The coronavirus disease 2019 (COVID-19) pandemic has reshaped our understanding of pathogen diagnosis. Wastewater surveillance has gained widespread acceptance as a practical strategy complementary to clinical reports to achieve accurate estimates of infection prevalence [
1], [
2]. Pathogens shed by symptomatic or asymptomatic carriers congregate in wastewater. By analyzing the concentration and sequence of pathogen nucleic acids, a wastewater surveillance system can provide early warning and an overview of epidemic propagation within catchments [
3] (
Fig. 1). To fully exploit wastewater surveillance as part of epidemic sentinel surveillance systems, a comprehensive interpretation of wastewater surveillance results and deeper insights into what additional information wastewater surveillance can offer is crucial.
There is a correlation between the concentration of pathogenic nucleic acids (
CDNA or
CRNA) in wastewater and their prevalence levels. To interpret wastewater surveillance data effectively, a reliable model between
CDNA or
CRNA and key indicators, such as new infections, should be established. The integrity of genetic materials, particularly RNA, is compromised during transit to municipal wastewater treatment plants and is influenced by factors such as pH, temperature, flow rate, and travel time from viral shedding to sampling. Additionally, factors such as water use per capita and viral concentration in excreta also play a role. Incorporating some of these factors as covariants notably contributes to model accuracy [
4], emphasizing that data related to weather, water quality, and clinical testing should be collected simultaneously with
CDNA or
CRNA. However, the influence of these parameters on the discrepancies between shed and detected viral load varies both spatially and temporally. Hence, machine learning has been proposed to develop spatiotemporally specific models. Multiple algorithms, including gradient-boosted trees, random forests, and artificial neural networks, are viable, with better performance than linear regression in predicting syndrome coronavirus 2 (SARS-CoV-2) positive cases [
5].
Pathogenic nucleic acids in wastewater come from asymptomatic and pre-symptomatic carriers, in addition to clinically reported cases. Therefore, although using wastewater surveillance data to predict clinical cases over time is effective, correlating CDNA or CRNA with the overall number of pathogen positives within a catchment can fully exploit the data as a complement to clinical monitoring. A random sampling of residents and nucleic acid detection—carried out by either local authorities or academic institutions—can be used to estimate overall positives. The holistic prevalence deduced from wastewater surveillance can support comprehensive public health strategies, such as the distribution of nonprescription drugs for mild cases—where medical help may not be sought—rather than just the input of clinical medical resources.
Wastewater surveillance combined with nucleic acid sequencing is used to assess the origin, evolution, and relative prevalence of pathogenic variants. In-depth mining and interpretation are required to fully utilize the genetic information of pathogens in municipal wastewater, which avoids biases in clinical sampling and requires far less sampling and testing to depict dynamic trends. Optimal wastewater genomic surveillance should robustly identify diverse pathogen lineages and assess their relative abundance to elucidate the dynamic transmission processes of variants within specific temporal and spatial contexts. Nevertheless, nucleic acid content in municipal wastewater is low, occurring mostly in fragments, which entails a heightened demand for nucleic acid extraction and enrichment technologies to achieve ideal sequencing coverage.
The influence of the wastewater matrix, polymerase chain reaction (PCR) biases, and genetic information on adulterated organisms can further increase the difficulty in sequencing and analyzing sequencing results [
6]. The inter-disturbance of diverse lineages causes difficulties in genome assembly and the assignment of mutations across reads to a specific genome. To address these problems, refinement is required for each step, including sample collection, nucleic acid extraction, amplification, sequencing, and data analysis. Noise originating from irrelevant genetic material can be relieved through high sampling density over time [
6]. Although China has published a standard method for the enrichment and detection of SARS-CoV-2, this method has limitations in terms of enriching low-concentration nucleic acids and facilitating scaled-up testing. Hence, there is a need to develop and apply more sophisticated nucleic acid concentration methods to enhance concentration aids in identifying lineages with low prevalence [
7].
Replicated sequencing contributes to higher sequence depth, leading to a robust capture of lineage-defining mutations. Targeted sequencing of smaller epidemiologically informative regions can promote the identification of signature mutations, and ultimately, variants of concern [
8]. Attempts have also been made to submit wastewater-derived nucleic acid fragments for nanopore sequencing whose nature of long-read may be conducive [
9]. Previous studies [
6], [
10] have developed reliable tools to robustly deconvolute lineage abundance using complex sequencing data. Nevertheless, in terms of wastewater genomic surveillance, further efforts are required to provide comparable results for the genetic detection and tracking of pathogenic variants in wastewater across regions. The results obtained—the relative prevalence trend of existing variants and the discovery of newly introduced variants—await more in-depth interpretation by epidemiologists to reveal the characteristics and geographical transmission routes of these variants.
