Abstract
Fast surveillance strategies are needed to control the spread of new emerging SARS-CoV-2 variants and gain time for evaluation of their pathological potential. This was essential for the Omicron variant (B.1.1.529) that replaced the Delta variant (B.1.617.2) and is currently the dominant SARS-CoV-2 variant circulating worldwide. RT-qPCR strategies complement whole genome sequencing, especially in resource lean countries, but mutations in the targeting primer and probe sequences of new emerging variants can lead to a failure of the existing RT-qPCRs. Here, we introduced an RT-qPCR platform for detecting the Delta- and the Omicron variant simultaneously using a degenerate probe targeting the key ΔH69/V70 mutation in the spike protein. By inclusion of the L452R mutation into the RT-qPCR platform, we could detect not only the Delta and the Omicron variants, but also the Omicron sub-lineages BA.1, BA.2 and BA.4/BA.5. The RT-qPCR platform was validated in small- and large-scale. It can easily be incorporated for continued monitoring of Omicron sub-lineages, and offers a fast adaption strategy of existing RT-qPCRs to detect new emerging SARS-CoV-2 variants using degenerate probes.
Introduction
The World Health Organization designated the SARS-CoV-2 variant B.1.1.529 as a variant of concern (VOC) with the name ‘Omicron’ on November 26th, 2021(1). Omicron, first identified in South Africa and Botswana in November 2021, has spread rapidly worldwide and is the current dominant strain in at least 166 nations worldwide, replacing the Delta variant(2). Since its designation as a VOC, Omicron has continued to evolve, leading to descendent lineages, termed BA.1-5, with different genetic constellations of mutations (1).
The Omicron variant is exceptional for carrying a large number of mutations not seen in combination before, including more than 40 mutations in the spike protein (3,4). The spike protein is the main target of antibodies generated by infection and of many vaccines widely administered, and these mutations are therefore associated with increased transmissibility (5) and immune escape (6,7).
Initially, the BA.1 sub-variant was the dominant circulating version of Omicron. However, since the beginning of 2022, the proportion of BA.2 sampled by whole genome sequencing (WGS) has increased dramatically, making BA.2 the dominant SARS-CoV-2 virus globally (2). BA.1 and BA.2 differ by 19 signature mutations located in the N-terminal and in the receptor-binding domain (RBD) of the spike gene, which makes them as different as some other major variants, e.g. the diversity between the original Wuhan and the Alpha variant in the spike gene is less than between BA.1 and BA.2 (3,4). One notable difference between BA.1 and BA.2 is that BA.2 lacks the characteristic S-gene target failure (SGTF)-causing deletion, ΔH69/V70. This key mutation, a 2-amino acid deletion, has appeared in multiple SARS-CoV-2 variants, including the Alpha variant (B.1.1.7) (8). BA.4 and BA.5 have an identical BA.2-like constellation of mutations in the spike protein, but carry additional signature mutations including S: ΔH69/V70 and S: L452R. The differences between BA.4 and BA.5 instead lie outside the spike gene. The first cases of BA.4 and BA.5 were reported from the Republic of South Africa in January 2022 (2) and subsequently have been found in at least 60 countries (2). In Denmark, 1,091 patient samples infected with BA.4 and 6,685 infected with BA.5 have been confirmed by whole genome sequencing (WGS) per 22nd of June, 2022 (9). BA.4 and BA.5 display higher transmission than other Omicron sub-variants and evade immunity from natural infections or vaccines to a higher extent than BA.1 and BA.2, although not completely, enabling BA.4 and BA.5 to infect people who were immune to prior sub-lineages of Omicron and other variants (10). As Omicron continuously evolves under immune pressure, it may develop mutations that would undermine most broadly-neutralizing therapeutic antibodies.
Denmark has one of the highest SARS-CoV-2 Reverse Transcription Real-time PCR (RT-qPCR) testing and WGS capacities in the world (11). As part of this, a flexible RT-qPCR platform termed Variant-PCR was developed for identification and monitoring of circulating SARS-CoV-2 variants of concern (VOCs) in near real-time, enabling rapid decision-making about testing, contact tracking and isolation (12). The Variant-PCR platform utilizes signature mutations or combinations hereof that allow differentiation of the most critical VOCs by detection of four signature mutations in the SARS-CoV-2 spike protein: ΔH69/V70, L452R, E484K and N501Y (12).
