Journal of Virological Methods 119 (2004) 151–158

Application of real-time RT-PCR for the quantitation and competitive replication study of H5 and H7 subtype avian influenza virus Chang-Won Lee, David L. Suarez∗ Southeast Poultry Research Laboratory, USDA-ARS, 934 College Station Road, Athens, GA 30605, USA Received 2 January 2004; received in revised form 24 March 2004; accepted 25 March 2004

Abstract Avian influenza (AI) viruses are endemic in wild birds and if transmitted to poultry can cause serious economic losses. In the study of AI, the quantitation of virus shed from infected birds is valuable in pathogenesis studies and to determine the effectiveness of vaccines, and is performed routinely by cultivation of virus containing samples using embryonating chicken eggs (ECE) and expressed by 50% egg infectious dose (EID50 ). Although, this assay is accurate and is the standard test for infectious virus titration, the method is laborious, requires a large number of ECE, and takes at least 7 days to determine results. In this study, a one-tube hydrolysis fluorescent probe based real-time RT-PCR (RRT-PCR) was applied for the quantitation of AI virus and compared with conventional virus titration method. A strong positive correlation was observed between the amount of RNA determined by quantitative RRT-PCR and the EID50 s determined by conventional methods. This RRT-PCR test was further applied in the study of competitive replication of co-infected H5 and H7 subtype viruses in chickens. Using hemagglutinin subtype specific probes, we were able to determine the amount of individual subtype virus, which could not have easily been done with conventional methods. This RRT-PCR based quantitation of AI virus, which is specific, sensitive, easy to perform, and rapid, will be useful for virological, pathogenesis, and protection studies. © 2004 Elsevier B.V. All rights reserved. Keywords: Avian influenza virus; Real-time RT-PCR; Virus titration; Quantification

1. Introduction Avian influenza (AI) is a disease caused by type A influenza virus, a member of the Orthomyxoviridae family. Influenza A viruses are responsible for major disease problems in birds, as well as in mammals including humans. Infection of domestic poultry by AI viruses typically produces syndromes ranging from mild, localized infection such as respiratory disease and drops in egg production to severe, systemic disease with near 100% mortality. Disease is usually absent with AI virus infection in most wild aquatic bird species, which is the primordial reservoir of all influenza A viruses (Swayne and Halvorson, 2003). In the US, avian influenza viruses have caused considerable economic losses due to decreased production, increased mortality, cost for eradication, quarantine restrictions, and ∗ Corresponding author. Tel.: +1-706-546-3479; fax: +1-706-546-3161. E-mail address: [email protected] (D.L. Suarez).

0166-0934/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jviromet.2004.03.014

trade embargoes on poultry and poultry products. The US government spent over $60 million in the Pennsylvania outbreak in 1983–1984 to eradicate a highly pathogenic (HP) H5N2 virus (Fichtner, 1987; Lasley, 1987). Although the HP form of the virus has not been detected since the Pennsylvania outbreak, periodic outbreaks of low pathogenic H5 and H7 subtype viruses have occurred in the US and continues to be a threat because of its potential to mutate to the HP form of the virus (Lee et al., in press-a; Spackman et al., 2003; Suarez et al., 2003). Retrospective genetic analysis of H5 and H7 subtype isolates from live bird markets (LBMs) provided evidence of association between LBMs and influenza outbreaks in commercial poultry operations (Suarez et al., 1999; Suarez and Senne, 2000; Spackman et al., 2003). In the LBMs located in the Northeastern US, a variety of birds such as chickens, turkeys, guinea fowl, quail, pheasant, peafowl, geese, and ducks are assembled in relatively high numbers with new birds being introduced daily, which provide a favorable environment for mutation of AI virus. Furthermore, because of

