Shift of human pathogen community composition and their potential human health risk after supply suspension in tap water

Scientific Reports volume 13, Article number: 12419 (2023) Cite this article

115 Accesses

1 Altmetric

Metrics details

Water supply suspension–restoration can occur frequently due to the overhauling of civil infrastructure in developing countries and the shutdown of commercial buildings during the pandemic. For comprehensive insights into the effects of water supply suspension–restoration, this study characterized the variations of the pathogen community composition of the tap water and their infection risk under different water supply scenarios. Metagenomic sequencing revealed a significant change of the human pathogen profiles, among which the most dominant pathogen changed from Pseudomonas aeruginosa (4.91%) to Acinetobacter johnsonii (0.59%). Furthermore, absolute quantification of pathogens by propidium-monoazide-qPCR revealed that the abundance of the three typical pathogens (Pseudomonas aeruginosa, Mycobacterium avium and Salmonella sp.) showed an increase of 2.44 log to 3.60 log immediately after water supply suspension–restoration and did not return to the normal level even after 2-h supply restoration, except for Pseudomonas aeruginosa. Quantitative microbial risk assessment suggested the infection risks of the three pathogens arising from direct utilization of tap water under stable water supply, including dermal exposure and oral intake, were all above the threshold of 10−4, and evidently increased after water supply suspension–restoration. This study warns us against the risk induced by the pathogens in tap water, especially after water supply suspension–restoration.

A variety of human microbial pathogens are present in drinking water and may be considered environmental contaminants that induce public health risk1. The urban drinking water distribution system (DWDS) is a transmission bridge between drinking water treatment plants and city dwellers2 and tap water (TW) is closely related to human living. Opportunistic pathogens are frequently detected in TW3, including Legionella pneumophila, Mycobacterium avium and Pseudomonas aeruginosa. Many countries or organizations have already included pathogens (such as Salmonella spp., Legionella spp. and Campylobacter jejuni) in drinking water safety guideline or standard4,5. Due to the large number of the population still living in water-scarce areas6, intermittent water supply is adopted in most low- and middle-income countries worldwide7,8, which is an important source of microbial infection in DWDS9. Besides, water supply suspension can also occur during the overhauling of civil infrastructure or the shutdown of commercial buildings during the pandemic10. During the water supply suspension, water quality undergoes serious deterioration and pathogenic bacteria can proliferate due to the consumption of residue chlorine11. What’s worse, the biofilm may detach from the pipes when the water flow in motion is suddenly forced to stop or start12, in which case the attached bacteria will enter the TW after the water supply restoration. Actually, the impact of pathogens in TW, especially after water supply suspension–restoration, on human health has become critical for risk control, which has not been well documented yet13.

For risk assessment, the detection methods of pathogens are of the essence14, both in terms of the coverage of diversity and the precision of absolute quantification. Traditional culture-dependent or molecular methods cannot cover all the pathogens in TW. Combined use of high throughput sequencing and metagenomic analysis is considered as a powerful tool to analyze the whole community of pathogens in the environments, including river15, sewage sludge16, human nasal cavity and feces17, which also serves as an effective and reliable approach for the comprehensive investigation of pathogens in clean environmental compartments, e.g. TW18. However, the relative abundance obtained from this method cannot be directly used for the assessment of the health risks induced by pathogens in TW, which relies on absolute quantification.

It is crucial to obtain the cell numbers of live pathogens in given volume of TW for precisely assessing their health risk with quantitative microbial risk assessment19 (QMRA), a method commonly employed in assessing health risk posed by waterborne pathogens20. Technically, quantitative real time polymerase chain reaction (qPCR) has been widely applied to absolutely quantify pathogens in water21,22, but cannot identify live or dead bacteria. Fortunately, propidium monoazide (PMA) is able to selectively combine with the naked DNA because of their nonpermeability for intact membrane cells among bacterial isolates and environmental samples23,24. After staining, DNA within dead cells and extracellular DNA would not be amplified during PCR due to the covalent link between dye and DNA molecule. PMA pretreatment of samples has been coupled with qPCR, namely PMA-qPCR, to trace live pathogens of concern in seawater25, agricultural produce26 and drinking water27.

This study aimed to investigate the composition, absolute abundance and health risk of pathogens in TW under stable water supply and suspended–restored water supply. Metagenome sequencing and bioinformatics analysis were conducted for comprehensive insights into the composition of the bacterial pathogens in TW. QMRA was employed to assess the health risks of the pathogens based on absolute quantification by PMA-qPCR. This study helps to extend our knowledge regarding the diversity and risks of pathogens in TW, which may facilitate the prevention and control of pathogen spread.

We detected a total of 339 pathogenic species (Table S1) in five TW sampling sites (Fig. 1A) under stable water (SW) supply in Nanjing with metagenomic sequencing. Eight pathogenic species within four different genera were identified to have high relative abundance (> 0.1% in the bacterial community) in SW samples (Fig. 1B). As the transmission distance increased, the relative abundances in the bacterial community of Pseudomonas aeruginosa (0.19–1.61%), Pseudomonas alcaligenes (0.24–4.04%), Sphingomonas paucimobilis (0.07–0.15%) and Mycobacterium avium (0.01–0.27%) increased gradually from JQM to XLW (Fig. 1B). Among all the sampling sites, site JHS showed to have the highest abundance of Pseudomonas stutzeri (0.29%), Pseudomonas fluorescens (0.20%), Pseudomonas putida (0.17%), Aeromonas hydrophila (0.04%) and Salmonella enterica (0.02%) (Fig. 1B and Fig. S1). However, only the relative abundance of P. aeruginosa increased and became the dominant pathogen (Fig. 1B) after re-chlorination tank treatment, from 1.61% in site XLW to 22.55% in site MQ. Moreover, the cluster analysis further illustrated that the structural similarity and alpha diversity of the pathogen community structure in SW decreased with increasing transport distance (Fig. 1B).

Geographic location of sampling sites, pathogen community composition and absolute abundance of the three typical pathogens in SW. (A) DWTP, re-chlorination tank, SW and suspension-restoration water (SRW) sampling locations are marked as black pentagram, black rhombus, red circle and blue triangle, respectively. (B) Relative abundance of abundant (mean value > 0.1%) pathogens at the different sites. Cluster analysis based on Bray–Curtis distance and Shannon index (mean ± standard error, n = 3) shows structural similarity and alpha diversity of pathogen community at the different sampling sites. (C) Absolute abundance (mean ± standard error, n = 10) of the three typical pathogens in SW at the different sites. *** means significant difference by t-test between sites XLW and MQ (p < 0.001).

