PATHOLOGY AND MOLECULAR DETECTION OF RABIES VIRUS IN FERRET BADGERS ASSOCIATED WITH A RABIES OUTBREAK IN TAIWAN Open Access

Rabies, possibly the oldest zoonotic disease, is caused by Rabies virus (RABV, family Rhabdoviridae, genus Lyssavirus). Rabies occurs worldwide, and only a few countries and regions are free from this disease. Rabies virus infects nearly all warm-blooded animals (Jackson and Fu 2013). There were >60,000 rabies-associated human deaths worldwide in 2010, primarily in Africa and Asia (WHO 2013). Dogs (Canis lupus familiaris) are the principal host of RABV in developing countries. Because of well-established canine vaccination programs, wildlife has become the most important source of RABV infection in many developed countries (Wandeler 1987; Blanton et al. 2007). Rabies is generally regarded as having a single-species reservoir, with spillover to other dead-end hosts (Smith and Wilkinson 2002; Krebs et al. 2003). Bats (Chiroptera), raccoons (Procyon lotor), skunks (Mephitidae), and foxes (Vulpes spp.) are regarded as the main reservoirs of rabies (Jackson 2008). Members of the weasel family (Mustelidae), including Melogale, Meles, and Mellivora species, are also susceptible to rabies (Smith 2002). Chinese ferret badgers (CNFBs, Melogale moschata moschata) are a primary rabies host and source of human rabies in southeast China (Zhang et al. 2009; Liu et al. 2010). Wildlife disease surveillance has long been carried out in Taiwan with government support; however, because of the limited species and populations of native free-ranging wild animals, surveillance was mainly focused on zoo animals. Because of the increased importance of free-ranging wildlife in emerging and reemerging diseases of animals and humans, the disease surveillance program was expanded in 2011 to include native free-ranging wildlife. Through this program, carcasses of native free-ranging wildlife have been submitted routinely to selected laboratories for necropsy and further examination since 2012. Until the first three rabid Taiwan ferret badgers (TWFBs) (Melogale moschata subaurantiaca) were diagnosed in mid-June 2013 through this program (Chiou et al. 2014), Taiwan had been considered rabies free for >50 yr, with the last autochthonous human and animal rabies cases recorded in 1959 and 1961, respectively.

In natural or experimental RABV infection, nonsuppurative meningoencephalomyelitis, ganglionitis, and sialadenitis are the characteristic histopathologic findings (Maxie and Youssef 2007). However, the severity and distribution of lesions and the amount and distribution of viral antigens are variable, depending on the host, viral strain, and clinical course (Hicks et al. 2009; Stein et al. 2010). Although ferret badger (FB)–associated rabies had been reported in China (Zhang et al. 2009; Liu et al. 2010), the associated pathologic changes and viral antigen distribution have not been described, to our knowledge. This information is essential for establishing more-accurate sampling and diagnosis modalities for FB-associated rabies, especially when composite sampling is not feasible. Herein, we characterize the pathologic changes and the pattern of viral antigen distribution during the recent rabies outbreak in TWFBs.

During May 2012 and January 2013, three adult, female TWFBs were rescued from central Taiwan and submitted to the wildlife first aid station for treatment. Clinically, TWFB 1 showed signs of emaciation, coma, paddling, loss of pain response, and reduced body temperature with a 2-cm skin wound on the chin; TWFB 2 was extremely weak and unable to move; and TWFB 3 displayed signs of weakness, labored and noisy breathing, hypersalivation, and exuding of foamy fluid from the mouth and nose. Despite supportive treatment, they died within 1–3 d and were submitted to National Taiwan University for routine disease surveillance.

Full necropsy was performed in TWFBs 1–3, and the entire brain and representative tissue samples from other major organs were collected. The first (C1) and second (C2) segments of cervical spinal cord and submandibular salivary gland were also collected from TWFB 3. All samples were fixed in 10% neutral-buffered formalin. Representative sections of various parts of the brain, C1/C2 spinal cord, and other tissues were processed for routine embedding in paraffin.

Retrospectively, various formalin-fixed, paraffin-embedded tissue blocks, containing cerebrum or cerebellum from six other TWFBs, including one necropsied in 2004 (TWFB 4), two in 2006, two in 2011, and one in 2012, provided by other diagnostic laboratories were also studied. Among the six TWFBs, TWFB 4 (female) and one obtained in 2011 (TWFB 5) (male) were extremely weak when found and died the next day on arrival at the rescue station; four other TWFBs were road-kills. All six carcasses had been frozen before sending for necropsy and displayed variable postmortem autolysis. Except for the paraffin blocks, no other frozen tissues were available.

