ABSTRACT
Replication of influenza A virus (IAV) from negative-sense viral RNA (vRNA) requires the generation of positive-sense RNA (+RNA). Most molecular assays, such as conventional real-time reverse transcriptase PCR (rRT-PCR), detect total RNA in a sample without differentiating vRNA from +RNA. These assays are not designed to distinguish IAV infection versus exposure of an individual to an environment enriched with IAVs but wherein no viral replication occurs. We therefore developed a strand-specific hybridization (SSH) assay that differentiates between vRNA and +RNA and quantifies relative levels of each RNA species. The SSH assay exhibited a linearity of 7 logs with a lower limit of detection of 6.0 × 102 copies of molecules per reaction. No signal was detected in samples with a high load of nontarget template or influenza B virus, demonstrating assay specificity. IAV +RNA was detected 2 to 4 h postinoculation of MDCK cells, whereas synthesis of cold-adapted IAV +RNA was significantly impaired at 37°C. The SSH assay was then used to test IAV rRT-PCR positive nasopharyngeal specimens collected from individuals exposed to IAV at swine exhibitions (n = 7) or while working at live bird markets (n = 2). The SSH assay was able to differentiate vRNA and +RNA in samples collected from infected, symptomatic individuals versus individuals who were exposed to IAV in the environment but had no active viral replication. Data generated with this technique, especially when coupled with clinical data and assessment of seroconversion, will facilitate differentiation of actual IAV infection with replicating virus versus individuals exposed to high levels of environmental contamination but without virus infection.
INTRODUCTION
Influenza A virus (IAV) is an enveloped RNA virus infecting a wide range of birds and mammals, including humans. The viral genome consists of eight segments of negative-sense, single-stranded RNA (vRNA) encoding 10 essential viral proteins and at least 4 strain-dependent accessory proteins (1). Each genome segment forms a viral ribonucleoprotein (vRNP) complex consisting of vRNA encapsulated by nucleoprotein (NP) associated with heterotrimeric polymerase (PB2, PB1, and PA). The RNA-dependent RNA polymerase lacks proofreading activity and is error-prone during RNA synthesis (2). High mutation rates resulting from RNA synthesis error and segment reassortment between coinfecting strains are responsible for a great diversity of genotypes and phenotypes. These viruses pose a pandemic threat if they are capable of infecting and transmitting between humans, especially when the population lacks preexisting immunity (3).
An infection of IAV begins with attachment to host cells. Following attachment, IAV enters cells generally through receptor-mediated endocytosis. The increasingly acidic environment of the endosome triggers hemagglutinin (HA)-mediated fusion of the viral envelope with the endosomal membrane followed by release of the vRNPs into the host cell cytoplasm. The vRNPs are then transported into the nucleus, where vRNAs are transcribed and replicated, generating two species of positive-sense RNAs (+RNA), mRNA, and cRNA. mRNA is then exported into the cytoplasm for translation using cellular machinery, while cRNA remains in the nucleus and serves as the template for generation of progeny vRNA. Synthesis of +RNA, mainly mRNA, dominates in the early phase of infection, whereas in the late phase of infection, vRNA is accumulated (4). Switching from transcription to replication that involves generation of cRNA can be regulated by viral proteins (NP, M1, and NS2/NEP), short viral RNAs (svRNA), or host factors (5–10). However, this complex mechanism remains to be fully elucidated.
Entry into host cells to initiate replication precedes an active infection with IAV. Defects in IAV entry, including host cell attachment, internalization, fusion of viral and endosomal membranes, and import of vRNP into the nucleus, can all jeopardize transcription and replication of the viral genome. For example, IAV has been found to be able to bind to cell surface receptors but not enter cells that are deficient in N-linked glycosylation (11). In general, avian IAV preferentially binds to glycoconjugates with terminal α2,3 linkage-sialic acids (SAα2,3). Scarcity of SAα2,3 receptors in the human upper respiratory tract is a restrictive factor for infection with most nonmammalian adapted avian IAVs. Apart from viral factors, differences in host genetic background, vaccination history, immune status, upper respiratory tract microbiome composition, nutritional imbalances, comorbid conditions, or even environmental factors, such as ozone exposure, can contribute to vast individual diversity in the susceptibility to influenza infection (12, 13). For example, using real-time reverse transcriptase PCR (rRT-PCR), influenza surveillance conducted in live animal markets detected IAV RNA in specimens collected from 76% of workers. However, attempts to recover IAV were not successful, and only one of the workers reported influenza-like illness (ILI) (14). Nasreen et al. estimated that, after exposure to highly pathogenic avian influenza A(H5N1) virus in live bird markets, 50 out of 721 enlisted poultry workers (6.9%) would be infected with the H5N1 IAV (15). Given that potential exposure to large amounts of IAV occurs in market environments and/or agricultural events with large numbers of infected animals, it is important to have molecular tools that can differentiate high levels of environmental exposure from bona fide infection.