Three typical pathogens including P. aeruginosa, M. avium and Salmonella sp. were chosen to detect their absolute abundance due to their prevalence and high relative abundance in SW samples. PMA-qPCR results showed that P. aeruginosa had the highest average concentration (7.40 × 103 cells/L) and detection frequency (96%, 48/50 samples), followed by M. avium (7.42 × 102 cells/L; 84%, 42/50 samples) and Salmonella sp. (27 cells/L; 36%, 18/50 samples). Furthermore, the absolute abundances of all the three pathogens increased generally from site JQM to site XLW and decreased from site XLW to site MQ, but an exception at site RHL for M. avium and Salmonella sp. (Fig. 1C). In particular, Salmonella sp. was significantly affected by the re-chlorination tank between site XLW and site MQ (p < 0.001), with a considerable decrease in absolute abundance. One-way ANOVA analysis showed that sampling sites greatly affected the absolute abundance of Salmonella sp. (p < 0.05).

Based on the metagenomic data, we compared differences in the communities of potential pathogens in TW during suspension–restoration water (SRW) supply and stable water supply. Statistical analysis showed that SRW and SW had similar alpha diversity (p > 0.05) of the pathogens (Fig. 2A), but some prevalent pathogens had different abundances between the SRW and SW samples. SRW was found to have seven pathogens with high relative abundance of over 0.1%, and SW had eight ones, of which five pathogens were shared with the two water types (Fig. 2B). P. aeruginosa (4.91%), P. alcaligenes (1.05%) and Ochrobactrum anthropi (0.13%) were considered highly abundant in SW exclusively, while Acinetobacter johnsonii (0.59%) and Mycobacterium kansasii (0.33%) had high abundance in SRW. The predominant pathogen changed from P. aeruginosa in SW to A. johnsonii in SRW (Fig. 2B). Moreover, SW and SRW were identified to have 78 pathogens with significant difference (p < 0.05, Fig. S2) and 18 pathogens with extremely significant difference (p < 0.001, Fig. 2C) in the relative abundance. Eight pathogens had higher proportions in SRW group, and 70 ones were highly abundant in SW group (p < 0.05, Fig. S2). When p < 0.001, all 18 pathogens had higher proportions in SW group (Fig. 2C).

Differences in pathogen community composition under different water supply scenarios. (A) Shannon index (mean ± standard error, n = 3 in SRW and n = 15 in SW) of pathogen community showing the difference of alpha diversity between SRW and SW. (B) Bubble plot showing the relative abundance of abundant (> 0.1%) pathogens in SW and SRW samples. (C) Extended error bar plot showing the pathogens with extremely significant differences (p < 0.001) between SRW and SW.

As for absolute quantification, PMA-qPCRs showed that the occurrence frequency of the three pathogens (P. aeruginosa, M. avium and Salmonella sp.) was 100% in SRW, which was substantially higher compared to SW. The largest increase in absolute abundance of Salmonella sp. (3.60 log, p < 0.05), followed by P. aeruginosa (2.58 log, p < 0.05) and M. avium (2.44 log, p < 0.05) was observed within 0–2 min after the restoration of water supply compared to the stable water supply. Moreover, the absolute abundances of the three pathogens in SRW (0–6 min) were significantly higher than those in SW (p < 0.05), and decreased with the extension of the water supply restoration time (Fig. 3). The absolute abundance of P. aeruginosa in SRW after 6 min of water supply restoration was similar to that in SW (p > 0.05). However, the absolute abundances of M. avium (1.09 × 104 cells/L) and Salmonella sp. (2.14 × 103 cells/L) were significantly higher than those in SW (p < 0.05) even after water supply restoration for 2 h.

Absolute abundance (mean ± standard error, n = 3) of the three typical pathogens in SRW after water supply restoration. * means significant difference by t-test between SRW and SW (p < 0.05).

We further calculated the annual infection probabilities P(inf, y) of the three pathogens in each of SW and SRW samples. The absolute abundance data in SW were fitted to probability distribution models with maximum likelihood estimation (Table S2, Fig. S3), which were found to be appropriate by Kolmogorov–Smirnov test (Table S2). Monte Carlo simulation showed that the median P(inf, y) of each of the three pathogens (Fig. 4A) exceeded the risk limit (10−4) for TW recommended by US EPA28. Furthermore, the median P(inf, y) of M. avium (1.24 × 10−2) and Salmonella sp. (5.54 × 10−2) through oral intake were one order of magnitude higher than that of P. aeruginosa (1.20 × 10−3) through dermal exposure (Fig. 4A). Among the three pathogens, a lower infection risk of P. aeruginosa in TW was observed compared to the other two pathogens (p < 0.0001, Fig. 4A). However, the difference of P(inf, y) between M. avium and Salmonella sp. was not significant (p > 0.05, Fig. 4A).

Annual infection probability of the three pathogens in SW (A) and SRW (B). (A) n = 5000. (B) n = 3. The exposure routes were dermal exposure of P. aeruginosa and oral intake of M. avium and Salmonella sp.; **** means extremely significant difference and ns means no significant difference by t-test between the two pathogens (p < 0.0001). * means significant difference by t-test between SRW and SW (p < 0.05). Dash line means the risk threshold 1 × 10−4.

The P(inf, y) of the three pathogens in each SRW sample increased significantly (P. aeruginosa increased by 3.02 × 102, M. avium increased by 1.86 × 102 and Salmonella sp. increased by 2.63 × 103, each p < 0.05) during 0–2 min after water supply restoration, and showed temporally decreasing trends (Fig. 4B). In comparison with M. avium and Salmonella sp., P. aeruginosa demonstrated a more remarkable decrease of P(inf, y) along with the time extension of water supply restoration. However, the P(inf, y) of each pathogen in SRW was above the threshold within the first 10 min after water supply restoration. Notably, the P(inf, y) of M. avium and Salmonella sp. remained significantly higher than the threshold in SW (p < 0.05) after two hours of the water supply restoration, while the P(inf, y) of P. aeruginosa (9.33 × 10−5) was lower than that in site SRWA.

