Introduction

The potato (Solanum tuberosum L.) ranks as the world’s fourth most important food crop, after maize, wheat and rice. Compared to other staple crops, a greater proportion of the potato crop is edible and a higher yield per hectare is achieved. This combined with the high nutritional value of the potato underlines its importance and explains on-going increases in global potato production. The most significant threat to potato production worldwide is the late blight disease, caused by the oomycete pathogen Phytophthora infestans (Mont.) de Bary, which led to Irish famines between 1,845 and 1,852. Late blight is mainly controlled by application of enormous amounts of chemicals, which has raised issues on environmental, healthy and economic impacts. Therefore, there is an urgent need for resistant potato cultivars. Host resistance is a more environmentally benign means of restricting late blight infection but the success of this and other management practices hinge on effective and durable host resistance [1].

The most sustainable strategy to manage late blight is to breed broad-spectrum disease resistance into potato. However, traditional disease resistance breeding that exploits resistance genes (R genes) has only been short-life successful. P. infestans has a remarkable capacity to rapidly adapt to resistant host plants [1–3], and introduced R genes were quickly defeated. Indeed, this feature has led authors to describe P. infestans as a pathogen with a “high evolutionary potential” and an “R gene destroyer” [1, 3].

In addition to R gene-based resistance, another type of resistance occurs in partially resistant potatoes [4, 5]. In order to reveal molecular events that are associated with such quantitative resistance to late blight, many technologies have been used [6]. In our lab, a cDNA library highly enriched for quantitative-resistance-related genes was constructed by suppression subtractive hybridization (SSH) from the quantitatively-resistant potato 386209.10 in previous studies [7]. The same quantitatively-resistant potato 386209.10 was also subjected to cDNA-AFLP analysis to compare gene expression profiles in response to P. infestans and to dl-beta-amino-butyric acid (BABA) [8], which can protect potato plants against late blight [9–12]. In this situation, a large amount of gene fragments that may be associated with quantitative-resistance were gathered.

Virus-induced gene silencing (VIGS) provides a powerful tool to enable large-scale genes functional analysis. VIGS describes a technique employing recombinant viruses to specifically reduce endogenous gene activity and is based on post-transcriptional gene silencing (PTGS) [13]. Until now, several RNA and DNA viruses have been modified to develop VIGS vectors [14]. For example, Potato virus X (PVX)-based VIGS was successfully applied in Nicotiana benthamiana to identify a potato gene, plastidic carbonic anhydrase, which was involved in resistance to P. infestans [15]. Tobacco rattle virus (TRV) is able to spread more vigorously throughout the entire plant, including meristem tissue, yet the overall symptoms of infection are mild compared with other viruses [16, 17]. N. benthamiana is a good model plant for VIGS studies [18]. VIGS studies have also been described for potato [19, 20], but are less routinely used.

In this study, we selected candidate genes based on both cDNA microarrays and cDNA-AFLP analysis and cloned them from the quantitatively-resistant potato 386209.10 [8, 21]. These candidate genes were subjected to TRV-based VIGS in both N. benthamiana and potato, and following inoculation with P. infestans, we monitored the late blight disease symptoms. By these means, we aim to identify target genes that play a role in quantitative resistance to P. infestans.

Materials and methods

Plant materials and growth conditions

The potato (Solanum tuberosum L.) genotype 386209.10 with quantitative resistance to late blight, but not containing R1–R11 genes, was kindly provided by the International Potato Centre. This genotype 386209.10 was used for cloning the candidate gene fragments for VIGS. Another potato clone GT12297-4 and N. benthamiana were used for functional VIGS studies. In vitro potato plantlets were propagated in sterile culture boxes containing MS medium supplemented with 4 % sucrose and 0.8 % agar and raised in a climate room under controlled conditions (16 h light/8 h dark cycle at 20 °C). Four-week-old plantlets were transplanted to a greenhouse under proper conditions, i.e. with temperature between 20 and 26 °C and the humidity above 80 %.