We designed a degenerate probe for the detection of the ΔH69/V70 key mutation in both Alpha and Omicron variants. By including additional results from L452R RT-qPCR, we were able to distinguish between the BA.1 and BA.2 sub-lineages, the Delta variant, and furthermore detect BA.4 and BA.5 as they emerged, although differentiation between BA.4 and BA.5 was not possible as they share the same spike protein mutations. Although BA.4 emerged before BA.5 in Denmark, BA.5 is now the dominant SARS-CoV-2 variant (9). Validation was performed in small-scale and performance was confirmed in large-scale diagnostic settings after adaptation of the Variant-PCR to include the degenerate ΔH69V/70 probe, using paired WGS consensus genomes.
This study proves that a flexible platform such as the Variant-PCR platform can be adapted quickly and easily with a degenerate probe to monitor and differentiate existing and upcoming SARS-CoV-2 variants as part of a national testing program to help policy-makers to make public health decisions.
Materials and methods
Ethics
Exemption for review by the ethical committee system and informed consent was given by the Committee on Biomedical Research Ethics – Capital region in accordance with Danish law on assay development projects.
Samples and controls
TWIST Synthetic SARS-CoV-2 RNA control (MT103907 England/205041766/2020) were purchased from TWIST bioscience in a known concentration (approx. 1.000.00 0copies/µl) and used as PCR standard Alpha (B.1.17) in 1:10 dilution series. The average Ct-values and standard deviations were calculated based on biological duplicates.
Patient samples infected with SARS-CoV-2 WT, Alpha variant B.1.1.7, Delta variant B.1.617.2 and Omicron variant B.1.529 were obtained from the Danish National Biobank to verify whether the RT-qPCR assays performs with all circulating VOCs (20). In short, throat swabs were collected and tested SARS-CoV-2 positive from community testing facilities (Test Center Denmark), which form part of the Danish national testing program (11). Primary diagnosis of SARS-CoV-2 infection was made by E-Sarbeco RT-qPCR (13).
Positive controls for large-scale RT-qPCR consisted of three heat-activated (56°C for 45 minutes) Danish virus isolates covering four key mutations present in the Alpha variant B.1.1.7 (E12), Delta variant B.1.617.2 (H12) and Beta variant B.1.351 (F12). Virus isolation and cultivation was performed as previously described (12). Sterile 1x PBS pH 7.2 (Gibco) was used as negative controls. Positive and negative controls were run in parallel with selected clinical samples throughout RNA isolation and RT-qPCR as described before (12).
RNA extraction
For small-scale experiments, viral genomes from throat swabs collected in 1xPBS (200 µL) were isolated using MagNa Pure 96 nucleic acid extraction system (Roche) with reagents from the MagNa Pure 96 DNA and Viral NA Small Volume kit with 200 µL sample in 1xPBS as input and 100 µL elution. As positive control material, supernatant from SARS-CoV-2 infected cells (120 µL) were mixed with 120 µL of MagNA Pure Lysis Buffer (Roche) followed by extraction as small-scale SARS-CoV-2 patient samples. Positive control RNA was stored at -80°C until use. For large-scale SARS-CoV-2 patient screening (Variant-PCR), viral RNA was extracted using Beckman Coulter RNAdvance Viral Reagent kit with 200 µl sample in 1xPBS as input and 50 µl Nuclease-free water for elution on Beckman Coulter Biomek i7 automated workstations.
RT-qPCR assays
The sequences of the primers and probes listed in Table 1 were synthesized and HPLC-purified by Biosearch Technologies, Denmark. To perform allelic discrimination analysis, two probes were designed for each key mutation in the spike gene: one detecting the wildtype nucleotide sequence and one detecting the mutation, based on sequence alignments using Geneious Prime 2021.0 (12). The assay targeting the E gene (E-Sarbeco) was used as unspecific control to determine the presence of the SARS-CoV-2 and estimate the viral load (13).
