the function of this gene in the potato–PVY interaction. In this study, we provide evidence that the gene PGSC0003DMG400014879 encodes a peptide that belongs to the PIP family, and named it StPIP1. Transgenic potato plants overexpressing StPIP1 produced clear rugose mosaic symptoms that were absent or very mild in control plants when infected with a compatible strain, PVY^NTN. Our transcriptomics data showed that marker genes of PTI and ETI were induced in infected StPIP1 overexpressors but not in non-transgenic plants. This study reveals a function for plant PIP peptides in antiviral immunity.

Materials and methods

Sequence analyses

Full-length genomic, transcript, and polypeptide sequences for PGSC0003DMG400014879 were retrieved from the Spud DB (http://solanaceae.plantbiology.msu.edu/) (Hirsch et al., 2014). Sequence data for Arabidopsis genes were retrieved from The Arabidopsis Information Resource (TAIR10) at www.arabidopsis.org (Berardini et al., 2015). The BLAST suite from the NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi) was used to find related genes and proteins (Altschul et al., 1990). For putative members of small secreted peptide families in tomato and potato, published peptide sequences of each family (CLV3/CLE, IDA/IDL, CEP, and PIP/PIPL) in Arabidopsis were used as queries for tBLASTN and BLASTp searches (Supplementary Table S1). Signal peptides were predicted using the programs Phobius (http://phobius.sbc.su.se/) and SignalP-5.0 (http://www.cbs.dtu.dk/services/SignalP/) (Kall et al., 2007; Almagro Armenteros et al., 2019). The programs Predotar (https://urgi.versailles.inra.fr/predotar/) (Small et al., 2004), PSORT (http://psort1.hgc.jp/form.html) (Nakai and Horton, 1999), and TargetP (http://www.cbs.dtu.dk/services/TargetP/) (Emanuelsson et al., 2000) were used to predict subcellular localizations. Three-dimensional structure prediction was done with PHYRE2 (http://www.sbg.bio.ic.ac.uk/phyre2/html/page.cgi?id=index) (Kelley et al., 2015). Cis-regulatory elements were searched in PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) (Lescot et al., 2002). Multiple sequence alignments were performed with Muscle using default settings (Edgar, 2004), and phylogenetic trees were constructed from those alignments using the maximum likelihood method, 1000 times bootstrapped with Mega7 (https://www.megasoftware.net/) (Kumar et al., 2016).

Plant growth

Potato plants of the cultivar PR were propagated in vitro on solid Murashige and Skoog (MS) medium (1× MS-modified BC potato salts, 2% sucrose, 100 mg l^–1 myo-inositol, 2 mg l^–1 glycine, 0.5 mg l^–1 nicotinic acid, 0.5 mg l^–1 pyridoxine, 0.1 mg l^–1 thiamine, pH 5.6). After 3–4 weeks, plants were transferred to 1 gallon pots filled with soil (four parts potting mix, one part sand) containing slow-release fertilizer (Osmocote Plus) in the greenhouse. Greenhouse temperature conditions were set at 21 °C day, 15 °C night. Supplemental light was provided by 400 W high-pressure sodium lamps to maintain a 14 h photoperiod. Plants were arranged in a randomized split-block design with six plants per treatment. Treatments included inoculation with two different strains of PVY (O or NTN) and a mock inoculation control. For assessing virus translocation to tubers, three progeny tubers from each plant were selected, treated with 7 ppm gibberellin (GA₃), incubated for 2–4 weeks at 27 °C to break dormancy, and planted to 1 gallon pots (three tubers per pot) filled with potting mix.

PVY stocks and inoculation

The PVY strain O isolate used in this study was first identified in a potato tuber from Aberdeen, ID, USA in 1999 by James Crosslin (USDA/ARS). The PVY NTN HR1 isolate (Genebank ID FJ204166) was donated by Dr Alexander Karasev (University of Idaho). Mechanical inoculation of potato leaves was done as previously described (Vinchesi et al., 2017).

PVY detection

PVY was detected by reverse transcription–PCR (RT–PCR). Nucleic acids were extracted using a protocol adapted from Dellaporta et al. (1983). Briefly, three upper leaflets per plant were excised, placed into a mesh bag (Agdia®), and pulverized in a buffer containing 100 mM Tris–HCl (pH 8.0), 50 mM EDTA, 500 mM NaCl, and 10 mM 2-mercaptoethanol. A 70 µl aliquot of 10% SDS was added to a 600 μl aliquot of the resulting slurry, mixed, and the sample was incubated at 65 °C for 10 min. To each sample, 200 μl of 5 M acidified potassium acetate (pH 5.7) was added and samples were incubated on ice for 10 min. After centrifugation at 15 900 g for 10 min, the supernatant was transferred to a new tube. After precipitation with 300 μl of cold isopropanol, samples were centrifuged, and the pellet was washed with 70% ethanol, and resuspended in 400 μl of deionized water. Nucleic acid extracts were used as templates to synthesize cDNAs with M-MuLV reverse transcriptase utilizing a mixture of random hexamers and oligo(dT)₁₈ primers. The resulting cDNAs were used as templates in a multiplex PCR assay as previously described (Lorenzen et al., 2006). Primers sequences are shown in Supplementary Table S2.

Molecular cloning

StPIP1-overexpressing plants

Total RNAs were extracted from leaves from the potato variety PR using the hot phenol method as described previously (Goyer et al., 2015) and treated with DNase (Ambion® DNA-free™ kit, LifeTechnologies). cDNAs were synthesized by M-MuLV reverse transcriptase (New England Biolabs) using an oligo(dT)₁₈ primer, and the PGSC0003DMG400014879-encoded cDNA was amplified using PrimeSTAR Max DNA Polymerase (Takara) using the forward and reverse primers shown in Supplementary Table S2. The 658 bp amplicon was directly cloned into pCR™Blunt TOPO^® vector (ThermoFisher Scientific), and the resulting construct was introduced into One Shot TOP10 Escherichia coli cells (ThermoFisher Scientific). Sixteen kanamycin-resistant isolated colonies were then cultured in LB medium supplemented with 50 mg l^–1 kanamycin. Plasmid DNA was extracted from each culture and sent for Sanger sequencing. Sequence alignment showed that PGSC0003DMG400014879 has four alleles encoding three protein isoforms in PR. Clone 1-1, which represents the most dominant allele, was used as template to amplify a 312 bp amplicon using the forward and reverse primers shown in Supplementary Table S2. The 312 bp amplicon was then ligated into the pDONR™/Zeo vector (ThermoFisher Scientific) using BP clonase following the manufacturer’s recommendations, and then subcloned into the pMDC32 plant binary vector (Curtis and Grossniklaus, 2003) by recombination using LR clonase. The final construct was verified by restriction digestion and Sanger sequencing.

