Plant material
Plant material consisted of the potato diploid population 15–1 (F1 progeny, N = 166) from a cross of Solanum hybrid DG 88–89 (seed parent) and a wild species S. chacoense (pollen parent). The maternal clone was generated within the diploid potato breeding program at the Plant Breeding and Acclimatization Institute—National Research Institute, Młochów, Poland. Paternal specie was obtained from National Centre for Plant Genetic Resources, Radzików, Poland (accession POL003:333133). DG 88–89 was the multigenerational hybrid originating from crosses of diploid potato clones. In terms of the genomes, the percentage of S. tuberosum in DG 88–89 was 78.3% and that of S. chacoense was 15.7%. Maternal clone and paternal species differ with TGA concentration in the PLE and PP. DG 88–89 exhibited a low TGA concentration (5.2 μg ml− 1) and nondetectable PP, while S. chacoense had a high TGA concentration (55.6 μg ml− 1) and high PP (70%) [15]. Three replicates of each of the progeny were grown in a greenhouse from May to October 2016. In full anthesis, leaflets were collected, mixed, portioned into 0.5 and 1 g portions, frozen in liquid nitrogen and stored until use at − 80 °C.
Preparation of potato leaf extract
PLE was prepared as previously described by [16]. Briefly, 0.5 g of frozen leaves was ground in liquid nitrogen, supplemented with 50 ml of distilled water (1% w/v) and shaken for 24 h on a laboratory shaker. Freshly prepared and filtered extract was used for biochemical and mass spectrometry (MS) analysis.
Evaluation of total glycoalkaloid, total phenolic and total flavonoid contents in potato leaf extract
The TGA concentration was measured for all 166 individuals in 2016 using the colorimetric method by [44] with modification as described by [16]. Briefly, 1% PLE was concentrated fourfold in a vacuum rotary evaporator (SpeedVac Appligene Refrigerated Aspirator, Germany). TGA was extracted using 10% acetic acid and precipitated with 5 M ammonium hydroxide. Samples were suspended in 100% methanol. The colour reaction was carried out using 98% sulfuric acid and 1% paraformaldehyde.
Absorbance was measured on a Hitachi U-1900 (Japan) spectrophotometer at a wavelength of 562 nm against a blank sample. The concentration of TGA was expressed in equivalents of α-solanine (Sigma-Aldrich, S3757).
TP content was determined as in [16] with a modification of the method described by [45]. Analysis was performed on twofold-concentrated PLE using freshly prepared tenfold-diluted Folin–Ciocalteu reagent (Sigma, F9252) and 7.5% sodium carbonate (w/v), and the absorbance was measured at 760 nm against a blank sample. The concentration of TP was expressed as equivalents of gallic acid (PhytoLab, 89,198).
TF was determined as in [16] with a modified method described by [46]. Analysis was performed on twofold-concentrated PLE using 10% aluminium chloride (w/v) and 1 M potassium acetate, and the absorbance was measured at 415 nm against a blank sample. The concentration of TF was expressed as equivalents of quercetin (PhytoLab, No. 89262).
TGA, TP and TF concentrations in selected genotypes were measured in 2016–2018. All measurements were performed in three biological repetitions, and each repetition had two technical replicates.
Evaluation of phytotoxic potential of potato leaf extract
PP was measured against the test plant - mustard cv. Rota (Vera-Agra Breeding Company, Cieszków, Poland) in three biological repetitions for all 166 individuals in 2016 and for the selected plants in 2017–2018 as previously described by [16]. Briefly, 15 mustard seeds after radicle protrusion (appx. 3 mm long) were transferred into Petri dishes (square, 12 cm) filled with filter paper and moisture with distilled water or 1% PLE (PLE-treated plants). After 5 days of incubation, the lengths of the control and PLE-treated mustard seedlings were measured. PP was expressed in % as the degree of seedling length inhibition/stimulation in relation to the length of control plants (grown in water) according to the formula
$$ Inhibition\ \left(\%\right)=\left(1-\frac{\mathrm{Treated}\ \mathrm{seedling}\ \mathrm{length}\ }{\mathrm{Control}\ \mathrm{seedling}\ \mathrm{length}}\right)x\ 100 $$
Construction of bulk samples
Based on TGA concentration and PP (Additional File 6), bulk samples C and D were constructed, each with three biological replications. Bulked sample analysis allows for more effective identification of genes underlying a trait. In this approach, contrasting individuals from a segregating population are pooled and then commonly screened to identify specific markers [47]. Sample C exhibited low TGA content (2.7 μg ml− 1 in PLE) and high PP (40%); D exhibited a low TGA concentration (2.6 μg ml− 1 in PLE) and nondetectable PP (PLE-treated plants were the same length as the control). In each sample, equal amounts of frozen leaves in liquid nitrogen from three F1 individuals were ground, mixed together and stored at − 80 °C.
