The S2 Subunit of Infectious Bronchitis Virus Beaudette Is a Determinant of Cellular Tropism

Infectious bronchitis remains a major problem in the global poultry industry, despite the existence of many different vaccines. IBV vaccines, both live attenuated and inactivated, are currently grown on embryonated hen's eggs, a cumbersome and expensive process due to the fact that most IBV strains do not grow in cultured cells. The reverse genetics system for IBV creates the opportunity for generating rationally designed and more effective vaccines. The observation that IBV Beaudette has the additional tropism for growth on Vero cells also invokes the possibility of generating IBV vaccines produced from cultured cells rather than by the use of embryonated eggs. The regions of the IBV Beaudette S glycoprotein involved in the determination of extended cellular tropism were identified in this study. This information will enable the rational design of a future generation of IBV vaccines that may be grown on Vero cells.


T he avian coronavirus infectious bronchitis virus (IBV) is a member of the genus
Gammacoronavirus in the order Nidovirales (1) and the etiological agent of the disease infectious bronchitis (IB) that affects domestic fowl (2)(3)(4)(5). IBV replicates primarily in the respiratory tract (6,7), causing a highly contagious respiratory disease characterized by nasal discharge, snicking, rales, and tracheal ciliostasis in chickens (8), but also in many other epithelial surfaces, including enteric surfaces (9), oviducts, and kidneys (10)(11)(12). 60) and for subsequent infection of the tracheal epithelium (61,62). The interaction between IBV and sialic acid has been confirmed by Madu et al. (63), who also identified a heparan sulfate (HS) binding site between amino acid residues 686 and 691 of the S2 subunit of the Beaudette S glycoprotein and showed that HS may be involved as a cofactor in Beaudette virus entry into host cells. Yamada and Liu (64) identified an additional furin cleavage motif within the putative HS binding site identified by Madu et al. (63) with a role in viral entry and syncytium formation in vitro. Cleavage was mapped to arginine residue 690. The susceptibility of cultured cells to infection with Beaudette was subsequently found to correlate with cellular furin expression levels (65). This additional proteolytic cleavage site within the S2 subunit, named S2=, has since been identified in SARS-CoV, MERS-CoV, and HCoV-229E and found to mediate membrane fusion and viral infectivity (66)(67)(68)(69).
Coronaviruses generally exhibit restricted cell and tissue tropism, which is dependent upon the S glycoprotein of individual coronavirus strains (70,71). However, some strains of IBV, following adaptation, are able to replicate in primary chicken cells, such as chick kidney (CK) cells. Interestingly, the Beaudette strain, after several hundred passages in embryonated eggs, was also discovered to have an extended host range with the ability to replicate in a mammalian cell line, Vero cells (72), and to a limited extent in baby hamster kidney (BHK-21) cell lines (73). We have previously demonstrated that the cellular tropism of IBV is determined by the S glycoprotein; replacement of the ectodomain of the IBV Beaudette S glycoprotein with the corresponding region from the pathogenic IBV M41-CK resulted in a recombinant IBV (rIBV), BeauR-M41(S), which had the tissue tropism associated with M41-CK (74). The present study aims to elucidate the specific regions of the IBV Beaudette S glycoprotein involved in the determination of extended cellular tropism, thus enabling the rational design and generation of IBV vaccines that may be grown on Vero cells.

RESULTS
Generation of rIBVs expressing S genes with chimeric subunits. Recombinant IBVs expressing chimeric S proteins composed of either the S1 subunit derived from M41-CK and the S2 subunit from CK cell-adapted strain Beaudette (Beaudette CK), BeauR-M41(S1), or the S1 subunit from Beaudette CK and the S2 subunit from M41-CK, BeauR-M41(S2), were produced ( Fig. 1) to determine which subunit of the S glycoprotein was responsible for the extended tropism of Beaudette CK for Vero cells. Sequence analysis following rescue of the rIBVs identified the presence of three nucleotide substitutions in BeauR-M41(S1) passaged three times (P 3 ) on CK cells [BeauR-M41(S1) P 3 -CK] at positions 22239 to 22241, GTT to AAC, corresponding to two adjacent codons, amino acids L 624 and F 625 , resulting in 1 amino acid substitution, F 625 T, within the Beaudette S2 subunit (Table 1). One nucleotide substitution, C 23378 T, was identified in BeauR-M41(S2) P 3 -CK, resulting in amino acid substitution A 452 T within the Beaudette S1 subunit. These substitutions could have arisen during rescue or passage of the virus on CK cells as they were not present in the parental recombinant vaccinia viruses (rVV).
