The importance of the “V”-Factor: Escherichia coli bacteremia and sepsis

In the previous issue of Virulence, Mora-Rillo et al.¹ publish a report investigating the role of virulence factors (VFs) on sepsis severity and survival outcomes in patients with extra-intestinal pathogenic E. coli infections (ExPEC). Mortality caused by E. coli bacteremia and sepsis remains high and is now one the top 10 causes of death in the western world.^2,3 Representing a significant health burden, E.coli is the gram-negative bacteria most often associated with sepsis and infection by uropathogenic E. coli account for 70–95% of community acquired and 50% of hospital acquired urinary tract infections.^2,3 The high degree of intra-strain genetic diversity represents a challenge in correct clinical management of patients. Coupled with this the level of patient risk factor diversity confounds this issue, highlighting the need for studies such as this to increase treatment options and survival outcomes for affected individuals.

These investigations have benefitted from the increasing access to and accuracy of high throughput sequencing as this technology has evolved. The decreasing costs of these techniques only serves to further enhance how they are utilized. Mora-Rillo et al make use of some of these including, single nucleotide polymorphism characterization using SNP pyro-sequencing allowing for creation of SNP allelic profiles (SAPs) to group and classify bacterial strains. Combined with MLST (Multi-locus sequence typing) this allows a robust delineation of phylogroups and bacterial strains.¹ In this study Mora-Rillo et al delineated strains from each common phylogroup (B2, A, D, B1) and 50 SAPs of which, 37 were unique and only 6 appeared in more than 5 isolates, but accounted for 50.8% of all the isolates.¹ The majority of human E. coli strains exist as commensal bacteria of the intestinal tract ⁴ of particular interest in the context of sepsis causing infections, however, are outgrowth or infection of populations of 2 additional categories, intestinal pathogenic, entero-pathogenic (of which there are 6 classifications, EPEC (enteropathogenic), EHEC (enterohaemorrhagic), ETEC (enterotoxigenic), EAEC (enteroaggregative), EIEC (enteroinvasive) and DAEC (diffusely adherent)) and ExPEC, which are responsible for urinary tract infections (UTI) and neonatal meningitis among other conditions.⁴ Inter-species transmission of E. coli strains is common and numerous studies have shown the high degree of similarity between human, avian and porcine virulence mechanisms and bacterial populations.^5,6

What drives bacterial virulence? Numerous studies throughout the late nineties and into this decade have sought to tease out solutions to this question so that clinicians may overcome the health burden such bacteremia exemplifies. While some studies have demonstrated that host risk factors are more detrimental to outcomes in sepsis severity, and Mora-Rillo's work presented here posits that virulence factors (VF) have a significant impact on survival outcomes ¹, it is likely to be more complex than either one of these hypotheses alone. Larger multi center studies are needed in order to address these complex interactions. Skewing toward a host driven response versus that owing to bacterial population expression of certain VFs can result from too small population size, sampling from one center where the “environmental microflora” may predicate presence or absence of particular strains and thus skew significant expression of certain virulence factors.

Different strains of E. coli utilize varying VFs depending on their phylogroup and strain. In EPEC infections for example, the locus of enterocyte effacement (LEE) pathogenicity island (PAI) encodes multiple genes including the open reading frame (Orf) for the Intimin protein. This protein has been shown to be a potent inducer of Th1 responses.^4,7 Most of the other 41 orfs at this locus encode for proteins involved in the type III secretion system, including both chaperones and effectors, one such as Tir (translocation intimin receptor) which has been shown to activate N-WASP and Arp2/3 complexes to induced actin polarization and microbe internalization.^4,7

A study by Firoozeh et al.⁸ examined the VFs associated with urinary tract infections (UTI) by UPEC and how their expression was correlated with development of pyelonephritis and cystitis. The surface VFs (adhesions) of UPEC have been demonstrated as some of most widely expressed genes associated with bacteria causing UTI. P-fimbriae are encoded by the Pap genes, S-fimbrial adhesions are encoded by Sfa genes and A-fimbrial adhesions are encoded by the Afa genes. In addition multiple secreted VF have been characterized including, α-hemolysin (HlyA) encoded by the hly gene and Cytotoxic necrotizing factor 1 (CNF), serum resistance ability of UTI bacteria is encoded by the traT gene and bacterial siderophores, such as aerobactin may also play a role in UPEC virulence. Using these genes this study demonstrated that presence of pap genes are more significant in pyelonephritis than in cystitis and cnf, sfa and afa genes were not found in any isolate. This is in contrast to Mora-Rillo's work that showed a strong correlation between presence of cnf and sepsis severity.¹ Hly genes were found in a higher proportion of isolates causing pyelonephritis than cystitis and finally that traT was shown to be the most prevalent virulence determinant. This study highlights the diversity of bacterial VF expression in various host contexts and indicates that understanding these is a useful tool for patient management.

Lefort et al.⁹ reported that host driven risk factors and mode of bacterial entry were more important determinants of severity of sepsis then VFs. This COLIBAFI study examined more than 1000 episodes of bacteremia from 15 centers and correlated severity of these episodes with bacterial determinants. Death was found to be associated with patient characteristics, older age, presentation of cirrhosis and previous hospitalizations. Presence of the ireA gene (a siderophore) was negatively correlated with death. While the group also presented data highlighting the route of entry of the bacterial strain as associated with death, one limitation of the study was that more than a quarter of all strains had no entry model. Furthermore the study focused on a limited number (18) of VFs and did not account for differences in host genetic background. These limitations highlight the vast diversity of factors that must be included in these investigations. More robust examination of the genomic sequence of ExPEC strains was performed by Salipante et al 2015.¹⁰ Using clinical isolates obtained from routine diagnostic culture of peripheral blood and or urine and focusing on more than 3 thousand genes determined no evidence of a phylogenomic division between infectious strains. This study also revealed that differences in VF content and antibiotic resistance phenotypes, which distinguish population level group of E. coli, do not persist when the overall population structure is taken into account. Some genes have multiple instances of independent acquisition or loss within lineages, identifying them as important to specific lineages or groups of infections. These findings highlight the difficulty, even when using next-generation sequencing, in interpreting results.

Further confounding these issues studies like Micenkova et al.'s ¹¹ from 2014 demonstrate that bacteriocin-encoding genes and ExPEC VFs are associated in human samples. Using data from both virulence determinants and bacteriocin encoding genes it was determined that these factors correlate positively in E.coli strains of human fecal origin. Other groups have published similar findings, looking at the relationship of microcins and bacteriocin synthesis with VFs in human disease.^12-14 The blaTEM findings from the following article examined antibiotic resistance gene expression and found that in their population blaTEM was associated with increased risk of severe sepsis.

While this study aligns itself neatly with previously published data in the field it revealed that BOTH patient risk factor AND presence of virulence factors have impact on disease severity and mortality rates in sepsis induced by E.coli bacteremia. This investigation and others mentioned above highlight the importance of studying the host microbe interaction from numerous contexts. Decreased costs and increased access to /accuracy of high throughput next generation sequencing platforms have greatly furthered our level of understanding of pathogenic bacteria and their evolved strategies to circumvent human immune responses and the complex interplay between commensal bacteria preventing pathogenic infection. So while the “V’-factor may be important it's value may be predicated in a context dependent manner.