Plants of Cucurbita pepo subsp. pepo cultivar-group Zucchini, variety True French, were grown in greenhouse facility at Department of Agricultural Science of University of Naples “Federico II” using standard horticultural practices. C. pepo cv. True French tissue samples were collected from young plants of about 10 cm high. Total RNA was isolated from ground, frozen leaf tissues using the SpectrumTM Plant Total RNA Kit (Sigma-Aldirch). A complete removal of traces of DNA was performed using On-Column DNase I Digest Set (Sigma-Aldirch). Quantity and integrity of the extracted total RNA were determined using NanoDrop ND-1000 Spectrophotometer (Thermo Fisher Scientific Inc., USA), on a denaturing formaldehyde gel and Agilent 2100 bioanalyzer (Agilent Technologies, USA) respectively, to be RIN > 8. Library preparation and sequencing were performed by the Genomix4Life S.r.l., spin-off of Salerno University. The sequencing library was prepared using the TruSeq RNA Sample Preparation Kit v2 (Illumina, San Diego, CA, USA) and paired-end reads of 100 bp were sequenced from the three independent samples on one lane of an illumine HiSeq 2000.

The quality control checks on raw sequence data (75,22 millions of paired reads totalling 15 e¹² bp) from all the three data sets was performed using FastQC (Andrews, 2010). Raw reads were filtered to remove the adapter sequences and the poorer quality regions with sequence pre-processing tool, Trimmomatic (Bolger et al., 2014). Paired-end read duplicates from the PCR amplification step in the sequencing process were removed and only those reads with a mapping score ≥ 30 were kept in the alignments. The high quality reads were aligned against the C. pepo reference genome sequence version 3.2 (https://cucurbigene.upv.es/) with STAR aligner (version 2.4.0j). The resulting alignment was used as input to Cufflinks (version 2.2.1) for transcript assembly. PASA pipeline (version 2.0.2) was used to combine Cufflinks results with the public transcriptome version 3.0 (https://cucurbigene.upv.es/).

The proteome functional annotation was performed through a match search against four database (TAIR10, SWISS-PROT, TrEMBL and GenBank-NR) using DIAMOND in sensitive mode with a cut-off e-value of 1 e⁻⁵ (Buchfink et al., 2015). To add information about protein function to our proteome, a Blast2GO (Conesa et al., 2005) annotation, using default parameters, were conducted. Finally, the zucchini proteome was scanned with InterProScan v.5.13 (Jones et al., 2014) against the InterPro protein signature databases to identify and finely characterize plant resistance proteins.

To identify orthologous gene groups among C. pepo, C. melo, C. sativus, C. lanatus and A. thaliana we used OrthoMCL software with default settings. The association between reference R-genes (http://prgdb.crg.eu/) and relative orthologous group (OG) was detected using Best BLAST Hit method (BlastP, E < 1 e⁻⁵) and the output was filtered for a query coverage and identity percentage, both >50%.

The sequencing produced a total of 69,5 millions of clean paired reads, obtaining 13,9 e¹² bp of RNA-Seq data for C. pepo (Supplementary Table 1). Transcriptome assembly yielded 68,720 transcripts, with mean length of 1,534 bp. The transcripts were translated and a high-quality proteome of 33,966 primary protein sequences, with mean length of 316 AA, were obtained. DIAMOND similarity-based searches were performed against the publically available databases (SWISS-PROT, TrEMBL, TAIR10, and GenBank-NR) to annotate C. pepo proteome (Supplementary Table 2). About 85% of proteins encoded by genes had homology with four principal databases and over 75% were functionally annotated (Table 1). In addition, a GO-annotation using Blast2GO were effected and a total of 256,138 GO-terms were assigned to about 65% (21480) of the predicted proteins (Supplementary Table 3).

