CORD-19:50f0a543bfb1a1e2da016d37676d18f2f45dad13 JSONTXT 7 Projects

Dual enteric and respiratory tropisms of winter dysentery bovine coronavirus in calves Abstract Although winter dysentery (WD), which is caused by the bovine coronavirus (BCoV) is characterized by the sudden onset of diarrhea in many adult cattle in a herd, the pathogenesis of the WD-BCoV is not completely understood. In this study, colostrumdeprived calves were experimentally infected with a Korean WD-BCoV strain and examined for viremia, enteric and nasal virus shedding as well as for viral antigen expression and virus-associated lesions in the small and large intestines and the upper and lower respiratory tract from 1 to 8 days after an oral infection. The WD-BCoV-inoculated calves showed gradual villous atrophy in the small intestine and a gradual increase in the crypt depth of the large intestine. The WD-BCoV-infected animals showed epithelial damage in nasal turbinates, trachea and lungs, and interstitial pneumonia. The WD-BCoV antigen was detected in the epithelium of the small and large intestines, nasal turbinates, trachea and lungs. WD-BCoV RNA was detected in the serum from post-inoculation day 3. These results show that the WD-BCoV has dual tropism and induces pathological changes in both the digestive and respiratory tracts of calves. To our knowledge, this is the first detailed report of dual enteric and respiratory tropisms of WD-BCoV in calves. Comprehensive studies of the dual tissue pathogenesis of the BCoV might contribute to an increased understanding of similar pneumoenteric CoV infections in humans. The coronaviruses (CoV) constitute a genus in the family Coronaviridae, large enveloped viruses ranging in diameter from 100 to 120 nm, that possess a single-stranded, positive-sense RNA genome of 27-32 kb in length and cause important diseases in animals [23] . Coronaviruses (CoVs) can be divided into three distinct groups according to their natural hosts, nucleotide sequences, and serologic relationships [23] . The bovine CoV (BCoV), which belongs to the CoV group 2a, contains five major structural proteins: the nucleocapsid (N), the transmembrane, the spike (S), the small membrane and the hemagglutinin=esterase (HE) proteins, the latter being unique to some group 2 CoVs [23] . Although both S and HE glycoproteins of the BCoV use Nacetyl-9-O acetyl neuraminic acid as a receptor determinant to initiate an infection, the S glycoprotein requires a smaller amount of these receptors on the surface of erythrocytes for agglutination than the HE protein [32, 33] . Therefore, the S glycoprotein is the major hemagglutinin of BCoV [32] and is believed to be responsible for the primary attachment of BCoV to other cell surface receptors [33] . Consequently, the variation in the host range and tissue tropism of CoVs may largely be due to variations in the S gycoprotein [15] . Bovine coronaviruses cause three different clinical diseases: severe calf diarrhea (CD), winter dysentery (WD) in adult cattle, and respiratory tract infections in calves and feedlot cattle [7, 24, 25, 31] . In addition, the virus has been associated with a new proposed clinical form, warmer season dysentery in adult cattle [27] . Because the porcine respiratory CoV emerged from porcine transmissible gastroenteritis CoV (TGEV) through deletion mutations with reduced enteropathogenicity and a loss of hemagglutinating activity [29] but pronounced tropism for lung tissue, many attempts have been made to isolate and identify a potential deletion mutant of BCoV with altered tissue tropism or virulence. Despite their involvement in the different clinical forms, only minor variations in the S and HE genes among the strains associated with CD, WD, and bovine respiratory CoV (BRCoV) have been reported [9] . Moreover, only one serotype has been identified in the BCoV isolates or strains, even though the data obtained from serological studies, using monoclonal and polyclonal antibodies, suggested the existence of serologically different subtypes of BCoV [40] . Animals that have recovered from an infection with the BRCoV, CD, or WD strains were protected from BCoV-associated diarrhea after a challenge exposure to either a hetero-logous or homologous strain of the BCoV [6] , which suggests that only one BCoV serotype exists [40] . Winter dysentery is characterized by a sudden onset of often bloody diarrhea that spreads rapidly in adult cattle in a herd and has been reported in many parts of the world [30] . The WD-affected cattle lose body condition, resulting in a dramatic decrease in milk production, and many animals show signs of respiratory disease (nasal discharge) [30] . Recently, WD was successfully reproduced experimentally in BCoV-seronegative adult cattle and calves [39] . However, the digestive and respiratory pathogenesis of WD-BCoV in calves is so far unknown. This study attempted to answer the following three questions: 1) where in the digestive tract of calves does WD-BCoV replicate and cause pathological changes? 