152

C.-W. Lee, D.L. Suarez / Journal of Virological Methods 119 (2004) 151–158

the complexity of the distribution channels for birds entering the LBMs, eradication of established AI subtypes have been difficult despite extensive surveillance and control programs (Senne et al., 1992, 1996; Panigrahy et al., 2002). In a previous study, a one-step real-time RT-PCR (RRT-PCR) assay using fluorogenic hydrolysis type probes was developed for the detection of type A influenza virus and the subsequent identification of the H5 and H7 subtypes (Spackman et al., 2002). This RRT-PCR assay demonstrated high correlation with virus isolation (VI) in embryonating chicken eggs (ECE) in a study with clinical samples from LBMs. Because of its capacity to deal with large number of samples in a rapid, sensitive, and specific manner, it became a feasible alternative to VI in ECE as a diagnostic or screening tool for AI virus. In the present study, the RRT-PCR test was further applied to the quantitation of AI virus for research purposes. The quantitation of virus shed from infected birds is a valuable tool in the study of viral pathogenesis. It also is valuable for determining the effectiveness of influenza vaccines, because effective vaccines not only prevent clinical disease but reduces viral shed, which reduces the likelihood of transmission of virus to new flocks (Swayne and Halvorson, 2003). Currently, virus quantitation is done by titration using ECE and is expressed as 50% embryo infective dose (EID50 ) (Villegas, 1998). However, virus titration using ECE is labor-intensive from sample treatment to virus identification, requires the availability of large numbers of ECE and takes at least 7 days to complete the test, which routinely limits the number of samples to be tested in the experiments. Thus, a correlation between the amount of viral RNA determined by quantitative RRT-PCR (qRRT-PCR) and the virus titer determined by conventional methods could simplify and accelerate the study of AI virus. The purpose of this study is three-fold. First, we sought to determine whether qRRT-PCR could replace the conventional virus titration method by comparing the two tests. Second, since RRT-PCR detection is based on sequence-specific probes, we sought to apply the qRRT-PCR assay for the quantitation of individual viruses from samples containing different hemagglutinin subtypes of viruses. Third, in addition to the development of qRRT-PCR as a titration method, we also wanted to study the replication and transmission competency of H5 and H7 subtype viruses to assess their potential risk to US poultry.

2. Materials and methods 2.1. Viruses The virus isolates used in this study are listed in Table 1. The viruses were obtained from the repository of the Southeast Poultry Research Laboratory (SEPRL) and were passaged one or two additional times in ECE to make working stocks of the virus.

Table 1 Virus strains used in this study Virus strain

Subtype

Abbreviation

Reference

CK/TX/167280-4/02 CK/NJ/118878-5/01 CK/PA/13609/93

H5N3 H7N2 H5N2

H5-02 H7-01 H5-93

TK/VA/55/02

H7N2

H7-02

Lee et al. (in press-a) Spackman et al. (2003) Horimoto and Kawaoka (1995) Spackman et al. (2003)

2.2. Primers and probes Sequences of the three primer and hydrolysis probe sets specific for matrix (M), H5, and H7 genes have been previously described (Spackman et al., 2002). The M gene-specific primer and probe set was designed for the detection of all type A influenza viruses. The H5- and H7-specific primer and probe sets were primarily targeted to North American H5 and H7 AI viruses. 2.3. RNA extraction RNA was extracted using the RNeasy mini kit (Qiagen, Valencia, CA). For virus-infected allantoic fluid and fluid containing tracheal swabs, the same procedure was applied as described previously (Spackman et al., 2002). For fluid containing cloacal swabs, 500 ␮l sample fluid was mixed with 500 ␮l of 70% ethanol and 500 ␮l of kit-supplied RLT buffer. This mixture of fluid was clarified by centrifugation at 12,000 × g for 1 min and the entire sample was applied to the RNeasy spin column. The rest of the procedure was followed as recommended by the manufacturer. 2.4. Production of control RNAs Generation of in vitro-transcribed matrix, H5, and H7 gene RNAs was previously described (Spackman et al., 2002). These control RNAs were used to determine the detection limits of the assay and to compare the amplification efficiency of in vitro-transcribed RNA and complete viral RNA. For the qRRT-PCR study, RNA was extracted from allantoic fluid containing each of the four viruses of known virus titer as described below. 2.5. RRT-PCR One-tube RRT-PCR was performed using the Qiagen one-step RT-PCR kit (Qiagen) in a 25-␮l reaction mixture containing 2 and 0.8 ␮l of kit-supplied enzyme mixture and dNTP mix, respectively, 10 pmol of each primer, 13 units of RNase inhibitor (Promega, Madison, WI), and 0.15 M of probe. Eight microliters of each sample and control RNAs were amplified using the SmartCycler (Cepheid, Sunnyvale, CA). The RT-PCR program consisted of 30 min at 50 ◦ C and 15 min at 95 ◦ C for all primer sets. A two-step PCR cycling protocol was used for the matrix gene primer set

