Conceived and designed the experiments: SRF MB. Performed the experiments: SRF DK RH AB TH. Analyzed the data: HW. Contributed reagents/materials/analysis tools: SRF ES MB CG RH EL. Wrote the paper: SRF ES TH.
The authors have declared that no competing interests exist.
The potential role of wild birds as carriers of highly pathogenic avian influenza virus (HPAIV) subtype H5N1 is still a matter of debate. Consecutive or simultaneous infections with different subtypes of influenza viruses of low pathogenicity (LPAIV) are very common in wild duck populations. To better understand the epidemiology and pathogenesis of HPAIV H5N1 infections in natural ecosystems, we investigated the influence of prior infection of mallards with homo- (H5N2) and heterosubtypic (H4N6) LPAIV on exposure to HPAIV H5N1. In mallards with homosubtypic immunity induced by LPAIV infection, clinical disease was absent and shedding of HPAIV from respiratory and intestinal tracts was grossly reduced compared to the heterosubtypic and control groups (mean GEC/100 µl at 3 dpi: 3.0×102 vs. 2.3×104 vs. 8.7×104; p<0.05). Heterosubtypic immunity induced by an H4N6 infection mediated a similar but less pronounced effect. We conclude that the epidemiology of HPAIV H5N1 in mallards and probably other aquatic wild bird species is massively influenced by interfering immunity induced by prior homo- and heterosubtypic LPAIV infections.
Migratory birds and members of the
Previous experimental studies with HPAIV H5N1 strains of different origins in various species of water birds including swans and geese
It has been hypothesized that a considerable number of these birds may have at least partially been protected by immunity induced by naturally occurring homosubtypic (HA homologous) infection with avian influenza viruses of low pathogenicity (LPAIV), and that cross reactive interference of even heterosubtypic (HA heterologous) LPAIV-induced immunity might have played a silencing role. LPAIV H5 strains are being continuously isolated from
We tested the effect of LPAIV-induced immunity by experimental inoculation of seronegative (for whole period before inoculation) captive mallards, with two different LPAIV subtypes, H5N2 and H4N6, and subsequent challenge infection with HPAIV H5N1. An H4 subtype virus was chosen because (i) the HA of this subtype is distantly related, by genetic and antigenic means, to that of the H5 subtype and (ii) subtype H4 viruses show a high prevalence in wild duck populations. Mallards represent the most abundant duck species in Eurasia and migrate over long distances, e.g., along the East-Atlantic flyway
The three AIV strains used in this study are maintained in the virus repository of the OIE and National Reference Laboratory for Avian Influenza (NRL AI) at the Friedrich-Loeffler-Institut (FLI). The LPAIV strains A/mallard/Föhr (Germany)/Wv1806-09K/03(H4N6) and A/duck/Potsdam/1402/86(H5N2) were used for pre-exposure inoculation of ducks. The HPAIV strain A/duck/Vietnam/TG24-01/05(H5N1) was used for challenge infection. This clade 1 isolate bears a PQRERRKKR/GLF motif at the HA0 cleavage site, and has an intravenous pathogenicity index (IVPI) of 2.9 in specific pathogen free (SPF) chickens; in addition, it has been found to induce clinically overt and lethal neurological disease in adult Pekin ducks (Harder et al., unpublished).
Thirty–two mallards (
Seven weeks after LPAIV inoculation, all 32 birds (including controls) were housed together in a BSL-3-Ag facility at the FLI. Oropharyngeal and cloacal swabs were collected from each bird to exclude active infection with and shedding of H4 or H5 LPAIV. Subsequently, all birds were challenged with 105 TCID50 of HPAIV H5N1 strain via ocular, nasal and oropharyngeal routes. The birds were then monitored daily for clinical signs of disease. Oropharyngeal and cloacal swabs were collected from all birds at 1, 2, 3, 4, 7, 10, 14 and 21 days post challenge (dpc) and tested by real-time RT-PCR. The experiment was terminated at 24 dpc when serum samples were collected for serological testing and tissue samples including brain, lungs, liver and pancreas were obtained for virological evaluation.
Hypothesis test of differences between groups are carried out by a Mann-Whitney-U-Test in R.
Swab and tissue samples were tested with TaqMan one-step real-time RT-PCR assays targeting the influenza A virus M gene
Log10 dilution series of quantified RNA run-off transcripts or replication-competent H5N1 virus (infectivity titrated on MDCK cell culture) were used.
The hemagglutination inhibition (HI) assay was performed as previously described
The serum samples were tested with a competitive ELISA targeting influenza A nucleoprotein antibodies following the manufacturer's instructions (ID Screen, Influenza A NP Antibody Competition, ID.VET).
