Source: PURDUE UNIVERSITY submitted to
WATER QUALITY: DEVELOPMENT OF GENOME SCAN AND OPTICAL SCATTERING IMAGING TECHNOLOGIES FOR MICROBIAL SOURCE TRACKING
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
NEW
Funding Source
Reporting Frequency
Annual
Accession No.
0219122
Grant No.
2009-34634-19916
Project No.
INDW-2009-04560
Proposal No.
2009-04560
Multistate No.
(N/A)
Program Code
DD-M
Project Start Date
Aug 15, 2009
Project End Date
Aug 14, 2012
Grant Year
2009
Project Director
Tseng, C.
Recipient Organization
PURDUE UNIVERSITY
(N/A)
WEST LAFAYETTE,IN 47907
Performing Department
Department of Biological Sciences
Non Technical Summary
High resolution microbial strain typing is vitally important for recognizing outbreaks of infection and essential for identifying microbial contamination in water which is a worldwide environmental and health problem. Despite recent scientific efforts, no current method meets the high stringent requirement for microbial source tracking. This project proposes to develop two innovative technologies for biomedical and environmental applications, with the specific aim of improving water quality. These include 1) the genomic scan technology and 2) the optical scattering imaging technology. E. coli, a widely used indicator for fecal contamination, will be used as a model system to develop these technologies. The genome scan technology is to generate host specific DNA markers for differentiating E. coli from different host species so that the source of fecal bacterial contamination of water can be identified. The optical scattering imaging technology utilizes laser scattering image patterns of bacterial colonies for distinguishing the source of bacterial contamination. In addition to the wild type E. coli, pathogenic E. coli O157:H7 will be subtyped. The two new technologies will be developed independently. The rationale for developing two technologies simultaneously is that both appear promising as an accurate and simple method for stringent bacterial identification. Data obtained from these technologies will be useful for confirming each other?s results. These two technologies will be compared and critically evaluated in terms of accuracy, reproducibility, simplicity, speed, equipment requirement, and cost. The overall rating will eventually be used to select one as a main technology and the other as a supplement with the aim of technology transfer to government regulatory agencies and private lab for microbial source tracking.
Animal Health Component
(N/A)
Research Effort Categories
Basic
30%
Applied
40%
Developmental
30%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
1330320104030%
1334010110030%
1330320202040%
Goals / Objectives
Microbial contamination of water is a worldwide environmental and health problem. No current technologies including DNA typing methods meet the resolution requirements for stringent strain typing; nor are current methods ideal for microbial source tracking. We propose to use E. coli, an indicator for fecal contamination, as a model system to develop two innovative technologies, one on genomic scan technology and the other on optical scattering imaging technology, for microbial source tracking. The genome scan technology is to generate host specific DNA markers for differentiating E. coli from different host species so that the source of fecal bacterial contamination of water can be identified. The optical scattering imaging technology utilizes laser scattering image patterns of bacterial colonies for distinguishing the source of bacterial contamination. In addition to the wild type E. coli, pathogenic E. coli O157:H7 will be subtyped. The outputs of this investigation are to generate data for comparing and evaluating the advantages and disadvantages of the two technologies in terms of accuracy, reproducibility, simplicity, speed, equipment requirements, and cost for final recommendations to the regulatory agencies and private laboratories for possible adoption.
Project Methods
The genome scan technology permits a comprehensive examination of entire E. coli DNA contents through sequencing and analyzing the genome-derived short fragments, each of which is represented by two tags (ditags) at both ends. The information is regained and genomic variation detected after mapping the ditags back to the reference sequence database. Genomic DNAs from the same host will be pooled to form a mixed genomic library as a tester. Comparison of the experimental ditags from the tester libraries to those from the reference databases and to each other will provide information for selecting candidates of host-specific markers. The candidate DNA markers will be used for mass screening of E. coli genomes of numerous isolates form known sources. The confirmed markers will be used as host-specific PCR primers, from which multiplex PCR will be developed for E. coli source tracking. The optical scattering imaging technology to be developed emphasizes the use of widely tunable laser sources to capture distinguishable patterns of bacterial colonies. The broad band laser spectrum for multiple-scale feature extraction will be used to provide a high degree of new formation. As a complete framework for MST, new and improved techniques in image pre-processing, feature extraction and classification for the E. coli identification will also be included. The hyperspectral characteristics or the spectral signature can be represented by the feature vectors decomposed by the wavelet transformation. A neural network will be implemented to classify the image features presented by wavelet coefficients. Two variations of multi-category support vector machine (M-SVM) as the main classifiers will be implemented to evaluate the performance in bacterial colony identification.

