Antigenic and Genetic Characterization of Influenza C Viruses Which Caused Two Outbreaks in Yamagata City, Japan, in 1996 and 1998

A total of 33 strains of influenza C virus (listed in Table 1) isolated from pediatric patients (<15 years of age) with acute respiratory illness who visited Katsushima Pediatric Clinic (located in Yamagata City) or Yamagata City Hospital Saiseikan during a 3-year period from January 1996 to December 1998 were used. All these viruses were initially isolated from throat swab specimens using the HMV-II line of human malignant melanoma cells or Madin-Darby canine kidney (MDCK) cells as a host according to described procedures (19, 23). For this study, however, viruses were reisolated from throat swabs by inoculating them into the amniotic cavity of 9-day-old embryonated hen's eggs, because isolation and passages of influenza C virus in tissue culture cells sometimes result in selection of antigenic variants (33).

Six older strains (KA176, Miyagi/77 [MI77], MS80, YA2681, AI181, and SP82) as well as 12 strains (Miyagi/1/93 [MI193], Miyagi/2/93 [MI293], Miyagi/3/93 [MI393], Miyagi/4/93 [MI493], Miyagi/6/93 [MI693], Miyagi/7/93 [MI793], Miyagi/1/94 [MI194], Miyagi/2/94 [MI294], Miyagi/3/94 [MI394], Miyagi/4/96 [MI496], Miyagi/7/96 [MI796], and Miyagi/8/96 [MI896]) which had been isolated in Sendai City between 1993 and 1996 and found to possess HE antigenicity identical to that of strain MS80 were also used for comparison. These viruses were also isolated, passaged, and propagated in eggs. HMV-II cells were grown in RPMI 1640 medium containing 10% fetal calf serum, and MDCK cells were cultured in Eagle's minimal essential medium supplemented with 10% fetal calf serum.

The hemagglutination inhibition (HI) test was done in microtiter plates with anti-HE monoclonal antibodies (MAbs) (30, 31) and 0.5% chicken erythrocytes. Briefly, 50 μl of 16 hemagglutinating units (HAU) of virus suspension per ml was added to each well containing 50 μl of twofold-diluted MAbs. After incubation for 30 min at room temperature, 100 μl of 0.5% chicken erythrocytes was added to all wells, and plates were stored for 60 min at 4°C. The HI titer was expressed as the reciprocal of the highest antibody dilution which completely inhibited hemagglutination. Precautions over inhibitors were not necessary because little or no inhibitor activity was detected in mouse ascitic fluid.

Viral RNA was extracted from 200 μl of the virus-containing amniotic fluid with the RNeasy minikit (Qiagen). The viral RNA was then transcribed into cDNA with avian myeloblastosis virus reverse transcriptase XL (Life Science) and an oligonucleotide primer complementary to positions 1 to 12 at the 3" end of all influenza C virus RNA segments. By using the resulting cDNA as a template, the individual RNA segments were amplified by PCR through 35 cycles of the thermocycler program described before (13). The PCR products were purified by rapid gel filtration with a Chroma Spin column (Clontech) and then sequenced by using the BigDye terminator cycle sequencing FS ready reaction kit on an ABI Prism 310 (Applied Biosystems) automatic sequencer.

The nucleotide sequences of the oligonuclotide primers used for PCR amplification and sequencing are available from the authors upon request. Sequence data were analyzed with the PHYLIP program (version 3.54c), and phylogenetic trees were constructed by the neighbor-joining method (29) using the same software.

The nucleotide sequences determined in this study have been submitted to the DDBJ/EMBL/GenBank databases and assigned accession numbers AB064397 to AB064488 .

During a 3-year survey from January 1996 to December 1998, a total of 10,726 throat swab specimens were collected from children with acute respiratory illness and examined for the presence of influenza C virus. The results obtained are shown in Fig. 1. In 1996, 20 strains of influenza C virus were isolated between 23 May and 6 August. It was impressive that 16 of the 20 isolates were obtained during a short period of 24 days (from 3 to 27 June; see Table 1), the isolation rate of June 1996 being as high as 4.9% (16 of 328).