C.-W. Lee, D.L. Suarez / Journal of Virological Methods 119 (2004) 151–158

which consisted of 40 cycles of 94 ◦ C for 1 s and 60 ◦ C for 20 s. For H5 and H7 specific primer sets, a three-step cycling protocol was used as 95 ◦ C for 10 s, 54 ◦ C for 30 s, and 72 ◦ C for 10 s for 40 cycles. Fluorescence data were acquired at the end of each annealing step. 2.6. Virus titration using ECE Virus containing allantoic fluid or swab samples were titrated and expressed as the 50% egg infectious dose (EID50 ) using ECE as previously described (Villegas, 1998). Briefly, 200 ␮l of each dilution of sample suspended in Brain Heart Infusion medium (BHI, Difco, Detroit, MI) was inoculated into five 10-day-old ECE. At 5 days post-inoculation (DPI), allantoic fluid was collected and the hemagglutination (HA) test was performed to determine the presence of the virus. The virus titer was determined using the Reed and Muench method (Reed and Muench, 1938). 2.7. qRRT-PCR For quantitation, swab samples were run together with known amounts of control viral RNA. To prepare control RNA, each of the four viruses used in this study (Table 1) were titrated using ECE as described above, and RNA were extracted from serially diluted viruses (101.0 –106.0 EID50 /ml). Standard curves were generated with those control viral RNAs and the amount of RNA in the samples was converted into EID50 /ml by interpolation. 2.8. Animal experiment I: comparison of virus titration using ECE and qRRT-PCR Two separate experiments were done for this study. In the first experiment, 3-week-old SPF White Rock (WR) chickens (n = 5 birds per group) obtained from flocks maintained at SEPRL were inoculated intranasally with H5-02 and H7-01 AI viruses with titer of 105.0 EID50 /0.2 ml. Tracheal and cloacal swabs were collected at 3 and 7 days post-infection (DPI). Individual tracheal and cloacal swabs were suspended in 1.5 ml BHI broth containing antibiotics (10,000 IU of penicillin G, 1000 ␮g of gentamicin, and 20 ␮g of amphotericin B/ml), and qRRT-PCR was performed with extracted RNA using matrix-specific primers and probe set. The second experiment was conducted the same way as the first experiment with H5-93 and H7-02 viruses. Exceptions are 10 4-week-old WR birds per group were used, and swabs were put in 1.2 ml of BHI broth. 2.9. Animal experiment II: competitive replication study of H5 and H7 subtype viruses Two groups of 10 birds were inoculated with mixtures of H5 and H7 AI virus containing the equivalent of 105.0 EID50 /0.2 ml per virus. Five infected birds were transferred and placed in the same cage with five uninoculated

153

birds. Tracheal swabs were collected at 3 and 5 DPI from both the infected and contact control birds. Individual swabs were suspended in 1.2 ml of BHI broth and qRRT-PCR was conducted with M-, H5-, and H7-specific primer and probe sets. The five infected birds that were reared in the same cage with the uninoculated birds were euthanized at 5 DPI to prevent overcrowding of cages. The remaining chickens were euthanized at 14 DPI after blood was drawn. Seroconversion was analyzed by hemagglutinin inhibition (HI) assay (Beard, 1989). 2.10. Statistical analysis Virus titration results determined by conventional and qRRT-PCR methods and viral titers obtained at different DPI were analyzed for significance (P < 0.05) using Student’s t-test. Correlation coefficient (r-value) between the amounts of virus shed determined by conventional and qRRT-PCR methods were obtained using the Excel (Microsoft) program. The minimum viral titer detected by conventional and qRRT-PCR methods in this study was determined to be 101.0 EID50 /ml. Thus, positive samples which was less than 101.0 EID50 /ml were given a numeric value of 100.9 EID50 /ml for statistical purposes.

