8.6.7 Enterobacteria and bacterial food poisoning

8.6.7 Enterobacteria and bacterial food poisoning 1032

section 8  Infectious diseases 1032 Nassif X, et al. (1999). Interactions of pathogenic neisseria with host cells. Is it possible to assemble the puzzle? Mol Biol, 32, 1124–​32. Newman L, et al. (2015). Global estimates of the prevalence and inci- dence of four curable sexually transmitted infections in 2012 based on systematic review and global reporting. PLoS One, 10, e0143304. Ohnishi M, et al. (2011). Is Neisseria gonorrhoeae initiating a future era of untreatable gonorrhea? Detailed characterization of the first strain with high-​level resistance to ceftriaxone. Antimicrob Agents Chemother, 55, 3538–​45. Public Health England (2014). Guidance for the detection of gon- orrhoea in England. 2014. https://​www.bashh.org/​documents/​ Guidance%20for%20the%20detection%20of%20gonorrhoea%20 in%20England%20Aug%202014.pdf Sherrard J, Barlow D (1996). Gonorrhoea in men: clinical and diag- nostic aspects. Genitourin Med, 72, 422–​6. Tabrizi SN, et al. (2011). Evaluation of six commercial nucleic acid amplification tests for the detection of Neisseria gonorrhoeae and other Neisseria species. J Clin Microbiol, 49, 3610–​16. Taylor SN, et al. (2018). Single-dose zoliflodacin (ETX0914) for treat- ment of urogenital gonorrhea. N Engl J Med, 379, 1835–45. Unemo M (2015). Current and future antimicrobial treatment of gonorrhoea—​the rapidly evolving Neisseria gonorrhoeae continues to challenge. BMC Infect Dis, 15, 364. Unemo M, Ison C (2013). Gonorrhoea. In:  Unemo M, et  al. (eds) Laboratory diagnosis of sexually transmitted infections, in- cluding human immunodeficiency virus, pp. 21–​54. World Health Organization (WHO), Geneva, Switzerland. Unemo M, Shafer WM (2014). Antimicrobial resistance in Neisseria gonorrhoeae in the 21st century: past, evolution, and future. Clin Microbiol Rev, 27, 587–​613. Whiley DM, et al. (2008). Exploring ‘best practice’ for nucleic acid de- tection of Neisseria gonorrhoeae. Sex Health, 5, 17–​23. 8.6.7  Enterobacteria and bacterial food poisoning Hugh Pennington ESSENTIALS The worldwide impact of food poisoning is very great. Such in- fections kill many children in the developing world, where diar- rhoeal diseases stunt their physical and cognitive development. The number of illnesses is also large elsewhere: in the United Kingdom the most common cause of food poisoning, Campylobacter, accounts for about 500 000 cases every year. The most common bacterial pathogens are Campylobacter and various members of the Enterobacteriaceae, a large family of Gram-​negative organisms, of which Escherichia coli, shigella, and salmonella are considered in this chapter. Escherichia coli Pathogenic E. coli include the following: Enteropathogenic—​virulence-​positive enteropathogenic E.  coli are now rare in industrialized countries; food-​ and water-​borne and person-​to-​person spread occur, resulting in diarrhoeal illness; fewer than 500 cases are recorded annually in the United Kingdom. Enteroaggregative—​first isolated from malnourished children in Chile suffering from chronic diarrhoea; not routinely tested for in industrialized countries but probably common. Enterotoxigenic—​an important cause of mortality in children under 5 years of age in developing countries, and causes travellers’ diar- rhoea; adheres to the mucosal surface of epithelial cells of the prox- imal small bowel, a process mediated by at least 12 different kinds of pili encoded by transferable plasmids, and produces enterotoxins. Enteroinvasive—​like shigella, for all practical purposes. Enterohaemorrhagic—​the most important enterohaemorrhagic E. coli is E. coli O157:H7, which produces a toxin virtually identical to that of Shigella dysenteriae. E. coli O157 is a normal non​pathogenic inhabitant of the gastrointestinal tract of cattle and sheep; most human infections are contracted either by the consumption of foods contaminated with animal manure or by its direct ingestion, prob- ably from hands that have touched contaminated surfaces. Clinical presentation is with diarrhoea (becoming bloody in 90% of cases) and abdominal pain, with a few cases (15% of children <10) going on to develop haemolytic uraemic syndrome. Diagnosis is by culture. Management is supportive. An enterohaemorrhagic E. coli /​Eagg E. Coli hybrid caused the ser- ious E. coli O104:H4 outbreak in Germany in May to July 2011; attrib- uted to an organic farm producing fenugreek seed sprouts. Shigella Infections are exclusively human, spread by the faecal–​oral route from person-​to-​person, and with a very low infectious dose. Shigellosis is endemic in developing countries in tropical areas, and it probably kills about 600 000 annually, mostly young children. Presentation is with watery diarrhoea, fever, and malaise, with se- vere infections (most often caused by S. dysenteriae) progressing to diarrhoea comprising mucus, blood, and pus, along with severe ab- dominal cramps and tenesmus. Management of mild cases is sup- portive; severe cases are given antibiotics (ampicillin, co-​trimoxazole, tetracycline, ciprofloxacin, others) as guided by local antimicrobial susceptibility data. Salmonella There are over 2000 salmonella serotypes, all belonging to the single species Salmonella enterica. Those that cause food poisoning infect both animals and humans, and most infections are food-​borne, most often by poultry, with S. enteritidis (strictly a serotype rather than a species) the paradigmatic organism. Clinical presentation is typic- ally with headache, vomiting (not usually a prominent feature), diar- rhoea, abdominal pain, and fever. Metastatic infection sometimes occurs, particularly osteomyelitis and in atherosclerotic vessels, ab- normal heart valves, and joint prostheses. Management of mild cases is supportive; severe cases are given antibiotics, usually a quinolone or macrolide. Campylobacter This is by far the most common cause of bacterial gastroenteritis in the industrialized world, with an annual incidence of infection perhaps as high as 1 per 100 in the United Kingdom. The organ- isms are very common in the intestines of wild birds, poultry, cattle, and sheep, but the source of infection in most human cases is unknown. A  prodrome of fever and general aching sometimes

8.6.7  Enterobacteria and bacterial food poisoning 1033 precedes abdominal pain (sometimes severe) and diarrhoea (fre- quently bloody). Complications include reactive arthritis (1% of cases) and Guillain–​Barré syndrome. Most infections are self-​limiting, but aside from supportive care, antibiotics (often erythromycin or ciprofloxacin) are given to severe cases. Prevention Prevention of food poisoning depends on Hazard Analysis and Critical Control Points, which identifies hazards, identifies the points in a pro- cess where they may occur, and decides which points are critical to control to ensure consumer safety (e.g. in milk pasteurization the crit- ical control points are the temperatures reached during heating, its dur- ation, and the measures taken to prevent subsequent contamination). Introduction Food poisoning denotes gastrointestinal diseases caused by microbes transmitted in food or by microbial toxins preformed there. Food spoilage by microbes also has important consequences for human health because of its impact on food supply. Each year 10–​20% of the world’s annual cereal crop of approximately 2 × 109 tonnes is lost through spoilage by moulds. Much of this loss occurs in the humid tropics and contributes there to the nutritional deficiencies caused by other factors. The terms food poisoning and food-​borne disease overlap but are not synonymous. Thus, variant Creutzfeldt–​Jacob disease (vCJD), contracted by eating meat products from cows with bovine spongi- form encephalopathy, only fits under the food-​borne rubric because of its very long incubation period despite the absence of gastrointes- tinal symptomatology. It is the