StPIP1-silenced plants

Artificial miRNAs (amiRNAs) targeting PGSC0003DMG400014879 were designed using the program WMD3 (http://wmd3.weigelworld.org). The full-length transcript (PGSC0003DMT400038539) was retrieved from SpudDB and used as a target for the WMD3 designer program. The transcript library ‘Solanum_tuberosum_v183.mRNA.PUT.fasta’ was used to check specificity, with the program set to accept no predicted off-targets. The amiRNA hairpin precursors were produced by overlapping PCR following a procedure recommended by the authors of WMD3 (Ossowski et al., 2008). The first fragment, a, was 424 bp and was amplified from the plasmid pRS300 (Addgene), and the second fragment, b, was 301 bp and was amplified from pRS300 using the primers shown in Supplementary Table S2. Fragments a and b were gel purified and a 1:1 ratio of each was used as template to amplify the 701 bp fragment c using pRS300a and pRS300b. Fragment c was gel purified and the 274 bp fragment d was amplified using fragment c as template and using the forward and reverse primers pRS300a and 5′-GAAAGATTGCTAACAACCGTTTATCTACATATATATTCCT-3′, respectively. Fragment c was again used as template to amplify the 451 bp fragment e using the forward and reverse primers 5′-GATAACGGTTGTTAGCAATCTTTCACAGGTCGTGATATG-3′ and pRS300b, respectively. Fragments e and d were gel purified and a 1:1 ratio of each was used as template to amplify the 705 bp fragment f using the forward and reverse primers 5′-CACCCTGCAAGGCGATTAAGTTGGGTAAC-3′ and pRS300b, respectively. Fragment f was cloned into pENTR™/D-TOPO^® (ThermoFisher Scientific) using the TOPO cloning reaction following the manufacturer’s recommendations. The insert was released by digestion with ApaI and SacI restriction enzymes, gel-purified, and subsequently ligated into the binary vector pMDC32 previously digested with ApaI and SacI under control of the Cauliflower mosaic virus (CaMV) 35S promoter. The final construct was named ‘pMDC32-PIPmiRNA’. Primers sequences are all shown in Supplementary Table S2.

Potato transformation

DNA constructs were introduced into the potato cultivar PR by Agrobacterium tumefaciens (strain EHA105)-mediated stable transformation as previously described (Chetty et al., 2015). Briefly, single isolated A. tumefaciens colonies containing the transformation vector were grown in 50 ml of YEP culture to saturation. A 10 ml aliquot was pelleted, and cells were resuspended in 40 ml of MS medium supplemented with 200 µM acetosyringone to an OD₆₀₀ of 0.8. Potato stem internodes (~5–10 mm long) were incubated for 15 min in the Agrobacterium suspension in a 50 ml Falcon tube with gentle shaking. Internodes were then blotted dry on Whatman paper and placed on Petri dishes containing callus-inducing medium (CIM) (MS medium supplemented with 0.2 mg l^–1 1-napthalenic acetic acid, 0.02 mg l^–1 GA₃, 2.5 mg l^–1trans-zeatin riboside) supplemented with 200 µM acetosyringone and overlaid with sterile Whatman filter paper in the dark for 2 d at room temperature. Internodes were then washed with water supplemented with 250 mg l^–1 cefotaxime, blotted dry, and transferred to Petri dishes containing CIM supplemented with 20 mg l^–1 hygromycin, 250 mg l^–1 cefotaxime, and 200 mg l^–1 carbenicillin. After incubation for 2 weeks, explants were transferred to shoot-inducing medium (SIM) (MS medium supplemented with 0.02 mg l^–1 1-napthalenic acetic acid, 0.02 mg l^–1 GA₃, and 2 mg l^–1trans-zeatin riboside) supplemented with 20 mg l^–1 hygromycin, 250 mg l^–1 cefotaxime, and 200 mg l^–1 carbenicillin. Explants were transferred to new SIM every 2 weeks. When shoots grew to ~1 cm in length (after ~8 weeks on SIM), they were excised and transferred to MS medium supplemented with 20 mg l^–1 hygromycin, 250 mg l^–1 cefotaxime, and 200 mg l^–1 carbenicillin. Plantlets that developed roots under hygromycin selection were then genotyped by PCR.

RT–qPCR

For RT–qPCR from whole leaflets, total RNAs were extracted as described before (Goyer et al., 2015). cDNAs were synthesized as described above except that oligo(dT)₁₈ only was used. cDNAs were used as template for quantitative PCR using Brilliant III Ultra-Fast SYBR® Green QPCR Master Mix (Agilent). Primers targeting StPIP1, PVY, and reference genes 18S rRNA, L2, and EF1α are shown in Supplementary Table S3. Details of RT–qPCR conditions are shown in Supplementary Table S5 following the Minimum Information for publication of Quantitative Real-Time PCR Experiments (MIQE) guidelines (Bustin et al., 2009; Graeber et al., 2011).

For RT–qPCR on potato leaf discs, total RNAs were extracted and cDNAs were synthesized as described before (Moroz et al., 2017). Primers used to measure expression of defense-related genes (StPR1b, StPR5, StWRKY, StERF3, StPAL1, and StJas) and reference genes (StUbq and StEF1-alpha) are described in Supplementary Table S4. Details of the workflow according to the MIQE guidelines are shown in Supplementary Table S6.

Calculations were done according to published methods (Schmittgen and Livak, 2008; Taylor et al., 2019). Statistical analyses were done using ANOVA or Student’s t-test from the log-transformed normalized expression.

QuantSeq analysis

Total RNAs were extracted from upper leaves (one leaflet from each of three plants) of PR and StPIP1-overexpressing (PIP-OE) plants infected or not with PVY^NTN [44 days post-inoculation (dpi), 70 d after transplantation] as described before (Goyer et al., 2015). RNAs were then sent to the Core Labs of the Oregon State University Center for Genome Research and Biocomputing for RNA quality control, library preparation, and sequencing. RNA quality was checked with an Agilent 2100 bioanalyzer (Plant RNA Nano Chip, Agilent). Libraries were prepared from 500 ng of RNA using the QuantSeq 3′ mRNA-Seq Library Prep Kit FWD for Illumina following the manufacturer’s recommendations (Lexogen). Library size was verified on an Agilent TapeStation 4200 using High Sensitivity D5000 Screen Tape®, and libraries were quantified by qPCR before sequencing on an Illumina HiSeq3000 (50 bp single end). Read quality was verified using FASTQC. Adaptors and poly(A) tails were removed from reads using cutadapt. Trimmed reads were then aligned to the potato reference genome (DM_v4.04) using the STAR aligner. Output alignment .bam files were used to calculate the number of reads mapping to exons using HTSeq (Anders et al., 2015) in the ‘Intersection (non-empty)’ mode. Differentially expressed exons were determined from HTSeq count tables using DESeq2 (Love et al., 2014) with parametric fit. Functional enrichment analysis of genes that were significantly differentially expressed was done with g:GOSt in g:Profiler (Raudvere et al., 2019) (https://biit.cs.ut.ee/gprofiler/gost). Genes were considered significantly differentially expressed if they had adjusted P-values (q) ≤0.05.

Measurements of second messengers Ca²⁺ and ROS

To measure cytosolic Ca²⁺ concentration, an aequorin-based luminescence assay was performed using the aequorin-expressing transgenic potato cultivar Désirée. Procedures for reconstitution, luminescence measurement, and data analysis and normalization were described in a previous publication (Moroz and Tanaka, 2020). Leaf discs (5 mm diameter) were harvested from 5- to 6-week-old plants and used for the assay. For ROS measurement, a luminol-based chemiluminescence assay was performed as described previously (Moroz and Tanaka, 2020). Leaf discs (5 mm diameter) and nodes (5 mm long) were harvested from 5-week-old plants and used for the assay. Results were expressed as relative light units (RLUs per tissue) after subtraction of the data at time 0 from those at each time point of the measurement.