2) does WD-BCoV replicate and cause damage to the respiratory tract of calves? 3) if so, in which cells of the respiratory tract does WD-BCoV replicate and cause damage? Accordingly, the digestive tract (small and large intestine) and the respiratory tracts including the nasal turbinates, trachea and lungs, were sampled from colostrumdeprived (Cols-D) calves inoculated with a Korean WD strain, KWD3, and analyzed for BCoV antigens by histopathology and immunohistochemistry, as well as by a scanning electron microscopy (SEM). In addition, the level of fecal and nasal virus shedding (detected by RT-PCR, nested PCR and ELISA) was assessed. The results of this study suggest that WD-BCoV has dual tropism for the digestive and respiratory tracts of calves and induces pathological changes in these tissues by destroying the epithelium. The KWD3 strain of WD-BCoV was originally isolated and identified from a fecal sample from an adult cow with WD [18, 19] . The KWD3 strain was passaged 6 times in human rectal adenocarcinoma cells (HRT-18G; ATCC). The virus titer of the KWD3 strain was determined using a cell culture immunofluorescence (CCIF) assay with fluorescein-isothiocyanate-conjugated hyperimmune antisera to the Mebus strain of the BCoV, using a slight modification of a method described previously [41] . The mock-infected HRT-18G tissue culture supernatant was used for the mock inoculation of a control Cols-D calf Pathogenesis of WD-BCoV in calves through the oral route. The possibility that the RT-PCR or nested PCR can detect the residual BCoV inoculum after inoculation was tested by inoculating another Cols-D calf with the inactivated KWD3 strain. The KWD3 strain was inactivated by chloroform treatment, as described previously [6] . A total of eight 2-to 4-day-old Cols-D Holstein calves were used in this study. They were fed commercial calf formula. Six calves were inoculated orally with 40 ml of the tissue culture supernatant from the KWD3 strain containing a virus titer of 1.5Â 10 8 immunofluorescence-forming units=ml by the CCIF assay (Table 1 ). One calf was inoculated orally with 40 ml of the mock-infected HRT-18G tissue culture supernatant, and the remaining calf was inoculated with 40 ml of the inactivated KWD3 strain (Table 1) . After inoculation, the rectal temperatures, respiratory signs, and color and consistency of the feces from each calf were evaluated daily. The consistency of the feces was scored on a scale of 0-4, with 0 representing firm; 1, pasty; 2, semimucoid; 3, liquid and 4, profuse diarrhea. Nasal swabs and fecal samples were collected daily before and after inoculation, as described previously [13] . The inoculated calves were euthanized at post-inoculation day (PID) 1 (one calf), 3 (one calf), and 5 (one calf). One calf died at PID 5 and two calves died at PID 8. Blood samples were collected twice from each calf, prior to inoculation and immediately before euthanasia. Blood samples were collected from the heart and aorta of the three dead calves. The calves inoculated with the mock-and inactivated-KWD3 strain were euthanized at PID 2 and 4, respectively. A necropsy of all the calves was performed immediately after death. At the necropsy, the intestinal tracts were removed and opened, and the small and large intestinal contents were collected. Samples from the intestinal segments were collected and immersed immediately in fixatives for the histological (10% buffered formalin) and SEM (see below) examinations. The intestinal segments examined included the duodenum (approximately 15 cm caudal to the pyloric valve), jejunum (mid region of the small intestine), ileum (approximately 15 cm cranial to the ileocecal junction), cecum, colon, and rectum. Samples from the nasal turbinates, trachea, lungs, liver, spleen, heart, brain and kidneys were also collected and placed immediately in 10% buffered formalin or SEM fixatives (see below). The formalin-fixed, paraffin-embedded tissue sections from each organ and tissue were stained with Mayer's hematoxylin and eosin and