P108 and T109 on E2 Glycoprotein Domain I Are Critical for the Adaptation of Classical Swine Fever Virus to Rabbits but Not for Virulence in Pigs

Historically, live attenuated vaccines produced by blind passage usually undergo adaptation in cell cultures or nonsusceptible hosts and attenuation in natural hosts, with a classical example being the classical swine fever virus (CSFV) lapinized vaccine C-strain, which was developed by hundreds of passages in rabbits. However, the mechanism of viral adaptation to nonsusceptible hosts and the molecular basis for viral adaptation and attenuation remain largely unknown. In this study, we demonstrated that P108 and T109 on the E2 glycoprotein together with the Erns glycoprotein of the rabbit-adaptive C-strain confer adaptation to rabbits on the highly virulent CSFV Shimen strain by affecting viral entry during infection but do not attenuate the Shimen strain in pigs. Our results provide vital information on the different molecular bases of CSFV adaptation to rabbits and attenuation in pigs.

examples of this are the Chinese hog cholera lapinized virus (HCLV, also known as C-strain) and the lapinized rinderpest virus (2,3). Remarkably, the viral infection is limited if one stage of the life cycle is blocked in nonsusceptible hosts, especially if the virus is unable to utilize a factor(s) which is necessary for infection or to evade a restriction factor(s). Hence, mechanisms of viral adaptation to nonsusceptible hosts by serial passaging are as follows: (i) improving entry efficiency (4,5), (ii) enhancing the ability to antagonize or evade the nonsusceptible host's restriction factor(s) (6)(7)(8), and (iii) promoting the assembly and release of virus particles (9). Entry is the first and vital step during viral infection, which is mediated by the interaction between viral envelope proteins and cellular receptors (10).
In general, CSFV infects only domestic pigs and wild boar (23). However, the species barrier of CSFV between pigs and rabbits was crossed by hundreds of passages of a highly virulent strain in rabbits. C-strain, a live attenuated vaccine against CSF, was developed by passaging a highly virulent CSFV strain in rabbits in the 1950s and is characterized by adaptation to rabbits (ATR) and attenuation in pigs (3). Until now, however, the mechanism of the C-strain ATR and the molecular basis for ATR and attenuation in pigs have remained elusive.
Our previous study found that vSM-HCLVE rns carrying the E rns of C-strain in the background of the highly virulent CSFV Shimen strain is not adaptive to rabbits, indicating that the E rns of C-strain alone could not render the Shimen strain adaptive to rabbits (24). In contrast, the E2 and E rns glycoproteins of C-strain together are sufficient for enabling the adaptation of the Shimen strain to rabbits. Furthermore, we observed that the C-strain backbone chimeras harboring E2 domain I (E2 DomainI ) or crucial mutation sites (E2 P108L-T109I ) on E2 DomainI of the Shimen strain do not adapt to rabbits, while the chimera harboring E2 domain II (E2 DomainII ) of the Shimen strain is adaptive to rabbits (24). Therefore, P108 and T109 on E2 DomainI (E2 P108-T109 ) of C-strain are responsible for its ATR, whereas E2 DomainII containing three different residues (R132, S133, and D191) of C-strain are not determinants of its ATR. Because of the above findings, we further investigated the effects of these critical amino acids of C-strain on the Shimen strain ATR and the virulence in pigs.
In the present study, we demonstrated that E2 P108-T109 in combination with the E rns glycoprotein (E2 P108-T109 -E rns ) of C-strain confers adaptation to rabbits on the Shimen strain. Remarkably, our data revealed that E2 P108-T109 -E rns of C-strain promotes viral entry during infection in the target cells using a series of pseudotyped viruses. Finally, we demonstrated that E2 P108-T109 -E rns of C-strain, responsible for the CSFV ATR, does not affect viral virulence in pigs, whereas E2 DomainII -E rns , which is not associated with the adaptation, attenuates the Shimen strain in pigs.