In 1997, influenza C virus was not isolated at all despite the examination of 3,550 throat swab specimens. In 1998, 13 strains were isolated between 27 March and 8 June, 7 of which were obtained in April, with an isolation rate of 2.8% (7 of 250). These observations suggest that during the 3 years (1996 to 1998), two outbreaks of influenza C may have occurred in Yamagata City, the first one in May to August 1996 and the second one in March to June 1998. Clinical details of the 33 children were largely similar to those described in our previous papers (11, 19). Briefly, fever, cough, and nasal discharge were observed in more than half of them, but lower respiratory tract infections such as bronchitis and pneumonia developed occasionally.

The 33 influenza C virus strains isolated in Yamagata City in 1996 and 1998 were examined in HI tests for reactivity with four different anti-HE MAbs characterized previously (30, 31). As shown in Table 1, the reactivity patterns of the 1996 and 1998 isolates were similar; all the viruses were highly reactive with MAbs J14 and Q5 but were unreactive or very weakly reactive with U4 and MS22. Comparison of this reactivity pattern with those of four reference strains (which represent each of the four virus groups shown to have circulated in Japan in the 1980s and the early 1990s [17]) revealed that all 33 isolates obtained from Yamagata City in 1996 and 1998 belonged to the YA2681 virus group.

In order to confirm the results of the antigenic analysis, which suggested that all the 1996 and 1998 isolates from Yamagata City have HE genes belonging to the YA2681-related lineage, the sequence of the HE gene (nucleotides 64 to 1989) was determined for 7 (Yamagata/3/96 [YA396], Yamagata/8/96 [YA896], Yamagata/9/96 [YA996], Yamagata/20/96 [YA2096], Yamagata/2/98 [YA298], Yamagata/6/98 [YA698], and Yamagata/13/98 [YA1398]) of the 33 isolates, and a phylogenetic tree was constructed by using these 7 sequences in addition to the 42 published sequences (1, 3, 4, 13, 15, 16, 17, 20, 21, 25, 26, 27, 28, 33) as well as the 12 sequences of the 1993 to 1996 isolates from Sendai City determined here (see below for details).

The nucleotide sequences of the HE genes of the seven 1996 and 1998 isolates from Yamagata were strikingly similar, with sequence identities of 99.9% among the 1996 strains, 99.7 to 100% among the 1998 strains, and 99.7 to 100% between the 1996 and 1998 strains. As shown in Fig. 2, the HE genes of the influenza C viruses analyzed were divided into six discrete lineages, represented by Taylor/47, AI181, SP82, KA176, YA2681, and MS80. The seven Yamagata strains of 1996 and 1998 were all within the YA2681-related lineage. We reported previously that virus strains which were isolated later than 1981 and had the HE genes of the YA2681 virus lineage can be divided into two distinct subgroups, represented by YA2681 and pig/Beijing/115/81 (PB11581), and that several strains (including Yamagata/10/89 [YA1089], Miyagi/3/91 [MI391], Miyagi/9/91 [MI991], and Miyagi/2/92 [MI292]) having HE genes (as well as other genes) closely related to those of PB11581 were isolated in Yamagata and Sendai cities between 1989 and 1992 (13). It is important to note that the 1996 and 1998 isolates from Yamagata possessed HE genes highly homologous to the genes of these PB11581-like viruses.

To ascertain whether the 1996 and 1998 strains in question have a close relationship with PB11581-like viruses (such as YA1089 and MI991) which circulated in Yamagata and Sendai cities between 1989 and 1992, partial nucleotide sequences of the PB2 (positions 52 to 520), PB1 (positions 50 to 425), P3 (positions 49 to 420), and NP (positions 71 to 670) genes as well as nearly complete sequences of the M (positions 26 to 1,147) and NS (positions 28 to 889) genes were determined for the seven 1996 and 1998 isolates listed above as well as for two old strains (KA176 and MI77), and the phylogenetic trees of individual genes were constructed by using the sequences of these nine strains as well as the previously reported sequences of the 15 strains isolated between 1980 and 1993 (2, 13, 17, 26, 27, 32). For construction of the P3 and NP gene trees, the sequences of the 12 strains isolated in Sendai City between 1993 and 1996 were also included for the reason described below.