Most molecular diagnostic assays that detect infection with IAV, such as conventional real-time reverse transcriptase PCR (rRT-PCR), rely on total viral RNA as a template to synthesize cDNA, which is then amplified by virus-specific primers. While these methods are optimized to detect small quantities of total viral RNA in a sample, they are not designed to differentiate between specimens collected from an infected individual (vRNA and +RNA species) versus an individual who was only environmentally exposed to IAV (vRNA). Strand-specific rRT-PCR is the most common diagnostic tool currently used to detect and differentiate RNA species, replacing traditional radioisotope-associated assays such as RNase protection, Northern blotting, and primer extension methods (4, 16). Although strand-specific rRT-PCR assays are highly specific and sensitive (4), they were validated for particular strains of IAV and may not detect all IAVs. Synthesis of new primers may be necessary for different IAVs and may require further optimization to minimize formation of primer dimers that can be detected nonspecifically during SYBR green rRT-PCR (4, 16). More critically, different primers could result in distinctive PCR amplification efficiency, which makes quantitative comparison between RNA species difficult (4, 16). Here, we developed a strand-specific hybridization (SSH) assay to detect and differentiate IAV vRNA from +RNA through strand-specific hybridization followed by PCR amplification using influenza-sequence-independent adaptor primers. Similar amplification efficiencies from a single pair of adaptor primers was observed for both IAV vRNA- and +RNA-specific assays. All hands-on procedures were performed at room temperature, which allowed for easy scale-up of testing throughput and automation of the procedures. Human specimens collected from individuals exposed to IAVs at live animal markets were tested to demonstrate the capability of the SSH assay to differentiate actual IAV infection with replicating virus versus exposure to high levels of environmental contamination but without virus infection.
MATERIALS AND METHODS
Strand-specific hybridization and real-time PCR.Strand-specific hybridization and real-time PCR for differentiation and detection of vRNA and +RNA were modified from a previously designed IAV subtyping assay (17). The procedures are illustrated in Fig. 1. Briefly, a hybridization probe (HP) of 35 nucleotides targeting either vRNA or +RNA of IAV was synthesized as a chimera with an adaptor of 22 nucleotides unrelated to IAV. The chimeric hybridization probe was covalently conjugated onto magnetic Dynabeads as described previously (17). Similarly, a bridge probe (BP) targeting vRNA or +RNA of IAV was also synthesized as a chimera with an IAV-independent adaptor but without bead conjugation. These two types of chimeric probes were then hybridized simultaneously to a cDNA molecule to form a complex consisting of the bridge probe, the hybridization probe, and the target cDNA that was converted from total RNA using random-primer-based SuperScript IV first-strand synthesis (Thermo Fisher Scientific). The duration of hybridization was 30 min at 45°C for a relatively high concentration of cDNA derived from cell culture isolates but was extended to 3 h or overnight for clinical specimens to improve sensitivity. Following hybridization, the gap between the bridge probe and hybridization probe was filled with deoxynucleoside triphosphates (dNTPs) at room temperature using T4 DNA polymerase and T4 DNA ligase. A 5-min denaturation using NaOH at room temperature was followed by on-bead purification steps to remove the input cDNA and excessive free probes. The resulting on-bead single-stranded DNA (ssDNA) was then used as a template for PCR amplification using a single pair of adapter primers irrespective of IAV strain or subtype. Five cycles of PCR amplification of on-bead target were performed to further enhance detection unless otherwise indicated. Identification of RNA species was done through TaqMan probes targeting the influenza A matrix gene. PCRs were prepared using 1× Brilliant III ultrafast multiplex quantitative PCR (qPCR) master mix (catalog no. 600880-051; Agilent Technologies). The PCR was carried out on an AriaMx real-time PCR thermocycler (PN G8830-6400; Agilent Technologies) at 95°C for 3 min for 1 cycle and 95°C for 5 s followed by 60°C for 15 s for 40 cycles. Probe and primer sequences and working concentrations are listed in Table S1 in the supplemental material.