examined microscopically. The histological evaluation was performed in a blind manner on coded samples, and a comparison was made with the tissues from the mock-infected control. Atrophy of the small intestinal villi and large intestinal glands was measured from histological sections of the duodenum, jejunum, ileum, cecum, colon and rectum. The small intestinal changes were scored according to the average villi= crypt (V=C) ratio plus the grade of epithelial cell desquamation, which was measured as follows: V=C ratio, 0 ¼ normal (V=C = 6:1), 1 ¼ mild (V=C ¼ 5.0 to 5.9:1), 2 ¼ moderate (V=C ¼ 4.0 to 4.9:1), 3 ¼ marked (V=C ¼ 3.0 to 3.9:1), 4 ¼ severe (V=C 5 3.0:1) and desquamation grade, 0 ¼ normal (no desquamation), 1 ¼ mild (cuboidal attenuation of tip villous epithelium), 2 ¼ moderate (desquamation of upper villous epithelium), 3 ¼ marked (desquamation of lower villous epithelium), 4 ¼ severe (desquamation of crypt epithelium). The large intestinal lesions were classified according to the grade of epithelial cell desquamation plus the grade of large intestinal gland atrophy as follows: desquamation grade, 0 ¼ normal (no desquamation), 1 ¼ mild (cuboidal attenuation of tip epithelium), 2 ¼ moderate (desquamation of tip epithelium), 3 ¼ marked (desquamation of upper middle epithelium), 4 ¼ severe (desquamation of lower middle epithelium). These changes were evaluated by measuring 10 randomly selected villi and large intestinal glands on the intestinal histological sections. This methodology is similar to that reported elsewhere [14] . Lesions in the nasal turbinates and trachea were classified according to the degree of degeneration (cell swelling and loss of cilia) and necrosis (characterized by pyknosis, karyorrhexis, karyolysis and the absence of nuclei associated with cell swelling and loss of cilia) of the epithelium as follows: 0 ¼ normal, 1 ¼ one to two degenerating cells, 2 ¼ three to five degenerating and necrotic cells, 3 ¼ many degenerating and necrotic cells. The lung lesions were scored according to the extent of interstitial pneumonia caused by the infiltration of macrophages and lymphoid cells in the interstitium, the accumulation of hyperplastic type II pneumonocytes, and the degree of the degeneration, necrosis and desquamation of the alveolar epithelial cells. The lesions were classified as follows: 0 ¼ normal, 1 ¼ scattered mild thickening of interstitium, and degeneration and necrosis of two or three, and desquamation of four to six alveolar epithelial cells, 2 ¼ multifocal moderate thickening of interstitium, and degeneration and necrosis of four to six, and desquamation of many alveolar epithelial cells, 3 ¼ diffuse severe thickening of the interstitium as well as degeneration, necrosis and desquamation of many alveolar epithelial cells. The distribution of the BCoV antigens in the tissues was evaluated through an immunohistochemical examination of the paraffin-embedded sections of each segment of the small and large intestines, nasal turbinates, trachea, and lungs, using a slight modification of the methodology described previously [8] . Three monoclonal antibodies [Mabs; BC C against the S protein (0.32 mg=ml), and BC 22 F8.3 C against the HE protein (0.2 mg=ml)] of the DB2 strain of BCoV, which were originally used for ELISA and Western blotting [36] , were used as the primary Mabs. These Mabs were diluted 1:200 with phosphate-buffered saline, pH 7.4, and applied to each section. The antigen distribution in the small and large intestines was evaluated based on the number of antigen-positive cells in the villi and large intestinal glands and was measured as follows: 0 ¼ no positive cells, 1 ¼ one to two positive epithelial cells in the tip or upper part of the villi and large intestinal glands, 2 ¼ three to five positive epithelial cells scattered in the tip or upper part of the villi and large intestinal glands, 3 ¼ many positive epithelial cells in the tip or upper part of the villi and large intestinal glands, 4 ¼ positive in almost all epithelial cells in the tip and upper part of the villi and large intestinal glands. The antigen distribution in the nasal turbinates, trachea and lungs was examined using a score based on the number of antigen-positive cells, and was measured as follows: 0 ¼ no antigen-positive cells, 1 ¼ two to three antigen-positive epithelial cells scattered in the epithelium of the nasal turbinates, trachea, and in the epithelium of the alveoli, bronchioles and bronchus of the lungs, 2 ¼ four to six antigen-positive epithelial