Dead and dying storks were collected from the area around the landing site 2 days after arrival of the flock on August 26. They were immediately placed in cold storage until transported to the laboratory on September 4, September 9, and October 8, when the birds were thawed and weighed, and necropsies were performed. A total of 13 frozen storks were received.

A small breeding colony of approximately 20 gulls was housed in a closed pen with wire-mesh walls at the Department of Zoology, University of Tel Aviv. In November 1999, several birds were found paralyzed, and two had died.

Brains were removed aseptically from the storks and gulls and homogenized by grinding in borate-buffered saline (pH=7.6). The homogenate was centrifuged at 1,500 rpm for 10 min. Other tissues were not examined. The supernatant was removed and used for intracranial inoculation of suckling mice and infection of Vero cells and yolk sac inoculation of 7-day-old embryonated eggs. Embryonic mortality occurred 3 to 4 days after inoculation and the embryos were colored cherry red but without hemagglutinating activity for chicken erythrocytes in the allantoic fluids. An arbovirus was therefore suspected. Likewise, morbidity of the mice was accompanied by spastic paralysis, a further sign of virally induced encephalitis. In Vero cell culture, a cytopathic effect (CPE) was seen within 3 to 5 days of inoculation.

Storks were aged either as fledglings (<1 year old) or mature birds by their wing feathering and intensity of the yellow pigment of the beaks and legs (E. Gorni, pers. comm..). Blood samples were taken from the Eilat flock 6 days after arrival. Other storks from this flock were placed in a nature reserve to recuperate, and blood samples were taken in September and October. In October 1998 and September 1999, storks were caught as they were migrating over Israel, and blood samples were taken. The January 2000 flock consisted of late migrators that were caught while roosting. The blood samples were taken from the White-eyed Gulls 1 week after isolation of WNV.

Sera were diluted twofold commencing at 1:10 in WNV suspension containing 100 50% tissue culture infective dose (TCID₅₀)/0.1 mL in microwell plates. The virus-serum mixture was allowed to stand at room temperature for 1 hr, and a suspension of 10⁵ Vero cells in 0.1 mL was added to each well. The plates were incubated at 37^°C in 5% CO₂/95% air for 4 days. The neutralizing titer of the serum was calculated as the highest dilution with 50% CPE.

When CPE of the infected Vero cell culture was observed, a glass coverslip was placed in the dish receiving the succeeding passage. Three days later the coverslip was removed and the cells fixed with cold acetone. Identification of the cytopathic agent was performed by indirect immunofluorescence. Panels of monoclonal antibodies to Alphavirus, and Flaviviruses (Table 1) were reacted with the Vero cells and then incubated with flourescein isothiocyanate (FITC)-labeled anti-mouse immunoglobulins. The Flavivirus monoclonal antibodies included 2B4, 6B8, and 6E12 from South Africa; 5F10, 1C9, and 1B4 from Israel, F7/101 from Oxford, U.K.; and 3H6 and 813 from Townsville, Australia. The Alphavirus (Sindbis virus [SINV]) antibodies were 30.11 and 30.12 from Oxford and 2F2 from Australia. Counter-staining was performed with Evans blue.

Brain extracts were performed as described above and the supernatant used to extract RNA with the QIAamp Viral RNA kit (Qiagen, Valencia, CA) and the RNAs resuspended in 60 μL of Viral Lysis Buffer (Qiagen, Valencia, CA) elution buffer according to the manufacturer’s protocol. Reverse transcription-polmenase chain reaction (RT-PCR) was performed on a gene fragment of the envelope protein using the primer pair WN132 (5′ GAAAACATCAAGTATGAGG 3′) and WN240 (5′ GAGGTTCTTCAAACTCCAT 3′) (genome positions 1,402 and 1,656), resulting in the synthesis of a 255-bp product (20). The resulting DNA fragment was visualized on 1.5% agarose gel stained with ethidium bromide.

WNV RNA was extracted from the second passage of Vero cell-infected culture fluid after isolation in a suckling mouse brain, using the QIAamp Viral RNA kit (Qiagen). Viral RNA was extracted from supernatant fluid according to the manufacturer’s protocol and the RNA resuspended in a final volume of 100 μLof RNase-free water. Six overlapping double-stranded cDNA templates were generated by using WNV-specific primer pairs provided by the Centers for Disease Control and Prevention (9). Additional primers with sequences designed from the sequence of WN-NY99 were used to amplify about 100 nucleotides from the far 3′ and 5′ extremities (Table 2). The envelope protein genes from WN viruses isolated in 1998 and 1999 from geese and in 1999 from a White-eyed Gull were amplified using primer pairs designed in the corresponding strands.

