Protective efficacy of vaccination against highly pathogenic avian influenza is dramatically suppressed by early infection of chickens with reticuloendotheliosis virus

Abstract

The present study was conducted to understand epidemiological factors that depress the protective efficacy of vaccination against highly pathogenic avian influenza (HPAI). Specific pathogen free chickens were infected at different ages with immunosuppressive reticuloendotheliosis virus (REV), then vaccinated with inactivated vaccine H5N2 and challenged with HPAI H5N1. Compared with control birds, early REV infection significantly inhibited antibody response to H5N2 vaccine and dramatically decreased protective efficacy. Immunosuppressive effects lasted for 4 months, and high mortality occurred in chickens receiving three vaccinations during 3 months prior to challenge with HPAI H5N1. Epidemiological studies indicated that REV infection and vertical transmission is common in chickens in some areas of southern China. We conclude that vaccination programmes against HPAI will not be fully effective if other immunosuppressive viral infections, such as REV, exist in chicken flocks.

Introduction

Highly pathogenic avian influenza (HPAI) has been recognized as a global problem, especially in certain Asian countries, and human infections with H5N1 avian influenza virus (AIV) were reported in Hong Kong in 1997 (Li et al., 2004). World authorities are concerned that an HPAI H5N1 strain will mutate and cause pandemic infection in humans. To decrease and prevent this possibility, poultry depopulation programmes have been conducted in infected areas at the beginning of HPAI outbreaks in Asia as recommended by the Office International des Epizooties (OIE). However, this has not successfully eradicated HPAI from South-east Asian countries and China, perhaps because of cultural and economic reasons in the implementation of such programmes. When HPAI was seen in chicken and duck flocks in China in 2005, a vaccination programme was enforced nationwide, as well as depopulation of affected poultry farms. However, HPAI cases still occurred occasionally in some vaccinated flocks. Although vaccination induced satisfactory antibody reactions to H5N1 in the bird experiments in laboratories (Jiang et al., 2005), some vaccinated flocks demonstrated poor, or no, antibody responses, especially after one vaccination in young flocks. In a previous study, we demonstrated that reticuloendotheliosis virus (REV) infection in 1-day-old specific pathogen free (SPF) chickens dramatically decreased antibody responses to H5N2 inactivated vaccine (Sun et al., 2006). However, it was not clear how long this immunosuppression lasted, and whether it decreased the protective efficacy of vaccines against the H5 subtype of AIV.

REV is one of the most common immunosuppressive viruses, and can depress immunity induced by vaccines against Marek's disease or Newcastle disease, as well as humoral responses to bacteria (Witter et al., 1979). Epidemiological studies in recent years have demonstrated that REV infections were very common in chicken flocks in China, especially in southern regions. REV was often isolated from chickens suffering different syndromes and lesions, such as tumours, growth retardation, pericarditis, perihepatitis, and even from eggs. Usually REV infection involved co-infection with other viruses (Sun & Cui, 2007). It is therefore reasonable to evaluate whether common REV infections interfere with vaccination programmes against HPAI in the field. In the present study we demonstrated that an early REV infection can reduce immunity, and can severely decrease the protective efficacy of vaccines, against HPAI.

Materials and Methods

Viruses. REV

A molecularly cloned strain REV-C99 (Ji et al., 2005) of the original SNV strain (Trager, 1959) was used to study its inhibitory effects on immune reactions of chickens to an inactivated vaccine of H5N2. The stock virus had a titre of 10⁶ median tissue culture infectious dose (TCID₅₀)/ml based on an indirect fluorescence antibody assay with REV-specific monoclonal antibody using chick embryo fibroblasts in 96-well plates.

HPAI H5N1 virus for challenge studies

A new field strain (A/chicken/ZW/06) was isolated from an outbreak in China (unpublished data) and used in the challenge studies. It has an FJ-like sublineage (Smith et al., 2006) based on the HA gene.

Birds and bird experimental facilities

All chickens were hatched from SPF eggs from the SPAFAS Company in Jinan. They were reared in separate isolators receiving filtered positive-pressure air in an SPF animal facility. All bird experiments were carried out following the regulations of the Laboratory Animal Management Committee of Shandong Province. For challenge studies, birds were moved into a biosafety level 3 (BSL3) facility.