cells scattered in the epithelium of nasal turbinates, trachea, and in the epithelium of the alveoli, bronchioles and bronchus of the lungs, and 3 ¼ many antigenpositive epithelial cells scattered in the epithelium of the nasal turbinates, trachea, and in the epithelium of alveoli, bronchioles and bronchus of lungs. The number of antigen-positive cells in the organs or tissues was determined by light microscopy in 10 fields per section at a magnification of 400Â . Segments of the duodenum, jejunum, ileum, colon, nasal turbinates and trachea were fixed in a mixture containing 3% glutaraldehyde, 2% paraformaldehyde and 1.5% acrolein in 0.1 M collidine buffer, pH 7.3, as described previously [14] . The specimens were dehydrated in an ethanol-dry ice series and gently vacuum dried. The dried tissues were sputter coated with approximately 150 Å of platinum, and were examined using a scanning electron microscope (JEOL, Tokyo, Japan). Detection of WD-BCoV in fecal, nasal swab and serum specimens 1) Detection of BCoV antigen by ELISA. Indirect antigencapture ELISA using the anti-BCoV Mabs employed for immunohistochemistry (see above) as the capture antibodies was used to detect the BCoV in the nasal swab fluids and fecal suspensions, as described elsewhere [36] . 2) Detection of WD-BCoV by IFA. The nasal epithelial cells were collected from the nasal swabs and processed for direct IF staining, as described elsewhere [35] . 3) RT-PCR and nested PCR. RNA was extracted and purified from the fecal, nasal swab and serum specimens of the inoculated Cols-D calves using the techniques described elsewhere [6] . RT-PCR and nested PCR assays with the primer pairs specific for the BCoV N gene were performed as described elsewhere [6] . The KWD3 strain caused severe diarrhea in all the inoculated calves by PID 1, which persisted until euthanasia or the death of the animal (Table 1) . Diarrhea was not observed in the inactivated KWD3-inoculated or the mock-inoculated calves. The calves inoculated with the KWD3 strain showed elevated body temperatures but the calves inoculated with the inactivated KWD3 or mock inoculum did not. None of the calves inoculated with the KWD3 strain showed any respiratory signs. Table 1 shows the data of fecal and nasal virus shedding in Cols-D calves inoculated with the KWD3-BCoV. ELISA detected the BCoV antigen in the fecal samples of the inoculated calves on PID 3-8. BCoV RNA was similarly detected by RT-PCR and nested PCR in the fecal samples on PID 1-8. IFA detected the nasal shedding of the BCoV in 3 out of 6 virus-inoculated calves at PID 3-4 for 1 or 2 days. ELISA revealed nasal shedding of the BCoV in 4 out of 6 virus-inoculated calves at PID 2-3 for 2-6 days. RT-PCR revealed nasal shedding of the BCoV in 5 out of 6 virus-inoculated calves at PID 3 for 1-6 days. Nested PCR detected nasal shedding of the BCoV in 5 out of 6 virus-inoculated calves at PID 2 for 2-7 days. Fecal and nasal virus shedding was not detected in the inactivated BCoV and mock-inoculated calves by the above-described methods. Histological findings (i) Intestinal tract. In mock-or inactivated-BCoVinoculated calves, no changes were observed. The small intestines exhibited long and slender villi with short crypts (Fig. 1A and B) . The large intestines contained intestinal glands, which were elongated, straight and open to the luminal surface (Fig. 1C ). In contrast, the BCoVinoculated calves showed mild to severe villous atrophy, widespread villous fusion and increased crypt depth in the small intestine ( Fig. 1D and E) , and moderate atrophy of the intestinal glands with severe epithelial desquamation in the large intestine (Fig. 1F ) in the multiple cross sections. Table 2 summarizes the histological changes in the small and large intestines from the virusinoculated calves. At PID 1, mild villous atrophy caused by necrosis and the desquamation of villous epithelial cells was confined in the duodenum. With time, the duodenal lesions became increasingly severe, and other parts of the intestine showed milder epithelial changes. Eventually, the duodenum showed severe villous atrophy and the other parts of the intestines including the jejunum, ileum, cecum, colon and rectum had moderate to marked mucosal lesions at PID 8 ( Fig. 1D-F) . (ii) Respiratory tract. The calves inoculated with mock-inactivated BCoV did not exhibit any changes. They had pseudostratified ciliated columnar epithelium with goblet cells in the nasal turbinates, trachea and bronchus ( Fig. 2A) . The large bronchioles