RESULTS
In vitro rescue and evaluation of chimeric viruses in the background of the Shimen strain. Based on previous results (24), we speculated that E2 P108-T109 or E2 DomainI but not E2 DomainII of C-strain is necessary for conferring adaptation to rabbits on the non-rabbit-adaptive Shimen mutant vSM-HCLVE rns . Since vSM-HCLVE rns E2, harboring the E rns and E2 glycoproteins of C-strain, showed better adaptation to rabbits than vSM-HCLVE1E2, containing the E1 and E2 glycoproteins of C-strain, according to the reproductive animal experiments in our previous study (24), three infectious clones (pSME2 L108P-I109T -HCLVE rns , pSM-HCLVE rns E2 DomainI , and pSM-HCLVE rns E2 DomainII ) were constructed based on the infectious clone pSM-HCLVE rns containing E rns of C-strain in the backbone of the Shimen strain (Fig. 1A). Chimeric viruses were rescued by transfecting individual infectious clones into swine kidney (SK6) cells, and transfected cells were serially passaged. Viral genomic sequencing results demonstrated that the sequences of the rescued chimeric viruses were as expected. The results of indirect immunofluorescence assay (IFA) demonstrated that the rescued viruses were infectious in the cells (Fig. 1B). The growth characteristics of the chimeras were determined in porcine kidney (PK-15) cells. Compared with the parental virus (Shimen strain), vSM-HCLVE rns E2 DomainI had lower viral titers at different time points, while viral titers of Purple boxes indicate genes from the highly virulent Shimen strain, while red boxes indicate genes derived from the lapinized vaccine C-strain (also known as HCLV). The infectious clones of chimeric viruses (vSM-HCLVE rns E2 DomainI , vSM-HCLVE rns E2 DomainII , and vSME2 L108P-I109T -HCLVE rns ) were developed based on the infectious clone of the non-rabbit-adaptive chimeric virus vSM-HCLVE rns carrying the E rns glycoprotein of C-strain. (B) Indirect immunofluorescence staining of the PK-15 cells infected by chimeric viruses and parental virus. The nuclei were stained with 4=,6-diamidino-2-phenylindole (DAPI). Bar, 400 m. (C) Growth kinetics of parental and chimeric viruses in PK-15 cells. The error bars represent the standard deviations for three replicates.
E2 P108-T109 -E rns of C-strain confers adaptation to rabbits on the Shimen strain. To investigate whether E2 P108-T109 -E rns of C-strain adapts the Shimen strain to rabbits, five groups of 24 rabbits were inoculated with different chimeric viruses (Table 1). Rectal temperature was recorded 24 h before and after inoculation and then every 6 h until 72 h postinoculation (hpi). According to the fever response standard described previously (24), vSME2 L108P-I109T -HCLVE rns and vSM-HCLVE rns E2 DomainII did not induce a fever response in rabbits, in contrast to C-strain (Table 1). Four rabbits from each group were randomly selected to be euthanized for determining copy numbers of the viral genome in the spleens by real-time reverse transcription-quantitative PCR (RT-qPCR). The viral genome was detected in the rabbits inoculated with vSME2 L108P-I109T -HCLVE rns , vSM-HCLVE rns E2 DomainI , or C-strain but not vSM-HCLVE rns E2 DomainII (Table 1). To further determine whether the progeny viruses were present in rabbit spleens, virus isolation was performed in SK6 cells. The IFA results demonstrated that the viruses were isolated from the spleens of rabbits inoculated with vSME2 L108P-I109T -HCLVE rns , vSM-HCLVE rns E2 DomainI , or C-strain but not with vSM-HCLVE rns E2 DomainII or Dulbecco's modified Eagle's medium (DMEM), and no mutation occurred, as confirmed by sequencing ( Fig. 2A). The immunohistochemistry results showed that the E2 glycoprotein was expressed in the spleen lymphocytes of the rabbits inoculated with C-strain, vSM-HCLVE rns E2 DomainI , or vSME2 L108P-I109T -HCLVE rns but not vSM-HCLVE rns E2 DomainII or the Shimen strain (Fig. 2B). These results indicated that chimeric viruses carrying E2 P108-T109 -E rns or E2 DomainI -E rns of C-strain are adaptive to rabbits. Additionally, at 10 days postinoculation (dpi), anti-E2 antibodies were detected in the remaining two rabbits of each group ( Table 1). As expected, we demonstrated that E2 P108-T109 or E2 DomainI but not E2 DomainII of C-strain could render vSM-HCLVE rns adaptive to rabbits, suggesting that E2 P108-T109 -E rns of C-strain could confer adaptation to rabbits on the Shimen strain, generating new rabbitadaptive CSFVs. Growth curves of rabbit-adaptive CSFV mutants in primary rabbit spleen lymphocytes and swine macrophages. To examine the growth property of rabbitadaptive CSFV mutants in rabbit cells, spleen lymphocytes were identified as target cells in the rabbits infected with C-strain based on the results of immunohistochemistry assay (Fig. 2B). To determine the exact cell type (T cells or B cells) in the spleens of the rabbits inoculated with C-strain, T cells and B cells were isolated from total spleen cells using the FACSAria cell-sorting system from rabbits inoculated with C-strain at 3 dpi (Fig. 3A). The viral genome copy numbers of C-strain in T cells and B cells were determined by RT-qPCR (Fig. 3B). Furthermore, viral genome copy numbers and NanoLuc activities were determined in spleen lymphocytes isolated from the rabbits inoculated with C-strain or the reporter virus vHCLV-NanoLuc (Fig. 3C). Importantly, luciferase activities were detected in the spleen lymphocytes infected with vHCLV-NanoLuc (Fig. 3D). These results indicated that C-strain infects the spleen lymphocytes (both T cells and B cells) of the rabbits.
Next, primary rabbit spleen lymphocytes were isolated and infected with the Shimen strain, C-strain, vSME2 L108P-I109T -HCLVE rns , vSM-HCLVE rns E2 DomainI , or vSM-HCLVE rns E2 DomainII . The viral genome copy numbers in the cells were analyzed at different time points. The results showed that viral genome copy numbers of rabbitadaptive CSFV mutants (vSME2 L108P-I109T -HCLVE rns and vSM-HCLVE rns E2 DomainI ) or the parental virus C-strain increased over time. In contrast, the viral genome copy numbers of non-rabbit-adaptive CSFVs (vSM-HCLVE rns E2 DomainII and the Shimen strain) decreased over time (Fig. 3E). Unfortunately, the progeny viruses in the supernatants were undetectable using IFA in SK6 cells due to low-level replication, which is consistent with observations in rabbits. These results further demonstrated that chimeric viruses vSME2 L108P-I109T -HCLVE rns and vSM-HCLVE rns E2 DomainI but not vSM-HCLVE rns E2 DomainII are adaptive to primary rabbit spleen lymphocytes.
Furthermore, primary swine macrophages were inoculated with a series of chimeric viruses. The viral genome copy numbers in macrophages or the titers of progeny viruses in the supernatants were determined at different time points. The viral genome copy numbers of the Shimen strain and three chimeric viruses were indistinguishable, except those of C-strain in primary swine macrophages (Fig. 3F). However, the viral titers in the supernatants of the macrophages infected with the Shimen strain were higher than those in cells infected with chimeric viruses (Fig. 3G).
