Profiling Total Viable Bacteria in a Hemodialysis Water Treatment System

Culture-dependent methods, such as heterotrophic plate counting (HPC), are usually applied to evaluate the bacteriological quality of hemodialysis water. However, these methods cannot detect the uncultured or viable but non-culturable (VBNC) bacteria, both of which may be quantitatively predominant throughout the hemodialysis water treatment system. Therefore, propidium monoazide (PMA)-qPCR associated with HPC was used together to profile the distribution of the total viable bacteria in such a system. Moreover, high-throughput sequencing of 16S rRNA gene amplicons was utilized to analyze the microbial community structure and diversity. The HPC results indicated that the total bacterial counts conformed to the standards, yet the bacteria amounts were abruptly enhanced after carbon filter treatment. Nevertheless, the bacterial counts detected by PMA-qPCR, with the highest levels of $2.14{\times}10^7copies/100ml$ in softener water, were much higher than the corresponding HPC results, which demonstrated the occurrence of numerous uncultured or VBNC bacteria among the entire system before reverse osmosis (RO). In addition, the microbial community structure was very different and the diversity was enhanced after the carbon filter. Although the diversity was minimized after RO treatment, pathogens such as Escherichia could still be detected in the RO effluent. In general, both the amounts of bacteria and the complexity of microbial community in the hemodialysis water treatment system revealed by molecular approaches were much higher than by traditional method. These results suggested the higher health risk potential for hemodialysis patients from the up-to-standard water. The treatment process could also be optimized, based on the results of this study.