Moreover, we also calculated the acceptable exposure volume of TW that meets the risk requirement proposed by EPA (10−4). The maximum acceptable volume of TW was 1.08 × 10−5 L in dermal exposure of P. aeruginosa, 2.94 × 10−2 L in oral intake of M. avium and 1.97 × 10−2 L in oral intake of Salmonella sp. (Fig. 5B). The acceptable exposure volumes of TW for all three bacteria were significantly lower (p < 0.0001, Fig. 5) compared to the exposure volumes adopted in Monte-Carlo simulations, implying high risks arising from direct utilization of TW.

Simulated exposure volume (A) and acceptable exposure volume (B) of TW for dermal exposure or oral intake. (A) n = 5000. (B) n = 18. The exposure routes were dermal exposure of P. aeruginosa and oral intake of M. avium and Salmonella sp. p < 0.0001 means extremely significant difference by t-test between simulated exposure volume and acceptable exposure volume of drinking water.

In order to accurately assess the risk for direct utilization of TW, this study gained comprehensive insights into the pathogen communities of TW, and chose key pathogens and optimized the method for absolute quantification. Since PMA pretreatment can eliminate the disturbance of dead bacteria23, this study applied PMA-qPCR to determine the cell numbers of bacterial pathogens in the given volume of water to meet the requirements of QMRA19. Single-copy genes were chosen to avoid complex conversions and uneven amplification of multi-copy genes, which was often overlooked29,30. Bacteria recovery efficiency, DNA extraction efficiency and PCR amplification efficiency were also under consideration to increase the accuracy of absolute quantification of the pathogens. Moreover, based on the infection characteristics of the three pathogens in TW, different exposure routes were also considered in this study to enhance the reliability of QMRA results. P. aeruginosa causes eye and skin infections31,32, and systemic infections through the injured skin into the patient33. Infection caused by Salmonella sp. begins with the ingestion of contaminated food or water34, allowing pathogens to reach the intestinal epithelium and cause intestinal diseases such as typhoid and paratyphoid fever35. M. avium can cause zoonotic infection by nasal or oral ingestion and can affect a variety of human tissues and organs, including the lungs, bone marrow and lymph nodes36.

In this study, both metagenomic analysis and PMA-qPCR revealed the presence of a variety of bacterial pathogens in TW. P. aeruginosa had the highest abundance in TW (especially in site MQ), which indicates that P. aeruginosa particularly adapts to the pipeline network environment compared to other pathogens. Both culture-dependent and molecular methods have confirmed the frequent occurrence of P. aeruginosa in TW37,38. The species is able to tolerate chlorine39 due to the superior barrier properties of outer membrane40 and remarkable release of extracellular polymeric substances41, resulting in its higher abundance than the other species within the genus of Pseudomonads42. Moreover, chlorination process can induce the over expression of the MexEF-OprN efflux pump43 and the promotion of horizontal transfer of plasmids44, which may greatly promote antibiotic resistance of P. aeruginosa and increased the infection risk.

We found that the abundance and community structure of pathogens varied considerably under different water supply scenarios, which is mainly driven by the bulk water growth and biofilm detachment. Many factors are likely to cause bulk water growth, such as sediment, residual chlorine and TOC. The sediment in the pipe will be stirred up when the water supply is restored after the water supply has been suspended, which may cause an increase in turbidity and the number of bacteria45. The concentration of chlorine and TOC was an important factor in limiting the number of bacteria. The residual chlorine concentration decreased and total cell counts increased significantly in TW after stagnation11. In addition, a decrease in TOC concentration was observed during water stagnation, which was partly caused by bacteria consumption for growth46.

Biofilm provides a possible habitat for pathogens, where microbes are protected against disinfectants47. These organisms include fecal indicator bacteria, obligate pathogens of fecal origin, opportunistic pathogens of environmental origin, enteric viruses and parasitic protozoa48. In fact, only less than 2% of the bacterial members in a drinking water system are found in the water phase49. Thus, bacteria in biofilms enter the pipeline due to the shear stress after water supply suspension–restoration50,51, increasing the absolute abundances in SRW of the three typical pathogens (P. aeruginosa, M. avium and Salmonella sp.) that tend to form biofilm in DWDS52,53,54,55 and can be persistent in biofilm for weeks56. Compared to SW, pathogens with increased relative abundance in SRW included Acinetobacter, Sphingomonas and Mycobacterium, the genera with high abundance in pipe biofilm48,57,58.

This study revealed that the annual infection risk of the three pathogens in DW were significantly higher than the EPA-designated acceptable threshold (10−4) in the case of direct utilization of TW without any treatment. Similarly, some previous studies have showed that the infection probability in many regions exceeded this threshold59,60,61,62. Besides, the exposure routes selected in this study were direct contact or drinking of TW, but residents in China usually boil TW before consumption. Also, some point-of-use commercial water purifiers like reverse membranes can further decrease the abundance and the health risks of pathogens, which were not taken consideration in this study. Since pathogens are greatly reduced by boiling or membrane treatment63, the above results therefore somewhat overestimated the potential human health risk of the pathogens arising from contact or consumption of tap water. Notwithstanding, the finding of this study that health risks of pathogens in SRW is significantly higher than that in SW still makes sense because of the changes in pathogen community and increases of their abundances. In addition, we further compared the simulated exposure volumes used in QMRA with the acceptable exposure volumes when utilized directly, resulting in significantly lower acceptable exposure volumes in this study. These warn us that the risk of TW is higher during a short period after the water supply has been restored. We suggest residents avoid direct consumption of TW if possible, including washing and direct drinking, during this period especially the first two hours.

As for all QMRA approaches, the input variables remain uncertain and limited, including exposure dose, assumptions underlying exposure assessment and dose–response parameters for pathogens used in the model9. We compensated for this uncertainty and limitation of exposure dose by several attempts. Considering assumptions underlying exposure assessment, we modeled daily drinking volume of TW as 1 L based on exposure scenarios described in EPA and WHO handbook5,64. This probability distribution is hardly representative of water consumption behavior in settings including a complex system of household water management. While for the dose–response parameters in the model, those parameters were collected in high-income setting with healthy adults, which may underestimate the infection risk for populations in developing countries and young children65. Drinking and rinsing after heating and differences in consumers were not under consideration due to the lack of relevant parameters. Besides the annual infection probability, many studies have also calculated disability-adjusted life years (DALYs) as the summary measure of disease burden9,60,66. However, this component was not taken into consideration in this study by virtue of the absence of relevant epidemiological data.

