Limits...
Comparative genomics of European avian pathogenic E. Coli (APEC)

View Article: PubMed Central - PubMed

ABSTRACT

Background: Avian pathogenic Escherichia coli (APEC) causes colibacillosis, which results in significant economic losses to the poultry industry worldwide. However, the diversity between isolates remains poorly understood. Here, a total of 272 APEC isolates collected from the United Kingdom (UK), Italy and Germany were characterised using multiplex polymerase chain reactions (PCRs) targeting 22 equally weighted factors covering virulence genes, R-type and phylogroup. Following these analysis, 95 of the selected strains were further analysed using Whole Genome Sequencing (WGS).

Results: The most prevalent phylogroups were B2 (47%) and A1 (22%), although there were national differences with Germany presenting group B2 (35.3%), Italy presenting group A1 (53.3%) and UK presenting group B2 (56.1%) as the most prevalent. R-type R1 was the most frequent type (55%) among APEC, but multiple R-types were also frequent (26.8%). Following compilation of all the PCR data which covered a total of 15 virulence genes, it was possible to build a similarity tree using each PCR result unweighted to produce 9 distinct groups. The average number of virulence genes was 6–8 per isolate, but no positive association was found between phylogroup and number or type of virulence genes. A total of 95 isolates representing each of these 9 groupings were genome sequenced and analysed for in silico serotype, Multilocus Sequence Typing (MLST), and antimicrobial resistance (AMR). The UK isolates showed the greatest variability in terms of serotype and MLST compared with German and Italian isolates, whereas the lowest prevalence of AMR was found for German isolates. Similarity trees were compiled using sequencing data and notably single nucleotide polymorphism data generated ten distinct geno-groups. The frequency of geno-groups across Europe comprised 26.3% belonging to Group 8 representing serogroups O2, O4, O18 and MLST types ST95, ST140, ST141, ST428, ST1618 and others, 18.9% belonging to Group 1 (serogroups O78 and MLST types ST23, ST2230), 15.8% belonging to Group 10 (serogroups O8, O45, O91, O125ab and variable MLST types), 14.7% belonging to Group 7 (serogroups O4, O24, O35, O53, O161 and MLST type ST117) and 13.7% belonging to Group 9 (serogroups O1, O16, O181 and others and MLST types ST10, ST48 and others). The other groups (2, 3, 4, 5 and 6) each contained relatively few strains.

Results: However, for some of the genogroups (e.g. groups 6 and 7) partial overlap with SNPs grouping and PCR grouping (matching PCR groups 8 (13 isolates on 22) and 1 (14 isolates on 16) were observable). However, it was not possible to obtain a clear correlation between genogroups and unweighted PCR groupings. This may be due to the genome plasticity of E. coli that enables strains to carry the same virulence factors even if the overall genotype is substantially different.

Conclusions: The conclusion to be drawn from the lack of correlations is that firstly, APEC are very diverse and secondly, it is not possible to rely on any one or more basic molecular or phenotypic tests to define APEC with clarity, reaffirming the need for whole genome analysis approaches which we describe here.

Conclusions: This study highlights the presence of previously unreported serotypes and MLSTs for APEC in Europe. Moreover, it is a first step on a cautious reconsideration of the merits of classical identification criteria such as R typing, phylogrouping and serotyping.

Electronic supplementary material: The online version of this article (doi:10.1186/s12864-016-3289-7) contains supplementary material, which is available to authorized users.

No MeSH data available.


The mean and the standard deviation of the total number of virulence factors detected in each phylogroup (A, A1, B1, B2, and D)
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC5120500&req=5

Fig8: The mean and the standard deviation of the total number of virulence factors detected in each phylogroup (A, A1, B1, B2, and D)

Mentions: The phylologroups A, B1, B2, D were determined in 2000 by Clermont [9] using the dichotomous approach that was enhanced by the addition of new subgroups described in 2010 by Carlos et al. [41]. Those were the groups A1, B3 (only found in humans) and D2 [41]. Using this existing classification [9, 41] for the 272 strains analysed we found (in decreasing order) that 132 grouped in B2 phylogroup, 61 in A1, 37 in group A and 21 in groups B1 and D. The remaining 21 were not ascribable to any group. Interestingly, previous studies by Walk et al. [42], demonstrated that the majority of E. coli strains that are able to persist in the environment belong to the B1 phylogenetic group. As relatively few of the strains examined here belonged to this ‘environmental’ group we can probably conclude these strains were less likely to be opportunistic pathogenic E. coli associated with, but not necessarily causing avian colibacillosis. No B3 (human only) strains were found, confirming host differentiation, a finding consistent with the incorrect view that APEC were associated with urinary tract infections in man that arose through dependence on analysis of carriage of some shared virulence determinants by UTI strains [4, 18, 22, 43, 44]. Johnson et al. [45] found that strains from phylogroups B2 and D harboured more virulence factors than strains from the phylogroups A and B1 [41], but the studies reported here differ as the average value of the virulence factors ranging between 6 to 8 factors was common to all the phylogroups found (Fig. 7). The average and the standard deviation of the number of virulence factors detected for each phylogroup is illustrated in Fig. 8. In the studies conducted in our laboratories we have noted that the carriage of virulence determinants (up to 5 maximum), by presumed commensal strains (not belonging to recognized APEC serotypes,unpublished findings) is notable in European avian E. coli isolates.. The Nolan laboratory [18] previously suggested that the detection of a minimum of 5 virulence factors could be used to define APEC, but the data produced here suggests this number is perhaps too low. Therefore, here we shall discuss other factors that must be considered before a definition of APEC can be authoritatively assigned to an isolate.Fig. 7