same for bovine tuberculosis trans- mitted by milk. The worldwide impact of food poisoning is very great, as rec- ognized by the World Health Organization Global Strategy for Food Safety (2002), and is the cause of death of many children in the developing world. Diarrhoeal diseases stunt the growth of children and impair their physical and cognitive development. Mortality rates are much lower in developed countries, but the number of illnesses is still large. Quantitation is difficult because of underreporting. A large national study of the number and causes of cases of infectious intestinal disease in the United Kingdom es- timated that in 2009 there were 16.9 million cases and over 1 mil- lion GP consultations due to infectious intestinal disease every year. Routes of transmission were not established in this study, but aetiologies indicate that food-​borne transmission occurred in a sig- nificant minority; for example, campylobacter was responsible for 500 000 cases. In the United States of America, it has been estimated that each year 3.645 million people contract bacterial food-​borne disease with 35 796 being hospitalized and 865 deaths. The human intestine is home to 1013 to 1014 microorganisms. Their collective genome contains at least 100 times as many genes as the human one. They metabolize glycans and amino acids, detoxify xenobiotics, and synthesize isoprenoids and vitamins. A prominent member of the distal gut and faecal flora is the methane synthe- sizer, Methanobrevibacter smithii. However, the taxonomic identity and precise properties of most gut microbes is unknown because they cannot be grown in the laboratory. Cultivable ones that cause disease comprise a small minority, even when those opportunistic pathogens which occur primarily as commensals are included, such as members of the Enterobacteriaceae. This family contains several important causes of disease. Salmonella Typhi and S. Paratyphi, the causes of typhoid and paratyphoid fevers, are members of the Enterobacteriaceae. They cause systemic disease and are described in detail in Chapter 8.6.9. Gastroenteritis caused by non-​Enterobacteriaceae (i.e. the Gram-​negative organisms Aeromonas, Plesiomonas, Vibrio para- haemolyticus and other non-​cholera vibrios), and, most important quantitatively by incidence, Campylobacter, are included here, to- gether with accounts of food poisoning caused by Bacillus spp. For descriptions of diseases caused by Clostridium botulinum and C. perfringens, see Chapter 8.6.25, C. difficile see Chapter 8.6.24, and Staphylococcus aureus see Chapter 8.6.4. An overview of infections of the intestinal tract is given in Chapter 15.18. Enterobacteriaceae The Enterobacteriaceae is a large family of Gram-​negative bacteria. Many species are free-​living, some are associated with plants and can be plant pathogens, and others live in the intestines of animals and humans. The pathogens considered in detail in this chapter be- long to the genera Escherichia, Shigella, and Salmonella. The formal bacteriological definition of the family is that its mem- bers are non-​sporing Gram-​negative rods that are often motile, usu- ally by peritrichous flagella. They are easily cultivable on ordinary laboratory media. They might or might not have capsules. All are aerobes, although many grow anaerobically as well. All ferment glu- cose with the formation of acid and sometimes gas, and most reduce nitrate to nitrite. They are oxidase negative and, with the exception of one type of Shigella dysenteriae, are catalase positive. For a century, species in the family have been identified by carbo- hydrate fermentation patterns and by testing the reactivity of the bacteria to antisera prepared against their surface structures. Salmonella and Shigella do not ferment lactose. With the exception of proteus, providencia, and morganella, all other Enterobacteriaceae ferment this sugar freely with acid production. After cultivation on agar medium containing lactose and a pH indicator, non-​lactose-​ fermenting colonies stand out because of their colour differ- ence, making the initial detection of salmonella or shigella a fairly straightforward task. A similar approach is used to detect the most frequently occurring enterohaemorrhagic Escherichia coli sero- type, E. coli O157:H7, most isolates of which do not ferment sorb- itol. Antigenic epitopes key in identification schemes reside in the thick outer bacterial layer (the O antigens) and in flagella (the H antigens). The outer layer is a complex of lipopolysaccharide protein and lipid. The lipopolysaccharide has a hydrophobic lipid A com- ponent (responsible for the pathological effects of endotoxin) and a hydrophilic polysaccharide made up of an O-​specific polysaccharide and a core oligosaccharide. K-​antigen epitopes reside on a capsular polysaccharide which when present covers the O antigens. It can be removed by boiling. Enterobacteriaceae have a clonal population structure. Each in- dividual pathogen has a common ancestor and the incidence of the disease it causes correlates with the population size of the clone that has grown from it. The task of the diagnostic laboratory is to detect

section 8  Infectious diseases 1034 and identify these clones as quickly and as cheaply as possible. In general, the traditional tests described here satisfy these require- ments. Enterobacteriaceae clones evolve in real time, however, so markers can often be found to distinguish strains, even those with a recent common clonal origin. Tests that determine the suscepti- bility of isolates to a range of bacterial viruses, bacteriophage typing, have been widely used for this purpose, particularly in the United Kingdom. Molecular methods that detect DNA sequence differences are widely used internationally and have universal applicability and high discriminatory power. The Enterobacteriaceae considered here live in the intestines of animals and humans. This environment facilitates gene exchange between individual bacteria. It has been known for many years that plasmids, bacteriophages, and transposons are mobile gen- etic elements. Studies on E. coli virulence factors in the 1990s led to the discovery of pathogenicity islands, large genomic regions that are present in pathogenic strains but not in related non​pathogens. They carry genes associated with virulence, are often associated with tRNA genes, and are frequently flanked by repeat sequences. Their G + C content is different from the rest of the bacterial chromosome. DNA sequencing studies have shown more recently that similar islands also occur in non​pathogenic strains. The functions they en- code contribute to increased adaptability, fitness, and competitive- ness. Genomic islands have been acquired from other bacteria, not necessarily closely related ones. With the other mobile genetic elem- ents they form part of a flexible gene pool which confers beneficial traits supplementing the essential functions encoded by the con- served core genome. Promiscuous plasmid exchange coupled with the spread of clones has been responsible for the emergence and spread of Enterobacteriaceae that produce New Delhi metallo-​β-​lactamase-​1 (NDM-​1). This inactivates all β-​lactam antibiotics except aztreonam. The blaNDM-​1 gene responsible usually occurs in isolates already resistant to most antibiotics; they only remain sensitive to colistin, fosfomycin, and, less consistently, to tigecycline. The first NDM-​1 infection was identified in 2008. Since then they have been iden- tified worldwide, mostly in E. coli and Klebsiella pneumoniae and usually in patients who have been treated in hospitals in the Indian subcontinent. In New Delhi the blaNDM-​1 gene has been found in Enterobacteriaceae (including shigella), aeromonads, and Vibrio cholerae, and in tap water and environmental water. Molecular genetics has shown that shigella and E. coli are so closely related that formally they belong to a single genus. Escherichia has