Tissue blocks were sectioned at 5 µm, stained with H&E, and examined by light microscopy. Based on frequency and severity, lesions of nervous and nonnervous system were graded as mild, moderate, or severe.

The immunohistochemical (IHC) staining was performed using the advanced Super Sensitive^™ Polymer-HRP [horseradish peroxidase] IHC Detection System (BioGenex, Fremont, California, USA) according to the manufacturer’s instructions with mouse anti-RABV glycoprotein immunoglobulin (Ig)G2a monoclonal antibody (Abcam, Boston, Massachusetts, USA) or rabbit anti-attenuated challenge virus strand (aCVSr) nucleoprotein polyclonal antibody kindly provided by the Laboratory of Transmission Control of Zoonosis, Department of Veterinary Science, National Institute of Infectious Diseases, Japan. Serial tissue sections at 5 μm were placed on coated slides, deparaffinized, rehydrated in ethanol successively from 100% to 70%, and washed in deionized water (DW). After treatment with 100 μg/mL proteinase K (AppliChem, St. Louis, Missouri, USA), the tissue slides were washed in Tris-buffered saline Tween-20 (TBST) buffer and incubated in 3% H₂O₂ to quench endogenous peroxidase. After rinsing with DW and washing in TBST buffer, the tissue slides were incubated with the PowerBlock reagent (BioGenex universal blocking reagent). The slides were washed and rinsed again and were incubated with the anti-RABV glycoprotein IgG2a monoclonal antibody at 1:500 dilution or with the anti-aCVSr nucleoprotein polyclonal antibody at 1:5000 dilution, washed in TBST buffer, incubated with the SuperEnhancer (BioGenex HRP kit), washed again, and drained. The slides were then incubated with the Polymer-HRP reagent (BioGenex HRP kit), rinsed with TBST buffer, and drained. Following treatment with chromogen-AEC [aminoethyl carbazole] (BioGenex HRP kit), the slides were rinsed with TBST buffer, counterstained with Mayer’s hematoxylin, rinsed with DW, and covered with coverslips using aqueous mounting media. Positive and negative controls were run parallel in each assay. The positive control was an IHC staining–positive brain-tissue block from a rabid bovine case, kindly provided by Dr. Chuen B. Hong at the University of Kentucky (Lexington, Kentucky, USA). Negative controls included corresponding tissue sections from RABV-negative TWFBs and substituting the monoclonal antibody with phosphate-buffered saline (PBS). Samples were considered positive for RABV only when distinct red-brown, finely to clumped granular cytoplasmic deposits were revealed in the samples and the positive control but not in the negative controls.

For the fluorescent antibody test (FAT), brain smears prepared with a frozen, ground, composite brain tissue suspension of cerebrum, hippocampus, thalamus, and hypothalamus left from a preparation for RNA extraction from each of TWFBs 1–3 were air-dried and fixed in acetone (Merck, Darmstadt, Germany). Following three PBS washes and air-dry sequences, direct (DFA) or indirect (IFA) immunofluorescence assay was performed. For DFA, the smears were incubated with a commercially available fluorescein isothiocyanate (FITC)–conjugated anti-rabies monoclonal antibody (Fujirebio Diagnostics, Malvern, Pennsylvania, USA) with three PBS washes. For IFA, the smears were incubated with the primary anti-RABV glycoprotein IgG2a monoclonal antibody (Abcam) at 1:500 dilution with three PBS washes and stained with an FITC-conjugated goat anti-mouse IgG antibody (Bethyl Laboratories, Montgomery, Texas, USA) with three PBS washes. The stained smears, with either DFA or IFA, were then counterstained with Hoechst stain at 1:200 dilution (Sigma-Aldrich, St. Louis, Missouri, USA) followed by immediate PBS wash. Positive-viral antigens appeared as apple-green fluorescent, finely to clumped granular aggregates under the fluorescence microscope (Optiphoto II, Nikon, Tokyo, Japan).

