Molecular approaches to determine the bacterial diversity of selected water samples

Abstract

Water contains many gastro-intestinal pathogens that could cause infectious diseases. The quality of water is usually measured by making use of various culturedependent methods. These culture-dependent methods are not always accurate as most bacteria are considered viable but not culturable. Molecular based methods and sequencing technologies are able to by-pass the culturing steps, thus being able to detect most bacteria in for example a water sample. The aim of the project was to test various molecular approaches for their suitability to analyse the species abundance of bacterial population in surface water samples, other aquatic environments like the South African hot water springs, and water that has been treated with solar disinfection. Accordingly, various combinations of a number of molecular techniques, including DGGE, pyrosequencing and PhyloChip analysis were tested for their ability to detect the various bacterial populations that are present in various water samples. The ability of the methods to provide information at different taxonomic levels was also investigated. Results showed that all three molecular techniques were able to detect the dominant Phyla in the water samples. The DGGE results were not able to classify the sequences any further than the Phylum level. Pyrosequencing was able to classify the resulting sequences successfully up to Family level. In most cases sequences obtained from the V4-7 region could be used to classify individuals to a lower taxonomic rank. However, even though region V4-7 was able to group more sequences to a lower taxonomic rank, both variable regions V1-3 and V4-7 gave a similar sample distribution. In other words, the same classification was found with both regions but region V4-7 was able to identify more of the same grouping. The PhyloChip was able to reveal the species that were present or absent but was not a good indication of abundance. Both pyrosequencing and PhyloChip analysis was able to reveal a comprehensive bacterial population, however, pyrosequencing was unable to detect bacterial populations that were present in the minority, like the typical water safety indicators, for example, E. coli and other coliforms. PhyloChip analysis was able to detect these pathogens but could only indicate presence / absence. The bacterial population detected by PhyloChip analysis correlated well with the bacterial population that the pyrosequencing results were able to detect. The DGGE technique was able to detect the most dominant Phyla that was also detected by the pyrosequencing and PhyloChip results, but was unable to accurately classify much further. These techniques were also able to identify the bacterial populations that were associated with the South African hot water springs, and although it could identify the populations in solar disinfected water, it was unable to provide information on the exact changes that occurred during treatment as DNA from nonviable cells may still have been amplified during the PCR reaction in spite of PMA treatment. Results from the SODIS experiments also showed that the environmental E. coli isolate’s die-off pattern was slower than that of the K12 isolate even though the initial levels of bacteria were fairly similar. To conclude, these techniques provide detailed information on the overall bacterial diversity of aquatic samples but would at present not be able to replace the more conventional culture-based techniques when analyzing the safety of water for human consumption.