Database searches

For the ESTs or TDFs, the Potato Genome Sequencing Consortium (PGSC) provides a promising new tool for identifying the corresponding genes [22]. Database searches were carried out using the Blast and Genome Browser Network and the results are shown in Table 1. The similarities of constructed sequences to N. benthamiana or N. tabacum (Table 1) were determined by Blastn searches in the Sol Genomics Network database (http://solgenomics.net/tools/blast/index.pl?db_id=194) and The Gene Index Project database (http://compbio.dfci.harvard.edu/cgi-bin/tgi/Blast/index.cgi), respectively.

Table 1 Characteristics of candidate genes
Full size table

Plasmid constructs and transformation of Agrobacterium tumefaciens for VIGS

The purified total RNA prepared from leaves of potato clone 386209.10 were used to synthesize single-stranded cDNA by the ReverTra Ace reverse transcriptase (TOYOBO, Japan) as template to amplify EST and TDF fragments. We used specific primers with restriction enzyme sites of EcoRI, BamHI, SacI or XbaI (Appendix S1). After purification by the DNA kit (Takara, Japan), the PCR fragments were cloned into accordingly digested pTRV2 (Km, provided by Dinesh-Kumar, Yale), the binary TRV RNA2 vector that facilitates VIGS in N. benthamiana [21]. As some ESTs contained internal cutting sites, the digested fragments of the corresponding genes were smaller than PCR products (Appendix S1). TRV: ESTs or TRV: TDFs binary constructs were then transformed into Agrobacterium sp. strain GV3101 by electroporation. The construct TRV: StPDS was constructed in the same way. Primers with KpnI and MluI sites (5′-AAAGGTACCTTGAATGAGGATGGGAGTGT-3′ and 5′-AAACGCGTAATGGCCGACAAGGTTCACA-3′) for amplifying the PDS fragment were derived from the potato PDS gene (AY484445). Both primers derived from the TRV2 vector (5′-GATTCTGTGAGTAAGGTTACC-3′ and 5′-TAATGTCTTCGGGACATGCC-3′) and specific genes were used to check the single clone of recombinant plasmids. Then the inserts of the positive clones were sequenced at BGI (Shenzhen, China).

VIGS using TRV: ESTs or TRV: TDFs recombinants and inoculation with P. infestans

For N. benthamiana, four-week-old plants were treated by coinfiltration of Agrobacterium tumefaciens strain GV3101 carrying pTRV1 and the various pTRV2 recombinants, in a 1:1 ratio [23]. For each of the recombinant clones, three plants were used. Three to four weeks after plants treated with TRV: StPDS showing initial photobleaching, the newly grown developed leaves compared to negative control plant were detached. They were incubated on the surface of wet filtrate paper in closed boxes and inoculated with 10 μl droplets of freshly produced zoospore-suspension of P. infestans isolates at 5 × 104 zoospores ml−1. The mixed isolates (race 1.3.4.7, 3.4.7.10.11 and 1.3.6.7.8.9.10.11) used for the inoculation were collected from west Hubei of China where late blight epidemics occur each growing season. The maintenance and inoculum preparation of P. infestans isolates followed the same method described by Vleeshouwers [24]. Inoculated leaves were incubated in a climate room at 18 °C, 90 % relative humidity and a 16 h light/8 h dark photoperiod provided by fluorescent lamps. For potato, two-week-old plants after transplanted in the greenhouse were inoculated in the same way as for N. benthamiana, except that agroinfiltration was performed in the stem instead of the leaf.

Estimation and analysis of lesion growth rate (LGR) and lesion size (LS)

Disease lesions were quantitatively assessed. For N. benthamiana, the disease lesion dimensions were measured at 3, 4 and 5 days post inoculation (dpi) using a caliper. The experiment was repeated three times and for each experiment eight leaflets were set up for each candidate gene. For potato, the disease lesion dimensions were measured just at 5 dpi, and three replicates with six leaflets for each candidate gene were used. The lesion size (LS) was calculated based on the following formula: LS = 1/4 × π × length × width. The lesion radius calculated from square root transformation of the area was converted into lesion growth rate (LGR, mm/day) [24]. LGR and LS were estimated and analyzed with REML and ANOVA respectively, using GenStat 13.2.