Illustration of strand-specific hybridization (SSH) for detection of vRNA or +RNA of influenza A virus. A chimera oligo consisting of an adaptor (AP) and a hybridization probe (HP) or bridge probe (BP) that is specific for negative single-strand viral RNA (vRNA, left panel) or positive single-strand mRNA/cRNA (+RNA, right panel). HP is conjugated to magnetic Dynabeads. These two chimera probes were hybridized simultaneously to the same target. The gap between the probes was then filled with dNTPs. After denaturation, the immobilized on-bead single-strand DNA was washed and used as a template for real-time PCR amplification in an influenza-sequence-independent manner using universal adaptor primers (UAP). Identification of vRNA or +RNA was done using TaqMan real-time PCR with the matrix-gene-specific TaqMan PCR probe (TMP).
Cell culture and viral infection.Madin-Darby canine kidney (MDCK) cells (ATCC, Manassas, VA) were grown in Dulbecco’s modified Eagle medium (DMEM) with 10% fetal bovine serum, 1 mM l-glutamine, 100 units/ml penicillin, and 100 μg/ml streptomycin and seeded onto 24-well plates. Confluent monolayers were washed twice with 10 mM phosphate-buffered saline (PBS) at pH 7.4 and inoculated with a reassortant H9N2 candidate vaccine virus, A/Hong Kong/33982/2006 x PR8 (IDCDC-RG26), at a multiplicity of infection (MOI) of 0.1 in 250 μl virus growth medium (VGM) with or without 2 μg/ml l-1-tosylamido-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin. The plate was incubated on ice for 1 h at 4°C to synchronize infection. After 1 h, inoculum was removed and cells were washed three times with ice-cold PBS. After washing, 3 wells of cells were lysed with 350 μl MagNA Pure total nucleic acid lysis buffer (Roche Diagnostics) and collected as the 0 h postinoculation time point. Then 500 μl VGM with 2 μg/ml TPCK-trypsin was added to the remaining wells, and the plate was incubated at 37°C with 5% CO2. After 2, 3, 4, 5, 6, and 24 h of incubation, the cells were rinsed with PBS and lysed with 350 μl lysis buffer prior to RNA extraction.
The cold-adapted influenza virus (caTX50) used in this study was a 6:2 reassortant virus containing the HA and neuraminidase (NA) gene segments from the wild-type A/Texas/50/2012 (H3N2) (wtTX50) virus and the six internal gene segments from a master donor virus, A/Leningrad/134/17/57 (caLen17), to provide temperature-sensitive, cold-adapted, and attenuated phenotypes (18). MDCK cells were inoculated with viruses at an MOI of 0.02 in the presence of 2 μg/ml TPCK-trypsin. After 1 h of incubation on ice in a refrigerator at 4°C, the inoculum was removed and cells were washed three times with ice-cold PBS. After 0, 2, 4, 6, 8, and 24 h of incubation in VGM with 2 μg/ml TPCK-trypsin at 37°C, the cells were rinsed with PBS and lysed with 350 μl of lysis buffer as above. Cell lysates without virus inoculation were used as mock controls.
Upper respiratory specimens.During July to August 2017, outbreaks of IAV A(H1N2) variant [A(H1N2v)] and A(H3N2) variant [A(H3N2v)] viruses were reported in Ohio in persons exposed to virus-infected swine at agricultural events. All persons identified with A(H1N2)v or A(H3N2)v infections during these outbreaks had fever or other influenza-like illness (ILI) and reported swine exposure. Seven specimens collected using nasopharyngeal swabs were stored in viral transport medium (VTM) at –80°C. Nasal swab specimens collected from live poultry market workers in Dhaka, Bangladesh, in February to April 2017 were stored in VTM in an insulated cooler (4 to 8°C) until storage at –80°C.