Callose analysis

Callose analysis was as previously described (Adam and Somerville, 1996; Gomez-Gomez et al., 1999), with minor modifications. Potato leaflets (~10 weeks after transfer from tissue culture to soil) were placed in a 3:1 (v/v) solution of ethanol and lactophenol (1:1:1:1 v/v/v/v phenol:glycerol:lactic acid:water) and left stationary for 1 week. The cleared leaflets were then incubated sequentially in 50% ethanol overnight, 67 mM K₂HPO₄ (pH 12) for 1 h, and 0.01% aniline blue in 67 mM K₂HPO₄ (pH 12) for 1 h. The midvein of each potato leaflet was then removed and half of the leaflet was mounted in 70% glycerol, K₂HPO₄ (pH 12) on a glass microscope slide. Callose deposits were detected by UV epifluorescence using a Leica MZFLIII stereomicroscope. The entire half leaflet was examined for callose deposits, and 2–3 representative pictures were taken from six leaflets per genotype. Callose spots were counted in a 1 mm² area in the center of each picture. Graphed data are the average callose spots in the 1 mm² area from 14 pictures per genotype, and error bars represent the SE of the data.

Results

Bioinformatics analyses predict that PGSC0003DMG400014879 belongs to the family of PAMP-induced peptides

In a previous study, we identified PGSC0003DMG400014879 as the most highly repressed gene in the cultivar PR in response to PVY^O inoculation (Goyer et al., 2015). This gene is annotated as an ATPase-binding cassette (ABC) transporter family protein in the potato genomics resource database Spud DB. However, ABC transporters are made of four major subunits with two transmembrane hydrophobic domains and two nucleotide-binding domains, and contain the amino acid signature sequence [LIVMFY]S[SG]GX3[RKA][LIVMYA]X[LIVFM] as consensus (Kang et al., 2011). In contrast, the protein predicted to be encoded by PGSC0003DMG400014879 does not contain the canonical signature sequence. Furthermore, three-dimensional structure prediction analysis using PHYRE2 indicated no secondary structure apart from an α-helix in the predicted N-terminal signal peptide region (see below) (Supplementary Fig. S1). These observations indicated that the PGSC0003DMG400014879 gene is not correctly annotated, possibly due to similarities between the C-terminal part of the PGSC0003DMG400014879-encoded protein and protein sequences of P-loop NTPase superfamily members, which include ABC transporters (Pathak et al., 2014).

A BLASTp search of non-redundant protein sequences using the predicted protein product of PGSC0003DMG400014879 as the query sequence resulted in many matches with annotations such as ‘hypothetical’ or ‘uncharacterized’. Eleven hits were from the Solanaceae (E-values ≤10^–20, coverage ≥89%, identity ≥55%) and were often predicted to be small (<100 amino acids) polypeptides, while there were no significant matches with ABC transporter-annotated sequences. One hit (47% identity, 61% similarity, 97% coverage) was annotated as ‘precursor of CEP16-like’ (from Hevea brasiliensis, sequence ID: XP_021642728.1). CEPs (C-terminally encoded peptides) are a class of secreted peptides (Roberts et al., 2013). These results suggested that PGSC0003DMG400014879 may belong to a family of genes encoding small secreted peptides. To confirm this hypothesis, we performed phylogenetic analyses with amino acid sequences from small secreted peptides including CLAVATA3 (CLV3/CLE) (Yamaguchi et al., 2016), CEP (Roberts et al., 2013), INFLORESCENCE DEFICIENT IN ABSCISSION (IDA)/IDA-Like (IDL) (Vie et al., 2015), and PIP/PIP-Like (PIPL) (Vie et al., 2015; Najafi et al., 2020) families from Arabidopsis and tomato, two nearby paralogs of PGSC0003DMG400014879 on chromosome 3, PGSC0003DMG400014880 and PGSC0003DMG400014833, and one gene, PGSC0003DMG400024991, on chromosome 2 with 63% similarity (Fig. 1). The encoded potato peptides grouped within the PIP and CEP branches, most closely to AtPIP2 and AtPIP3. Based on these results, we named PGSC0003DMG400014879, PGSC0003DMG400014880, PGSC0003DMG400014833, and PGSC0003DMG400024991 as StPIP1, StPIP2, StPIP3, and StPIP4, respectively.

Fig. 1.

Maximum-likelihood phylogenetic tree of the small secreted peptide families CLV3/CLE, CEP, IDA/IDL, and PIP/PIPL. The sequences of CLV3/CLE, CEP, IDA/IDL, and PIP/PIPL proteins from Arabidopsis were retrieved from Goad et al. (2017) (CLV3/CLE), Roberts et al (2013) (CEP), and Vie et al (2015) (IDA/IDL, PIP/PIPL). Sequences for 10 CLE peptides from tomato (IDs starting with ‘Sl’) were retrieved from Zhang et al. (2014). Sequences of four previously uncharacterized paralogous secreted peptides from potato (StPIP1, StPIP2, StPIP3, and StPIP4) were retrieved from SpudDB (http://solanaceae.plantbiology.msu.edu/). Genes encoding StPIP1, StPIP2, and StPIP3 are paralogs located on chromosome 3 [PGSC0003DMG400014879 (StPIP1), PGSC0003DMG400014880 (StPIP2), and PGSC0003DMG400014874 (StPIP3)]. Gene names ending with an asterisk are putative members of small secreted peptide families from tomato and potato that were found using the BLAST suite from the NCBI. The tree was constructed using MEGA7 with 1000 bootstraps. A few CEP-annotated peptides grouped with PIP or IDL/IDA families.

A multiple sequence alignment highlights the similarity and identity between the Arabidopsis PIPs and the potato proteins (Fig. 2). Particularly conserved are the ‘RPL’ motif defining the predicted signal peptide cleavage point (see below), and the ‘GPS(P)xGxGH’ motif within the propeptides (Hou et al., 2014). The propeptides of StPIP1, 2, and 3 have two conserved ‘GPS(P)xGxGH’ motifs, while AtPIP1 and StPIP4 have only one (Hou et al., 2014; Vie et al., 2015; Najafi et al., 2020).

Fig. 2.

Multiple sequence alignment comparing the PIP pre-propeptides of Arabidopsis (AtPIP1–AtPIP3) with those of potato (StPIP1–StPIP4). Genes encoding StPIP1, StPIP2, and StPIP3 are paralogs located on chromosome 3 [PGSC0003DMG400014879 (StPIP1), PGSC0003DMG400014880 (StPIP2), and PGSC0003DMG400014874 (StPIP3)]. The dotted line indicates the predicted signal peptide cleavage motif, and solid lines indicate the core PIP motifs, ‘GPSP’ and ‘GxGH’, that mark the C-terminal end of the mature PIP peptides. The alignment was done with ClustalW and shading was done with BoxShade.

In agreement with a functional prediction as a secreted peptide, the programs Predotar, PSORT, and TargetP predicted a subcellular localization in the ‘endoplasmic reticulum’ (99% probability), ‘outside’ (74% probability), or the ‘secretory pathway’ (98% probability), respectively. The programs Phobius and SignalP5.0 predicted that the N-terminal signal peptide is cleaved between the 24th and 25th residues at the SEARP motif between the alanine and the arginine (Supplementary Fig. S2).