Experimental design. Experiment I

All chickens were divided into four groups (about 30 birds per group) after hatch and placed in separated SPF isolators. Groups 1 and 2 were inoculated intraperitoneally with REV-C99 of 10³ TCID₅₀ at 1 day or 15 days of age. Groups 3 and 4 were not infected with REV. At 15 days of age, birds in Groups 1 to 3 were inoculated with inactivated H5N2 oil–water emulsion vaccine (0.5 ml each subcutaneously in the neck, License No. 2006-150132077; YEBIO Company, Qindao, China). Serum samples were collected 5 weeks after vaccination for determining antibody titres to H5N2. One-half of the birds from each group were moved into the BSL3 facility for the first challenge test. Each bird from each of the four groups was inoculated with 10⁶ chicken median lethal dose (LD₅₀) of the HPAI strain A/chicken/ZW/06(H5N1) intramuscularly. One-half of the birds in Groups 1 to 3 remained in the SPF facility until the second vaccination against H5N2 at 7 weeks. At 3 and 5 weeks after the second vaccination, serum samples were collected and then these birds were moved into the BSL3 facility for challenge with 10⁶ chicken LD₅₀ doses of the same HPAI strain. In both challenges, death was recorded each day, and dead birds were checked in necropsy during the first 8 days after challenge according to the standard procedure described by the OIE Manual of Standards 1996. Mortality comparisons between different groups were analysed statistically by the chi-squared test.

Experiment II

This was similar to the Experiment I, but for a longer period. Groups 1 and 2 were inoculated intraperitoneally with REV-C99 of 10³ TCID₅₀ respectively at 1 day or 2 weeks of age. Groups 3 and 4 were not infected with REV. Birds in Groups 1 to 3 were vaccinated three times (8 days, 7 weeks and 12 weeks of ages), as in Experiment I. One-half of the birds from each group were moved into the BSL3 facility for the first challenge with 10⁵ LD₅₀ HPAI H5N1 5 weeks after the second vaccination. The other half of the birds in Groups 1 to 3 were still kept in the SPF facility until the third vaccination at 12 weeks of age. Five weeks after the third vaccination, all birds were moved into BSL3 facility for challenge with 10⁶ LD₅₀ H5N1. Serum samples were collected for antibody titres 5 weeks after the first vaccination and 3 weeks after the second and third vaccinations. Mortality rates of different groups were compared and the difference was statistically analysed by the chi-square test.

Determination of haemagglutination inhibition antibody titres

Haemagglutination inhibition (HI) titres of serum samples to subgroup H5 AIV were determined at different times after vaccination using the same isolate of H5N2 used for vaccination as the antigen for HI tests according to the OIE standard procedures. The HI titres between REV-infected and control groups were statistically analysed by Student's t test.

Results

Inhibitory effects of REV infection on HI antibody titres after vaccination

REV infection at 1 day of age strongly inhibited the antibody response to H5 AIV after vaccination (Table 1). The HI antibody titres in REV-infected chickens were significantly lower than those of control birds. This immunosuppressive effect persisted for at least 4 months (in Experiment II). However, the immunosuppressive effects of REV infection were age related. A late REV infection with 1000 TCID₅₀ at 8 days or 15 days of age induced much weaker immunosuppression than early REV infection. HI titres in chickens inoculated with REV at 8 days of age were only significantly lower than the control after one vaccination, and the immunosuppressive effects were not seen after two or three vaccinations (Table 1).

Table 1.  Inhibitory effects of REV infection on HI antibody titres to H5 AIV after vaccination

REV inoculation			HI titres (log2)
Group	Age (days)	Dose (TCID50)	5 weeks after first vaccination at 2 weeks of age	3 weeks after second vaccination at 7 weeks of age	3 weeks after third vaccination at 12 weeks of age
Experiment I
1	1	1000	0.62±1.5(26)^A	3.55±3.3(20)^A	ND
2	15	1000	5.77±2.9(26)^B	6.69±2.0(13)^B	ND
3	Control	–	6.25±2.0(28)^B	6.69±1.4(13)^B	ND
Experiment II
1	1	1000	0.13±0.4 (31)^A	2.44±2.8(27)^A	5.41±3.7(27)^A
2	8	1000	4.1±3.5(28)^B	6.74±3.4(27)^B	8.75±1.5(32)^B
3	Control	–	8.52±1.5(34)^C	9.35±2.7(34)^B	9.32±1.3(22)^B
Data presented as the mean±standard error. Values followed by the same uppercase superscript letter in the same column of the same experiment do not differ significantly (P > 0.05) by t-test analysis. ND, not done.

