Source: PENNSYLVANIA STATE UNIVERSITY submitted to
DEVELOPMENT OF A NOVEL MULTILOCUS SEQUENCE TYPING (MLST) SUBTYPING STRATEGY FOR TRACKING THE FARM TO FORK TRANSMISSION OF ESCHERICHIA COLI O157:H7
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
EXTENDED
Funding Source
Reporting Frequency
Annual
Accession No.
0220923
Grant No.
2010-65201-20619
Project No.
PEN04375
Proposal No.
2009-03611
Multistate No.
(N/A)
Program Code
93231
Project Start Date
Jul 1, 2010
Project End Date
Jun 30, 2014
Grant Year
2010
Project Director
Dudley, E. G.
Recipient Organization
PENNSYLVANIA STATE UNIVERSITY
208 MUELLER LABORATORY
UNIVERSITY PARK,PA 16802
Performing Department
Food Science
Non Technical Summary
Escherichia coli O157:H7 is foodborne pathogen that is responsible for an estimated 73,000 illnesses each year within the United States. This organism causes diarrhea that may become bloody, and approximately 5-15% of cases progress into a life-threatening condition called haemolytic uremic syndrome. E. coli O157:H7 can be isolated from the intestinal tract of cattle, and is transmitted to humans when contaminated fecal material or cattle hides come into contact with food materials. Most foodborne infections are the result of contaminated ground beef, however a growing number of outbreaks are linked to fresh produce such as lettuce and spinach. As part of an epidemiologic investigation, DNA-based methods are used to track the spread of specific strains of E. coli O157:H7 from farm-to-fork. Only once the route(s) of transmission are defined can effective intervention strategies be implemented to stop this spread. For E. coli O157:H7, this is currently done using a technique called pulsed field gel electrophoresis (PFGE). While PFGE has proven effective this method suffers from many limitations including the length of the protocol, technical expertise needed, and the ambiguous data that sometimes results. Techniques that differentiate between strains of E. coli O157:H7 based upon differences in DNA sequence, such as multilocus sequence typing (MLST), have potential to be more rapid and cost effective. The development of such methods has suffered due to the large similarity found within the DNA of different E. coli O157:H7 strains. We hypothesize that sequencing the DNA of a large number of E. coli O157:H7 strains will identify genomic targets for discriminating between different strains and lead to the development of a tracking method that can complement or even replace PFGE. Once we have developed an MLST approach, we will test its effectiveness in microbial source tracking within beef processing facilities. This research therefore has the potential to improve the safety of our food supply by improving our ability to rapidly trace the source of contamination and to apply control points to break this chain of transmission.
Animal Health Component
20%
Research Effort Categories
Basic
80%
Applied
20%
Developmental
(N/A)
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
7124010110090%
7123320110010%
Goals / Objectives
The pathogen Escherichia coli O157:H7 is responsible for a large number of infections yearly in the United States and other countries, which sometimes result in life-threatening diseases such as hemolytic uremic syndrome. The reservoir of E. coli O157:H7 is the gastrointestinal tract of cattle, and while many routes of transmissions from animal to humans are reported, the majority of human infections are foodborne, and ground beef represents the major food vehicle. Molecular subtyping methods are DNA-based approaches for tracking the spread of foodborne pathogens from farm to fork. Only once the route(s) of transmission are defined can effective intervention strategies be implemented to stop this spread. For E. coli O157:H7, this is currently done using pulsed field gel electrophoresis (PFGE). However this method suffers from many limitations including the length of the protocol, technical expertise needed, and the ambiguous data that sometimes results. Multilocus sequence typing (MLST) is a more rapid, cost effective method that has been applied to other pathogens, but has not been effectively developed for O157:H7 because of the high sequence identity found among genomes of different strains. We hypothesize that high-throughput sequencing of a large number of E. coli O157:H7 strains will identify genomic targets for discriminating between different strains, and lead to the development of a tracking method that is superior to PFGE. This research project has three objectives: 1) Sequence the genomes of 24 Escherichia coli O157:H7 strains using low cost, next-generation technology. 2) Identify sequence variations within these genomes that effectively distinguish between all current and previously sequenced strains, and develop an MLST subtyping approach for strain discrimination. 3) Test the effectiveness of the MLST method to track and control the spread of E. coli O157:H7 strains within meat-processing facilities. The goal of this research is the development of an MLST scheme for E. coli O157:H7 that would be useful for tracking the source of contamination during epidemiologic investigations. Additionally, the data generated will provide a great deal of information concerning single nucleotide polymorphisms (SNPs) that exist between strains of E. coli O157:H7. Identifying SNPs and/or other genomic regions that can be used as markers of highly pathogenic strains would provide risk assessment data concerning the potential pathogenicity of E. coli O157:H7 isolate(s) found within the food system.
Project Methods
Twenty-four strains of E. coli O157:H7 that were previously characterized as coming from distinct phylogenetic groups, as well as strains that cannot be distinguished using current DNA sequence-based typing methods, will be shotgun sequenced using SOLiD sequencing. The sequences will be uploaded into the Galaxy system, which will be used for data storage and for quality analysis of sequence reads. The Mapping and Assembly with Quality tool (MAQ), which is integrated into the Galaxy system, will be used to align shotgun sequences from each strain to a reference genome (E. coli O157:H7 strain Sakai), and to identify SNPs between the two genomes. Additionally, the program LASTZ will be used to identify SNPs between E. coli strain Sakai and other previously sequenced strains deposited within GenBank. This data will be imported into Microsoft Excel, and we will identify approximately 50 genomic regions that are conserved among all strains, and show the greatest sequence variation between strains. PCR primers will be designed that amplify these targeted regions, and these genomic loci will be amplified and sequenced from a collection of 48 E. coli O157:H7 strains obtained from the E. coli Reference Center at Penn State. We will identify no more than 15 loci that will permit the greatest discrimination between strains. This newly developed MLST method will next be used to determine whether it is able to effectively type strains of O157:H7 obtained from beef processing plants in Pennsylvania. Strains obtained from plant surveys will be typed using MLST, as well as PFGE. Discriminatory power will be calculated to determine if this MLST method is as effective as PFGE at E. coli O157:H7 source tracking within processing facilities.

