Ecology and Epidemiology of Essherichia Coli O157:H7 in Fresh Produce Production Regions of the Central California Coast

NON-TECHNICAL SUMMARY: Consumption of fresh fruits and vegetables is growing in the United States, but has coincided with produce-associated outbreaks. Our hypothesis is that key biotic and abiotic processes link the primary environmental reservoirs. Preventing on-farm contamination of produce with enteric pathogens would enhance the safety of the Nation's food supply. We propose to sample vertebrate animals, water, soil and produce, for EcO157, and commensal and non-O157 shigatoxin+ E. coli, and use epidemiological approaches to determine if: (1) vertebrate populations are sources of EcO157 contamination of watersheds, soil, and plants; (2) climate, landscape, irrigation, or management practices correlate with contamination; and (3) in-field contamination of leafy vegetables is associated with production practices and environmental risk factors. This information will be used to inform growers about strategies to prevent contamination, to educate the livestock community about potential impacts of their operations, minimizing wild animal contact with fields.

<P>
APPROACH: Our overall study design combines epidemiological and microbiological methods followed by outreach and education activities to disseminate prevention and control information. We will conduct an in-depth longitudinal study that identifies the key biotic and abiotic processes that sufficiently load, then hydrologically link and disseminate, primary environmental reservoirs of EcO157 within and between LV fields, resulting in bacterial contamination of this raw agricultural commodity. For each node (vertebrate sources, water, soil, lettuce) of the system, we will collect a detailed set of covariates that will be used to identify critical control points, points of environmental amplification, and management practices that either elevate or decrease the risk of in-field contamination and dissemination of EcO157 on LV. Strains will be analyzed by Multi-Locus Variable number tandem repeat Analysis (MLVA) and Pulsed Field Gel Electrophorsis (PFGE) to characterize isolates by genetic differences, determine molecular epidemiologic linkages among samples isolated or obtained from public health collaborators, and thus, pinpoint the mechanisms that link and disseminate vertebrate sources of EcO157 within and between fields of LV. Hierarchical regression models will identify significant covariates for our outcomes of interest. We will also screen for non-O157 enterohemorrhagic E. coli, to determine if relevant serovars are also present in these environments.