This study provided a comprehensive insight into the effect of water supply suspension–restoration on human pathogen community composition and their potential human health risks. P. aeruginosa was the dominant pathogen in SW samples. Water supply suspension–restoration caused the obvious shift of the community composition of pathogens and the evident increase in the absolute abundances of P. aeruginosa, M. avium and Salmonella sp. The infection risks of the three pathogens arising from direct utilization of drinking water under stable water supply, including dermal exposure and oral intake, were all above the threshold of 10−4 recommended by EPA. Water supply suspension–restoration further increased the health risk, demonstrating the necessity that more attempts should be dedicated to controlling the pathogens proliferation induced by water supply suspension.

Sampling TW under stable water supply (no water cut-off for over one week, SW samples) and suspended–restored water supply (water supply restored after more than six-hours suspension, SRW samples) was carried out to characterize the composition and its variation of bacterial community. The coordinate positions of the sampling sites were shown in Table S3. We collected TW samples with filter elements (MK2-EG-BG, Toray, Japan) for the maximum biomass along the distribution pipeline according to the linear distance from drinking water treatment plant (DWTP) (Fig. 1A) for 10 times (every two weeks from June 14th 2019) and obtained a total of 50 SW samples. Three water supply suspension events (Fig. 1A) were selected by inquiring about the water suspension announcement in advance. Before water supply suspension, 10 L of tap water were collected in SRW site A–C as SW samples. When drinking water supply was restored, we continuously collected tap water (10 L) at five time points within the first ten minutes (0–2 min, 2–4 min, 4–6 min, 6–8 min and 8–10 min), considering that the flow rate was 5.48 L/min. We also collected samples after two hours (2 h later) and the taps were kept running until collection. TW samples in SRW site A–C were filtered with 0.22 μm micropore membrane (BKMAN, China) to collect microorganism, because of limited TW volume. For DNA extraction, the filter elements and micropore membranes were separately soaked with 1 × PBS overnight to elute microorganism and the eluent was centrifuged at 14,000×g for 15 min. DNA was extracted using the FastDNA® Spin Kit for Soil (MP Biomedicals, USA), and the DNA concentration and purity were measured by micro-spectrophotometry (NanoDrop One, Thermo Fisher Scientific, China).

DNA extracted from the three SW samples (randomly chosen from the 10 time points) collected at each site and the three SRW samples (0–2 min) were subject to metagenome sequencing on the Illumina HiSeq 2500 platform (Novogene Bioinformatic Technology, Beijing, China). The sequencing strategy was Index 150 PE (paired-end sequencing, 150-bp reads) to generate nearly equal number of reads for each sample. The final size of the raw data in FASTQ format was approximately 100 M reads for each sample (Table S4). The raw reads with low quality (more than 10 “N” or 50% bases with Q ≤ 5) or contaminated by adapter were removed by fastp67 (version 0.19.7), and the filtered clean reads were used for metagenomic analysis. The sequencing reads of TW samples were annotated by Metaphlan268 with default parameters to obtain species profiles and their relative abundance. In order to identify HPB, the species profiles were compared with the self-established HPB database composed of 534 species (Table S5)69,70,71.

The single-copy genes, oaa, 16S rRNA gene (V1–V2) and invA (Table S6) have been identified to be specifically located on the genomes of P. aeruginosa72, M. avium73 and Salmonella sp.74,75, respectively. Thus, the three genes in all the TW samples were quantified by PMA-qPCR to determine their copy numbers. An optimization assay was also designed and conducted to determine the optimal concentration of PMAxx (Biotium, USA), which was finally achieved at 15 μM (Text S1, Fig. S4). qPCRs were conducted with a final volume of 25 μL, containing 12.5 μL of 2 × SYBR Green Mix (Vazyme Biotech, Nanjing, China), 1 μL of each primer (10 μM), 2 μL of template DNA and 8.5 μL of ddH2O. Thermal cycling and fluorescence detection were performed under different reaction conditions (Table S7) on QuantStudio 3 Real-Time PCR Systems with QuantStudio Design and Analysis Software (version 1.4) (Thermo Fisher Scientific, China). Each reaction was performed in triplicate. Standard curves (Fig. S5) were obtained with tenfold serial dilutions of the recombinant plasmids carrying target genes76, to generate the amplification efficiency and correlation coefficient (R2) of the PCRs. Based on the standard curve and the Ct value of each sample, the gene abundance was calculated and normalized against volume (L) of water samples. Stability tests and sensitivity tests were carried out to ensure the feasibility of the optimized PMA-qPCR (details in Text S1, Table S8, Figs. S6 and S7).

For absolute quantification of the three pathogens, tap water samples spiked with the three typical pathogens were tested to obtain the recovery efficiencies during the sample pretreatment (details in Text S1). Briefly, the given concentrations of pathogens (Table S9) were added to TW and filtered through filter elements or micropore membrane to collect microorganism. After quantified with the optimized PMA-qPCR method, the concentration of recovered pathogens was correlated to the spiked bacterial concentrations after logarithm transformation to obtain the recovery efficiency for the water filtration and DNA extraction methods (Table S9). Each gradient was set up in triplicate for the experiment.

The number of pathogens per liter (cells/L) in DW, namely the absolute concentration (C), was calculated by the formula:

where, Q is concentration of extracted DNA eluent quantified with qPCR (copies/μL DNA eluent), Va is final volume of extracted DNA eluent (μL), k is recovery efficiency of targeted bacteria for water filtration and DNA extraction, and Vb is volume of drinking water samples to extract DNA (L).