priority; however, Shigella is a useful name and is likely to continue in use for the foreseeable future. Likewise, the enthusiasm of those who gave hundreds of specific names to salmonella strains distin- guished by serotyping was misplaced. The strains are so closely related they are now referred to as serovars of the single species, Salmonella enterica. Escherichia coli Theodor Escherich was the first to grow E. coli in pure culture. He employed the ‘Plattenmethode’ described by Robert Koch in 1881 in his investigation in Munich in 1885 of the intestinal bacterial flora of newborns. It was the first detailed study of human commensal bacteria; appropriately so, because an overwhelming majority of the hundred billion billion E. coli bacteria that live in the world at any time are normal inhabitants of the intestines of healthy humans and animals. The perception in the 1940s by molecular biologists of its harmless nature coupled with its non​fastidious cultural require- ments and rapid growth (2–​3 generations/​h in the laboratory) led to its choice as a model organism. The strain most often studied is K12, isolated from the faeces of an American convalescent diphtheria pa- tient in 1922. More Nobel prizes have been won by researchers on E. coli than any other species (with the exception of Homo sapiens). The identification of E.  coli using traditional bacteriological methods is straightforward. It is a non-​sporing Gram-​negative rod, usually motile with peritrichous flagella, facultatively anaerobic, and a gas producer from fermentable carbohydrates. The methyl red re- action is positive and the Voges–​Proskauer reaction negative. Many strains have a polysaccharide capsule or microcapsule and most rapidly ferment lactose. Finding such an organism in a normally sterile site such as cerebrospinal fluid or in larger numbers in urine than can be accounted for by contamination is sufficient to indicate an aetiological role. Different approaches have to be used to detect enterovirulent E. coli in stools. Selective indicator media have been developed for E. coli O157. Other kinds are not looked for routinely; the best methods for detection use DNA probes or polymerase chain reaction amplification procedures that are only available in refer- ence laboratories. Few studies have been done on the carriage of commensal E. coli by healthy individuals but it is known that some carry a single clone for long periods, whereas others carry several simultaneously, acquiring and losing different clones rapidly. Some clones have a worldwide distribution; others seem to be only local. The genome sequence of strain K12 was published in 1997 and since then the genomes of representative pathogenic clones have been sequenced. A general principle has emerged that within the spe- cies there is an enormous amount of genetic diversity. Comparison of K12, a uropathogenic isolate, and an E. coli O157 showed that only 39.2% of the combined set of proteins was common to all. The genomes of the pathogens were as different from each other as each pathogen was from the commensal strain. Another E. coli charac- teristic is that different clones share a common genomic backbone of vertically evolved genes which is punctuated by many islands that have been acquired by different horizontal transfer events in each strain. All pathogenic E. coli are sticky, in that they produce structures on their surfaces that act as organelles of attachment. The pro- teins that make them sticky are adhesins, which recognize host cell structures—​receptors—​with stereochemical specificity. This fit is an important determinant of host specificity and tissue tropism. Adhesins are often assembled into hair-​like fibres, pili. Some take the form of a fuzzy mass on the bacterial surface, curli. Others form no particular oligomeric structures. The genomes of uropathogenic strains are also rich in genes coding for autotransporters, phase-​switch recombinases, and iron-​ sequestration systems. Enteropathogenic E. coli (EPEC) The isolation of antigenically identical E. coli strains during the in- vestigation of outbreaks of diarrhoea in young babies in the 1940s in London, Aberdeen, and Liverpool provided the first clear evidence that E. coli could be an intestinal pathogen. Subsequent serotyping showed the isolates to be O111 and O55. The disease they caused had a mortality of about 50% and mostly occurred in babies aged 6 months or less. Although volunteer studies in Liverpool showed

8.6.7  Enterobacteria and bacterial food poisoning 1035 that isolates from babies caused gastroenteritis in adults, a dose of 2 × 109 organisms only led to a mild short illness. Typical EPEC cause illness in infants and children under 2 years old; the hospital outbreaks that occurred in the 1940s are no longer seen. Intestinal colonization by typical EPEC involves virulence plasmid-​encoded type IV bundle-​forming pili which mediate bacterium to bacterium adherence and the formation of compact microcolonies on the surface of host cells, a pattern called localized adherence. EPEC fall into two related groups. Each contains several clones, some of which have been circulating for many years and have been found on several continents. O type does not always correlate with clonal type; thus type O142 marked two clones, one respon- sible for a high mortality outbreak in a Mexico City hospital in 1965 and the other for much less lethal infections of infants in Indonesia in 1960, hospital outbreaks in England, Scotland, and Ireland from 1969 to 1972, and sporadic cases in Canada in 1972 and Arizona in 1975. Virulence-​positive EPEC are now rare in industrialized coun- tries. Surveys in Brazil showed that they were common there in the 1980s and 1990s. In Europe and North America, EPEC lacking the virulence plasmid are now much more frequent causes of diarrhoea. These atypical EPEC are now becoming proportionally commoner in Brazil as well. A mechanism central to EPEC pathogenesis is the attaching and effacing (A/​E) lesion. At the sites of adhesion in the colon, intes- tinal cell microvilli disappear. Actin accumulates beneath the bac- teria, which become seated on pedestal-​like structures. The bacterial genes for the production of attaching and effacing lesions are located on the locus of enterocyte effacement pathogenicity island. It codes for intimin, an outer membrane protein responsible for adherence of the bacteria to enterocytes, Esp molecules, which are involved in the machinery that translocates bacterial proteins into enterocytes, and tir, which is translocated and inserts into the enterocyte cell mem- brane to act as the receptor for intimin. The incubation period of the diarrhoeal illness caused by EPEC ranges from 12 to 72 h, and the illness can last for several days. Food-​ borne, water-​borne, and person-​to-​person spread occur. Fewer than 500 cases are recorded annually in the United Kingdom. Enteroaggregative E. coli (EAggEC) These adhere to cell cultures in a ‘stacked brick’ pattern, a property often encoded on a 60-​MDa plasmid. Enteroaggregative E. coli were first isolated from malnourished children in Chile who had chronic diarrhoea and have been found since in Brazil, Mexico, India, and Zaire. They are not routinely tested for industrialized countries but are probably common. They have diverse O and H types. Little is known about their virulence factors or their precise pathogenic potential. Enterotoxigenic E. coli (ETEC) ETEC are an important cause of mortality in children under 5 years old in developing countries, and a significant cause of travellers’ diar- rhoea; 31 to 75% of Peace Corps volunteers in Africa with diarrhoea have been found to have ETEC in their stools. An incubation period of 12 to 72 h is followed by diarrhoea and vomiting lasting 3 to 5 days. ETEC adhere to the mucosal surface of epithelial cells of the prox- imal small bowel, a process mediated in different strains by at least 12 different kinds of pili