were lined with ciliated d Blood samples were collected twice from each calf, prior to inoculation and immediately before euthanasia. Blood samples were collected from the heart and aorta from three dead calves (calf numbers 4-6). Pathogenesis of WD-BCoV in calves columnar cells, whereas the smallest (terminal) bronchioles were lined by ciliated cuboidal cells and, distally, with noncilated cells. The respiratory bronchioles were lined by a cuboidal epithelium, which became flattened distally. The alveoli were lined by flat alveolar epithelial cells (Fig. 2B) . The changes in the nasal turbinates and trachea of the virus-inoculated calves were represented by multifocal epithelial cell degeneration (cell swelling and loss of cilia) and necrosis of a few epithelial cells (Table 3) . At PID 1, no lesions were observed in the upper respiratory tract. However, at PID 3, few epithelial cells showing swelling with patch loss of cilia in the nasal turbinates and trachea were evident. Epithelial degeneration and occasional necrosis was detected in multifocal areas from PID 5 to 8 (Fig. 2C) . A few lymphoid cells infiltrated the submucosa in some of the virus-inoculated calves. The lungs from virus-infected calves showed no changes at PID 1 (Table 3) . At PID 3, necrosis and sloughing of the alveolar lining cells was observed, resulting in the denudation of basal lamina (Fig. 2D) . There was patch hyperplasia of type II pneumocytes and mild interstitial infiltration of macrophages. The alveolar sacs contained some foamy macrophages and desquamated alveolar lining cells, as well as some fibrin clots and eosinophilic proteinaceous fluid. At PID 5 and 8, there were an intense infiltration of macrophages with a few lymphoid cells in the alveolar septa and lower numbers of hyperplastic type II pneumocytes in the alveolar wall, leading to interstitial pneumonia ( Fig. 2E and F ). The alveolar cavities had a few foamy macrophages and desquamated alveolar lining cells, as well as some fibrin and eosinophilic proteinaceous fluid. The epithelia of the alveoli, bronchi and bronchioles were often desquamated or necrotic. These changes are listed in Table 3 . Fig. 3A and B, Table 2 ). There was a correlation between BCoV antigen expression and the histological lesions. At PID 1, the BCoV antigen was detected in the epithelium of the duodenum and jejunum, but no histological changes were observed in the jejunum ( Table 2) . At PID 3, there was a significant increase in the number of antigen-positive cells in the small intestines, which began to appear in some surface epithelia and three to five macrophages of the large intestinal glands (Table 2) . At PID 5, the number of antigen-positive cells decreased in the villi of the small intestine but Table 2 ). (ii) Respiratory tract. Staining of BCoV antigens was detected in the cytoplasm of the degenerated and necrotic epithelial cells of the nasal turbinates, trachea and lungs ( Fig. 3C and D, Table 3 ). No BCoV antigen-positive cells were detected in these tissues at PID 1. From PID 3 on, the number increased gradually until PID 8. There was a consistently higher number of BCoV antigen-positive cells in the lungs than in the nasal turbinates and trachea. In the lungs, the positive cells included the epithelium of alveoli, bronchi and bronchioles (Fig. 3D ). Viral antigens were not detected in the respiratory tract of the inactivated-KWD3-strain-and mock-inoculated calves using the SAB-PO method. RT-PCR detected BCoV RNA in the serum samples from two calves at PID 8 (Table 3) . Nested PCR amplified BCoV RNA in the serum samples of all of the inoculated calves at PID 3, 5, and 8 (Table 3) . RT-PCR and nested PCR did not detect BCoV RNA in the serum of the inactivated-KWD3-strain-and mock-inoculated calves (Table 3) . The inactivated-KWD3-strain-and mock-inoculated calves exhibited no changes. Long, finger-like small intestinal villi with circular transverse grooves on their surface, and a uniform and densely packed microvillous coat were seen ( Fig. 4A and B) . The large intestinal epithelium was ridged, with a uniform and densely packed microvillus coat. In the calves inoculated with the KWD3 strain, denuding and shortening of the small intestinal villi and large intestinal glands were observed, and the microvillus coat was sloughed. At PID 1, in the duodenum patch, exfoliation of the microvilli and epithelial desquamations was observed, with a slight shortening of the villi (Fig. 4C and D) . The severity of these morphological changes increased with