RNA extraction, screening, and cDNA synthesis.One hundred microliters of human specimens or cell-culture lysates were subjected to total RNA extraction using the MagNA Pure 96 viral NA small-volume kit (catalog no. 06542588001; Roche Diagnostics) following the manufacturer’s instructions. The purified RNA was eluted into 100 μl nuclease-free water. For human specimens, each RNA was initially screened for the detection of the influenza A virus matrix gene and human RNase P (RNP) housekeeping gene using a TaqMan real-time RT-PCR assay (19). cDNA was generated from total RNA using a random-primer-based SuperScript RT IV kit (catalog no. 11756500; Thermo Fisher Scientific) following the manufacturer’s instructions. The seasonal influenza positive control (SIPC) consisted of three beta-propiolactone (BPL) inactivated influenza viruses grown in 10- to 11-day-old embryonated chicken egg allantoic fluid for 24 to 48 h without further purification [influenza A(H3N2), A(H1N1)pdm09, and influenza B] and cultured human cells (A549) that served as a positive control for the human RNP assay. RNA was extracted from 100 μl SIPC following the procedures described above.
The 2014–2015 formulation of FluMist quadrivalent influenza vaccine (purchased from MedImmune, LLC) consisted of two live attenuated influenza A viruses (LAIV) [A/California/7/2009 (H1pdm)09 and A/Texas/50/2012 (H3N2)] and two live attenuated influenza B viruses (B/Massachusetts/2/2012 and B/Brisbane/60/2008). RNA was extracted from 200 μl of FluMist vaccine, and aliquots of 10−5 dilutions of the RNA were tested in this study.
Standard curve analysis.The template for standard curve analysis was synthetic ssDNA of 98 nucleotides (nt) (Integrated DNA Technologies, Inc., Iowa). The sequences of the +RNA template corresponded to the IAV matrix 1 (M1) gene at position 238 to 328 of the consensus sequence aligned from 6,965 matrix gene sequences of influenza A virus as described previously (17) (Fig. S2). The sequences of the synthetic +RNA template are listed in Table S1. The nucleotide sequences of the vRNA template were reverse complementary to the +RNA sequences. The concentration of the template was determined using a NanoDrop spectrophotometer and further normalized using TaqMan real-time PCR. Tenfold serial dilutions of the ssDNA ranging in concentration from 10 to 1 × 109 copies per reaction were prepared in triplicate for each dilution and processed simultaneously with the samples tested. PCRs were prepared using 1× Brilliant III ultrafast multiplex qPCR master (catalog no. 600880-051; Agilent Technologies). The PCR was carried out on an AriaMx real-time PCR thermocycler (PN G8830-6400; Agilent Technologies) at 95°C for 3 min for 1 cycle and then 95°C for 5 s followed by 60°C for 15 s for 40 cycles. Probe and primer sequences and working concentrations are listed in Table S1.
Statistical analysis.Multiple t tests were performed using GraphPad Prism version 7.05 (GraphPad Software, California) to compare the average mean value of fold increases in vRNA or +RNA expression across different hours postinoculation. The individual P values were computed with fewer assumptions; i.e., each time point was analyzed individually without assuming consistent standard deviation.
RESULTS
SSH assay design and validation.To assess sensitivity, linearity, and reproducibility of the SSH assay, synthetic ssDNA corresponding to the sequences of vRNA or +RNA of IAV M1 gene was used as a hybridization template. Standard curve analysis suggested that both the vRNA and +RNA assays performed equally well with similar amplification efficiency close to 100%. The linear range of the assays spanned 7 logs, ranging from 1 × 102 to 1 × 109 copies of ssDNA molecules per reaction (Table 1; Fig. S1). The lower limit of detection (LLoD) was 6 × 102 copies per reaction for the vRNA assay and 1 × 102 copies per reaction for the +RNA assay (Table 1). The intra-assay coefficient variation (CV) was less than 5%, and the interassay CV was less than 15% (Table 1). The specificity of the assays was assessed from three different perspectives—no template control (NTC), high concentration of influenza B RNA, and high concentration of nontarget synthetic ssDNA (1 × 108 copies/reaction). The assays only detected their targets; no signals were reported in any of nonspecific controls (Table 2).