The reaction mixtures for the RT-qPCR experiments were prepared as follows:
For each 25 µL reaction in 96-well format used in small-scale experiment, 5 µL extracted RNA were added to 20 µL reaction mix containing 12.5 µL Luna® Universal Probe One-step RT-qPCR reaction buffer (New England Biolabs Inc), 1.25 µL Luna® Warmstart RT Enzyme mix, primers and probes (100 µM, see volumes in Table 1) and nuclease-free water.
For each 15 µL reaction in 384-well format used for large-scale screening, 5 µL extracted RNA were added to 10 µL reaction mix containing 7.5 µL Luna® Universal Probe One-step RT-qPCR reaction buffer (New England Biolabs Inc), 0.75 µL Luna® Warmstart RT Enzyme mix, primers and probes (100 µM, see volumes in Table 1) and nuclease-free water. The setup in 384-well mastermix plates were arranged in a quadratic pattern enabling analysis of each patient sample with four different mastermixes in parallel, using first quadrate to detect Δ69/70, the second to detect L452R, the third to detect E-Sarbeco and the final well was not in use (water). This allowed for easy transfer of purified RNA from a 96-well format used during extraction to a 384-well mastermix plate for PCR (e.g. isolated RNA in A1 in the 96-well template plate was transferred to A1, A2, B1 and B2 of the mastermix plate).
The PCR program consists of reverse transcription at 55°C for 10 min., initial denaturation at 95°C for 3 min., followed by 45 cycles of denaturation and annealing/extension at 95°C for 15 sec., and 58°C for 30 sec., respectively, regardless of PCR plate format.
All RT-qPCR reactions were performed in a calibrated Bio-Rad CFX real-time PCR instrument. The raw data was analysed with the Bio-Rad CFX Maestro Software using a predefined threshold cut-off value of 100 RFU as a quality step, in case one of the probes in the allelic discrimination pair failed. Recorded Ct-values and end-RFU were calculated in Maestro version 5.2.8.222 and exported for further data analysis. For the mutations ΔH69/V70 and L452R, detection was based on allelic discrimination where the end-RFU values were utilized to determine the presence of a mutation. A sample is considered positive based on the following criteria: Ct-values between 10-38 and end-RFU > 200 at Ct = 45.
Whole Genome Sequencing
Identities of mutation signatures were confirmed by Whole genome sequences generated by The Danish COVID-19 Genome Consertium (DCGC) from PCR-positive samples as described by (12).
Results
A degenerate ΔH69/V70 probe is specific for Omicron detection
We evaluated the analytical performance characteristics of the ΔH69/V70 RT-qPCR included in our Variant-PCR for the qualitative detection of Omicron in clinical samples. First, SARS-CoV-2 positive clinical samples containing Omicron (B.1.529), Alpha (B.1.1.7) as positive control, as well as Wuhan (SARS-CoV-2 WT), Delta (B.1.617.2) and negative SARS-CoV-2 patient samples as negative controls, were included into the ΔH69/V70 RT-qPCR (Fig 1A). The presence of SARS-CoV-2 was confirmed in positive samples by E-Sarbeco RT-qPCR (13) multiplexed with the ΔH69/V70 RT-qPCR (Fig 1B).
(A) Detection of SARS-CoV-2 variants with the original probe, including SARS-CoV-2 alpha variant as positive control and neg. SARS-CoV-2 patient samples and a water control as negative controls. (B) Samples were tested as multiplex RT-qPCRs; E-Sarbeco in parallel with the ΔH69/V70 RT-qPCR. Error bars in (A)-(B) indicate SEM for three biological replicates.
The original ΔH69/V70 RT-qPCR (probe design based on the SARS-CoV-2 alpha variant sequence) was able to detect the ΔH69/V70 mutation in Omicron confirmed samples (Fig 1A). However, the end fluorescence intensity was reduced about five times in Omicron BA.1 (265 RFU) compared to samples with SARS-CoV-2 Alpha (1407 RFU). Sequence analysis revealed a single nucleotide substitution from C-to-T at nucleotide position 21696 in the Omicron BA.1 sequence, resulting in a mismatch in the middle of ΔH69/V70 probe binding region, which led to the observed reduction of the fluorescence intensity.