Progress 07/01/12 to 06/30/13

Outputs
Target Audience: Academic scientists and industry personnel interested in current methods of differentiating strains of pathogenic bacteria (particularly E. coli O157:H7). Individuals working in such areas within state and federal public health departments. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided? In addition to the research progress, the funds from this project have been used to support the stipends of one M.S. and one Ph.D student in the Department of Food Science; the former student is currently pursuing a Ph.D at Iowa State University. This grant has also supported research projects of two undergraduate students in PI Dudley’s laboratory. One of these students had a first-author publication (Hartzell et al., 2011), finished an M.S. degree in Food Science at the University of Nebraska-Lincoln in 2012 and is now employed at Nestle in Solon, OH. The second student (C. Lewis) was a secondary author on two publications (Hartzell et al., 2011; Chen et al., 2013), and is currently pursuing a Ph.D in Cell and Developmental Biology at Penn State University. How have the results been disseminated to communities of interest? Through peer-reviewed publications, and through posters at scientific meetings. What do you plan to do during the next reporting period to accomplish the goals? The three main goals remaining are: 1) to complete the analysis of all E. coli O157:H7 genomes; 2) to complete validation of SNPs identified in genome sequences and test novel DNA-based subtyping method on clinical isolates and beef production facility isolates of E. coli O157:H7 and; 3) submission of the above for publication.