Inhibitory effects of REV infection on protective efficacy against an HPAI H5 field strain after vaccination

Challenge studies demonstrated that an early REV infection could significantly decrease the protective efficacy of vaccinations against H5 AIV. As indicated in Table 2, chickens could not be completely protected from a challenge with the heterologous field strain A/chicken/ZW/06(H5N1) 5 weeks after the first vaccination, but the mortality in Group 1 inoculated with REV at 1 day was significantly higher than that in vaccinated control Group 3 (22/23 versus 7/13, P < 0.05). This immunosuppression on protective efficacy after vaccination lasted for at least 3 to 4 months. Mortality rates after virus challenge in chickens infected with REV at 1 day of age (Group 1 in both Experiments I and II) were significantly higher than the control chickens without REV infection (Group 3) 5 weeks after the second vaccination (14/19 versus 1/13 in Experiment I and 4/22 versus 0/14 in Experiment II, P < 0.05), and even 5 weeks after the third vaccination (12/26 versus 1/13 in Experiment II, P < 0.01). These effects were also influenced by the age at which REV infection occurred. Mortality rates after virus challenge in chickens infected with REV at 15 days (Group 2 in Experiment I) or 8 days (Group 2 in Experiment II) of age were not significantly higher than the control (Group 3) after the first, second and third vaccinations.

Table 2.  Inhibitory effects of REV infection on the protective efficacy of vaccination against HPAI H5N1 virus

REV inoculation			Mortality 8 days after challenging at 5 weeks after vaccination
Group	Age (days)	Dose (TCID50)	First vaccination at 2 weeks of age	Second vaccination at 7 weeks of age	Third vaccination at 12 weeks of age
Experiment I
1	1	1000	22/23^A	14/19^A	ND
2	15	1000	5/12^B	2/13^B	ND
3	–	Vaccinated control	7/13^B	1/13^B	ND
4	–	Unvaccinated control	10/10^A	9/10^A	ND
Experiment II
1	1	1000	ND	4/22^A	12/26^A
2	8	1000	ND	0/28^B	3/17^AB
3	–	Vaccinated control	ND	0/14^B	1/13^B
4	–	Unvaccinated control	ND	10/10^C	8/8^C
Chickens in Groups 1 to 3 were vaccinated with inactivated H5N2 at 2 weeks, 7 weeks and 12 weeks of age. Group 4 was an unvaccinated control. In Experiment I, birds were challenged with HPAI virus field strain A/chicken/ZW/06(H5N1) using 10^6 chicken LD50 intramuscularly 5 weeks after the first and second vaccinations. In Experiment II, birds were challenged with A/chicken/ZW/06(H5N1) using 10^5 chicken LD50 5 weeks after the second vaccination and with 10^6 chicken LD50 5 weeks after the third vaccination. Values followed by same uppercase superscript letter in the same column of the same experiment do not differ significantly (P > 0.05) by the chi-squared test. ND, not done.