3. Results 3.1. Sensitivity and amplification efficiency of the RRT-PCR The detection limit of the RRT-PCR assay in this study was as low as 10−1 fg or 10 gene copies of in vitro transcribed RNA with M- and H7-specific probes. The sensitivity of H5-specific probe was about 10-fold lower and the detection limit was 1 fg or 102 gene copies of RNA. To verify that the amplification efficiencies of the in vitro-transcribed RNA and the viral RNA are equal, RRT-PCR was performed with both types of RNA that were serially diluted. Standard curves for those RNAs were generated by plotting their cycle threshold numbers (CT ) versus their dilution factors (Fig. 1). The efficiencies can be considered as equal if the difference of the slopes (s) of standard curves is smaller than 0.1 (Gut et al., 1999). In the RRT-PCR assay with three different gene-specific probes, high correlation (r 2 > 0.99) between cycle number and dilution factors were observed with all three different probes used, and s values from each comparison with different probes was less than 0.1, showing equal amplification efficiency of complete viral RNA compared to in vitro-transcribed RNA and verifying that the standard curve generated with viral RNA is reliable and can be applied in AI virus quantitation. Further, s of less than 0.1 were also observed among standard curves generated with different gene specific probes. Examples of standard curves generated with each gene-specific probes are shown in Fig. 1.

154

C.-W. Lee, D.L. Suarez / Journal of Virological Methods 119 (2004) 151–158

Fig. 1. Amplification efficiencies of in vitro-transcribed RNA (䊉) and complete viral RNA (). Three independent RRT-PCR was performed with M-, H5-, and H7-specific probes and calibration curves were generated by plotting the threshold cycle numbers (CT ) against the corresponding dilution factors. s indicate the differences of the slopes.

3.2. Comparison of virus titration using ECE and qRRT-PCR

3.3. Competitive replication study of H5 and H7 subtype viruses

To determine the replication competency of four recent H5 and H7 viruses, chickens were infected with one of four viruses and the amount of viral shedding was determined using conventional and qRRT-PCR methods (Table 2). At 3 DPI, high titers of virus (more than 104.0 EID50 /ml) were detected from tracheal swabs from birds in all four groups. However, the titer of virus in the trachea decreased significantly (P < 0.001) and the number of virus positive birds also decreased at 7 DPI. Only the H7-01 virus infected chickens retained a relatively high titer in the trachea and all five birds tested were positive. From cloacal swabs, virus was only detected from birds infected with the H7 subtype viruses. A strong positive correlation (r = 0.972) between the EID50 s measured by conventional method and qRRT-PCR assay were observed. Furthermore, in each group there was no significant difference (P > 0.21) in mean virus titer obtained with tracheal swabs taken at 3 DPI between the two methods and lower values of standard deviation were consistently observed with titer measured by qRRT-PCR than the conventional virus titration method.

H5 and H7 subtype viruses were evaluated for their competitive replication competency and ability to spread from infected chickens to contact control birds (Table 3). Since quantitation of the amount of individual subtype viruses using conventional methods is difficult and laborious, qRRT-PCR with H5- and H7-sepcific probe was applied to determine the virus titer. There was no statistically significant difference between H5 and H7 subtype virus titer, although H5-02 replicated slightly better than the H7-01 in co-infected birds. Furthermore, both H5 and H7 viruses were highly efficient in transmission of the virus to the contact control birds and large amounts of virus were detected from tracheal swabs taken at 3 and 5 DPI. Efficient replication and transmission ability of both subtype viruses were further confirmed by high HI antibody titer detected 2 weeks after post-infection or post-transmission. The accuracy of the virus titers determined by H5and H7-specific qRRT-PCR were evaluated by comparing the sum of the H5 and H7 titers with the titer obtained with M-specific qRRT-PCR. There was a good correlation (r = 0.902) and no significant difference in mean virus titer between the two amounts, which indicate quantitation of

C.-W. Lee, D.L. Suarez / Journal of Virological Methods 119 (2004) 151–158

155

Table 2 Comparison of conventional virus titration and qRRT-PCR Viruses

Experiment I CK/TX/167280-4/02 (H5)

Bird number

Mean ± S.D.