We previously showed that StPIP1 transcripts are present at low levels in PR leaves (Goyer et al., 2015). To find out whether StPIP genes are expressed in other potato tissues, we searched gene expression data in the Expression Atlas (https://www.ebi.ac.uk/gxa/home) (Papatheodorou et al., 2020). Only StPIP1 and StPIP4 had detectable expression levels. Both genes are expressed in the petiole [transcripts per million (TPM) values of 1 and 0.6, respectively], while StPIP1 is also expressed in the shoot apex (TPM=0.7). In addition, searches for potential stimuli in the available data sets showed that StPIP1, StPIP2, and StPIP4 are all induced in response to Phytophthora infestans (log₂ fold changes of 1.8, 2.3, and 2.4, respectively) in a Russet Burbank (RB) line expressing the Phytophthora infestans resistance gene Rpi-blb1 (Gao et al., 2013).

To find further clues as to its function, we searched for cis-regulatory elements in the 1000 bp region upstream of the start codon of StPIP1 and found several (biotic) stress responsive cis-regulatory elements (Supplementary Table S7), indicating a possible function of StPIP1 in (biotic) stress response.

Together, sequence analyses indicated that the gene PGSC0003DMG400014879 encodes a putative PIP family member and is expressed in response to biotic stresses.

Exogenous application of StPIP1 induces ROS production and expression of defense-related genes in potato

To test for possible bioactivity of the StPIP1-derived peptides, we measured changes in the early stages of defense responses to pathogens, which usually involve, amongst others, increases in Ca²⁺ cytosolic concentrations, increases in ROS production, and changes in expression of plant immunity-associated genes (Yu et al., 2017), upon exogenous application of chemically synthesized StPIP1. Since we did not know the exact composition of possible final mature StPIP1 peptides, we used four chemically synthesized variants of the StPIP1 propeptide: StPIP1_long that contains both of the conserved ‘GPSPxGxGH’ motifs and has the first proline of the ‘GPSP’ motif hydroxylated (hydroxylation of the first proline of the ‘GPSP’ motif has been shown to increase peptide activities in Arabidopsis; Hou et al., 2014); StPIP1_short that contains only the second ‘GPSPxGxGH’ motif and has the first proline of the motif hydroxylated; StPIP1_short_NoHY which is identical to StPIP1_short but has no hydroxylated proline; and StPIP1_short_NoC which is identical to StPIP1_short but has the last eight C-terminal amino acids deleted (Supplementary Table S8).

First, we measured changes in cytosolic Ca²⁺ concentration in leaves of 5- to 6-week-old aequorin-expressing potato cultivar Désirée plants in response to exogenous treatment with the synthetic StPIP1-derived peptides. None of the peptides stimulated cytosolic Ca²⁺ transients, whereas a known elicitor peptide, StSystemin, triggered increased cytosolic Ca²⁺ 5–7 min after peptide application (Supplementary Fig. S3), showing a lack of function of StPIP1 in Ca²⁺ signaling. Second, we measured apoplastic ROS production in nodes and leaves of potato cultivars RB, which is susceptible to all PVY strains, and PR. In nodes, all peptides tested induced ROS production in RB, except StPIP1_short, while StPIP1_short_NoC and StPIP1_short_NoHY induced ROS in PR (Fig. 3A, B; Supplementary Fig. S4). Similar to the Ca²⁺ measurements in leaves, none of the peptides stimulated ROS production in leaves, except for StPIP1_short_NoHY in PR (Supplementary Fig. S4). Overall, the variant StPIP1_short_NoHY elicited the highest levels of ROS production. Last, we used RT–qPCR to measure changes in the abundance of mRNA transcripts derived from a selection of known defense-related marker genes in response to peptide treatment (Fig. 4). Overall, if considering only transcripts with a >2-fold change in abundance, RB was more responsive to all peptide treatments than PR (Fig. 4). In RB, levels of all transcripts significantly increased in response to all StPIP1-derived peptides (Fig. 4A), and the magnitude of response was comparable with responses induced by StSystemin (Supplementary Fig. S5), except StERF3 and StJas that were more weakly induced by StPIP1 peptides. In PR, StPIP1_short_NoC induced the expression of StPR1b by at least 4-fold (Fig. 4B). These results show that StPIP1 peptide variants are able to induce the expression of defense-related genes in potatoes.

Fig. 3.

ROS production in potato leaves and nodes in response to StPIP1 and its variants. (A) Nodes of Russet Burbank (RB) and (B) nodes of Premier Russet (PR). All peptides were added at the final concentration of 1 µM. RLU is presented as a result of subtraction of RLU₀ (at time 0) from RLU_t (at each time point of the measurement). Line graphs are shown as mean values ±SE (n=8).

Fig. 4.

Expression of defense-related genes in leaves in response to StPIP1 and its variants. (A) Russet Burbank (RB) and (B) Premier Russet (PR). Leaf discs were treated for 30 min with the indicated peptides at the final concentration of 1 µM. The expression levels of potato defense-related genes were monitored by RT–qPCR. Data are shown as normalized fold expression compared with mock control (2^-ΔΔCt). Two reference genes, StEF1α and StUbq, were used for normalization. Histogram bars are mean values ±SE. Unpaired Student’s t-test from log-transformed values was used for statistical analysis (*P<0.05; **P<0.01; ***P<0.001). Note the difference of scale between the two graphs.

Overexpressors of StPIP1 show differences of symptoms in compatible reactions with PVY

To further investigate the role of StPIP1 in the defense response to PVY, we generated transgenic PR potato plants that either overexpress or silence the expression of StPIP1. We identified three independent StPIP1-overexpressing lines, PIP-OE1, PIP-OE8, and PIP-OE14, that have different increased levels of expression of StPIP1 compared with control PR, as determined by RT–qPCR (Supplementary Fig. S6). When grown in vitro, PIP-OE1 and PIP-OE14 plantlets were shorter than control PR and had short internodes and relatively small leaves (Supplementary Fig. S7). However, this phenotype was not apparent when plants were grown in soil in a greenhouse (Supplementary Fig. S7). We also identified three independent amiRNA lines, Pami5.2, Pami8, and Pami9, that had lower transcript levels as determined by RT–qPCR (Supplementary Fig. S8). Under standard greenhouse conditions, lines Pami5.2 and Pami9 showed some novel phenotypic characteristics, such as mild chlorosis and leaf wrinkling in the upper leaves. Line Pami8 showed a severe developmental phenotype characterized by stunting, a prostrate growth habit, and severe wrinkling of the leaves when grown in vitro or in a greenhouse (Supplementary Fig. S8).

We then inoculated StPIP1-overexpressing and -silenced plants with PVY^O (incompatible interaction) or PVY^NTN (compatible interaction), and monitored the rate (i.e. number of plants showing localized necrotic lesions) and onset (i.e. time of first appearance of localized necrotic lesions) of HR on the inoculated leaves. We performed two repeated experiments with StPIP1-overexpressing lines, and one experiment with StPIP1-amiRNA lines. As expected, in all three experiments, most or all of the non-transgenic PR plants developed localized round necrosis characteristic of a HR on leaves inoculated with PVY^O, but no HR-like symptoms were observed on PR leaves inoculated with PVY^NTN (Supplementary Fig. S9; Supplementary Tables S9–S11). The mean time for onset of HR varied between experiments from ~9 to ~20 dpi (Supplementary Tables S9–S11). However, in StPIP1-overexpressing and -silencing lines, the rate and onset of HR caused by PVY^O were similar to those in untransformed PR, the only significant difference being a delay in the onset of HR in Pami5.2 (Supplementary Tables S9–S11). These results showed that modulating the expression of StPIP1 had no or little effect on rate and onset of HR, indicating that StPIP1 may not play an important role in PVY-induced HR. It is noteworthy that the HR in PR seemed to be mild because the initial localized necrosis usually did not expand from the initial point of appearance and did not lead to leaf drop as would be observed for a robust HR.

