Limits...
Benzalkonium tolerance genes and outcome in Listeria monocytogenes meningitis

View Article: PubMed Central - PubMed

ABSTRACT

Objectives: Listeria monocytogenes is a food-borne pathogen that can cause meningitis. The listerial genotype ST6 has been linked to increasing rates of unfavourable outcome over time. We investigated listerial genetic variation and the relation with clinical outcome in meningitis.

Methods: We sequenced 96 isolates from adults with listerial meningitis included in two prospective nationwide cohort studies by whole genome sequencing, and evaluated associations between bacterial genetic variation and clinical outcome. We validated these results by screening listerial genotypes of 445 cerebrospinal fluid and blood isolates from patients over a 30-year period from the Dutch national surveillance cohort.

Results: We identified a bacteriophage, phiLMST6 co-occurring with a novel plasmid, pLMST6, in ST6 isolates to be associated with unfavourable outcome in patients (p 2.83e-05). The plasmid carries a benzalkonium chloride tolerance gene, emrC, conferring decreased susceptibility to disinfectants used in the food-processing industry. Isolates harbouring emrC were growth inhibited at higher levels of benzalkonium chloride (median 60 mg/L versus 15 mg/L; p <0.001), and had higher MICs for amoxicillin and gentamicin compared with isolates without emrC (both p <0.001). Transformation of pLMST6 into naive strains led to benzalkonium chloride tolerance and higher MICs for gentamicin.

Conclusions: These results show that a novel plasmid, carrying the efflux transporter emrC, is associated with increased incidence of ST6 listerial meningitis in the Netherlands. Suggesting increased disease severity, our findings warrant consideration of disinfectants used in the food-processing industry that select for resistance mechanisms and may, inadvertently, lead to increased risk of poor disease outcome.

No MeSH data available.


Manhattan plot for single-nucleotide polymorphisms (SNPs) associated with unfavourable outcome. Manhattan plot for SNPs associated with unfavourable outcome calculated by a univariate linear mixed model analysis. Sequences from 85 isolates were mapped against a ST6 reference and SNPs present in between 5% and 95% of isolates were called (n = 166 839). The −log10 of the p-value is plotted on the y-axis and the location of SNPs in the reference genome on the x-axis. The red horizontal line is the threshold for correction of multiple testing (Bonferroni correction). SNPs in a region around 1.65 Mbp on the reference genome have the lowest p-values. None of the SNPs reach statistically significant levels of association after correction for multiple testing.
© Copyright Policy - CC BY
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5392494&req=5

fig2: Manhattan plot for single-nucleotide polymorphisms (SNPs) associated with unfavourable outcome. Manhattan plot for SNPs associated with unfavourable outcome calculated by a univariate linear mixed model analysis. Sequences from 85 isolates were mapped against a ST6 reference and SNPs present in between 5% and 95% of isolates were called (n = 166 839). The −log10 of the p-value is plotted on the y-axis and the location of SNPs in the reference genome on the x-axis. The red horizontal line is the threshold for correction of multiple testing (Bonferroni correction). SNPs in a region around 1.65 Mbp on the reference genome have the lowest p-values. None of the SNPs reach statistically significant levels of association after correction for multiple testing.

Mentions: There were 1952 gene groups and 166 839 SNPs present in between 5% and 95% of isolates. None of the gene groups or SNPs were significantly associated with unfavourable outcome or mortality after Bonferroni correction (Fig. 2; see Supplementary material, Figs. S3 and S4). However, because of statistical overcorrection for population stratification (see Supplementary material, Fig. S5) and extensive linkage disequilibrium in the listerial genome, Bonferroni correction is too strict [18]. The region around 1.65 million base pairs (Mbp) in the reference genome showed the strongest signal in the SNP association analysis with unfavourable outcome and was therefore explored further (Fig. 2). In this region in the reference genome we identified a prophage, which interrupts the master regulator for competence gene expression, comK, suggested to be a listerial virulence factor, which is not involved in DNA uptake [25]. The structural configuration of a prophage-interrupted comK gene was present in 22 of 85 de novo aligned clinical isolate sequences (26%) but was not associated with unfavourable outcome (Fisher’s exact test, p 0.62) or mortality (Fisher’s exact test, p 0.43) in our cohort. However, another prophage with high homology to the comK-interrupting phage (median sequence identity, 66%; range, 65%–70%) was not present in the reference isolate and identified in 12 clinical isolates and drove the observed association with unfavourable outcome. This novel phage (phiLMST6; accession number Hx2000053476) was present in ST6 isolates only and in the phylogenetic tree of ST6, 11 of 12 (92%) isolates carrying phiLMST6 were clustered and formed a clonal expansion (Fig. 1c). These isolates were associated with unfavourable outcome (occurring in 11 of 12 patients infected with isolates carrying phiLMST6 (92%) versus 33 of 73 patients infected with isolates without phiLMST6 (45%); Fisher’s exact test, p 0.004) but not mortality (6 of 12 (50%) versus 20 of 53 (27%); Fisher’s exact test, p 0.17). The phiLMST6 phage consisted of 66 genes, of which 47/66 (71%) encode hypothetical proteins, 14/66 (21%) encode phage structural proteins, 2/66 (2%) are recombinases and 3/66 (5%) encode annotated genes: a methylase, an amidase and a DNA-binding protein (see Supplementary material, Fig. S6).