7 DPI

3 DPI

7 DPI

ECE

qRRT-PCR

ECE

qRRT-PCR

ECE

qRRT-PCR

1 2 3 4 5

4.78 4.23 4.23 4.90 3.57

4.53 4.83 3.95 4.39 4.26

– <1 2.40 – –

–

– – – – –

– – – – –

– – – – –

– – – – –

4.34 + 0.53

4.39 + 0.33

1 2 3 4 5

4.78 4.23 4.90 4.78 4.23

4.99 4.22 4.45 4.72 4.66

1.57 <1 1.78 1.90 2.03

– – – – –

– – – – –

3.03 – – <1 –

2.53 – – <1 –

4.58 + 0.33

4.61 + 0.29

1.64 + 0.44

1.83 + 0.28

3.90 4.40 3.90 5.03 4.03 5.24 4.90 5.03 5.03 4.90

3.91 4.16 3.83 5.12 4.15 4.65 4.76 4.66 5.15 4.65

– – – – – – – – – –

– – – – – – – – – –

– – – – – – – – – –

– – – – – – – – – –

– – – – – – – – – –

– – – – – – – – – –

4.64 + 0.52

4.50 + 0.47

5.47 5.08 5.57 4.72 5.23 5.57 4.90 5.03 4.90 4.40

5.12 4.66 5.24 5.28 5.01 5.23 4.48 4.89 4.97 4.66

– – – – – –

– <1 – <1 – – 2.92 – – –

– – <1 – – – – – 3.57 –

– – 1.25 – – – – – 3.84 –

– – – – – – – –

– – – – – – – –

5.09 + 0.38

4.95 + 0.28

1 2 3 4 5 6 7 8 9 10

Mean ± S.D. TK/VA/55/02 (H7)

3 DPI qRRT-PCR

Mean ± S.D. Experiment II CK/PA/13609/93 (H5)

Cloacal swab

ECE

Mean ± S.D. CK/NJ/118878-5/01 (H7)

Tracheal swab

1 2 3 4 5 6 7 8 9 10

3.03 – – –

1.58 2.26 – –

1.54 1.67 1.75 2.25 1.95

2.23 –

1.95 –

Virus titer was obtained either by conventional method using ECE or by qRRT-PCR with swabs taken at 3 and 7 days post-infection (DPI). Virus titer is expressed as log10 50% egg infective doses/ml.

individual subtype from a mixture of different viruses can also be determined with subtype-specific qRRT-PCR. 4. Discussion Real-time PCR is a relatively new technique, but it has been widely used for the diagnosis and study of avian pathogens (Kim et al., 2002; Jackwood and Sommer, 2002; Carli and Eyigor, 2003; Jackwood et al., 2003). It also has been successfully applied in the screening of AI virus field samples and is a reliable alternative to virus isolation in ECE (Spackman et al., 2002). In this study, we further applied this technique for influenza virus research. Conventionally,

AI virus titration is performed by cultivating virus containing samples in ECE, which is labor intensive, requires a large supply of ECEs and takes at least 7 days. Furthermore, since this system uses ECE as a growth medium, the degree of adaptation of individual viral strains to eggs can result in misinterpretation of the actual infectious virus titer in vivo. Our results shows qRRT-PCR is a feasible alternative to traditional virus titration methods for many reasons. First, qRRT-PCR is a sensitive test and was able to detect as low as 1 fg of in vitro transcribed RNA with all three-gene specific probes. In terms of EID50 , qRRT-PCR was able to detect viral titers lower than 101.0 EID50 which was more sensitive than ECE method (Table 2). This

156

C.-W. Lee, D.L. Suarez / Journal of Virological Methods 119 (2004) 151–158

Table 3 H5 and H7 subtype AI virus co-infection study Group

H5/02 + H7/01 infected

H5/02 + H7/01 contacted

H5/93 + H7/01 infected

H5/93 + H7/01 contacted

a b c d

Testsa

Mean ± S.D.