Trypan blue staining of infected leaves

The trypan blue staining was employed to visualize the colonization with sporulating mycelium of P. infestance. Trypan blue stock solution was prepared by mixing 10 g phenol, 10 ml glycerol, 10 ml lactic acid, 10 ml water and 0.02 g of trypan blue together. A working solution was prepared by diluting the stock solution with 75 % ethanol (1:2 v/v). Infected leaves were transferred into a plastic jar with diluted trypan blue solution. The jar (lid slightly unscrewed) was placed in a heated water bath and the staining solution was boiled for one minute. After that, the leaves were left overnight in the staining solution. The next day, leaves were distained by replacing the staining solution with 75 % ethanol and the distaining solution was replaced several times.

Results

Candidate genes were selected for functional screening

In total, 63 candidate genes highly expressed during the early biotrophic infection were selected for the functional screening, which include 35 ESTs from the cDNA microarrays, 27 TDFs from the cDNA-AFLP profile and an extra gene STMCD18 from the publication (Table 1). We compared the similarities of the constructed parts of the candidate genes with N. benthamiana and most candidate genes showed high identities (above 80 %) to the orthologues of N. benthamiana (Table 1).

VIGS is effective on N. benthamiana as well as potato

To test whether VIGS studies are suitable for our purpose, we tested N. benthamiana as well as potato with control treatments. The photobleaching phenotype was used by suppressing the expression of the endogenous phytoene desaturase gene (PDS). A fragment about 524 bp of the potato StPDS (91 % identity to N. benthamiana) was cloned into pTRV2 and transformed into A. tumefaciens strain GV3101. For N. benthamiana, we used four-week-old plants and agroinfiltrated StPDS control construct into the lower-side of leaves. Photobleaching was observed after approximately one week and the top leaves became totally white at 12 dpi (Fig. 1a). This indicates that VIGS with the heterogenous PDS construct worked well on N. benthamiana. For potato, StPDS control fragment was agroinfiltrated into two-week-old stems of potato clone GT12297-4. The top leaves were almost white at 25 dpi in a proper greenhouse condition, with temperature between 20 and 26 °C and the humidity above 80 % (Fig. 1b). During initial VIGS test, we found that VIGS developed slowly on potato, which normally took three weeks or longer to show photobleaching. Besides, the VIGS treatment on potato often resulted in less uniform and weaker silencing of the gene throughout an infected plant compared to N. benthamiana. After we consistently observed photobleaching symptoms with StPDS on potato in repeated experiments, we conclude, therefore, that potato could be used for VIGS to screen our candidate genes.

Fig. 1
figure 1

Photobleaching phenomenon on Nicotiana benthamiana and potato after VIGS treatment of StPDS. a VIGS effect of StPDS on N. benthamiana at 12 days post inoculation; b. VIGS effect of StPDS on potato at 25 days post inoculation

Full size image

Three candidate genes cause morphological changes in N. benthamiana

The phenotypic effects of the candidate genes on plant morphology were also investigated by VIGS. We cloned all 63 candidate gene fragments into pTRV2 and treated four-week-old N. benthamiana together with pTRV1 in a 1:1 ratio by agro-coinfiltration. Then we monitored the plants for altered morphology over time. Over all, VIGS-treated N. benthamiana plants grew slower compared to non-treated plants, but generally, most constructs did not cause morphological changes compared to the healthy plants. In contrast, three different constructs, i.e. TRV: 7, TRV: 35 and TRV: 50 caused morphologically different phenotypes (Fig. 2). The candidate genes 7, 35 and 50 revealed similarity to a nucleolar protein nop56, an ethylene-responsive proteinase inhibitor 1 and a ribulose bisphosphate carboxylase small chain 2C, respectively (Table 1) by Blastn searches against S. tuberosum cv phurjea genome.