QMRA of the target pathogens (P. aeruginosa, Salmonella sp. and M. avium) was conducted by following the most commonly used four-step risk assessment process77, including hazard identification, exposure assessment, dose–response assessment and risk characterization (details in Text S2). We chose the three pathogens for QMRA by considering frequency of detection, reliability of detection methods, and availability of dose–response information77, based on the data obtained from this study and previous studies (details in Text S2). Two exposure routes, namely oral intake (direct drinking) and dermal exposure (eye contact during washing), were opted for exposure assessment. Beta-Poisson model was used to simulate the dose–response correlations for the three pathogens, and the relevant information has been summarized in Tables S10 and S11. The annual infection probability P(inf, y), the descriptive endpoint of risk assessment, was calculated with the formula shown in Text S2. We ran 5000 simulations using Monte Carlo in Matlab to predict the overall annual infection risk.

Cluster analysis was performed with PAST (version 4.03) for high-abundance pathogens (average relative abundance > 0.1% in all SW samples) identified by metagenomic analysis based on Bray–Curtis distance. Welch’s t-test was used to identify pathogens with significant differences (p < 0.05) between SRW and SW samples with STAMP78 (version 2.1.3). One-way ANOVA and t-test were performed using SPSS (IBM, USA) to determine significant difference in absolute abundance and annual infection probability of the pathogens at the different sampling sites, and the results were considered statistically significant when p < 0.05.

All the metagenomic data in this study have been deposited in NCBI Sequence Read Archive under accession number PRJNA807827.

Leclerc, H., Schwartzbrod, L. & Dei-Cas, E. Microbial agents associated with waterborne diseases. Crit. Rev. Microbiol. 28, 371–409. https://doi.org/10.1080/1040-840291046768 (2002).

Article CAS PubMed Google Scholar

Brookes, J. D. et al. Emerging challenges for the drinking water industry. Environ. Sci. Technol. 48, 2099–2101. https://doi.org/10.1021/es405606t (2014).

Article ADS CAS PubMed Google Scholar

Falkinham, J. O., Hilborn, E. D., Arduino, M. J., Pruden, A. & Edwards, M. A. Epidemiology and ecology of opportunistic premise plumbing pathogens: Legionella pneumophila, Mycobacterium avium, and Pseudomonas aeruginosa. Environ. Health Perspect. 123, 749–758. https://doi.org/10.1289/ehp.1408692 (2015).

Article CAS PubMed PubMed Central Google Scholar

EPA. National Primary Drinking Water Regulations (Office of Ground Water and Drinking Water, 2009).

Google Scholar

WHO. Guidelines for Drinking Water-Quality 4th edn. (The World Health Organization, 2011).

Google Scholar

Liu, X.-Y., Hong, Y. & Gu, W.-P. Influence of light quality on Chlorella growth, photosynthetic pigments and high-valued products accumulation in coastal saline-alkali leachate. Water Reuse 11, 301–311. https://doi.org/10.2166/wrd.2021.088 (2021).

Article ADS CAS Google Scholar

Erickson, J. J., Quintero, Y. C. & Nelson, K. L. Characterizing supply variability and operational challenges in an intermittent water distribution network. Water 12, 2143. https://doi.org/10.3390/w12082143 (2020).

Article Google Scholar

Danilenko, A., van den Berg, C., Macheve, B. & Moffitt, L. J. The IBNET Water Supply and Sanitation Blue Book 2014. The International Benchmarking Network for Water and Sanitation Utilities Databook (World Bank Group, 2014).

Book Google Scholar

Bivins, A. W. et al. Estimating infection risks and the global burden of diarrheal disease attributable to intermittent water supply using QMRA. Environ. Sci. Technol. 51, 7542–7551. https://doi.org/10.1021/acs.est.7b01014 (2017).

Article ADS CAS PubMed Google Scholar

Ye, C. et al. Recovery of microbiological quality of long-term stagnant tap water in university buildings during the COVID-19 pandemic. Sci. Total Environ. 806, 150616–150616. https://doi.org/10.1016/j.scitotenv.2021.150616 (2021).

Article CAS PubMed PubMed Central Google Scholar

Zhang, H. et al. Stagnation trigger changes to tap water quality in winter season: Novel insights into bacterial community activity and composition. Sci. Total Environ. https://doi.org/10.1016/j.scitotenv.2022.157240 (2022).

Article PubMed PubMed Central Google Scholar

Leishear, R. A. Our water mains contaminate us with E. coli, lead and copper—Preventable illness and death follow. Am. J. Water Sci. Eng. 7, 72–96 (2021).

Article Google Scholar

Szewzyk, U., Szewzyk, R., Manz, W. & Schleifer, K. H. Microbiological safety of drinking water. Annu. Rev. Microbiol. 54, 81–127 (2000).

Article CAS PubMed Google Scholar

Ramirez-Castillo, F. Y. et al. Waterborne pathogens: Detection methods and challenges. Pathogens (Basel, Switzerland) 4, 307–334. https://doi.org/10.3390/pathogens4020307 (2015).

Article PubMed Google Scholar

Godoy, R. G. et al. Bacterial community composition and potential pathogens along the Pinheiros River in the southeast of Brazil. Sci. Rep. 10, 9331. https://doi.org/10.1038/s41598-020-66386-y (2020).

Article ADS CAS PubMed PubMed Central Google Scholar

Ju, F. et al. Antibiotic resistance genes and human bacterial pathogens: Co-occurrence, removal, and enrichment in municipal sewage sludge digesters. Water Res. 91, 1–10. https://doi.org/10.1016/j.watres.2015.11.071 (2016).

Article CAS PubMed Google Scholar

Nakamura, S. et al. Direct metagenomic detection of viral pathogens in nasal and fecal specimens using an unbiased high-throughput sequencing approach. PLoS One https://doi.org/10.1371/journal.pone.0004219 (2009).

Article PubMed PubMed Central Google Scholar

Huang, K., Zhang, X.-X., Shi, P., Wu, B. & Ren, H. A comprehensive insight into bacterial virulence in drinking water using 454 pyrosequencing and Illumina high-throughput sequencing. Ecotoxicol. Environ. Saf. 109, 15–21. https://doi.org/10.1016/j.ecoenv.2014.07.029 (2014).

Article CAS PubMed Google Scholar

Schijven, J. F. & Husman, A. M. D. R. Applications of Quantitative Microbial Source Tracking (QMST) and Quantitative Microbial Risk Assessment (QMRA) (Springer, 2011).

Book Google Scholar

Owens, C. E. L. et al. Implementation of quantitative microbial risk assessment (QMRA) for public drinking water supplies: Systematic review. Water Res. 174, 115614. https://doi.org/10.1016/j.watres.2020.115614 (2020).

