Genome assembly of L. monocytogenes strain S2542
We sequenced and assembled the genome of strain S2542 and compared it with the genomes of the previously studied strains; RO15 and ScottA. Assembly and annotation resulted in a 2.9 Mbp genome consisting of 14 contigs (N50=477482 bp) with 2839 predicted coding sequence (CDS). GC-content of strain S2542 was 37.9%, which is identical to the GC-content of ScottA. Based on Blast Average Nucleotide Identity (ANIb), the genome of strain S2542 was more similar to ScottA than to RO15 (ANIb scores S2542/ScottA: 99.97, S2542/RO15: 94.55). Genome alignment of strains RO15, ScottA, and S2542 indicated high similarities between the genomes. Ortholog gene prediction between three strains revealed that 31 genes were present in S2542 but absent in strains ScottA and RO15 (Table S1). Most of the genes specific to strain S2542 were annotated to encode hypothetical proteins (Table S1). In total, 49 genes were found in both ScottA and RO15 but not in S2542, and 47 of these were prophage genes (Table S1). Thus, the differences in genome sequence most likely do not explain the differences in gene expression response to HPP described in our previous study on RO15 and ScottA and study presented for strain S2542 [8,9].
Based on PHASTER prophage prediction, two small regions in S2542 (Contig1:183471-206364: 22.8 kb; Contig2: 99648-114057: 14.4 kb) were annotated as a prophage with a low confidence score. Homolog of these regions were seen in both strain RO15 and ScottA. We did not observe CRISPR-Cas or anti-CRISPR genes within the strain S2542 genome.
Based on genome sequence analysis using BIGSdb , multilocus sequence typing (based on seven loci) assigned strain S2542 to ST-145. In addition, based on genome sequence, the clonal complex (CC) of strain S2542 was CC2 and it belongs to the Lineage I, as ScottA. Serotype prediction based on the genome suggested that strain S2542 belongs to PCR-serogroup 4b. By contrast S2542 was reported to belong to serogroup 1/2a based on an antigen test . Pan-genome analysis showed that strain S2542 harboured similar genes as serotype 4b strains (Figure 1a). Core genome alignment tree visualization also supported assignment of S2542 to serotype 4b, since two clear clusters were observed for serotype 4b strains (F2365, ScottA, S2542) and serotype 1/2a strains (2HF33, RO4, RO15, MB5, C7, EGD-e, AB199, AB120) (Figure 1a.)
Viable cell count after HPP
To assess cell recovery capacity after HPP, all three strains were treated at 400 MPa and 8°C for 8 min and colony forming units (CFU) were enumerated immediately after treatment and after storage for 24 and 48 h at 8 °C (Figure 1b). For all three strains CFU/ml were higher for the samples plated immediately after HPP (t = 0 min) compared to the samples stored for 24 and 48 h. Also, the reduction in CFU/ml at t = 0 min after HPP compared to untreated controls was significantly higher for S2542 than for ScottA and RO15. Moreover, after 24h and 48 of storage no viable cells could be detected for S2542, i.e. CFU/ml for all replicates of the high pressure treated samples were below the limit of detection, whereas for ScottA and RO15 CFU were detected at least for some of the replicates at these timepoints.
HPP-induced changes in expression of representative genes
In order to investigate the contradictory results of HPP-induced gene expression between the Bowman et al.  study and our previous results , a number of genes that showed strong negative correlation between the two studies were selected and their expression levels were analysed by ddPCR approach. In contrast, we did not observe any difference in the HPP-induced expression changes of the selected genes between these strains. Furthermore, gene expression changes at 400 MPa after 24 h for strains ScottA and S2542 were significantly correlated (Pearson Correlation; r = 0.97) (Table 1). This indicates that the gene expression responses of strains S2542 and ScottA are similar in these experimental conditions. No significant differences were observed between RO15 and S2542 (Table 1). Thus, expression profiles of the selected genes in response to HPP appear to be not significantly different between the strains. The contradictory results between our previous RNA-seq study  and the microarray and qPCR data reported by Bowman et al.  are difficult to explain, but could be potentially explained by the different experimental conditions (8°C-8 min HPP exposure in Duru et al  and 15°C-5 min HPP exposure in Bowman et al ) or difference of the growth of the bacteria during the experiments.