Fig. 2
figure 2

Morphological changes of Nicotiana benthamiana after VIGS treatment at 16 days post inoculation. a A healthy plant. b A negative control plant treated with TRV: 00. c After treatment with TRV: 7, the stem had stopped growing and leaves became thick and small. d After treatment with TRV: 35, plants had stopped differentiating the terminal buds. e After treatment with TRV: 50, plants grew slowly and leaves became thin, small and chlorosis

Full size image

Two candidate genes cause altered resistance to P. infestans in both potato and N. benthamiana

To test whether the 63 candidate genes have an effect on resistance to P. infestans, we subjected them to VIGS on four-week-old N. benthamiana. Plants treated with TRV: StPDS were used as negative control. Three to four weeks after the negative control plants showed initial photobleaching, newly grown developed leaves were detached and inoculated with P. infestans zoospore suspension. We rated the resistance by measuring disease lesions on 3 dpi, 4 dpi and 5 dpi and estimated LGR. Statistical analysis showed that resistance to P. infestans was significantly reduced after VIGS treatment with TRV vectors containing candidate gene fragments 22, 44, 48, 52, 55 and 63 (Table 2). According to Blastn searches against S. tuberosum cv phurjea genome database, these genes are hinted as a lipoxygenase, a double WRKY type transcription factor, a suberization-associated anionic peroxidase, an ATP binding protein, a TATA binding protein associated factor and another ATP binding protein, respectively (Table 1).

Table 2 Decreasing resistance to Phytophthora infestans in potato and Nicotiana benthamiana after VIGS treatment of some genes
Full size table

After scoring lesions, we treated infected leaves with trypan blue. The infected areas of plants that were treated with TRV: 22, TRV: 44, TRV: 48, TRV: 52, TRV: 55 and TRV: 63 were much larger than those of healthy and negative control plants. Representative pictures of P. infestans infections on N. benthamiana are shown in Fig. 3 (TRV: 22, TRV: 63 not shown). From the microscopic observation, most of the blue-stained tissue is colonized with sporulating mycelium (Fig. 4). These data confirm that VIGS treatment with TRV: 22, TRV: 44, TRV: 48, TRV: 52, TRV: 55, and TRV: 63 resulted in enhanced susceptibility to P. infestans.

Fig. 3
figure 3

Susceptibility to Phytophthora infestans after VIGS treatment of some genes in Nicotiana benthamiana. a–f Comparisons of the lesions infected by P. infestans at 6 days post inoculation on N. benthamiana before and after trypan blue staining

Full size image
Fig. 4
figure 4

Microscopic observation of an infected leaf after trypan blue staining. sp sporangium, hy hypha

Full size image

Functional screening of candidate genes was also performed on potato. In this case, LS was investigated at 5 dpi. Statistical analysis showed that VIGS treatments with several candidate genes enhanced susceptibility, i.e. TRV: 12, TRV: 17, TRV: 21, TRV: 22, TRV: 24 and TRV: 48 (Table 2). From our results, we concluded that two candidate genes 22 and 48, identified as lipoxygenase and suberization-associated anionic peroxidase were found to have a significant effect on disease resistance in N. benthamiana and potato (Table 1). The other genes 12, 17, 21, 24 are hinted as a Tonoplast dicarboxylate transporter, a UDP-arabinose 4-epimerase (EC 5.1.3.5), a 2-oxoglutarate dehydrogenase and a 2-deoxyglucose-6-phosphate phosphatase, respectively (Table 1).

Discussion

To identify genes involved in quantitative resistance to potato late blight, 63 candidate genes were selected based on their high expression levels during the early infection stages with P. infestans and the genes were subjected to VIGS treatment and detached leaf assay in both N. benthamiana and potato. Candidate genes were first subjected to VIGS treatment and after photobleaching caused by TRV: StPDS occurred, the leaves were detached and inoculated with P. infestans. We found 10 candidate genes caused enhanced susceptibility after VIGS treatment in N. benthamiana and potato (Tables 1, 2). This suggests that enhanced expression of these genes may cause enhanced resistance to late blight.