Amplified cDNAs were purified by ion-exchange chromatography and precipitated with 2v of isopropanol. Both strands of the purified cDNAs were sequenced by using the Taq Dye Deoxy Terminator Cycle sequencing kit (Perkin Elmer Corp./Applied Biosystem, Norwalk, CN) using primers spaced about 400 bases apart on the genome (9). Cycle sequencing was performed by combining about 0.2 pmols of purified cDNA and 30 pmols of primer and following the manufacturer’s protocol. Sequences were aligned with the Clustal W alignment (21).

A total of four Flaviviruses were isolated from storks in Vero cells—one from a pool of three brains from the first group of dead storks collected on August 28, 1998; one isolate from one of six individual brains; and two isolates from four brains also collected on August 28. The three pooled stork brains from the first group were also injected into three litters of baby mice, causing paralysis. As noted previously, all the storks had been collected 2 days after landing and were frozen immediately. The four isolates were confirmed as WNV by indirect immunofluorescent antibody (IFA) using monoclonal antibodies and by RT-PCR. No Alphavirus reactivity was detected in any of the brains by IFA (Table 3). One WNV was isolated from the brain of one of the White-eyed Gulls found dead in November 1999 in Tel Aviv (data not shown).

All three homogenates of the mouse brains inoculated with the brains of the first group of storks were RT-PCR positive. Two of the six brains of the second group and all four of the third group were RT-PCR positive. Thus, a total of one pool of 3 brains and 6 individual brains from 10 storks tested were RT-PCR positive (Table 3).

The full-length RNA genome of the WNV isolate from one of the stork-1998 (IS98-ST1) has been sequenced. With the exception of about 30 nucleotides at each end of the genome used as primers for cDNA amplification (Table 2), the complete nucleotide sequence of the viral RNA has been determined (data not shown). The WNV genome arrangement is similar to those published for WNV-Nigeria, WNV-NY99, and HNY1999 (9,22,23). The genome is 11,029 nucleotides in length and contains one long open reading frame of 10,302 nt starting at nt 97. Nucleotide sequence comparison of IS98-ST1 with WNV-NY99 (9) showed 28 mutations in the coding region (99.75% similarity). Of the mutations, five were transversions, and the rest occurred at the third codon. Ten amino acid changes were found. The change at position 51 (A51 to V) in WNV-NY99 E protein was not found in the HNY1999 sequence. Other mutations were observed in NS1 (N17 to S), NS2A (R165 to G), NS2B (G82 to D and E83 to G), NS3 (P496 to L and E521 to D), and NS5 (S54 to P, N280 to K, and A372 to V).

The envelope gene sequence of IS98-ST1 (GenBank accession No. AF 481864) was compared to those from goose98 (GenBank accession No. AY033388) and goose99 (GenBank accession No. AY033391) and gull99 (GenBank accession No. AY033390) and included HNY 99 as the consensus sequence (Figure). The comparison showed 3 nt changes, of which one was unique for IS98-ST1 strain (position 1,179), one common to goose99 and gull99 (position 729) and one shared by gull99 and IS98-ST1 (position 3). The 501 amino acids of the E genes of the four strains (IS98-ST1, goose98, goose99, and gull99) were identical (data not shown).

Table 4 presents the WNV neutralization titers of groups of storks according to the month and year of the blood samples, the storks’ age, and geographic location. In 1998 sera were collected from the Eilat flock on September 2, and from eight older storks in a wildlife sanctuary during September and October. Late migrating storks were also captured in northern Israel during their autumn migration in October, and 10 birds from various bird sanctuaries were also sampled. Between January and May 1999, all five storks resident in northern Israel had WNV antibodies. With the fall migration, blood samples were taken from a group of six storks, consisting of four fledglings and two older birds. One stork in each age group had WNV antibodies.

Storks examined through June 30, 2000, consisted of two groups, one was a flock of overwintering adults that had migrated in September 1999 and the others were four fledglings hatched in April 2000 from parents that had overwintered and bred on the Golan Heights. Antibodies were detected in 9 of the 12 adults but in none of the young birds.

Thus, a total of 65 stork sera were examined of which 19 were from birds <1 year old. In the 19 birds <1 year, neutralizing antibodies were found in 4 of them (21.1%), whereas in the older group of 46 storks, 33 (71.7%) were seropositive.

Of the 11 White-eyed Gull sera examined 1 week after the isolation of WNV, 8 were seropositive, 6 had titers > 1:1280, one of 1:640, and one of 1:80, and three were < 1:10.