Transboundary and Emerging Diseases
ORIGINAL ARTICLE
Bovine Arboviruses in Culicoides Biting Midges and Sentinel Cattle in Southern Japan from 2003 to 2013 T. Kato1, H. Shirafuji1, S. Tanaka1, M. Sato2, M. Yamakawa2, T. Tsuda2 and T. Yanase1 1 2
Kyushu Research Station, National Institute of Animal Health, NARO, Kagoshima, Japan National Institute of Animal Health, NARO, Tsukuba, Japan
Keywords: arbovirus; sentinel cattle; Culicoides biting midge; orthobunyavirus; orbivirus; rhabdovirus Correspondence: T. Yanase. Kyushu Research Station, National Institute of Animal Health, NARO, 2702 Chuzan, Kagoshima 891-0105, Japan. Tel.: +81 99 268 2078; Fax: +81 99 268 3088; E-mail: [email protected] Received for publication August 20, 2014 doi:10.1111/tbed.12324
Summary Epizootic congenital abnormalities, encephalomyelitis and febrile illnesses in cattle caused by arthropod-borne viruses (arboviruses) are prevalent in Japan. Causative viruses including orthobunyaviruses, orbiviruses and rhabdovirus are thought to be transmitted by Culicoides biting midges. Recently, the incursions of several arboviruses, potentially Culicoides-borne, were newly confirmed in Japan. However, their spread pattern and exact vector species are currently uncertain. Attempts to isolate arboviruses from Culicoides biting midges and sentinel cattle were conducted in Kagoshima, located at the southernmost end of the main islands of Japan, a potentially high-risk area for incursion of arboviral diseases and outbreak of endemic ones. Seventy-eight isolates comprising Akabane, Peaton and Sathuperi viruses of the genus Orthobunyavirus of the family Bunyaviridae, bluetongue virus serotype 16, D’Aguilar virus, Bunyip Creek virus and epizootic haemorrhagic disease virus serotype 1 of the genus Orbivirus of the family Reoviridae, a potentially novel rhabdovirus of the genus Ephemerovirus and unidentified orbivirus-like viruses were obtained from Culicoides biting midges and sentinel cattle between 2003 and 2013. Akabane, Sathuperi, D’Aguilar and Bunyip Creek viruses were selectively isolated from Culicoides oxystoma, suggesting this vector’s responsibility for these arbovirus outbreaks. The results of virus isolation also implied that C. tainanus, C. jacobsoni and C. punctatus are competent for the transmission of bluetongue virus serotype 16, Peaton virus and epizootic haemorrhagic disease virus serotype 1, respectively. Our monitoring in Culicoides biting midges and sentinel cattle detected the circulation of Akabane virus just prior to the accumulations of bovine congenital abnormalities and encephalomyelitis by it around study sites in 2003, 2006, 2008 and 2013. Silent circulations of the other arboviruses, including potentially new viruses, were also detected during the study period.
Introduction Arthropod-borne viruses (Arboviruses) are transmitted by haematophagous arthropod vectors, such as mosquitoes, ticks and Culicoides biting midges. Over 500 arboviruses have been described in the world to date, with some being etiological agents of human and animal diseases (Hart, 2001). Substantial trade restrictions in the livestock industry have occasionally resulted from the © 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
presence of certain arboviruses, such as bluetongue virus (BTV) (OIE, 2014). In Japan, arbovirus infections in cattle are frequently reported and have caused serious damage to the beef and dairy industries. Large outbreaks of abnormal deliveries in cattle, such as abortion, stillbirth, premature birth and congenital malformations, have been periodically caused by arboviruses, such as Akabane and Aino viruses (AKAV and AINOV) of the genus Orthobunyavirus of the family 1
Bovine Arboviruses in Southern Japan
Bunyaviridae, and Chuzan virus (CHUV) of the genus Orbivirus of the family Reoviridae (Forman et al., 2008). As recently reported, AKAV was also associated with bovine epizootic encephalomyelitis (Kono et al., 2008). Seasonal epidemics of acute febrile illness in cattle caused by bovine ephemeral fever virus (BEFV) of the genus Ephemerovirus of the family Rhabdoviridae have been sporadically observed (Kato et al., 2009). Before the 1980s, Ibaraki virus (IBAV), a strain of epizootic hemorrhagic disease virus (EHDV) serotype 2, was involved in large outbreaks of diseases in cattle characterized by fever and deglutitive disorder (Forman et al., 2008). Recently, Peaton, Sathuperi and Shamonda viruses (PEAV, SATV and SHAV) of the genus Orthobunyavirus and D’Aguilar virus (DAGV) of the genus Orbivirus were newly confirmed and were thought to be involved in congenital malformations of cattle (Matsumori et al., 2002; Ohashi et al., 2004a; Yanase et al., 2004, 2005b). D’Aguilar virus is regarded as a strain of Palyam virus (PALV) together with CHUV, and a genetic reassortant was probably generated between the two viruses (Ohashi et al., 2004a). Although other PALVs, such as Bunyip Creek virus (BCV) and CSIRO Village virus, were frequently isolated from cattle blood, none of these viruses has been associated with diseases in Australia (Cybinski and St George, 1982; Cybinski and Muller, 1990). An EHDV strain, which was regarded as an IBAV variant, was widely spread in the western part of Japan in 1997 and caused a large outbreak of bovine abortion in the spread area (Ohashi et al., 1999). Although the variant is serologically related to IBAV, its outer capsid protein, VP2, is poorly identical to that of IBAV (Ohashi et al., 2002), suggesting that it is sorted into a different serotype of EHDV than serotype 2. The above-mentioned viruses are thought to be transmitted by Culicoides biting midges (Diptera: Ceratopogonidae) and are probably introduced with the infected midges from overseas to Japan by seasonal winds every summer (Yanase et al., 2005a). However, details of their natural transmission cycles and spread patterns are uncertain. Subclinical infections are often caused by arboviruses in ruminants. In the case of congenital abnormalities by AKAV, AINOV and CHUV, the affected calves are generally noted several months after the virus has spread (Kurogi et al., 1975; Goto et al., 1988; Tsuda et al., 2004). Therefore, an early warning system is needed to detect the virus incursion and spread before clinical cases are apparent. Monitoring in vectors and sentinel animals is certainly an important component of systems for rapid and definite detection of arboviral activity. Kagoshima, located at the southern end of the main islands of Japan, is thought to be one of the gateways for arbovirus incursion from overseas. Indeed, the nationwide surveillance for bovine arboviral diseases in Japan indicated that the transmission was often started in the southwestern 2
T. Kato et al.
region including Kagoshima (Yanase et al., 2005a; Forman et al., 2008). In a previous investigation between 1985 and 2002, 85 arbovirus isolates including AKAV, AINOV, CHUV, DAGV and EHDV were obtained from Culicoides biting midges (Yanase et al., 2005a), supporting that Kagoshima is an ideal point for the monitoring. Herein, we report the monitoring of arboviral activity in the same area for an additional 11 years. Virus isolation was attempted from Culicoides biting midges and sentinel cattle, and characterization of isolated viruses was conducted. Materials and Methods Study sites The sampling of Culicoides biting midges was conducted in Kagoshima city (31.17–31.45°N, 128.23–131.12°E) of Kagoshima prefecture in southern Japan between May and November from 2003 to 2012. The sampling period was shortened to 5 months (July–November) in 2013. Sentinel cattle were also bled around the same period (May to early December in 2003–2012 and July to early December in 2013). During the study period, the annual mean temperature of Kagoshima city was 18–19°C, annual precipitation was 1500–3000 mm, and humidity was more than 60% throughout the year. Average temperatures of May and November, which were the initial and the last month of the monitoring period, were about 21 and 16°C, respectively. The hottest season was from July to August, and its average temperature was 28–29°C. The rainy season was from the end of May to the middle of July. Culicoides collections were conducted at two cowsheds, one (cowshed A; 31.54°N, 130.51°E, altitude 12 m) located in the coastal plain and the other (cowshed B; 31.58°N, 130.48°E, altitude 150 m) in the mountainous area 5 km northwest from cowshed A (Fig. 1). Cowshed A, holding 5–10 head of cattle and several sheep, was surrounded by a residential area, paddy rice fields and hills covered by evergreens and bamboos. Two head of sentinel cattle were placed there. Cowshed B, housing about 10 cattle, was built on the side of a forested mountain. There was a small paddy rice field 250 m from the cowshed. Insect collections Adult Culicoides were collected twice a week using a light trap equipped with a 6-W blacklight and a downdraught suction motor. The light trap was set in the cowsheds as close to the cattle as possible, at 1.4 m from the ground, between 4:00 and 6:00 p.m., and was collected the following morning between 8:00 and 9:00 a.m. Collected midges were kept for ≥3 days by feeding them 10% sucrose solution at 25°C until the sucked blood was fully digested. All of the surviving midges from cowshed A were counted and pooled by species based on the identification keys for Japa© 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
T. Kato et al.
Bovine Arboviruses in Southern Japan
tion before inoculation. Prepared samples were inoculated to tube cultures that were thereafter maintained by rotation at 37°C and observed for cytopathic effect (CPE) over 7 days. The cultures were passaged twice in the same manner if CPE was not observed. Dot immunobinding assay
Fig. 1. Maps showing the cowsheds for arbovirus monitoring. Satellite image was provided by Japan Aerospace Exploration Agency (JAXA). Bar indicates 1 km.
