Despite advances in laboratory design, professional training, and workplace biosafety guidelines, laboratory-acquired infections continue to occur. Effective tools are required to investigate cases and prevent future illness. Here, we demonstrate the value of whole-genome sequencing as a tool for the identification and source attribution of laboratory-acquired salmonellosis.
Salmonella is a leading cause of bacterial gastroenteritis, with an estimated global incidence of 93.8 million cases per year, including 155,000 deaths from diarrheal disease and 690,000 deaths from invasive disease (1, 2). Illness due to nontyphoidal Salmonella strains is usually self-limiting, but symptoms can be severe. Invasive disease is less common but more frequently fatal (3). Most cases of Salmonella enterocolitis are food-borne. Contaminated eggs and poultry are common sources of infection, but outbreaks have also been triggered by meat, fruit, nuts and even bread (4,5). Approximately 15% of cases involve other routes of transmission, including contaminated water, direct person-to-person contact, and laboratory-acquired infections.The Saskatchewan Disease Control Laboratory (SDCL) is a public health laboratory that provides clinical diagnostic and microbiology services to the people of Saskatchewan, Canada. In 2011, a medical laboratory technologist from the SDCL became ill with enterocolitis. Salmonella enterica serovar Typhimurium was isolated and, although there was no record of a laboratory accident, occupational exposure was investigated as a possible source of illness. By pulsed-field gel electrophoresis (PFGE), the technologist's isolate was indistinguishable from ATCC 14028, a commercially available strain of S. enterica serovar Typhimurium that is widely employed in clinical and educational settings. At the SDCL, a derivative of ATCC 14028, called SDCL 14028, is used for routine laboratory procedures, including quality control of Salmonella antisera and culture media. By PFGE, the XbaI profiles of SDCL 14028 and ATCC 14028 are identical. However, SDCL 14028 has a distinct BlnI profile (Fig. 1). Because in vitro passaging can result in mutations, archived aliquots of SDCL 14028 and a new aliquot purchased from the original supplier (Cedarlane, Ontario, Canada) were examined. All aliquots had the same variant BlnI profile (data not shown), indicating that the altered PFGE pattern was not an artifact of laboratory passaging. For SDCL 14028 to be the source of the technologist's infection, the strain would have had to revert to the wild type during the course of the clinical illness. Although theoretically possible, this was considered highly unlikely. As such, the initial investigation concluded that the infection was not acquired from a laboratory, and its source remained unknown.We have now revisited this case using a whole-genome sequencing approach. A review of the SDCL strain archive uncovered four clinical isolates (STM-A to STM-D) of S. enterica serovar Typhimurium with ATCC 14028-like PFGE profiles (Table 1). STM-A is...