The serum samples of all ducks have been tested by serum neutralization test (SNT), to quantify the serologic response, based on a previously described procedure
The titre of HPAIV in the swab samples was extrapolated from Ct-values on basis of calibration experiments using different log10 dilution series of A/duck/Vietnam/TG24-01/05 (H5N1) virus. Infectivity is expressed as TCID50 per 100 µl of the swab sample fluids (
Tissues samples including trachea, lungs, heart, cerebrum, cerebellum, spinal cord, proventriculus, gizzard, small and large intestine, liver, pancreas and kidney of two birds from the control group, which died at 5 and 6 dpc, were collected, formalin fixed and processed for paraffin embedding according to standard procedures, and immunohistochemistry for influenza virus A nucleoprotein (NP) was performed. Briefly, after dewaxing sections were microwave irradiated for antigen retrieval (2×5 min, 600 W, 10 mM citrate buffer pH 6.0) and were incubated with a rabbit anti-NP serum (1∶750). A biotinylated goat anti-rabbit IgG1 (Vector, Burlingame, CA, USA) was applied (1∶200) as secondary antibody. By means of the avidin-biotin-peroxidase complex method, a bright red intracytoplasmatic and nuclear signal was observed. Positive control tissues of chickens experimentally infected with HPAI virus (H5N1) and additionally, a control primary rabbit serum against bovine papillomavirus (BPV 1∶2000) were included.
The cloacal and oropharyngeal swab samples collected from 32 ducks during 8 weeks prior to LPAIV inoculation revealed negative results by real-time RT-PCR, indicating that the ducks were not shedding AIV before experimental infection. In addition, ducks were serologically negative to influenza A antigens tested by ELISA and HI tests (with using H4N6, H5N2 and H5N1 antigens), indicating that the birds were not exposed to AIV before inoculation.
All birds remained clinically healthy during seven weeks after inoculation of H4 and H5 LPAIV. The results of serological evaluation of ducks by ELISA and HI tests using the homologous antigens at 1, 2, 4 and 7 weeks after inoculation are summarized in
Group | ELISA |
H4N6 HI |
H5N2-HI | H5N1-HI | SN-test |
||||||||||||||
B.I. |
1 |
2 | 4 | 7 | Post C. | B.I. | 1 | 2 | 4 | 7 | B.I. | 1 | 2 | 4 | 7 | Post C. |
Before C. | Post C. | |
H5 | 0/12 | 12/12 | 12/12 | 7/12 | 6/12 | 12/12 | <2 | <2 |
<2 | <2 | <2 | <2 | 5.1±0.8 | 5.8±0.8 | 5.7±0.5 | 4±0.4 | 6.7±1.1 | 5.4 | 9.5 |
H4 | 0/12 | 12/12 | 12/12 | 9/12 | 3/12 | 12/12 | <2 | 2.6±2.2 | 4.3±1.4 | 4±0.9 | 2.3±1 | <2 | <2 | <2 | <2 | <2 | 7.2±0.9 | 2 | 8.4 |
Control | 0/8 | 0/8 | 0/8 | 0/8 | 0/8 | 7/7 | <2 | <2 | <2 | <2 | <2 | <2 | <2 | <2 | <2 | <2 | 8.4±0.9 | <2 | 8.7 |
The numbers indicate weeks after LPAIV inoculation.
The ELISA results indicated as the number of positives out of the number of tested birds.
Hemagglutination inhibition results indicate the geometric mean titre of (log2) serum samples and mean log2±standard deviation.
Carried out against AI virus A/cygnus cygnus/Germany/R65/06 (H5N1) and expressed as geometric mean titre of (log2) serum samples.
Before LPAIV H4N6 or H5N2 inoculation.
Post C. = 24 days post challenge. Before C. = the day of challenge, before inoculation of HPAIV.
None of the ducks with mean titre of <2, showed reactivity higher than 1 log2.