nese Culicoides species (Wada, 1999). One hundred midges at cowshed B were selected randomly and counted by species to calculate the composition rate and then preserved as mixed species pools based on the collection date. The Culicoides samples were stored at 80°C until needed for arbovirus isolation. Virus isolation The collected midges were processed as described previously for virus isolation (Yanase et al., 2005a). Briefly, each pool of midges was manually homogenized by a plastic pestle and was suspended in Eagle’s minimum essential medium (MEM; Nissui, Tokyo, Japan) supplemented with 0.295% tryptose phosphate broth, 0.015% sodium bicarbonate, 10 lg/ml gentamicin sulphate and 2.5 lg/ml amphotericin B. The medium volume varied with the number of midges: 2 ml for ≤10 midges, 3 ml for 11–40 midges and 5 ml for ≥41 midges. The suspension was clarified by centrifugation at 10 000 g for 3 min, and the supernatant was used for inoculation to cells. Sentinel cattle were free from AKAV, AINOV, IBAV and CHUV antibodies at the time of selection. Seropositive cattle were replaced to na€ıve ones at the start of each season. Heparinized blood samples were collected twice a week and were separated into plasma and blood cells by centrifugation. Blood cells were washed three times with phosphatebuffered saline (PBS) to eliminate the antibodies. All blood samples were stored at 80°C until use. Baby hamster kidney (BHK-21) and hamster lung (HmLu-1) cells were incubated in a test tube with MEM supplemented with 0.295% tryptose phosphate broth, 0.015% sodium bicarbonate and 10% bovine serum overnight at 37°C and were washed three times with Earl’s solu© 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
A dot immunobinding assay was performed as described previously (Yoshida and Tsuda, 1998). Briefly, the supernatants of cell culture showing CPE were sucked into an Immobilon PVDF membrane (Millipore, Billerica, MA, USA) by a slot blotting apparatus under negative pressure. After rinsing with Tris-buffered saline (TBS) [20 mM Tris–HCl (pH 7.5), 0.15 M NaCl], the membrane was immersed in blocking buffer (8% skim milk in TBS) for 2 h and, then, reacted with monoclonal antibodies (MAbs) to the nucleocapsid protein and the viral surface glycoprotein (Gc) of AKAV, and the Gc protein of AINOV and PEAV, and with the mouse antisera to IBAV, CHUV and BTV serotype 21. After washing three times with TBS, the membrane was further reacted with horseradish peroxidase-conjugated goat antibody to mouse IgG (MP Biochemicals, Solon, OH, USA) for 30 min and then washed five times with TBS. The immune complexes were detected by colour development with 0.027% 3,3-diaminobenzidine tetrahydrochloride and 0.016% H2O2 in TBS. RT-PCR and sequence analysis RNA was extracted from the supernatant of the inoculated cell cultures showing CPE by the High Pure Viral RNA Kit (Roche Diagnostics, Mannheim, Germany) and was tested for RT-PCRs targeting the S RNA segment of the orthobunyaviruses, segment 3 of orbiviruses, the L RNA polymerase gene of rhabdoviruses and segment 2 of EHDV serotype 1, respectively, as previously described (Matsumori et al., 2002; Ohashi et al., 2004b; Bourhy et al., 2005; Maan et al., 2010) (Table S1). The RT-PCR assay and sequencing analysis targeting segment 2 of DAGV and BCV were carried out with the primers listed in Table 1 using the Titan One Tube RT-PCR Kit in accordance with the manufacturer’s instructions (Roche). A primer pair, DAGVL2F and DAGVLR2-2, was designed for specific detection of DAGV segment 2. The cDNA synthesis was conducted at 50°C for 30 min followed by 2 min at 94°C after pre-incubation with the primers at 94°C for 4 min and then on ice. The PCR temperature profile was 10 cycles of 30 s at 94°C, 30 s at 55°C and 45 s at 68°C, followed by 25 cycles of 30 s at 94°C, 30 s at 55°C and 45 s at 68°C, with the latter time increased by 5 s per cycle. Specific detection of BCV segment 2 was carried out by a primer pair, BCVseg2-F1 and BCVseg2-R1. The RT-PCR conditions were modified from 3
Bovine Arboviruses in Southern Japan
T. Kato et al.
Table 1. Oligonucleotide primers designed for detection and sequencing of segment 2 of D’Aguilar virus (DAGV) and Bunyip Creek virus (BCV) Primer For DAGV DAGVL2F DAGVL2R-2 For BCV BCVseg2-F BCVseg2-F1 BCVseg2-F2 BCVseg2-F3 BCVseg2-F4 BCVseg2-R4 BCVseg2-R3 BCVseg2-R2 BCVseg2-R1 BCVseg2-R
Sequence (50 –30 )
Position
Purpose
CAAAGGATCTGCAGGTTGATG CCCATCCTTTCTGACTCTGC
338-358 1016-997
RT-PCR, sequencing RT-PCR, sequencing
GTTAAAATCTCGCAGG ATTGAAATTGGACAAGGAGG TCGATCTAAATCCAAGGGTG GTGGTCGCACATCTATCTGA TTATCGGATGATACGAAGCC ATCACCCTCCACCATATTCC ATGACAGCTCGTCGTTCTTC TTAGATAGTTGCGGAAAATC TCTACCATATGGTATCACAG GTAAGTTGATTCCGCAGGTA
1-16 680-699 1380-1399 2041-2060 2774-2793 271-252 996-977 1692-1673 2413-2394 3064-3045
RT-PCR, sequencing RT-PCR, sequencing Sequencing RT-PCR, Sequencing Sequencing Sequencing RT-PCR, Sequencing Sequencing RT-PCR, sequencing RT-PCR, sequencing
those of DAGV detection by changing the annealing temperature to 50°C and the extension time to 2 min. The 50 and 30 terminals of coding region of BCV segment 2 were amplified with primer pairs, BCVseg2-F and BCVseg2-R3, and BCVseg2-F3 and BCVseg2R, respectively. The PCR products were purified with the QIAquick PCR Purification Kit (Qiagen, Hiden, Germany) and then directly sequenced with the BigDye Terminator Cycle Sequencing Kit v3.1 (Life technologies, Carlsbad, CA, USA) on an ABI 3100-Avanti Genetic Analyzer (Life Technologies). The nucleotide sequences were edited by DNASIS Pro Ver. 3.0 (Hitachi Solutions, Tokyo, Japan). Sequence similarity was searched with the basic local alignment search tool (BLAST), and pairwise nucleotide (nt) and amino acid (aa) sequence identities were calculated with GENETYX software version 10 (GENETYX, Tokyo, Japan). Neighbour-joining (NJ) consensus tree was constructed using MEGA 6 from MUSCLE sequence alignments with default parameters (Tamura et al., 2013). The reliability of the inferred consensus tree was tested by bootstrap resampling of 1000 pseudoreplicate data set. Transmission electron microscopy The infected cells were fixed with 2% glutaraldehyde and 2% paraformaldehyde in 0.1 M phosphate buffer for 2 h and post-fixed in 2% OsO4 in 0.1-M phosphate buffer adding 7.5% sucrose for an hour. The specimens were embedded in low-viscosity resin. Ultrathin sections were stained with uranyl acetate and lead citrate and examined with a JEM-1010 electron microscope (JEOL, Tokyo, Japan). Results A total of 198 367 Culicoides biting midges comprising 12 species collected at cowshed A were divided into 2440 pools 4
and were screened for arboviruses (Table 2). The captured numbers of Culicoides biting midges gradually increased from May onward. Higher numbers of midges were observed between July and October, while low numbers were collected in November (data not shown). The most dominant species was Culicoides oxystoma (71.3%), followed by C. tainanus (13.0%), C. punctatus (9.6%), C. jacobsoni (5.0%), C. matsuzawai (0.5%) and C. sumatrae (0.4%) (Table 2). The other six species were occasionally captured at the study site. Culicoides tainanus was the most abundant in May and June, but decreased after July (data not shown); C. punctatus showed a similar trend. The C. oxystoma population increased in July and comprised most of the processed midges (>85%) between August and October. At cowshed B, 12 species were trapped during the study period. The counting of measured fractions revealed that the percentage of C. tainanus in the catches was the highest (the mean ratio is 40.2%). Five species, C. punctatus, C. oxystoma, C. jacobsoni, C. matsuzawai and C. sumatrae, were caught at somewhat high ratios (the mean ratio ranged from 9.3% to 16.0%). The other six species, C. humeralis, C. arakawae, C. cylindratus, C. actoni, C. lungchiensis and C. ohmorii, were contained in the catches at lower ratios (< 1.0%). Totally, 1202 blood samples were collected from sentinel cattle during the study period. Because the blood samples were separated into plasma and blood cells, 2404 inocula were attempted for virus isolation. A total of 77 virus isolates were obtained during the study period; 33 isolates from Culicoides biting midges (Table 3) and 44 isolates from the sentinel cattle (Table 4). The occasions of arbovirus isolations were limited from the end of August to the beginning of December (Table 5). Through initial screening with the dot immunobinding assay, 18 and 5 isolates were identified as AKAV and PEAV, respectively. Two isolates, KSB-1/P/08 and KSB-2/C/08, © 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
T. Kato et al.
Bovine Arboviruses in Southern Japan
Table 2. Numbers of each Culicoides species processed for virus isolation from cowshed A in Kagoshima between 2003 and 2013 Culicoides species
No. of processed
No. of pools
No. of positive pool
Isolated virus
C. oxystoma C. tainanus C. punctatus C. jacobsoni C. matsuzawai C. sumatrae C. arakawae C. cylindratus C. actoni C. lungchiensis C. humeralis C. ohmorii Total
141 341 25 864 19 002 9854 1040 859 200 106 60 30 8 3 198 367
566 408 547 378 154 213 63 58 26 19 5 3 2440
18 1 1 1 0 0 0 0 0 0 0 0 21
Akabane, Sathuperi, Bunyip Creek, D’ Aguilar BTV Serotype 16 EHDV Serotype 1 Peaton
Table 3. Isolation of arboviruses from Culicoides biting midges in Kagoshima between 2003 and 2013
Virus
Year
Bunyaviridae, Orthobunyavirus Akabane 2008 Peaton 2010 Sathuperi 2008 Reoviridae, Orbivirus Bunyip Creek
D’ Aguilar
2009 2009 2013 2013
BTV serotype 16 EHDV serotype 1 Unassigned virus KSB-8/C/09 KSB-3~5/C/10
2008 2008 2013 2009 2010
Day/Month
No. of isolates
Culicoides species
Cowshed
Sensitive cell line
Genebank accession no.†
17 October 14 September 26 September
1 1 1
C. oxystoma C. jacobsoni C. oxystoma
A A A
BHK-21 BHK-21 BHK-21
29 September– 13 November 29 September 27 August– 1 November 27 September– 25 October 31 October 4 November 29 October
4
C.oxystoma
A
BHK-21
AB973438 (KSB-4/C/09)
1 12
Culicoides spp. C. oxystoma
B A
BHK-21 BHK-21
AB973439 (KSB-7/C/09) AB978375 (KSB-1/C/13)
6
Culicoides spp.