E2 P108-T109 -E rns of C-strain affects viral entry in the target cells. The E2 glycoprotein has been reported to be responsible for viral entry and tropism (14)(15)(16), while RNA replication is mediated by nonstructural proteins (NS3, NS4A-4B, and NS5A-5B) (25). Therefore, we speculated that the distinct adaptation potential of these chimeric viruses in rabbits may be due to different viral entry efficiency. A series of pseudotyped viruses (pps) bearing the E rns , E1, and E2 glycoproteins from the Shimen strain (SMpps) or C-strain (HCLVpps) were generated to investigate whether their differences occur in viral entry. We infected rabbits with SMpps or HCLVpps and isolated the spleen lymphocytes at 48 hpi. To amplify the fluorescence signals, an anti-EGFP (enhanced green fluorescent protein) antibody was used to measure EGFP expression. Meanwhile, mouse IgG was used to exclude nonspecific signals. We detected EGFP in the spleen lymphocytes from the rabbits inoculated with HCLVpps but not SMpps (Fig. 4A), suggesting that the C-strain ATR results from improved entry efficiency.
E2 P108-T109 -E rns of C-strain does not attenuate the Shimen strain in pigs. To examine whether the key amino acids contributing to viral ATR affect virulence in pigs, four groups of pigs were inoculated intramuscularly (i.m.) with vSME2 L108P-I109T -HCLVE rns , vSM-HCLVE rns E2 DomainI , vSM-HCLVE rns E2 DomainII , and the Shimen strain, using 10 5 50% tissue culture infective doses (TCID 50 ). Clinical signs and rectal temperatures were monitored. All the pigs showed typical clinical signs (26) of CSF starting at 3 to 5 dpi, and the pigs inoculated with vSME2 L108P-I109T -HCLVE rns , vSM-HCLVE rns E2 DomainI , or the Shimen strain died at 11 to 16 dpi (Table 2). However, the temperatures of the pigs inoculated with vSM-HCLVE rns E2 DomainII normalized at 11 dpi, and the pigs sur-FIG 4 Entry was affected by the E2 P108-T109 -E rns of C-strain. (A) HCLVpps but not SMpps could enter spleen lymphocytes of rabbits. Spleen lymphocytes were isolated from rabbits inoculated with 10 7 transducing units (TU) of HCLVpps or SMpps or with 1 ml of DMEM. EGFP expression was detected in the spleen lymphocytes of the rabbits inoculated with HCLVpps but not SMpps or DMEM. (B) E2 P108-T109 -E rns of C-strain affected viral entry into lymphocytes. Spleen lymphocytes were isolated from rabbits inoculated with 10 7 TU of SME1-HCLVE rns E2 DomainI pps, SME1-HCLVE rns E2 DomainII pps, or SME1E2 L108P-I109T -HCLVE rns pps or with 1 ml of DMEM. EGFP expression was detected in the spleen lymphocytes of the rabbits inoculated with SME1-HCLVE rns E2 DomainI pps or SME1E2 L108P-I109T -HCLVE rns pps but not those inoculated with SME1-HCLVE rns E2 DomainII pps or DMEM. EGFP expression was detected using anti-EGFP antibody, and an irrelevant mouse IgG served as negative control. (C) The E2 P108-T109 -E rns of C-strain did not affect viral entry into swine SK6 cells. EGFP expression was measured in SK6 cells infected with parental and chimeric pseudotyped viruses at an MOI of 2 at 48 hpi.

Determinants of CSFV Adaptation and Attenuation
Journal of Virology vived until the end of this experiment. Viremia kinetics in animals inoculated with vSME2 L108P-I109T -HCLVE rns or vSM-HCLVE rns E2 DomainI were indistinguishable from those induced by the parental Shimen strain (Fig. 5A and B). In contrast, the pigs inoculated with vSM-HCLVE rns E2 DomainII showed lower viremia than the parental virus (Fig. 5A), while virus was undetectable in the blood samples from the pigs inoculated with vSM-HCLVE rns E2 DomainII using IFA in SK6 cells (Fig. 5B). The pigs inoculated with vSME2 L108P-I109T -HCLVE rns , vSM-HCLVE rns E2 DomainI , or the Shimen strain displayed CSF-specific pathological changes (including enlargement and hemorrhages in the lymph nodes, petechiae in the kidneys, and infarcts in the spleen) (25). In contrast, no obvious or slight pathological lesions were observed in the pigs inoculated with vSM-HCLVE rns E2 DomainII (Fig. 5C).