Expression of chimeric S genes does not alter growth on primary chicken cells. Analysis of the growth kinetics of the rIBVs BeauR-M41(S1) P 3 -CK and BeauR-M41(S2) P 3 -CK in primary chick kidney (CK) cells identified peak titers equivalent to those of the parental viruses, Beau-R, M41-CK, and BeauR-M41(S), at 48 h postinfection. Interestingly, the growth of the rIBVs was initially slower over the first 24 h ( Fig. 2A). Immunofluorescence analysis of IBV-infected CK cells showed that the rIBVs were able to infect and spread to neighboring cells, as observed for the parental viruses (data not shown).
The S2 subunit of Beaudette is responsible for Vero cell tropism. Immunofluorescence analysis showed that Vero cells could be infected with Beau-R or BeauR-M41(S1) P 3 -CK ( Fig. 2D and G). However, only a small number of individual Vero cells were infected with M41-CK or BeauR-M41(S) (Fig. 2E and F), confirming previous results that infectious progeny virus does not infect neighboring cells and that there is no virus spread (74), indicating that progeny virus either is not released from infected Vero cells or is not infectious. Although the rIBV BeauR-M41(S2) P 3 -CK formed infectious centers on CK cells, the rIBV had a phenotype on Vero cells similar to that observed with M41-CK and BeauR-M41(S) (Fig. 2H).
Analysis of the growth kinetics of the rIBVs on Vero cells confirmed the findings of the immunofluorescence studies. BeauR-M41(S1) P 3 -CK demonstrated growth similar to that of Beau-R in Vero cells, albeit with a lower peak titer (Fig. 2B). M41-CK, BeauR-M41(S), and BeauR-M41(S2) P 3 -CK did not reach titers greater than 1.5 log 10 PFU/ml in Vero cells. Our results show that BeauR-M41(S1) P 3 -CK and Beau-R produce infectious progeny in both CK and Vero cells, whereas the tropism of BeauR-M41(S2) P 3 -CK resembled that of M41-CK and BeauR-M41(S) for growth in Vero cells. Overall, replacement of the Beaudette S1 subunit with the corresponding S1 sequence from M41-CK did not affect the ability of the rIBV to replicate and produce progeny virus in Vero cells. In contrast, replacement of the Beaudette S2 subunit with the M41 S2 sequence resulted in loss of the ability of Beau-R to grow in Vero cells. The ability of Beau-R and BeauR-M41(S1) P3-CK to grow on Vero cells resides within the Beaudette S2 subunit and not within the S1 subunit, known to contain the RBD.
Adaptation of rIBVs to growth on Vero cells. The P 3 -CK rIBVs were passaged three times on Vero cells (P 3 -Vero rIBVs) to determine whether virus replication could be maintained by passage on this mammalian cell line. Passage of BeauR-M41(S1) resulted in a cytopathic effect (CPE) from 24 h postinfection, whereas passage of BeauR-M41(S2) did not result in an observable CPE. Additional passages of the rIBV BeauR-M41(S1) resulted in syncytium formation from P 5 , indicating that the virus, as previously observed for Beaudette, had become adapted for growth on Vero cells (data not shown).
Analysis of the growth kinetics of BeauR-M41(S1) P 3 -CK passaged on Vero cells showed an increase in titer following passage to within 1 log 10 of the titers observed for Beau-R. Overall, the growth of the P 7 -Vero isolate matched the growth kinetics observed for Beau-R (Fig. 3).
Sequence analysis of the S gene from BeauR-M41(S1) P 7 -Vero identified one nucleotide difference, located within the S2 subunit, from the parental viruses, Beau-R and BeauR-M41(S). The substitution occurred at nucleotide position C 23378 T, resulting in an amino acid change of T 1004 I (Table 1). This change may contribute to the further adaptation of BeauR-M41(S1) for growth in Vero cells and may be involved in syncytium formation.