time. At PID 3 and 5, stunted small intestinal villi and large intestinal glands were fused, and the villi appeared conical or leaf-shaped. At PID 8, some severely stunted villi appeared which exhibited complete exfoliation of microvilli ( Fig. 4E and F) . The inactivated-KWD3-strain-and mock-inoculated calves exhibited no respiratory epithelial changes. They showed cilia that completely covered the luminal surface of nasal turbinates and trachea ( Fig. 5A and B) . However, nasal turbinates and trachea samples from the calves inoculated with the KWD3 strain initially showed patchy exfoliation of the cilia, with complete exfoliation of cilia occurring later ( Fig. 5C and D) . It should be noted that the WD-BCoV-induced small and large intestinal lesions were different from those caused by TGEV, which causes only small intestinal lesions [29] . Although BCoV is believed to replicate in the epithelium of the respiratory tract [5] [6] [7] 31 ], the precise regions or cells where BCoV replicates are still unknown. This study demonstrated that WD-BCoV has tropism for the epithelium of the nasal turbinates, trachea and lungs. Moreover, mild to severe interstitial pneumonia was detected in the BCoV-inoculated calves, suggesting that the WD-BCoV induces interstitial pneumonia. Some pneumotropic viruses, e.g. feline CoV and canine distemper virus, cause necrosis of the alveolar epithelium, which can lead to compensatory hyperplasia of the type II pneumocytes as well as the infiltration of macrophages and lymphoid cells into the interstitium, resulting in interstitial pneumonia [10, 21] . Since this study showed that WD-BCoV replicated and damaged the alveolar epithelium, the interstitial pneumonia induced by WD-BCoV might occur through a similar mechanism. However, unlike SARS-CoV, WD-BCoV showed stronger enterotropism than respiratory tro-pism [34] . Moreover, WD-BCoV caused dual lesions in the small and large intestines, even though the lesions in small intestine were more severe than those of the large intestine. Phylogenetically, the Korean WD-BCoV strains, including the KWD3, are closely related to both respiratory and enteric strains among the other known BCoVs studied in paired comparisons [18] . Moreover, the recently isolated American BCoVs clustered into the same branch with other recently isolated respiratory strains and were distinct from the ancestral enteric Mebus strain [16] . This means that the BCoV strains might be diverging from an enteric tropism to a dual (respiratory and enteric) tropism over time through intermediates [16, 18] . This study demonstrated that WD-BCoV exhibited both enteric and respiratory tropisms. Previously, it was not known if BCoV induced viremia. In this study, WD-BCoV RNA was detected transiently (PID 3) in the serum samples of calves showing that an oral infection leads to viremia. However, evidence for BCoV replication was first detected in the epithelial cells of the small intestine of a calf at PID 1, but not in the respiratory tract, where evidence of virus replication was first detected at PID 3. The onset and duration of viremia was parallel to nasal virus shedding and the presence of viral antigen and associated lesions in the respiratory tract. These results suggest that the cell-free viremia originated after the intestinal damage, which may have led to the development of respiratory infection and lesions after viremia [2] . Many viruses induce viremia during which the viruses circulate in the blood or blood cells and can spread to the target organs to initiate an infection [34] . However, it is unclear how the BCoV reaches the circulatory system from intestinal lesions after oral inoculation. One possible path is that cell-free BCoV penetrates the gut barrier from the luminal side after the destruction of enterocytes in the villi, which would expose the basement membrane, followed by transit into the circulatory system, which can spread to the nasal turbinates, trachea and lungs with replication in the epithelium [2] . Although monocyte-associated BCoV viremia was not examined in this study, antigen-positive macrophages for BCoV were detected in the lamina propria of the large intestinal glands at PID 3. Recent studies have shown that a feline CoV infection leads to the systemic viral spread through monocyte-associated viremia [21] . Therefore, it is possible that BCoV infections of the respiratory tract may also occur through monocyte-associated viremia, which may originate from the intestines. Further studies will be needed