The virus for inoculation was originally isolated from the brain of a scarlet macaw (Ara macao) that died from PDD; the inoculation was passaged 6 times in the quail cell line CEC-32 (24). The persistently ABV-infected CEC-32 cells were suspended in medium with 2% fetal bovine serum, sonicated, clarified by centrifugation at 3,000 × g for 10 min, and assayed to determine infectivity. For inoculations, we used an infectivity titer of 10⁴ 50% infectious dose/mL.

By using the RNeasy Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions, we isolated total RNA for sequencing from 200 μL of virus-containing supernatant. Total RNA was reverse transcribed by using random hexamer primers. PCR for parts of the large viral polymerase and the nucleocapsid protein genes of avian bornavirus was performed as described (11). We analyzed PCR products by using gel electrophoresis and purified the products for sequencing by using the QIAquick PCR Purification Kit (QIAGEN). Sequencing was carried out by LGC Genomics (Berlin), and BlastN (http://blast.ncbi.nlm.nih.gov/Blast.cgi) was used to align generated sequence information.

For inoculation purposes, we divided the cockatiels into 2 groups (groups ic and iv), each consisting of 9 animals. We intracerebrally (IC) inoculated birds in group ic and intravenously (IV) inoculated birds in group iv with 0.1 mL of the inoculum described above. All birds were under isoflurane anesthesia when inoculated. One bird remained untreated and served as a sentinel bird in group ic. Another group of 9 birds from the same flock served as controls; they remained untreated and were kept separate from the inoculated birds during the investigation period.

Over a period of 230 days, we surveyed the birds daily to determine their health status. We obtained swab samples from the crop and cloaca to test for the presence of ABV RNA by using real-time reverse transcription PCR as described (10). Cycle thresholds (C_t) >36.0 were considered negative (10). Swab samples were obtained every other day until all birds of the respective group had ABV-positive test results, then specimens were obtained weekly. In parallel, once a week we obtained 0.3-mL blood samples for indirect immunofluorescence assay (IIFA) detection of antibodies against ABV, as described (25). For humane reasons, we euthanized birds with clinical signs typical of PDD (emaciation, undigested seed in the feces, neurologic signs) and reduced general condition. All remaining birds, including control birds, were euthanized 115 or 116 dpi or at the end of the trial (229 or 230 dpi).

We obtained samples of brain, eye, spinal cord, ischiadic nerve, adrenal gland, heart, liver, kidney, spleen, pancreas, crop, proventriculus, gizzard, intestine, pectoral muscle, and skin with feathers from all birds that died or were euthanized. For histopathologic analysis immunohistochemical testing, and other immunohistologic procedures, samples were fixed in 5% buffered formalin, embedded in paraffin wax, and used for preparation of 5-μm sections stained with hematoxylin and eosin. Samples from similar organs were frozen fresh for subsequent real-time PCR to detect ABV RNA. In addition, infectious virus was isolated by using samples from brain and retina as described (25). To exclude the presence of any other infection, we examined samples of blood, liver, and lung for bacteria; we used mycological staining to examine samples of proventriculus for infection with yeast; and we examined samples from all intestinal parts for parasites.

Antibodies against ABV were detected by use of an IIFA on persistently Borna disease virus (BDV)–infected Madin-Darby canine kidney cells. For the assay, we used a 1:50 dilution of fluorescein isothiocyanate–conjugated goat anti-bird IgG (Bethyl Laboratories, Inc., Montgomery, TX, USA), as described (25).

We performed immunohistochemical testing according to the avidin-biotin complex method, as described (25,26). This method uses a polyclonal rabbit antibody directed against the phosphoprotein and the X protein of BDV.

For virus isolation, we performed infectivity assays as described by Narayan et al. (27) and Herzog et al. (25). Organ samples were homogenized, and 10-fold dilutions were prepared in GIBCO Glasgow Minimum Essential Medium BHK-21 1× (Invitrogen, Paisley, UK) with 10% fetal bovine serum. The mixture was then mixed with equal volumes of freshly dispersed cells of the quail cell line CEC-32 and then incubated on Lab-Tek Chamber Slides (Nunc, Roskilde, Denmark) for 6 days at 37°C. Virus replication was demonstrated by indirect immunofluorescence by using polyclonal serum specimens, which cross-reacted reliably with ABV antigen, from rats experimentally infected with BDV (25).

We used the Wilcoxon-Mann-Whitney-test to ascertain differences between the IC- and IV-inoculated groups. U <11 (critical value with α = 0.005) was considered highly significant for calculated test statistics.

We used BlastN (http://blast.ncbi.nlm.nih.gov/Blast.cgi) to compare newly generated sequences of parts of the nucleocapsid protein and large viral polymerase genes with sequences available from GenBank. Both genes showed highest accordance (99%) with ABV genotype 4 (EU781959.1 and FJ002323.1).