Comparative genomics of European avian pathogenic E. Coli (APEC)
The mean and the standard deviation of the total number of virulence factors detected in each phylogroup (A, A1, B1, B2, and D)
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC5120500&req=5

Fig8: The mean and the standard deviation of the total number of virulence factors detected in each phylogroup (A, A1, B1, B2, and D)
Mentions: The phylologroups A, B1, B2, D were determined in 2000 by Clermont [9] using the dichotomous approach that was enhanced by the addition of new subgroups described in 2010 by Carlos et al. [41]. Those were the groups A1, B3 (only found in humans) and D2 [41]. Using this existing classification [9, 41] for the 272 strains analysed we found (in decreasing order) that 132 grouped in B2 phylogroup, 61 in A1, 37 in group A and 21 in groups B1 and D. The remaining 21 were not ascribable to any group. Interestingly, previous studies by Walk et al. [42], demonstrated that the majority of E. coli strains that are able to persist in the environment belong to the B1 phylogenetic group. As relatively few of the strains examined here belonged to this ‘environmental’ group we can probably conclude these strains were less likely to be opportunistic pathogenic E. coli associated with, but not necessarily causing avian colibacillosis. No B3 (human only) strains were found, confirming host differentiation, a finding consistent with the incorrect view that APEC were associated with urinary tract infections in man that arose through dependence on analysis of carriage of some shared virulence determinants by UTI strains [4, 18, 22, 43, 44]. Johnson et al. [45] found that strains from phylogroups B2 and D harboured more virulence factors than strains from the phylogroups A and B1 [41], but the studies reported here differ as the average value of the virulence factors ranging between 6 to 8 factors was common to all the phylogroups found (Fig. 7). The average and the standard deviation of the number of virulence factors detected for each phylogroup is illustrated in Fig. 8. In the studies conducted in our laboratories we have noted that the carriage of virulence determinants (up to 5 maximum), by presumed commensal strains (not belonging to recognized APEC serotypes,unpublished findings) is notable in European avian E. coli isolates.. The Nolan laboratory [18] previously suggested that the detection of a minimum of 5 virulence factors could be used to define APEC, but the data produced here suggests this number is perhaps too low. Therefore, here we shall discuss other factors that must be considered before a definition of APEC can be authoritatively assigned to an isolate.Fig. 7

View Article: PubMed Central - PubMed

ABSTRACT

Background: Avian pathogenic Escherichia coli (APEC) causes colibacillosis, which results in significant economic losses to the poultry industry worldwide. However, the diversity between isolates remains poorly understood. Here, a total of 272 APEC isolates collected from the United Kingdom (UK), Italy and Germany were characterised using multiplex polymerase chain reactions (PCRs) targeting 22 equally weighted factors covering virulence genes, R-type and phylogroup. Following these analysis, 95 of the selected strains were further analysed using Whole Genome Sequencing (WGS).

Results: The most prevalent phylogroups were B2 (47%) and A1 (22%), although there were national differences with Germany presenting group B2 (35.3%), Italy presenting group A1 (53.3%) and UK presenting group B2 (56.1%) as the most prevalent. R-type R1 was the most frequent type (55%) among APEC, but multiple R-types were also frequent (26.8%). Following compilation of all the PCR data which covered a total of 15 virulence genes, it was possible to build a similarity tree using each PCR result unweighted to produce 9 distinct groups. The average number of virulence genes was 6–8 per isolate, but no positive association was found between phylogroup and number or type of virulence genes. A total of 95 isolates representing each of these 9 groupings were genome sequenced and analysed for in silico serotype, Multilocus Sequence Typing (MLST), and antimicrobial resistance (AMR). The UK isolates showed the greatest variability in terms of serotype and MLST compared with German and Italian isolates, whereas the lowest prevalence of AMR was found for German isolates. Similarity trees were compiled using sequencing data and notably single nucleotide polymorphism data generated ten distinct geno-groups. The frequency of geno-groups across Europe comprised 26.3% belonging to Group 8 representing serogroups O2, O4, O18 and MLST types ST95, ST140, ST141, ST428, ST1618 and others, 18.9% belonging to Group 1 (serogroups O78 and MLST types ST23, ST2230), 15.8% belonging to Group 10 (serogroups O8, O45, O91, O125ab and variable MLST types), 14.7% belonging to Group 7 (serogroups O4, O24, O35, O53, O161 and MLST type ST117) and 13.7% belonging to Group 9 (serogroups O1, O16, O181 and others and MLST types ST10, ST48 and others). The other groups (2, 3, 4, 5 and 6) each contained relatively few strains.

Results: However, for some of the genogroups (e.g. groups 6 and 7) partial overlap with SNPs grouping and PCR grouping (matching PCR groups 8 (13 isolates on 22) and 1 (14 isolates on 16) were observable). However, it was not possible to obtain a clear correlation between genogroups and unweighted PCR groupings. This may be due to the genome plasticity of E. coli that enables strains to carry the same virulence factors even if the overall genotype is substantially different.

Conclusions: The conclusion to be drawn from the lack of correlations is that firstly, APEC are very diverse and secondly, it is not possible to rely on any one or more basic molecular or phenotypic tests to define APEC with clarity, reaffirming the need for whole genome analysis approaches which we describe here.

Conclusions: This study highlights the presence of previously unreported serotypes and MLSTs for APEC in Europe. Moreover, it is a first step on a cautious reconsideration of the merits of classical identification criteria such as R typing, phylogrouping and serotyping.

Electronic supplementary material: The online version of this article (doi:10.1186/s12864-016-3289-7) contains supplementary material, which is available to authorized users.

No MeSH data available.