Article CAS PubMed Google Scholar

Aw, T. G. & Rose, J. B. Detection of pathogens in water: From phylochips to qPCR to pyrosequencing. Curr. Opin. Biotechnol. 23, 422–430. https://doi.org/10.1016/j.copbio.2011.11.016 (2012).

Article CAS PubMed Google Scholar

Wallmann, L. et al. Fate and persistence of antibiotic-resistant bacteria and genes through a multi-barrier treatment facility for direct potable reuse. Water Reuse 11, 373–390. https://doi.org/10.2166/wrd.2021.097 (2021).

Article MathSciNet CAS Google Scholar

Nocker, A., Cheung, C. Y. & Camper, A. K. Comparison of propidium monoazide with ethidium monoazide for differentiation of live vs. dead bacteria by selective removal of DNA from dead cells. J. Microbiol. Methods 67, 310–320. https://doi.org/10.1016/j.mimet.2006.04.015 (2006).

Article CAS PubMed Google Scholar

Luo, L.-W. et al. Aggravated biofouling caused by chlorine disinfection in a pilot-scale reverse osmosis treatment system of municipal wastewater. Water Reuse 11, 201–211. https://doi.org/10.2166/wrd.2021.108 (2021).

Article ADS CAS Google Scholar

Bae, S. & Wuertz, S. Rapid decay of host-specific fecal Bacteroidales cells in seawater as measured by quantitative PCR with propidium monoazide. Water Res. 43, 4850–4859. https://doi.org/10.1016/j.watres.2009.06.053 (2009).

Article CAS PubMed Google Scholar

Chen, S. Y., Wang, F., Beaulieu, J. C., Stein, R. E. & Ge, B. L. Rapid detection of viable Salmonellae in produce by coupling propidium monoazide with loop-mediated isothermal amplification. Appl. Environ. Microbiol. 77, 4008–4016. https://doi.org/10.1128/aem.00354-11 (2011).

Article ADS CAS PubMed PubMed Central Google Scholar

Chiao, T. H., Clancy, T. M., Pinto, A., Xi, C. W. & Raskin, L. Differential resistance of drinking water bacterial populations to monochloramine disinfection. Environ. Sci. Technol. 48, 4038–4047. https://doi.org/10.1021/es4055725 (2014).

Article ADS CAS PubMed Google Scholar

Fewtrell, L. & Bartram, J. In Water Quality: Guidelines, Standards and Health : Assessment of Risk and Risk Management for Water-Related Infectious Diseases (eds Fewtrell, L. & Bartram, J.) (World Health Organization, 2001).

Google Scholar

Cui, Q., Fang, T., Huang, Y., Dong, P. & Wang, H. Evaluation of bacterial pathogen diversity, abundance and health risks in urban recreational water by amplicon next-generation sequencing and quantitative PCR. J. Environ. Sci. 57, 137–149. https://doi.org/10.1016/j.jes.2016.11.008 (2017).

Article CAS Google Scholar

Fang, T. et al. Distribution comparison and risk assessment of free-floating and particle-attached bacterial pathogens in urban recreational water: Implications for water quality management. Sci. Total Environ. 613, 428–438. https://doi.org/10.1016/j.scitotenv.2017.09.008 (2018).

Article ADS CAS PubMed Google Scholar

Chiller, K., Selkin, B. A. & Murakawa, G. J. Skin microflora and bacterial infections of the skin. J. Investig. Dermatol. Symp. Proc. 6, 170–174. https://doi.org/10.1046/j.0022-202x.2001.00043.x (2001).

Article CAS PubMed Google Scholar

Mena, K. D. & Gerba, C. P. Risk assessment of Pseudomonas aeruginosa in water. Rev. Environ. Contam. Toxicol. 201, 71–115 (2009).

CAS PubMed Google Scholar

Cogen, A. L., Nizet, V. & Gallo, R. L. Skin microbiota: A source of disease or defence?. Br. J. Dermatol. 158, 442–455. https://doi.org/10.1111/j.1365-2133.2008.08437.x (2008).

Article CAS PubMed PubMed Central Google Scholar

Fabrega, A. & Vila, J. Salmonella enterica serovar Typhimurium skills to succeed in the host: Virulence and regulation. Clin. Microbiol. Rev. 26, 308–341. https://doi.org/10.1128/cmr.00066-12 (2013).

Article CAS PubMed PubMed Central Google Scholar

Stanaway, J. D. et al. The global burden of typhoid and paratyphoid fevers: A systematic analysis for the Global Burden of Disease Study 2017. Lancet Infect. Dis. 19, 369–381. https://doi.org/10.1016/s1473-3099(18)30685-6 (2019).

Article Google Scholar

Uwamino, Y. et al. Showering is associated with Mycobacterium avium complex lung disease: An observational study in Japanese women. J. Infect. Chemother. 26, 211–214. https://doi.org/10.1016/j.jiac.2019.09.003 (2020).

Article PubMed Google Scholar

Felfoldi, T., Tarnoczai, T. & Homonnay, Z. G. Presence of potential bacterial pathogens in a municipal drinking water supply system. Acta Microbiol. Immunol. Hung. 57, 165–179. https://doi.org/10.1556/AMicr.57.2010.3.2 (2010).

Article CAS PubMed Google Scholar

Schiavano, G. F. et al. Prevalence and antibiotic resistance of Pseudomonas aeruginosa in water samples in central Italy and molecular characterization of oprD in imipenem resistant isolates. PLoS One 12, e0189172. https://doi.org/10.1371/journal.pone.0189172 (2017).

Article CAS PubMed PubMed Central Google Scholar

Luo, L. W. et al. Evaluating method and potential risks of chlorine-resistant bacteria (CRB): A review. Water Res. 188, 116474. https://doi.org/10.1016/j.watres.2020.116474 (2021).

Article CAS PubMed Google Scholar

Cloete, T. E. Resistance mechanisms of bacteria to antimicrobial compounds. Int. Biodeterior. Biodegrad. 51, 277–282. https://doi.org/10.1016/s0964-8305(03)00042-8 (2003).