encoded by transferable plasmids. There they produce enterotoxins, either a heat-​labile (LT) or a heat-​stable (ST) one, or both. LTs resembles cholera toxin in structure, mode of entry into cells, and toxic effects therein (see Chapter 8.6.12). There are different forms but are all made up of one A subunit and five B subunits. There are two kinds of the low molecular weight ST; ST-​1 increases intestinal secretion through a route that involves the acti- vation of cyclic guanosine monophosphate. ETEC enterotoxins are often plasmid encoded. Many E. coli O serotypes have ETEC virulence factors; different clones vary in pilus type and in the enterotoxins they express. Enteroinvasive E. coli (EIEC) For all practical purposes EIEC are like shigella (see next); they have the same virulence factors and cause watery diarrhoea. Enterohaemorrhagic E. coli (EHEC) The most important EHEC is E. coli O157:H7. Because it produces a toxin which is lethal to cultured African green monkey (Vero) cells and is virtually identical to that of Shigella dysenteriae serotype 1 it is often called VTEC or STEC. Epidemiology E.  coli O157 is a new pathogen. It came to notice abruptly and dramatically in the United States of America in 1982 where it in- fected consumers of beef burgers at a well-​known chain of fast food restaurants. The first outbreak in England was in 1983. There is a rough correlation between closeness to the north and south poles and the national incidence of infection, which is higher in Scotland than England, in Canada than the United States of America, and in Argentina than Brazil. Accurate figures on its incidence in tropical countries are not available; it is probably uncommon. E. coli O157 is a normal non​pathogenic inhabitant of the gastrointestinal tract of cattle and sheep. A significant minority of animals, up to 9%, carry it at any one time. Most tissue-​associated E. coli O157 adhere to mucosal epi- thelium in a region extending up to 5 cm proximal to the rectoanal junction characterized by a high density of lymphoid follicles. Transmission of infection in humans is by the faecal–​oral route. Person-​to-​person spread between young children occurs, and most infections are contracted either by the consumption of foods con- taminated with animal manure or by its direct ingestion, probably from hands that have touched contaminated surfaces. Prevention of the contamination of carcasses in slaughter houses is difficult, which explains why transmission by meat occurs. Transmission by burgers has been significant in the United States of America be- cause they are often consumed rare; maintaining 60° C for 2 min in their centre makes them safe. Many ready-​to-​eat foods have been vectors (e.g. lettuce). Poorly pasteurized milk, unpasteurized apple juice, and untreated drinking water have been important vehicles of transmission. Contamination of meats after cooking was important in the big Scottish outbreak in 1996, in which about 500 people were infected and 17 died. About 80% of infections are sporadic. In North America and Europe, they are more common in people who live in or who have visited rural areas; in a majority of infections a food vehicle cannot be identified and direct transmission probably occurs.

section 8  Infectious diseases 1036 Pathogenesis E. coli O157 has the locus of enterocyte effacement pathogenicity island and adheres to enterocytes with the production of attaching and effacing lesions. In this respect it resembles EPEC; it may be that the latter was its progenitor. It also produces Shiga toxins (Stx1, Stx2). They are made of a single A subunit and a B pentamer. Stx1 is almost identical to the toxin produced by Shigella dysenteriae type 1; there are several allelic variants of Stx2 which are 50% homolo- gous to Stx1 in amino acid sequence. The B subunit binds to the glycosphingolipid globotriaosylceramide on the surface of host cells; the A subunit enters and turns off protein synthesis by disrupting the large ribosomal subunit in a ricin-​like fashion. Shiga toxins induce apoptosis in human renal cells as well. Most pathogenic E. coli O157 are Stx2 gene positive; about two-​thirds are positive for Stx1. Clinical features After an incubation period ranging from 2 to 12 days, most com- monly 3 days, diarrhoea starts. In up to 90% of cases it becomes bloody after another 1 to 3 days. Asymptomatic infections are not rare. Most symptomatic cases are afebrile; abdominal pain is more severe than in other forms of bacterial gastroenteritis and abdom- inal tenderness is common. After between 5 and 13 days of diar- rhoeal onset, a minority of cases develop haemolytic uraemic syndrome (HUS). The risk is much greater at the extremes of age; about 15% of children under 10 years develop HUS. Other risk fac- tors are antibiotic administration and the use of antimotility agents. Thrombocytopenia is the first abnormality to develop. There is in- creased activity of plasminogen activator inhibitor 1 and the concen- tration of fibrin D-​dimers and thrombin fragments 1 and 2 becomes high. In full HUS (some cases never progress beyond thrombocyto- penia) the kidneys fail. Neurological complications—​thrombotic or haemorrhagic strokes, seizures, and coma—​occur in 10% of HUS cases and cardiac dysfunctions occur in about the same propor- tion; they are important determinants of mortality. No treatment has been shown to prevent the development of haemolytic uraemic syndrome or specifically affect its course; the vascular damage that causes it is almost certainly well underway when patients present with diarrhoea. There is no bacteraemia and at this time the Shiga toxin has probably already reached its target organs via the blood stream. Management is supportive rather than specific. Antibiotics, antimotility agents, and non​steroidal anti-​inflammatory drugs should not be given. Fluid balance should be monitored and treated carefully to avoid cardiac overload. Platelet monitoring will indicate whether the haemolytic uraemic syndrome risk period has passed. Anaemia sometimes requires transfusion. Renal failure requires specialist management, and renal function returns in the majority. The sequelae of E. coli O157 haemolytic uraemic syndrome mostly relate to renal function; risk factors for long-​term problems are the severity of the HUS itself and the need for dialysis. In most cases of haemolytic uraemic syndrome, long-​term problems have not been described. Laboratory diagnosis The diagnosis of E. coli O157 infection is by culture. Growth on se- lective media containing sorbitol leads to the formation of colourless colonies that are provisionally identified using O157 antiserum. For the detection of small numbers or organisms (e.g. in food suspected to be a vehicle of transmission), a procedure using enrichment cultures followed by a specific concentration step using magnetic beads covered with O157 antiserum is carried out. Direct tests for Shiga toxin have been developed. Subtyping by phage typing and by DNA sequence-​based profiling is an essential tool in outbreak investigation. E. coli O157 is the most common EHEC and cause of haemolytic uraemic syndrome. Other serotypes fall into these categories, and O26:H11, O103:H2, O111:H−, and O113:H21 have caused outbreaks in Australia and in continental Europe. Some pathogenic strains of E. coli O157 ferment sorbitol. Routinely used selective media detect none of these. Control The inability to influence the outcome of EHEC infections once es- tablished means that prevention is paramount. The development and implementation of preventive policies has been driven by the impact of big dramatic outbreaks, particularly those associated with burger chains in the United States of America in the 1990s and a butcher’s shop in Scotland in 1996. In the United States of America, the Food and Drug Administration classifies E. coli O157 as a food adulterant; in consequence its detection has very bad commercial effects. In the