Moreover, histopathological examination of the various organs from the pigs inoculated with vSM-HCLVE rns E2 DomainII showed slight histopathological lesions, including depletion in lymph nodes and reactive hyperplasia in spleens. However, the pigs inoculated with vSM-HCLVE rns E2 DomainI or vSME2 L108P-I109T -HCLVE rns showed histopathological lesions similar to those in pigs inoculated with the Shimen strain in tonsils (lymphopenia), kidneys (denaturalization and necrosis in some tubular epithelia), and bladder (epithelial degeneration) (Fig. 5D).
Taken together, these observations indicate that E2 P108-T109 -E rns of C-strain contributes to the ATR but does not alter the virulence of the Shimen strain in pigs, whereas E2 DomainII -E rns of C-strain attenuates the Shimen strain, suggesting that there are different molecular determinants of CSFV ATR and virulence in pigs.

DISCUSSION
C-strain, which was developed by passaging a highly virulent CSFV strain in rabbits, is adaptive to rabbits and attenuated in pigs. It has been established that for the Shimen strain, acquiring the adaptation to rabbits depends on E2 together with E rns or E1 of C-strain. Furthermore, P108 and T109 in E2 DomainI are crucial amino acids for C-strain to be adaptive to rabbits (24). In the present study, three chimeric viruses containing E2 P108-T109 , E2 DomainI , or E2 DomainII of C-strain in the backbone of the non-rabbit-adaptive Shimen mutant vSM-HCLVE rns were generated and evaluated for CSFV ATR and virulence in pigs. Our study shows for the first time that E2 P108-T109 or E2 DomainI but not E2 DomainII of C-strain could render vSM-HCLVE rns adaptive to rabbits, suggesting that E2 P108-T109 -E rns of C-strain confers adaptation to rabbits on the Shimen strain. Importantly, we demonstrated that E2 P108-T109 -E rns of C-strain mediated the adaptation of CSFV by prompting viral entry during infection in rabbit spleen lymphocytes. However, pathogenicity analysis in pigs showed that E2 P108-T109 -E rns of C-strain did not alter the virulence of the Shimen strain in pigs.
It has been demonstrated that the E2 glycoproteins of pestiviruses are associated with viral tropism. A chimeric pestivirus containing border disease virus or CSFV E2 glycoprotein in the background of bovine viral diarrhea virus (BVDV) alters the tropism to different cells in contrast to the parental BVDV (13,27). Recombinant E2 glycoproteins derived from three different pestiviruses have different abilities to modify BVDV and CSFV to be able to infect permissive cells, suggesting that the E2 glycoprotein is involved in host tropism of pestiviruses at the entry stage (28). Our findings demonstrated that E2 P108-T109 -E rns of C-strain is responsible for the CSFV ATR, which further  indicates the important roles of E2 and E rns in the tropism of pestiviruses. However, the amount of replication that takes place in primary rabbit spleen lymphocytes is lower than that in primary swine macrophages. The distinct replication ability of CSFV in rabbits and pigs was observed. The possible explanations are as follows: the rabbit is not the natural host for CSFV, the adaptive ability of CSFV is low in primary rabbit spleen lymphocytes, and there is no cell-cell junction in suspended primary rabbit spleen lymphocytes. The mechanism of viral adaptation to a heterogeneous host by passaging in nonsusceptible hosts can be associated with mutations in the viral genome, which can improve entry efficiency and promote viral genome replication and virion assembly and release (5,29). Among these possible mechanisms, entry is the first and most important step for viral infection, and it is determined by the viral surface proteins. It has been reported that the highly conserved residue Q226 is associated with H2N2 and H3N2 viral adaptation to human receptors (4,30). Three mutations in the viral glycoproteins E1 (L216R) and E2 (V388G and M405T) improve the efficiency of hepatitis C virus (HCV) entry into Lunet N mCD81 cells, which are likely to promote exposure of the CD81binding site (5). Mutation of A281 was observed during HIV adaptation in macaques and affected the ability of the HIV-1 Env to use macaque CD4 (31). For pestiviruses, increasing evidence has proved that entry is mediated by several glycoproteins. The E rns glycoprotein of pestiviruses attaches to the virion envelope by directly interacting with the E2 glycoprotein to form E rns -E2 heterodimers and is indispensable for virus attachment and infection of target cells (32). BVDV and CSFV belong to the genus Pestivirus, sharing similar properties (33), and E1-E2 heterodimers are essential for BVDV entry (34). In our study, E2 P108-T109 -E rns of C-strain was demonstrated to be associated with the adaption of CSFV to rabbits by affecting viral entry during infection, which expands our knowledge of the entry of pestiviruses.