Next, we used RT–PCR to assess PVY systemic infection. We tested systemic leaves from inoculated plants (i.e. in-season infection) and leaves from tuber progeny (i.e. seedborne infection) (Supplementary Tables S9–S11; Supplementary Figs S10–S12). In two of three trials (Supplementary Tables S10, S11; Supplementary Figs S11, S12), we could not detect the virus in any of the PR plants inoculated with PVY^O. In the first trial (Supplementary Table S9; Supplementary Fig. S10), although all PR plants inoculated with PVY^O tested positive in in-season leaves, only half of the progeny plants tested positive. Together, these results indicate a certain degree of resistance of PR to PVY^O, consistent with the mild HR observed. However, there was no significant difference in infection rates between PR and either StPIP1-overexpressing or -silenced lines in response to PVY^O(Supplementary Tables S9–S11; Supplementary Figs S10–S12). In the case of PVY^NTN, in two of the trials (Supplementary Tables S9, S11; Supplementary Figs S10, S12), all PR plants tested positive for the virus in both in-season leaves and tuber progeny, consistent with a lack of resistance of PR to this strain. In the second trial, only two out of six plants inoculated with PVY^NTN tested positive for the virus in both in-season leaves and tuber progeny (Supplementary Table S10; Supplementary Fig. S11), which may be due to technical failure, but the infection rate was still higher than in plants inoculated with PVY^O (i.e. zero out of six plants tested positive). This may explain the significant differences observed between PR and overexpressing lines in that trial (Supplementary Table S10; Supplementary Fig. S11). Otherwise, there was no significant difference in infection rates between PR and either overexpressing or silencing lines (Supplementary Tables S9, S11; Supplementary Figs S10, S12). These results showed that modulating the expression of StPIP1 had no effect on the rate of systemic infection with either PVY^O or PVY^NTN strains.

Finally, in PVY-inoculated PR, StPIP1-overexpressing, and StPIP1-silenced lines, we monitored the development of systemic symptoms on systemic leaves. In all experiments, there were no symptoms or very mild mosaic symptoms observed on non-inoculated leaves in either transgenic or control PR plants that were inoculated with PVY^O (Supplementary Tables S9–S11). Likewise, PR plants inoculated with PVY^NTN did not show symptoms or produced only mild symptoms in three out of six plants at >50 dpi in one of the trials (Supplementary Table S9). In contrast, PIP-OE1 and PIP-OE14 (and PIP-OE8 in our first trial) inoculated with PVY^NTN produced clearly visible rugose mosaic symptoms starting as early as ~30 dpi (Fig. 5; Supplementary Tables S9, S10). This observation was consistent throughout experiments. We hypothesized that the strong phenotypic reaction of StPIP1-overexpressing plants infected with PVY^NTN is due to higher virus amounts in leaf tissues. To test this hypothesis, we measured the amount of viral RNA relative to two reference genes, L2 and EF1α, by RT–qPCR in leaves of PR and PIP-OE1 infected with PVY^NTN 44 dpi. The relative amount of viral RNA was >2-fold lower in leaves of PIP-OE1 compared with that in leaves of PR (Fig. 6; Supplementary Fig. S13), and this difference was statistically significant (P<0.05), rejecting our initial hypothesis that rugose mosaic symptoms are due to higher viral load. Instead, our results indicated that overexpression of StPIP1 decreased the viral load during the compatible interaction with PVY^NTN, suggesting that the associated rugose mosaic symptoms may be due to an increased plant defense response.

Fig. 5.

Symptoms of StPIP1-overexpressing lines compared with Premier Russet (PR) upon inoculation with PVY^NTN. Symptoms including rugose mosaic, stunting, and chlorosis were clearly visible in StPIP1 overexpressors infected with PVY^NTN. (A) Mock- and PVY^NTN-inoculated PR and StPIP1-overexpressing line #1 (PIP-OE1) 44 d after inoculation. (B) Close-up images of canopy leaves of mock- and PVY^NTN-inoculated PR and PIP-OE1 47 d after inoculation. (C) Three independent StPIP1-overexpressing lines (PIP-OE14, PIP-OE8, and PIP-OE1) infected with PVY^NTN showing clearly visible symptoms compared with a relatively asymptomatic PR control 69 d post-inoculation.

Fig. 6.

Quantification of PVY in systemically infected Premier Russet and PIP-OE1 plants. Leaf samples were harvested 44 d after inoculation with PVY^NTN. PVY titer was determined by RT–qPCR relative to the reference gene L2. Data are means ±SE (n=4 for Premier Russet, n=5 for PIP-OE1). The asterisk indicates a significant difference (P<0.05) as determined by Student t-test.

Infection with PVY^NTN induces major changes in expression of plant defense response genes in StPIP1-overexpressing lines

To test our hypothesis that foliar symptoms in StPIP1-overexpressing lines infected with PVY^NTN were due to changes in plant defense response, we analyzed changes in gene expression at the transcriptome level in systemic leaves infected with PVY^NTN in PR and PIP-OE1 at 44 dpi using QuantSeq (Moll et al., 2014). Four comparisons were made: mock-inoculated PR versus mock-inoculated PIP-OE1; PVY^NTN-infected PR versus mock-inoculated PR; PVY^NTN-infected PIP-OE1 versus mock-inoculated PIP-OE1; and PVY^NTN-infected PR versus PVY^NTN-infected PIP-OE1 (Fig. 7; Supplementary Tables S12–S15). When PVY^NTN-infected PR was compared with mock-inoculated PR, only six genes were differentially expressed (q≤0.05) (Fig. 7A; Supplementary Table S12), indicating that PVY^NTN infection had little effect on the overall transcriptome of PR in systemic leaves. When mock-inoculated PIP-OE1 was compared with mock-inoculated PR, 92 genes were differentially expressed in addition to StPIP1 itself (Fig. 7B; Supplementary Table S13). This indicates that the overexpression of StPIP1 had a mild effect on the overall transcriptome of PR under normal growth conditions. Strikingly, when comparing PVY^NTN-infected versus mock-treated-PIP-OE1 plants, 3500 genes were differentially expressed (Fig. 7C; Supplementary Table S14). Likewise, when comparing PVY^NTN-infected PIP-OE1 versus PVY^NTN-infected PR plants, 1921 genes were differentially expressed (Fig. 7D; Supplementary Table S15). Amongst DEGs with a q≤0.05 and a |log2(foldchange)| cut-off ≥2, about half (47%) were common between the two comparisons, while ~32% were specific to the comparison of PVY^NTN-infected and mock-inoculated PIP-OE1, and 20% were specific to the comparison of PVY^NTN-infected PIP-OE1 versus PVY^NTN-infected PR (Fig. 7E).

Fig. 7.

QuantSeq Volcano plots (A–D) and Venn diagram (E) of DEGs in PR and the StPIP1 overexpressor PIP-OE1 44 d after inoculation with PVY^NTN. Data points in red are significantly differentially expressed, with adjusted P-values (q) ≤0.1 (default for DEseq2); however, we use the threshold of q ≤0.05 to define genes that were significantly differentially expressed. (A) Comparison between PVY^NTN-treated and mock-treated PR. The plot represents each gene with a dot. (B) Comparison between mock-inoculated PIP-OE1 and PR. (C) Comparison between PVY^NTN-treated and mock-treated PIP-OE1. (D) Comparison between PVY^NTN-inoculated PIP-OE1 and PR. Arrows show the overexpressed StPIP1 in (B) and (D). Note the change of scale on the y-axis between plots. (E) Venn diagram showing the number of common and unique DEGs [|log₂(FC)|≥2; q≤0.05] between (C) and (D) comparisons.