RNA isolation and RNA-seq analysis
RNA was isolated from samples C and D according to the protocol described in [48] using TRIzol reagent. Briefly, 0.1 g of tissue ground in liquid nitrogen was supplemented with 1 ml of TRIzol reagent. Extraction was performed twice in chloroform. The RNA was precipitated in 0.3 ml of salt solution (0.8 M sodium citrate and 1.2 M sodium chloride) and 0.3 ml of isopropanol and resuspended in sterile water. The quality and quantity of RNA were determined using a NanoDrop spectrophotometer (Thermo Scientific) at 260 nm and 280 nm and on a 2% agarose gel. Next, RNA was treated with DNase I (Thermo Scientific, EN0521) to degrade double-stranded and single-stranded DNA contaminants in RNA samples.
The mRNA was isolated using the Dynabeads® mRNA Purification Kit for mRNA enrichment (Ambion, 61,006), and a library was prepared using the MGIEasy RNA Directional Library Prep Set (MGI, 1000006386), both according to the manufacturer’s protocols.
The established cDNA libraries were sequenced on the BGISEQ-500 sequencing platform (BGI Genomics, China) to generate 100-bp paired-end reads. RNA-seq reads were generated by Genomed® (Warsaw, Poland). After filtering of adaptor sequences and low-quality reads, data were obtained for subsequent analysis. Then, the index of the reference genome (https://www.ncbi.nlm.nih.gov/assembly/GCF_000226075.1) was built using Bowtie v2.1.0, and clean reads obtained for samples C and D were aligned to the reference genome using TopHat v2.0.9 (Broad Institute, Boston, MA). Next, HTSeq v0.5.3 was used to count the number of reads mapped to each gene. The DEGs were identified by the DESeq package.
Analysis of gene ontology term enrichment
To study the biological functions of the DEGs, gene set enrichment with GO terms was performed using the topGO package. To extract the significant GO categories, Fisher’s exact test was performed with the elim algorithm. To prepare circle diagram of all significant GO terms, we used as query for finding the ontology in various functional categories on the basis of GOslim categories as:
$$ \frac{\mathrm{annotations}\ \mathrm{to}\ \mathrm{terms}\ \mathrm{in}\ \mathrm{GOslim}\ \mathrm{category}\ }{\mathrm{total}\ \mathrm{annotations}\ \mathrm{to}\ \mathrm{terms}\ \mathrm{in}\ \mathrm{this}\ \mathrm{ontology}} \times 100. $$
Analysis of the glycoalkaloids profile in potato leaf extract
The GA fraction was isolated from 1% PLE of bulked samples C and D using a solid-phase extraction method (QuEChERS). First, for each PLE sample, α-solamarine (ChemFaces, CFN93102) dissolved in methanol was added as an internal standard to a final concentration of 10 ng μl− 1 to calculate the percentage of recovery of GA. In the control sample, α-solamarine was added to distilled water to the same final concentration. Both types of samples (control and PLE) were passed through sterilizing filters (0.2 μm, Nalgene™). To 750 μl of a sample, an equal amount of acetonitrile (ACN) with 1% formic acid was added, and the sample was applied onto the solid phase of QuEChERS (UTC, ECQUCHL12CT) and shaken for 30 s on a vortex mixer. Then, the supernatant obtained after GA isolation was diluted 10-fold with methanol. HPLC-MS analysis was performed on a Dionex 3000 RS-HPLC equipped with a DGP-3600 pump, a WPS-3000 TLS TRS autosampler, a TCC-3000 RS column compartment (Dionex Corporation, USA) and a Bruker micrOTOF-QII mass spectrometer (Bruker Daltonics, Germany). The chromatography column was a 50 × 3.1 (i.d)-millimetre Thermo Scientific Hyperil GOLD with 1,9-μm particles (Part No. 25002–052130, Serial No. 0110796A6, Lot No. 10922).
