Bacillary Dysentery: Shigella and Enteroinvasive Escherichia coli

Isolation Techniques

The infecting strain of Shigella is generally present in stool in concentrations between 10 ³ and 10 ⁹ viable cells per gram of stool, depending on the stage of illness. During the postconvalescent shedding period, counts fall to 10 ² to 10 ³ viable cells per gram of stool. Recovery of the agent microbiologically is not usually difficult in the early stages of disease because of the higher counts present; it is more difficult during later stages of illness because of the lower counts of viable bacteria. Patients with shigellosis at the height of their illness can have negative stool cultures. Careful selection of feces and processing on appropriate media give a higher yield of organisms. The sooner after passage the specimen is processed, the higher is the yield. Stool that stands at room temperature for more than 24 hours has a profound drop in the number of viable cells, and recovery of the pathogen is less likely. A rectal swab obtained and seeded immediately at the bedside is the optimal way to perform a stool culture.

For bacteriologic identification of Shigella , a bit of blood or mucus is seeded onto at least two different media. In general, stool is plated lightly onto a medium with only mild inhibiting factors for gram-negative growth, such as MacConkey agar, xylose-lysine-deoxycholate agar, Tergitol 7, or eosin–methylene blue (EMB) agar, whereas a separate specimen is plated heavily onto a more inhibitory medium, such as Shigella-Salmonella medium. The more plates used, the greater the recovery yield. After overnight incubation at 37°C, lactose-negative colonies are tested biochemically and then serologically identified with Shigella grouping and typing antisera.

Group and Type Identification

Shigella are divided into four groups, depending on serologic similarity and fermentation reactions: group A (Shigella dysenteriae), with 15 serotypes; group B (Shigella flexneri), with 19 serotypes and subserotypes; group C (Shigella boydii), with 20 serotypes; and group D (Shigella sonnei), with a single serotype . Commercial antiserum is available for determining group- and type-specific antigenicity. S. sonnei accounts for 60% to 80% of the cases currently reported in the United States and other industrialized areas.

Invasive Escherichia coli

Certain strains of E. coli can cause a clinical illness indistinguishable from shigellosis and should be considered as causative agents of bacillary dysentery. Almost all the Shigella -like E. coli strains have been shown to possess somatic antigens related to Shigella serotypes, further demonstrating the similarity of these two groups of organisms. EIEC strains that cause bacillary dysentery have been shown to belong serologically to the following E. coli O groups: 28, 29, 112, 115, 124, 136, 143, 144, 147, 152, 164, and 167. Serotyping may ultimately prove to be useful in detecting these strains. At the genomic level, most EIEC strains are distributed across three lineages. The classic laboratory test for determining the virulence of a bacterial isolate ( Shigella or EIEC strain) was the Sereny test. Keratoconjunctivitis develops after 1 to 7 days in guinea pigs (or rabbits) when an invasive bacterial strain ( E. coli or Shigella ) is dropped into the conjunctival sac of the animal ( Fig. 224.1 ). This test is no longer used for diagnostic purposes. As discussed earlier, Shigella and EIEC have evolved from common ancestors and conceptually form a single pathovar within E. coli, with different lineages.

A different form of bacillary dysentery has been shown to be caused by an O157:H7 strain of E. coli and other Shiga toxin (STx)–producing E. coli strains (see Chapter 218 ).

Communicability and Infectivity

Shigellosis is the most communicable of the bacterial diarrheas. Experiments in volunteers have demonstrated that shigellosis is unique among bacterial enteropathogens in that fewer than 100 viable cells can readily produce the disease in healthy adults. Dose-response data obtained in volunteers for virulent strains from three species of Shigella are given in Table 224.1 . When volunteers ingested 500 or fewer viable cells of S. flexneri, S. sonnei , or S. dysenteriae type 1 (the Shiga bacillus), essentially the same rate of clinical illness resulted—27% to 45%. In general, the laboratory-passaged strains used in volunteer studies are less infectious than naturally transmitted strains, implying that the actual infectious dose in the field is considerably lower than 500 organisms. This low dose of organisms probably explains how the illness can be so readily transferred from person to person, why the secondary attack rate is so high when an index case is introduced into a family, and why recurrent bacillary dysentery is an important problem in institutionalized or crowded populations.

The reasons for this low dose response are not completely clear. One possible explanation is that virulent shigellae can withstand the low pH of gastric juice. In a study of adult Bangladeshi men admitted to the hospital with diarrhea, normal gastric acid levels were seen in patients with shigellosis, amebiasis, and pathogen-negative diarrhea, whereas patients with secretory diarrhea caused by Vibrio cholerae and enterotoxigenic E. coli had low gastric acid levels, offering evidence that Shigella did not require reduced gastric acidity to produce enteric disease. In another study, Shigella isolates were able to survive at a pH of 2.5 for at least 2 hours, whereas Salmonella was not. Shigella strains were also shown to be able to survive in acidic apple juice and tomato juice stored at 7°C and 22°C, respectively, for up to 14 days, showing its resistance to acid. A study in Kenyan children suggested that unabsorbed iron could be a direct risk for pathogens or for intestinal dysbiosis. Although EIEC and Shigella possess the same virulence determinates, infection with IEC requires an average dose 1000 times higher to cause infection. EIEC strains have not been compared with Shigella to determine if relative acid susceptibility might explain the different dose response. Nonpathogenic E. coli strains appear to have similar acid susceptibility to strains of Shigella, suggesting that this is not the reason for the difference in dose response.

Mucosal Invasion and Inflammation

Virulent Shigella and other nontoxigenic invasive E. coli strains produce disease after invading the intestinal mucosa via the basolateral surface of the colonic enterocyte. Genes required for bacterial entry into epithelial cells are present on a 30-kb region of a 220-kb virulence plasmid. Shigella infection is superficial, and only rarely does the organism penetrate beyond the mucosa, which explains the rarity of obtaining positive blood cultures in patients with shigellosis, despite the common occurrence of hyperpyrexia and toxemia. Shigella and EIEC invade colonic and rectal cells, including M cells of the follicle-associated epithelium, macrophages, and epithelial cells; invasion is followed by intracellular multiplication, spread of infection to adjacent cells, severe inflammation, and destruction of colonic mucosa. Apoptotic destruction of macrophages in subepithelial tissue allows survival of the invading shigellae, and inflammation facilitates further bacterial entry. Once the organisms are intracellular, they multiply within the cytoplasm and move from cell to cell by an actin-dependent process.

Pathogenic strains of Shigella and other bacterial enteropathogens have evolved a complex type III secretion mechanism that enables them to invade the intestinal mucosa. Bacterial proteins (including toxins) are injected from the bacterial cytoplasm into the cytosol of host mucosal cells, where they modulate the functions of the host cells and dictate how the host and pathogen relate. Each type III system consists of the secretion apparatus, secreted effector proteins, cytoplasmic chaperones (specialized for transporting the specific effector proteins), and specific transcriptional regulators. The secreted proteins—there are approximately 20, including VirA, OspB to OspG, IpaA to IpaD, and IpgD—facilitate bacterial entry into nonphagocytic cells and induce apoptosis and are targets for vaccine development. Shigella strains decrease production of host antimicrobial peptides, thus facilitating persistence by avoiding the host immune system. Strains of bacterial pathogens that use type III secretion systems can be detected by screening for virulence genes directly.