Typically, live attenuated vaccines developed by blind passage in cell cultures or nonsusceptible hosts are adaptive to the nonsusceptible host while being attenuated in the primary cell or host. For example, the adaptation of African swine fever virus to Vero cells leads to a gradual attenuation of virulence in pigs (35). Attenuation of the strain PC22A of porcine epidemic diarrhea virus was achieved by cell culture passage (36). Live attenuated vaccines such as C-strain and the lapinized rinderpest virus were developed by passaging a highly virulent strain in rabbits. However, the molecular basis for adaptation and attenuation remains largely unclear. Mutations in the surface glycoprotein E increase the adaptation of tick-borne encephalitis virus to BHK-21 cells and significantly attenuate neuroinvasiveness in adult mice (37). A single mutation in VP2 (A221T) confers the adaptation of infectious pancreatic necrosis virus to CHSE cells and attenuates virulence in Atlantic salmon fry (38). In contrast, D279N and A284T mutations can confer infectious bursal disease virus adaptation in cell culture but do not lead to virus attenuation (39). It was reported that the E2 glycoprotein of the CS vaccine strain derived from the LK VNIIVIM parental vaccine strain (40) could markedly attenuate the CSFV Brescia strain (18). Furthermore, many residues on the E2 were identified as being associated with CSFV virulence, including a discrete epitope (TAVS PTTLR) (19), W871T, W875D, and V878T in the internal fusion peptide (17), T830A (41), and T745I and M979K (22). The N269A/Q substitution, which removed the putative glycosylation site in the E rns glycoprotein, decreased virulence of the Brescia strain (21). Notably, our results showed that E2 P108-T109 -E rns of C-strain, which is associated with CSFV ATR, did not affect CSFV virulence in pigs. The E rns glycoprotein of C-strain does not affect the virulence of the chimeric viruses, which may be due to the N269 present on both C-strain and the Shimen strain (21). However, E2 glycoprotein domain II, which is irrelevant to adaptation, alters virulence, further demonstrating that the molecular determinants of the ATR are different from those of attenuation of CSFV in pigs.
In summary, we demonstrated that E2 P108-T109 -E rns of C-strain determines the adaptation of CSFV by facilitating viral entry during infection of rabbits. Furthermore, the molecular determinants of CSFV ATR do not affect virulence in pigs. Our findings contribute to our understanding of the molecular basis for adaptation and attenuation of live attenuated vaccines developed by blind passage in cell cultures or hosts. This study also implies that novel live attenuated vaccines against CSF may be developed by targeting genetic modifications instead of random evolution through blind cell passage.
Three chimeric viruses were rescued as described previously with a slight modification (24). Six micrograms of each plasmid mixed with 6 l of X-tremeGENE HP DNA transfection reagent (catalog no. 6366546001; Roche) was transfected into SK6 cells cultured in a 6-well plate. The transfected cells were passaged several times, and the supernatants were subjected to detection of the E rns glycoprotein by a CSFV antigen test kit (catalog no. 99-40939; IDEXX). The positive samples were identified by RT-PCR, sequencing, and IFA.