Recently, considerable advances have been made in wastewater surveillance systems worldwide. In Singapore, the National Environment Agency published a weekly wastewater viral load index during the COVID-19 pandemic as a public reference [
11]. Researchers in Thailand established a correlation between wastewater viral load and COVID-19 infection trends, recommending regular sampling in large-scale water treatment plants for effective routine monitoring [
12]. In rural Bangladesh, researchers identified optimal wastewater sampling sites for COVID-19 monitoring based on the evaluation of sanitation facilities. They confirmed the correlation between viral loads in these samples and nationally reported COVID-19 cases [
13]. In China, Hong Kong Special Administration Region has emerged as a pioneer in this field where researchers have conducted extended sampling at three wastewater treatment plants, providing support for the long-term sentinel monitoring of pathogens at these sites [
14]. However, China, as well as many other nations, faces challenges owing to its large size and uneven development. Implementation of national-scale wastewater surveillance requires the development of multidimensional strategies.
In anticipation of future epidemiological monitoring of wastewater pathogens, we hope to build a firewall not only against SARS-CoV-2 but also other public health threats—such as
Mycoplasma pneumonia, influenza, and norovirus—as well as antimicrobial resistance. Herein, we propose a four-dimensional strategy to fully exploit this information in wastewater (
Fig. 2). In regions with constrained resources and inadequate services,
in-situ diagnostics must be established at wastewater sites and should integrate sampling, nucleic acid extraction, and nucleic acid detection into a streamlined operational process that does not require specialized faculty or equipment and delivers reliable results. In this context, we believe that clustered regularly interspaced short palindromic repeats (CRISPR) detection—integrated with cost-effective paper-based devices not requiring specialized PCR equipment—could be a feasible option [
15].
During periods without large-scale epidemic outbreaks, qualitative nucleic acid tests conducted on-site at wastewater treatment plants are sufficient for monitoring and early warning. However, during large-scale outbreaks, professionals and competent laboratories must intervene, perform quantitative nucleic acid testing and sequencing, and further detailed analyses. In competent testing laboratories, a high-throughput analytical platform should be developed to discern subtle shifts in viral abundance accurately. Experiences during the SARS-CoV-2 pandemic have demonstrated the competency of existing nucleic acid diagnostic laboratories in conducting high-throughput nucleic acid tests. To enable high-throughput analysis of pathogenic nucleic acids in wastewater, it is necessary to integrate high-throughput enrichment and nucleic acid extraction methods for relevant pathogens [
7]. This will facilitate the establishment of a network of proficient wastewater-based epidemiological laboratories. By introducing a swift nucleic acid sequencing framework, a profound analysis of the relative abundance of various mutation subtypes can be achieved, thereby effectively characterizing the progression of epidemics.
The rapid evolution of pathogens is more likely to occur in regions with poor health and hygiene conditions, often found in less developed areas [
16]. For global monitoring systems, the deployment of additional monitoring stations in underdeveloped regions may be crucial. Therefore, with the assistance of developed countries, it is essential to implement cost-effective wastewater monitoring programs to establish local pathogen warning mechanisms. However, the absence of wastewater collection systems in rural areas makes it challenging to obtain representative samples for sequencing and analysis of local strains [
17]. Instead, collecting samples from high-traffic locations, such as hospitals, markets, and transportation hubs, is often more efficient and economical. Ultimately, using machine learning, a wastewater-based epidemiological surveillance model can be built to correlate wastewater-derived data with public health information on a broader spatial and temporal scale [
18]. This necessitates data sharing between regions and nations as well as improving normalization across datasets.
It should be noted that wastewater-based epidemic surveillance, though more cost-effective and easier to conduct than clinical screening—especially for underserved areas and low-resource settings—cannot pinpoint infected individuals. Although this can be partially overcome by sampling from sewers upstream of the wastewater treatment plant to monitor epidemic dynamics in institutions of concern, such as sanatoriums and universities, wastewater surveillance of pathogens cannot fully supplement clinical testing. On the contrary, clinical testing data, such as vaccination rates or shedding profiles of different variants, can assist in the better interpretation and utilization of wastewater-derived data.
The COVID-19 pandemic has emphasized the peculiar role of wastewater-based epidemiology. Wastewater contains comprehensive and abundant prevalence information that can be used for epidemiological surveillance. To provide a reliable basis for public health interventions, robust models for infection prevalence prediction using CDNA or CRNA need to be developed. Effective mining of additional information on viral variants and their epidemiological importance must be achieved. In this regard, we propose a four-dimensional methodology to establish an informative and regionally suitable wastewater surveillance system. With the help of a robust wastewater surveillance system, chemical and biological markers besides pathogens—such as illicit drugs, pharmaceuticals, and personal care products—in wastewater can be explored as an overall indicator of the health conditions and living habits of residents within a catchment.
Acknowledgments
The authors wish to thank the National Natural Science Foundation of China (51878638, 51821006, and 52192684) for supporting this work.
PDF
(542KB)
