1. Introduction
OPV is a live attenuated vaccine that protects well against poliomyelitis by stimulating both humoral and mucosal immunity. However, OPV is composed of three poliovirus Sabin strains that are genetically unstable and can revert to neurovirulence during replication in both cell cultures and vaccine recipients, leading to vaccine-associated paralytic poliomyelitis (VAPP) in vaccinees or their contacts, as well as the emergence of virulent circulating vaccine-derived poliovirus (cVDPV) strains that cause polio outbreaks [
1].
Since, in 2015, wild poliovirus type 2 was declared eradicated by the World Health Organization (WHO), the routine use of trivalent OPV (tOPV) was stopped to reduce the risk of VAPP and cVDPV. It was replaced by bivalent OPV (bOPV), which contains only Sabin 1 and 3 viruses, supplemented with at least one dose of inactivated poliovirus vaccine (IPV) [
2,
3]. IPV is known to produce inadequate mucosal immunity and, thus, the vaccinees receiving bOPV with IPV are vulnerable to being infected with type 2 poliovirus [
4,
5,
6]. Since the switch, cVDPV2 outbreaks increased significantly [
7]. To stop these outbreaks, monovalent OPV2 (mOPV2) was used, but it resulted in triggering new cVDPV2 outbreaks [
8,
9].
The main attenuating site in Sabin strains is located in domain V of the 5′ untranslated region (UTR), in which mutations determining reversion to virulence occur (in A
481G in the case of Sabin 2) [
10,
11,
12,
13]. The degree of attenuation is determined by the thermal stability of domain V: the stronger hairpin structure correlates with higher neurovirulence. The A
481G reversion occurs rapidly, within a week post-vaccination [
14,
15,
16]. For Sabin 2 virus, the initial steps of the de-attenuation also involve nucleotide changes U
398C in domain IV of 5′UTR and the amino acid I
143T mutation in VP1 capsid protein [
13].
Recently, two genetically stable novel OPV2 (nOPV2) vaccine candidates have been developed [
17,
18]. Both candidates have their domain V genetically stabilized by the replacement of all weak U-G and selected strong C-G nucleotide pairs with intermediately strong U-A pairs [
19], resulting in similar thermal stability and level of attenuation of domain V as in the original Sabin 2 strain. However, the reconstituted domain V is more genetically stable because it takes at least two simultaneous mutations for the base pair to be strengthened. In addition, the critically important cis-acting replicative element (cre) was relocated from the center of the genome to the 5′ UTR, protecting the reconstituted domain V from being recombined out. Finally, nOPV2 candidate 1 (nOPV2-c1) also contains D
53N and K
38R amino acid changes in 3D polymerase to improve replication fidelity and reduce the recombination rate, respectively [
18]. In November 2020, the WHO issued an Emergency Use Listing (EUL) recommendation for the nOPV2c1. This allows the implementation of the vaccine in countries affected by cVDPV2 outbreaks, and, currently, more than 200 million doses of nOPV2 have been used, with the anticipated safety profile confirmed [
20].
In addition to monitoring of the above mutations that may affect vaccine safety, it is also important to ensure that efficacy of the new vaccine is not compromised by genetic changes that can take place during virus growth in cell culture. Preliminary experiments revealed that growth in Vero cells may result in accumulation of E
295K mutation in the capsid protein VP1, which reduces immunogenicity. Additionally, VP1-I
143T and VP1-N
171D, which both modestly increase neurovirulence in transgenic mice, were also observed in vaccine lots and considered important to monitor [
21].
In this communication, we describe quantitative, multiplex, one-step reverse-transcriptase polymerase chain reactions (qmosRT-PCR) for detection and quantitation of VP1-I143T, VP1-N171D, and VP1-E295K, which may impact the neurovirulence or immunogenicity of the nOPV2.
The results demonstrate high specificity, linearity, and sensitivity from the assays, demonstrating that they are suitable for quality control of nOPV2 vaccine.