United Kingdom and in Europe as a whole, the implementa- tion of Hazard Analysis and Critical Control Points (HACCP)—​the evidence-​based food safety system—​has probably been driven more rapidly than it otherwise would have been. With occasional excep- tions these measures have worked well. For example, in Scotland rural/​environmental risk factors for infection now far outweigh food ones. Further reductions in the number of cases will be difficult to achieve. No effective measures for reducing E. coli O157 in rumin- ants have been devised, and ruminants shed large numbers into the environment. In north-​east Scotland (human population 5 × 105) it has been estimated that cattle and sheep drop about 3 × 1013 live E. coli O157 on the ground every day; the infectious dose of E. coli O157 for humans is very small, less than 100. Fortunately, the chain of events that leads to transmission from manure to mouth only oc- curs infrequently. In most years since the mid-​1990s the annual in- cidence of infection in Scotland by E. coli O157 has been the highest in the world, but it is usually about 4 per 100 000, so infections are uncommon. EHEC/​EAggEC hybrid Between May and July 2011 more than 3500 cases of gastroenteritis caused by E. coli O104:H4 occurred in Germany. A small outbreak in France at the same time caused by the same organism assisted epi- demiological investigations which showed a very clear association with the consumption of fenugreek seed sprouts. Big differences with E. coli O157 were that many (c.25% of those with gastroenter- itis) went on to develop haemolytic uraemic syndrome and that most of these were adults. The incubation period (median 8 days) was longer but the interval between the onset of diarrhoea (median 5 days) was shorter. The causative organism had EHEC character- istics (the Stx2 gene, a high pathogenicity island encoding an iron uptake system, and adhesin genes) as well as those characteristic of EAggEC (the virulence plasmid pAA, the aggA (coding for the pilin subunit of aggregative adherence fimbriae), aggR (a transcriptional regulator), plc (coding for an intestinal colonization protein) and

8.6.7  Enterobacteria and bacterial food poisoning 1037 set1 (coding for Shigella enterotoxin) genes. The fenugreek was con- taminated with faeces during growth, or harvesting, or processing. The source—​animal or human—​is not known. Shigella Bacteriologists working in Japan, Germany, and the Philippines in the early 20th century demonstrated the bacterial aetiology of many cases of dysentery, and that the causative organisms belonged to a group of related but different non​motile, non​capsulate Gram-​ negative bacilli closely resembling E. coli but differentiated from it by their inability to ferment lactose on overnight incubation. The names of the genus, Shigella, and three of the four species, S. flex- neri, S. boydii, and S. sonnei, commemorate them. The pioneer was Kiyoshi Shiga, who discovered S. dysenteriae in Tokyo in 1898. Epidemiology As countries become more affluent there is a fall in the number of shigella types circulating as common causes of disease. There is also a relative shift towards types that cause milder disease. S. dysente- riae type 1 causes the most severe disease. In the United Kingdom it had disappeared by the mid-​1920s, when several S. flexneri types and S. sonnei were endemic. In England and Wales after 1950, 95 to 98% of infections were caused by S. sonnei, although S. flexneri was still more common in Scotland. In the United States of America S. flexneri became less common than S. sonnei in 1968; currently in Thailand S. sonnei is becoming more common than S. flexneri. However, the propensity of shigella to cause epidemics has meant that the change in incidence of infection has not been one of unre- mitting reduction. Thus in England and Wales after a postwar peak of 49 000 notifications of S. sonnei dysentery in 1956, the incidence declined steadily to an annual average of 3000 notifications between 1970 and 1990. However, they rose sharply in 1991 and again in 1992, peaking at 17 000 cases and then falling again. S. dysenteriae type 1 became more common in Mexico and Central America in 1968, in the Indian subcontinent in 1975, and Central Africa during 1985. Shigella infections are exclusively human (monkeys are suscep- tible but very probably catch their infections from humans) and are spread by the faecal–​oral route. Volunteer studies and information from outbreaks caused by the faecal contamination of water and food on cruise liners have shown that the infectious dose is very low; dys- entery can follow the ingestion of 10 viable organisms. Most spread is person-​to-​person and infection is greatly facilitated by bringing people close together in institutions and circumstances where un- sanitary defaecation and inadequate hand washing is common; well-​ described examples are prisons in England in the early 19th century, mental hospitals in the United Kingdom, Germany, Denmark, and the United States of America later in the 19th century and in the early 20th century, British soldiers in Greece and Mesopotamia (now Iraq) in the First World War, and children in nursery and primary (elementary) schools in the United Kingdom in the early 1990s. It is considered that those with diarrhoea are by far the most effective transmitters of infection. After recovery many individuals continue to excrete organisms for a few weeks; temporary carriers of this kind are not thought to be important sources of infection, even if they are food handlers. Large and dramatic water-​borne outbreaks have occurred occasionally in industrialized countries; milk and ice cream have also been vectors. Vegetables contaminated with human faeces during growth, harvesting, or preparation have also caused outbreaks. Molecular typing has revealed the international nature of some; for example, more than 100 cases in the United Kingdom, Denmark, Norway, and Sweden in 1994 were shown in this way to be due to lettuce contaminated with an identical strain of S. sonnei. Shigellosis is endemic in developing countries in tropical areas; a long-​standing estimate is that it kills about 600 000 people—​mostly young children—​annually. More recently, there have been outbreaks of S. sonnei and S. flexneri in men who have sex with men, with global spread of antibiotic-​resistant strains. Pathogenesis Central to the pathogenesis of shigellosis (including that of enteroinvasive E. coli, which can be regarded as a variant of S. son- nei) is invasion of the colonic mucosa. Organisms gain access to the basolateral pole of enterocytes through M cells, components of in- testinal lymphoid follicles (Peyer’s patches). Bacteria infect macro- phages in these structures and kill them by apoptosis. Their release allows direct invasion and is associated with a cytokine-​induced in- flammatory response that facilitates bacterial invasion by disrupting the epithelial architecture. The entry of shigella into intestinal cells is actin-​microfilament dependent. Shortly after entry the bac- terium lyses its phagocytic vacuole and grows in the cytoplasm at a rate of about 40 min/​generation; most of the bacterial proteins responsible are plasmid encoded. Bacteria then spread from cell to cell. Infected cells die, and bacterial spread continues deep into the lamina propria. There is an acute inflammatory response dominated by polymorphonuclear leucocytes. Proctocolitis with epithelial des- quamation and purulent necrosis with ulcers leads to the produc- tion of bloody mucus. Spread of bacteria from the intestines to other parts of the body is rare. Shiga toxin is only produced by S. dysenteriae type 1 (see earlier). As with EHEC, infections with S. dysenteriae type 1 lead, in a mi- nority of cases, to haemolytic uraemic syndrome. The increased severity of S. dysenteriae type 1 proctocolitis compared with that caused by other Shigella spp. is probably due to the local effects