Commentary: Campylobacter and Hemolytic Uremic Syndrome

Introduction

C ampylobacter spp. trigger the largest number of cases of bacterial-mediated gastrointestinal disease annually in the developed world (CDC, 2006; Humphrey et al., 2007). Campylobacter are also associated with serious postinfection neurological disorders such as Guillian-Barré and Miller Fisher syndromes (Hughes and Cornblath, 2005; Takahashi et al., 2005). In addition to these extensively researched diseases, a small number of clinical studies have implicated Campylobacter in the development of hemolytic uremic syndrome (HUS) (Table 1). HUS is characterized by a nonimmune hemolytic anemia, a low platelet count, and renal impairment (Noris and Remuzzi, 2009).

In a recent publication, Bowen et al. (2016) described a case of Campylobacter-induced HUS in a 22-year-old English woman. The patient was transferred to a hospital renal unit suffering with watery, nonbloody diarrhea, vomiting, and headache; her stool culture was positive for Campylobacter jejuni. Laboratory workup indicated that the patient had acute kidney damage, anemia, thrombocytopenia, and a blood smear showed red blood cell fragments; therefore, the patient was diagnosed with diarrhea-associated HUS (D+HUS) (Canpolat, 2015).

Bowen et al. (2016) and other clinicians (Table 1) do not discuss mechanisms by which a Campylobacter infection induces HUS. At the present time, ∼90% of HUS cases worldwide are caused by Shiga toxin-producing Escherichia coli (Abdullah et al., 2014; Kavanagh et al., 2014) and a few cases are induced by an invasive Shigella dysenteriae or Streptococcus pneumoniae infection (Copelovitch and Kaplan, 2008; Taylor, 2008; Smith et al., 2013).

Dasti et al. (2010) state that the only verified toxin in Campylobacter is cytolethal distending toxin (CDT). The toxin is made up of three subunits: CdtA, CdtB, and CdtC (Dasti et al., 2010; Silva et al., 2011; Bolton, 2015). CdtA and CdtC are dimeric subunits, which deliver CdtB to the mammalian cell cytoplasm; CdtB displays DNase I-like action and induces DNA double-stranded breaks followed by cell cycle arrest, cell distension, and cell death (Smith and Bayles, 2006). However, CDT does not appear to be associated with the induction of HUS.

Toxin-Mediated HUS

S. dysenteriae type 1 produces Shiga toxin (Stx) and Shiga toxin-producing E. coli (STEC) produce Stx1 and Stx2. There is only one amino acid difference between Stx1 and Stx, whereas Stx2 is 56% homologous to Stx/Stx1 at the deduced amino acid sequence level (Jackson et al., 1987; Pal, 2015). Stx1 and Stx2 are antigenically distinct, but have the same mode of action as Stx. Both Stx1 and Stx2 have several amino acid subtypes (variants) (Doughari et al., 2009; Fagerquist et al., 2014). Stx of S. dysenteriae type 1 has been associated with a small number of D+HUS cases, especially in developing countries (Taylor, 2008).

STEC accounts for ca. 90% of D+HUS cases in Europe and America (Abdullah et al., 2014), and those strains producing Stx2 are the major inducers of HUS (Scheiring et al., 2008). Flagler et al. (2010) stated that Stx2 is more toxic in vivo than Stx1, but the difference is not clear. Differences in binding of the toxins to the receptor Gb3 (globotriaosylceramide) may play a role in determining toxicity.

To date no Campylobacter strain has been identified that possesses a homolog to any known Shiga toxin-producing genes. An intergenus horizontal gene transfer event necessary to bring a stx gene into Campylobacter from a non-Campylobacter donor appears to be a very unlikely event. The recognition and degradation of invading foreign nucleic acids by the CRISPR-Cas and restriction-modification systems are what probably act as barriers to this type of horizontal gene transfer in C. jejuni and could explain why the genetic transfer of HUS induction has not occurred (Gardner and Olson, 2012).

Stx is an AB₅ toxin: the A subunit causes the toxic effects, and the B₅ subunit promotes binding of Stx to endothelial cells lining the interior of blood vessels of the kidney, brain, liver, pancreas, heart, and hematopoietic cells possessing the Gb3 receptor (Melton-Celsa, 2014). Following infection and colonization of the intestinal mucosa by STEC, patients develop a profuse diarrhea which may be bloody, indicating hemorrhagic colitis. The bacterial cells lyse and release Stx into the intestinal lumen and are then transported to underlying tissues (Melton-Celsa, 2014). The B subunit of Stx binds the toxin to Gb3 with cellular internalization of the receptor-holotoxin complex. The A subunit is released from the holotoxin and inhibits cellular protein synthesis, leading to endothelial cell death. The kidney is the main target of STEC HUS, but the central nervous system can be affected, as well. D+HUS is due to Stx-facilitated injury to the kidney microvascular endothelial cells with the initiation of the coagulation cascade, which creates a blockage in the kidney venous system with resultant obstruction of kidney activity (Zoja et al., 2010; Keir et al., 2012; Cheung and Tractman, 2014; Melton-Celsa, 2014).