2. Materials and Methods
2.1. Plasmids and Viruses
Plasmids containing the entire genomes of nOPV2 VP1-N171D, and VP1-E295K amino acid mutants were prepared by Dr. Andrew Macadam’s group at the National Institute for Biological Standards and Control in the UK and the plasmid that contains the nOPV2 VP1-I143T genome was kindly provided by Professor Raul Andino from the University of California, San Francisco, CA, USA. These plasmids were used for the recovery the nOPV2c1 mutant viruses as described below.
Monovalent bulks of nOPV2 candidate 1 (nOPV2-c1) batches nPOL 2016C, nPOL 2018C, nPOL2 B0419, and nPOL2 B0519 as well as monovalent bulks of nOPV2 candidate 2 (nOPV2-c2) batches nPOL 2038C and nPOL 2056C and drug product nOPV2-c1 batches 2060119C, 2060219C, and 2060319C were provided by P.T. Bio Farma (Indonesia) and were used for quantification of the level of the mutant variants.
2.2. Recovery of nOPV2-c1 Mutant Variants
nOPV2-c1 mutant viruses were recovered from the plasmids containing T7 promotor and genomes of nOPV2 VP1-I
143T, VP1-N
171D, and VP1-E
295K mutant variants as described previously [
5,
22] with some changes. Briefly, plasmid clones of nOPV2-c1 mutants were digested with Hind III, purified with SPRI beads (1.8×), and eluted in 45 µL H
2O. RNA transcripts were generated from the linearized plasmids using the MEGAscript™ T7 Transcription Kit (Thermo Fisher Scientific, South San Francisco, CA, USA). To remove unincorporated nucleotides and most proteins, the nOPV2-c1 -mutant transcripts were precipitated with LiCl and the pellet was washed once with 1 mL of 70% ethanol and re-centrifuged to maximize removal of unincorporated nucleotides. Then, RNA was dissolved in sterile diethylpyrocarbonate-treated water and stored at −80 °C for downstream use. The purified RNA transcripts were analyzed by electrophoresis on 1% agarose. Then, 2–3 µg of RNA transcript was added to 10
6 HEp-2C cells and subjected to electroporation. The transfected cells were planted in 6-well plates and incubated at 34 °C and 5% CO
2 until the appearance of a complete cytopathic effect (CPE). After the appearance of the CPE, the plate was freeze-thawed three times, and the supernatants were used to transfect a monolayer of HEp-2C cells into 25 cm
2 flasks. The cells were incubated until the appearance of the CPE, and the supernatants were prepared as mentioned above, aliquoted, and stored at −80 °C.
The recovered nOPV2-c1 mutant viruses were sequenced with the Illumina sequencing platform as described below to confirm their nucleotide sequences.
2.3. Primers and TaqMan Oligo Probes Used for nOPV2-c1 Mutant Variant Detection and Quantification
The forward and reverse primers for each of the nOPV2 VP1-I
143T (corresponding to U
2970C nucleotide change), VP1-N
171D (corresponding to A
3053G nucleotide change), and VP1-E
295K (corresponding to G
3425A nucleotide change) mutant variants were selected to flank the targeted mutations (
Table 1). PCR amplification with these primers resulted in DNA fragments of 74, 60, and 61 nucleotides long for nOPV2 VP1-143T, VP1-171D, and VP1-295K mutants, respectively. The TaqMan oligoprobes were designed to be short (less than or equal to 13 nucleotides), with a conjugated minor groove binder (MGB) and low melting temperature, to allow discrimination of a single point mutation (
Table 1). The primers were synthesized by Integrated DNA Technologies IDT (Durham, NC, USA) and TaqMan probes were synthesized by Thermo Fisher Scientific (South San Francisco, CA, USA).
2.4. qmosRT-PCR Amplification
Viral RNA was extracted from nOPV2 viruses using a QIAamp viral RNA mini kit (QIAGEN, Chatsworth, CA, USA) according to the manufacturer’s protocol. The qmosRT-PCR reactions were prepared with the QuantiFast Multiplex RT-PCR Kit (QIAGEN, Valencia, CA, USA) in a final volume of 25 μL using 2 μL of viral RNA. Mutant and non-mutant TaqMan probes were used at final concentrations of 25 nM each in a mixture with the forward and reverse primers (
Table 1) at concentrations of 0.8 μM each. The RNA sample of the nOPV2-c1 virus was used as positive control, and water was used as negative control. RNAs and/or plasmids containing the genomes of nOPV2-c1 mutant and non-mutant viruses were used as reference standards: reference standards, mutant, and non-mutant plasmids/RNAs with known genome copy (GC) numbers per mL were used for extrapolation of GC numbers for mutants and non-mutants of the nOPV2 test samples.