In order to detect the ΔH69/V70 mutation in all known SARS-CoV-2 variants, we designed a degenerate ΔH69/V70 probe (Fig 2). Replacing the original probe targeting ΔH69/V70 with the degenerate probe led to almost a full recovery of the end fluorescence intensity (Fig 2A) in both SARS-CoV-2 variants (Alpha and Omicron) (Fig 2A). Specificity of the degenerate probe was confirmed by the lack of signal from SARS-CoV-2 variants lacking the H69/V70 deletion (SARS-CoV-2 WT, Delta, Omicron BA.2), as well as for the negative controls (Fig2A). In order to discriminate between BA.1 and BA.2 sub-variants, we included a WT probe in the ΔH69/V70 RT-qPCR. All SARS-CoV-2 variants lacking the ΔH69/V70 (SARS-CoV-2 WT, Delta, Omicron BA.2) were detected correctly by the H69/V70 WT probe (Fig 2B). Moreover, the fluorescence intensity was slightly reduced for the degenerate probe compared to the original probe targeting the ΔH69/V70 present in the SARS-CoV-2 Alpha variant (synthesized standard TWIST control) (Suppl, Fig1), but the sensitivity was not reduced, as judged by dilutions of a commercial standard (Tab.1; Suppl, Fig1).
(A) Detection of SARS-CoV-2 variants with a degenerate probe targeting the ΔH69/V70. The SARS-CoV-2 Alpha- and BA.1 variants were included as positive controls and neg. SARS-CoV-2 patient samples and a water control as negative controls. (B) Detection of SARS-CoV-2 variants with the WT probe. SARS-CoV-2 WT, Delta and BA.2 variants were included as positive controls and neg. SARS-CoV-2 patient samples and a water control as negative controls.
Large-scale performance confirmation of allelic discrimination RT-qPCR with degenerate ΔH69/V70 probe
The ΔH69/V70-WT RT-qPCR with the degenerate probe was validated on 745 PCR-positive SARS-CoV-2 patient samples collected from 3rd to 16th of January 2022 as part of the Danish national testing program. This period was selected since Omicron BA.1 and BA.2 sub-lineages were present in approximately equal proportions (9).
The H69/V70 deletion is not present in the SARS-CoV-2 variants Delta and Omicron BA.2; therefore the L452R RT-qPCR was included into the RT-qPCR large-scale screening platform (12,14). Running two RT-qPCRs, each directed against one specific key mutation, e.g. ΔH69/V70 and L452R, made it possible to detect four different variants (Table 2).
Validation was performed by pairwise comparison of results from the Variant-PCR with WGS using the WGS consensus genome from the same samples as reference standard. We obtained valid RT-qPCR results and a consensus genome sequence from each of the 745 selected SARS-CoV-2 PCR-positive patient samples (Fig 3). In total, 339 patient samples were characterized with BA.1 infection while 399 samples was identified as BA.2 by WGS analysis, showing a nearly even distribution between BA.1 and BA.2 genomes. The Delta genome was found in the last seven patient samples.
(A) Allelic discrimination analysis to differentiate between the H69/V70 deletion and the WT sequence. (B) Allelic discrimination analysis to differentiate between the 452R mutation and the L452 WT sequence.
The ΔH69/V70 mutation was detected by the degenerate ΔH69/V70 probe in 339 patient samples, which correlated 100% with Omicron BA.1 WGS confirmed samples (Fig 3). Furthermore, 406 SARS-CoV-2 positive clinical samples had a WT signal being either of Delta or Omicron BA.2 origin. In the L452R RT-qPCR, seven samples were found positive for the 452R mutation characteristic for the Delta variant in correlation with WGS results. The L452-WT sequence was detected in 738 out of the 745 patient samples corresponding to the total consensus sequences identified as Omicron by WGS, regardless of BA.1 and BA.2 sub-lineage identity. Thus, none of the samples differed between the degenerate ΔH69/V70-WT RT-qPCR and WGS, demonstrating high specificity (>99.9%) and high sensitivity (>99.9%) in large-scale as well as small-scale.
In summary, we developed a RT-qPCR system for rapid differentiation of Omicron BA.1 and BA.2 in large-scale, validated by comparison of RT-qPCR and WGS data from 745 patient samples analyzed in parallel.