Generation of rIBVs with modified Beaudette S2 subunit-specific motifs. Once it was established that the S2 subunit is responsible for the extended tropism of Beaudette in Vero cells, the amino acid sequence of Beaudette S2 was compared to the S2 amino acid sequences in other IBV strains to identify potential amino acids, unique to the Beaudette S2, which may play a role in the tropism for growth in Vero cells. The positions of the amino acid differences between Beaudette, M41, and two pathogenic field strains of IBV, QX and 4/91, are shown in Fig. 1B. A Beaudette-specific motif (BSM), 686 SRRKRSLIE 694 , was identified in the Beaudette S2 sequence surrounding the S2= cleavage site at arginine residue 690 and was not found in the S glycoprotein of any other IBV strain; the Beaudette-specific amino acids within the BSM are underlined. In order to determine whether this motif plays a role in the Vero cell tropism of IBV The S2 Subunit Determines IBV Tropism Journal of Virology Beaudette, two full-length IBV cDNAs were generated: BeauR-S-MM, which has the Beau-R genomic background (75) in which the BSM in the S2 subunit was replaced with nucleotides encoding the corresponding M41 motif (MM) sequence, 686 SPRRRSFIE 694 (Fig. 1B), and BeauR-M41-S-BSM, which was based on BeauR-M41(S) (74), in which the MM in the M41 S2 subunit was replaced with the BSM (Fig. 1B). Following rescue and growth in CK cells, sequence analysis identified the presence of one nucleotide substitution in BeauR-M41-S-BSM P 3 -CK, A 22226 C, resulting in the amino acid substitution D 620 A in the S2 subunit (Table 1), which could have arisen during rescue or passage of the virus on CK cells, as the substitution was not present in the rVV. No mutations were identified in the BeauR-S-MM P 3 -CK S gene. Analysis of the growth kinetics of BeauR-S-MM P 3 -CK and BeauR-M41-S-BSM P 3 -CK in CK cells showed that they displayed peak titers equivalent to those of the parental viruses, Beau-R, M41-CK, and BeauR-M41(S), at 48 h postinfection, although the growth of BeauR-M41-S-BSM was slower over the first 24 h (Fig. 4A). Immunofluorescence analysis of infected CK cells confirmed that both viruses were able to infect neighboring cells (data not shown).
The Beaudette-specific motif in the S2 subunit is sufficient to confer the tropism for growth in Vero cells. Immunofluorescence analysis of infected Vero cells infected with BeauR-M41-S-BSM P 3 -CK (Fig. 4B) demonstrated that the introduction of the BSM into the M41 S glycoprotein resulted in the ability of an rIBV expressing the modified M41 S glycoprotein to grow on Vero cells. Conversely, rIBV BeauR-S-MM P 3 -CK, in which only the BSM in the Beaudette S2 subunit was replaced with the corresponding MM sequence from M41, lost its ability to grow in Vero cells (Fig. 4C), with a growth phenotype being observed for M41 and BeauR-M41(S). This observation was confirmed by analysis of the growth kinetics of the rIBVs in Vero cells (Fig. 4D). Production of progeny BeauR-S-MM P 3 -CK virus from Vero cells was undetectable. Although BeauR-M41-S-BSM P 3 -CK produced progeny virus in Vero cells, the growth was considerably lower than that observed for Beau-R. Progeny BeauR-M41-S-BSM P 3 -CK virus was observed only from 24 h postinfection and reached a peak titer at 72 h postinfection that was about 6 log 10 lower than that of Beau-R. These results show that the loss of the BSM abolished the ability of Beau-R to grow on Vero cells and that the accruement of the BSM resulted in the ability of an rIBV expressing the M41 S to have a tropism for growth on Vero cells, albeit to a lesser extent than replacement with the entire Beaudette S2. Overall, it appeared that either a complete Beaudette S2 subunit or other additional Beaudette S2-specific amino acids may be required for the growth characteristics observed with the complete Beaudette S2.