The SSH assay efficiency, linearity, reproducibility, and lower limit of detection
Specificity of the SSH assay
Kinetics of +RNA and vRNA expression of IAV in MDCK cells.Following inoculation of MDCK cells with the A(H9N2) virus at an MOI of 0.1, IAV +RNA was detected 2 h postinoculation (hpi) in the presence of 2 μg/ml TPCK-trypsin but was not detectable until 4 hpi in the absence of trypsin (Fig. 2A). Also, the increase in +RNA levels appeared to be more robust in the presence of trypsin than in the absence of trypsin. The expression of +RNA at 6 hpi with trypsin increased by ∼50-fold (mean ± SD, 48 ± 3.3; P < 0.005) and ∼20-fold (mean ± SD, 19.8 ± 1.3; P < 0.005) relative to that at 3 hpi and 4 hpi, respectively (Fig. 2A). In contrast, in the absence of trypsin, the +RNA level at 6 hpi increased by only ∼15-fold (mean ± SD, 14 ± 4.9; P < 0.05) relative to that at 4 hpi, when it first became detectable (Fig. 2A). IAV +RNA expression at 6 hpi and 24 hpi was significantly different between the two groups (–trypsin and +trypsin) (P ≤ 0.05; Fig. 2A), but the difference in vRNA expression was not statistically significant between the groups across all time points tested (Fig. 2B). Although transcription, as measured by an increase in +RNA levels, occurred earlier and more robustly with trypsin than without trypsin (Fig. 2A), the increase in the +RNA level relative to the vRNA level (+RNA/vRNA ratio) exhibited a similar pattern between the treatment groups. The ratio levels peaked around 5 hpi but quickly decreased at 6 hpi and decreased by about 90% by 24 hpi (P < 0.05), regardless of trypsin addition, suggesting that trypsin enhances both transcription and replication (Fig. 2C).
Kinetics of transcription and replication of influenza A virus in MDCK cells. The MDCK cells were inoculated with H9N2 virus (IDCDC-RG26) at an MOI of 0.1 and incubated at 37°C. The fold increase in copy numbers of vRNA or +RNA was relative to the copy numbers of the RNA (#) that was above the limit of detection, noted with a red dashed line. Error bars were calculated from triplicate treatments for each time point. (A) Expression of +RNA with (+) or without (–) 2 μg/ml TPCK-trypsin. (B) Expression of vRNA with (+) or without (–) 2 μg/ml TPCK-trypsin. (C) The vRNA/+RNA ratio was calculated using the fold increase in RNA copy numbers shown in panels A and B.
Kinetics of +RNA and vRNA synthesis of wild-type influenza A viruses and cold-adapted strains in MDCK cells.Cold-adapted influenza virus is known to replicate less efficiently at 37°C than its parent wild-type virus (18) due to mutations in the RNA polymerase proteins PB2 and PB1 (20). Thus, the kinetics of +RNA and vRNA expression of caTX50 compared to its parental strain, wtTX50, were investigated. For the caTX50, +RNA was detectable 2 hpi and kept increasing slowly until 8 hpi and then plateaued without significant reduction at 24 hpi (P = 0.76) (Fig. 3A). In contrast, for the wtTX50, although +RNA was not detected until 4 hpi, it increased rapidly by ∼30-fold (mean ± SD, 32.4 ± 5.3; P < 0.001) at 6 hpi and by ∼300-fold (mean ± SD 326 ± 29.2; P < 0.001) at 8 hpi and continuously accumulated through 24 hpi (Fig. 3A). The peak level of +RNA expression of the caTX50 was less than 10% that of the wtTX50 (Fig. 3A), indicating that transcription of the caTX50 was significantly restricted at 37°C. Although +RNA levels of caTX50 increased by ∼60-fold (P < 0.05) at 6 hpi (Fig. 3A), there was no statistically significant change in the vRNA level up to 6 hpi (Fig. 3B). Likewise, +RNA levels of the wtTX50 increased by over 300-fold at 8 hpi (Fig. 3A), but vRNA only increased by 30-fold (P < 0.05) (Fig. 3B). IAV +RNA was predominant over vRNA until 24 hpi, when synthesis of +RNA was subdued, while vRNA continuously accumulated, resulting in approximately equimolar amounts of both RNA species (Fig. 3C). A(H3N2) and A(H1N1)pdm09 viruses inactivated with beta-propiolactone had a very small amount of +RNA (+RNA/vRNA ratio = 0.51%), although the vRNA levels were relatively high at around 1 million copies per reaction (Table S3). In contrast, a much higher +RNA/vRNA ratio (18.3%) was detected in the FluMist preparation, a cold-adapted live attenuated influenza vaccine that has been licensed for use in the United States since 2003 (21) (Table S3).
Kinetics of transcription/replication of wild-type influenza A virus (A/Texas/50/2015 [wtTX50]) and its cold-adapted strain (caTX50) in MDCK cells incubated at 37°C. The fold increase in copy numbers of +RNA (A) or vRNA (B) was relative to the copy numbers of the RNA (#) that was above the limit of detection, noted with a red dashed line. The vRNA/+RNA ratio (C) was calculated using the fold increase in RNA copy numbers shown in panels A and B. Error bars were calculated from triplicate treatments for each time point. Shown here are the results from one of the reproducible and independent experiments.