A lipoxygenase (LOX) and a suberization-associated anionic peroxidase, represented by candidates 22 and 48, respectively, were identified to affect the late blight resistance in both potato and N. benthamiana background. In addition, we found that candidate genes 44, 52, 55 and 63 decreased the resistance to late blight in N. benthamiana and candidate genes 12, 17, 21 and 24 in potato. These candidates share similarities with various enzymes, e.g. UDP-arabinose 4-epimerase, 2-oxoglutarate dehydrogenase and 2-deoxyglucose-6-phosphate phosphatase, two different ATP binding proteins, a Sodium-dicarboxylate cotransporter, a double type WRKY transcription factor, and a binding protein associated factor (Table 1). We hypothesize that overexpression of those genes in potato may contribute to quantitative resistance to late blight.

LOXes are dioxygenases that catalyze the hydroperoxidation of polyunsaturated fatty acids or their esters that contain a cis, cis-1, 4-pentadiene moiety. Candidate gene 22 was identified as a LOX. In plants, LOXes have been implicated in the biosynthesis of stress-responsive signaling molecules, such as traumatin [25] and jasmonic acid [26]. Another LOX gene from potato, POTLX-3 (GenBank/EMBL accession no. U60202), was also reported to be involved in defense to P. infestans [27]. This is agreement with our finding that decreased expression of the putative LOX candidate gene 22 decreased resistance to P. infestans in both potato and N. benthamiana.

Suberization-associated anionic peroxidases were reported to play an important role of specific cell wall suberization in plant defense [28–30]. Cell wall suberization can result in strengthening the cell wall, and thereby inhibit pathogen invasion. We identified the candidate gene 48 as a putative Suberization-associated anionic peroxidase. This candidate gene was also found to decrease resistance to late blight after VIGS treament in potato and N. benthamiana, and we consider it as a good candidate to contribute to quantitative resistance.

With respect to cell wall strengthening, a UDP-arabinose 4-epimerase (EC 5.1.3.5), alternately named UDP-d-xylose 4-epimerase, was reported to play a rate-limiting role in the control of cell wall biosynthesis [31]. In our study, a putative UDP-arabinose 4-epimerase was identified, which shared high similarity with candidate gene 17. VIGS treatment of this gene resulted in decreased resistance to late blight in potato, which supports the hypothesis that modifying the cell wall may play a prominent role in quantitative resistance to late blight [32].

WRKYs are a large family of plant-specific transcription factors that bind to the W-box of promoter regions of many pathogenesis-related (PR) genes [33, 34]. PR genes are thought to be associated with regulating defense responses to both abiotic and biotic stresses [35, 36]. Phosphorylation of the NbWRKY8 (GenBank/EMBL accession no. AB445392.1), a double WRKY transcription factor from N. benthamiana, was recently reported to play a role in the defense response [37]. Candidate gene 44 showed high identity to NbWRKY8, and VIGS treatment of it resulted in a rapid development of the disease.

Candidate gene 63 showed similarity to an ATP binding protein, which is co-localizing with a QTL for late blight resistance on chromosome XI of potato population PCC1 (developed at International Potato Center) [38]. Most plant disease resistance (R) genes identified today encode proteins with a central nucleotide binding site (NBS) and a C-terminal Leu-rich repeat (LRR) domain. The NBS contains three ATP/GTP binding motifs known as the kinase-1a or P-loop, kinase-2, and kinase-3a motifs. Perhaps this candidate gene is involved in both a quantitative resistance response and R-gene related defense responses [39, 40].

We also identified the other candidate genes 12, 21, 24 and 55, representing a Sodium-dicarboxylate cotransporter, a 2-oxoglutarate dehydrogenase, a putative 2-deoxyglucose-6-phosphate phosphatase and a putative TATA binding protein, respectively, in our study. To our knowledge, there are no reports on functions in plant disease resistance for these genes and they might represent novel genes associated with quantitative resistance to late blight.

The fact that silencing a same gene in potato and N. benthamiana showed different levels of susceptibility to P. infestans might for example be explained by the fact TRV-based silencing in potato is weaker and less uniform, as shown from the photobleaching control experiments.

In conclusion, the combination of VIGS and detached leaf assays on both N. benthamiana and potato has proven to be a fast and powerful tool for identifying genes that may play a role in quantitative resistance to P. infestans in potato. Further studies of these identified target genes functions will help us clarify the resistance mechanisms involved in the quantitative resistance to P. infestans.