Virus titersb from individual swabs 1

2

3

4

5

qRRT-PCR, 3 DPI H5 H7 H5 + H7 M

4.92 3.65 4.94 5.05

4.76 3.55 4.79 5.06

5.53 3.93 5.54 5.53

4.78 5.29 5.41 5.36

4.46 2.37 4.47 4.67

4.89 3.76 5.03 5.13

± ± ± ±

0.39 1.04 0.44 0.33

qRRT-PCR, 5 DPI H5 H7 H5 + H7 M

2.83 0.95 2.84 3.10

4.09 3.71 4.24 4.45

3.83 2.44 3.85 4.16

3.21 3.75 3.86 3.83

2.30 2.11 2.52 2.73

3.49 2.59 3.46 3.79

± ± ± ±

0.57 1.18 0.74 0.75

HIc , 2 DPI H5 ag H7 ag

6d 8

7 7

6 9

6 7

7 7

qRRT-PCR, 3 DPI H5 H7 H5 + H7 M

4.71 4.28 4.85 5.54

3.84 3.87 4.16 4.09

4.96 1.73 4.97 5.47

4.65 3.37 4.67 4.71

4.25 4.41 4.64 4.88

4.48 3.53 4.66 4.93

± ± ± ±

0.44 1.09 0.31 0.59

qRRT-PCR, 5 DPI H5 H7 H5 + H7 M

3.32 2.35 3.37 4.79

4.65 3.45 4.68 5.00

4.69 3.31 4.70 4.92

5.34 3.35 5.35 5.67

3.86 4.04 4.26 5.38

4.37 3.30 4.42 5.19

± ± ± ±

0.79 0.61 0.83 1.01

HI, 2 WPI H5 ag H7 ag

6 9

8 6

6 7

7 8

6 10

qRRT-PCR, 3 DPI H5 H7 H5 + H7 M

4.64 5.04 5.20 4.95

3.61 3.54 3.59 3.63

4.92 5.02 5.27 5.11

4.41 4.79 4.94 4.62

4.72 4.74 5.04 5.12

4.46 4.63 4.71 4.69

± ± ± ±

0.51 0.62 0.76 0.62

qRRT-PCR, 5 DPI H5 H7 H5 + H7 M

3.19 2.49 3.27 3.32

2.93 2.50 3.07 3.00

3.11 2.71 3.26 3.02

1.85 2.53 2.62 2.14

3.26 2.91 3.42 3.36

2.87 2.66 3.13 2.96

± ± ± ±

0.58 0.19 0.31 0.57

HI, 2 WPI H5 ag H7 ag

9 8

9 10

8 7

9 8

9 8

qRRT-PCR, 3 DPI H5 H7 H5 + H7 M

2.94 3.74 3.80 3.35

3.91 1.32 3.91 3.59

0.95 3.01 3.01 2.93

5.34 2.57 5.35 5.63

3.62 1.46 3.62 3.58

3.35 2.42 3.97 3.82

± ± ± ±

1.60 1.03 0.99 1.05

qRRT-PCR, 5 DPI H5 H7 H5 + H7 M

2.35 3.34 3.39 2.83

2.52 4.83 4.83 4.44

<1.0 4.99 4.99 4.72

3.70 4.43 4.51 4.21

2.68 4.93 4.94 4.59

2.11 4.50 4.53 4.16

± ± ± ±

0.82 0.69 0.67 0.77

HI, 2 WPI H5 ag H7 ag

6 5

6 8

4 9

7 8

6 9

qRRT-PCR was performed with H5, H7, and matrix (M) specific primer sets at 3 and 5 days post-infection (DPI). Virus titer is expressed as log10 50% egg infective doses/ml. Hemagglutinin inhibition (HI) test were done with H5 and H7 antigen (ag) with sera obtained at 2-week post-infection (WPI). HI antibody titer is expressed as log 2 reciprocal of the endpoint in two-fold sera dilution.