Detection of IAV RNA species in upper respiratory specimens.All individuals were positive for IAV RNA by M gene rRT-PCR and exhibited typical ILI symptoms, including fever. Among 15 upper respiratory specimens collected from outbreaks of A(H1N2) variant (H1N2v, n = 2) and A(H3N2) variant virus (H3N2v, n = 13) in Ohio during July to August 2017, 7 specimens with good sample quality (RNP threshold cycle [CT] < 27) and small variations in RNP CT value (<2 CT) were analyzed for vRNA and +RNA expression using our SSH assay (Table S2). IAV vRNA was detected in all 7 cases, whereas IAV +RNA was detected in all but 1 case (patient 7; Fig. 4A and B). This patient was treated with oseltamivir for 3 days prior to collection of the nasopharyngeal swab sample (IAV CT = 35.3; Table S2).
Detection of influenza A RNA species in human specimens. (A) Detection of influenza vRNA or +RNA in patients. RNA copy numbers per reaction were normalized against the RNP CT value to account for variations in sampling. The limit of detection is marked by a dashed line in color representing the vRNA or +RNA assay. Error bars were calculated from duplicate reactions. (B) Ratio of +RNA/vRNA copy numbers per reaction calculated from panel A.
Detection and differentiation of IAV RNA species among live poultry market workers.Live poultry market (LPM) workers with minimal use of personal protective equipment or low standard of hygiene practice (e.g., not washing hands after working with sick poultry) may be exposed to avian IAVs from infected poultry or contaminated environments, especially those who are engaged in cleaning feeding trays and waste pens and slaughtering and eviscerating poultry (15, 22). Nasal swab specimens collected from two LPM workers in Bangladesh in February 2017 to April 2018 were available for analysis (Table 3). Both individuals tested weakly positive for IAV RNA, and one individual was weakly positive for H9 hemagglutinin RNA by rRT-PCR (Table 3). Neither reported ILI, and no IAV was isolated from their specimens. All other influenza A HA subtype-specific rRT-PCR assays (H1, H3, H5, and H7) failed to generate any signal (data not shown). Initial tests of these two samples were negative for both IAV vRNA and +RNA following the procedures illustrated in Fig. 1. However, after increasing PCR preamplification from 5 cycles to 15 cycles, IAV vRNA became positive, but IAV +RNA remained undetectable (Table 3).
Detection and differentiation of RNA species of influenza A virus in nasal swab specimens collected from two live poultry market workers in Bangladesha
DISCUSSION
IAV RNA polymerase carries out transcription and replication of the viral RNA genome to produce mRNA, cRNA, and progeny vRNA (6). Differentiation and quantification of IAV RNA species (vRNA, mRNA, and cRNA) have been used for developing LAIV (23, 24), assessing the efficacy of antiviral drugs that target influenza polymerase complex (25, 26), or understanding genetic determinants of thermostability of IAV RNA polymerase (27). In order to distinguish virus infection versus environmental exposure, we developed a procedure that allows differentiation of replicating versus nonreplicating IAV through quantitative detection and differentiation of RNA species. Although the NP and nonstructural (NS) genes of IAV are reported to be transcribed in early phases of infection (4, 28), we chose to design our assay against the commonly used M1 segment of the IAV matrix gene due to its slower evolutionary rate and high sequence conservation across all IAV subtypes (29, 30). Our SSH assay was capable of detecting IAV +RNA 2 h (with trypsin) or 4 h (without trypsin) postinoculation of MDCK cells infected with IAV, similar to what was reported previously (4, 16). Using a replication-compromised attenuated virus, we showed that both +RNA expression and the +RNA/vRNA ratio of the wild-type virus (wtTX50) were higher than those of its cold-adapted strain (caTX50) when incubated at 37°C. These replication kinetic experiments also demonstrated the assay’s ability to quantify +RNA and vRNA over multiple time points. Finally, the assay was used to test specimens collected from individuals exposed to animal influenza viruses at agricultural events in the United States and live poultry markets in Bangladesh in order to assess differences between vRNA and +RNA species. Detection of +RNA was consistent with active infection in symptomatic individuals but not in asymptomatic individuals or a specimen collected postantiviral treatment.