The histopathologic changes and the results of IHC and immunofluorescent staining were scored in a semiquantitative fashion according to the severity and distribution of the lesion as well as to the intensity and frequency of the positive signals of a specific staining. The score was determined based on the percentage of the entire section of a particular tissue with lesion involvement and the percentage of neurons in the entire section of a particular tissue or smear showing positive staining. The scoring system consisted of four categories: −, no pathologic changes or no positive staining; +, <25% of the entire section of a particular tissue containing mild to moderate pathologic changes or <25% of the neurons in the entire section of a particular tissue or in the entire smear containing weak to moderate-positive staining; ++, 25–50% of the entire section of a particular tissue containing mild to severe pathologic changes or 25–50% of the neurons in the entire section of a particular tissue or in the entire smear containing weak to strong positive staining; and +++, >50% of the entire section of a particular tissue containing severe pathologic changes or >50% of the neurons in the entire section of a particular tissue or in the entire smear containing strong positive staining.

For fresh brain, 25 mg of tissue were mixed with 1 mL of TRIzol reagent (Invitrogen, Carlsbad, California, USA) and homogenized. After adding 0.2 mL of chloroform and being centrifuged at 12,000 × G for 15 min, the upper aqueous phase was collected for total RNA extraction by using RNeasy Mini Kit (Qiagen, Valencia, California, USA) followed by cDNA synthesis with the Transcriptor first strand cDNA [complementary DNA] synthesis kit (Roche Diagnostics, Indianapolis, Indiana, USA) according to the manufacturer’s instructions. The RABV nucleoprotein (N) and glycoprotein (G) genes were then amplified by PCR using the primer sets RV-N-F (1–25): 5′-ACGCTTAACAACAAAACCATAGAAG-3′/RV-N-R (1515–1538): 5′-CGGATTGACGAAGATCTTGCTCAT-3′ and RV-G-F (3291–3315): 5′-CATCCCTCAAAAGACTTAAGGAAAG-3′/RV-G-R (4918–4941): 5′-CCGAGGAGATGAGGTCTTCGGGAC-3′, respectively, as described (Liu et al. 2010); the amplicon sizes were 1,300 and 500 base pairs (bp), respectively. For paraffin-embedded brain tissue, 10 pieces of 10-μm sections from each of the six archival cases and TWFBs 1–3 were deparaffinized with xylene. After centrifugation, the cell pellets were collected for RNA extraction by using the RNeasy FFPE kit (Qiagen) and subsequent cDNA synthesis according to the manufacturer’s instructions. The RABV N gene was then amplified by reverse transcription (RT)-PCR using two primer sets P4(509): 5′-GAGAAGGAACT(C/T)CA(G/T)GAGTA-3′/P4(110BT): 5′-(G/T)TTCACATGTTCGAGTAT-3′ and S4(CN12): 5′-AATCTCACCGC GAGAGAGG-3′/S4(CN13): 5′-GTGGCATTAA GAGACCTGAC-3′ (Wacharapluesadee et al. 2006); the amplicon sizes were 150 and 236 bp, respectively. The amplicons were further sequenced (Tri-I Biotech, Taipei, Taiwan) and compared with those viral sequences deposited in GenBank/EMBL/DDBJ (National Center for Biotechnology Information 2015).

Aside from poor nutritional condition (TWFB 1), moderate lung edema (TWFB 3), and congested leptomeninges (TWFBs 1–3), no other significant changes were observed.

In the nervous tissue, nonsuppurative meningoencephalomyelitis and ganglionitis with formation of Negri bodies were seen in TWFBs 1–3 and in archival TWFB 4 and 5 (Table 1). The lesions consisted of multifocal, mild-to-severe infiltration of lymphocytes and plasma cells mixed with some macrophages, eosinophils, and neutrophils in the leptomeninges and Virchow-Robin spaces of the gray matter (Fig. 1A) and part of the white matter throughout the brain and C1/C2 spinal cord. Except for cerebellum, variable gliosis was present throughout the neuropil (Fig. 1A). Many neurons were undergoing degeneration and necrosis, characterized by variable cytoplasmic hypereosinophilia, atrophy and vacuolization, karyorrhexis, and loss of cell integrity. Segmental loss of Purkinje cells was noted in the cerebellum. Varying numbers of round, oval, oblong-to-irregular, discrete-to-amorphous, eosinophilic Negri bodies were readily detected in the cytoplasm of the neuronal bodies and nervous processes throughout the brain (Fig. 1B), C1/C2 spinal cord, and ganglia (Tables 1, 2 and Fig. 1C).