We then used the statistical tool g:GOSt from g:Profiler to perform functional profiling of the DEGs that were significantly expressed (q≤0.05, |log₂(foldchange)|≥2) in PVY^NTN-inoculated PIP-OE1 compared with mock PIP-OE1 or PVY^NTN-inoculated PR (Supplementary Figs S14, S15). Several significantly enriched terms (P<0.05), common to both comparisons, were clearly associated with plant defense responses (e.g. ‘Plant pathogen interaction’, ‘Systemic acquired resistance’, ‘Chitinase activity’) and signaling (e.g. ‘Protein phosphorylation’, ‘Calcium ion binding’, ‘Regulation of DNA binding factors activity’). Genes associated with pathogen responses are listed in Tables 1 and 2. In addition to genes identified by gProfiler, many other genes were found to have annotations related to plant defense responses such as several WRKY transcription factors, Hsr203J (an HR marker gene), glucanases, and Hcr2-0A-annotated resistance genes (Supplementary Tables S14, S15). Interestingly, StPIP2 (PGSC0003DMT400038540), a paralog of StPIP1, was also induced in both comparisons.

Table 1.

Selected DEGs with immune-related GO-enriched terms in PVY^NTN-inoculated PIP-OE1 compared with PVY^NTN-inoculated PR

Transcript ID	Log2(FC)	SE	Adjusted P-value	Gene annotationa	GO terms
PGSC0003DMT400037745c	3.470	0.541	9.76×10–9	Conserved gene of unknown functionb	Regulation of systemic acquired resistance (GO:0010112)
PGSC0003DMT400037744	3.208	0.586	1.99×10–6	Conserved gene of unknown functionb
PGSC0003DMT400041025c	2.691	0.606	2.45×10–4	Conserved gene of unknown functionb
PGSC0003DMT400014779	2.492	0.606	8.59×10–4	NPR1/NIM1-interacting protein NIMIN2c
PGSC0003DMT400046345	3.210	0.348	7.95×10–18	Cyclic nucleotide-gated calmodulin-binding ion channel	Plant–pathogen interaction (KEGG:04626)
PGSC0003DMT400059272	2.964	0.372	2.77×10–13	Calcium-binding protein
PGSC0003DMT400027849	2.845	0.407	2.56×10–10	Ethylene responsive transcription factor ERF4
PGSC0003DMT400074364	2.774	0.490	7.17×10–7	SGT1 (suppressor of the G2 allele of skp1)
PGSC0003DMT400083031c	2.616	0.347	5.75×10–12	Enhanced disease susceptibility 1 protein (EDS1)
PGSC0003DMT400025744c	2.583	0.185	3.32×10–41	Endoplasmin homolog
PGSC0003DMT400021079	2.541	0.270	1.75×10–18	Calcium-binding allergen Ole e
PGSC0003DMT400041342	2.151	0.339	1.54×10–8	Calmodulin
PGSC0003DMT400024594	2.148	0.208	2.94×10–22	Heat shock protein 83
PGSC0003DMT400003364	2.531	0.388	5.01×10–9	Calcium ion-binding protein	Plant–pathogen interaction (KEGG:04626), MAPK signaling pathway (KEGG:04016)
PGSC0003DMT400053338c	2.478	0.462	3.42×10–6	Serine-threonine protein kinase, plant-type
PGSC0003DMT400070945	2.357	0.409	4.16×10–7	Serine-threonine protein kinase, plant-type
PGSC0003DMT400083727	2.347	0.583	1.20×10–3	Calmodulin
PGSC0003DMT400044026	2.272	0.632	5.14×10–3	Calcium-binding EF hand family protein
PGSC0003DMT400061478c	2.266	0.503	1.92×10–4	Cytoplasmic small heat shock protein class I
PGSC0003DMT400044379c	2.128	0.328	6.67×10–9	Transcription factor TSRF1	MAPK signaling pathway (KEGG:04016)
PGSC0003DMT400034487c	2.085	0.601	7.78×10–3	ERF transcription factor 5
PGSC0003DMT400003888	2.756	0.589	9.26×10–5	Chitinase 134	MAPK signaling pathway (KEGG:04016), Chitinase activity (GO:0004568)
PGSC0003DMT400003877	2.663	0.277	2.02×10–19	Class II chitinase	Chitinase activity (GO:0004568)
PGSC0003DMT400022352	2.652	0.325	6.31×10–14	Endochitinase (chitinase)
PGSC0003DMT400069033	2.011	0.280	7.05×10–11	Endochitinase 2

^a Annotations come from the Potato Genome Sequencing Consortium (PGSC; http://solanaceae.plantbiology.msu.edu/).

^b Although the PGSC does not list an annotation for these genes, BLAST analysis shows that they code for proteins with sequence similarity to NPR1/NIM1-interacting (NIMIN) proteins in Arabidopsis (Weigel et al., 2001) and tobacco (Zwicker et al., 2007).

^c Genes which are differentially expressed in this specific comparison.

Table 2.

Selected DEG with immune-related GO-enriched terms in PVY^NTN-inoculated PIP-OE1 compared with mock-inoculated PIP-OE1

Transcript ID	Log2(FC)	SE	Adjusted P-value	Gene annotationa	Go terms
PGSC0003DMT400037744	4.026	0.549	1.79×10–11	Conserved gene of unknown functionb	Systemic acquired resistance (GO:0009627), Defense response, incompatible interaction (GO:0009814)
PGSC0003DMT400014779	3.66	0.472	7.54×10-13	NPR1/NIM1-interacting protein NIMIN2c
PGSC0003DMT400059031c	3.610	0.471	1.67×10–12	Aspartate aminotransferase
PGSC0003DMT400051169c	2.390	0.637	0.002	Phytoalexin-deficient 4-2 protein
PGSC0003DMT400055847c	–2.088	0.654	0.010	Lipid-binding protein
PGSC0003DMT400063776c	2.257	0.505	1.27×10–4	Calmodulin-binding protein	Systemic acquired resistance (GO:0009627), Defense response, incompatible interaction (GO:0009814), Calmodulin binding (GO:0005516)
PGSC0003DMT400074364	3.329	0.516	5.87×10–9	SGT1 (suppressor of the G2 allele of skp1)	Plant–pathogen interaction (KEGG:04626)
PGSC0003DMT400013094c	3.164	0.629	1.05×10–5	PR1 protein
PGSC0003DMT400083027c	2.837	0.418	6.91×10–10	Enhanced disease susceptibility 1 protein (EDS1)
PGSC0003DMT400027849	2.624	0.419	1.65×10–8	Ethylene responsive transcription factor ERF4
PGSC0003DMT400059272	3.167	0.338	1.49×10–18	Calcium-binding protein	Plant–pathogen interaction (KEGG:04626), Calcium ion binding (GO:0005509)
PGSC0003DMT400003364	3.142	0.357	2.42×10–16	Calcium ion-binding protein
PGSC0003DMT400021079	2.731	0.303	3.15×10–17	Calcium-binding allergen Ole e
PGSC0003DMT400044026	2.403	0.655	0.002	Calcium-binding EF hand family protein
PGSC0003DMT400052233c	2.276	0.652	0.004	Polcalcin Jun o
PGSC0003DMT400041342	2.173	0.327	1.79×10–9	Calmodulin
PGSC0003DMT400083727	2.03	0.5777	0.004	Calmodulin
PGSC0003DMT400046345	3.117	0.509	3.92×10–8	Cyclic nucleotide-gated calmodulin-binding ion channel	Plant–pathogen interaction (KEGG:04626), Integral component of membrane (GO:0016021)
PGSC0003DMT400070945	2.250	0.430	4.24×10–6	Serine-threonine protein kinase, plant-type
PGSC0003DMT400024594	3.127	0.481	4.12×10–9	Heat shock protein 83	Plant–pathogen interaction (KEGG:04626), Protein processing in endoplasmic reticulum (KEGG:04141)
PGSC0003DMT400025743c	2.631	0.232	4.00×10–27	Endoplasmin homolog
PGSC0003DMT400037335c	2.177	0.192	4.00×10–27	Heat shock cognate protein 80
PGSC0003DMT400003877	2.986	0.399	5.87×10–12	Class II chitinase	Chitinase activity (GO:0004568)
PGSC0003DMT400022352	2.715	0.230	3.07×10–29	Endochitinase (chitinase)
PGSC0003DMT400069033	2.201	0.210	3.80×10–23	Endochitinase 2
PGSC0003DMT400003888	2.004	0.559	0.003	Chitinase 134

