Strategy to Evaluate Changes in Bacterial Community Profiles and Bacterial Pathogen Load Reduction After Sewage Disinfection

Abstract

Sewage effluent discharge is a major source of pathogenic contamination to the environment. The disinfection process is critical for the elimination of pathogens in sewage. In this study, we examined the impact of chlorine disinfection on the total, viable, and culturable populations of indicator bacteria, pathogens, and bacterial communities in two contrasting types of effluents (primarily treated saline and secondarily treated freshwater). Effluents collected bimonthly over 1 year were examined using cultivation, quantitative PCR (qPCR), and 16S rRNA gene amplicon sequencing coupled with or without propidium monoazide (PMA) treatment. The results showed that each type of effluent was characterized by a specific set of representative genera before disinfection. Salinity appeared to be the major abiotic factor associated with the differences in bacterial community compositions. The pathogen analysis pipeline revealed over 20 viable clinically important pathogenic species in the effluents. Although the bacterial communities differed markedly between the two types of effluents before disinfection, the species of pathogens persisting after disinfection were similar, many of them were members of Enterobacter and Vibrio. The relative abundances of all pathogens identified in the amplicon sequences were multiplied by the 16S rRNA gene copy numbers of total bacteria detected by PMA-qPCR to estimate their concentrations. Pathogens remained viable after disinfection reached 8 log₁₀ 16S rRNA copies ml^-1 effluent. Meanwhile, around 80 % of the populations of three indicator bacteria including Escherichia coli, Enterococcus, and Bacteroidales were viable after disinfection, but over 99 % of the viable E. coli and Enterococcus were in the non-culturable state. We estimated the total pathogen load by adding the concentrations of all viable pathogens and examined their correlations with indicator bacteria of different types, physiological states, and effluents. The results showed that the PMA-qPCR measurement of E. coli is a reliable proxy of bacterial pathogen loads in both types of effluents. The utility of viable indicator bacteria as a biological index to assess the overall bacteriological hazards in effluents is discussed.

Keywords: PMA treatment; amplicon sequencing; bacterial community; chlorination; effluent; indicator bacteria; pathogen; quantitative PCR.

Figure 1

Concentrations of E. coli, Enterococcus, Bacteroidales, and total bacteria before (Pre) and after (Post) chlorine disinfection in SC and ST. The samples collected bimonthly over a year were represented by individual data points, and the arithmetic means of the samples were indicated by horizontal lines. The concentrations of total, viable and culturable cells of (A) E. coli and (B) Enterococcus in each sewage treatment work were compared. The total and viable 23S rRNA gene copy numbers detected in PMA-qPCR assays were converted to numbers of cell equivalent by dividing the median numbers of gene copies per cell (log₁₀ cell equivalent ml⁻¹). The concentrations of culturable cells were determined using the plate counting method (log₁₀ CFU ml⁻¹). The concentrations of total and viable 16S rRNA gene copy numbers of (C) Bacteroidales and (D) total bacteria in each sewage treatment work (log₁₀ gene copies ml⁻¹) were shown. # Less than 1 colony was found per ml of effluent.

Figure 2

Alpha-diversity indices (number of sequence variants, Shannon index and Pielou's evenness index) measuring the richness, diversity, and evenness of total (No PMA) and viable (PMA) bacterial communities before and after chlorine disinfection. The samples collected bimonthly over a year were represented by individual data points with error bars (mean ± 1 S.D). All samples were rarefied to the minimum depth of filtered reads and the sequence variants were used for calculation.

Figure 3

Linear Discriminant Analysis Effect Size (LEfSe) plot of significantly discriminative genera identified in the viable populations of effluents in SC and ST (Kruskal–Wallis sum-rank test, q < 0.05, absolute log LDA score > 2). Absolute LDA score indicates the discriminatory power of the genus in the dedicated group. Genera with reported pathogenic species were marked with asterisks.

Figure 4

The average concentrations of total and viable pathogenic species before and after chlorine disinfection in each sewage treatment work over the year. Color scale represented the Z-scores in rows. Paired t-test for comparing the concentrations of species in all PMA-treated samples before and after chlorine disinfection: *p < 0.05, **p < 0.01, ***p< 0.001.

Figure 5

Pearson correlations between total pathogen loads and fecal indicator bacteria with different measurement strategies. The culturable plate counts, total and viable qPCR gene copies of E. coli, Enterococcus, and Bacteroidales were compared to the total pathogen loads (all viable pathogens) for all samples collected before and after chlorine disinfection throughout the year in each sewage treatment work.