Impacts
What was accomplished under these goals? During the past year we have begun analyzing in detail the genomes of over 225 sequenced isolates of E. coli O157:H7, and have extensively studied the utility of genomic loci called Clustered Regularly Interspaced Short Palindromic Repeats as subtyping markers. Much of this work has been published in the peer-reviewed literature this year (see reference list). As part of the genome analyses performed to date, we have made two novel discoveries. First, we have discovered that genomic loci called CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) could be used for the development of molecular methods of identifying Shiga toxin-producing E. coli (O157:H7 and other serogroups such as O26, O45, O103, O111, O121, and O145). This manuscript was accepted recently in Applied and Environmental Microbiology (see Yin et al reference). Secondly, we discovered spontaneous deletions that occur in the genome of E. coli O157:H7 that may affect the accuracy of pulsed field gel electrophoresis, the current gold-standard for following the transmission of this pathogen during foodborne outbreaks (Chen et al., 2013). Therefore, although Specific Aim 2 is still ongoing, we have already made significant advances in our knowledge of E. coli O157:H7 molecular subtyping that would be useful for source-tracking this pathogen during epidemiologic investigations. We anticipate concluding this project next year with a thorough analysis of the genomes The goal of Aim 3 was to isolate E. coli O157:H7 from small and very-small beef production facilities, and to test the ability of the method developed in Specific Aim 2 to classify and identify potential sources of these isolates. We completed a large screen of such facilities in Pennsylvania (See Svoboda et al. reference). In this study, environmental swabs, carcass swabs, hide swabs, fecal samples, and ground beef from seven small and very small beef processing plants in Pennsylvania and one in New Jersey were obtained from October 2010 to December 2011. A multiplex polymerase chain reaction assay was used to determine the presence of Shiga toxin-producing E. coli (STEC) O-groups: O157, O145, O121, O113, O111, O103, O45, and O26 in the samples. Results demonstrated that 56.6% (154/272) of the environmental samples, 35.0% (71/203) of the carcass samples, 85.2% (23/27) of the hide samples, 37.5% (12/32) of the fecal, and 17.0% (20/118) of the ground beef samples tested positive for one or more of the serogroups. However, only 0.07% (20/272) of the environmental samples, 0.04% (9/203) of the carcass samples, 0% (0/27) of the hide samples, 0% (0/32), and 0% (0/118) ground beef samples tested positive for both the serogroup and Shiga toxin genes. Based on this survey, small and very small beef processors may be a source of the 6 non-O157:H7 STEC. This information would be of interest to regulatory officials, researchers, public health personnel, and the beef industry. However, to date we have not isolated sufficient numbers of E. coli O157:H7 to effectively test the method developed in Specific Aim 2. We continue to collect and characterize isolates as part of other funded studies, and will also obtain recent clinical isolates from our collaborators at the PA Department of Health to accurately quantify the effectiveness of our DNA-based subtyping method. It has become increasingly apparent over the past several years that there is tremendous genetic variation among strains of E. coli O157:H7, which impacts the ecology and virulence potential of the organism. This project represents the largest publically available analysis of this pathogen to date (all genome sequences have, or are in the process, of being released in the DNA sequence database called GenBank). This project is defining in detail the genetic diversity of O157:H7 which is critical especially as full genome sequencing of pathogens is becoming a powerful molecular epidemiology tool for outbreak investigations. This project is also one of the first surveys of Shiga toxin-producing E. coli in very small beef production facilities, which are increasing in numbers. A greater understanding of the ecology of these pathogens in such facilities is crucial for devising critical control points during slaughter.

Publications

  • Type: Journal Articles Status: Accepted Year Published: 2013 Citation: Yin, S., M. Jensen, J. Bai, C. DebRoy, R. Barrangou, and E. G. Dudley. Evolutionary divergence of Shiga toxin-producing Escherichia coli is reflected in CRISPR spacer composition. Accepted in Appl. Environ. Microbiol.
  • Type: Journal Articles Status: Published Year Published: 2013 Citation: Chen, C., C. A. Blumentritt, M. M. Curtis, V. Sperandio, A. G. Torres, and E. G. Dudley. Restrictive streptomycin-resistant mutations decrease the formation of attaching and effacing lesions in Escherichia coli O157:H7 strains. Antimicrob. Agents Chemother. 57:4260-4266.
  • Type: Journal Articles Status: Accepted Year Published: 2013 Citation: Svoboda, A. L., E. G. Dudley, C. DebRoy, E. W. Mills, and C. N. Cutter. Presence of Shiga toxin-producing Escherichia coli O-groups in small and very small beef processing plants and resulting ground beef detected by a multiplex polymerase chain reaction assay. Accepted in Foodborne Pathogens and Disease
  • Type: Journal Articles Status: Published Year Published: 2013 Citation: Eppinger, M., S. Daugherty, S. Agrawai, K. Galens, N. Sengamalay, L. Sadzewicz, L. Tallon, T. A. Cebula, M. K. Mammel, P. Feng, R. Soderlund, P. I. Tarr, C. DebRoy, E. G. Dudley, C. M. Fraser, J. Ravel. 2013. Whole-genome draft sequences of 26 enterohemorrhagic Escherichia coli O157:H7 strains. Genom. Announc. 1:e00134-12.
  • Type: Journal Articles Status: Published Year Published: 2013 Citation: Chen C., C. R. Lewis, K. Goswami, E. L. Roberts, C. Debroy, and E. G. Dudley. 2013. Identification and Characterization of Spontaneous Deletions within the Sp11-Sp12 Prophage Region of Escherichia coli O157:H7 Sakai. Appl. Environ. Microbiol. 79:1934-1941.