of Shiga toxin on the colonic vasculature. Clinical features After an incubation period ranging from 12 h to 7 days, but most commonly 2 to 3 days, symptoms usually start suddenly, often with abdominal colic. Watery diarrhoea follows, usually with fever and malaise. The symptomatology of most S. sonnei infections progresses no further and, usually, the number of watery stools is small. The most severe infections are caused by S. dysenteriae. After 1 to 3 days the diarrhoea becomes bloody and very frequent, being composed of mucus, blood, and pus. Abdominal cramps and tenesmus are se- vere. Serious complications, sometimes lethal, are hyponatraemia, hypoglycaemia, septic shock, and haemolytic uraemic syndrome. Recovery in complicated cases is slow. More straightforward but se- vere illnesses, such as those not infrequently caused by S. flexneri and S. boydii, usually last about 4 days but might continue for 10 days or more. Laboratory diagnosis Diagnosis is by culture and traditional bacteriological methods work well; faeces are the best samples. Shigella dies rapidly when swabs dry and such samples should be transported to the labora- tory quickly. Inoculation of enrichment cultures from broths or

section 8  Infectious diseases 1038 direct inoculation onto special media gives colonies recognizable as shigella by morphology. Further identification is by biochemical tests and type-​specific antisera. DNA probes for plasmids are avail- able and Shiga toxin can be looked for. Treatment S. sonnei infections in healthy individuals other than those at the extremes of age do not benefit from antibiotic treatment. Agents re- ducing gut motility should be avoided. Antibiotic treatment of se- vere infections must be guided by antimicrobial susceptibility data; antibiotic-​resistant strains are common in areas where these infec- tions have a high incidence. Ampicillin, co-​trimoxazole, tetracycline, or ciprofloxacin have worked well; ceftriaxone and pivmecillinam have been successfully used to treat infections in children caused by antibiotic-​resistant strains. Control The occurrence of urban epidemic shigellosis in countries like the United Kingdom long after the universal provision of treated town water shows that, while the provision of safe water in parts of the world where serious shigella infections are common is a necessary general public health measure, it will not be sufficient. Interrupting faecal–​oral spread needs the provision of toilets and wash hand ba- sins in homes—​more a concomitant of economic development than of public health programmes. Salmonella The number of different salmonella clones is very large, but they all belong to the single species Salmonella enterica. Traditional bacterio- logical methods—​serotyping using O and H antigens and simple biochemical tests—​have been used to identify different kinds of sal- monella since the 1930s and they are good markers of clonal identity. The custom of referring to the entities they define as though they were species, for example, Salmonella enteritidis is taxonomically in- correct (they are serotypes) but operationally useful. A minority of salmonella serotypes has a host range limited to a single species, for example, for humans Salmonella typhi (see Chapter 8.6.9), and these serotypes not considered further here. The serotypes that cause food poisoning infect both animals and humans, and well over 2000 have been described. Epidemiology Person-​to-​person spread is uncommon and the infected/​carrier food handler is not an important source; faecal–​oral spread after contact with carrier animals such as terrapins and other reptiles occurs from time to time, but most infections are food-​borne. In the United Kingdom a big increase in microbiologically confirmed salmonella infection rates and the number of serotypes causing in- fections occurred in the late 1940s and early 1950s. A common pat- tern, which continues, is that a serotype appears, persists, and then declines. Their source for humans is food animals. Cattle, sheep, and pigs are far less important than poultry, although S. Typhimurium of bovine origin has remained quite common for many years. However, poultry dominate and the paradigmatic organism is S. enteritidis. It caused a panzootic in broiler and layer chicken flocks in Europe and the United States of America starting in the 1980s and concomi- tantly a human pandemic. In England and Wales it peaked in 1993; more than 525 000 fell ill during its course. In chickens, S. enteritidis not only grows in the intestines but also invades the reproductive tract leading to egg contamination. Since the early 1990s, control in flocks by slaughter, vaccination, and heightened biosecurity in hen houses has markedly reduced carriage levels in poultry, accounting for the decline in the number of human cases; in England and Wales in 2004 there were 2201 infections. The propensity of certain Salmonella serotypes to expand their population size has been enor- mously facilitated by the scale and nature of the poultry industry. The increase in the number of human infections since the 1940s has followed the expansion of broiler production. In 1950, United States broiler production was 631 million heads and per capita consump- tion was 8.7 pounds ready to cook; in 1990, it was 5864 million and 61.0 pounds, respectively. Cross-​contamination, where organisms from chicken car- casses have been transferred to ready-​to-​eat foods in the kitchen, has caused large outbreaks. Undercooked egg products are im- portant vectors of S. Enteritidis. Many other foods have been ve- hicles of transmission: unpasteurized milk, dried milk, desiccated coconut, alfalfa sprouts, mung bean sprouts, lettuce, and chocolate. Multicontinental outbreaks occur because of international trade (e.g. 4000 cases of S. Agona infection were caused by a contamin- ated kosher snack in the United Kingdom, Israel, the United States of America, and Canada in 1996). Pathogenesis Volunteer studies give an infectious dose ranging from 125 000 to 50 million organisms. For some foods, particularly those with much fat (e.g. cheese, potato chips, peanut butter, and chocolate), it is much less and ranges from fewer than 10 to 100. Organisms attack the distal small intestine and large intestine. At points of contact there is a transient denaturation of brush border microvilli, bacteria are internalized, and they remain in membrane-​bound compart- ments. Replication is necessary for virulence; their presence triggers a transepithelial migration of neutrophils. These processes need the action of many bacterial genes, some of which are in pathogenicity islands. Clinical features The incubation period ranges from 4 to 48 h but most commonly it is between 8 and 24 h. Onset is often sudden, with headache, vomiting (not usually a prominent feature), diarrhoea, and abdom- inal pain; fever is common. The clinical course is usually short, up to 2 to 3 days; in a minority it is severe and prostrating with dehy- dration. Mortality rates are low but are higher in infants (meningitis sometimes occurs) and elderly people with preexisting pathologies. Bloodstream infections are common in Africa in association with malaria and malnutrition in children and HIV infections in adults. Some serotypes (e.g. S. Dublin) are more virulent; bacteraemia with any serotype is usually transient but sometimes leads to metastatic infection, particularly in atherosclerotic vessels, abnormal heart valves, and joint prostheses. Osteomyelitis most frequently occurs in long bones, costochondral junctions, and the spine. Sickle cell anaemia is an important predisposing condition. Arthritis can be septic or reactive; the latter follows more than 1% of infections. It is most commonly seen in those with the HLA-​B27 haplotype. Faecal excretion of organisms continues for 4 to 8 weeks and is longer for infants; the number of organisms excreted is usually low. Carriage for longer than 6 months is rare.