The P 3 -CK viruses BeauR-M41-S-BSM and BeauR-S-MM were passaged on Vero cells to determine whether virus replication could be maintained. BeauR-S-MM isolates were unable to maintain replication on Vero cells for three passages, as assessed by reverse transcription-PCR (RT-PCR). The isolates caused no CPE, and no infected cells were observed on either CK or Vero cells (data not shown). The rIBV BeauR-M41-S-BSM isolates were successfully passaged on Vero cells, and the growth kinetics of one of the BeauR-M41-S-BSM P 7 -Vero passaged isolates were analyzed. The virus had a growth pattern very similar, within 1 log 10 , to the growth pattern of Beau-R on Vero cells (Fig.  5). No additional mutations were identified in the BeauR-M41-S-BSM P 7 -Vero S gene, other than the nucleotide substitution identified at P 3 -CK, A 22226 C, resulting in the amino acid substitution D 620 A in the S2 subunit. The genetic stability of BeauR-M41-S-BSM observed over seven passages on Vero cells indicates that the spike glycoprotein is not under pressure to mutate under these conditions. Generation of rIBVs with additional Beaudette-specific amino acid modifications within the M41-CK S2 subunit. Once it was established that the BSM within the S2 subunit is able to confer the extended tropism of IBV in cell culture, other Beaudettespecific amino acids, identified in the Beaudette S2 subunit in comparison to other IBV S2 sequences (Fig. 1B), were introduced into the M41-CK S2 subunit, in addition to the BSM. In order to determine the specific amino acids and minimum number of amino acid changes required to fully confer tropism, full-length IBV cDNAs with various combinations of the Beaudette-specific amino acids in the S2 subunit of M41 based on the BeauR-M41-S-BSM genomic background were generated; L 578 F and N 617 S were introduced by pGPT-M41-S2-L 578 F-N 617 S, and N 826 S, L 857 F, and I 1000 V were introduced by pGPT-M41-S2-N 826 S-L 857 F-I 1000 V. Following homologous recombination into the IBV  (Table 1), which could have arisen during rescue or passage of the virus on CK cells, as the substitutions were not present in the rVV.
Analysis of the growth kinetics of the rIBVs with the modified M41 S2 subunit on CK cells showed that most of the viruses, apart from BeauR-M41-S-BSM-L 578 F-N 617 S-N 826 S-L 857 F-I 1000 V, had growth patterns that were representative of the rIBV BeauR-M41(S) growth pattern, in that they produced less progeny virus than Beau-R in the first 24 h postinfection but had titers that reached or exceeded the titer of Beau-R over the next 48 h postinfection (Fig. 6A). Interestingly, rIBV BeauR-M41-S-BSM-L 578 F-N 617 S-N 826 S-L 857 F-I 1000 V had a growth pattern similar to that of Beau-R, in that it had a peak titer at 24 postinfection which decreased over the 48 h postinfection, and the peak titer was about 1.2 log 10 less than that of Beau-R at 24 h postinfection. All the rIBVs had a similar growth pattern in CK cells by immunofluorescence analysis of the infected cells (data not shown).
Assessment of tropism. Analysis of the growth kinetics of the P 3 -CK rIBVs on Vero cells showed that they all grew with patterns similar to those of Beau-R, but with lower peak titers and lower titers throughout infection, whereas BeauR-M41(S), expressing the donor S protein, did not grow in Vero cells (Fig. 6B). Interestingly, BeauR-M41-S-BSM-L 578 F-N 617 S-N 826 S-L 857 F-I 1000 V P 3 -CK, which contains the most Beaudette-specific amino acids inserted into the M41 S2 subunit, grew the most similarly to Beau-R, indicating that only 8 Beaudette-derived amino acids are required to confer the ability to grow on Vero cells with growth kinetics and a growth phenotype similar to those of Beau-R in Vero cells. The other rIBVs with modified M41-CK S2 subunits replicated to a titer similar to that of BeauR-M41-S-BSM P 3 -CK, reaching a peak titer at 72 h postinfection.
