Wastewater-Based Epidemiology and Long-Read Sequencing to Identify Enterovirus Circulation in Three Municipalities in Maricopa County, Arizona, Southwest United States between June and October 2020

Abstract

We used wastewater-based epidemiology and amplicon-based long-read high-throughput sequencing for surveillance of enteroviruses (EVs) in Maricopa County, Arizona, Southwest United States. We collected 48 samples from 13 sites in three municipalities between 18 June and 1 October 2020, and filtered (175 mL each; 0.45 µm pore size) and extracted RNA from the filter-trapped solids. The RNA was converted to cDNA and processed through two workflows (Sanger sequencing (SSW) and long-read Illumina sequencing (LRISW)) each including a nested polymerase chain reaction (nPCR) assay. We subjected the ~350 bp amplicon from SSW to Sanger sequencing and the ~1900-2400 bp amplicon from LRISW to Illumina sequencing. We identified EV contigs from 11 of the 13 sites and 41.67% (20/48) of screened samples. Using the LRISW, we detected nine EV genotypes from three species (Enterovirus A (CVA4, EV-A76, EV-A90), Enterovirus B (E14) and Enterovirus C (CVA1, CVA11, CVA13, CVA19 and CVA24)) with Enterovirus C representing approximately 90% of the variants. However, the SSW only detected the five Enterovirus C types. Similarity and phylogenetic analysis showed that multiple Enterovirus C lineages were circulating, co-infecting and recombining in the population during the season despite the SARS-CoV-2 pandemic and the non-pharmaceutical public health measures taken to curb transmission.

Keywords: Arizona; Enterovirus C; environmental monitoring; high-throughput nucleotide sequencing; human; wastewater-based epidemiological monitoring.

Figure 1

(A) Summary of the workflow used in this study. (B) The region of enterovirus genomes amplified in each of the assays. Abbreviation: LRI = long-read Illumina.

Figure 2

Relative abundance of enterovirus variants recovered and unambiguously identified on the four (0, 28, 70 and 105) sampling days using the SSW. Numbers in chart are percentages reflecting the number of sites where the different EV types were detected relative to the number of sites where EVs were detected and unambiguously typed on any specific sampling day. Please note that d0, d28, d70 and d105 mean EV types detected in samples collected on days 0, 28, 70 and 105, respectively. Additionally, note that 100% in d28 refers to the single CVA24 detected on that day (please see Table 2 and Table S1).

Figure 3

Relative abundance of enterovirus variants recovered on three (3) of the sampling days (0, 70 and 105) using assays 3a and 3b (LRISW). Numbers in chart are percentages reflecting the cumulative number of variants across all sites where the different EV types were detected on any specific sampling day. Note that the image above only displays variants recovered using assays 3a and 3b (LRISW). Particularly, the CVA24 and CVA13 recovered on days 28 and 105, respectively, using assay 2 (SSW) are not shown here. Note that d0, d70 and d105 mean EV types detected in samples collected on days 0, 70 and 105, respectively.

Figure 4

Maximum likelihood phylogenetic tree of CVA1 (A) VP4-VP1 [from assay 3a contigs] and (B) 2A-2C [from assay 3b contigs] genomic regions. The phylogenetic trees are based on an alignment of CVA1 contigs recovered from assay 3a and 3b (LRISW) in this study and those present in GenBank. The contigs are coded based on the sampling sites and days. Filled triangle, circle, square and diamond represent ST9-d0, ST7-d0, ST4-d0 and ST11-d105, respectively. ST5-d105 is denoted using empty diamond. Bootstrap support is shown if >50. Abbreviation: ST = site. Please note that the region amplified in the SSW is present in the contig used for making the tree in Figure 4A but not in Figure 4B. Specifically, all VP1 portions of amplicon 3b were removed from the contigs used to make the tree in Figure 4B.

Figure 5

Neighbor joining phylogenetic tree of Enterovirus Cs recovered in this study (A) VP4-VP1 [from assay 3a contigs] (B) VP1-2C [assay 3b contigs] and (C) 2A-2C [from assay 3b contigs] genomic regions. The phylogenetic trees are based on an alignment of Enterovirus C contigs recovered from assay 3a and 3b (LRISW) in this study. In Figure 5C, the CVA1 contigs that do not cluster with other CVA1s are indicated with black circles. Bootstrap support is shown if >50. Abbreviation: ST = site. Collapsed taxa denote all contigs that belong to the same EV type.

Figure 6

Maximum likelihood phylogenetic tree of CVA1 and CVA19 contigs recovered from assay 3b (LRISW) in this study (A) VP1-2C (B) VP1 only and (C) 2A-2C. In Figure 6C, the CVA1 contigs that do not cluster with other CVA1s are indicated with black circles. Bootstrap support is shown if >50. Abbreviation: ST = site. Collapsed taxa denote all contigs that belong to the same EV type.