8.6.7  Enterobacteria and bacterial food poisoning 1039 Laboratory diagnosis Diagnosis is by culture. Direct plating of faeces onto selective media and testing of suspicious colonies by slide agglutination for O antigens can give a presumptive diagnosis in 24 h. Enrichment broth cultures increase test sensitivity and are used to search for small numbers of bacteria in faeces or food. Phage typing schemes, available for S. enteritidis, S. typhimurium, and S. virchow, and DNA sequence-​based profiling methods have high resolution and are used to type isolates from patients and other sources in outbreaks. Treatment Fluid and electrolyte replacement is the management mainstay. Drugs that reduce gut motility are contraindicated. In uncom- plicated cases antibiotics have no place and they may prolong the excretion of organisms. In patients with a high risk of bac- teraemia and invasive disease (infants under 3 months, im- munosuppressed patients, patients with cancer, and those with haemoglobinopathies) antibiotics should be considered. Ciprofloxacin is usually the agent of choice but antibiotic-​ resistant strains have emerged and therapy must be guided by susceptibility testing. Cefotaxime and ceftriaxone have been of value in treating meningitis in infants. Prevention Preventing the infection of food animals is central and it has been successful in poultry in Northern Europe and North America. HACCP has been adopted worldwide; refrigeration and adequate cooking are very important critical control points. Campylobacter Campylobacter was discovered as a pathogen of sheep at the begin- ning of the 20th century, but 70 years elapsed before it was recog- nized as a common cause of human gastroenteritis. Its high optimum growth temperature (42°C), need for a microaerobic atmosphere, and requirement for help from selective medium to inhibit other competing gut bacteria hindered its detection. Campylobacter shares about 50% of its genes with helicobacter; both have a spiral shape and flagella. Most human infections are caused by Campylobacter jejuni and some by C. coli. Occasional infections are caused by C. fetus, an important pathogen of cattle and sheep, and sometimes in patients with immune deficiency. Epidemiology Campylobacter is by far the most common cause of bacterial gastro- enteritis in the industrialized world. In England and Wales in 2016, 52 381 laboratory isolates were recorded, 7.3 times more than for non​typhoidal salmonellae. In the United Kingdom it is estimated that for every case reported to national surveillance there are 9.3 cases in the community. Campylobacteriosis is a zoonosis. The organisms are very common inhabitants of the intestines of wild birds, poultry, cattle, and sheep. Mechanized processes in chicken abattoirs mean that most carcasses leave with surface contamination. However, the source of infection in most human cases is unknown; outbreaks, an invaluable epidemiological investigative tool, are rare, and the very great genotypic and phenotypic diversity of the C. jejuni genome caused by frequent horizontal gene exchange seriously impedes the development of epidemiologically useful typing systems. Multilocus sequence typing allows the identification of genetically related clonal complexes. The most common, ST-​21 has been iso- lated from human cases and healthy cattle, broiler chickens, wild birds, and sheep. Unlike Salmonella, the organisms do not grow on contamin- ated food, so outbreaks are uncommon. They have been associ- ated with failures in milk pasteurization and water chlorination. The incidence of sporadic human cases in the United Kingdom rises sharply in weeks 21 to 24 (May and June); the reason for this is unknown. Pathogenesis The infectious dose is low, fewer than 1000 viable organisms. The jejunum and ileum are colonized first, with extension distally, often to the colon and rectum. Infection is invasive; the mesenteric lymph glands enlarge and become inflamed and neutrophil polymorpho- nuclear leucocytes accumulate in the intestinal mucosa. A cytolethal distending toxin, phospholipase A, and flagellar structural proteins as well as other bacterial proteins with unknown functions are pro- duced by all pathogenic isolates. Clinical features The incubation period ranges from 1 to 7 days and averages 3 days. A prodrome of fever and general aching sometimes precedes ab- dominal pain and diarrhoea; vomiting is not a prominent feature. Abdominal pain can be severe and acute appendicitis is a frequent differential diagnosis. The diarrhoea contains leucocytes, is fre- quently bloody, and seldom lasts more than 2 to 3 days. Most pa- tients have culture-​negative stools after 5 weeks. Ten to 15% of patients have a recurrence of symptoms. About 1% of patients develop reactive arthritis 1 to 3 weeks after the onset of illness. It is indistinguishable from that which follows Salmonella infections. Campylobacter gastroenteritis is the most frequent event that leads to the development of Guillain–​Barré syndrome (Chapter 24.16); 26–​41% of cases have a history of its occurring 1 to 3 weeks after the onset of diarrhoea. Laboratory diagnosis Laboratory diagnosis is by culture. Stools are plated onto selective media and incubated for 48 h at 42–​43°C in 5–​15% oxygen and 1–​ 10% CO2. Infectivity is labile; if delays in transport to the laboratory are expected, faeces should be refrigerated or placed in transport medium. Diagnosis of recent infections is by serology. Treatment Most Campylobacter infections are self-​limiting. Fluid and elec- trolyte replacement may be needed. Most strains are sensitive to erythromycin; ciprofloxacin and other fluoroquinolones are also ef- fective in more severe infections, but resistant strains are becoming more common. Miscellaneous food poisoning bacteria Listeria monocytogenes See Chapter 8.6.38.