Large-scale performance confirmation of BA.4 and BA.5 Omicron variants
The Danish national SARS-CoV-2 genomic surveillance system identified the first case of BA.4 in a sample from 2nd of March, 2022. On 15th of April, 2022, the first case of BA.5 case was detected. At this point in time, BA.2 accounted for more than 99% of all sequenced variants in Denmark (9). The prevalence of BA.4 increased at a linear rate from 0.1% to 4.7% during the period from 11th of April to 1st of June, 2022 with 327 identified BA.4 genomes in total. In contrast to BA.4, the prevalence of BA.5 increased more rapidly and on 1st of June, 2022, 30.7% of all sequenced genomes from SARS-CoV-2 positive throat samples were confirmed as BA.5 consensus sequences increasing the total number to 1155 identified BA.5 cases (9). Since BA.5 clearly was heading to replace BA.2 as the dominant SARS-CoV-2 lineage in Denmark, we investigated if our adapted RT-qPCR platform was able to detect Omicron BA.4 and BA.5 sub-variants as a group, since this RT-qPCR per se will not able to discriminate between them due to identical genomic sequences in the spike region.
Data obtained from WGS revealed that 92.5% of BA.4 consensus sequences (184/199) and 97.6% of BA.5 consensus sequences (328/336) carried the ΔH69/V70 mutation. The prevalence of the mutation 452R is 93% and 98.8% in BA.4 and BA.5, respectively, which confirmed that both BA.4 and BA.5 carry both signature mutations. Based on alignment of unique sequences, 86 BA.4 and 53 BA.5, respectively, the primers and probes in our L452R RT-qPCR had a complete match in all BA.4 sequences. Of the 53 analyzed BA.5 sequences, one showed an A-to-T mutation in the middle of the L452R probe-binding region whereas primer-binding regions were found to be conserved. The BA.5 genomes had a perfect match to the primers and probe in the degenerate ΔH69/V70 RT-qPCR whereas a single nucleotide mismatch, an A-to-G substitution located in the middle of the ΔH69/V70 forward primer (ACATTCAACTCGGGACTTGTTCT), was observed in 10 of the 86 investigated BA.4 sequences (12%), which may influence the performance of the RT-qPCR reaction.
Validation of the allelic ΔH69/V70-WT and the L452-WT discrimination RT-qPCRs were performed on 66 SARS-CoV-2 positive patient samples collected from 24th to 26th of May, 2022 at Danish community test facilities. Of these selected samples, 28 viral genomes were characterized as BA.4 and 38 as BA.5 by WGS. Results from our adapted Variant-PCR showed that both key mutations were detected in combination in all selected samples as expected; e.g. all samples were infected with a SARS-CoV-2 virus being ΔH69/V70-L452R double-positive in accordance with either a BA.4 or BA.5 viral genotype (Fig 4).
(A) Allelic discrimination analysis to differentiate between the ΔH69/V70 mutation and the WT sequence. (B) Allelic discrimination analysis to differentiate between the 452R mutation and the L452 WT sequence.
Similar Ct-values were obtained in the E-Sarbeco RT-qPCR, indicating that all three RT-qPCRs perform equally well with BA.4 and BA.5 variants, confirming high specificity of the degenerate ΔH69/V70 probe for an Alpha VOC-like sequence, and providing evidence for high specificity of this method in large-scale studies. Furthermore, we observed no pronounced differences in amplification and detection of BA.4 viral genomes with a mismatch mutation in the sequence recognized by the ΔH69/V70 forward primer compared to patient samples with a conserved sequence (Fig 4).
In summary, our adapted Variant-PCR could rapidly detect Omicron BA.4 and BA.5 validated by comparison of RT-qPCR and WGS data from 66 patient samples analyzed in parallel.
Discussion
Control of the pandemic caused by SARS-CoV-2 is challenged by the emergence of new variants, emphasizing the need for rapid detection methods that complement WGS in order to support control of transmission chains and assist public health response and decision-making. RT-qPCR is widely deployed in diagnostic virology (15). In case of a public health emergency, laboratories can rely on this robust technology to establish new diagnostic tests for screening within their routine services before commercial assays become available. RT-qPCR strategies are of special interest in resource lean countries to complement WGS, the golden standard for SARS-CoV-2 variant surveillance. However, mutations in the targeting primer and probe sequences of new emerging variants can lead to a failure of the existing RT-qPCRs.