The specificity of the SSH assay was ensured at two separate steps, target enrichment/binding through cDNA strand-specific hybridization and target detection via TaqMan PCR. Formulation of the hybridization buffer allowed hybridization at a universal temperature for all probes with the same number of nucleotides, regardless of sequence variations (17). Both HP- and BP-nucleotides, therefore, could be highly degenerative to potentially include all conserved mutations found in IAV M1 genes without the need of frequent updates on probe sequences (17). Conventional TaqMan real-time PCR requires three conserved regions for selection of a forward primer, a reverse primer, and a TaqMan probe. This SSH assay uses universal adapter primers for PCR amplification. Without the constraints of primer selection, a TaqMan probe can be designed anywhere along the entire gene in conserved regions with the fewest sequence variations. Different IAVs from multiple genetic lineages were evaluated in this study. Both the +RNA and vRNA assays were highly specific, as no signals were detected in samples with high content of influenza B RNA (CT = 12.6) or nontarget synthetic templates (100 million copies per reaction).
Following the uncoating step, the vRNP complexes enter the nucleus, where primary transcription takes place. During primary transcription, which is independent of protein synthesis, mRNA is produced by the polymerase carried on the parental vRNP (1, 31). Viral protein synthesis following mRNA production promotes further synthesis of mRNA (secondary transcription), cRNA, and progeny vRNA (replication) (1, 31). Quantitative real-time RT-PCR analysis suggested that the cRNA level was about 1% of the mRNA level of the A/WSN/33 virus at 4 hpi in MDCK cells (4). As such, in serving the purpose of distinguishing replicating versus nonreplicating IAV, we did not attempt to differentiate mRNA from cRNA. Also, based on the analysis using primer extension and an RNA-RNA hybridization technique, mRNA predominates over cRNA and vRNA in the early phase of infection (27, 32). In line with these previous reports, a similar pattern was observed in this study, as the +RNA/vRNA ratio of IDCDC-RG26 (H9N2) virus peaked at the early stage of cell infection (∼5 hpi) and was reduced by 90% at 24 hpi. Likewise, both the cold-adapted and wild A(H3N2) viruses had higher levels of +RNA expression at 8 hpi than at 24 hpi, demonstrating that +RNA dominates in the early phase of infection and starts decreasing afterwards, while vRNA gradually increased throughout the course of infection (4). Coinciding with these data, BPL-inactivated IAV isolates grown beyond 24 h without purification had a small portion of +RNA (+RNA/vRNA ratio = 0.5%) compared to a significantly higher +RNA/vRNA ratio (18%) detected in the FluMist LAIV vaccine, suggesting that the FluMist LAIV vaccine contained IAV with active transcription.
It has long been recognized that trypsin enhances multicycle virus replication by proteolytic activation of viral hemagglutinin (HA) that is critical for membrane fusion (33, 34). Recent studies suggested that trypsin can also promote a higher yield of influenza vaccine production in MDCK cells through damping cellular antiviral defenses by proteolytic degradation of secreted interferon (IFN) (35). IFNs activate IFN-stimulated genes (ISGs) in infected cells and in neighboring noninfected cells, preventing them from potential viral invasion (36). Expression of ISGs results in diverse modes of action, including degradation of viral RNA and proteins, inhibition of vRNA transcription and translation, and inactivation of budding viruses (36, 37). Here, we observed that IAV +RNA was detected 2 hpi in the presence of trypsin but was undetectable until 4 hpi in the absence of trypsin (Fig. 1A). Trypsin induced a significant increase in +RNA levels at 6 hpi (P < 0.005) (Fig. 1A) but not in vRNA levels (P = 0.15) (Fig. 1B). Given that it takes about 1.5 h for the vRNPs to enter the nucleus and 3 h to initiate replication (38, 39), our observation may in part reflect the dual effects of trypsin on proteolytic HA activation and IFN degradation, both of which could ultimately enhance IAV transcription and replication.