The positive control and reference standard samples were run in duplicate, and the negative control and test samples were run in three repeats. The qmosRT-PCR procedure was performed using a real-time PCR System ViiA7 (Thermo Fisher Scientific, South San Francisco, CA, USA) at the following thermocycler conditions: one cycle of incubation for 20 min at 50 °C and 5 min at 95 °C, followed by 40 cycles each consisting of 15 s at 95 °C, 15 s at 50 °C, and 30 s at 60 °C.
2.5. Quantitative RT-PCR for Virus Genome Copy Number Determination
To quantify the genome copy (GC) number in each sample, a qmosRT-PCR was used. Briefly, the qmosRT-PCR reactions were prepared with a QuantiFast Multiplex RT-PCR Kit (QIAGEN, Valencia, CA, USA) in 96-well optical plates in a final volume of 25 μL using 2 μL of RNA of test and control samples and 2 μL of DNA-plasmid for standard references. The RNAs of nOPV2-c1 virus were used as positive controls, and water was used as negative control. Plasmid containing the genome of nOPV2-c1 mutant virus with a known GC number was used as a standard reference for extrapolation of the GC for nOPV2 mutants as test samples, and plasmid containing the genome of nOPV2-c1 non-mutant virus with a known GC number was used as standard reference for extrapolation of the GC number for nOPV2-c1 non-mutants as test samples.
All control and standard reference samples were run in duplicate, and the test samples were run in three repeats. The specific primer pairs and probes used for each virus are presented
Table 1. The qmosRT-PCR procedure was performed as described above.
2.6. Deep Sequencing Analysis
RNA was extracted from nOPV2 viruses as described above and used for RNA library preparation. The RNA library was prepared using the TruSeq NEBNext Ultra II RNA Library Prep Kit for Illumina (New England BioLabs, Ipswich, MA, USA). Briefly, 100 ng of RNA was fragmented to generate a mean fragment distribution of 500 nt and priming was performed in one reaction using the buffer provided in the kit. The first and second strands of DNA were synthesized according to the manufacturer’s protocol. The resulting DNA fragments were subjected to end repair and ligated to Illumina paired end adaptors. The ligated products were size selected using AMPure XP beads, then amplified using eight cycles of PCR with multiplex indexed primers and purified by magnetic beads (Agencourt AMPure PCR purification system, Beckman Coulter, Brea, CA, USA). Then, the DNA libraries were analyzed for the size and quality with a 4200 Tape Station system (Agilent Technologies, Santa Clara, CA, USA) using a high-sensitivity D1000 Screen Tape. After obtaining the Qubit concentration for the libraries and the mean peak sizes from the Agilent Tape Station profile, paired-end sequencing was performed using a MiSeq System (Illumina, San Diego, CA, USA), producing 250 nt reads. The raw sequencing reads were analyzed with the specialized High-Performance Integrated Virtual Environment (HIVE) platform developed in-house [
23]. The RNA sequences of nOPV2-c1 and nOPV2-c2 with GenBank accession numbers MZ245455 and MN654096, respectively, were used as genome references for HIVE alignment and mutations profiling.
4. Discussion
Virus strains used in the manufacture of viral vaccines undergo mutational changes, which may cause reversion of attenuated strains to virulence, as well as changes in the antigenic properties that may affect the immunogenicity of vaccines [
26,
27,
28]. Therefore, monitoring of the genetic consistency of vaccines is an important task to ensure their quality, safety, and potency.