As shown in Fig. 3, the PB2 genes analyzed appeared to be split into six lineages, represented by strains MS80, KA176, AI181, SP82, YA2681, and PB11581. The Yamagata isolates of 1996 and 1998, together with YA1089 and MI991, were all located on the PB11581 virus lineage. In the PB1 gene tree, four lineages, represented by AI181, MS80, KA176, and YA2681, were identified, the 1996 and 1998 strains of Yamagata as well as PB11581, YA1089, and MI991 being in the YA2681 virus lineage.

In the tree of the M gene sequences, there were two major branch clusters (previously designated lineages I and II by Tada et al. [32]), one (lineage I) containing KA176, YA2681, and PB11581, and the other (lineage II) containing AI181 and MS80. The 1996 and 1998 isolates in question were within the former branch cluster, which also contained YA1089 and MI991. The NS gene sequences were also split into two discrete lineages (designated A and B by Alamgir et al. [2]), lineage A containing AI181, KA176, and MS80 and lineage B including YA2681 and PB11581. The 1996 and 1998 strains from Yamagata, like YA1089 and MI991, were in the latter lineage. These results indicate that the Yamagata isolates of 1996 and 1998 had PB2, PB1, M, and NS genes (as well HE gene) closely related to those of the PB11581-like viruses that circulated in Yamagata and Sendai cities during 1989 to 1992.

Interestingly, however, the phylogenetic positions of the 1996 and 1998 Yamagata strains were different from those of the PB11581-like viruses in the P3 and NP gene trees. The P3 genes were divided into five different lineages, represented by MS80, KA176, AI181, SP82, and YA2681. The 1996 and 1998 isolates from Yamagata were located on the MS80 virus lineage, separately from PB11581, YA1089, and MI991, which were on the YA2681 virus lineage. The NP genes were split into six distinct lineages, five represented by MS80, YA2681, AI181, KA176, and PB11581 and one (designated the MI193-related lineage) made up by six 1993 and 1994 isolated from Sendai (MI193, MI293, MI693, MI194, MI294, and MI394). The 1996 and 1998 Yamagata strains were located on the MS80 virus lineage, whereas YA1089 and MI991 were on the PB11581 virus lineage. The results shown in Fig. 2 and Fig. 3, taken together, strongly suggest that influenza C virus strains isolated in Yamagata City in 1996 and 1998 are reassortant viruses which acquired HE, PB2, PB1, M, and NS genes from a PB11581-like virus (such as YA1089 and MI991) and P3 and NP genes from an MS80-like virus.

As stated above, the Yamagata isolates of 1996 and 1998 are reassortants which obtained P3 and NP genes from an MS80-like parent. In Japan, influenza C viruses having the same HE antigenicity as MS80 were first isolated in the Kinki district between 1982 and 1983 (Kyoto/41/81, Nara/82, and Hyogo/1/83) (1). In the Tohoku district (which includes Yamagata and Miyagi prefectures), however, influenza C virus antigenically indistinguishable from MS80 was isolated for the first time in 1992 (Yamagata/5/92 [YA592]) (27), which was followed by isolation of 12 strains of the MS80 virus group in Sendai City between 1993 and 1996 (six in 1993, three in 1994, and three in 1996; see Materials and Methods).

To obtain information about the origin of the P3 and NP genes of the 1996 and 1998 isolates from Yamagata, the nucleotide sequences of the P3 and NP genes were determined for the 12 MS80-related isolates from Sendai (sequences of YA592 were reported previously [27]) and then compared with those of the Yamagata isolates in question. Figures 2 and 3 show that although all 13 MS80-like isolates (including YA592) had HE genes belonging to the MS80-related lineage, only three (YA592, MI393, and MI493) possessed the P3 and NP genes on the lineages represented by MS80. Moreover, it was found that the older strains (Kyoto/41/82 and Nara/82) isolated in the Kinki district in 1982, rather than those of the three Yamagata and Sendai strains, had P3 and NP genes more closely related to those of the 1996 and 1998 strains isolated in Yamagata City, raising the possibility that influenza C virus similar to the Kinki strains described above may have supplied the Yamagata strains of 1996 and 1998 with the P3 and NP genes.