Clinical symptoms varied significantly among members of the three groups. From day two after inoculation onwards, up to seven ducks in the control group became severely sick, but only one of the control birds died (6 dpc) while others recovered slowly. One more duck died at 5 dpc. Unfortunately due to loss of the wing tag of this bird and also another bird from the H4 group at the same day, it could not be unambiguously assigned to either H4 or control groups (see also footnote 6 in
Group | Cloacal swabs (dpc) | Tracheal swabs (dpc) | |||||||||||||||
1 |
2 | 3 | 4 | 7 | 10 | 14 | 21 | 1 | 2 | 3 | 4 | 7 | 10 | 14 | 21 | ||
H5 | Pos. No. |
0/12 | 0/12 | 0/12 | 0/12 | 0/12 | 0/12 | 0/12 | 0/12 | 3/12 | 3/12 | 2/12 | 1/12 | 0/12 | 0/12 | 0/12 | 0/12 |
Average Ct |
neg | neg | neg | neg | neg | neg | neg | Neg | 34.8 | 34.07 | 34.25 | 34.24 | neg | neg | neg | neg | |
Ct Range |
- | - | - | - | - | - | - | - | 34.2–35.1 | 33.5–34.8 | 33.3–35.3 | - | - | - | - | - | |
Average GEC |
<1 | <1 | <1 | <1 | <1 | <1 | <1 | <1 | 1.1×103 | 1.7×103 | 1.8×103 | 1.5×103 | <1 | <1 | <1 | <1 | |
GEC range |
- | - | - | - | - | - | - | - | 9.0×102–1.6×103 | 1.1×103–2.4×103 | 8.0×102–1.8×103 | - | - | - | - | - | |
Average TCID50 | <1 | <1 | <1 | <1 | <1 | <1 | <1 | <1 | 1.0×101 | 1.3×101 | 1.7×101 | 1.5×103 | <1 | <1 | <1 | <1 | |
TCID50 Range | - | - | - | - | - | - | - | - | 7.0–1.3×101 | 7.0–2.0×101 | 7.0–2.6×101 | - | - | - | - | - | |
H4 | Pos. No. | 0/12 | 0/12 | 0/12 | 0/12 | 0/11 |
0/11 | 0/11 | 0/11 | 12/12 | 12/12 | 12/12 | 10/12 | 4/11 | 1/11 | 0/11 | 0/11 |
Average Ct | neg | neg | neg | neg | Neg | neg | neg | Neg | 29.84 | 29.8 | 30.67 | 32.39 | 33.99 | 34.16 | neg | neg | |
Ct Range | - | - | - | - | - | - | - | - | 24.2–34.5 | 24.1–33 | 27–33.2 | 29.8–35.1 | 33.2–35.1 | - | - | - | |
Average GEC | <1 | <1 | <1 | <1 | <1 | <1 | <1 | <1 | 7.5×104 | 7.1×104 | 2.3×104 | 8.6×103 | 1.9×103 | 1.6×103 | <1 | <1 | |
GEC Range | - | - | - | - | - | - | - | - | 1.3×103–5.8×105 | 3.1×103–6.1×105 | 2.9×103–1.1×105 | 8.9×102–2.1×104 | 8.9×102–2.8×103 | - | - | - | |
Average TCID50 | <1 | <1 | <1 | <1 | <1 | <1 | <1 | <1 | 1.4×103 | 1.3×103 | 3.1×102 | 7.5×101 | 1.7×101 | 1.3×101 | <1 | <1 | |
TCID50 Range | - | - | - | - | - | - | - | - | 1.3×101–1.2×104 | 2.6×101–1.3×104 | 2.6×101–1.8×103 | 7.0×100–2.4×102 | 7.0×100–2.6×101 | - | - | - | |
Control | Pos. No. | 4/8 | 4/8 | 6/8 | 4/8 | 1/6 |
0/6 | 0/6 | 0/6 | 8/8 | 8/8 | 8/8 | 8/8 | 6/6 | 4/6 | 2/6 | 1/6 |
Average Ct | 33.91 | 32.71 | 32.15 | 32.17 | 35.2 | neg | neg | Neg | 28.12 | 27.54 | 27.83 | 29.44 | 33.26 | 33.9 | 34.45 | 34.3 | |
Ct Range | 32.7–34.8 | 30.33–34.37 | 26.9–34.6 | 28.4–35.2 | - | - | - | - | 22.7–33.6 | 24.6–31.6 | 26.3–30.1 | 27.0–31.4 | 29.3–35.6 | 31.8–34.4 | 34.0–34.9 | - | |
Average GEC | 2.1×103 | 5.9×103 | 2.2×104 | 1.6×104 | 8.6×102 | <1 | <1 | <1 | 2.5×105 | 1.4×105 | 8.7×104 | 3.6×104 | 5.8×103 | 2.9×103 | 1.4×103 | 1.5×103 | |
GEC range | 1.1×103–3.9×103 | 1.8×103–1.5×104 | 1.2×103–1.1×105 | 8.2×102–4.9×104 | - | - | - | - | 2.1×103–1.4×106 | 7.3×103–4.6×105 | 1.8×104–1.7×105 | 8.2×103–1.1×105 | 6.6×102–2.8×104 | 7.3×102–6.6×103 | 1.0×103–1.7×103 | - | |
Average TCID50 | 1.8×101 | 6.1×101 | 3.3×102 | 2.1×102 | 7.0×100 | <1 | <1 | <1 | 5.7×103 | 2.5×103 | 1.4×103 | 5.0×102 | 6.8×101 | 2.9×101 | 1.0×101 | 1.1×101 | |
TCID50 range | 7.0×100–3.3×101 | 1.3×101–1.7×102 | 2.0×101–1.8×103 | 1.0×101–6.7×102 | - | - | - | - | 2.1×101–3.4×104 | 7.3×101–9.4×103 | 2.1×102–2.9×103 | 7.9×101–1.7×103 | 1.3×101–3.5×102 | 1.3×101–6.6×101 | 7.0×100–1.3×101 | - |
The numbers indicate days post-challenge (dpc).
