Wastewater sequencing — an early warning system for infectious disease outbreaks

Monitoring the genomic characteristics of pathogens circulating in a population can reveal important insights into the epidemiological dynamics of an outbreak. Unfortunately, sequencing every confirmed positive sample in a densely populated area is both challenging and expensive. Since viruses are shed in faeces, routine sequence analysis of wastewater samples could act as an early warning system for the silent transmission of viruses, helping public health bodies make quick, informed decisions to reduce their spread. Wastewater-based epidemiological analysis using nanopore sequencing could provide a rapid, accurate, and cost-effective solution to augment pathogen surveillance.

‘We ... built in nanopore sequencing so we could ... remove that requirement for shipping a sample to a specialised facility. Nanopore sequencing can be performed within countries where the samples are collected’¹

Outbreaks of poliovirus — wild-type or vaccine derived — risk reversing the colossal progress of global eradication programmes. Successful control of any pathogen requires effective surveillance, involving rapid and accurate detection. At the Vaccine Epidemiology Research Group, Imperial College London, UK, Alex Shaw and colleagues were concerned with the time required to transport poliovirus samples to specialist sequencing facilities¹. RNA viruses, such as poliovirus and SARS-CoV-2, are fast evolving, so they continually accumulate changes in their genome. These mutations can provide essential information on virus transmission, spread, and evolution but only if samples are analysed quickly enough. To overcome these challenges, they developed Direct Detection by Nanopore Sequencing (DDNS), which utilises the accessibility, scalability, and portability of nanopore sequencing to deliver faster access to results. DDNS could provide actionable results ‘from sample to sequence in as little as 2 days’, at a cost of around ‘$15 per sample, [including] reagents and kits’¹. Their approach used a nested PCR2 to detect poliovirus in stool and wastewater samples (Figure 1). The two-step reaction first amplified the entire capsid region of the enteroviruses present in the sample. In the second step, the team used poliovirus-specific primers with barcodes³ allowing the amplified products from multiple samples to be pooled and sequenced on a single flow cell. Such sample multiplexing enables very high-throughput analyses and reduces cost. Leveraging the streamlined Ligation Sequencing Kit library preparation workflow, the team then used both the MinION and GridION devices to sequence samples, identifying poliovirus in real-time using RAMPART⁴ software, which maps reads according to serotype. The poliovirus-specific primers target the VP1 region of poliovirus, which is ‘required for informing a vaccination response’ and can help ‘map the movement of wild viruses’¹.

Figure 1: Poliovirus direct detection by nanopore sequencing (DDNS). Image kindly provided by Dr. Alex Shaw, Imperial College London, UK.

Long, accurate sequencing reads enable easier analysis of complex environmental surveillance samples. DDNS identifies poliovirus from metagenomic samples, providing a sample-by-sample report of which polioviruses are present and any mutations that have been acquired, compared to a reference sequence. For added flexibility, the first round RT-PCR product can be used to identify species A and C enteroviruses. The second primer set amplifies species B and D enteroviruses from the same RNA sample — providing cost-effective enterovirus sequencing when combined with amplicon pooling and native barcoding. Global training efforts are enabling other labs to perform DDNS for genomic surveillance, such as at the Institut National de Recherche Biomédicale (INRB) in the Democratic Republic of the Congo. The team at INRB compared the results obtained for 2,365 samples analysed using the traditional cell culture method with the metagenomic DDNS technique, reporting ‘excellent accuracy when comparing the two methods’¹. The sensitivity between the two approaches was as expected, based on previous work, and the specificity was ‘excellent’, indicating the absence of contamination despite performing nested PCRs. The turnaround time was much faster using DDNS (Figure 2) and the median sequence identity was 100% when comparing DDNS with Sanger sequencing.

Figure 2: Comparison of the time taken for the traditional cell culture and Sanger sequencing method to detect poliovirus compared with DDNS at the INRB. Median time between stool collection and sequence generation was 36 days using cell culture and Sanger sequencing, and 14 days by DDNS. Image kindly provided by Dr. Alex Shaw, Imperial College London, UK.

Key samples that were used to confirm three poliovirus outbreaks were identified by DDNS 23 days faster than the cell culture method. This rapid turnaround time suggests DDNS could replace the traditional cell culture and Sanger sequencing method, which ‘can take weeks’. DDNS could provide a rapid, scalable, and cost-effective response for effective poliovirus surveillance, guiding public health initiatives in countries where the virus is still endemic.

‘[DDNS can] potentially … lead to the ability to respond to outbreaks a lot more quickly. Quicker responses lead to smaller outbreaks’¹

Also leveraging the low cost and rapid turnaround times of nanopore sequencing in wastewater pathogen surveillance are Xuan Lin and Ryan Ziels with their colleagues from The University of British Columbia, Vancouver, Canada. Using a PCR enrichment approach to target SARS-CoV-2, the team assessed different multiplex primer schemes and wastewater sample types — influent wastewater and primary sludge⁵. Accessible, rapid, and cost-effective workflows to detect the emergence of variants of concern (VoC) within municipal areas would be a powerful tool during any infectious disease outbreak. Using a MinION to deliver real-time results, the team identified the use of influent wastewater and a 400 bp primer scheme as optimal conditions to accurately detect VoC within SARS-CoV-2 circulating in the Metro Vancouver region of British Columbia (Figure 3). The rapid turnaround time of approximately three days (from sampling to data generation), low capital cost, and high portability of nanopore sequencing, combined with the highly multiplexed tiling PCR sequencing approach, showed great promise in wastewater surveillance to complement genomic epidemiology efforts.

Figure 3: SARS-CoV-2 whole-genome sequencing coverage for three multiplex tiling PCR primer schemes (150 bp, 400 bp, and 1,200 bp), including breadth of genome coverage (A–C) and median depth of coverage across the genome (D–F). Use of the 400 bp tiling scheme on influent wastewater provided optimal breadth and depth of genome coverage. Image taken from Lin et al.⁵ and available under Creative Commons license (creativecommons.org/licenses/by/4.0).