Chromatographic conditions: For the mobile phase, solvent A was water, and solvent B was ACN. The flow program was as follows: 0 min – 5% solvent B; 1.4 min – 5% solvent B; 22.9 min – 95% solvent B; 24.4 min – 95% solvent B; 24.5 min – 5% solvent B; 29 min – 5% solvent B. The injection sample volume was 1.5 μL. The flow rate was 0.2 ml min− 1, and the eluent was monitored by MS. The analysis was performed in negative ESI mode. Scan range: 50–1500 m/z, end plate offset: − 500 V, capillary: 4500 V, nebulizer gas (N2): 1,2 bar, dry gas (N2): 10 L/min, dry temperature: 220 °C.
Qualitative analysis of GA content was performed using the frequency of each compound in the TGA found in the sample. The qualitative GA profile was calculated using the following formula:
$$ F\left[\%\right]=\frac{A_n}{A_{A\iota \iota}}\times 100\%, $$
where
F – Frequency of compounds
An – Area of analysis compound
AAll – Area of all GAs in the sample.
Quantitative analysis was performed using the GA standards α-solanine (ChemFaces, CFN90560), α-chaconine (ChemFaces, CFN00450), α-solamargine (ChemFaces, CFN90159), and α-solasonine (PhytoLab 83,271). To quantify the compounds, calibration curves for each GA standard were generated over the concentration range of 0.1 μg ml− 1 to 10 μg ml− 1. For leptine II, a curve for α-solanine was used due to the lack of a standard. The results are expressed in μg ml− 1 and take into account the percent recovery of each compound in relation to the control sample concentration.
Protein extraction
Proteins were isolated from C and D as described in [49] with a minor modification. Then, 0.1 g of powdered tissue was suspended in 350 μl of the extraction buffer and incubated on ice for 30 min. Then, phenol solution (Roti®-Aqua-Phenol) was added in a 1:1 (v/v) ratio and incubated at room temperature for 10 min. The phenol phase was recovered twice by centrifugation at 4 °C, transferred to new tubes with extraction buffer 1:1 (v/v) and precipitated in cold methanol containing 0.1 M ammonium acetate 1:4 (v/v). The mixture was incubated overnight at − 20 °C and centrifuged at room temperature. The liquid phase was removed, and the pellet was washed once with 100% methanol pre-chilled to − 20 °C, centrifuged with 80% acetone, and centrifuged at the highest speed. The final protein pellet was air-dried and dissolved in 200 μl of 25 mM ammonium bicarbonate. The sample protein content was determined according to the method described by [50] using the bicinchoninic acid assay and bovine serum albumin as a standard. Five independent biological replicates were analysed in this study. A total of 120 μg of protein from each probe was sent to the Mass Spectrometry Laboratory at the Institute of Biochemistry and Biophysics, Polish Academy of Sciences (Warsaw, Poland), for nano-LC-MS-MS/MS (nanoliquid chromatography coupled to tandem mass spectrometry) analysis.
Comparative analysis of differentially expressed proteins
Peptide mixtures were analysed by nano-LC-MS-MS/MS using a nano-Acquity (Waters) LC system and a Q-Exactive mass spectrometer (Thermo Electron Corp., San Jose, CA) with the same equipment, buffers and parameters like in [51]. The raw data were processed by Mascot Distiller followed by Mascot Search (Matrix Science, London, UK, on-site licence) against the UniProt Solanum tuberosum database (February 2018 release). The search parameters for precursor and product ion mass tolerances were 30 ppm and 0.1 Da, respectively enzyme specificity: trypsin and missed cleavage sites. Peptides with Mascot scores exceeding the threshold value corresponding to < 5% of the expectation value as calculated by the Mascot procedure were considered positively identified. Quantitative analysis was performed as described by [52]. Shortly, the mass calibration and data filtering described above were carried out like in [51]. At the end, lists of identified peptides with corresponding abundances were exported for statistical analysis carried out with in-house developed Diffprot Software (version 1.5.19; 3.01.2013) [53]. Prior to analysis, abundances were normalized with LOWESS. Proteins with more than 90% common peptides were clustered, and only peptides unique for the cluster were used for statistical analysis. Only proteins with q-values ≤0.05 were considered differentially expressed.