Raphanus sativus

Soil sampling occurred in July 2019 in a commercial field situated in Gabbioneta-Binanuova (45°12’03.0” N 10°12’27.8” E), Cremona, Po Valley (Northern Italy). At the selected field site conservation agriculture practices have been adopted since 2011. Specifically: (i) A mixture of cover crops (CCs), composed by rye (Secale cereale L.), hairy vetch (Vicia villosa Roth), and radish (Raphanus sativus L.), was sown each year after harvesting the previous main crop and terminated right before the next seedbed preparation; (ii) reduced-tillage (RT) operations, which consisted of a ripper passage (25 cm depth) and one/two spring harrowing (10–15 cm depth), were annually performed before planting the main crop. The crop sequence was a 3-year crop rotation including maize (Zea mays L.), soybean (Glycine max (L.) Merr.), and processing tomato (Solanum lycopersicum L.).
Soil samples adhered to the roots of tomato plants were manually separated from the surrounding bulk soil and collected. The plants were carefully uprooted from soil, collected in sterile polybags and stored at 4 °C for the isolation of rhizobacteria. The rhizosphere and rhizoplane soil were separated from the bulk soil following the method proposed by Barillot et al. [40 (link)]. Briefly, bulk soil was removed shaking plants by hand for 10 min vigorously, paying attention to the roots’ integrity, as long as the roots’ non-adhering soil particles were completely removed. In order to collect rhizosphere and rhizoplane soil, the root system was washed with 500 mL of 0.9% NaCl added with Tween 80 (0.01% v/v) and afterwards 150 mL of bacterial suspension were incubated at 25 °C for 90 min with shaking at 180 rpm.

To improve gene prediction, we further obtained transcriptomes by sequencing high-quality RNA from mixed fresh leaf, flower, and stem tissues and sequenced them by the Illumina HiSeq X Ten platform. We removed adapters and discarded reads with >10% N bases or reads having more than 20% bases of low quality (below 5) using NGS QC Toolkit v.2.3.3^{72 (link)} and finally generated 19.87 Gb of clean data. We assembled the de novo and genome-guided transcriptomes with clean reads by Trinity v.2.4.0⁷³ . We also mapped the RNA-sequencing (RNA-seq) reads to the assembled genome to obtain the mapping rate through HISAT2 v.2.1.0^{74 (link)} to evaluate the completeness of the genome.
We run PASA pipeline v.2.1.0^{75 (link)} to align the transcripts to the assembled genome to carry out ORF prediction and gene prediction. To train the HMM model for Augustus, we extracted complete, multiexon genes, removed redundant high-identity genes (cut-off all-to-all identity of 70%), and finally generated the best candidate and low-identity gene models for training. We aligned the RNA-seq data to the hard-masked genome assembly by HISAT2^{74 (link)} and used bam2hints in Augustus to generate the intron hint file. We used this hint file to carry out ab initio gene prediction by Augustus v.3.2.2^{76 (link)}. For homologous prediction, the reference protein sequences of Brassica rapa, Brassica napus, Raphanus sativus, Brassica juncea, and Brassica nigra were downloaded and aligned against the I. indigotica genome using TBLASTN v.2.2.31^{77 (link)} and searched with an e value of 1e⁻⁵. After filtering low-quality results, gene structure was predicted using GeneWise v.2.4.1^{78 (link)}. We combined the results from PASA, Augustus and GeneWise to generate the final protein-coding gene set using EVidenceModeler v.1.1.1^{75 (link)}. To obtain the untranslated regions and alternatively spliced isoforms, we used PASA to update the gff3 file for two rounds and obtain the final gene models.
We annotated the functions of the predicated genes against public databases by NCBI BLAST+ v.2.2.31^{77 (link)} with a cut-off e value of 1e⁻⁵ and maximum number of target sequences of 20, including the Swiss-Prot and TrEMBL databases^{79 (link)}. Best-hit BLAST results were then used to define gene functions. We used InterProScan v.5.25-64.0^{80 (link)} to identify motifs and domains by matching against public databases. We identified GO annotations by using Blast2GO v.4.1⁸¹ according to the blast results and combined them with InterPro GO entries. We mapped the existing GO terms to enzyme codes by Blast2GO and submitted the predicted proteins to the KEGG (Kyoto Encyclopedia of Genes and Genomes) Automatic Annotation Server (KAAS)^{82 (link)} to obtain KO numbers for KEGG pathway annotation.

Twenty-six samples (approximately 500 g each) of edible portions of leafy vegetables including lettuce (Lactuca sativa), mint (Mentha piperita), jarjir (Arugula, Eruca sativa), radish (Raphanus sativus) and parsley (Petroselinum crispum) were collected from five locations (Jazan, Abuarish, Sabya, Damad and Darb) across the Jazan province of Saudi Arabia. The vegetable samples were collected in the months of April and May 2022 and kept in clean polyethylene bags until further use. From each sampling location, 3–6 cultivation plots were chosen, and approximately 100–150 g of each vegetable was collected from four different sites of every plot, and composite samples were made by mixing thoroughly. The number of composite samples collected from each location were as follows: Jazan, n = 4; Sabya, n = 5; Abuarish, n = 6; Damad, n = 5; and Darb, n = 6. The collected samples were directly brought to the laboratory, and the adhered soil and airborne dust were cleaned by washing first with running tap water and then with deionized water. The collected vegetables were chopped into small pieces, dried in an oven at 60 °C until a constant weight, crushed to pass through a 40-mesh sieve and stored in a refrigerator for further processing.

The phytotoxic activity was studied on radicle elongation and germination of the seeds of two crops Raphanus sativus L. (radish), Sinapis arvensis L. (wild mustard), and one weed Lolium multiflorum Lam. (Italian ryegrass). R. sativus seeds were purchased from Blumen group s.r.l. (Bologna, Italy); L. multiflorum seeds were bought from Fratelli Ingegnoli s.p.a. (Milano, Italy); S. arvensis seeds were picked from wild populations in Sidi Ismail, Beja, Tunisia on June 2021. These seeds, usually employed in phytotoxic tests, are easily germinable and are well known from a histological point of view. The seeds were sterilized in 95% ethanol for 15 s and sown in Petri dishes (Ø = 90 mm), on three layers of Whatman filter paper, impregnated with distilled water (7 mL, control) or the essential oil solution (7 mL), at different doses. The germination conditions were 20 ± 1 °C, with natural photoperiod. The EOs, solubilized in water–acetone mixture (99.5:0.5 v/v) to increase their solubility, were tested at the doses of 1000, 500, 250, 100 μg/mL. No differences in activity were registered between controls performed with water–acetone mixture and controls with water alone. Seed germination was observed in Petri dishes every 24 h. A seed was considered germinated when the protrusion of the root became evident [61 (link)]. After 120 h for R. sativus and S. arvensis seeds and 168 h for L. multiflorum seeds, the radicle lengths were determined and are expressed in cm. Each determination was conducted three times, using Petri dishes with 10 seeds inside.