Benzalkonium tolerance genes and outcome in Listeria monocytogenes meningitis
Manhattan plot for single-nucleotide polymorphisms (SNPs) associated with unfavourable outcome. Manhattan plot for SNPs associated with unfavourable outcome calculated by a univariate linear mixed model analysis. Sequences from 85 isolates were mapped against a ST6 reference and SNPs present in between 5% and 95% of isolates were called (n = 166 839). The −log10 of the p-value is plotted on the y-axis and the location of SNPs in the reference genome on the x-axis. The red horizontal line is the threshold for correction of multiple testing (Bonferroni correction). SNPs in a region around 1.65 Mbp on the reference genome have the lowest p-values. None of the SNPs reach statistically significant levels of association after correction for multiple testing.
© Copyright Policy - CC BY
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5392494&req=5

fig2: Manhattan plot for single-nucleotide polymorphisms (SNPs) associated with unfavourable outcome. Manhattan plot for SNPs associated with unfavourable outcome calculated by a univariate linear mixed model analysis. Sequences from 85 isolates were mapped against a ST6 reference and SNPs present in between 5% and 95% of isolates were called (n = 166 839). The −log10 of the p-value is plotted on the y-axis and the location of SNPs in the reference genome on the x-axis. The red horizontal line is the threshold for correction of multiple testing (Bonferroni correction). SNPs in a region around 1.65 Mbp on the reference genome have the lowest p-values. None of the SNPs reach statistically significant levels of association after correction for multiple testing.
Mentions: There were 1952 gene groups and 166 839 SNPs present in between 5% and 95% of isolates. None of the gene groups or SNPs were significantly associated with unfavourable outcome or mortality after Bonferroni correction (Fig. 2; see Supplementary material, Figs. S3 and S4). However, because of statistical overcorrection for population stratification (see Supplementary material, Fig. S5) and extensive linkage disequilibrium in the listerial genome, Bonferroni correction is too strict [18]. The region around 1.65 million base pairs (Mbp) in the reference genome showed the strongest signal in the SNP association analysis with unfavourable outcome and was therefore explored further (Fig. 2). In this region in the reference genome we identified a prophage, which interrupts the master regulator for competence gene expression, comK, suggested to be a listerial virulence factor, which is not involved in DNA uptake [25]. The structural configuration of a prophage-interrupted comK gene was present in 22 of 85 de novo aligned clinical isolate sequences (26%) but was not associated with unfavourable outcome (Fisher’s exact test, p 0.62) or mortality (Fisher’s exact test, p 0.43) in our cohort. However, another prophage with high homology to the comK-interrupting phage (median sequence identity, 66%; range, 65%–70%) was not present in the reference isolate and identified in 12 clinical isolates and drove the observed association with unfavourable outcome. This novel phage (phiLMST6; accession number Hx2000053476) was present in ST6 isolates only and in the phylogenetic tree of ST6, 11 of 12 (92%) isolates carrying phiLMST6 were clustered and formed a clonal expansion (Fig. 1c). These isolates were associated with unfavourable outcome (occurring in 11 of 12 patients infected with isolates carrying phiLMST6 (92%) versus 33 of 73 patients infected with isolates without phiLMST6 (45%); Fisher’s exact test, p 0.004) but not mortality (6 of 12 (50%) versus 20 of 53 (27%); Fisher’s exact test, p 0.17). The phiLMST6 phage consisted of 66 genes, of which 47/66 (71%) encode hypothetical proteins, 14/66 (21%) encode phage structural proteins, 2/66 (2%) are recombinases and 3/66 (5%) encode annotated genes: a methylase, an amidase and a DNA-binding protein (see Supplementary material, Fig. S6).

View Article: PubMed Central - PubMed

ABSTRACT

Objectives: Listeria monocytogenes is a food-borne pathogen that can cause meningitis. The listerial genotype ST6 has been linked to increasing rates of unfavourable outcome over time. We investigated listerial genetic variation and the relation with clinical outcome in meningitis.

Methods: We sequenced 96 isolates from adults with listerial meningitis included in two prospective nationwide cohort studies by whole genome sequencing, and evaluated associations between bacterial genetic variation and clinical outcome. We validated these results by screening listerial genotypes of 445 cerebrospinal fluid and blood isolates from patients over a 30-year period from the Dutch national surveillance cohort.

Results: We identified a bacteriophage, phiLMST6 co-occurring with a novel plasmid, pLMST6, in ST6 isolates to be associated with unfavourable outcome in patients (p 2.83e-05). The plasmid carries a benzalkonium chloride tolerance gene, emrC, conferring decreased susceptibility to disinfectants used in the food-processing industry. Isolates harbouring emrC were growth inhibited at higher levels of benzalkonium chloride (median 60 mg/L versus 15 mg/L; p <0.001), and had higher MICs for amoxicillin and gentamicin compared with isolates without emrC (both p <0.001). Transformation of pLMST6 into naive strains led to benzalkonium chloride tolerance and higher MICs for gentamicin.

Conclusions: These results show that a novel plasmid, carrying the efflux transporter emrC, is associated with increased incidence of ST6 listerial meningitis in the Netherlands. Suggesting increased disease severity, our findings warrant consideration of disinfectants used in the food-processing industry that select for resistance mechanisms and may, inadvertently, lead to increased risk of poor disease outcome.

No MeSH data available.