Development of a quadruple qRT-PCR assay for simultaneous identification of highly and low pathogenic H7N9 avian influenza viruses and characterization against oseltamivir resistance

The strains of influenza A and B were previously isolated, and preserved by our group. Viral stocks were prepared in embryonated chicken eggs (Xinxing Dahuanong Biotechnology, Guangzhou, China) with the approval of Animal Ethics Committees from Shenzhen Third People’s Hospital. Viral RNA was extracted from stocks using the QIAamp RNA Viral Kit (Qiagen, Heiden, Germany) according to the manufacturer’s instruction. Viral RNA was eluted in 50 μL of RNA-free buffer and stored at − 80 °C for subsequent use.

MDCK cells (ATCC, Manassas, USA) in 96-well plates were grown to 90% confluence and infected with 10-fold serial dilutions of the H7N9 chicken embryo allantoic fluid for 1 h at 37 °C. Then the cells were overlaid with fresh DMEM plus TPCK-treated trypsin (2 μg/ml). At 3 days post infection (dpi), plates were assessed for the lowest dilution in which 50% of the wells exhibited cytopathology and HA titer. The 50% tissue culture infectious dose (TCID₅₀) values were calculated according to the Reed-Muench method [28].

All the HA and NA sequences of H7N9 used in the present study were downloaded from the NCBI (National Center for Biotechnology Information) and GISAID (Global Initiative on Sharing All Influenza Data), and aligned using the MEGA 7.0 software. Regions covering the insertion mutation in HA cleavage site and the NAI-resistance mutation (R292K) in NA were chosen as the targets for the primer design. The primers and probes were designed using the Primer Premier 5 software. The probes were labeled with four different fluorescent reporter dyes including FAM, VIC, CY5, and ROX at the 5′-end. The FAM and VIC were labeled with the fluorescent quencher dye NFQ-MGB at the 3′-end, while CY5 and ROX were labeled with BHQ. The sequences and genome positions of the primer and probe set are shown in Table 1.

A section of the HP-H7 and N9-K292 genes were amplified with the primers H7-F/H7-R and N9-F/N9-R, respectively. The PCR products were purified using the Gene JET PCR Purification Kit (Thermo, MA, USA) and ligated to the pGEM-T vector (Promega, Madison, USA). Plasmids were quantified using the Nanodrop 2000 Spectrophotometer (Thermo, MA, USA). The copy number (molecules/mL) was calculated using the following equation: [C × A / 660 × L], in which C represents the concentration of plasmid (g/mL) assessed by the optical density measurement; A is the Avogadro number (6.023 × 10²³); L is the length of the plasmid (number of nucleotides); and 660 is an approximation of the molecular weight of a nucleotide (g/mol).

The samples (both clinical and cultured) were tested by a one-step multiplex qRT-PCR in an ABI QuantStudio Dx Real-Time cycler (Applied Biosystems, Foster City, USA). The One Step PrimeScript™ RT-PCR Kit (Takara, Dalian, China) was used as follows: 0.8 μL enzyme mixture (including reverse transcriptase [RT] and Taq polymerase), 10 μL 2 × One Step RT-PCR buffer III, 0.4 μL of each primer and probe for NA gene, 0.6 μL of each primer for HA gene, 0.6 μL of the universal probe for HA gene (H7-P-W), 0.8 μL of the specific probe for HP-H7 gene (H7-P-M), 0.8 μL RNase free water, and 5 μL RNA (total 20 μL/reaction mixture). The concentration of all the primers and probes was 20 μM. The qRT-PCR assay conditions were as follows: reverse transcription for 5 min at 42 °C; 10 s at 95 °C for reverse transcriptase inactivation and DNA polymerase activation followed by 40 cycles of 5 s at 95 °C and 30 s at 60 °C (annealing-extension step). The data were analyzed using the QuantStudio™ Real-Time PCR Software (Applied Biosystems, Foster City, USA). Commercial qRT-PCR kits for the detection of HP-H7N9 viruses were also used (Mabsky Biotech Co., Ltd., Shenzhen, China; Zhengzhou Zhongdao Biotechnology Co., Ltd., Henan, China) following the manufacturers’ instructions. All samples were analyzed in triplicate with three independent runs.