Sabin strains used in the manufacture of OPV vaccine represent a case study of reversion of attenuated vaccine strains to virulence. Molecular studies performed in the 1980s showed that U
472C mutation in Sabin type 3 virus [
29], A
481G mutation in Sabin type 2 virus [
30], and G
480A and U
525C mutations in Sabin type 1 virus [
31] accumulate during virus growth in vitro and in vivo and are responsible for reversion to virulence. Similar mutations are found in vaccine-derived strains isolated from cases of vaccine-associated paralytic polio and, therefore, the presence of these revertants in batches of polio vaccine is considered a safety risk. A method based on PCR and restriction enzyme cleavage (MAPREC) was developed to identify batches of oral polio vaccine (OPV) with unacceptable levels of neurovirulent revertants [
32,
33] and recommended by the WHO as an in vitro method of choice for lot release of OPV. However, MAPREC suffers from some shortcomings, including various technical challenges at multiple steps of its complex protocol.
Previously, we developed a method for quantification of mutants in Sabin 3 virus based on allele-specific primers that consist of three segments: the mutant-discriminating 3′-end, linked by a flexible polyinosine stretch to a specific segment that serves as an anchor to increase stability of binding and prevent non-specific priming [
34]. Such a method needs the design of tethered allele-specific primers, extensive screening for the working primers, and optimization before it starts to work properly.
Quantitative PCR is an obvious alternative, but most PCR-based genotyping protocols quantify only one SNP allele per reaction [
35,
36,
37,
38,
39,
40].
nOPV2 was developed as a genetically stable vaccine to control outbreaks caused by cVDPV2. A recent study [
41] comparing Sabin 2 virus and nOPV2 demonstrated significantly lower neurovirulence, as determined in a transgenic mouse assay, in the virus shed by recipients of nOPV2 compared to recipients of conventional OPV2. Furthermore, while shed Sabin 2 virus had the anticipated A481G reversion in the primary attenuation site in domain V in the 5′ UTR, associated with increased mouse neurovirulence, the stabilized domain V in the nOPV2 viruses did not show polymorphisms consistent with reversion to neurovirulence [
41].
We previously reported on the use of NGS for quality control of nOPV2, with quantitative specifications imposed for the three mutations of interest studied in this work [
21]. While NGS may be a viable approach for lot release, maintaining NGS assays and equipment in a validated state for indefinite routine release is, to our knowledge, unprecedented. As such, we sought to explore alternatives.
Quantitative PCR and NGS are broadly used to detect mutant variants in viruses and viral vaccines. Quantitative PCR has the advantage of being easy, faster, and cheaper to use to evaluate up to about four SNPs. NGS, in contrast, allows simultaneous analysis of many genomic loci; it is, however, more technically demanding and expensive.
The method proposed in this communication is based on the use of two short minor groove binder probes; one specific for the mutants and the second specific for the non-mutants. Both are run in the same quantitative RT-PCR reaction with forward and reverse primers (
Table 1), as described above. The results of the qmosRT-PCR, used to determine quantities of VP1 I
143T, N
171D, and E
295K mutant variants that may affect the safety and immunogenicity of nOPV2, demonstrated a very good correlation with the results of the Illumina sequencing. This suggests that the qmosRT-PCR assay can be used for quantification of mutant variants in nOPV2 batches. Even though the method was designed to specifically quantify the mutants in the nOPV2-c1, the results for the quantification of the mutants in the nOPV2-c2 were comparable to those generated by NGS.
The method enables the quantification of both mutants and non-mutants in the same sample and in the same reaction, reducing time and labor and simplifying the method.
The proposed method is simple, sensitive, and could be used to test about 18 samples simultaneously, along with appropriate controls and reference samples. In addition, the use of different dyes to discriminate between mutants and non-mutants makes this method much easier to perform. The qmosRT-PCR method produces consistent results (with less than twofold difference from run to run) and is able to detect about 1% of mutant variants.
The quantification of the mutant variants in nOPV2 lots correlated well with the results of the NGS, indicating that the qmosRT-PCR method could be used for quality control of the nOPV2 vaccine.
In conclusion, the quantitative RT-PCR for quantification of undesirable mutants in the nOPV2 vaccine described in this communication offers a simple and rapid method for quantification of mutants in nOPV2 virus. The assays for quantitation of nOPV2 VP1-I143T, VP1-N171D, and VP1-E295K mutants were designed specifically to be applied for quality control during the manufacture of the nOPV2 vaccine.