6.4 ± 0.6 7.6 ± 0.9

6.6 ± 0.9 8.0 ± 1.6

8.8 ± 0.5 8.2 ± 1.1

5.8 ± 1.1 7.8 ± 1.6

C.-W. Lee, D.L. Suarez / Journal of Virological Methods 119 (2004) 151–158

increased sensitivity of RRT-PCR is likely due to detection of RNA from incompletely packaged virus particles or viral RNA from infected cells. Potentially this could result in false positives, but influenza viral RNA from samples from live animals is an important finding. Second, qRRT-PCR is a rapid method that can be done within a half-day while the conventional titration method takes at least a week to complete. Third, in addition to its speed, qRRT-PCR is easy to perform and reduces the handling of infectious material as compared to traditional methods. This may reduce the risk of cross-contamination and increase reproducibility. We consistently observed a lower standard deviation with qRRT-PCR than the ECE method, and this uniformity will be invaluable in vaccine efficacy or protection studies that are routinely done for AI virus studies (Lee et al., in press-b,c). Fourth, and most importantly, a good correlation was observed between the two methods, which verifies that qRRT-PCR is a suitable replacement for viral titration in ECEs. In addition to titration of total influenza virus present in a sample, we were able to quantitate different viral strains in a mixed infection. This discriminative property, which is not easily available with the traditional method, provides a new tool for AI virus study. We applied this tool to compare the replication and transmission competency of H5 and H7 subtype viruses. In the LBMs in the Northeast US, H5N2 influenza was the predominant strain isolated before 1994, but after that time H7N2 became the dominant isolate (Panigrahy et al., 2002). Furthermore, the H7 subtype virus has been associated with several commercial poultry outbreaks in the Northeastern US (Spackman et al., 2003). This observation implied that the H7N2 virus had better fitness in poultry as compared to the other subtypes of avian virus that were introduced in the LBM. Our in vivo comparison showed little difference between the H5 and H7 viruses studied, although H7 virus was shed for longer in the trachea and was detected more often from cloacal samples (Table 2). These results indicate that the chicken origin H5 viruses examined in our study were also well adapted to poultry and can be a threat to our poultry population. Further studies regarding the susceptibility of different avian species to those two subtypes and the persistence of the virus in the host, are needed to better understand the true risk of those viruses. In conclusion, we tested and validated the feasibility of qRRT-PCR as a means to titrate the AI virus. Furthermore, with HA specific probes, viruses with different subtypes could be quantitated from a mixture of different viruses. This RT-PCR based quantitation of AI virus will be an invaluable tool for many aspects of virological, pathogenesis, and vaccine protection studies.

Acknowledgements The authors wish to thank Suzanne Deblois for technical assistance and Roger Brock for animal care assistance. The