Spongiform degeneration, characterized by focal to multifocal, variable neuronal perikaryon and neuropil vacuolization, was noted in the cerebral cortex, thalamus, brainstem, and junction of the granular and Purkinje cell layers of cerebellum of TWFBs 1–3 (Fig. 1D). The vacuoles were round, oval-to-polygonal, and variable in size; they were empty or contained some amorphous membranous to granular structure and proteinaceous substance. In TWFBs 4 and 5, postmortem autolysis precluded satisfactory evaluation of the spongiform degeneration.

Mild to severe ganglionitis, characterized by lymphocytic infiltration mixed with other inflammatory cells, degeneration, and necrosis of ganglion cells similar to those seen in the neurons, and formation of Negri bodies was observed in the periadrenal, perirenal, and myenteric plexus ganglia (Fig. 1C). Areas of necrosis accompanied by infiltration of mixed inflammatory cells were seen in the adrenal cortex (Fig. 1E, F) and medulla of TWFBs 1–3. In addition, the submandibular salivary gland of TWFB 3 had scattered mild lymphocyte infiltration in the interstitium.

Apparent, but variable, IHC-positive signals for RABV antigens were revealed in TWFBs 1–3 (Fig. 2A) and TWFBs 4 and 5, in which the earliest positive case, TWFB 4, occurred in 2004 (Fig. 2B). There were no apparent differences in the distribution, intensity, and frequency of positive signals by using the mouse anti-RABV glycoprotein IgG2a monoclonal antibody or rabbit anti-aCVSr nucleoprotein polyclonal antibody (data not shown). Positive signals were observed almost exclusively in the gray matter of cerebrum and cerebellum and in the hippocampus, thalamus, hypothalamus, brainstem, and gray matter of the C1/C2 spinal cord (Table 1 and Fig. 2A, B). The positive signals were characterized as sparsely to widely distributed, finely to clumped granules throughout the perikaryon of the neurons and axonal and dendritic processes. Similar positive signals were detected in the ganglion cells of periadrenal, perirenal (Fig. 2C), sublingual, and myenteric plexus ganglia, skin, trachea, bronchi, heart, and pancreas; nerve bundles of the submucosa and tunica muscularis of gastrointestinal (GI) tract; acinar epithelial cells and ganglion cells of the submandibular salivary gland; the remaining parenchymal cells and infiltrating macrophages of the adrenal necrotic regions (Fig. 1F, inset); and macrophages (lysozyme positive, toluidine-blue negative) randomly scattered in the mucosa, submucosa, and muscular layer of the tongue and GI track, mesenteric adipose tissue, dermis, urinary bladder, heart, and Kupffer cells (Table 2 and Fig. 2D, E).

By DFA or IFA, apparent immunofluorescent positive signals were readily observed in the brain smears of TWFBs 1–3. Strongly positive, finely-to-clumped granular-to-globular signals were seen in the neuronal perikaryon, nervous processes, and amorphous tissue debris (Table 1 and Fig. 2F).

By using fresh brain tissue, the results of RT-PCR (Fig. 3) and subsequent sequence analysis of the N and G genes confirmed that TWFBs 1–3 were RABV infected. By using paraffin-embedded brain tissues, the RT-PCR also demonstrated specific amplicons of the N gene with the anticipated sizes of 150 and 236 bp in TWFBs 1–3 but failed to demonstrate any detectable amplicons in the six archival cases that were run in parallel (data not shown). The strain of RABV of TWFBs (RABV-TWFB) belongs to Lyssavirus genotype 1, and the nucleotide identity of the N gene within the three isolates of TWFB 1–3 was 97–99%. The N gene of the three RABV-TWFB isolates was 89–90% identical to the CNFB isolates (F02, F04, JX08-45, JX08-48, JX09-18), 91% to the dog-related RABV within the China I lineage (HN10, GD-SH-01, CTN-1, CTN-181), 88% to the Southeast Asia dog isolate (QS-05), 88% to the dog-related RABV isolates within the China II lineage (JX09-17, SH06), 86–87% to the cosmopolitan isolates (serotype 1, FluryLEP, DRV-NG11, 9147FRA, SAD B19), 87% to the golden palm civet (Paradoxurus zeylonensis) and human Sri Lanka isolates (H-08-1320, H-1413-09), and 84% to the North American bat (Lasionycteris noctivagans)-related RABV (SHBRV-18).

