For more details, please see my thesis 'Ozone-Mediated Control of Food Spoilage and Food-Borne Pathogens'. Aim: to characterise the bacterial community of the tomato calyx and pedicel and to investigate how this changed over time post-harvest, after ozone treatment had been applied or not. Materials and methods: Tomato cv. Brioso RZ F1 (72 - 130) (Rijk Zwaan) seeds (British Tomato Growers Association, UK) were planted on 4th March 2020 in a mixture of Levington Advance M2 compost (BHGS Ltd) and Sinclair Pro coarse vermiculite medium grade 2.0 - 5.0 mm (BHGS Ltd). They were germinated in a propagator within a glasshouse (22 - 24oC, 16 hour photoperiod) at the Department of Biology, University of Oxford. After germination, plants remained in the glasshouse and were re-potted, irrigated and treated with Levington Tomorite liquid plant feed (B&Q, UK) when necessary. Tomatoes were collected and packed between 19th June and 5th October 2020. Throughout this time period, a total of 25 tomatoes were collected from each plant, in the form of 5 punnets, which each contained 5 tomatoes collected at one time. For tomato collection, sterile gloves and scissors were used to cut tomatoes from vines, so that each tomato had approximately 1 cm of vine above the natural breakpoint. This was completed to mimic conditions of tomatoes being kept on the vine, since Tomato cv. Brioso RZ F1 (72 - 130) (Rijk Zwaan) is sold on the vine (Rijk Zwaan, 2021) and there were not enough tomatoes per vine to use complete vines in experiments. Only tomatoes which were red-ripe, firm, and looked healthy were collected. Tomatoes from the same plant were collected into a punnet (DUNI, UK), which was wrapped in cling-film for transportation to the laboratory. Five tomatoes from the same plant were then weighed and collectively transferred into a new punnet (DUNI, UK), which was sealed with 5-lane custom microperforated film (Coveris, UK) at 100oC before being ozone-treated or not. For ozone treatment, gaseous ozone was generated by the in-pack ozone system (Anacail, UK, model: F250, serial no: 6)), for 3 seconds at a command voltage of 3.2 V (~350 – 700 peak ppm ozone). The ozone which was generated would have taken up to a few hours to completely decay so bacteria would have remained exposed to ozone during this time. After ozone treatment or not, punnets which contained tomatoes were stored for 0, 3 or 10 days before calyxes and pedicels in each punnet were collected, pooled and stored at -80oC before further sample processing. Samples were homogenised in liquid nitrogen. DNA was extracted from ~110 mg of each homogenised sample using the Zymo quick-DNA fecal/soil microbe miniprep kit (Zymo Research, USA, product code: D6010) according to the manufacturer’s instructions (version 2.4.5, revised 04/2021) with deviations: bead beating conditions were 2 x 2.5 minutes at maximum speed with a TissueLyser (Qiagen, UK) (protocol step 2), bashing bead lysis tubes were centrifuged at 14,000 x g for 1 minute (protocol step 3), as much supernatant as possible was transferred to the Zymo-Spin™ III-F Filter tube (protocol step 4), Zymo-Spin™ IICR Column was centrifuged at 11,000 x g for 1 minute (protocol steps 8 and 9), 50 µl DNA elution buffer was used for elution’s at 10,000 x g for 1 minute (protocol step 10). A DNA extraction negative control was collected by extracting DNA from buffers only (sample NC.1). DNA concentrations were measured on a Qubit 4 Fluorometer (ThermoFisher Scientific, UK) with an Invitrogen Qubit dsDNA HS and BR assay kit (ThermoFisher Scientific, UK, product code: Q32850), according to the manufacturer’s instructions (revision A.0, 02/2015). DNA concentrations were adjusted to 4 ng/µl in sterile water, however since the DNA concentration of the DNA extraction negative control was below the detection limit (<0.2 ng/µl), the DNA concentration was not adjusted for this sample. DNA was stored at -20oC. For 16S rRNA library preparation, the Illumina ‘16S Metagenomic Sequencing Library Preparation (15044223B)’ protocol (revised 11/2013) was generally followed. Variable region 4 of the 16S rRNA gene was PCR-amplified from 12 ng of each DNA extraction with Invitrogen Platinum SuperFI DNA Polymerase (ThermoFisher Scientific, UK, product code: 12351010), according to the manufacturer’s instructions (revision B, 2017), using primers 515F (Parada) and 806R (Apprill). Mitochondria rRNA blocker (mPNA) (PNA BIO INC, USA, product code: MP01-50) and chloroplast rRNA blocker (pPNA) (PNA BIO INC, USA, product code: PP01-50) were dissolved by a 55oC incubation at 1400 rpm for 10 minutes, and then added to PCR reactions at a final concentration of 5 µM to prevent amplification of plant DNA. A PCR negative control was collected by PCR-amplifying sterile water only (NC.2). PCR products were cleaned up according to the Illumina protocol, however the beads used were Macherey-Nagel NucleoMag NGS Clean-up and Size Select (FisherScientific, USA, product code: 15889167). All PCR products were checked on a 1% agarose gel. Amplicons within a single sample then required to be tagged with the same index. Indexes were added according to the Illumina protocol with KAPA HiFi HotStart ReadyMix (Roche Sequencing Store, USA), however the indexes used were from Nextera XT Index Kit v2 Set C (96 indexes, 384 samples) (Illumina, UK, product code: FC-131-2003). PCR products were cleaned up and checked as previously described. The DNA concentration of each sample was measured as previously described and adjusted to 4 nM in 10 mM Tris pH 8.5. The DNA concentration of the DNA extraction and PCR negative controls were below the detection limit of 0.2 ng/µl, (ThermoFisher Scientific, UK) therefore the DNA concentrations were not adjusted for these samples. 5 µl of each sample was pooled and subjected to high-throughput sequencing with an Illumina Miseq sequencing platform at the James Hutton Institute, UK. All code that was used for the data analysis is available at https://github.com/AshleighMacdonald/tomato_calyx_and_pedicel_bacterial_community_analysis. Adapters were removed from reads using Cutadapt version 4.3 (Martin, 2011) and then reads were filtered, denoised and merged using DADA2. Taxonomy was assigned to the amplicon sequence variants (ASVs) using the naïve Bayesian classifier method, which was trained on the Silva nr99 dataset version 138.1 (Quast et al., 2013). An ‘ASV_sample_information’ file has been provided which presents the ASVs detected in each sample. Samples are named in a format like ‘B1.UT0’ where the B number is biological replicate, T or UT is ozone-treated or untreated and the subsequent 0, 3 or 10 refers to post-harvest/post-treatment storage days (The only exception is NC.1 or NC.2 which are the negative controls). Metadata for samples has been provided in the metadata file. Note that the quant variable was the DNA concentration of samples after initial DNA extractions (ng/µl).