Immunofluorescence analysis of infected Vero cells showed that the P 3 -CK rIBVs containing the additional Beaudette-specific modifications within the M41-CK S2 sub- The S2 Subunit Determines IBV Tropism Journal of Virology unit formed infectious centers on Vero cells (Fig. 6C to H). At 24 h postinfection, only small foci of infected Vero cells were generated by some of these rIBVs during the initial passage on Vero cells (Fig. 6E and G). This is reflected in the delayed growth during the first 24 h postinfection, as shown in Fig. 6B. As shown above, passage of BeauR-M41-S-BSM P 3 -CK on Vero cells improved the growth of this virus in Vero cells; therefore, the rIBVs with the additional Beaudettederived amino acids in the M41 S2 subunit were passaged on Vero cells, and the P 7 -Vero isolates were reexamined for growth on Vero cells. Analysis of the growth kinetics of the P 7 -Vero isolates on Vero cells clearly demonstrated that all the rIBVs now replicated with patterns similar to and peak titers equivalent to those observed for Beau-R (Fig. 7). Interestingly, three of the P 7 -Vero rIBVs, BeauR-M41-S-BSM-N 617 S, BeauR-M41-S-BSM-L 578 F-N 617 S, and BeauR-M41-S-BSM-N 826 S-L 857 F-I 1000 V, had higher titers of progeny virus, almost 2 log 10 higher to 24 h postinfection, than Beau-R; at 72 h postinfection, the peak titers were similar to those of Beau-R. Sequence analysis of the S genes from the P 7 -Vero isolates identified several nucleotide substitutions that may be involved in further conferment of tropism for Vero cells (Table 1); BeauR-M41-S-BSM-L 857 F-I 1000 V P 7 -Vero C 22255 T results in the amino acid substitution P 630 S, BeauR-M41-S-BSM-L 578 F-N 617 S-N 826 S-L 857 F-I 1000 V P 7 -Vero G 20584 A results in the amino acid substitution G 73 S, and BeauR-M41-S-BSM-N 617 S, -L 578 F-N 617 S, and -N 826 S-L 857 F-I 1000 V P 7 -Vero were all found to have a common substitution, A 22962 C, resulting in the amino acid substitution Q 866 H in the M41 S2 subunit. BeauR-M41-S-BSM-N 826 S-L 857 F-I 1000 V P 7 -Vero had several other substitutions; G 21709 A resulted in V 448 I, and C 22342 A resulted in P 659 T and a mixed population of C 22099 T corresponding to the first engineered amino acid modification in the M41 S2 subunit L 578 F. The rIBV appears to be losing the L 857 F modification, resulting in a mixed population at C 22936 T. Interestingly, BeauR-M41-S-BSM-I 1000 V P 7 -Vero was found to gain the Beaudettespecific amino acid L 578 F, which was introduced into the M41 S2 subunit in some of the rIBVs.
The rIBVs containing the Beaudette-specific motif are proteolytically cleaved at the S2= cleavage site. As the Beaudette-specific motif surrounds the S2= cleavage site, the susceptibility of rIBVs containing the BSM to proteolytic cleavage was investigated. CK cells were infected with BeauR-M41-S-BSM P 3 -CK, BeauR-S-MM P 3 -CK, or parent virus Beau-R, BeauR-M41(S), or M41-CK for 24 h and then lysed, and the proteolytic cleavage of the spike glycoprotein was analyzed by Western blotting with anti-S2 antibody (Fig. 8). A clear S2= cleavage product was observed in the Beau-R sample, but this was absent in the BeauR-M41(S), M41-CK, and BeauR-S-MM samples, which do not contain the BSM and are unable to grow in Vero cells. A faint band corresponding to the S2= cleavage product was also observed in the BeauR-M41-S-BSM sample, indicating that proteolytic cleavage at the S2= site within the BSM may play a role in extending the host tropism of Beau-R. The faint band generated by BeauR-M41-S-BSM in comparison with that produced by Beau-R may indicate that proteolytic processing is not as efficient in the recombinant spike glycoprotein. This could be due to subtle differences in the tertiary or quaternary structures of the BeauR-M41-S-BSM spike glycoprotein resulting in steric hindrance of the protease acting at the S2= site. The size of the S2= cleavage product is smaller than expected when analyzed by Western blotting. It is possible that the band that we observed at about 30 kDa may be a breakdown product of the S2= cleavage product or that the size of the S2= cleavage product may appear to be smaller than expected due to distortion of protein coils in the gel. Nevertheless, it is important to note that this band was observed only in samples infected with rIBV containing the Beaudette-specific motif surrounding the S2= cleavage site, indicating that this is where cleavage occurs, so its production may be involved in the extended tropism of Beau-R.