Ardissino et al. (2014) have shown that children with bloody diarrhea caused by STEC are often (35.6–39.9%) coinfected with either Campylobacter or Salmonella. The limited number of cases of typical D+HUS postulated to have a non-STEC etiology are potentially STEC-associated, but the patients are coinfected with another enteropathogen such as Salmonella or Campylobacter. Bowen et al. (2016) do not mention testing for the presence of STEC, Stx1, and/or Stx2, or the genes for Shiga toxin synthesis. Therefore, it is not completely certain that C. jejuni was responsible for the induction of HUS. Only two of the HUS cases described in Table 1 (cases described by Carter and Cimolai, 1996; Soper et al., 2000) indicated that STEC strains were not found. Therefore, the isolation of a bacterium (other than STEC) in HUS cases that is stated to be the HUS-inducer should be carefully evaluated.

Pneumococcal HUS (pHUS) is induced by an invasive S. pneumoniae infection, and similarly to STEC-induced HUS, pHUS shows the triad of acute kidney injury, microangiopathic hemolytic anemia, and thrombocytopenia and accounts for 5–15% of all HUS cases (Copelovitch and Kaplan, 2008; Smith et al., 2013). Diarrhea is rarely associated with pHUS. While not demonstrated in vivo, it has been postulated that the Thomsen–Friedenreich (TF) cryptantigen is involved in HUS induction. The TF antigen is part of the surface structure of erythrocytes, platelets, and glomerular endothelial cells and is concealed by a layer of neuraminic acid. S. pneumoniae neuraminidase A cleaves n-acetyl neuraminic acid from the surface of the cells with exposure of the TF antigen. Preformed host IgM antibodies bind the exposed TF antigen leading to endothelial injury initiating a cascade of events resulting in HUS (Copelovitch and Kaplan, 2008; Coats et al., 2011; Smith et al., 2013). This type of neuraminidase activity has not been observed in Campylobacter.

Atypical HUS

A number of mutations in genes encoding the complement regulatory proteins have been described in patients with atypical hemolytic uremic syndrome (aHUS). Most of the mutations occur in the complement genes CFH, MCP, and CFI that encode the regulatory proteins H, membrane cofactor protein, and protein I, respectively (Hirt-Minkowski et al., 2010; Kavanagh et al., 2013). The disorder is caused by a chronic uncontrolled activation of the alternative complement system leading to the HUS triad. Gene mutations in complement regulatory proteins, which normally protect host cells from complement activation render the regulatory proteins nonprotective. The uncontrolled complement activation involves an initial injury of kidney endothelial cells, followed by obstruction of small arterioles and capillaries of the kidneys by platelet plugs/fibrin thrombi, leading to loss of kidney function (Hirt-Minkowski et al., 2010; Loirat and Frémeaux-Bacchi, 2011).

aHUS accounts for only 5–10% of all HUS cases (Hirt-Minkowski et al., 2010). Diarrhea/gastroenteritis can trigger aHUS as demonstrated in French pediatric patients and Italian pediatric and adult patients; however, the diarrhea/gastroenteritis-inducing organisms were not identified (Loirat and Frémeaux-Bacchi, 2011). Kavanagh et al. (2013) has suggested that a Campylobacter upsaliensis diarrhea triggered aHUS in the case report of Carter and Cimolai in Table 1; however, Carter and Cimolai (1996) did not determine that there were aHUS-mutations in their patient. Thus, it is not certain that aHUS was triggered by C. upsaliensis. Reliable reports implicating Campylobacter-induced gastroenteritis as a trigger for aHUS were not found.

Conclusions

To determine if Campylobacter infections are associated with HUS, a framework is proposed to help guide future clinical investigations to more clearly define any role that Campylobacter spp. may play in the development of HUS. Therefore, future studies observing a possible relationship between Campylobacter and HUS should take steps to address the following concerns. The presence of known Stx toxins, the genes that code for these toxins, and the E. coli types that produce these toxins should be excluded. The patient should be screened for genetic mutations that can produce aHUS as well as for any physical abnormalities of the kidneys that could produce a dysfunction that could be mistaken for HUS. Finally, any Campylobacter strain implicated in the development of HUS must be isolated and fully sequenced to produce a record of the organism's genetic makeup. This information is necessary to determine if the specific Campylobacter strain carries genetic elements not commonly found in most other Campylobacter isolates that could play a role in HUS. It is necessary to exclude more likely culprits for the induction of HUS as well as determining a mechanism by which Campylobacter could cause HUS. Thus, at the present time, any role for Campylobacter in HUS warrants further investigation.