SpikeSeq: A rapid, cost efficient and simple method to identify SARS-CoV-2 variants of concern by Sanger sequencing part of the spike protein gene

Abstract

In 2020, the novel coronavirus, SARS-CoV-2, caused a pandemic, which is still raging at the time of writing this. Here, we present results from SpikeSeq, the first published Sanger sequencing-based method for the detection of Variants of Concern (VOC) and key mutations, using a 1 kb amplicon from the recognized ARTIC Network primers. The proposed setup relies entirely on materials and methods already in use in diagnostic RT-qPCR labs and on existing commercial infrastructure offering sequencing services. For data analysis, we provide an automated, open source, and browser-based mutation calling software (https://github.com/kblin/covid-spike-classification, https://ssi.biolib.com/covid-spike-classification). We validated the setup on 195 SARS-CoV-2 positive samples, and we were able to profile 85% of RT-qPCR positive samples, where the last 15% largely stemmed from samples with low viral count. We compared the SpikeSeq results to WGS results. SpikeSeq has been used as the primary variant identification tool on > 10.000 SARS-CoV-2 positive clinical samples during 2021. At approximately 4€ per sample in material cost, minimal hands-on time, little data handling, and a short turnaround time, the setup is simple enough to be implemented in any SARS-CoV-2 RT-qPCR diagnostic lab. Our protocol provides results that can be used to choose antibodies in a clinical setting and for the tracking and surveillance of all positive samples for new variants and known ones such as Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1) Delta (B.1.617.2), Omicron BA.1(B.1.1.529), BA.2, BA.4/5, BA.2.75.x, and many more, as of October 2022.

Keywords: BA.1; BA.2; BA.2.75.x; BA.4/5; COVID-19; CoV; Contact tracing; Coronavirus; Mutations; Omicron; Profiling; SARS-CoV-2; Sequencing; Spike protein; Surveillance; Variants of concern.

Fig. 1

Profiling of selected SARS-CoV-2 variants based on key mutations from N440K to T716I in a single amplicon in the spike protein. Even without the near universal D614G mutation, all current and previous VOC variants have a unique mutation profile in this amplicon window, except for BA.4/5, which are commonly reported together. While not shown, the only mutation in the primer sequences in any VOC is on the likely inconsequential K417.

Fig. 2

Workflow of SpikeSeq typing of SARS-CoV-2 variant mutations. Time estimates are based on a single 96 well PCR plate and no automation. The total hands-on time is one hour and 30 min, and only requires the set up of an RT-PCR from already extracted RNA from a diagnostics lab followed by transfer of 1.5 μl of the product to a new 96 well PCR plate. The plate is subsequently sent for Sanger sequencing at a commercial provider. The total cost per sample including plastware, enzymes, transport, and sequencing is approx. 4 €. After the raw data is received, data analysis takes less than five minutes for 96 samples. If Sanger sequencing is available in-house, the time to result could be less than six hours.

Fig. 3

The sensitivity of the SpikeSeq method versus the Ct values of the diagnostic RT-qPCR test (n = 195). The number above each bar refers to the number of samples in that bin. Overall, Sanger sequencing data from 85% of samples were of sufficient quality for mapping. The non-template control is not depicted, as it did not produce a Sanger read that mapped to the reference genome. * : Ct values stem from a 2-step PCR where the first 7 cycles are not fluorescence registered, possibly meaning that the Sanger sequencing assay is even more sensitive than suggested by this figure.