DISCUSSION
Infectious bronchitis remains a major problem in the global poultry industry, despite the existence of many different vaccines. IBV vaccines, both live attenuated and inactivated, are currently grown in embryonated hen's eggs, due to the fact that most IBV strains do not grow in cultured cells. Production of IBV vaccines in embryonated eggs is expensive and cumbersome, with each egg producing only a small volume of allantoic fluid from which the virus is isolated. Furthermore, the supply of embryonated eggs is not guaranteed to be reliable, which may seriously affect the production of required vaccines. As the demand for seasonal and pandemic influenza vaccines rises, the supply of embryonated eggs for the production of other vaccines may be reduced. Another concern about the use of embryonated eggs is the possible presence of adventitious viruses, which may compromise the vaccine stocks or cause pathology in the vaccinated chickens.
In light of the numerous disadvantages of egg-based vaccine production, the ability to produce live vaccines in vitro would be beneficial to the vaccine industry as well as the poultry industry. Vaccine production in vitro is faster, more efficient, and able to produce larger volumes of vaccine than production in ovo. This is particularly important given the very competitive vaccine price points in the poultry sector. This approach would also reduce the number of embryonated hen's eggs utilized, an important consideration under the principles of the 3Rs for the more ethical use of animals in testing (replacement, reduction, refinement).
Coronavirus S glycoproteins have recently been shown to undergo large-scale conformational changes upon fusion with the host cell membrane (30). We have Cells were lysed at 24 h postinfection, and spike glycoproteins were analyzed by Western blotting with anti-S2 antibody. Actin was detected with anti-beta-actin antibody. Two replicates of BeauR-S-MM P 3 -CK and BeauR-M41-S-BSM P 3 -CK were used. Bands corresponding to uncleaved spike, the S2 subunit, and the S2= cleavage product are labeled. The S2= cleavage product was detected in lanes 2, 7, and 8 only.
demonstrated that it is possible to generate rIBVs expressing chimeric S glycoproteins comprised of S1 and S2 subunits from two different strains of IBV that are able to maintain productive infections in vitro and generate peak titers similar to those of the parent viruses. This indicates that the chimeric spikes are still able to undergo the conformational changes required during entry into host cells. It is possible that inefficient conformational changes leading to a reduction in the rate of viral entry into host cells may be responsible for the lag in viral growth kinetics for some of the rIBVs with modified spike glycoproteins at early times in the course of infection. These may be corrected by the additional mutations accrued during adaptation of the rIBVs to Vero cells, as the growth kinetics of all rIBVs increased upon serial passage.
Although live attenuated and inactivated vaccines are universally used in the control of IBV, they do not offer cross-protection between the different circulating serotypes of IBV. The advent of a reverse genetics system for IBV (75)(76)(77) creates the opportunity for generating rationally designed and more effective vaccines. The prospect of swapping spike genes from emerging strains of IBV into an attenuated backbone, such as Beau-R, is promising (8,78).
The observation that IBV Beaudette, a highly attenuated laboratory strain, has the additional tropism for growth in Vero cells invokes the further possibility of generating IBV vaccines produced from cultured cells rather than by the use of embryonated eggs. Vero cells were first isolated in 1962 from kidney epithelial cells extracted from an African green monkey. They have already been validated for virus growth and diagnostic purposes and are licensed for use in human vaccine manufacture. Vero cells are currently used in the production of polio and rabies vaccines (79,80), and several influenza virus vaccines have been developed for growth on Vero cells (81,82). Vero cells have been extensively tested for tumorigenic properties and can be grown in suspension or a flat bed, and it is possible to achieve consistent virus yields.
Beaudette is itself too attenuated to be an efficacious vaccine (8,83); however, we have identified the specific region in the Beaudette genome, within the spike glycoprotein, responsible for its ability to grow in Vero cells. While the S1 subunit of IBV contains the receptor binding domain and is responsible for binding to host cells, it was determined that infectivity for Vero cells is conferred by the Beaudette S2 subunit, in particular, the Beaudette-specific motif 686 SRRKRSLIE 694 surrounding the S2= cleavage site. Although there is a clear indication of the involvement of the Beaudette-specific motif in the ability of Beau-R to replicate on Vero cells, the addition of other Beaudettespecific amino acids within the S2 subunit or mutations introduced by serial passage in Vero cells were able to further increase the growth kinetics of the rIBVs on Vero cells. These additional mutations may serve to optimize the structure of the recombinant S glycoprotein. Interestingly, the Beaudette-specific motif does not appear to confer the ability to grow on BHK-21 cells in the same way (data not shown), indicating that there may be other regions of the S glycoprotein responsible for the extended host range of Beau-R in mammalian cells other than Vero cells.