<P>

PROGRESS: 2007/09 TO 2008/08 <BR>
OUTPUTS: The start date for this proposal was Oct. 1, 2008. Many of the objectives of this project are similar and very complementary to those of another CSREES funded project (2006-01240) starting in Oct 2007 and involving the Salinas Valley region of California. The major differences in this project are the region to be sampled (San Benito County) and addition of methods for isolation and characterization of non-O157 shiga toxin producing E. coli (STEC). Thus, efforts expended in year 1 and part of year 2 for project 2006-01240 was very beneficial for identifying study sites for this project. At least three growing sites and multiple ranches appropriate for studies of risk factors were identified and sampling and testing was initiated in April, 2008. A PCR method for measurement of the presence of shigatoxin in enrichment broths was developed and tested for sensitivity with isolates of O157 and non-O157 E. coli spiked into broths cultured with environmental samples (water, lettuce, soil). Soil, leafy vegetable, water samples and deer fecal samples collected during the 2007 and 2008 deer hunting seasons, and feral pig, wild bird and small wild mammal samples collected by USDA Wildlife Services personnel, were provided to the ARS lab for testing for O157. More than 2500 samples have been processed to date for E. coli O157:H7, generic E. coli and non-O157 STECs for both projects 1 and 2. The deer sample results have been of interest to the CA Dept of Fish and Game, growers and produce representatives as a result of deer being considered a "significant risk" in the Leafy Greens Marketing Agreement. Clinical strains of E. coli O157:H7 from multiple states (CA, HI, OR, ID, WA, MN), cattle isolates, and meat isolates have been obtained also and genotyped. A genotyping database was created as one of the objectives of the project. The database contains genotyping (multilocus variable number tandem repeat analysis, MLVA) and source data results for >1100 strains as of October-2008. The ARS lab has become a member of CDC PulsNet and will be submitting profiles as they are obtained. Sampling and testing will continue into the rainy season with an emphasis on livestock, wild animal and watershed sampling. Communications and meetings with other appropriate federal and state agencies expressing interest in collaboration are continuing. <BR>PARTICIPANTS: The PI (RM) and Co-PI (RA) continued meeting in Year 2 with growers and ranchers to identify farms and ranches to enroll. Two people were hired to sample farms and ranches two to three times each week. The PI has organized two meetings of the project collaborators. A collaborator, M. Jay-Russell, moved to the Western Institute for Food Safety and Security at UC Davis, and leads the sampling team and coordinates interactions with cooperators. All laboratory analyses are being done at the ARS laboratory (M. Cooley); three technicians were hired to assist. The PI communicated extensively with other public health agencies to obtain epidemiological information and relevant strains (outbreak, sporadic, produce-related) for genotyping. The Co-PI has consulted with K. Tate and R. Larsen on hydrology- and cattle-related issues relevant to the project. CDF&G has requested the PI's lab continue testing deer samples for E. coli O157:H7. Two USDA-APHIS wildlife staff members provide samples through an interagency agreement, in place of staff to be hired under the UC Davis specific cooperative agreement for this purpose. <BR>TARGET AUDIENCES: Outreach activities include numerous contacts of the PI and Co-PI with growers for enrollment of fields for study. Additional outreach activities include numerous meetings to explain the objectives and needs of the project with, for example, the CA Lettuce Research Board, CA Strawberry Commission, CA Dept. Fish and Game, Monterey County Resource Conservation District, USDA National Resource Conservation District, local and national cattlemen associations. The preliminary studies leading to the project, and project objectives and rationale, have been presented at numerous national and international meetings including IAFP, IFT, UK-Society for General Microbiology, FoodMicro-2008 and other meetings. There is broad awareness of the project by the major produce processors, trade organizations and many growers. <BR>PROJECT MODIFICATIONS: The objectives of the proposal were developed around the concept that flooding of fields during heavy rain events may explain some of the frequent outbreaks of E. coli O157. However, recognition of these results by processors and buyers has resulted in restrictions on purchase of produce from these types of fields. Therefore, our objectives have been modified slightly to minimize testing of the flooded field hypothesis, and concentrate more on testing wildlife near and on fields, as they may transport pathogens to fields directly. The other sampling objectives proposed remain the same.
<BR><BR>
IMPACT: 2007/09 TO 2008/08<BR>
Methods for measuring MPN of generic E. coli and methods for automation have been evaluated for sensitivity and ease of use for high-throughput testing. A PCR method for screening enrichment broths for presence of potential STEC isolates was developed based on different primers to increase the sensitivity of detection. This method proved to be useful, but further development will be needed to increase sensitivity. A new commercial master mix for environmental samples was tested and increased detection of STECs by >50%. Results with >2000 samples have revealed that the Colilert and EC Chromagar methods each have strengths and weaknesses relative to the large number of samples that must be processed during this project. Additional methods developed for detecting STECs during this project are: (1) a novel shigatoxin (stx) capture method and detection of stx with anti-stx monoclonal antibody; (2) a 12 antisera immunochemical screening assay for detecting strain differences among multiple stx-positive isolates; this assay identifies rapidly isolates to retain for further analysis. Our method of non-selective enrichment of fecal/water/plant/soil samples, stx-PCR of enrichment sample, plating on selective media, picking of multiple suspect colonies, and immunochemical and PCR analysis of isolates has resulted in isolation of >200 strains of STEC to date. An important objective of this project is to develop methods for isolation and characterization of STECs for purposes of measuring incidence in the same environments as assessed for O157:H7 incidence, and to provide another opportunity for source tracking pathogens in the produce production region. STEC isolates will be analyzed by PFGE, but also a second method developed from next generation sequence data. Genotyping of strains isolated from watersheds, wildlife and livestock indicates transport of pathogens may be occurring during heavy rain events, through roaming wildlife and indicates persistent incidence of pathogen in specific locations. Sampling and testing will continue for 12-18 months.