Currently, most commercial kits for SARS-CoV-2 variant determination utilize a so-called S-gene target failure phenomenon(16), where a detected variant positive for the N and ORF1ab genes, but negative for S gene is interpreted as Omicron, due to the large number of mutations in the S gene of this particular variant. The BA.1 sub-lineage showed S-gene target failure in RT-qPCRs due to the presence of the ΔH69/V70 mutation, however, an S-gene target failure strategy may not be able to detect the Omicron BA.2 lineage, which is currently the dominant SARS-CoV-2 virus circulating worldwide, as BA.2 is missing this particular deletion.
Recently, several single one-step RT-qPCRs reported for the specific detection of Omicron in clinical and environmental samples have been developed (17–19). As Omicron continues to spread and evolve into several lineages and sub-lineages, these methods suffer, as they are not able to differentiate between multiple Omicron sub-lineages, although they have been described as highly specific and sensitive for Omicron.
The allelic discrimination RT-qPCR we propose using four probes targeting two signature mutations in parallel has a very high accuracy enabling detection of known VOCs almost in real-time (within 24h) after sample collection. We provide evidence for simultaneous detection of the L452R mutation and ΔH69/V70 deletion in spike protein in large-scale screening, using a degenerate probe to detect and differentiate between circulating Omicron- and Delta variants. To the best of our knowledge, this is the first RT-qPCR platform for specific detection and discrimination of Omicron sub-variants using a degenerate probe. This strategy enables detection of a key deletion, ΔH69/V70, regardless of a single nucleotide substitution or not, and hence, increases the robustness of the RT-qPCR. The samples for large-scale performance evaluation were chosen to challenge the robustness of these RT-qPCRs since lineage replacement is expected to occur during a pandemic transition period; the BA.1 sub-variant and its signature mutation, ΔH69/V70 with a single nucleotide substitution, was replaced by BA.2 with a wildtype genotype at both targets, ΔH69/V70 and L452R. Later, the appearance of BA.4 and BA.5 variants was observed in an increasing number of patient samples being positive for both key mutations. In fact, Omicron being the predominant variant worldwide, negative results in allelic discrimination RT-qPCR, but positive in other SARS-CoV-2 methods targeting conserved elements of the SARS-CoV-2 genome such as E-Sarbeco, may indicate the emergence of novel potential VOCs with S-gene mutations in target regions hindering recognition by the primers and/or probes.
In summary, after adapting our Variant-PCR platform with a degenerate ΔH69/V70 probe, we were able to detect the four most common Omicron variants in circulation, BA.1, BA.2, BA.4 and BA.5, and the Delta VOC, by combined analysis of two different signature mutations in parallel, ΔH69/V70 and L452R, validated by a comparison of the RT-qPCR and WGS data. Furthermore, this method enables us to differentiate between Omicron and Delta VOCs, between BA.1 and BA.2 sub-lineages, and between BA.2 and emerging BA.4/BA.5 sub-lineages. If SARS-CoV-2 continues to evolve very contagious VOCs with potential to escape humeral immune response, which may lead to a new wave of COVID-19 infection, monitoring and screening might have to last for a prolonged period of time, making the use of degenerate probes a useful strategy to adapt existing RT-qPCRs for SARS-CoV-2 detection.
Data Availability
All data produced in the present work are contained in the manuscript
Supporting information
(A) Dilution of the TWIST contol standard (SARS-CoV-2 variant Alpha) to a concentration of 10−3 copies/µl detected by the original and the degenerate probe targeting the ΔH69/V70. (B) Dilution row of the TWIST contol standard (SARS-CoV-2 variant Alpha) detected by the original and the degenerate probe targeting the ΔH69/V70.
Acknowledgements
We would like to extend our gratitude to Halenur Bayhan Ari, Kristina Finneisen and Caroline Mølsted Benfeldt for their assistance and technical support.
Footnotes