The Beaudette-specific motif has two putative mechanisms to permit infectivity for Vero cells: either by facilitating binding to additional host attachment factors (63) or as an additional protease cleavage site (64). It is possible that both activities may play a role in the extended tropism of Beaudette in vitro. The results of this study suggest that IBV Beaudette employs a mechanism of entry into Vero cells different from that into CK cells. Western blot analysis demonstrated that rIBVs containing the Beaudette-specific motif are susceptible to cleavage at the S2= site, whereas the rIBVs containing the equivalent sequence from M41 are not cleaved at the S2= site. The efficiency of this cleavage appears to be lower in the rIBVs containing the BSM than in Beau-R, however, which may be due to differences in the tertiary or quaternary structure between the spike glycoproteins, resulting in the S2= site being less accessible to the protease. The S2= cleavage product of 472 amino acids equates to about 52 kDa if you assume a linear relationship between molecular weight and the size of the protein. The S2 is glycosylated and palmitoylated, which alters the size and which can distort the protein coils that form when the protein is denatured using SDS. We consistently see a band of about 30 kDa in Beau-R-infected samples. It is possible that this band may be a breakdown product of the S2= cleavage product or that the size of the S2= cleavage product may appear to be smaller than expected due to distortion of protein coils in the gel. Nonetheless, this study contributes to the mounting evidence from other coronaviruses that cleavage at the S2= site is involved in entry and that this region plays a role in fusion activation (64,66,67,69).
Inclusion of this Beaudette-specific region of the spike glycoprotein in recombinant viruses confers the ability to grow in Vero cells to incompetent strains, such as BeauR-M41(S). We have previously demonstrated that this recombinant IBV expressing the M41 spike glycoprotein in the genetic background of Beaudette is capable of inducing protective immunity against challenge with virulent M41 (8). These findings are significant, as they allow the development of vaccine strains expressing different spike glycoproteins that may be propagated in cell culture, saving costs and reducing the number of eggs required for vaccine production.

Cells and viruses.
The IBV strains used were (i) M41-CK, derived from the pathogenic M41 strain (84) after multiple passages in CK cells (8,85,86) (this isolate is able to produce infectious virus in CK cells but not Vero cells); (ii) Beau-R (75), a molecular clone of CK cell-adapted Beaudette (87), Beau-CK (88) (Beau-R is able to produce infectious virus on both CK and Vero cells); and (iii) BeauR-M41(S), in which the ectodomain of the Beau-R S gene was replaced with the corresponding sequence of the M41-CK S glycoprotein (74). This rIBV based on the genome of Beaudette, but with the S glycoprotein ectodomain from M41-CK, has the same tropism as M41-CK and is therefore able to produce infectious virus on CK cells but not Vero cells. IBV strains were propagated and titrated in CK cells as described previously (89)(90)(91). Growth curves and confocal microscopy of IBV infection were carried out using CK and Vero cells (74). Vaccinia viruses (VVs) were propagated in Vero cells, and large stocks for DNA extraction were prepared in BHK-21 cells as described previously (76,77).
Although the Beaudette and M41 viruses belong to the same serogroup, Massachusetts, it is important to note that the two viruses are not genetically related, in the sense that one was derived from the other. Comparison of their nucleotide and amino acid sequences show many differences between the two viruses. Unfortunately, there are instances in the literature where it is implied that Beaudette was derived from M41 (63,92), leading to the incorrect conclusion that mutations arising in M41 may have led to the ability of Beaudette to produce infectious progeny in Vero cells. This may have arisen from IBV Beaudette being referred to as IBV-42 (sometimes abbreviated to M42) and M41 as IBV-41 (72,93,94). IBV Beaudette was isolated in the 1930s (87), and M41 was isolated in the 1940s (84). IBV Beaudette is able to grow and produce infectious virus in CK cells and can be also be adapted to Vero cells by repeat passage, as observed by syncytium formation (72). Beau-R, the molecular clone of Beaudette CK, does not produce syncytia when grown on Vero cells (75); however, repeated passage of Beau-R on Vero cells resulted in adaption, as previously observed for Beaudette CK, for syncytium formation on Vero cells (95).