^a Annotations come from the Potato Genome Sequencing Consortium (PGSC; http://solanaceae.plantbiology.msu.edu/).

^b Although the PGSC does not list an annotation for this gene, BLAST analysis shows that it codes for a protein with sequence similarity to NPR1/NIM1-interacting (NIMIN) proteins in Arabidopsis (Weigel et al., 2001) and tobacco (Zwicker et al., 2007).

^c Genes which are differentially expressed in this specific comparison.

Together, these results show that PVY^NTN triggers the expression of defense-related genes in StPIP1-overexpressing lines, responses that were absent in PR control. Further, the defense-related DEGs probably account for the stark difference in symptom presentation between PR and StPIP1-overexpressing lines.

Callose deposition is higher in StPIP1-overexpressing plants

Callose deposition at the cell wall is a defense response that restricts pathogen penetration through the cell wall and movement through plasmodesmata (Iglesias and Meins, 2000; Ellinger et al., 2013). To further assess the status of plant defense response in StPIP1-overexpressing plants, we analyzed callose deposition in systemic leaves in virus-free and fully infected plants at 44 dpi with PVY^NTN. Using aniline blue staining to detect callose in leaf tissue, we observed a significantly increased amount of callose in virus-free PIP-OE1 compared with virus-free PR leaves (~4-fold) (Fig. 8). Leaves of PVY-infected PIP-OE1 plants also had higher levels of callose compared with leaves of PVY-infected PR, although this was not statistically significant (Fig. 8). These results revealed at the cellular level that StPIP1-overexpressing plants are in a primed state with enhanced defense response.

Fig. 8.

Callose spots in virus-free (mock) and PVY^NTN-infected systemic leaves of the StPIP1-overexpressing PIP-OE1 line and Premier Russet. Data are the average callose spots in the 1 mm² area from 14 pictures per genotype, and error bars represent the SE of the data. Shared letters indicate that there was no significant difference as determined by ANOVA (P=0.05).

Discussion

We report here evidence that the potato gene StPIP1 is involved in the antiviral defense response against PVY in potato, and provide evidence that a peptide encoded by this gene can elicit plant defenses. We provide several lines of evidence for its function: (i) phylogenetic and amino acid sequence analyses (Figs 1, 2) show that the StPIP1 clusters with PIP family members from Arabidopsis and that the encoded protein contains the signature motif GPS(P)xGxGH along with a putative transit peptide for targeting to the apoplast; (ii) the promoter of StPIP1 contains several putative cis-regulatory elements that are involved in biotic stress response (Supplementary Table S7); (iii) exogenous application of synthetic StPIP1-derived peptides triggered ROS accumulation (Fig. 3) and increased expression of plant immunity marker genes (Fig. 4); and (iv) plants that overexpress StPIP1 showed a phenotypic reaction (rugose mosaic) (Fig. 4), major changes in the expression of genes related to plant immunity response (Fig. 7; Tables 1, 2; Supplementary Figs S14, S15; Supplementary Tables S12–S15), higher levels of callose deposition (Fig. 8), and a lower viral titer when systemically infected with a compatible strain of PVY (Fig. 6; Supplementary Fig. S13).

Despite evidence of an increased plant defense response in StPIP1-overexpressing plants infected with PVY^NTN, overexpression of StPIP1 did not fully protect the plants from systemic infection, but rather decreased the virus titer. Our results are similar to those in Arabidopsis where overexpression of AtPIP1 increased resistance to P. syringae but did not stop infection (Hou et al., 2014; Najafi et al., 2020). This can be explained by the relatively slow and weak response mediated by StPIP1. Indeed, increases in ROS production after exogenous application of StPIP1 were small and delayed relative to those observed with StSystemin (Supplementary Fig S4), and few defense genes responded to StPIP1 treatment in cultivar PR (Fig. 4; Supplementary Fig S5). Although StPIP1-overexpressing plants had higher callose deposition even before encountering the virus (Fig. 8), which seems to indicate that part of the plant defense system was already primed, the virus was still capable of overcoming those defenses and spread systemically (Supplementary Tables S9, S10). In agreement with our observation, a recent report has also shown that callose deposition is not a guarantee of virus restriction (Lukan et al., 2018). Once the virus became systemic, StPIP1-overexpressing plants were able to increase the expression of a large number of plant defense genes, as shown by QuantSeq (Fig. 7; Tables 1, 2; Supplementary Figs S14, S15; Supplementary Tables S12–S15), and able to keep the virus titer at a lower level than in non-transgenic plants (Fig. 6; Supplementary Fig. S13). The strong foliar phenotypic reaction (i.e. rugose mosaic) in PVY^NTN-infected StPIP1-overexpressing plants (Fig. 5) may be due to an excess of energy devoted to plant defense responses to the detriment of the overall plant fitness. In other words, the overexpression of StPIP1 broke the tolerance of PR, a ‘Typhoid Mary’ cultivar that displays little to no symptoms while still developing a systemic infection, to PVY^NTN, and turned it into a sensitive cultivar whose defense response is too weak to stop virus multiplication and/or movement completely.

While StPIP1-overexpressing lines showed severe symptoms (i.e. rugose mosaic) that were quasi-absent in PR when infected with a compatible strain, PVY^NTN (Fig. 5), they showed no such difference when infected with PVY^O, an incompatible strain that sometimes is able to overcome HR and become systemic. A possible explanation is that the PVY^O-triggered ETI response overshadows the StPIP1-mediated PTI response. Testing the effect of overexpressing StPIP1 against compatible strains of PVY in a cultivar lacking resistance genes (e.g. RB) would help to validate this hypothesis. If confirmed, a crosstalk between resistance responses (i.e. ETI and PTI) may explain why StPIP1 is transiently down-regulated in the early stages of the incompatible PR–PVY^O interaction, but not in the compatible PR–PVY^NTN interaction (Goyer et al., 2015). In this context, StPIP1 may be a target of a strain-specific PVY effector that triggers ETI, such as HCPro (Chowdury et al., 2019), in a similar way to the capsid protein of Plum pox virus, an avirulence factor that triggers ETI but also suppresses PTI (Nicaise and Candresse, 2017). Another plausible explanation is that virus strain-specific features (e.g. RNA and/or protein sequence compositions) determine recognition by and activation of the StPIP1-mediated response. Future studies should focus on identifying these important features and the mechanism of recognition.