section 8  Infectious diseases 1040 Vibrio parahaemolyticus Vibrio parahaemolyticus is the most common bacterial cause of diar- rhoea (usually watery, sometimes explosive) in Japan. Infection fol- lows the consumption of seafoods, particularly those prepared raw in the Japanese style. The incubation period is usually 10 to 20 h (range 4–​9 h) and the illness lasts 1 to 2 days. Pathogenic strains produce a heat-​stable toxin and are Kanagawa positive (produce haemolysis on Wagatsuma’s agar). Other vibrios that cause seafood-​ associated gastroenteritis are V. fluvialis, V. hollisae, V. mimicus, and V. vulnificus. Aeromonas hydrophila This Gram-​negative rod is frequently isolated from diarrhoea. Virulence factors remain unidentified. Exotoxin producers See Chapters 8.6.24 and 8.6.25 for Clostridium difficile, C. botu- linum, and C. perfringens, and Chapter 8.6.4 for Staphylococcus aureus. Bacillus cereus This Gram-​positive saprophyte produces heat-​resistant spores. It is common in raw foods, especially rice, and causes two kinds of food poisoning, emetic and diarrhoeic. Vomiting occurs 6 h or less after eating food containing preformed toxin, usually lightly cooked rice that has then been stored at room temperature and reheated, condi- tions which stimulate the bacterium to produce the low molecular weight heat-​, acid-​, and protein-​resistant peptide toxin. Diarrhoea occurs 8–​24 h after eating contaminated food. A heat-​labile entero- toxin is produced in the intestine. Both kinds of illness are short lived. Other Bacillus spp., B. licheniformis, B. pumilis, and B. subtilis, have caused B. cereus-​like illnesses. Prevention of food poisoning The production of safe food rests on evidence-​based practical tech- nologies and management systems; HACCP is central to their de- livery. The system was developed by the National Aeronautics and Space Administration (NASA) and others in the 1960s to prevent food poisoning in space; the notion of diarrhoea and vomiting in zero gravity was too awful to contemplate. HACCP is now used worldwide and in many countries for some food businesses it is a legal requirement. It identifies hazards, identifies the points in a pro- cess where they may occur, and decides which points are critical to control to ensure consumer safety. A good example is milk pasteur- ization; critical control points are the temperatures reached during heating, its duration, and the measures taken to prevent subsequent contamination. As a written scheme testable by food law enforcers, HACCP stops at the farm gate and the dwelling door. However, its prin- ciples apply on the farm and in the home, and their promulgation there currently exercises all promoters of food safety. Ignorance of them is not restricted to these environments; large food poisoning outbreaks have followed failures of food processors to follow them. Milk pasteurization is again a good example. Political resistance to its implementation in England meant that 65 000 died there from milk-​borne bovine tuberculosis between 1912 and 1937. Thirty-​ nine milk-​borne salmonella outbreaks with deaths in Scotland between 1970 and 1981 drove legislation preventing the sale of un- pasteurized milk there, and now nearly all United Kingdom milk is pasteurized. However, pasteurization failures or postpasteurization contamination still lead to campylobacter and E.  coli O157 outbreaks. FURTHER READING Advisory Committee on the Microbial Safety of Food (2005). Second report on campylobacter. Food Standards Agency, London. Carlin F, Nguyen-​The C (2013). Pathogen update:  bacillus species. In: Sofos J, (ed). Advances in microbial food safety, Vol 1. Woodhead Publishing, Cambridge. Cheasty T, Smith HR (2010). Escherichia. In: Borriello SP, Murray PR, Funke G (eds) Topley and Wilson’s microbiology and microbial infec- tions (bacteriology). Wiley. DOI:10.1002/​9780470688618. taw 0052. Maskell D, Mastroeni P (eds) (2006). Salmonella infections:  clin- ical, immunological and molecular aspects. Advances in molec- ular and cellular microbiology. Cambridge University Press, Cambridge, UK. Nair GB, et al. (2007). Global dissemination of Vibrio parahaemo- lyticus serotype 03; K6 and its serovariants. Clin Microbiol Rev, 20, 39–​48. Nordmann P, Naas T, Poirel L (2011). Global spread of carbapenemase-​ producing Enterobacteriaceae. Emerg Infect Dis, 17, 1791–​8. Pennington TH (2010). Review. Escherichia coli O157. Lancet, 376, 1428–​35. Ricke SC, Dawoud TM, Kwon YM (2015). Application of molecular methods for traceability of foodborne pathogens in food safety systems. In:  Ricke SC, Donaldson JR, Phillips CA (eds) Food Safety Emerging Issues, Technologies, and Systems. Academic Press, London. Schroeder GN, Hilbi H (2008). Molecular pathogenesis of shigella spe- cies: controlling host cell signaling, invasion and death by type III secretion. Clin Microbiol Rev, 21, 134–​56. Tarr PI, Gordon CA, Chandler WI (2005). Shiga-​toxin-​producing Escherichia coli and haemolytic uraemic syndrome. Lancet, 365, 1073–​86. The Pennington Group (1997). Report on the circumstances leading to the 1996 outbreak of infection with E. coli O157 in Central Scotland, the implications for food safety and the lessons to be learned. The Stationery Office, Edinburgh. The Public Inquiry into the September 2005 outbreak of E. coli O157 in South Wales 2009. http://​gov.wales/​docs/​dhss/​publications/​ 150618ecoli-​reporten.pdf Threlfall EJ (2010). Salmonella. In: Borriello SP, Murray PR, Funke G (eds) Topley and Wilson’s microbiology and microbial infections (bac- teriology). Wiley. DOI:10.1002/​9780470688618. taw0054. Toro C, et al. (2015). Shigellosis in subjects with traveler’s diarrhea versus domestically acquired diarrhea: implications for antimicro- bial therapy and human immunodeficiency virus surveillance. Am J Trop Med Hyg, 93, 491–​6. Young KT, Davis LM, Dirita VJ (2007). Campylobacter jejuni:  mo- lecular biology and pathogenesis. Nat Rev Microbiol, 5, 665–​79.