Progress 07/01/11 to 06/30/12

Outputs
OUTPUTS: This past year saw the continued sequencing of additional genomic sequences from E. coli O157:H7 and more focused work on the Shiga toxin-encoding phage, the development of new identification methods, and the conclusion of the survey of the small (10-500 employees) and very small (< 10 employees or total annual revenue of <$2.5 million) beef plants in Pennsylvania. The full genome sequencing of E. coli O157:H7 strains is being done in collaboration with Dr. Mark Eppinger (University of Maryland Genome Science Institute) and other investigators from Penn State, Washington University, and the FDA. Approximately 240 strains, are being sequenced by 454 and/or Illumina technology, and we expect all to be completed by this Fall. Of interest, we have sequenced ATCC strain 700728, which is the toxin-negative strain commonly used by researchers who perform field experiments, such as inoculation of fresh produce, to determine the survival of E. coli O157:H7 in the environment. In collaboration with Dr. Maria Marco (Univ. California-Davis) we will determine the relatedness of this strain to those common to human clinical cases. Additionally, we have been using the same sequencing technologies to generate genomes of Shiga toxin 2 (stx2)-encoding bacteriophage. The phage were induced by growing bacteria in the presence of ciprofloxacin, and the DNA from phage was harvested by standard protocols. We have additionally been using genomic data to develop methods of distinguishing strains based upon their DNA sequence, and this year targeted CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) as potential molecular subtyping markers. We sequenced these regions from >40 E. coli O157:H7 strains and additionally from >100 non-O157:H7 strains of the "Big Six" serotypes (O26, O111, O103, O121, O45, and O145). Lastly, we investigated whether small and very small beef processing plants are a potential source of STEC, utilizing the enrichment and detection methods optimized in the first objective. In this survey, environmental swabs, carcass swabs, hide swabs, fecal samples, and ground beef from small and very small beef processing plants were obtained to determine the presence of STEC Results of this research have been disseminated through poster presentations at the 2012 annual meetings of the American Society for Microbiology and International Association of Food Protection. All full genome sequences generated to data have been deposited within GenBank. This data will also be submitted for publication in peer-reviewed jourals. PARTICIPANTS: Dr. Edward G. Dudley, lead PI; Dr. Catherine Cutter, co-PI; Chun Chen, graduate student; Kakolie Goswami, graduate student; Shuang Yin, graduate student; Amanda Svoboda, graduate student; Dr. Mark Eppinger, collaborator (University of Maryland Institute of Genome Sciences). This grant has provided stipend and tuition support for the four graduate students listed above, as well as travel to scientific meetings. TARGET AUDIENCES: Academic and industry scientists interested in developing methods to differentiate and detect strains of Shiga toxin producing Escherichia coli, especially serotype O157:H7. PROJECT MODIFICATIONS: Nothing significant to report during this reporting period.