IFA and virus titration. The viral titers in the rabbit spleens or the pig blood samples at different time points were determined as previously described (44). SK6 or PK-15 cells were inoculated with serial 10-fold dilutions of the samples and cultured in 37°C for 48 h. Cold absolute ethanol was used for fixing cells at Ϫ20°C for 20 min. The fixed cells were washed three times with phosphate-buffered saline (PBS) and incubated with an anti-E2 polyclonal antibody at 37°C for 2 h (45). After five washes with PBS, cells were incubated with Alexa Fluor 488 goat anti-rabbit IgG (catalog no. A11034; Roche) at 37°C for 1 h and then washed five times with PBS. Then, the cells were stained with 0.5 g/ml 4=,6-diamidino-2phenylindole (DAPI; catalog no. C0060; Solarbio) for 15 min and washed three times with PBS. The cells were analyzed for green fluorescence using an inverted fluorescence microscope (EVOS FL; Life Technologies). The viral titers were calculated according to the Reed-Muench method and expressed as median tissue culture infective doses (TCID 50 ) per milliliter (46).
Growth curves of the rescued viruses. To determine the multistep growth curves of the rescued viruses, PK-15 cells cultured in 24-well plates were infected with these rescued viruses at a multiplicity of infection (MOI) of 0.1. Two hours later, the supernatants were removed, and the cells were washed three times with PBS. Then, fresh DMEM supplemented with 2% FBS was added to each well, and the GATGGATCCTCTCCACTATAATAG pSME2 L108P-I109T -HCLVE rns -1F CCACCTCGAGATGCTATGTGG pSME2 L108P-I109T -HCLVE rns -1R CTCGAATGTCACGGAAGTGGGTAAAGCCCCCTTATGC pSME2 L108P-I109T -HCLVE rns -2F GCATAAGGGGGCTTTACCCACTTCCGTGACATTCGAG pSME2 L108P-I109T -HCLVE rns -2R TAGATGGATCCTCTCCACTAT cells were cultured at 37°C and 5% CO 2 . The cells were harvested at 12-h intervals and used to determine viral titers.
RNA extraction and RT-qPCR. Total RNA from tissues or cells was extracted using RNAiso Plus (catalog no. 9109; TaKaRa) according to the manufacturer's protocol. cDNA synthesis was processed in a 20-l volume with avian myeloblastosis virus (AMV) reverse transcriptase XL (catalog no. 2621; TaKaRa). The copy numbers of the CSFV genome were determined using a previously described RT-qPCR assay (47).
Luciferase assay. The reporter virus vHCLV-NanoLuc expressing the NanoLuc protein fused with the N pro protein (between the amino acids 13 and 14) based on C-strain was generated. Rabbit spleen lymphocytes infected with vHCLV-NanoLuc or C-strain or treated with DMEM only were lysed with 100 l of passive lysis buffer (catalog no. N1150; Promega) and incubated on a shaker for 1 h at 4°C. The supernatants were collected by centrifuging at 12,000 ϫ g for 10 min at 4°C. NanoLuc activities were measured with EnVision multilabel plate readers (PerkinElmer).
Flow cytometry. The spleens from the rabbits inoculated with C-strain were collected after rabbits were euthanized at 3 dpi, and spleen cells were obtained by smearing the organs on 200-mesh copper wire mesh with RPMI 1640 medium. The red blood cells were lysed using red blood cell lysis buffer (catalog no. R1010; Solarbio), and the lymphocyte suspensions were washed with PBS. The following antibodies were used for flow cytometry, according to the manufacturers' instructions: goat anti-rabbit IgM chain preadsorbed to secondary antibody (DyLight 488) (catalog no. ab98454; Abcam) for isolating B cells, mouse anti-rabbit T lymphocytes (catalog no. MCA800GA; Bio-Rad) for isolating T cells, and Alexa Fluor 633 goat anti-mouse IgG (heavy plus light chain [HϩL]) (catalog no. A21052; Invitrogen) as the secondary antibody. Stained mononuclear cells from the spleens of the rabbits infected with C-strain were added to cytometry tubes and sorted with a high-speed cell sorter (MoFlo XDP; Beckman Coulter).