Progress 08/15/10 to 08/14/11

Outputs
OUTPUTS: Outputs completed during the reporting period: Part I. Identification of host specific E. coli DNA markers for microbial source tracking (MST). 1. E. coli genomic DNA prepared and pooled: E. coli genomic DNA from human, cow, horse, pig, chicken, seagull, and gooses were prepared and pooled. 2. Mass genomic sequencing based on 75 base long DNA fragments was carried out using the Illumina Genome Analysis II technology. 3. E. coli stain 536 was used as a reference for mapping and identifying SNPs. 4. All 75 base long DNA fragments were matched to the reference E. coli genome to identify host specific SNPs. 5. Host specific PCR primers were designed based on SNPs. Foward and reverse primers of 200 base long were designed for validity testing using electrophoresis. Part II. Development of a laser technology for microbial source tracking. 1. A red laser system were developed for distinguishing E. coli from varius host species. 2. A database of laser imaging pattern for E. coli colonies from different host species have been established. 3. To increase the accuracy and applicability of the laser imaging system, a multispectral laser imaging technology is being contucted. 4. Current efforts are being made to develop an automated system capable of 1) directing laser beams to the precise locations of E. coli colonies for image generation and 2) detecting and capturing the laser image with high speed so that a large amount of data can be registered. 5. New Algarithms are being developed to process the multispectral imaging data for distinguishing E. coli from various host species. PARTICIPANTS: Charles C. Tseng, PI, Professor of Biology, Purdue University Clumet, is in charge of the overall project. Bin Chen, CoPI, Assistant Professor of Electrical and Computer Engineering, Purdue University Calumet, is charge of the laser imaging technology. Hao Gong, was a graduate student in the Electrical and Computer Engineering Department, Purdue University Calumet. He assisted in the laser imaging work. PRithviraj P. Patankar, was a graduate student in the Biological Science Department, Purdue University Calumet. He assisted in the PCR work. Xu Zhang, is a rsearch scientist in the Biological Science Department, Purdue University Calumet. He is responsible for E. coli culture and DNA preparation research. Michael J. Sadowsky is a professor of microbiology, University of Minnesota. He is a consultant and provided E. coli sample. TARGET AUDIENCES: Several presentations have been given to Lake Michigan beach managers and related organizations for our innovative technology for microbial source tracking. Hundreds of E. coli samples have been received from several beaches in Indiana, Michigan, and Illiniois for identifying the sources. PROJECT MODIFICATIONS: A automated multispectral laser imaging technology is currently being developed. This technology can be used for MST for both water and food.

Impacts
Outcomes of the project during the reporting period are as follows: Part I. Identification of host specific E. coli DNA markers for microbial source tracking (MST). Data for sequences and mapping resutls have been completed. For the E. coli genome from each host species, over 28 million sequences were obtained with the size betwseen 5 and 6 GB. All of the mapping positions and SNPs detected were recorded. These data were further grouped according host species. Human E. coli genome has a larger SNP than other host species. Over 100 primer pairs per host species have been obtained and are currently being tested for host specific DNA markers. Part II. Development of a laser technology for microbial source tracking. E. coli isolates of 10 difference host individuals each from cow, goose, horse, human and seagull were grown to form colonies which were then imaged. The size of E. coli colonies was approximately 0.8 mm for laser imaging to achieve consistent scatter patterns. Approximately 30 pictures were taken for each individual. Approximately 86.3% of images were correctly assigned to host species, i.e., human, horse, cow, seagull or goose. The classifier also shows very promising potentials for individual host identification. It is expected that when the multispectral system is fully developed and operational, the accuracy will be greatly increased.

