Baseline Sequencing Surveillance of Public Clinical Testing, Hospitals, and Community Wastewater Reveals Rapid Emergence of SARS-CoV-2 Omicron Variant of Concern in Arizona, USA

Abstract

The adaptive evolution of SARS-CoV-2 variants is driven by selection for increased viral fitness in transmissibility and immune evasion. Understanding the dynamics of how an emergent variant sweeps across populations can better inform public health response preparedness for future variants. Here, we investigated the state-level genomic epidemiology of SARS-CoV-2 through baseline genomic sequencing surveillance of 27,071 public testing specimens and 1,125 hospital inpatient specimens diagnosed between November 1, 2021, and January 31, 2022, in Arizona. We found that the Omicron variant rapidly displaced Delta variant in December 2021, leading to an "Omicron surge" of COVID-19 cases in early 2022. Wastewater sequencing surveillance of 370 samples supported the synchronous sweep of Omicron in the community. Hospital inpatient COVID-19 cases of Omicron variant presented to three major hospitals 10.51 days after its detection from public clinical testing. Nonsynonymous mutations in nsp3, nsp12, and nsp13 genes were significantly associated with Omicron hospital cases compared to community cases. To model SARS-CoV-2 transmissions across the state population, we developed a scalable sequence network methodology and showed that the Omicron variant spread through intracounty and intercounty transmissions. Finally, we demonstrated that the temporal emergence of Omicron BA.1 to become the dominant variant (17.02 days) was 2.3 times faster than the prior Delta variant (40.70 days) or subsequent Omicron sublineages BA.2 (39.65 days) and BA.5 (35.38 days). Our results demonstrate the uniquely rapid sweep of Omicron BA.1. These findings highlight how integrated public health surveillance can be used to enhance preparedness and response to future variants. IMPORTANCE SARS-CoV-2 continues to evolve new variants throughout the pandemic. However, the temporal dynamics of how SARS-CoV-2 variants emerge to become the dominant circulating variant is not precisely known. Genomic sequencing surveillance offers unique insights into how SARS-CoV-2 spreads in communities and the lead-up to hospital cases during a surge. Specifically, baseline sequencing surveillance through random selection of positive diagnostic specimens provides a representative outlook of the virus lineages circulating in a geographic region. Here, we investigated the emergence of the Omicron variant of concern in Arizona by leveraging baseline genomic sequence surveillance of public clinical testing, hospitals, and community wastewater. We tracked the spread and evolution of the Omicron variant as it first emerged in the general public, and its rapid shift in hospital admissions in the state health system. This study demonstrates the timescale of public health preparedness needed to respond to an antigenic shift in SARS-CoV-2.

Keywords: Omicron variant; SARS-CoV-2; emergence dynamics; hospital-associated mutations; wastewater surveillance.

FIG 1

SARS-CoV-2 variants circulating in Arizona from November 1, 2021 – January 31, 2022. (A) Arizona county map showing the geographic distribution of specimens sequenced for baseline surveillance. (B) Sublineage frequencies of circulating SARS-CoV-2 genomes shown by Arizona counties. (C) Circulating SARS-CoV-2 variants in Arizona detected through community and hospital baseline surveillance shown by week. (D) Number of COVID-19 cases per day in Arizona, USA. The map in panel A was generated iwth the open source software package plotly.

FIG 2

Characteristics of SARS-CoV-2 hospital surveillance. (A) Number of COVID-19 hospitalizations and reported COVID-19 diagnostic test positivity rate in Arizona. (B) Prevalence of Omicron variant in community and hospital surveillance samples. Nonlinear (sigmoidal) curves were fitted to data. The number of days between curve midpoints (50% proportion) is shown. (C) SARS-CoV-2 viral load, as measured by diagnostic RT-PCR, between Delta and Omicron variants for community and hospital surveillance samples. Statistical significance was assessed by Mann-Whitney U test. (D) SARS-CoV-2 genome map shows locations of amino acid substitutions found to be statistically associated (P < 0.05) with hospital or community specimens for Delta (top) and Omicron (bottom) variants. Statistical significance was assessed by Fisher’s exact test with multiple testing correction using the Benjamini/Hochberg algorithm.

FIG 3

Network transmission analysis of SARS-CoV-2 genomes. (A) Overview of the bioinformatic workflow used to generate transmission networks. (B) Visualization of the largest connected Omicron BA.1 cluster of the generated network. Nodes are colored by county, generated from participant provided zip code. (C) Incoming and outgoing edges from the four most populous counties in the entire Omicron network is shown. Edges are colored by their source county and have arrowheads directed toward their target county. Outgoing edges contain an inset bar colored by the county of the targeted node.

FIG 4

Wastewater surveillance detected Omicron variant sweeping through the community. (A) Map of wastewater collection sites monitored in the Greater Tempe area, Arizona. (B) SARS-CoV-2 viral load of wastewater samples measured by RT-PCR (line charts) and variant abundance of next generation sequencing data determined by Freyja (heatmaps) is shown for each collection site over time. (C) Representative technical replicates of next generation sequencing and variant analysis (Freyja) from three wastewater samples. Three independent extractions were performed on the wastewater samples. (D) SARS-CoV-2 genome map showing the 15 Delta- and 17 Omicron-specific mutations analyzed. (E) The number of Delta-specific mutations is indicated by the circle size, and relative abundance of the Delta-specific mutations is indicated by the color intensity for each wastewater sample. (F) The number of Omicron-specific mutations is indicated by the circle size, and relative abundance of the Omicron-specific mutations is indicated by the color intensity for each wastewater sample. The map in panel A was made using the arcGIS web interface.

FIG 5

Wastewater detection of Omicron variant using a digital PCR assay. (A) The Omn143 assay design is shown, dPCR primers (arrows) and probe (box) sequences are indicated. Sequence alignment shows Omicron, Delta, Alpha, Beta and Gamma VOCs. Sequence differences in the probe binding is highlighted in red. (B) Representative dPCR results of Omn143 assay using water negative control, synthetic DNA constructs of Delta and Omicron. Three independent experiments were performed. (C) Omicron variant abundance of wastewater samples by RT-dPCR. Gray boxes indicate no sampling at that time point.

FIG 6

Emergence rate of Delta, Omicron BA.1, BA.2 and BA.5. Nonlinear (sigmoidal) curves were fitted for each variant using 5-day sliding window. The time it took a variant to increase from 10% to 90% case proportion (Δ) and Hillslope coefficient (n_H) is indicated. (Delta: n = 4072, Daily Average = 26.27, n_H 95% CI = 0.045 – 0.052; Omicron BA.1: n = 27396, Daily Average = 304.40, n_H 95% CI = 0.11 – 0.13; Omicron BA.2: n = 23971, Daily Average = 171.22, n_H 95% CI = 0.046 – 0.056; Omicron BA.5: n = 4597, Daily Average = 39.97, n_H 95% CI = 0.050 – 0.058). Asterisk indicates calculation of emergence to 80% prevalence.