Article CAS Google Scholar

Wang, Y.-H. et al. Effects of chlorine disinfection on the membrane fouling potential of bacterial strains isolated from fouled reverse osmosis membranes. Sci. Total Environ. 693, 133579. https://doi.org/10.1016/j.scitotenv.2019.133579 (2019).

Article ADS CAS PubMed Google Scholar

Jia, S. et al. Bacterial community shift drives antibiotic resistance promotion during drinking water chlorination. Environ. Sci. Technol. 49, 12271–12279. https://doi.org/10.1021/acs.est.5b03521 (2015).

Article ADS CAS PubMed Google Scholar

Hou, A.-M. et al. Chlorine injury enhances antibiotic resistance in Pseudomonas aeruginosa through over expression of drug efflux pumps. Water Res. 156, 366–371. https://doi.org/10.1016/j.watres.2019.03.035 (2019).

Article CAS PubMed Google Scholar

Jin, M. et al. Chlorine disinfection promotes the exchange of antibiotic resistance genes across bacterial genera by natural transformation. ISME J. 14, 1847–1856. https://doi.org/10.1038/s41396-020-0656-9 (2020).

Article CAS PubMed PubMed Central Google Scholar

Hu, D. et al. Microbiological quality of roof tank water in an urban village in southeastern China. J. Environ. Sci. 125, 148–159. https://doi.org/10.1016/j.jes.2022.01.0361001-0742 (2023).

Article Google Scholar

Zlatanovic, L., van der Hoek, J. P. & Vreeburg, J. H. G. An experimental study on the influence of water stagnation and temperature change on water quality in a full-scale domestic drinking water system. Water Res. 123, 761–772. https://doi.org/10.1016/j.watres.2017.07.019 (2017).

Article CAS PubMed Google Scholar

Flemming, H. C., Percival, S. L. & Walker, J. T. in International Conference on Innovations in Conventional and Advanced Water Treatment Processes 271–280 (2002).

Wingender, J. & Flemming, H. C. Biofilms in drinking water and their role as reservoir for pathogens. Int. J. Hyg. Environ. Health 214, 417–423. https://doi.org/10.1016/j.ijheh.2011.05.009 (2011).

Article PubMed Google Scholar

Liu, G. et al. Pyrosequencing reveals bacterial communities in unchlorinated drinking water distribution system: An integral study of bulk water, suspended solids, loose deposits, and pipe wall biofilm. Environ. Sci. Technol. 48, 5467–5476. https://doi.org/10.1021/es5009467 (2014).

Article ADS CAS PubMed Google Scholar

Kumpel, E. & Nelson, K. L. Mechanisms affecting water quality in an intermittent piped water supply. Environ. Sci. Technol. 48, 2766–2775. https://doi.org/10.1021/es405054u (2014).

Article ADS CAS PubMed Google Scholar

Kumpel, E. & Nelson, K. L. Intermittent water supply: Prevalence, practice, and microbial water quality. Environ. Sci. Technol. 50, 542–553. https://doi.org/10.1021/acs.est.5b03973 (2016).

Article ADS CAS PubMed Google Scholar

Rasamiravaka, T., Labtani, Q., Duez, P. & El Jaziri, M. The formation of biofilms by Pseudomonas aeruginosa: A review of the natural and synthetic compounds interfering with control mechanisms. BioMed Res. Int. https://doi.org/10.1155/2015/759348 (2015).

Article PubMed PubMed Central Google Scholar

Mann, E. E. & Wozniak, D. J. Pseudomonas biofilm matrix composition and niche biology. FEMS Microbiol. Rev. 36, 893–916. https://doi.org/10.1111/j.1574-6976.2011.00322.x (2012).

Article CAS PubMed Google Scholar

Falkinham, J. O., Norton, C. D. & LeChevallier, M. W. Factors influencing numbers of Mycobacterium avium, Mycobacterium intracellulare, and other mycobacteria in drinking water distribution systems. Appl. Environ. Microbiol. 67, 1225–1231. https://doi.org/10.1128/aem.67.3.1225-1231.2001 (2001).

Article ADS CAS PubMed PubMed Central Google Scholar

Steenackers, H., Hermans, K., Vanderleyden, J. & De Keersmaecker, S. C. J. Salmonella biofilms: An overview on occurrence, structure, regulation and eradication. Food Res. Int. 45, 502–531. https://doi.org/10.1016/j.foodres.2011.01.038 (2012).

Article CAS Google Scholar

Torvinen, E., Lehtola, M. J., Martikainen, P. J. & Miettinen, I. T. Survival of Mycobacterium avium in drinking water biofilms as affected by water flow velocity, availability of phosphorus, and temperature. Appl. Environ. Microbiol. 73, 6201–6207. https://doi.org/10.1128/aem.00828-07 (2007).

Article ADS CAS PubMed PubMed Central Google Scholar

Ren, H. X. et al. Pyrosequencing analysis of bacterial communities in biofilms from different pipe materials in a city drinking water distribution system of East China. Appl. Microbiol. Biotechnol. 99, 10713–10724. https://doi.org/10.1007/s00253-015-6885-6 (2015).

Article CAS PubMed Google Scholar

Shentu, H.-B. et al. Radial distribution shifts of microbial communities in biofilms developed under stagnation condition in water supply pipes. J. Zhejiang Univ. (Eng. Sci.) 51, 1317–1323 (2017).

Google Scholar

George, J. et al. Quantitative microbial risk assessment to estimate the health risk in urban drinking water systems of Mysore, Karnataka, India. Water Qual. Expo. Health 7, 331–338. https://doi.org/10.1007/s12403-014-0152-4 (2015).

Article CAS Google Scholar

Machdar, E., van der Steen, N. P., Raschid-Sally, L. & Lens, P. N. L. Application of quantitative microbial risk assessment to analyze the public health risk from poor drinking water quality in a low income area in Accra, Ghana. Sci. Total Environ. 449, 134–142. https://doi.org/10.1016/j.scitotenv.2013.01.048 (2013).

Article ADS CAS PubMed Google Scholar

Razzolini, M. T. P., Lauretto, M. D., Hachich, E. M., Sato, M. I. Z. & Nardocci, A. C. Giardia and Cryptosporidium infection risk by simultaneous exposure to drinking water. Microb. Risk Anal. 4, 1–6. https://doi.org/10.1016/j.mran.2016.01.001 (2016).