Impacts
Data from the genome sequencing will be analyzed in total once the final datasets are generated. The sequences from the stx2-encoding phage has revealed that there are at least three distinct variants with respect to their early replication and regulation genes. One of these phage was isolated from a strain previously characterized as a "clade 8", which is a group previously suggested to be associated with high human virulence. This phage exhibited a high infectivity to other E. coli strains, and in this process produces high levels of Shiga toxin. This information may help explain why certain clade 8 strains, when growing in the presence of intestinal E. coli, may produce higher levels of Shiga toxin than when they are grown as pure cultures. Our genome sequencing data has also directed us to analyze the CRISPR sequences found within E. coli O157:H7 as well as other Shiga toxin-producing E. coli (STEC) strains. We have successfully used these loci as molecular subtyping markers for Salmonella enterica. Our data suggests that these sequences do not vary significantly within isolates from a given serogroup, meaning that they may be useful signatures or even defining markers for each of the Big Six STEC. We are currently investigating this hypothesis further. A previously optimized multiplex PCR assay was used to detect the presence of STEC O-groups and the presence of Shiga toxin (stx) and intimin (eae) genes were determined using PCR primers described by Paton and Paton (1998). Results demonstrated that 55.5% (151/272) of the environmental samples, 36.9% (75/203) of the carcass samples, 85.2% (23/27) of the hide samples, 37.5% (12/32) of the fecal, and 18.6% (22/118) of the ground beef samples tested positive for one or more of the serogroups. However, only 7.7% (21/272) of the environmental samples, 5.9% (12/203) of the carcass samples, 0% (0/27) of the hide samples, 0% (0/32) of the fecal, and 0% (0/118) ground beef samples tested positive for stx1 and/or stx2 genes. In addition, 13.6% (37/272) of the environmental samples, 7.9% (16/203) of the carcass samples, 0% (0/27) of the hide samples, 0% (0/32) of the fecal samples, and 0.8% (1/118) of the ground beef samples tested positive for eae gene.

Publications

  • Svoboda, A., C. DebRoy, E. G. Dudley, E. Mills, and C. N. Cutter. 2012. Presence of Shiga-toxin producing Escherichia coli in small and very small beef processing plants and resulting beef products detected by a multiplex polymerase chain reaction assay. IAFP Program and Abstract Book Annual Meeting. Providence, Rhode Island. P3-16.
  • Chen, C., C. Lewis, K. Goswami, and E. G. Dudley. 2012. Spontaneous excisions within the Sp11-Sp12 prophage region of Escherichia coli O157:H7 Sakai. Presented at the 2012 International Association for Food Protection meeting. Providence, Rhode Island. P2-141.
  • Goswami, K., C. Chen, and E. G. Dudley. 2012. Identification of variable regions within genomes of Shiga toxin prophage from Escherichia coli O157:H7. IAFP Program and Abstract Book Annual Meeting. Providence, Rhode Island. P1-121.
  • Yin, S., C. DebRoy, and E. G. Dudley. 2012. CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) as a potential molecular subtyping marker for high-risk Shiga toxin-producing Escherichia coli (STEC) Isolates. IAFP Program and Abstract Book Annual Meeting. Providence, Rhode Island. P1-88.
  • Chen, C., K. Goswami, and E. G. Dudley. 2012. Investigating the role of the toxin-encoding bacteriophage and host background on the expression of Stx2 in Escherichia coli O157:H7. Abstract book of the 110th General Meeting of the American Society for Microbiology. P-1129.
  • Dudley, E. G., C. Chen, and K. Goswami. 2012. Investigating the role of the toxin-encoding bacteriophage and host background on the expression of Stx2 in Escherichia coli O157:H7. Abstract book of the 8th International Symposium on Shiga Toxin (Verotoxin) producing Escherichia coli Infections. Amsterdam, the Netherlands. P-099.