After China, Taiwan is the second region in the world in which FB-associated rabies has been diagnosed. The TWFB-associated rabies was diagnosed based on 1) characteristic pathologic findings of nonsuppurative meningoencephalomyelitis, ganglionitis, and Negri bodies; 2) positive IHC staining and DFA/IFA tests for RABV antigens; and 3) positive viral nucleic acid detection by RT-PCR followed by genomic sequencing (GenBank accession numbers KF620487–KF620489; Chiou et al. 2014) in TWFBs 1–3. Recent phylogeographic data indicate that RABV-TWFB is a distinct lineage within the Asian group and differentiated from the RABV of CNFBs 158–210 yr ago, with the most recent common ancestor of RABV-TWFB originating 91–113 yr ago; and the data suggest that RABV-TWFB could have been cryptically circulating without being recognized for a long time (Chiou et al. 2014). Our retrospective work tracing the occurrence of TWFB-associated rabies back >10 yr, further supports this speculation. However, the postmortem autolysis and limited paraffin-blocks of archival brain tissues preclude satisfactory morphologic and molecular evaluation. Following the diagnosis of rabies in TWFBs 1–3, rabies has been diagnosed by FAT in another 459 TWFBs, five gem-faced civets (Paguma larvata), one shrew (Suncus murinus), and one FB-bitten puppy as of 7 May 2015 (Animal Health Research Institute, Council of Agriculture, Executive Yuan, Taiwan, Republic of China, unpubl. data).

In rabid TWFBs, lesions and RABV antigens were widely distributed in the brainstem, cerebrum, cerebellum, anterior cervical spinal cord, periadrenal, perirenal, and myenteric plexus ganglia, and adrenal gland. Similar extensive lesions and RABV antigen distribution in the central nervous system (CNS) has also been seen in naturally acquired rabid raccoons and skunks (Stein et al. 2010; Hamir 2011). However, hippocampus, brainstem/cerebellum, and cervical spinal cord/adjacent brainstem were the best site for RABV antigen detection in dogs/cats, cattle (Bos taurus), and horses (Equus caballus), respectively (Stein et al. 2010). In carnivores and herbivores/humans, hippocampus and Purkinje cells of cerebellum are the common sites for finding Negri bodies, respectively (Jackson et al. 2001; Maxie and Youssef 2007). However, TWFBs, similar to raccoons, have widely distributed, large numbers of Negri bodies readily observed in both central and peripheral neural tissues, although brainstem has a better detection rate.

The formation of Negri bodies in the neural tissue of a rabid animal seems related to the type of viral isolate to which the animal was exposed. Raccoons infected with the raccoon isolate developed numerous neuronal Negri bodies in the CNS and ganglion cells of non-CNS tissues; however, dog or bat isolates resulted in the formation of no, or only occasional, Negri bodies in raccoons (Hamir 2011). Suspicious natural canine isolate-induced rabies cases in CNFBs have been reported in China (Zhang et al. 2009). If the RABV strain-dependent phenomenon observed in raccoons also occurs in FBs, the presence of many Negri bodies in the neural tissue of the rabid TWFBs, along with the results of nucleotide identity of the N gene between RABV-TWFB isolates and other known RABV isolates, may further support that RABV-TWFB as a distinct lineage that has been circulating in the TWFB population for a significant period, instead of a recent spillover from a canine or bat source. However, this assumption requires further clarification.

In some animal species, care must be taken to differentiate Negri bodies from nonspecific inclusions, such as Hirano, Pick, Lewy, Lafora, Bunina, and motor neuron disease inclusion bodies. These have been found in the pyramidal cells of the hippocampus or the lateral geniculate nucleus neurons in cats, skunks, dogs, and horses and in the larger neurons of the medulla and spinal cord of old sheep (Ovis aries) and cattle (Maxie and Youssef 2007). Because TWFBs are susceptible to Canine distemper virus (CDV) infection (Chen et al. 2008), intracytoplasmic CDV inclusion bodies should also be considered as a possible differentiation. However, aside from nonsuppurative meningoencephalitis and neuronal intracytoplasmic inclusions, as seen in rabid TWFBs, CDV-infected TWFBs may also have neuronal or glial intranuclear inclusions (Chen et al. 2008). Based on the results of pathologic findings and RABV antigen detection, brainstem seems ideal for sampling and diagnosing FB-associated RABV infection, although cerebral cortex, hippocampus, thalamus, and hypothalamus are also suitable.