B
BHK-21
1 1 1
Culicoides spp. C. tainanus C. punctatus
B A A
BHK-21 BHK-21 BHK-21
1 3
Culicoides spp. Culicoides spp.
B B
BHK-21 BHK-21
10 November 24 September– 12 October
AB698469, AB698474, AB698482 (KSB-2/C/08)‡
AB686226 (KSB-7/C/08)‡ AB686225 (KSB-6/C/08)‡ AB978374 (KSB-36/C/13)
†
Parentheses indicate a virus isolate used for sequence analysis. Sequences determined in previous studies.
‡
from sentinel cattle and C. oxystoma in 2008 were bound to MAbs to the AKAV nucleocapsid protein, but not to the other MAbs and the antisera, indicating that these isolates are a member of the former Simbu serogroup, but not AKAV, AINOV and PEAV. The RT-PCR targeting the S RNA segment of orthobunyaviruses amplified a fragment with the predicted size from the extracted RNA, and their determined sequences were highly homologous (97.7– 99.4% nt identity) with the S RNA segment of Japanese SATV isolates since 1999. Further sequence analysis conducted for the M RNA segment that encodes the outer© 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
surface glycoproteins confirmed that these isolates could be assigned to SATV (Yanase et al., 2012). Forty-eight isolates were sorted into BTV, PALV or EHDV by species-specific dot immunobinding and RT-PCR assays for orbivirus detection. For further characterization, genetic analysis for segment 2 encoding the serotype-/ strain-specific outer capsid was conducted for the obtained orbiviruses. In 2008, two isolates, KSB-6/C/08 and KSB-7/ C/08, bound to the anti-BTV serum and the sequence analysis of their segment 2 indicated that these isolates belonged to BTV serotype 16 as described previously (Shirafuji et al., 5
Bovine Arboviruses in Southern Japan
T. Kato et al.
Table 4. Isolation of arboviruses from sentinel cattle at cowshed A in Kagoshima between 2003 and 2013 No. of isolates Virus
Year
Bunyaviridae, Orthobunyavirus Akabane 2003 2006 2008 2013 Peaton 2006 2010 Sathuperi 2008 Reoviridae, Orbivirus Bunyip Creek 2009 D’ Aguilar 2013 Rhabdoviridae, Ephemerovirus KSB-1/P/03 2003
Day/Month
Plasma
10–24 October 27–30 October 14–17 October 13 September–8 October 19–22 September 17 September 22 September
4 2 1 3 2 1 1
13–17 November 15 October–3 December
1
22 August
1
Blood cells
2 1 2 2 1
2 18
Sensitive cell line
Genebank accession no.†
BHK-21, HmLu-1 BHK-21, HmLu-1 BHK-21, HmLu-1 BHK-21, HmLu-1 BHK-21 BHK-21 HmLu-1
AB973363 (KSB-1/P/08)
BHK-21 BHK-21
AB973437 (KSB-1/E/09)
BHK-21
AB978373
†
Parentheses indicate a virus isolate used for sequence analysis.
Table 5. Monthly prevalence rate (%) of arboviruses in blood samples of sentinel cattle and pools of Culicoides biting midges Year†
May
June
July
Blood samples of sentinel cattle 2003 0 0 0 2006 0 0 0 2008 0 0 0 2009 0 0 0 2010 0 0 0 2013 – – 0 Pools of Culicoides biting midges at cowshed A 2008 0 0 0 2009 0 0 0 2010 0 0 0 2013 – – 0 Pools of Culicoides biting midges at cowshed B 2008 0 0 0 2009 0 0 0 2010 0 0 0 2013 – – 0
August
September
October
November
December
5.6 0 0 0 0 0
0 11.1 5.6 0 6.3 12.5
22.2 11.1 11.1 0 0 50.0
0 0 0 12.5 0 50.0
0 0 0 0 0 50.0
0 0 0 5.0
2.9 5.6 2.5 5.0
2.6 10.5 0 45.5
4.2 10.0 0 7.7
– – – –
0 12.5 28.6 12.5
14.3 0 12.5 55.6
0 20.0 0 0
– – – –
0 0 0 0
†
Years when viruses were isolated.
2012). Seven isolates in 2009 reacted with the antiserum to CHUV, which is specific for strains of PALV. However, no successful amplification was observed in CHUV- and DAGV-specific RT-PCR assays for them (data not shown). An orbivirus which was obtained in 2008 from cattle in Okinawa prefecture, located to the south of Kagoshima prefecture, was identified as BCV, because its segment 2 (accession numbers: AB973440) has significant identity (90.7%) with that of Australian prototype isolate (CSIRO58, accession number: AB973436). The products by RT-PCR assays with BCV segment 2-specific primer pairs from three orbiviruses isolated in 2009 had 99.8 and 90.7% nt identities with the corresponding sequence of BCV Okinawan and 6
Australian isolates, respectively. This result clearly showed that the Kagoshima isolates could be assigned to BCV. The other 4 orbiviruses isolated in 2009 were also identified as BCV by RT-PCR assay with its specific primer pair (BCVseg2-F1 and BCVseg2-R1). A PALV-specific sequence was detected from 37 isolates obtained in 2013 by RT-PCR targeting segment 3. These isolates were identified as DAGV by further RT-PCR assay for specific detection of segment 2. The sequence of the RT-PCR product from an obtained isolate, KSB-1/C/13, revealed high nt identities (92.6–95.9%) with the corresponding region of the past DAGV Japanese isolates. One isolate, KSB-36/C/13, from C. punctatus in 2013, was classified into EHDV by RT-PCR targeting © 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
T. Kato et al.
segment 3. Further RT-PCR assays for specific detection of IBAV and the IBAV 1997 variant were unsuccessful with KSB-36/C/13 (data not shown). In contrast, the partial sequence of segment 2 (position on genome between 1324 and 1984 nt) was amplified from the isolate by RT-PCR with the primer set specific for EHDV serotype 1. Sequence analysis of the RT-PCR product showed that the closest relative of KSB-36/C/13 is an Australian isolate (DPP2209) of EHDV serotype 1, because of their significant similarity (94.8% nt identity). The remaining five isolates were not identified by the above-mentioned methods and thus underwent physicochemical and morphological characterizations. In ultrathin sections of infected BHK-21 cells with KSB-1/P/03 isolated in 2003, bullet-shaped virions characteristic of rhabdoviruses were observed within the rough endoplasmic reticulum and at the surface of the plasma membrane (Fig. 2). The RT-PCR assay targeting the L RNA polymerase gene of rhabdoviruses amplified a 431-bp fragment from KSB-1/ P/03. The sequence of the PCR product was significantly related to members of the genus Ephemerovirus with 70.1‒ 73.5% nt and 72.8‒79.4% aa identities and showed the maximum nt and aa identities with Berrimah virus and Kotonkan virus (KOTV), respectively. NJ tree based on the 135 aa residues of L RNA polymerase of selected rhabdoviruses revealed that KSB-1/P/03 clearly clusters with members of the genus Ephemerovirus (Fig. 3). The trees constructed by maximum likelihood, maximum evolution and maximum parsimony methods supported the NJ tree (data not shown). Four cytopathic agents, KSB-8/C/09, KSB-3/C/10, KSB-4/C/10 and KSB-5/C/10, isolated from Culicoides spp. in 2009 and 2010, were not sensitive to chloroform (data not shown). By electron microscopy, spherical virus particles of 50‒100 nm in diameter and tubule structures were observed in the cytoplasm of BHK21 cells infected with these cytopathic agents (Fig. 4a,b). Of 21 isolates from Culicoides biting midges collected at cowshed A, 18 isolates comprising AKAV, SATV, DAGV and BCV were obtained from C. oxystoma. Bluetongue virus serotype 16, EHDV serotype 1 and PEAV were isolated from C. tainanus, C. punctatus and C. jacobsoni, respectively (Table 3). Twelve isolates comprising BTV serotype 16, BCV, DAGV and the unidentified viruses were obtained from mixed midge pools collected at cowshed B. Akabane virus, PEAV, SATV, DAGV and BCV were also isolated from sentinel cattle. Sathuperi virus and BCV were obtained from bovine plasma and blood cells, respectively, although the other viruses were isolated from both materials. The isolation of the presumed rhabdovirus was made only from bovine plasma, not from midges. BHK-21 cells could be used for the isolation of all viruses obtained in this study. However, the virus isolation in HmLu-1 cells was successful only for AKAV and SATV. © 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
Bovine Arboviruses in Southern Japan
Fig. 2. Electron micrographs of BHK-21 cells infected with the presumed rhabdovirus, KSB-1/P/03. Bullet-shaped virus particles are shown budding from the plasma membrane. Bar indicates 200 nm.