Progress 07/01/10 to 06/30/11

Outputs
OUTPUTS: This past year, we have extensively characterized a collection of clinical E. coli O157:H7 isolates for sequencing, are currently in the process of generating sequence information for these isolates, and have been surveying small (10-500 employees) and very small (< 10 employees or total annual revenue of <$2.5 million) beef plants in Pennsylvania for Shiga toxin producing E. coli. The clinical E. coli collection consisted of 52 isolates obtained from the PA Department of Health, and come from sporadic and outbreak cases during 2006-2008. We used a molecular subtyping technique called Lineage-Specific Polymorphism Assay (LSPA) to categorize isolates into defined lineages, determined the level of Shiga toxin produced, and screened these isolates for a variety of virulence factors including Stx1, Stx2, hlyA, eae, and katA. We additionally categorized isolates into Clades, using a previously described method from Thomas Whittam's laboratory. We determined from this study that isolates categorized as Clade 8 constituted approximately 1/3 of all E. coli O157:H7 isolates submitted to the PA Department of Health during this time frame, which is significant as it is currently believed that this clade defines an especially virulent group. From this and other work in the lab, we identified a collection of 25 isolates to sequence for identifying single nucleotide polymorphisms (SNP). These isolates have been submitted for sequencing on the Illumina platform, and two of these will additionally be sequenced using the 454 platform. As suggested by one reviewer of our proposal, we did experience difficulties obtaining quality data using SOLiD sequencing, and therefore have changed our approach to these more widely used methodologies. As most shotgun E. coli sequencing projects do not obtain quality sequence information for chromosomally inserted bacteriophage, we are additionally isolating ciprofloxacin-inducible phage from many of these isolates to sequence by 454. During the course of this work, we have also established a collaboration with Dr. Mark Eppinger at the University of Maryland Genomic Sciences Institute, who is sequencing >150 additional E. coli O157:H7 isolates as part of an NIH contract; data from both of these projects will be analyzed together to take full advantage of these large datasets. Lastly, 316 environment, carcass, hide, ground beef, and fecal samples from PA beef plants have been analyzed both by traditional culturing and by molecular methods. Results to date indicate 24, 5, 6, 8, 0, 8, 11 and 8 enrichments were PCR-positive for E. coli of serotypes O157, O145, O121, O113, O111, O103, O45, and O26 respectively. Additionally, 0 and 5 enrichments were PCR-positive for stx1 and stx2, respectively. We are currently screening isolates from agar plates to determine whether we can identify stx1/stx2 positive isolates that are also from one of the eight serogroups listed. We have also screened 90 enrichments from farmer's market and retail samples, and found that 4 and 1 retail samples were O157 and O121 positive, respectively, however no enrichments were positive for stx1 or stx2. PARTICIPANTS: Dr. Edward G. Dudley (PI), Dr. Catherine N. Cutter (co-PI), Chun Chen (Ph.D candidate), Kakolie Goswami (Ph.D candidate), Amanda Svoboda (M.S. candidate), Annette Hartzell (undergraduate student in Penn State's Honors program), Carrie Lewis (undergraduate student). This project has provided scientific training for multiple individuals. Annette Hartzell graduated in 2010, and is currently an M.S. student in the University of Nebraska-Lincoln Department of Food Science. Carrie Lewis is currently a Ph.D student in Penn State's Department of Cell and Developmental Biology. TARGET AUDIENCES: Industry, government, and basic researchers interested in approaches of tracking and controlling the spread of foodborne pathogens. PROJECT MODIFICATIONS: Nothing significant to report during this reporting period.

Impacts
Prior to this research, it was appreciated that Clade 8 strains of E. coli O157:H7 may exhibit differences in virulence levels, however the two published reports on the prevalence of these strains suggested either a low incidence, or a frequent and increasing distribution. This work supported the suggestion that these strains are common clinical isolates of E. coli O157:H7 in Pennsylvania. This suggests that we need to increase our knowledge and monitoring of these strains to determine whether this poses an increased threat to the US food supply. The sequencing data being generated builds on our knowledge of E. coli O157:H7 diversity, and will be applied in the near future to develop improved subtyping methods for this pathogen. The survey work reported here is providing insight into the prevalence of E. coli O157:H7 and non-O157 STEC in beef production facilities, which have received little attention to date. Our preliminary data suggests that these organisms are present at low levels in these environments. This data will be used to either validate that control points used by these plants are working, or to assist these producers identify improvements they can make in their HAACP plan.

Publications

  • Hartzell, A., C. Chen, C. Lewis, K. Liu, S. Reynolds, and E. G. Dudley. 2011. Escherichia coli O157:H7 of lineage-specific polymorphism assay 211111 and clade 8 are common clinical isolates within Pennsylvania. Foodborne Path. Dis. 8:763-8.