Sputum samples were obtained from ten confirmed cases of H7N9 infection in Shenzhen Third People’s Hospital and Yunnan Center for Disease Control and Prevention. Viral RNA was extracted using the QIAamp RNA Viral Kit (Qiagen, Heiden, Germany) according to the manufacturer’s instruction.

First, we downloaded and compared the HA genes of the HP- and LP-H7N9 virus. The HA genes of the HP-H7N9 virus possessed either a KRTA, KRIA, or KRAA insertion with a nucleic acid 12 bp in size, 5′-AAA CGG A(G)C(T)T GCG-3′. Thus, we designed the forward (H7-F), reverse (H7-R) primers covering this region, and a probe (H7-P-M) which can specifically recognize the specific insertion. Additionally, a universal probe (H7-P-W) targeting both the HP- and LP-H7N9 within the same region was designed to identify the presence of H7. We then downloaded and compared the NA genes of the H7N9 viruses, selected the forward (N9-F), reverse (N9-R) primers covering the R292K mutation region, and designed a probe (N9-P-M) which can specifically recognize the R292K mutation. Meanwhile, a universal probe (N9-P-W) within the same region was designed to detect the presence of N9 genes. The four probes were labeled with four types of different fluorescence. The oligonucleotide sequences, genome positions, labeled fluorescence of the primers and probes for the multiplex assay are listed in Table 1.

The primer-probe set was optimized as a single-well method for H7N9 detection by one-step multiplex qRT-PCR method as described in the Methods section, and the optimized concentrations of the primers and probes are shown in Table 1. To test the specificity of our multiplex qRT-PCR assay, four different H7N9 viruses (LP-H7N9 without NAI-resistance mutation, LP-H7N9 with NAI-resistance mutation, HP-H7N9 without NAI-resistance mutation and HP-H7N9 with NAI-resistance mutation) were used (the identities of viruses used are listed in Table 2). The universal probes of H7 and N9 successfully detected all four phenotypes of H7N9 virus (Fig. 1). Meanwhile, specific probes targeting HP-H7 or N9-K292 specifically recognized the corresponding phenotypes of H7N9 viruses without any nonspecific amplification. In addition, there were also no nonspecific amplification when tested against other influenza viruses (including H1N1, H3N2, H5N6, H6N6 and H9N2 subtype of influenza A and influenza B) (Table 2).

The detection limit of the multiplex qRT-PCR assay was evaluated by quantitative standards from a pGEM-T vector expressing the target sequence of the HP-H7 and N9-K292 genes, or viral RNAs prepared from serial ten-fold dilutions of the four different phenotypes of H7N9 viral stocks. The multiplex qRT-PCR method using DNA standards as template showed that Ct values were similar among the four targets, and the linear range was between 1 × 10⁶ to 1 × 10¹ molecules, respectively. Moreover, the regression coefficients (R²) values were above 0.99, indicating that the assay was both accurate and precise over this range (Fig. 2). The limit of detection was determined to be 50 genome-equivalent copies per reaction, based on specific amplification curves (Fig. 2a) and the standard curve (Fig. 2b).

Meanwhile, RNA samples were extracted from ten-fold serial dilutions of four phenotypes of stock H7N9 viruses ranging from 2 × 10⁶ to 2 × 10^− 2 TCID₅₀/mL, and were used to test the detection limit (Figs. 3 and 4). Results showed that the detection limit of the multiplex qRT-PCR assay was similar among the four H7N9 viruses (2.8 × 10^− 2 TCID₅₀) per reaction, except for LP-H7N9 without NAI-resistance, which was 2.8 × 10^− 3 TCID₅₀ per reaction. The titers correlated well with the obtained Ct values (R² > 0.99) (Figs. 3 and 4).

To validate the assay with clinical samples, we tested 10 sputum specimens obtained from 10 patients confirmed to be infected with influenza A (H7N9) virus during the fifth wave (Table 3). As expected, all samples tested positive for H7N9 viruses with similar Ct values compared to two commercial kits (Table 3). Furthermore, samples from patients SZ05, SZ07 and SZ08 were determined to be HP-H7N9; and samples from patients SZ04, SZ07 and SZ08 were determined to have NAI-resistance mutations. These results were confirmed by Sanger sequencing.