Discussion The species compositions of Culicoides biting midges at cowshed A and B were different. The differences of species components and abundances of Culicoides biting midges might depend on the environments surrounding the cowsheds. The paddy rice fields close to the study sites were probably breeding grounds for C. oxystoma, as described in previous studies (Kitaoka and Morii, 1963; Yanase et al., 2013). At present, the larval habitats of C. tainanus, which occupied the majority of the collected midges at cowshed B, are uncertain. The efficacy of arbovirus transmission is supposed to be much related with the presence and abundance of exact vector species. Environmental factors affecting the habitat of each Culicoides species should be elucidated for better choice of monitoring sites. Akabane virus, SATV, DAGV and BCV were selectively isolated from C. oxystoma at cowshed A, indicating that this species plays a principal role for the transmission of multiple arboviruses, as described previously (Yanase et al., 2005a). Culicoides oxystoma is widely distributed through tropical to temperate zones in the Palaearctic, Oriental, Australasia and Afrotropic regions (Wirth and Hubert, 1989; Bakhoum et al., 2013). However, trials of arbovirus isolation from C. oxystoma have not been extensively conducted in its range except Japan and Australia (Standfast et al., 1984; Yanase et al., 2005a). It should be further considered of its vector role in the transmission of veterinary arboviruses. Isolation of EHDV serotype 1, IBAV and AINOV from C. punctatus during the past and present surveillances, suggests that it has some vector competence for multiple arboviruses such as C. oxystoma. Bluetongue virus serotype 16 and PEAV were isolated from C. tainanus and 7
Bovine Arboviruses in Southern Japan
T. Kato et al.
Berrimah virus (AY854636)
100
Bovine ephemeral fever virus (AY854671)
99
Kimberley virus (AY854637) 58
100
Malakal virus (JQ941707) Adelaide River virus (AY854635)
70
100
Ephemerovirus
Obodhiang virus(HM856902) Yata virus (KM085030)
KSB-1/P/03 (AB978373)
61
Koolpinyah virus (KM085029)
60 100
Kotonkan virus (AY854638) Coastal Plains virus (GQ294473)
99
Tibrogargan virus (NC_020804) Durham virus (FJ952155)
96
Tupaia virus (NC_007020) Drosophila melanogaster sigmavirus (GQ375258) Oak-Vale virus (AY854670) Humpty doo virus (AY854643) Charleville virus (AY854644)
97 88
Almpiwar virus (JN882642) Fukuoka virus (AY854651)
88
Le Dantec virus (AY854650) Drosophila obscura sigmavirus (NC_022580) 100
Pike fry rhabdovirus (FJ872827) Spring viraemia of carp virus (NC_002803) Vesicular stomatitis New Jersey virus (NC_024473)
73
Chandipura virus (HM627187)
69
Vesiculovirus
Isfahan virus (NC_020806) Vesicular stomatitis Indiana virus (NC_001560) Cocal virus (EU373657)
81 92
Vesicular stomatitis Alagoas virus (EU373658) Ngaingan virus (NC_013955) Flanders virus (JX431885)
77 54
Wongabel virus (NC_011639) Rabies virus (NC_001542) European bat lyssavirus 1 (NC_009527)
100 78
European bat lyssavirus 2 (NC_009528)
0.05
C. jacobsoni, respectively. Although these species have not been recognized as arbovirus vectors, the result implies that they also have some vector competence. Although the number of monitoring sites was limited, our attempt to detect arbovirus spreads linked to the occurrence of diseases was sensitive. Sporadic outbreaks of bovine congenital abnormalities by AKAV were reported in southern Japan in winters 2003/2004 and 2008/2009, and our monitoring successfully detected the AKAV spread before clinical cases accumulated. Because AKAV is difficult to detect from the affected bovine foetuses which are often infected before several months (Kurogi et al., 1975, 1977), virological monitoring in sentinel animals and vector insects might be the most effective way to gain the causative viruses. When a large outbreak of bovine encephalomyelitis occurred in the southern part of Japan in 2006 (Kono et al., 2008), a causative AKAV variant was isolated from both Culicoides biting midges and sentinel cattle at the monitoring sites. Several clinical cases of AKAV encephalomyelitis were also reported in the area including the study sites in 8
Lyssavirus
Fig. 3. Neighbour-joining consensus tree constructed from the aligned sequences of the L RNA polymerases (135 aa residues) of selected animal rhabdoviruses. Genbank accession numbers for each retrieved sequence are shown in parentheses. Bootstrap values (%) were calculated from 1000 psuedoreplicates of the data set, and values above 50% are indicated at the appropriate nodes. Scale bar indicates estimated number of aa substitution per site.
2013 (MAFF, 2014), and AKAV isolates were gained at the same time. The viral sequences detected from the diseased cattle were almost identical with those of the isolated viruses during the monitoring (data not shown), supporting our findings. Silent circulation of the other arboviruses could be visualized by the screening in Culicoides biting midges and sentinel cattle during the study period. Some of them can be supposed to be involved in congenital abnormalities in ruminants. Although no diseased cattle were observed around the study sites in 2008, a few suspected cases of bovine malformations associated with SATV were reported previously (Yanase et al., 2012). Interestingly, Schmallenberg virus (SBV), which recently emerged in northern Europe and caused a large-scale outbreak of malformation in ruminants, was thought to be a variant of SATV (Goller et al., 2012; Hoffmann et al., 2012; Yanase et al., 2012). Because of the serious impact of SBV on the livestock industry, it becomes a target for monitoring in many countries (EFSA, 2013). It may be important to consider the role © 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
T. Kato et al.
(a)
(b)
Fig. 4. Ultrastructural analysis of BHK-21 cells infected with an unidentified virus, KSB-3/C/10. (a) Viral particles in cells infected with KSB-3/C/ 10. (b) Tubules (indicated by arrows) are generated in KSB-3/C/10 infected cells. Similar virus particles and tubule structures were observed in cells infected with KSB-8/C/09, KSB-4/C/10 and KSB-5/C/10. Bars indicate 100 nm in (a) and 200 nm in (b), respectively.
of C. oxystoma in the transmission and spread of SBV in its range, if the virus is introduced there. Peaton virus appeared to cause congenital abnormalities in a lamb by experimental infection (Parsonson et al., 1982) and was possibly involved in foetal abnormalities in calves in the field (Matsumori et al., 2002). The isolation of PEAV from C. jacobsoni was the first field isolation from this species. If this species has a vector competence for PEAV, the risk of virus transmission exists in wide ranges of Asia and Ocea© 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
Bovine Arboviruses in Southern Japan
nia, where it is distributed (Wirth and Hubert, 1989; Dyce et al., 2007). D’Aguilar virus had been isolated from cattle and Culicoides biting midges in Australia and Japan, but its involvement in malformations was confirmed only in the latter country (Knudson et al., 1984; Ohashi et al., 2004a). No obvious increase in bovine malformations was reported around the study sites between 2013 autumn and 2014 spring after the DAGV spread. The nationwide serological surveillance showed that BTV sometimes spreads to the northern part of Japan, and sporadic outbreaks of bluetongue occurred in the central region (Goto et al., 2004). Although BTV has been isolated from various species of Culicoides biting midges in the world (Maclachlan, 2011), its detection in identified Culicoides species on the main islands of Japan had been unsuccessful for a long time. Culicoides tainanus, from which BTV serotype 16 was isolated, is widely distributed all through Japan and often becomes predominant in midges collected at cowsheds (Wada, 1999). Although the single isolation seems to be insufficient to prove the vector competence of C. tainanus, the findings suggested its vector role in transmission and spread of BTV. So far, it has been uncertain whether BCV and EHDV serotype 1 are involved in cattle diseases. Bunyip Creek virus was isolated repeatedly in the 1970s and 1980s in Australia (Cybinski and St George, 1982; Standfast et al., 1984; Cybinski and Muller, 1990). There was no evidence of any association between BCV and bovine malformations after the virus spread in the southern part of Japan between 2009 autumn and 2010 spring, suggesting that the virus has weak or no teratogenicity to cattle, unlike the closely related viruses, CHUV and DAGV. Also, no illness associated with EHDV serotype 1 has been observed in Japan, either in past studies (Miura et al., 1988) or the present one. Almost all EHDV strains, except several virulent strains such as IBAV and the IBAV 1997 variant, probably produce subclinical infection in cattle (Savini et al., 2011). However, comprehensive characterization of EHDV isolates may help to identify pathogenicity factors included in the particular strains. Several isolates obtained in this study are potentially novel arboviruses. The presumed rhabdovirus, KSB-1/P/03, from sentinel cattle in 2003 was genetically close to the viruses belonging to the genus Ephemerovirus, but not identical with them. This isolate is considered to be a new ephemerovirus. The sentinel cattle infected with KSB-1/P/ 03 exhibited no clinical symptoms. However, this may not exclude the possibility of a disease associated with this rhabdovirus, because related BEFV and KOTV are known to cause severe febrile illness in cattle (Walker, 2005). Attempts at identification by dot immunobinding assays and RT-PCRs did not produce any positive result for four isolates from pools containing mixed Culicoides species at 9
Bovine Arboviruses in Southern Japan
cowshed B in 2009 and 2010. Although no isolation was made from cattle, the combination of the materials (Culicoides biting midges) and susceptive cell line (mammalian BHK-21 cells) in the virus isolation indicates that they are a kind of arbovirus. In addition, their physicochemical (chloroform-resistant property) and morphological features (viral particle size and appearance of tubules in the infected cells) suggest that they belong to the genus Orbivirus (Roy, 2007). Genetic analysis of their genomes will be necessary to clarify their taxonomic position and relationship with other known viruses. Our results clearly showed that BHK-21 cells are more sensitive for virus isolation than HmLu-1. Especially, it appeared that BHK-21 cells only have enough susceptibility for orbivirus infection in the virus isolation from the fieldcollected materials. However, other susceptible cell lines, such as mosquito and Culicoides cells, should be tested to enhance arbovirus isolation. It was likely that blood cells are more favourable than plasma for isolation of BCV and DAGV. Intimate association of orbiviruses with erythrocytes was probably responsible for this finding as suggested before (Brewer and MacLachlan, 1994; Aradaib et al., 1997). The occasion of orthobunyavirus isolation from sentinel cattle was more frequent than that from Culicoides biting midges. In contrast, BTV and EHDV were found only in Culicoides biting midges. Furthermore, there were differences in the viruses obtained from Culicoides biting midges at cowsheds A and B. This might depend on Culicoides fauna at each collection site. Data from the current study indicate the importance of selecting the most accurate materials, methods and monitoring sites for sensitive virus detection. Three orthobunyaviruses, AKAV, PEAV and SATV, and two orbiviruses, DAGV and BCV, were parallely isolated from Culicoides biting midges and sentinel cattle at cowshed A. Short (
ORIGINAL ARTICLE
Bovine Arboviruses in Culicoides Biting Midges and Sentinel Cattle in Southern Japan from 2003 to 2013 T. Kato1, H. Shirafuji1, S. Tanaka1, M. Sato2, M. Yamakawa2, T. Tsuda2 and T. Yanase1 1 2
Kyushu Research Station, National Institute of Animal Health, NARO, Kagoshima, Japan National Institute of Animal Health, NARO, Tsukuba, Japan
Keywords: arbovirus; sentinel cattle; Culicoides biting midge; orthobunyavirus; orbivirus; rhabdovirus Correspondence: T. Yanase. Kyushu Research Station, National Institute of Animal Health, NARO, 2702 Chuzan, Kagoshima 891-0105, Japan. Tel.: +81 99 268 2078; Fax: +81 99 268 3088; E-mail: [email protected] Received for publication August 20, 2014 doi:10.1111/tbed.12324
Summary Epizootic congenital abnormalities, encephalomyelitis and febrile illnesses in cattle caused by arthropod-borne viruses (arboviruses) are prevalent in Japan. Causative viruses including orthobunyaviruses, orbiviruses and rhabdovirus are thought to be transmitted by Culicoides biting midges. Recently, the incursions of several arboviruses, potentially Culicoides-borne, were newly confirmed in Japan. However, their spread pattern and exact vector species are currently uncertain. Attempts to isolate arboviruses from Culicoides biting midges and sentinel cattle were conducted in Kagoshima, located at the southernmost end of the main islands of Japan, a potentially high-risk area for incursion of arboviral diseases and outbreak of endemic ones. Seventy-eight isolates comprising Akabane, Peaton and Sathuperi viruses of the genus Orthobunyavirus of the family Bunyaviridae, bluetongue virus serotype 16, D’Aguilar virus, Bunyip Creek virus and epizootic haemorrhagic disease virus serotype 1 of the genus Orbivirus of the family Reoviridae, a potentially novel rhabdovirus of the genus Ephemerovirus and unidentified orbivirus-like viruses were obtained from Culicoides biting midges and sentinel cattle between 2003 and 2013. Akabane, Sathuperi, D’Aguilar and Bunyip Creek viruses were selectively isolated from Culicoides oxystoma, suggesting this vector’s responsibility for these arbovirus outbreaks. The results of virus isolation also implied that C. tainanus, C. jacobsoni and C. punctatus are competent for the transmission of bluetongue virus serotype 16, Peaton virus and epizootic haemorrhagic disease virus serotype 1, respectively. Our monitoring in Culicoides biting midges and sentinel cattle detected the circulation of Akabane virus just prior to the accumulations of bovine congenital abnormalities and encephalomyelitis by it around study sites in 2003, 2006, 2008 and 2013. Silent circulations of the other arboviruses, including potentially new viruses, were also detected during the study period.