Article Google Scholar

Rodriguez-Alvarez, M. S. et al. Development of a relative risk model for drinking water regulation and design recommendations for a peri urban region of Argentina. Int. J. Hyg. Environ. Health 218, 627–638. https://doi.org/10.1016/j.ijheh.2015.06.007 (2015).

Article PubMed PubMed Central Google Scholar

Brown, J. & Sobsey, M. D. Boiling as household water treatment in Cambodia: A longitudinal study of boiling practice and microbiological effectiveness. Am. J. Trop. Med. Hyg. 87, 394–398. https://doi.org/10.4269/ajtmh.2012.11-0715 (2012).

Article PubMed PubMed Central Google Scholar

EPA. Development and Evaluation of Novel Dose-Response Models for Use in Microbial Risk Assessment (National Center for Environmental Assessment Office of Research and Development, 2008).

Google Scholar

Kotloff, K. L. et al. Burden and aetiology of diarrhoeal disease in infants and young children in developing countries (the Global Enteric Multicenter Study, GEMS): A prospective, case–control study. Lancet 382, 209–222. https://doi.org/10.1016/s0140-6736(13)60844-2 (2013).

Article PubMed Google Scholar

Schoen, M. E. et al. Quantitative microbial risk assessment of antimicrobial resistant and susceptible Staphylococcus aureus in reclaimed wastewaters. Environ. Sci. Technol. 55, 15246–15255. https://doi.org/10.1021/acs.est.1c04038 (2021).

Article ADS CAS PubMed PubMed Central Google Scholar

Chen, S. F., Zhou, Y. Q., Chen, Y. R. & Gu, J. fastp: An ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34, 884–890. https://doi.org/10.1093/bioinformatics/bty560 (2018).

Article CAS Google Scholar

Truong, D. T. et al. MetaPhlAn2 for enhanced metagenomic taxonomic profiling. Nat. Methods 12, 902–903. https://doi.org/10.1038/nmeth.3589 (2015).

Article CAS PubMed Google Scholar

Cantet, F. et al. Quantification of Vibrio parahaemolyticus, Vibrio vulnificus and Vibrio cholerae in French Mediterranean coastal lagoons. Res. Microbiol. 164, 867–874. https://doi.org/10.1016/j.resmic.2013.06.005 (2013).

Article CAS PubMed PubMed Central Google Scholar

Song, H.-M. et al. Study on pathogenic bacteria distribution in water environment of Shijiazhuang. J. Environ. Health 32, 140–142 (2015).

Google Scholar

Ward, L. N. & Bej, A. K. Detection of Vibrio parahaemolyticus in shellfish by use of multiplexed real-time PCR with TaqMan fluorescent probes. Appl. Environ. Microbiol. 72, 2031–2042. https://doi.org/10.1128/aem.72.3.2031-2042.2006 (2006).

Article ADS CAS PubMed PubMed Central Google Scholar

Choi, H. J. et al. Improved PCR for identification of Pseudomonas aeruginosa. Appl. Microbiol. Biotechnol. 97, 3643–3651. https://doi.org/10.1007/s00253-013-4709-0 (2013).

Article CAS PubMed Google Scholar

Wilton, S. & Cousins, D. Detection and identification of multiple mycobacterial pathogens by DNA amplification in a single tube. PCR Methods Appl. 1, 269–273 (1992).

Article CAS PubMed Google Scholar

Fey, A. et al. Establishment of a real-time PCR-based approach for accurate quantification of bacterial RNA targets in water, using Salmonella as a model organism. Appl. Environ. Microbiol. 70, 3618–3623. https://doi.org/10.1128/aem.70.6.3618-3623.2004 (2004).

Article ADS CAS PubMed PubMed Central Google Scholar

Malorny, B., Hoorfar, J., Bunge, C. & Helmuth, R. Multicenter validation of the analytical accuracy of Salmonella PCR: Towards an international standard. Appl. Environ. Microbiol. 69, 290–296. https://doi.org/10.1128/aem.69.1.290-296.2003 (2003).

Article ADS CAS PubMed PubMed Central Google Scholar

Shi, P. et al. Metagenomic insights into chlorination effects on microbial antibiotic resistance in drinking water. Water Res. 47, 111–120. https://doi.org/10.1016/j.watres.2012.09.046 (2013).

Article CAS PubMed Google Scholar

Hamouda, M. A. et al. Scenario-based quantitative microbial risk assessment to evaluate the robustness of a drinking water treatment plant. Water Qual. Res. J. Canada 51, 81–96. https://doi.org/10.2166/wqrjc.2016.034 (2016).

Article CAS Google Scholar

Parks, D. H., Tyson, G. W., Hugenholtz, P. & Beiko, R. G. STAMP: Statistical analysis of taxonomic and functional profiles. Bioinformatics 30, 3123–3124. https://doi.org/10.1093/bioinformatics/btu494 (2014).

Article CAS PubMed PubMed Central Google Scholar

Download references

This study was supported by the National Natural Science Foundation of China (52192682 and 52025102) and Postdoctoral Science Foundation of China (2022M711555).

State Key Laboratory of Pollution Control and Resource Reuse, School of the Environment, Nanjing University, 163 Xianlin Road, Nanjing, 210023, China

Shengnan Liu, Ruiming Jiang, Peng Liu & Xu-Xiang Zhang

China Three Gorges Construction Engineering Corporation, Beijing, 100048, China

Qisheng Li

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

P.L. designed the project, reviewed the manuscript and supervised the study. Q.L. collected samples, analyzed metagenomic data and prepared Figs. 1, 2. R.J. collected samples and improved the quantification methods. S.L. analyzed the data and prepared Figs. 3, 4, 5, wrote the main manuscript text. X.-X.Z. wrote part of the manuscript text, reviewed the manuscript and supervised the study.

Correspondence to Peng Liu or Xu-Xiang Zhang.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

Liu, S., Li, Q., Jiang, R. et al. Shift of human pathogen community composition and their potential human health risk after supply suspension in tap water. Sci Rep 13, 12419 (2023). https://doi.org/10.1038/s41598-023-39225-z

Download citation

Received: 30 May 2023

Accepted: 21 July 2023

Published: 01 August 2023

DOI: https://doi.org/10.1038/s41598-023-39225-z

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.