Mitochondrial Multiplex Real-Time PCR as a Source Tracking Method in Fecal-Contaminated Effluents JANE M. CALDWELL,* MORGAN E. RALEY, AND JAY F. LEVINE Department of Population Health and Pathobiology, College of Veterinary Medicine, North Carolina State University, 4700 Hillsborough Street, Raleigh, North Carolina 27606 Multiplex real-time PCR amplifying fecal mitochondrial DNA (mtDNA) combined with rapid, crude DNA preparations are promising additions to surface water source tracking methods. Amplification of eukaryotic mitochondrial DNA identifies the fecal source directly and can be used in conjunction with other intestinal microbial methods to characterize effluents. Species-specific primers and dual- labeled probes for human, swine, and bovine NADH dehydrogenase subunit 5 (ND5) genes were created for multiplex real-time PCR in feces and effluent slurries. The linear range of the multiplex assay was 10 2 -10 7 mtDNA copies for human, bovine, and swine effluent in combination (equal volumes). PCR amplification efficiencies for bovine, human, and swine mtDNA when assayed in combination were 93, 107, and 92% respectively. Linear regression correlation coefficients (r 2 ) were 0.99 for all standard curves except for human mtDNA in combination (r 2 ) 0.95). Multiplex amplification of bovine, human, and swine mtDNA (ND5) exhibited no cross-reactions between the effluents from three species of interest. Also, no cross- reactions were observed with effluents of other verte- brates: sheep, goat, horse, dog, cat, Canada goose, broiler, layer, turkey, and tilapia. Performed as a blind test, the PCR operator was able to correctly identify all but two effluent challenge samples (10/12 or 83% correct) with no false positives (22/22 or 100% correct). The multiplex assay had a tendency to detect the species of highest mtDNA concentration only. Better detection of all three species in a combination of human, bovine, and swine effluents was accomplished by running each real-time PCR primer/ probe set singly. Real-time PCR detection limit was calculated as 2.0 × 10 6 mitochondrial copies or 0.2 g of human feces per 100 mL effluent. Some carry-over mtDNA PCR signal from consumed beef, but not pork, was found in feces of human volunteers. 1. Introduction Contamination with enteric pathogens from animal and human feces limits the use of surface waters for recreation, commercial shellfish harvesting, and the irrigation of agri- cultural crops for human consumption. The potential sources of fecal contamination are numerous and reflect the breadth of mammalian activity on the planet. Spills from municipal and residential waste handling, runoff from fields used for spreading livestock waste, livestock waste lagoon leaks, feedlot runoff, and wildlife contamination add to the total fecal load in surface waters. Fecal pollution of surface waters is considered a non- point source problem. However, mitigating the effects of fecal waste contamination in surface waters requires identification of its source. Source tracking methods have traditionally used indicator organisms (IO) such as thermotolerant coliforms, Escherichia coli and Enterococcus spp., to detect fecal contaminants. However, fecal coliforms such as E. coli are not diagnostic of particular animal species and are not good indicators of viruses and other pathogens that may be present in surface waters (1-3). Finding host-specific microorganisms has been the holy grail of microbial source tracking (MST). Bifidobacterium adolescentis (4, 5), Bifidobacterium dentium (6), human adenovirus and human enterovirus (7), human polyomavirus (8), Enterococcus spp (9-13), and enterococcal surface protein from Enterococcus faecium (14) have been used as indicators of human fecal pollution. Teschovirus has been used as an indicator of porcine fecal contamination (15) and bovine enteric virus as indicator of bovine fecal contamination (3). Antibiotic resistance of various indicator organisms for source tracking has also been examined (16-18) but requires the generation of an extensive library of isolates. Many expensive and time-consuming nucleic acid- based molecular methods for source tracking have been tested, including ribotyping, length heterogeneity PCR, terminal-restriction fragment length polymorphism, repeti- tive PCR, pulsed-field gel electrophoresis, and amplified fragment length polymorphism (19). However, no single indicator organism or molecular test works for all eukaryotic species. Biochemical methods evaluated for source tracking include fecal sterols (20), stable isotopic ratios (21), and whitening agents (22). Blanch and co-workers (23) suggested that no single parameter is able to discriminate between human and non-human sources of fecal pollution and recommended a “basket” of four or five parameters including microbiological, phage, and chemical assays. A combination of microbial and chemical indicators has been suggested (24) and other researchers have recommended identifying new genomic targets and quantification methods for mi- crobial source tracking (25). Mitochondrial DNA (mtDNA) in conjunction with PCR or real-time PCR is used extensively in the fields of phylogenetics (26, 27), forensics (28, 29), and medicine (30). Martellini and co-workers (31) first used mtDNA with conventional and nested PCR to differentiate between human, bovine, porcine, and ovine sources in surface water source tracking. The advantages of targeting mtDNA as a source tracking tool are substantial. Polymerase chain reaction of mitochondrial DNA can be used to identify the animal species directly rather than microbial species it may host. Feces contain large amounts of exfoliated epithelial cells (32) and mtDNA has many copies per cell (33, 34). Healthy human peripheral blood mononuclear cells were reported to have 1000 (35) and normal human endometrial cells have 158-2625 mtDNA copies/cell (36). Therefore, mtDNA genes give robust PCR signals similar to 16S rRNA genes. Detectable DNA persists even after cellular death. Martellini et al. (31) were able to detect human DNA for up to 15 days after addition of mtDNA to a water source. Since we are detecting host DNA, debates about indicator organism host relevance, cultivability, and viability are moot. * Corresponding author phone: 919-513-6368; fax: 919-513-6464; e-mail: jane_caldwell@ncsu.edu. Environ. Sci. Technol. 2007, 41, 3277-3283 10.1021/es062912s CCC: $37.00  2007 American Chemical Society VOL. 41, NO. 9, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 3277 Published on Web 03/31/2007