Receptors for SGP-rich peptides are typically leucine-rich repeat-containing receptor-like kinases (LRR-RLKs) (Yamaguchi et al., 2006; Stenvik et al., 2008; Yamaguchi et al., 2010). In Arabidopsis, AtRLK7 is the receptor for AtPIP1 and AtPIP2 (Hou et al., 2014). Hou et al (2014) initially identified RLK7 as a promising receptor candidate for AtPIP1 because it was one of the few class XI LRR-RLKs that were induced in response to pathogen infection or PAMP elicitation. To identify candidate receptors of StPIP1, we searched the QuantSeq data for LRR-RLKs that were induced in PVY^NTN-inoculated PIP-OE1 compared with either mock-inoculated PIP-OE1 or PVY^NTN-inoculated PR. The expression of the closest homolog of AtRLK7 in potato, PGSC0003DMG400004966 (transcript ID PGSC0003DMT400012744), was not induced. However, six genes encoding putative LRR-RLKs, based on the presence of an LRR domain, a single-pass transmembrane domain, and an intracellular Ser/Thr kinase domain (Shiu and Bleecker, 2001), were found to be significantly induced in PVY^NTN-inoculated PIP-OE1 (Supplementary Table S16). Two of them, PGSC0003DMG400011989 and PGSC0003DMG400027586, belong to class XI LRR-RLKs, ~50% similar (>95% coverage) to RLK7 (Pitorre et al., 2010) as well as the AtPEP1 receptors AtPEPR1/2 (Yamaguchi et al., 2006, 2010), and are expressed in leaf and petiole like StPIP1 (as well as the shoot apex and tuber in the case of PGSC0003DMG400027586) (https://www.ebi.ac.uk/gxa/home), making them attractive candidate receptors for StPIP1. Future studies are warranted to assess if these genes encode StPIP1 receptors. In addition, because PIP1-RLK7-induced responses are partially dependent on BAK1 in Arabidopsis (Hou et al., 2014), it would be interesting to investigate whether BAK1 orthologs in potato are involved in StPIP1 signaling.

Supplementary data

The following supplementary data are available at JXB online.

Table S1. Amino acid sequences of small secreted peptides from Arabidopsis, tomato, and potato.

Table S2. Primers used for cloning, genotyping, and detection of PVY by RT–PCR.

Table S3. Primers used for RT–qPCR for determination of StPIP1 gene expression and PVY levels.

Table S4. Primers used for RT–qPCR on defense genes from potato leaf discs.

Table S5. MIQE corresponding to Supplementary Table S3.

Table S6. MIQE corresponding to Supplementary Table S4.

Table S7. Cis-acting regulatory elements in the 1000 bp promoter region of StPIP1.

Table S8. Sequences of StPIP1 propeptides used in this study.

Table S9. Rate and onset of hypersensitive response, symptom presentation, and PVY infection in PR and StPIP1-overexpressing lines in Experiment 1.

Table S10. Rate and onset of hypersensitive response, symptom presentation, and PVY infection in PR and StPIP1-overexpressing lines in Experiment 2.

Table S11. Rate and onset of hypersensitive response, symptom presentation, and PVY infection in PR and StPIP1 artificial miRNA lines in Experiment 3.

Table S12. List of DEGs in systemic leaves of PVY^NTN-infected versus mock Premier Russet 44 d after inoculation as determined by QuantSeq.

Table S13. List of DEGs in systemic leaves of non-infected (mock) PIP-OE1 versus Premier Russet 44 days after treatment as determined by QuantSeq.

Table S14. List of DEGs in systemic leaves of PVY^NTN-infected versus mock-PIP-OE1-infected 44 d after inoculation as determined by QuantSeq.

Table S15. List of DEGs in systemic leaves of PVY^NTN-infected-PIP-OE1 versus PVY^NTN-infected Premier Russet 44 d after inoculation as determined by QuantSeq.

Table S16. Leucine-rich repeat domain-containing genes induced in PIP-OE1 infected with PVY^NTN compared with both mock-inoculated PIP-OE1 and PVY^NTN-inoculated Premier Russet.

Fig. S1. PHYRE structural model prediction of StPIP1.

Fig. S2. SignalP-5.0 prediction of signal peptide and cleavage for StPIP1.

Fig. S3. Cytosolic Ca²⁺ concentration in leaves of cultivar Désirée after application of variants of StPIP1 and StSystemin.

Fig. S4. ROS production in potato leaves and nodes in response to StPIP1 and its variants.

Fig. S5. Expression of defense-related genes in leaves in response to StPIP1 and its variants.

Fig. S6. StPIP1 gene expression increase in StPIP1-overexpressing lines as determined by RT–qPCR.

Fig. S7. Growth phenotype of in vitro grown StPIP1-overexpressing line PIP-OE1 compared with control Premier Russet plantlets.

Fig. S8. Characterization of artificial miRNA lines silencing StPIP expression.

Fig. S9. Hypersensitive response (HR) on the adaxial (upper row) or abaxial (lower row) sides of PVY-inoculated leaves of potato cultivar Premier Russet.

Fig. S10. PVY testing by RT–PCR of leaves from in-season infected and seedborne infected plants for Experiment 1.

Fig. S11. PVY testing by RT–PCR of leaves from in-season infected and seedborne infected plants for Experiment 2.

Fig. S12. PVY testing by RT–PCR of leaves from in-season infected plants (A–C) in the StPIP1 artificial miRNA experiment.

Fig. S13. Quantification of PVY relative to the reference gene EF1α in systemically infected Premier Russet and PIP-OE1 plants.

Fig. S14. Manhattan plot illustrating the enrichment analysis based on Gene Ontology from the comparison between PVY^NTN-infected versus mock-PIP-OE1.

Fig. S15. Manhattan plot illustrating the enrichment analysis based on Gene Ontology from the comparison between PVY^NTN-infected PIP-OE1 versus PVY^NTN-infected Premier Russet.

Acknowledgements

We thank Dr Alex Karasev for the gift of PVY strains. MMC was supported by grants from the U.S. Department of Agriculture National Institute of Food and Agriculture (2016-38420-25286) and the Western SARE (2018-38640-28418-WS1GS). CJR was supported in part by a grant from the U.S. Department of Agriculture ARS-State Partnership Potato Research program (2092-22000-021-35-S). LT was supported by a grant from the U.S. Department of Agriculture National Institute of Food and Agricultural (2019-67032-29072). This work was supported by the Northwest Potato Research Consortium to AG and the U.S. Department of Agriculture National Institute of Food and Agriculture (2019-67013-29963 and 1015621) to KT.

Author contributions

Conceptualization: MMC and AG; formal analysis: MMC, NM, KT, CJR, LT, and AG; funding acquisition: MMC, KT, JCA, and AG; investigation: MMC, NM, CJR, and AG; project administration: AG; supervision: KT, JCA, and AG; visualization: MMC, NM, CJR, and AG; writing—original draft preparation: MMC and AG; writing—review and editing: MMC, KT, JCA, AMR, and AG.

Data availability

Raw Illumina sequencing data are available at the NCBI Sequence Read Archive under the accession PRJNA669287.