Number of positive birds as indicated by real-time RT-PCT. Ct-values<36 considered as positive.
Average ct-value of positive ducks per 15 µl of swab fluid.
The max and min range of positive ct-values.
Average genome equivalent copy (GEC) number of positive samples/100 µl of swab fluid.
Average tissue culture infectious dose (TCID50) titre/100 µl of swab fluid.
At this stage one of the control birds died.
Two ducks lost the wing tags at 5 dpc, and one of them died at 5 dpc, since their assignment to either H4 or control groups was not possible, both were excluded from calculations after day 5.
The results of the real-time RT-PCR testing of cloacal and oropharyngeal swab samples taken on days 1, 2, 3, 4, 7, 10, 14, 21 after challenge are summarized in
Average of virus excretion and 95% confidence intervals are depicted.
Cloacal shedding was not observed in ducks from groups with previous LPAIV infection. Clear differences were also seen regarding the oropharyngeal shedding of the H4 and H5 groups, especially on day 3 and 4 significant differences in tracheal shedding is observed between all three groups (
Tissue samples comprising brain, lung, liver and pancreas from the one control duck which died at 6 dpc were highly positive in real-time RT-PCR (2.0×107, 5.1×104, 4.4×104 and 1.1×106 GEC/100 µg, respectively). No viral RNA/infectivity was detected in the same tissues from any of the ducks surviving until 24 dpc.
Surviving ducks in all three groups developed high levels of HPAIV H5-specific antibodies post-challenge according to NP-ELISA, H5-specific HI and serum neutralization tests (
The control duck, which died at 6 dpc showed moderate congestion of the liver and edema of the brain. In histopathology, the cerebrum was severely congested, multifocally there was neuropil degeneration with mild vacuolation (
(A) Brain, Cerebrum; Duck at 5 dpc. Congestion shown by hematoxylin-eosin staining. Bar 100 µm. (B) Brain, Cerebrum; Duck at 6 dpc. Intense intranuclear and intracytoplasmic AIV antigen staining within neurons and neuroglia. Immunohistochemistry. ABC method using anti-NP monoclonal antibody HB65, hematoxylin counterstain. Bar 100 µm.
Here we show that pre-existing immunity induced by infection with homo- or heterosubtypic LPAIV modifies the course of an experimental challenge infection with HPAIV H5N1 in mallards. Clinical signs as well as amplitude and tissue tropism of virus shedding was affected.
Seven (out of eight) control ducks became severely sick. In contrast, only three ducks (out of 12) with previous H4N6 infection showed mild clinical symptoms but recovered fast, and no clinical symptoms were obvious in ducks with previous H5N2 LPAIV infection. Viral shedding from the respiratory tract was most pronounced in control ducks. Preferential viral shedding via the oropharynx has been consistently demonstrated with HPAIV H5N1 viruses
Clinical symptoms in ducks of the control group seemed to be more severe that has previously been reported for experimental inoculation of naive mallards with HPAIV H5N1
Long-distance migration is one of the most demanding physiologic activities in the animal world
In summary, the results of our study show that, in captive mallards, heterosubtypic cross reactive immunity can derogate clinical symptoms of an HPAIV H5N1 infection, reduce the amount and duration of viral shedding from the respiratory tract and prevent viral shedding from the intestinal tract. Homosubtypic immunity may fully abrogate clinical symptoms and viral shedding from the intestinal tract, and drastically reduce viral shedding from the respiratory tract. Therefore, mallards with prior exposure to homologous LPAI viruses may remain healthy and might be suitable for long-distance transposition of HPAIV, but probably only shed very low titers of virus. Mallards with prior exposure to heterosubtypic LPAI viruses might pose a greater risk for transmission and spread of HPAIV, because they can shed higher amounts of virus (but only via the respiratory route) without developing severe clinical disease. Still, the potential role of respiratory shedding compared to intestinal shedding in the efficacy of bird-to-bird transmission of HPAIV in the nature needs to be clarified.
We like to thank Bianca Kinnemann, Kathrin Steffen and Diana Wessler for technical assistance.