Application of a high-resolution melt assay for monitoring SARS-CoV-2 variants in Burkina Faso and Kenya

Abstract

The rapid emergence and global dissemination of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) highlighted a need for robust, adaptable surveillance systems. However, financial and infrastructure requirements for whole-genome sequencing mean most surveillance data have come from higher-resource geographies, despite unprecedented investment in sequencing in low- and middle-income countries (LMICs). Consequently, the molecular epidemiology of SARS-CoV-2 in some LMICs is limited, and there is a need for more cost-accessible technologies to help close data gaps for surveillance of SARS-CoV-2 variants. To address this, we have developed two high-resolution melt (HRM) curve assays that target variant-defining mutations in the SARS-CoV-2 genome, which give unique signature profiles that define different SARS-CoV-2 variants of concern (VOCs). Extracted RNA from SARS-CoV-2-positive samples collected from 205 participants (112 in Burkina Faso, 93 in Kenya) enrolled in the MALCOV study (Malaria as a Risk Factor for COVID-19) between February 2021 and February 2022 were analyzed using our optimized HRM assays. With next-generation sequencing on Oxford Nanopore MinION as a reference, two HRM assays, HRM-VOC-1 and HRM-VOC-2, demonstrated sensitivity/specificity of 100%/99.29% and 92.86%/99.39%, respectively, for detecting Alpha, 90.08%/100% and 92.31%/100% for Delta, and 93.75%/100% and 100%/99.38% for Omicron BA.1. The assays described here provide a lower-cost approach to conducting molecular epidemiology, capable of high-throughput testing. We successfully scaled up the HRM-VOC-2 assay to screen a total of 506 samples from which we were able to show the replacement of Alpha with the introduction of Delta and the replacement of Delta by the Omicron variant in this community in Kisumu, Kenya.IMPORTANCEThe rapid evolution of the severe acute respiratory syndrome coronavirus 2 variants of concern (VOCs) demonstrated the need for accessible surveillance tools so all communities can conduct viral surveillance. Sequencing, the gold standard, is still a largely inaccessible methodology in low-resource settings. Here, we present a quick, low-cost tool to screen for the common VOCs, designed to support surveillance efforts in low-resource settings. This tool was used to screen samples from Burkina Faso and Western Kenya throughout the pandemic. We show through comparison to sequencing that our assay can generate highly similar data on the different variants circulating in a population, therefore showing the effectiveness of our tool. While not a replacement for sequencing, we present a method of screening and prioritizing samples for further investigation and reduce overburdening sequencing capacity. Our findings provide insight into one potential tool that could be further applied to pathogen screening in the absence of robust sequencing infrastructure.

Keywords: Africa; Burkina Faso; COVID-19; HRM; Kenya; SARS-CoV-2; diagnostics; surveillance; variants of concern.

Fig 1

SARS-CoV-2 variants identified in Burkina Faso using different methods. (A) Number of samples collected in Burkina Faso from July 2021 to January 2022 and the variants that were identified by nanopore sequencing. (B) Number of samples in the Burkina Faso cohort and the variant identified by using the HRM-VOC-1 assay. Negative results represent those with no amplification observed; undetermined samples had amplification but no identifiable VOC peak. (C) Number of samples in the Burkina Faso cohort and the variant identified by using the HRM-VOC-2. Negative results are those where no amplification was observed, and undetermined results are those with a control peak without an identifiable VOC peak.

Fig 2

SARS-CoV-2 variants identified in Kenya using different methods. (A) Number of samples collected in Kenya from February 2021 to February 2022 and the variants that were identified by nanopore sequencing. (B) Number of samples in the Kenyan cohort and the variant identified by using the HRM-VOC-1 assay. Negative results represent those with no amplification observed; undetermined samples had amplification but no identifiable VOC peak. (C) Number of samples in the Kenyan cohort and the variant identified using the HRM-VOC-2. Negative results represent those with no amplification observed; undetermined samples had amplification but no identifiable VOC peak.

Fig 3

Time series of variants identified by HRM-VOC-2 when tested on 506 SARS-CoV-2 PCR-positive samples collected in Kenya throughout the study period. This is a combined data set including the 86 successfully sequenced samples. Undetermined represents samples that produced a control peak but no identifiable VOC peaks. Invalid represents samples where there was no control peak. No data represents weeks of the year where no positive samples had been collected.