Introduction Arthropod-borne viruses (Arboviruses) are transmitted by haematophagous arthropod vectors, such as mosquitoes, ticks and Culicoides biting midges. Over 500 arboviruses have been described in the world to date, with some being etiological agents of human and animal diseases (Hart, 2001). Substantial trade restrictions in the livestock industry have occasionally resulted from the © 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
presence of certain arboviruses, such as bluetongue virus (BTV) (OIE, 2014). In Japan, arbovirus infections in cattle are frequently reported and have caused serious damage to the beef and dairy industries. Large outbreaks of abnormal deliveries in cattle, such as abortion, stillbirth, premature birth and congenital malformations, have been periodically caused by arboviruses, such as Akabane and Aino viruses (AKAV and AINOV) of the genus Orthobunyavirus of the family 1
Bovine Arboviruses in Southern Japan
Bunyaviridae, and Chuzan virus (CHUV) of the genus Orbivirus of the family Reoviridae (Forman et al., 2008). As recently reported, AKAV was also associated with bovine epizootic encephalomyelitis (Kono et al., 2008). Seasonal epidemics of acute febrile illness in cattle caused by bovine ephemeral fever virus (BEFV) of the genus Ephemerovirus of the family Rhabdoviridae have been sporadically observed (Kato et al., 2009). Before the 1980s, Ibaraki virus (IBAV), a strain of epizootic hemorrhagic disease virus (EHDV) serotype 2, was involved in large outbreaks of diseases in cattle characterized by fever and deglutitive disorder (Forman et al., 2008). Recently, Peaton, Sathuperi and Shamonda viruses (PEAV, SATV and SHAV) of the genus Orthobunyavirus and D’Aguilar virus (DAGV) of the genus Orbivirus were newly confirmed and were thought to be involved in congenital malformations of cattle (Matsumori et al., 2002; Ohashi et al., 2004a; Yanase et al., 2004, 2005b). D’Aguilar virus is regarded as a strain of Palyam virus (PALV) together with CHUV, and a genetic reassortant was probably generated between the two viruses (Ohashi et al., 2004a). Although other PALVs, such as Bunyip Creek virus (BCV) and CSIRO Village virus, were frequently isolated from cattle blood, none of these viruses has been associated with diseases in Australia (Cybinski and St George, 1982; Cybinski and Muller, 1990). An EHDV strain, which was regarded as an IBAV variant, was widely spread in the western part of Japan in 1997 and caused a large outbreak of bovine abortion in the spread area (Ohashi et al., 1999). Although the variant is serologically related to IBAV, its outer capsid protein, VP2, is poorly identical to that of IBAV (Ohashi et al., 2002), suggesting that it is sorted into a different serotype of EHDV than serotype 2. The above-mentioned viruses are thought to be transmitted by Culicoides biting midges (Diptera: Ceratopogonidae) and are probably introduced with the infected midges from overseas to Japan by seasonal winds every summer (Yanase et al., 2005a). However, details of their natural transmission cycles and spread patterns are uncertain. Subclinical infections are often caused by arboviruses in ruminants. In the case of congenital abnormalities by AKAV, AINOV and CHUV, the affected calves are generally noted several months after the virus has spread (Kurogi et al., 1975; Goto et al., 1988; Tsuda et al., 2004). Therefore, an early warning system is needed to detect the virus incursion and spread before clinical cases are apparent. Monitoring in vectors and sentinel animals is certainly an important component of systems for rapid and definite detection of arboviral activity. Kagoshima, located at the southern end of the main islands of Japan, is thought to be one of the gateways for arbovirus incursion from overseas. Indeed, the nationwide surveillance for bovine arboviral diseases in Japan indicated that the transmission was often started in the southwestern 2
T. Kato et al.
region including Kagoshima (Yanase et al., 2005a; Forman et al., 2008). In a previous investigation between 1985 and 2002, 85 arbovirus isolates including AKAV, AINOV, CHUV, DAGV and EHDV were obtained from Culicoides biting midges (Yanase et al., 2005a), supporting that Kagoshima is an ideal point for the monitoring. Herein, we report the monitoring of arboviral activity in the same area for an additional 11 years. Virus isolation was attempted from Culicoides biting midges and sentinel cattle, and characterization of isolated viruses was conducted. Materials and Methods Study sites The sampling of Culicoides biting midges was conducted in Kagoshima city (31.17–31.45°N, 128.23–131.12°E) of Kagoshima prefecture in southern Japan between May and November from 2003 to 2012. The sampling period was shortened to 5 months (July–November) in 2013. Sentinel cattle were also bled around the same period (May to early December in 2003–2012 and July to early December in 2013). During the study period, the annual mean temperature of Kagoshima city was 18–19°C, annual precipitation was 1500–3000 mm, and humidity was more than 60% throughout the year. Average temperatures of May and November, which were the initial and the last month of the monitoring period, were about 21 and 16°C, respectively. The hottest season was from July to August, and its average temperature was 28–29°C. The rainy season was from the end of May to the middle of July. Culicoides collections were conducted at two cowsheds, one (cowshed A; 31.54°N, 130.51°E, altitude 12 m) located in the coastal plain and the other (cowshed B; 31.58°N, 130.48°E, altitude 150 m) in the mountainous area 5 km northwest from cowshed A (Fig. 1). Cowshed A, holding 5–10 head of cattle and several sheep, was surrounded by a residential area, paddy rice fields and hills covered by evergreens and bamboos. Two head of sentinel cattle were placed there. Cowshed B, housing about 10 cattle, was built on the side of a forested mountain. There was a small paddy rice field 250 m from the cowshed. Insect collections Adult Culicoides were collected twice a week using a light trap equipped with a 6-W blacklight and a downdraught suction motor. The light trap was set in the cowsheds as close to the cattle as possible, at 1.4 m from the ground, between 4:00 and 6:00 p.m., and was collected the following morning between 8:00 and 9:00 a.m. Collected midges were kept for ≥3 days by feeding them 10% sucrose solution at 25°C until the sucked blood was fully digested. All of the surviving midges from cowshed A were counted and pooled by species based on the identification keys for Japa© 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
T. Kato et al.
Bovine Arboviruses in Southern Japan
tion before inoculation. Prepared samples were inoculated to tube cultures that were thereafter maintained by rotation at 37°C and observed for cytopathic effect (CPE) over 7 days. The cultures were passaged twice in the same manner if CPE was not observed. Dot immunobinding assay
Fig. 1. Maps showing the cowsheds for arbovirus monitoring. Satellite image was provided by Japan Aerospace Exploration Agency (JAXA). Bar indicates 1 km.