Publications

  • Bin Chen, Hao Gong, Xu Zhang and Prithviraj P. Patankar, Michael J. Sadowsky, Charles C. Tseng. Laser imaging for rapid Microbial Source Tracking. Int. J. Computational Biology and Drug Design 3(3):177-186, 2010


Progress 08/15/09 to 08/14/10

Outputs
OUTPUTS: Two new technologies for microbial source tracking (MST) are being developed: Part l: Mass genome sequencing based technology for generating host specific E. coli DNA markers. Work accomplished include a)genomic DNA preparation, b)genomic sequencing and mapping, c)detection of host specific SNPs, d) groupings of E. coli SNPs according host species for a genomic variation study, and e)SNP PCR primer design and validity tests. In short, mass sequencing of E. coli genomes were carried out using Illumina Genome Analyzer II based on the Solexa technology. E. coli genomes with 75-base sequences were collected from eight sequencing lanes (2 lanes for human E. coli and 1 lane for E. coli from each of the other six host species). Mapping results to the E. coli reference genome were recorded and single nucleotide polymorphisms (SNPs)for E. coli from each host species were identified. SNP PCR primers were generated for a lengh of 200bp. These primers are potential host specific E. coli DNA markers. Part 2: Optical scattering imaging technology for differentiating host specific E. coli. Work accomplished include a)bacterial growth and colony preparation, b) use of the Rayleigh-Sommerfield diffraction formula for describing the laser scattering patterns of E. coli colonies,and c) image processing and analysis. E. coli isolates of 10 difference host individuals each from human, horse, cow, seagull, and goose were grown to form colonies which were then imaged. The size of E. coli colonies was approximately 0.8mm for laser imaging to achieve consistent scatter patterns. Approximately 30 pictures were taken for each individual. Preliminary tests of the applicability of these two technologies for MST have been undertaken. The results are promising (see Outcomes/Impact) below. PARTICIPANTS: Charles C. Tseng (PI)is responsible for the overall project. His major work includes the development of mass sequencing based technology to generate E. coli host specific DNA markers. He is also reponsible for preparation of E. coli colonies and work closely with Bin Chen who is reponsible for the laser scattering method. Bin Chen (CoPI) is responsible for developing laser scattering technology and analyzing the image patterns for differentiating different E. coli host species. Alex Xu Zhang (technical assistant) is reponsible for performing laboratory experiments and supervising several graduate students who work on this project. TARGET AUDIENCES: Not relevant to this project. PROJECT MODIFICATIONS: Not relevant to this project.

Impacts
Part 1. Mass sequencing based technology for generating E. coli host specific DNA markers: E. coli genomic sequences of 75-base long from seven different host species were collected using the second generation Illumina sequencer. All sequences ranging from 28 - 30 million reads per sample were mapped to the reference E. coli genome. Host specific SNP PCR primers identified include 2 for human, 3 for pig, 3 for horse, 4 for cow, 4 for chicken, and 4 for seagull, and 1 for goose. Part 2. Opttical scattering technology for generating host specific E. coli: The preliminary resuts (based on a limited database) show that the accuracy of using laser image patterns for identifying host specific E. coli are 88.4% for human, 82.3% for horse, 82.8% for cow, 89.6% for seagaul, and 88.7% for goose. It is It is expected that by the end of the project, we will have these two new technologies assessed in terms of their accuracy, reproducibility, simplicity, speed, equipment requirement, and cost.

Publications

  • No publications reported this period