Human infection following exposure to IAV is dependent on numerous variables, including, but not limited to, the site of infection (e.g., upper versus lower respiratory tract), virion droplet size/dose, receptor binding specificity, polymerase compatibility, host adaptation, and immune status (12, 13). Swine workers were reported to have higher titers of serum-neutralizing antibodies against swine A(H1N1) and higher levels of nasal immunoglobulin A (IgA) against swine A(H1N1) and A(H3N2) viruses, which could reduce their chance of infection by related or heterologous swine IAVs (40). In an influenza surveillance study in a colony in Canada, where the residents practiced communal farming relatively isolated from urban areas, 208 (25%) participants tested positive for IAV RNA by rRT-PCR; about 10% were asymptomatic (41). Similarly, in a community-based study in Hong Kong in 2008, IAV RNA was detected with RT-PCR in 8/59 (14%) asymptomatic individuals (42). Asymptomatic infection is possible, especially among children and young adults or those individuals who are serologically positive for certain types of IAV (43–45). Several studies suggest that healthy volunteer participants in experimental IAV infection were partially immune to the contemporary strain even though they had a very low level of hemagglutination inhibition (HAI) antibodies in their serum before challenge (46, 47). For example, among 11 adult healthy volunteers with low HAI antibody titers who were challenged with A(H2N2) virus, only 6 were infected (48). A study to assess the seroprevalence of highly pathogenic avian influenza A(H5N1) virus among LPM workers in Bangladesh indicated that only 6 of 284 (2%) LPM workers were seroconverted (15). In the present study, we found that IAV vRNA, but not +RNA, was detected in two asymptomatic individuals who worked in live poultry markets in Bangladesh, suggesting that these individuals were exposed to IAV but had no active infection with replicating IAV. In environments where viral load is high, such as in live bird markets with IAV-infected poultry, IAV rRT-PCR positive swabs collected from individuals may reflect the level of environmental exposure, not active infection.
There is limited yet compelling evidence suggesting a positive correlation between +RNA levels and pathogenesis or transmissibility of IAV. In one study, the authors revealed that replacement of the A/chicken/Beijing/1/1994 (H9N2)-like M gene with the A/quail/Hong Kong/G1/1997 (H9N2)-like M gene through natural reassortment led to an early surge in mRNA expression, which was associated with more severe pathogenesis and extrapulmonary virus spread in chickens (49). However, other studies did not find a correlation between IAV total RNA levels and clinical symptom scores, as the RNA level varied by up to 5 logs even with similar symptom scores (41). Our data also revealed variation in +RNA levels among the clinical specimens tested in individuals infected with A(H1N2)v and A(H3N2)v viruses while attending agricultural events (Fig. 3; Table S2). Although +RNA was detected in all but one individual, +RNA levels varied greatly among the specimens, suggesting a high degree of variability in viral replication in infected persons and/or various stages of replication following infection. The high degree of variation in +RNA levels could also result from when samples are collected following infection and the inherent differences between when individuals are infected versus when samples are collected in either the field or clinic. Nevertheless, specimens should be collected prior to antiviral treatment. The specimen with undetectable IAV +RNA and a low level of vRNA was collected 3 days after oseltamivir treatment, suggesting that virus replication was significantly impaired when antiviral therapy was given within 48 h of illness onset, as previously reported (50). Finally, the number of clinical and field samples available for analysis that also had adequate clinical data and a range of symptoms was low. A larger number of clinically relevant and field specimens is needed to demonstrate the potential value of the SSH assay in clinical applications and field studies. In addition to further testing of the clinical applications of the assay, future disease pathogenesis studies would benefit from such an analysis by comparing levels of influenza virus replication in various anatomical sites to that of disease severity.
In summary, the novel technique described here based on strand-specific hybridization to differentiate vRNA from +RNA is both sensitive and specific. It can be used for quantification and comparison of IAV RNA species and is universally optimized, as the amplification efficacy is independent of IAV-strain sequence. The SSH assay has been tested using laboratory-cultured IAV isolates, replication-compromised vaccine strains, and field samples collected from live poultry market workers and clinically symptomatic individuals exposed to swine at agricultural events. Simple on-bead washing steps allow easy, scalable automation to accommodate the high-throughput often required for IAV surveillance. The SSH assay can facilitate differentiation of IAV RNA species to rule out environmental exposure from actively replicating virus infections.
ACKNOWLEDGMENTS
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention/Agency for Toxic Substances and Disease Registry.
N.Z. and E.N.H. were supported by Chickasaw Nation Industries.
FOOTNOTES
- Received 10 February 2020.
- Returned for modification 15 March 2020.
- Accepted 27 March 2020.
- Accepted manuscript posted online 3 April 2020.
Supplemental material is available online only.
- Copyright © 2020 American Society for Microbiology.