nese Culicoides species (Wada, 1999). One hundred midges at cowshed B were selected randomly and counted by species to calculate the composition rate and then preserved as mixed species pools based on the collection date. The Culicoides samples were stored at 80°C until needed for arbovirus isolation. Virus isolation The collected midges were processed as described previously for virus isolation (Yanase et al., 2005a). Briefly, each pool of midges was manually homogenized by a plastic pestle and was suspended in Eagle’s minimum essential medium (MEM; Nissui, Tokyo, Japan) supplemented with 0.295% tryptose phosphate broth, 0.015% sodium bicarbonate, 10 lg/ml gentamicin sulphate and 2.5 lg/ml amphotericin B. The medium volume varied with the number of midges: 2 ml for ≤10 midges, 3 ml for 11–40 midges and 5 ml for ≥41 midges. The suspension was clarified by centrifugation at 10 000 g for 3 min, and the supernatant was used for inoculation to cells. Sentinel cattle were free from AKAV, AINOV, IBAV and CHUV antibodies at the time of selection. Seropositive cattle were replaced to na€ıve ones at the start of each season. Heparinized blood samples were collected twice a week and were separated into plasma and blood cells by centrifugation. Blood cells were washed three times with phosphatebuffered saline (PBS) to eliminate the antibodies. All blood samples were stored at 80°C until use. Baby hamster kidney (BHK-21) and hamster lung (HmLu-1) cells were incubated in a test tube with MEM supplemented with 0.295% tryptose phosphate broth, 0.015% sodium bicarbonate and 10% bovine serum overnight at 37°C and were washed three times with Earl’s solu© 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
A dot immunobinding assay was performed as described previously (Yoshida and Tsuda, 1998). Briefly, the supernatants of cell culture showing CPE were sucked into an Immobilon PVDF membrane (Millipore, Billerica, MA, USA) by a slot blotting apparatus under negative pressure. After rinsing with Tris-buffered saline (TBS) [20 mM Tris–HCl (pH 7.5), 0.15 M NaCl], the membrane was immersed in blocking buffer (8% skim milk in TBS) for 2 h and, then, reacted with monoclonal antibodies (MAbs) to the nucleocapsid protein and the viral surface glycoprotein (Gc) of AKAV, and the Gc protein of AINOV and PEAV, and with the mouse antisera to IBAV, CHUV and BTV serotype 21. After washing three times with TBS, the membrane was further reacted with horseradish peroxidase-conjugated goat antibody to mouse IgG (MP Biochemicals, Solon, OH, USA) for 30 min and then washed five times with TBS. The immune complexes were detected by colour development with 0.027% 3,3-diaminobenzidine tetrahydrochloride and 0.016% H2O2 in TBS. RT-PCR and sequence analysis RNA was extracted from the supernatant of the inoculated cell cultures showing CPE by the High Pure Viral RNA Kit (Roche Diagnostics, Mannheim, Germany) and was tested for RT-PCRs targeting the S RNA segment of the orthobunyaviruses, segment 3 of orbiviruses, the L RNA polymerase gene of rhabdoviruses and segment 2 of EHDV serotype 1, respectively, as previously described (Matsumori et al., 2002; Ohashi et al., 2004b; Bourhy et al., 2005; Maan et al., 2010) (Table S1). The RT-PCR assay and sequencing analysis targeting segment 2 of DAGV and BCV were carried out with the primers listed in Table 1 using the Titan One Tube RT-PCR Kit in accordance with the manufacturer’s instructions (Roche). A primer pair, DAGVL2F and DAGVLR2-2, was designed for specific detection of DAGV segment 2. The cDNA synthesis was conducted at 50°C for 30 min followed by 2 min at 94°C after pre-incubation with the primers at 94°C for 4 min and then on ice. The PCR temperature profile was 10 cycles of 30 s at 94°C, 30 s at 55°C and 45 s at 68°C, followed by 25 cycles of 30 s at 94°C, 30 s at 55°C and 45 s at 68°C, with the latter time increased by 5 s per cycle. Specific detection of BCV segment 2 was carried out by a primer pair, BCVseg2-F1 and BCVseg2-R1. The RT-PCR conditions were modified from 3
Bovine Arboviruses in Southern Japan
T. Kato et al.
Table 1. Oligonucleotide primers designed for detection and sequencing of segment 2 of D’Aguilar virus (DAGV) and Bunyip Creek virus (BCV) Primer For DAGV DAGVL2F DAGVL2R-2 For BCV BCVseg2-F BCVseg2-F1 BCVseg2-F2 BCVseg2-F3 BCVseg2-F4 BCVseg2-R4 BCVseg2-R3 BCVseg2-R2 BCVseg2-R1 BCVseg2-R
Sequence (50 –30 )
Position
Purpose
CAAAGGATCTGCAGGTTGATG CCCATCCTTTCTGACTCTGC
338-358 1016-997
RT-PCR, sequencing RT-PCR, sequencing
GTTAAAATCTCGCAGG ATTGAAATTGGACAAGGAGG TCGATCTAAATCCAAGGGTG GTGGTCGCACATCTATCTGA TTATCGGATGATACGAAGCC ATCACCCTCCACCATATTCC ATGACAGCTCGTCGTTCTTC TTAGATAGTTGCGGAAAATC TCTACCATATGGTATCACAG GTAAGTTGATTCCGCAGGTA
1-16 680-699 1380-1399 2041-2060 2774-2793 271-252 996-977 1692-1673 2413-2394 3064-3045
RT-PCR, sequencing RT-PCR, sequencing Sequencing RT-PCR, Sequencing Sequencing Sequencing RT-PCR, Sequencing Sequencing RT-PCR, sequencing RT-PCR, sequencing
those of DAGV detection by changing the annealing temperature to 50°C and the extension time to 2 min. The 50 and 30 terminals of coding region of BCV segment 2 were amplified with primer pairs, BCVseg2-F and BCVseg2-R3, and BCVseg2-F3 and BCVseg2R, respectively. The PCR products were purified with the QIAquick PCR Purification Kit (Qiagen, Hiden, Germany) and then directly sequenced with the BigDye Terminator Cycle Sequencing Kit v3.1 (Life technologies, Carlsbad, CA, USA) on an ABI 3100-Avanti Genetic Analyzer (Life Technologies). The nucleotide sequences were edited by DNASIS Pro Ver. 3.0 (Hitachi Solutions, Tokyo, Japan). Sequence similarity was searched with the basic local alignment search tool (BLAST), and pairwise nucleotide (nt) and amino acid (aa) sequence identities were calculated with GENETYX software version 10 (GENETYX, Tokyo, Japan). Neighbour-joining (NJ) consensus tree was constructed using MEGA 6 from MUSCLE sequence alignments with default parameters (Tamura et al., 2013). The reliability of the inferred consensus tree was tested by bootstrap resampling of 1000 pseudoreplicate data set. Transmission electron microscopy The infected cells were fixed with 2% glutaraldehyde and 2% paraformaldehyde in 0.1 M phosphate buffer for 2 h and post-fixed in 2% OsO4 in 0.1-M phosphate buffer adding 7.5% sucrose for an hour. The specimens were embedded in low-viscosity resin. Ultrathin sections were stained with uranyl acetate and lead citrate and examined with a JEM-1010 electron microscope (JEOL, Tokyo, Japan). Results A total of 198 367 Culicoides biting midges comprising 12 species collected at cowshed A were divided into 2440 pools 4
and were screened for arboviruses (Table 2). The captured numbers of Culicoides biting midges gradually increased from May onward. Higher numbers of midges were observed between July and October, while low numbers were collected in November (data not shown). The most dominant species was Culicoides oxystoma (71.3%), followed by C. tainanus (13.0%), C. punctatus (9.6%), C. jacobsoni (5.0%), C. matsuzawai (0.5%) and C. sumatrae (0.4%) (Table 2). The other six species were occasionally captured at the study site. Culicoides tainanus was the most abundant in May and June, but decreased after July (data not shown); C. punctatus showed a similar trend. The C. oxystoma population increased in July and comprised most of the processed midges (>85%) between August and October. At cowshed B, 12 species were trapped during the study period. The counting of measured fractions revealed that the percentage of C. tainanus in the catches was the highest (the mean ratio is 40.2%). Five species, C. punctatus, C. oxystoma, C. jacobsoni, C. matsuzawai and C. sumatrae, were caught at somewhat high ratios (the mean ratio ranged from 9.3% to 16.0%). The other six species, C. humeralis, C. arakawae, C. cylindratus, C. actoni, C. lungchiensis and C. ohmorii, were contained in the catches at lower ratios (< 1.0%). Totally, 1202 blood samples were collected from sentinel cattle during the study period. Because the blood samples were separated into plasma and blood cells, 2404 inocula were attempted for virus isolation. A total of 77 virus isolates were obtained during the study period; 33 isolates from Culicoides biting midges (Table 3) and 44 isolates from the sentinel cattle (Table 4). The occasions of arbovirus isolations were limited from the end of August to the beginning of December (Table 5). Through initial screening with the dot immunobinding assay, 18 and 5 isolates were identified as AKAV and PEAV, respectively. Two isolates, KSB-1/P/08 and KSB-2/C/08, © 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
T. Kato et al.
Bovine Arboviruses in Southern Japan
Table 2. Numbers of each Culicoides species processed for virus isolation from cowshed A in Kagoshima between 2003 and 2013 Culicoides species
No. of processed
No. of pools
No. of positive pool
Isolated virus
C. oxystoma C. tainanus C. punctatus C. jacobsoni C. matsuzawai C. sumatrae C. arakawae C. cylindratus C. actoni C. lungchiensis C. humeralis C. ohmorii Total
141 341 25 864 19 002 9854 1040 859 200 106 60 30 8 3 198 367
566 408 547 378 154 213 63 58 26 19 5 3 2440
18 1 1 1 0 0 0 0 0 0 0 0 21
Akabane, Sathuperi, Bunyip Creek, D’ Aguilar BTV Serotype 16 EHDV Serotype 1 Peaton
Table 3. Isolation of arboviruses from Culicoides biting midges in Kagoshima between 2003 and 2013
Virus
Year
Bunyaviridae, Orthobunyavirus Akabane 2008 Peaton 2010 Sathuperi 2008 Reoviridae, Orbivirus Bunyip Creek
D’ Aguilar
2009 2009 2013 2013
BTV serotype 16 EHDV serotype 1 Unassigned virus KSB-8/C/09 KSB-3~5/C/10
2008 2008 2013 2009 2010
Day/Month
No. of isolates
Culicoides species
Cowshed
Sensitive cell line
Genebank accession no.†
17 October 14 September 26 September
1 1 1
C. oxystoma C. jacobsoni C. oxystoma
A A A
BHK-21 BHK-21 BHK-21
29 September– 13 November 29 September 27 August– 1 November 27 September– 25 October 31 October 4 November 29 October
4
C.oxystoma
A
BHK-21
AB973438 (KSB-4/C/09)
1 12
Culicoides spp. C. oxystoma
B A
BHK-21 BHK-21
AB973439 (KSB-7/C/09) AB978375 (KSB-1/C/13)
6
Culicoides spp.