During the 230-day investigation period, 5 of the 18 inoculated birds (3 from group iv and 2 from group ic) showed clinical signs typical of PDD: birds iv1 and iv3 had gastrointestinal signs, birds ic1 and iv5 showed neurologic signs, and bird ic2 had gastrointestinal and neurologic signs. In bird ic2, the following signs developed 33, 37, and 41 dpi, respectively: general symptoms, e.g., apathy; undigested seeds in the feces; and epileptic-like seizures. In birds iv3 and iv2, gastrointestinal signs first became obvious 116 and 126 dpi, respectively. In birds iv5 and ic1, which were only affected neurologically, signs were first noticed 159 and 199 dpi, respectively.

In addition, birds ic6 and iv2 died suddenly and unexpectedly, without obvious signs, 66 and 120 dpi, respectively. Three apparently healthy birds from each group (birds iv3, iv4, iv7, ic3, ic4, and ic8) were euthanized 115 or 116 dpi. At the end of the investigation period, 229 or 230 dpi, the remaining birds (iv1, iv5, iv8, iv9, ic5, ic7, and ic9) and the sentinel bird (se1), which all appeared clinically healthy, were euthanized (Table 1). The control birds were in a good health status during the entire investigation period and were euthanized with the other birds after 230 days.

Necropsy revealed a dilated proventriculus in 7 of the 18 inoculated birds (ic1, ic2, ic3, iv1, iv2, iv3, and iv4), 4 of which had shown signs typical of PDD (Table 2). Dilatation of the proventriculus was most severe in bird ic2, which had both gastrointestinal and neurologic signs. No macroscopic alterations were detected in the remaining 11 inoculated birds, the sentinel bird, or the 9 control birds.

Histopathologic examination revealed that all infected birds and the sentinel bird had mononuclear infiltrates characteristic of PDD in a wide range of organs and of considerable severity. Most of the immune cell infiltrates were in the central nervous system and the gastrointestinal tract, but they were also present in heart; liver; kidney; pancreas; skin with feathers; and, in 1 case, the spleen. In some of the animals, the infiltrates had a follicle-like appearance. Infiltrates were not detected in the pectoral muscle of any birds. No other infections were detected during bacteriologic, mycologic, and parasitologic examinations. The control group did not have histopathologic alterations in organs and did not have any other infection.

We detected ABV RNA in swab specimens from all inoculated birds and the sentinel bird. We obtained the first positive test results for group ic 19–29 dpi, whereas we first amplified ABV RNA from group iv samples 25–72 dpi (Figure 1). By using the Wilcoxon-Mann-Whitney test, we determined that ABV RNA was detected significantly earlier in group ic than in group iv (calculated value for test statistic U = 5.5, which is <11 [critical value with n₁ = 9 and n₂ = 9] and therefore considered to be highly significant, with α = 0.005). We obtained the first positive ABV RNA test results for the sentinel bird 76 dpi (Figure 1). The C_t for all birds constantly decreased during the trial, from 35.6 at first ABV RNA detection to 19.8 at last sampling (geometric mean values).

Antibodies against ABV were first detected in group ic 7–63 dpi, and group iv birds seroconverted 29–57 dpi (Figure 2). Statistical analysis did not show any noticeable difference between the ic and iv group (calculated value for test statistic U = 39.5, which is >21 [critical value with n₁ = 9 and n₂ = 9] and thus considered not significant, with α = 0.05). Antibody titers in groups ic and iv steadily increased to <20,480 during the investigation period (Figure 3). We did not detect ABV RNA or antibodies against ABV in control group birds during the investigation period.

We detected ABV RNA in all organs that were examined 115 or 116 dpi and 229 or 230 dpi. A high amount of ABV RNA (geometric mean C_t 12.0–17.0) was found in the central nervous system, gastrointestinal tract, and in skin with feathers. Moderate amounts of ABV RNA (geometric mean C_t 17.0–23.0) were found in heart, kidney, spleen, and pancreas. The lowest amounts of ABV RNA (geometric mean C_t 23.0–29.0) were found in pectoral muscle and liver. In the sentinel bird, only skin with feathers had positive test results (C_t 28.77), but brain, crop, gizzard, and intestine had high C_t (>36.0); all other organs remained negative for ABV RNA (Figure 4). We did not detect ABV RNA in the organs of control birds.

The detection of viral phosphoprotein and X-protein was mostly restricted to the central nervous system and the heart. In some cases, ABV antigen was also found in some parts of the gastrointestinal tract. We detected X-protein in all group ic birds but not in the sentinel bird. We could not detect phosphoprotein p24 in bird iv3 or in the sentinel bird. ABV antigen was not found in any of the organs of the control birds.

We isolated infectious ABV from all inoculated birds. No infectious virus was detected from the organs of the sentinel bird or birds in the control group.