Discussion

Since authorities throughout the world are concerned about the potential of HPAI H5N1 to mutate and cause a pandemic disease in humans (Li et al., 2004), one of the most important aims might be to minimize H5N1 spreading and its circulation in chickens and other poultry. Although depopulation programmes have been enforced in many parts, H5N1 HPAI outbreaks have been frequently reported since 2003 in some Asian countries, including China, Indonesia, Viet Nam, Thailand and South Korea (Swayne, 2007; Yu, 2007). It appears that the H5N1 HPAI infection become an endemic disease in Asia. Due to the coexistence of both large-scale poultry farms and backyard-raised chickens, ducks and geese, it is very difficult to successfully conduct eradication programmes in these areas. That is why vaccination programmes are also officially enforced or encouraged in some Asian countries in conjunction with depopulation programmes in the HPAI-infected farms in limited areas in campaigns against HPAI. In many farms, inactivated oil–emulsion vaccines of subgroup H5 gave satisfactory results but HPAI outbreaks still occurred in some vaccinated flocks. As Smith et al. (2006) reported, the appearance of antigenic variants might be a major reason responsible for the failures of vaccination. Our study demonstrated that early REV infection strongly suppressed HI antibody titres to the H5 subtype of AIV and significantly decreased protective efficacy of vaccination, implying that virus-induced immunosuppression in chickens might contribute to the vaccination failures. More importantly, such immunosuppressive effects could last even up to 4 months even after two or three repeated vaccinations (Tables 1 and 2). There are many immunosuppressive viruses circulating widely in chickens in addition to REV, and we suggest that subclinical immunosuppressive viral infections should also be considered as an important epidemiological aspect in the control and prevention of HPAI.

Immunosuppression induced by REV is age related (Witter & Fadly, 2003), and vertical transmission (Motha & Egerton, 1987) and REV-contaminated vaccines (Fadly et al., 1996) are the major routes for the early infection of REV. The results in the present study demonstrate that only early REV infection could strongly and persistently decrease HI antibody titres and protective efficacy of vaccination, whereas late REV infection induced only weak and temporary immunosuppression.

Our epidemiological studies indicate that REV infections are very common in chickens in southern China. In serological surveys conducted in 2003/2004, all 23 investigated “yellow chicken” flocks raised in southern China were positive to REV antibodies and the total average positive rate was 34.4% (230/668 birds). However, REV antibody-positive rates are quite variable from flock to flock. Among the seven “yellow chicken” breeder flocks older than 30 weeks investigated, the positive rates ranged from 3/30 to 12/14 (unpublished data), indicating that all flocks have a certain percentage of birds that were sensitive to REV infection and would also produce eggs without maternal antibody to REV. More importantly, vertical transmission rates of REV through fertilized eggs could be as high as 21.7% (REV was isolated from 3/14, 5/17, 3/12 and 4/26 embryos collected from four separate breeder farms with high tumour mortality) in southern China (Sun & Cui, 2007; also additional unpublished data). This vertical transmission may cause more severe immunosuppression than the REV infection at 1 day of age seen in this study. Furthermore, REV was commonly isolated and identified in co-infection with other viruses such as Marek's disease virus, subgroup J avian leucosis virus, and chicken anaemia virus (Cui et al., 2000; Zhang et al., 2003; Sun & Cui, 2007). These viruses are also capable of inducing immunosuppression. Co-infection of REV with these viruses has been shown to enhance immunosuppression and to more severely inhibit antibody reactions to vaccines against H5N1 (Li & Cui, 2007).

In the REV-infected groups, there were not only significantly lower HI antibody mean titres but also a larger variation among individuals, as indicated by the higher coefficients of variation (Table 1). This suggested that there were some individuals with very severely damaged immune function, which may become easy targets for H5 AIV infection in vaccinated flocks and provide exceptional sources for the infection within the flocks. This would also help the viruses to continuously circulate in flocks. More importantly, such vaccinated birds with partial immunity will become carriers for viral evolution under immune selective pressures. It is already very well documented that REV infection can decrease immune reaction to various vaccines (Witter & Fadly, 2003; Sun et al., 2006), and the present study further demonstrates that immunosuppression induced by early REV infection can last for at least 4 months, even in chickens that are vaccinated three times with inactivated H5 subgroup AIV vaccines as is commonly done in chicken farms in China and some other South-east Asia countries in recent years. Considering that some immunosuppressive viral infections and their co-infections are common in these areas as mentioned above (Cui et al., 2000; Zhang et al., 2003; Sun & Cui, 2007), this study suggests that vaccination programmes against HPAI would not be fully effective if other immunosuppressive viral infections such as REV exist in chickens.

Acknowledgements

The authors wish to thank Dr Larry Bacon from Avian Disease and Oncology Laboratory, East Lansing, Michigan, USA for excellent editorial comments and assistance. The study was supported by a grant from National Natural Science Foundation of China to Dr Z. Cui (#30330450).