B
BHK-21
1 1 1
Culicoides spp. C. tainanus C. punctatus
B A A
BHK-21 BHK-21 BHK-21
1 3
Culicoides spp. Culicoides spp.
B B
BHK-21 BHK-21
10 November 24 September– 12 October
AB698469, AB698474, AB698482 (KSB-2/C/08)‡
AB686226 (KSB-7/C/08)‡ AB686225 (KSB-6/C/08)‡ AB978374 (KSB-36/C/13)
†
Parentheses indicate a virus isolate used for sequence analysis. Sequences determined in previous studies.
‡
from sentinel cattle and C. oxystoma in 2008 were bound to MAbs to the AKAV nucleocapsid protein, but not to the other MAbs and the antisera, indicating that these isolates are a member of the former Simbu serogroup, but not AKAV, AINOV and PEAV. The RT-PCR targeting the S RNA segment of orthobunyaviruses amplified a fragment with the predicted size from the extracted RNA, and their determined sequences were highly homologous (97.7– 99.4% nt identity) with the S RNA segment of Japanese SATV isolates since 1999. Further sequence analysis conducted for the M RNA segment that encodes the outer© 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
surface glycoproteins confirmed that these isolates could be assigned to SATV (Yanase et al., 2012). Forty-eight isolates were sorted into BTV, PALV or EHDV by species-specific dot immunobinding and RT-PCR assays for orbivirus detection. For further characterization, genetic analysis for segment 2 encoding the serotype-/ strain-specific outer capsid was conducted for the obtained orbiviruses. In 2008, two isolates, KSB-6/C/08 and KSB-7/ C/08, bound to the anti-BTV serum and the sequence analysis of their segment 2 indicated that these isolates belonged to BTV serotype 16 as described previously (Shirafuji et al., 5
Bovine Arboviruses in Southern Japan
T. Kato et al.
Table 4. Isolation of arboviruses from sentinel cattle at cowshed A in Kagoshima between 2003 and 2013 No. of isolates Virus
Year
Bunyaviridae, Orthobunyavirus Akabane 2003 2006 2008 2013 Peaton 2006 2010 Sathuperi 2008 Reoviridae, Orbivirus Bunyip Creek 2009 D’ Aguilar 2013 Rhabdoviridae, Ephemerovirus KSB-1/P/03 2003
Day/Month
Plasma
10–24 October 27–30 October 14–17 October 13 September–8 October 19–22 September 17 September 22 September
4 2 1 3 2 1 1
13–17 November 15 October–3 December
1
22 August
1
Blood cells
2 1 2 2 1
2 18
Sensitive cell line
Genebank accession no.†
BHK-21, HmLu-1 BHK-21, HmLu-1 BHK-21, HmLu-1 BHK-21, HmLu-1 BHK-21 BHK-21 HmLu-1
AB973363 (KSB-1/P/08)
BHK-21 BHK-21
AB973437 (KSB-1/E/09)
BHK-21
AB978373
†
Parentheses indicate a virus isolate used for sequence analysis.
Table 5. Monthly prevalence rate (%) of arboviruses in blood samples of sentinel cattle and pools of Culicoides biting midges Year†
May
June
July
Blood samples of sentinel cattle 2003 0 0 0 2006 0 0 0 2008 0 0 0 2009 0 0 0 2010 0 0 0 2013 – – 0 Pools of Culicoides biting midges at cowshed A 2008 0 0 0 2009 0 0 0 2010 0 0 0 2013 – – 0 Pools of Culicoides biting midges at cowshed B 2008 0 0 0 2009 0 0 0 2010 0 0 0 2013 – – 0
August
September
October
November
December
5.6 0 0 0 0 0
0 11.1 5.6 0 6.3 12.5
22.2 11.1 11.1 0 0 50.0
0 0 0 12.5 0 50.0
0 0 0 0 0 50.0
0 0 0 5.0
2.9 5.6 2.5 5.0
2.6 10.5 0 45.5
4.2 10.0 0 7.7
– – – –
0 12.5 28.6 12.5
14.3 0 12.5 55.6
0 20.0 0 0
– – – –
0 0 0 0
†
Years when viruses were isolated.
2012). Seven isolates in 2009 reacted with the antiserum to CHUV, which is specific for strains of PALV. However, no successful amplification was observed in CHUV- and DAGV-specific RT-PCR assays for them (data not shown). An orbivirus which was obtained in 2008 from cattle in Okinawa prefecture, located to the south of Kagoshima prefecture, was identified as BCV, because its segment 2 (accession numbers: AB973440) has significant identity (90.7%) with that of Australian prototype isolate (CSIRO58, accession number: AB973436). The products by RT-PCR assays with BCV segment 2-specific primer pairs from three orbiviruses isolated in 2009 had 99.8 and 90.7% nt identities with the corresponding sequence of BCV Okinawan and 6
Australian isolates, respectively. This result clearly showed that the Kagoshima isolates could be assigned to BCV. The other 4 orbiviruses isolated in 2009 were also identified as BCV by RT-PCR assay with its specific primer pair (BCVseg2-F1 and BCVseg2-R1). A PALV-specific sequence was detected from 37 isolates obtained in 2013 by RT-PCR targeting segment 3. These isolates were identified as DAGV by further RT-PCR assay for specific detection of segment 2. The sequence of the RT-PCR product from an obtained isolate, KSB-1/C/13, revealed high nt identities (92.6–95.9%) with the corresponding region of the past DAGV Japanese isolates. One isolate, KSB-36/C/13, from C. punctatus in 2013, was classified into EHDV by RT-PCR targeting © 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
T. Kato et al.
segment 3. Further RT-PCR assays for specific detection of IBAV and the IBAV 1997 variant were unsuccessful with KSB-36/C/13 (data not shown). In contrast, the partial sequence of segment 2 (position on genome between 1324 and 1984 nt) was amplified from the isolate by RT-PCR with the primer set specific for EHDV serotype 1. Sequence analysis of the RT-PCR product showed that the closest relative of KSB-36/C/13 is an Australian isolate (DPP2209) of EHDV serotype 1, because of their significant similarity (94.8% nt identity). The remaining five isolates were not identified by the above-mentioned methods and thus underwent physicochemical and morphological characterizations. In ultrathin sections of infected BHK-21 cells with KSB-1/P/03 isolated in 2003, bullet-shaped virions characteristic of rhabdoviruses were observed within the rough endoplasmic reticulum and at the surface of the plasma membrane (Fig. 2). The RT-PCR assay targeting the L RNA polymerase gene of rhabdoviruses amplified a 431-bp fragment from KSB-1/ P/03. The sequence of the PCR product was significantly related to members of the genus Ephemerovirus with 70.1‒ 73.5% nt and 72.8‒79.4% aa identities and showed the maximum nt and aa identities with Berrimah virus and Kotonkan virus (KOTV), respectively. NJ tree based on the 135 aa residues of L RNA polymerase of selected rhabdoviruses revealed that KSB-1/P/03 clearly clusters with members of the genus Ephemerovirus (Fig. 3). The trees constructed by maximum likelihood, maximum evolution and maximum parsimony methods supported the NJ tree (data not shown). Four cytopathic agents, KSB-8/C/09, KSB-3/C/10, KSB-4/C/10 and KSB-5/C/10, isolated from Culicoides spp. in 2009 and 2010, were not sensitive to chloroform (data not shown). By electron microscopy, spherical virus particles of 50‒100 nm in diameter and tubule structures were observed in the cytoplasm of BHK21 cells infected with these cytopathic agents (Fig. 4a,b). Of 21 isolates from Culicoides biting midges collected at cowshed A, 18 isolates comprising AKAV, SATV, DAGV and BCV were obtained from C. oxystoma. Bluetongue virus serotype 16, EHDV serotype 1 and PEAV were isolated from C. tainanus, C. punctatus and C. jacobsoni, respectively (Table 3). Twelve isolates comprising BTV serotype 16, BCV, DAGV and the unidentified viruses were obtained from mixed midge pools collected at cowshed B. Akabane virus, PEAV, SATV, DAGV and BCV were also isolated from sentinel cattle. Sathuperi virus and BCV were obtained from bovine plasma and blood cells, respectively, although the other viruses were isolated from both materials. The isolation of the presumed rhabdovirus was made only from bovine plasma, not from midges. BHK-21 cells could be used for the isolation of all viruses obtained in this study. However, the virus isolation in HmLu-1 cells was successful only for AKAV and SATV. © 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
Bovine Arboviruses in Southern Japan
Fig. 2. Electron micrographs of BHK-21 cells infected with the presumed rhabdovirus, KSB-1/P/03. Bullet-shaped virus particles are shown budding from the plasma membrane. Bar indicates 200 nm.