Infection of primary cells from rabbits or pigs with chimeric viruses. Primary rabbit spleen lymphocytes or swine macrophages were isolated from the animals and infected with parental or chimeric viruses with an MOI of 0.01. After 2 h of incubation in a CO 2 incubator, the infected cells were washed three times with PBS and further incubated for 4, 24, 48, or 72 h (17). The viral genome copy numbers or the viral titers at different time points were measured.
Experimental infection of rabbits with chimeric viruses. Twenty-four 14-week-old New Zealand White rabbits were divided into 5 groups and inoculated intravenously (i.v.) via the marginal ear vein with the viruses indicated in Table 1. The rectal temperature of all rabbits was monitored every 6 h from 24 to 72 hpi as described previously (24). Four rabbits were selected randomly from each group and euthanized at 3 hpi. The viral genome copy numbers in the spleens of the rabbits were determined and virus isolation was performed as described previously (24). The anti-E2 antibodies of the remainder rabbits were tested at 10 hpi using the classical swine fever virus antibody test kit (catalog no. 99-43220; IDEXX) according to the manufacturer's manuals.
Experimental infection of rabbits with pseudotyped viruses. The rabbits were inoculated i.v. with different chimeric pseudotyped viruses. Spleen lymphocytes were isolated at 48 hpi. Mouse anti-coral green fluorescent protein (cGFP)-tagged monoclonal antibody (MAb) (catalog no. A00185; GenScript) was used to detect the expression of EGFP in lymphocytes, and irrelevant mouse IgG (catalog no. A7028; Beyotime) was used as a negative control. As the secondary immunoreagent, fluorescein isothiocyanatelabeled goat anti-mouse IgG (catalog no. A11029; Invitrogen) was used. All antibodies mentioned were diluted at 1:200 in PBS. For flow cytometry analysis, 10 6 cells in each sample were permeabilized with 0.15% Triton X-100. The cells were further incubated with primary antibody for 1 h and washed three times for 5 min with PBS. Then the cells were incubated with secondary antibody for 45 min and washed three times for 5 min with PBS. The fluorescence signal was analyzed with an Accuri C6 Plus flow cytometer (BD Biosciences).
Experimental infection of pigs with chimeric viruses. To assess the virulence of chimeric viruses relative to the Shimen strain, 5-week-old healthy pigs were randomly divided into 4 groups (groups 1 to 3, n ϭ 3; group 4, n ϭ 2), and each group was housed in an individual room. Group 1 was inoculated i.m. with vSM-HCLVE rns E2 DomainI , group 2 was inoculated i.m. with vSM-E2 L108P-I109T -HCLVE rns , group 3 was inoculated i.m. with vSM-HCLVE rns E2 DomainII , and group 4 was inoculated i.m. with the Shimen strain. The inoculation dose of the virus for each group was 10 5 TCID 50 (17). Clinical signs and rectal temperature were monitored daily, and anticoagulated blood samples of pigs were collected every 3 or 4 days. CSFV RNA was determined in anticoagulated blood samples by RT-qPCR. The chimeric viruses in the blood samples were titrated in SK6 cells using IFA.
Pathological examinations. Macroscopic and microscopic pathological changes of the pig tissues were examined as described previously (50). The tonsils, lymph nodes, kidneys, bladders, and spleens were fixed with 10% formalin and then embedded in paraffin wax. For histopathological examinations, prepared tissue sections were stained with hematoxylin and eosin (H&E).