The spongiform change seen in the brain of rabid TWFBs is similar to that reported in RABV-infected skunks and foxes (Charlton 1984; Charlton et al. 1987) and a heifer (Bos taurus; Foley and Zachary 1995). The pathogenesis of such change in the rabid animals is unclear. Electron-microscopic findings suggest that the lesion development starts gradationally from small to large, membrane-bound vacuoles in the neuronal processes, primarily dendrites, with no need of incorporation of viral components (Charlton et al. 1987).

The changes seen in the submandibular salivary gland of the rabid TWFB 3 are similar to those seen in the parotid salivary gland of experimentally RABV-infected skunks and foxes, characterized as an interstitial sialadenitis (Balachandran and Charlton 1994). They also share a similar scattered distribution pattern of RABV antigens in the ganglia, nerve bundles, and acinar epithelial cells (Balachandran and Charlton 1994). The positive result of the RT-PCR of a saliva swab from TWFB 3 further demonstrated that RABV nucleic acid was also produced and released (H.-Y.C. unpubl. data).

To our knowledge, adrenal necrosis has not been described as a rabies-associated histopathologic change in rabid animals, although Negri bodies have been found in the ganglion cells of the adrenal medulla (Hamir et al. 1992). Scattered to significant amounts of IHC staining–positive signals for RABV antigens were seen in the remaining parenchymal cells and some macrophages in the necrotic areas of both cortex and medulla but not in the neuron or nerve fibers within the parenchyma. Although RABV antigens were detected in the remaining adrenal parenchymal cells in the necrotic regions, whether the necrosis is directly induced by RABV requires further elucidation. Aside from the adrenal gland, sporadic IHC staining–positive macrophages were seen in a variety of tissues (Table 2). Although macrophages are less likely to support RABV production, they can be a source of infectious virus upon transfer in mice (Mus musculus, C57BL/6; Naze et al. 2013). This may explain why we detected IHC-positive RABV antigens in the macrophages widely distributed in various tissues.

The accumulated data show that FB-associated RABV infection in Taiwan occurs mainly in TWFBs, although sporadic spillover to other species occurs. The gradual increase in the population of TWFBs in the past 10 yr may have had a role in the recent outbreak; however, whether there are other cofactors or other causes is not known and requires further investigation. The origin of the TWFB-associated RABV remains unanswered. The earliest written record of rabies in Taiwan was in the early 19th century, when Taiwan was under Japanese colonization. Human-mediated animal translocation and animal migration are thought to have had a major role in the dispersal of RABV (Fèvre et al. 2006; Bourhy et al. 2008). The domestic dog has been suggested to serve as the main source of interspecies RABV transmission through which viral lineages are generated and spread to other taxa (Bourhy et al. 2008). Aside from being ruled by the Chinese, the island of Taiwan has been colonized by the Spanish, Dutch, and Japanese for various periods. Thus, it is reasonable to speculate that human-mediated animal (especially dog) translocation was one of the most likely sources of RABV in TWFBs.

These reported pathologic changes and viral antigen distribution in FBs naturally infected with the RABV-TWFB strain may provide a basis for future studies, such as cross-species disease spread and vaccine development. The diagnosis of rabies in TWFBs has clearly demonstrated that known or unknown diseases may circulate cryptically in free-ranging wildlife sharing habitat with humans. These diseases may have serious effects on human and other animal health, on wildlife conservation, and on the economy. We are reminded once again of the essentiality of systematic wildlife disease surveillance.

We thank Jen-Tzu Yang, Ying-Hui Wu, and staff of the Taiwan Endemic Species Research Institute, Council of Agriculture, for the collection and submission of carcasses of free-range wildlife. Special thanks to Tsung-Chou Chang, from National Pingtung University of Science and Technology, for his kind help in the study and in memory of his recent passing away. The study was supported in part by grants 102AS-10.1.1-BQ-B1(3), 103AS-10.1.1-BQ-B1(1), 104AS-10.1.2-BQ-B3(1), MOST 103-3114-Y-518-001, and MOST 104-3114-Y-518-002 from the Bureau of Animal and Plant Health Inspection and Quarantine, Council of Agriculture, Executive Yuan, Taiwan.