Discussion The species compositions of Culicoides biting midges at cowshed A and B were different. The differences of species components and abundances of Culicoides biting midges might depend on the environments surrounding the cowsheds. The paddy rice fields close to the study sites were probably breeding grounds for C. oxystoma, as described in previous studies (Kitaoka and Morii, 1963; Yanase et al., 2013). At present, the larval habitats of C. tainanus, which occupied the majority of the collected midges at cowshed B, are uncertain. The efficacy of arbovirus transmission is supposed to be much related with the presence and abundance of exact vector species. Environmental factors affecting the habitat of each Culicoides species should be elucidated for better choice of monitoring sites. Akabane virus, SATV, DAGV and BCV were selectively isolated from C. oxystoma at cowshed A, indicating that this species plays a principal role for the transmission of multiple arboviruses, as described previously (Yanase et al., 2005a). Culicoides oxystoma is widely distributed through tropical to temperate zones in the Palaearctic, Oriental, Australasia and Afrotropic regions (Wirth and Hubert, 1989; Bakhoum et al., 2013). However, trials of arbovirus isolation from C. oxystoma have not been extensively conducted in its range except Japan and Australia (Standfast et al., 1984; Yanase et al., 2005a). It should be further considered of its vector role in the transmission of veterinary arboviruses. Isolation of EHDV serotype 1, IBAV and AINOV from C. punctatus during the past and present surveillances, suggests that it has some vector competence for multiple arboviruses such as C. oxystoma. Bluetongue virus serotype 16 and PEAV were isolated from C. tainanus and 7
Bovine Arboviruses in Southern Japan
T. Kato et al.
Berrimah virus (AY854636)
100
Bovine ephemeral fever virus (AY854671)
99
Kimberley virus (AY854637) 58
100
Malakal virus (JQ941707) Adelaide River virus (AY854635)
70
100
Ephemerovirus
Obodhiang virus(HM856902) Yata virus (KM085030)
KSB-1/P/03 (AB978373)
61
Koolpinyah virus (KM085029)
60 100
Kotonkan virus (AY854638) Coastal Plains virus (GQ294473)
99
Tibrogargan virus (NC_020804) Durham virus (FJ952155)
96
Tupaia virus (NC_007020) Drosophila melanogaster sigmavirus (GQ375258) Oak-Vale virus (AY854670) Humpty doo virus (AY854643) Charleville virus (AY854644)
97 88
Almpiwar virus (JN882642) Fukuoka virus (AY854651)
88
Le Dantec virus (AY854650) Drosophila obscura sigmavirus (NC_022580) 100
Pike fry rhabdovirus (FJ872827) Spring viraemia of carp virus (NC_002803) Vesicular stomatitis New Jersey virus (NC_024473)
73
Chandipura virus (HM627187)
69
Vesiculovirus
Isfahan virus (NC_020806) Vesicular stomatitis Indiana virus (NC_001560) Cocal virus (EU373657)
81 92
Vesicular stomatitis Alagoas virus (EU373658) Ngaingan virus (NC_013955) Flanders virus (JX431885)
77 54
Wongabel virus (NC_011639) Rabies virus (NC_001542) European bat lyssavirus 1 (NC_009527)
100 78
European bat lyssavirus 2 (NC_009528)
0.05
C. jacobsoni, respectively. Although these species have not been recognized as arbovirus vectors, the result implies that they also have some vector competence. Although the number of monitoring sites was limited, our attempt to detect arbovirus spreads linked to the occurrence of diseases was sensitive. Sporadic outbreaks of bovine congenital abnormalities by AKAV were reported in southern Japan in winters 2003/2004 and 2008/2009, and our monitoring successfully detected the AKAV spread before clinical cases accumulated. Because AKAV is difficult to detect from the affected bovine foetuses which are often infected before several months (Kurogi et al., 1975, 1977), virological monitoring in sentinel animals and vector insects might be the most effective way to gain the causative viruses. When a large outbreak of bovine encephalomyelitis occurred in the southern part of Japan in 2006 (Kono et al., 2008), a causative AKAV variant was isolated from both Culicoides biting midges and sentinel cattle at the monitoring sites. Several clinical cases of AKAV encephalomyelitis were also reported in the area including the study sites in 8
Lyssavirus
Fig. 3. Neighbour-joining consensus tree constructed from the aligned sequences of the L RNA polymerases (135 aa residues) of selected animal rhabdoviruses. Genbank accession numbers for each retrieved sequence are shown in parentheses. Bootstrap values (%) were calculated from 1000 psuedoreplicates of the data set, and values above 50% are indicated at the appropriate nodes. Scale bar indicates estimated number of aa substitution per site.
2013 (MAFF, 2014), and AKAV isolates were gained at the same time. The viral sequences detected from the diseased cattle were almost identical with those of the isolated viruses during the monitoring (data not shown), supporting our findings. Silent circulation of the other arboviruses could be visualized by the screening in Culicoides biting midges and sentinel cattle during the study period. Some of them can be supposed to be involved in congenital abnormalities in ruminants. Although no diseased cattle were observed around the study sites in 2008, a few suspected cases of bovine malformations associated with SATV were reported previously (Yanase et al., 2012). Interestingly, Schmallenberg virus (SBV), which recently emerged in northern Europe and caused a large-scale outbreak of malformation in ruminants, was thought to be a variant of SATV (Goller et al., 2012; Hoffmann et al., 2012; Yanase et al., 2012). Because of the serious impact of SBV on the livestock industry, it becomes a target for monitoring in many countries (EFSA, 2013). It may be important to consider the role © 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
T. Kato et al.
(a)
(b)
Fig. 4. Ultrastructural analysis of BHK-21 cells infected with an unidentified virus, KSB-3/C/10. (a) Viral particles in cells infected with KSB-3/C/ 10. (b) Tubules (indicated by arrows) are generated in KSB-3/C/10 infected cells. Similar virus particles and tubule structures were observed in cells infected with KSB-8/C/09, KSB-4/C/10 and KSB-5/C/10. Bars indicate 100 nm in (a) and 200 nm in (b), respectively.
of C. oxystoma in the transmission and spread of SBV in its range, if the virus is introduced there. Peaton virus appeared to cause congenital abnormalities in a lamb by experimental infection (Parsonson et al., 1982) and was possibly involved in foetal abnormalities in calves in the field (Matsumori et al., 2002). The isolation of PEAV from C. jacobsoni was the first field isolation from this species. If this species has a vector competence for PEAV, the risk of virus transmission exists in wide ranges of Asia and Ocea© 2015 Blackwell Verlag GmbH • Transboundary and Emerging Diseases.
Bovine Arboviruses in Southern Japan
nia, where it is distributed (Wirth and Hubert, 1989; Dyce et al., 2007). D’Aguilar virus had been isolated from cattle and Culicoides biting midges in Australia and Japan, but its involvement in malformations was confirmed only in the latter country (Knudson et al., 1984; Ohashi et al., 2004a). No obvious increase in bovine malformations was reported around the study sites between 2013 autumn and 2014 spring after the DAGV spread. The nationwide serological surveillance showed that BTV sometimes spreads to the northern part of Japan, and sporadic outbreaks of bluetongue occurred in the central region (Goto et al., 2004). Although BTV has been isolated from various species of Culicoides biting midges in the world (Maclachlan, 2011), its detection in identified Culicoides species on the main islands of Japan had been unsuccessful for a long time. Culicoides tainanus, from which BTV serotype 16 was isolated, is widely distributed all through Japan and often becomes predominant in midges collected at cowsheds (Wada, 1999). Although the single isolation seems to be insufficient to prove the vector competence of C. tainanus, the findings suggested its vector role in transmission and spread of BTV. So far, it has been uncertain whether BCV and EHDV serotype 1 are involved in cattle diseases. Bunyip Creek virus was isolated repeatedly in the 1970s and 1980s in Australia (Cybinski and St George, 1982; Standfast et al., 1984; Cybinski and Muller, 1990). There was no evidence of any association between BCV and bovine malformations after the virus spread in the southern part of Japan between 2009 autumn and 2010 spring, suggesting that the virus has weak or no teratogenicity to cattle, unlike the closely related viruses, CHUV and DAGV. Also, no illness associated with EHDV serotype 1 has been observed in Japan, either in past studies (Miura et al., 1988) or the present one. Almost all EHDV strains, except several virulent strains such as IBAV and the IBAV 1997 variant, probably produce subclinical infection in cattle (Savini et al., 2011). However, comprehensive characterization of EHDV isolates may help to identify pathogenicity factors included in the particular strains. Several isolates obtained in this study are potentially novel arboviruses. The presumed rhabdovirus, KSB-1/P/03, from sentinel cattle in 2003 was genetically close to the viruses belonging to the genus Ephemerovirus, but not identical with them. This isolate is considered to be a new ephemerovirus. The sentinel cattle infected with KSB-1/P/ 03 exhibited no clinical symptoms. However, this may not exclude the possibility of a disease associated with this rhabdovirus, because related BEFV and KOTV are known to cause severe febrile illness in cattle (Walker, 2005). Attempts at identification by dot immunobinding assays and RT-PCRs did not produce any positive result for four isolates from pools containing mixed Culicoides species at 9
Bovine Arboviruses in Southern Japan
cowshed B in 2009 and 2010. Although no isolation was made from cattle, the combination of the materials (Culicoides biting midges) and susceptive cell line (mammalian BHK-21 cells) in the virus isolation indicates that they are a kind of arbovirus. In addition, their physicochemical (chloroform-resistant property) and morphological features (viral particle size and appearance of tubules in the infected cells) suggest that they belong to the genus Orbivirus (Roy, 2007). Genetic analysis of their genomes will be necessary to clarify their taxonomic position and relationship with other known viruses. Our results clearly showed that BHK-21 cells are more sensitive for virus isolation than HmLu-1. Especially, it appeared that BHK-21 cells only have enough susceptibility for orbivirus infection in the virus isolation from the fieldcollected materials. However, other susceptible cell lines, such as mosquito and Culicoides cells, should be tested to enhance arbovirus isolation. It was likely that blood cells are more favourable than plasma for isolation of BCV and DAGV. Intimate association of orbiviruses with erythrocytes was probably responsible for this finding as suggested before (Brewer and MacLachlan, 1994; Aradaib et al., 1997). The occasion of orthobunyavirus isolation from sentinel cattle was more frequent than that from Culicoides biting midges. In contrast, BTV and EHDV were found only in Culicoides biting midges. Furthermore, there were differences in the viruses obtained from Culicoides biting midges at cowsheds A and B. This might depend on Culicoides fauna at each collection site. Data from the current study indicate the importance of selecting the most accurate materials, methods and monitoring sites for sensitive virus detection. Three orthobunyaviruses, AKAV, PEAV and SATV, and two orbiviruses, DAGV and BCV, were parallely isolated from Culicoides biting midges and sentinel cattle at cowshed A. Short (