to determine which of these pathways causes viremia. Another possibility is that a minor aerosol infection might have occurred during oral inoculation or that the virus within the feces of infected animals was inhaled at PID 1, which might have caused virus replication, but with delayed kinetics. Indeed, intravenous BCoV inoculation results in enteric and nasal virus shedding [Cho and Saif, unpublished data] . This suggests that whatever the cause of the viremia, the viruses in the blood can reach and cause lesions in the epithelial cells in both the intestinal and respiratory tracts. This supports our hypothesis that cell-free viremia from the intestinal tract may causes respiratory tract lesions. Many viruses that are passaged on a variety of tissue culture cell lines may have amino acid substitutions at the surface of the virus particles, which then may acquire the capacity to enter the cells through alternative receptors [3] . It is likely that cell culture propagation of BCoV affects the antigenic composition [17] , hemagglutination activity [37] , and intestinal replication of the virus [20] . Besides, mutations of SARS-CoV were observed in passage 0 [28] . In order to limit the potential effect of cell culture-related nucleotide mutations, this study used a WD-BCoV strain that had been passaged 6 times in HRT-18G cells. However, this cell culture propagation of WD-BCoV might result in amino acids changes, particularly the S protein, leading to a modification of WD-BCoV cell tropism [3] . In this study, the mild villous atrophy caused by necrosis and desquamation of the villi lining cells was confined to the duodenum sampled at PID 1 from a BCoV-infected calf in comparison with normal duodenal villi. Although the present BCoV caused cytopathic effects to the digestive and respiratory epithelia, its mechanisms are poorly understood, particularly in early infections. The BCoV was reported to cause a loss of microvilli in the in-testinal epithelium of a BCoV-inoculated calf sacrificed 3 h after inoculation [26] , resulting in diarrhea in experimentally inoculated and naturally infected calves. However, the present histological, IHC and SEM studies do not allow a rigorous reconstruction of the sequence of the loss of microvilli during the initial WD-BCoV infections of the digestive tract because transmission electron and immuno-electron microscopic studies were not performed. These analyses would be particularly useful with the specimens sampled 3 h after inoculation, which presumably would have provided more detailed data on their multiplication and fine structural changes. It has been reported [22] that some mouse hepatitis coronavirus (MHV) strains rapidly reduce the level of actin mRNA, whose protein interacts with microtubules. As a consequence, the virus-induced impairment of the actin filaments interacting with microtubules might lose their connection with the plasma membrane covering the microvillus or anchored cilia, resulting in the loss of microvilli or cilia [4] . In addition, TGEV and some MHV strains induce apoptosis in some cell types [1, 12] . Therefore, it is possible that the loss of microvilli and cilia induced by WD-BCoV in the digestive and respiratory tracts might occur through a similar mechanisms such as a combination of MHV-or TGEV-induced effects [23] . Our results suggest that the intestinal lesions caused by WD-BCoV spread from the small intestine to the large intestine. With time, the morphological changes in the duodenum became increasingly severe, which led to villous atrophy and fusion. Eventually, the duodenum showed severe villous atrophy, and the other parts of the intestines including the jejunum, ileum, cecum, colon and rectum exhibited moderate to marked mucosal lesions at PID 8. These changes in the lesions are similar to those of calves inoculated with the CD-BCoV strain in which the lesions were first observed in the small intestine and then extended to the large intestine [13, 25, 26, 31] . In the present study, WD-BCoV induced pathological changes in the respiratory tract. These lesions may play as a predisposing factor for secondary bacterial pathogens such as Mannheimia spp. [10, 38] . In addition, a concurrent infection can exacer-bate the clinical signs of respiratory tract disease [11] . Considering the acknowledged roles of other viruses in the bovine respiratory disease complex [11, 38] , it is possible that WD-BCoV may also act synergistically with other infectious agents and stressors to cause pneumonia in calves [38] .