A fine characterization of genes encoding domains similar to plant resistance (R) proteins, in C. pepo proteome was conducted. R-proteins can be categorized according to the presence and organization of protein domains, such as Toll/Interleukin-1 receptor (TIR), coiled coil (CC), the nucleotide-binding site (NBS), leucine-rich repeats (LRRs). A total of 64 R-proteins (also called NLR, NB-LRR, NBS-LRR, or NB-ARC-LRR proteins) were identified (Supplementary Table 4). The CNL (Coiled coil, Nucleotide-binding site, Leucine-rich repeats) class was divided into sub-classes based on sequence similarity with the canonical CNLs that contain an EDVID amino-acid motif, and the RPW8-like proteins (Andolfo et al., 2014b). Interestingly, an expansion of RPW8-NLR genes (11 out of 64) in C. pepo was discovered. Diversely, C. melo, C. sativus, and C. lanatus presented only three RPW8-NLRs for each species (Figure 1A). It is now well-known that RPW8-NLRs can function as helper NLRs for well-defined NLR–mediated resistance responses. Thus, they may enhance the C. pepo defense system to offset its reduced number of NRL receptors available (Sanseverino and Ercolano, 2012). In addition, C. pepo RPW8-NLRs showed a very high homology to ADR1 (activated disease resistance 1), R-gene that confer resistance again Erysiphe cichoracearumi, the causal agent of Powdery Mildew (PM) in A. thaliana (Micali et al., 2008). PM disease, caused by Podosphaera xanthii (syn. Sphaerotheca fuliginea) has an important economic impact on C. pepo varieties. ADR1-like proteins expansion, identified in C. pepo, could suggest an adaptive diversification induced by specie-specific pathogen pressure (Andolfo and Ercolano, 2015).

A comparative analysis among C. pepo, C. melo, C. sativus, C. lanatus, and A. thaliana were performed to obtain functional information on our proteome. A total of 18,742 orthologous groups (OGs), which included 107,386 sequences, were identified (Supplementary Table 5). 9,453 OGs enclosed 69,982 sequences and were highly conserved in all analyzed genomes (Figure 1B). A core Cucurbitaceae proteins (9,465 sequences) clustered in 2099 OGs were detected. About 65% (22,214) of C. pepo predicted proteins were grouped in 13,953 OGs (Figure 1B). Several C. pepo gene family expansions associated to transmembrane R-genes were discovered. One hundred zucchini proteins, annotated as Receptor-like Kinase (RLK) and Receptor-like Protein (RLP) were clustered in five OGs (OG_00004, OG_00027, OG_00038, OG_00053, OG_01889, and OG_00579) and associated a well-characterized R-genes (Supplementary Table 5). Probably the expansion of cell surface receptors (RLKs and RLPs) and relative strengthening of first defence line represent adaptive dynamics to balance the limited number of cytoplasmic receptors (NRLs) (Andolfo and Ercolano, 2015). The C. pepo gene family expansions could be associated to the cucurbit-common tetraploidization recently identified by Wang et al. (2017). Furthermore, we identified a number of gene families related to important agronomical traits. Fourteen OGs related to OVATE gene family grouped 19 zucchini genes. OVATE is an important locus for fruit shape determination and plant development (Rodríguez et al., 2011). We identified three zucchini ortholog genes to PSY1 and PSY3 melon genes that putatively involved in carotenoid metabolism and fruit ripening (Garcia-Mas et al., 2012). The Cup000085g037789 is the ortholog gene of OR, a cloned gene that governs the fruit flesh colur in melon and in other important crops (Tzuri et al., 2015). Comparative analysis of C. pepo proteome can be used to identify orthologous genes for functional study. Our dataset represented a very important resource to reduce the plant breeding work for the identification of candidates for important agronomical traits (Supplementary Table 5).

The dataset submitted to NCBI include the raw read sequences of three biological replicates of Cucurbita pepo subsp. pepo cultivar-group Zucchini, variety True French, in FASTQ format. The raw reads of C. pepo can be accessed at NCBI with the following BioSample accession number: SAMN07426850 (www.ncbi.nlm.nih.gov/Traces/study/?acc=SRP114337). The C. pepo transcriptome annotation, in GTF format, and primary protein sequences in FASTA format can be accessed at FIGSHARE with the following link (https://figshare.com/s/8a083f60df238acdbc19). The Supplementary Material (Supplementary Tables 1–5) for this article can be found online at: (https://figshare.com/s/8a083f60df238acdbc19). Users can download and use the data freely for research purpose only with acknowledgment to us and quoting this paper as reference to the data.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.