157

authors also wish to thank Dr. Erica Spackman for critical review of this manuscript. This research was supported by USDA, ARS CRIS project 6612-32000-039. References Beard, C.W., 1989. Serological procedures. In: A Laboratory Manual for the Isolation and Identification of Avian Pathotypes. Kendall-Hunt Publishing, Dubuque, pp. 192–200. Carli, K.T., Eyigor, A., 2003. Real-time polymerase chain reaction for Mycoplasma gallisepticum in chicken trachea. Avian Dis. 47, 712–717. Fichtner, G.J., 1987. The Pennsylvania/Virginia experience in eradication of avian influenza (H5N2). In: Beard, C.W., Easterday, B.C. (Eds.), Proceedings of the Second International Symposium on Avian Influenza. US Animal Health Association, Richmond, VA, pp. 33–38. Gut, M., Leutenegger, C.M., Huder, J.B., Pedersen, N.C., Lutz, H., 1999. One-tube fluorogenic reverse transcription-polymerase chain reaction for the quantitation of feline coronaviruses. J. Virol. Methods 77, 37– 46. Horimoto, T., Kawaoka, Y., 1995. Molecular changes in virulent mutants arising from avirulent avian influenza viruses during replication in 14-day-old embryonated eggs. Virology 206, 755–759. Jackwood, D.J., Sommer, S.E., 2002. Identification of infectious bursal disease virus quasispecies in commercial vaccines and field isolates of this double-stranded RNA virus. Virology 304, 105–113. Jackwood, M.W., Hilt, D.A., Callison, S.A., 2003. Detection of infectious bronchitis virus by real-time reverse transcriptase-polymerase chain reaction and identification of a quasispecies in the Beaudette strain. Avian Dis. 47, 718–724. Kim, Y., Gharaibeh, S.M., Stedman, N.L., Brown, T.P., 2002. Comparison and verification of quantitative competitive reverse transcription polymerase chain reaction (QC-RT-PCR) and real time RT-PCR for avian leukosis virus subgroup J. J. Virol. Methods 102, 1–8. Lasley, F.A., 1987. Economics of avian influenza: control vs. noncontrol. In: Beard, C.W., Easterday, B.C. (Eds.), Proceedings of the Second International Symposium on Avian Influenza. US Animal Health Association, Richmond, VA, pp. 390–399. Lee, C.W., Senne, D.A., Linares, J.A., Woolcock, P., Stallknecht, D., Spackman, E., Swayne, D., Suarez, D.L., in press-a. Characterization of recent H5 subtype avian influenza viruses from US poultry. Avian Pathol. Lee, C.W., Senne, D.A. Suarez, D.L., in press-b. Generation of reassortant influenza vaccines by reverse genetics that allows utilization of a DIVA (differentiating infected from vaccinated animals) strategy for the control of avian influenza. Vaccine. Lee, C.W., Senne, D.A., Suarez, D.L., in press-c. Effect of vaccine use in the evolution of Mexican-lineage H5N2 avian influenza virus. J. Virol. Panigrahy, B., Senne, D.A., Pedersen, J.C., 2002. Avian influenza virus subtypes inside and outside the live bird markets, 1993–2000: a spatial and temporal relationship. Avian Dis. 46, 298–307. Reed, L.J., Muench, H., 1938. A simple method for estimating fifty percent endpoints. Am. J. Hyg. 27, 493–497. Senne, D.A., Pearson, J.E., Panigrahy, B., 1992. Live poultry markets: a missing link in the epidemiology of avian influenza. In: Easterday, B.C. (Ed.), Proceedings of the Third International Symposium on Avian Influenza. University of Wisconsin, Madison, WI, pp. 50–58. Senne, D.A., Panigrahy, B., Kawaoka, Y., Pearson, J.E., Suss, J., Lipkind, M., Kida, H., Webster, R.G., 1996. Survey of the hemagglutinin (HA) cleavage site sequence of H5 and H7 avian influenza viruses: amino acid sequence at the HA cleavage site as a marker of pathogenicity potential. Avian Dis. 40, 425–437. Spackman, E., Senne, D.A., Davison, S., Suarez, D.L., 2003. Sequence analysis of recent H7 avian influenza viruses associated with three different outbreaks of commercial poultry in the US. J. Virol. 77, 13399–13402.

158

C.-W. Lee, D.L. Suarez / Journal of Virological Methods 119 (2004) 151–158

Spackman, E., Senne, D.A., Myers, T.J., Bulaga, L.L., Garber, L.P., Perdue, M.L., Lohman, K., Daum, L.T., Suarez, D.L., 2002. Development of a real-time reverse transcriptase PCR assay for type A influenza virus and the avian H5 and H7 hemagglutinin subtypes. J. Clin. Microbiol. 40, 3256–3260. Suarez, D.L., Spackman, E., Senne, D.A., 2003. Update on molecular epidemiology of H1, H5, and H7 influenza virus infections in poultry in North America. Avian Dis. 47, 888–897. Suarez, D.L., Garcia, M., Latimer, J., Senne, D.A., Perdue, M., 1999. Phylogenetic analysis of H7 avian influenza viruses isolated from the live bird markets of the Northeast United States. J. Virol. 73, 3567– 3573.

Suarez, D.L., Senne, D.A., 2000. Sequence analysis of related low-pathogenic and highly pathogenic H5N2 avian influenza isolates from United States live bird markets and poultry farms from 1983 to 1989. Avian Dis. 44, 356–364. Swayne, D.E., Halvorson, D.A., 2003. Influenza. In: Saif, Y.M., Barnes, H.J., Glisson, J.R., Fadly, A.M., McDougald, L.R., Swayne, D.E. (Eds.), Diseases of Poultry. Iowa State University Press, Ames, IA, pp. 135–160. Villegas, P., 1998. Titration of biological suspensions. In: Swayne, D.E. (Ed.), A Laboratory Manual for the Isolation and Identification of Avian Pathogens, fourth ed. American Association of Avian Pathologists, Kennett Square, PA, pp. 248–254.