key: cord-0701379-mvh4ig2g
authors: Suñé, Carlos; Jiménez, Gustavo; Correa, Isabel; Bullido, María J.; Gebauer, Fatima; Smerdou, Cristian; Enjuanes, Luis
title: Mechanisms of transmissible gastroenteritis coronavirus neutralization
date: 1990-08-31
journal: Virology
DOI: 10.1016/0042-6822(90)90521-r
sha: 7018a22df61dff269b8f16195ef95b118158463d
doc_id: 701379
cord_uid: mvh4ig2g

Abstract Transmissible gastroenteritis virus (TGEV) was neutralized more than 109-fold with antibodies of a single specificity [monoclonal antibodies (MAbs)]. Most of the virus was neutralized in the first 2–3 min of a reversible reaction, which was followed by a second phase with a decreased neutralization rate and, in some cases, by a persistent fraction, which was a function of the MAb and of the antibody-to-virus ratio. Neutralization of TGEV is a specific event that requires the location of the epitope involved in the neutralization in the appropriate structural context, which is present in the wild-type virus but not in certain MAb escaping mutants. In neutralization of TGEV by binary combinations of MAbs specific for the same or for different antigenic sites, either no cooperation or a synergistic effect, respectively, was observed. Mechanisms of TGEV neutralization by MAbs were characterized at high, intermediate, and low antibody-to-virus ratios. Under these conditions, mainly three steps of the replication cycle were inhibited: binding of virus to the cell, internalization, and a step that takes place after internalization. In addition, virus aggregation could be responsible for the neutralization of 10 to 20% of virus infectivity.

Porcine transmissible gastroenteritis (TGE) is an enteric disease of swine associated with a high mortality in neonates, caused by a coronavirus that consists of a single-stranded positive-sense RNA molecule of 26-30 kb and three structural proteins S, N, and M, with 1447, 382, and 262 amino acids, respectively (Siddell et al., 1983 ; Sturman and Holmes, 1983 ; Salt and Bahl, 1986 ; Kapke and Brian, 1986 ; Laude et al., 1987 ; Rasschaert and Laude, 1987) . In coronaviruses the S protein is responsible for the induction of neutralizing antibodies (Garwes et al., 1976 (Garwes et al., , 1978 Sturman and Holmes, 1983) . Studies on the antigenic structure of prototype coronaviruses (mouse hepatitis virus, infectious bronchitis virus, and TGEV) have shown that at least two critical sites are involved in their neutralization (Talbot et al., 1984 ; Wage et al., 1984 ; Mockett et al ., 1984 ; Fleming at al., 1986 ; Delmasatal., 1986) . In TGEV only the S glycoprotein has been involved in virus neutralization in the absence of complement . In this protein, a minimum of seven antigenic sites have been defined ; site A is the major inducer of neutralizing antibodies, although other sites (B and D) , less immunogenic, are also involved in TGEV neutralization (Enjuanes et al., 1990 ; Delmas et al., 1986 ; Correa et al., 1990 ; . In addition, antibody-mediated complement-dependent virus neutralization has been reported with M-specific ' To whom requests for reprints should be addressed . 559 monoclonal antibodies (MAbs) (Woods et al., 1987 ; Laude et al., 1986) .

Virus neutralization can be achieved by inhibition of any step of virus replication (Dimmock, 1984 (Dimmock, , 1987 . In TGEV, very little is known about the mechanisms of neutralization . One report (Nguyen et al ., 1986) has suggested that the neutralization of TGEV by secretary IgA from milk or IgG from serum is due to inhibition of the internalization of the adsorbed virus, but not to blockage of virus binding to the host cell .

In this paper, we characterize the mechanisms for neutralization of TGEV by monoclonal and polyclonal antibodies, at high, intermediate, and low antibody-tovirus ratios .

Cells and virus. The epithelial swine testicle cell line ST developed by McClurkin and Norman (1966) was used to grow the Purdue strain of TGEV (PUR46) (Bullido et al ., 1988 ; Sanchez et al., 1990) , which was plaque purified (Jimenez et al., 1986) .

Virus neutralization and purification . Two procedures were followed in the plaque reduction assay . In one, a fixed amount of TGEV PFU were incubated with the antibody, and dilutions of the virus-antibody mixture were plaqued . In the second, 10-fold dilutions of the virus were incubated with a fixed amount of antibody, and the mixtures were applied to the cells without further dilution . Both plaque reduction assays were performed by incubating, at 37° for 30 min (unless otherwise indicated), the TGEV in 50 pl of phosphate-SUWE ET AL. buffered saline (PBS) containing 2% fetal calf serum and 1 vol of the MAb or antiserum dilution . Portions containing 50 µl of each mixture were applied to cell monolayers grown in 24-well microplates . After 1 hr of virus adsorption, the inoculum was replaced with medium containing 2% fetal calf serum, 40 µg of DEAEdextran/ml, and 0 .1 % agarose, and the cells were incubated at 37° for 2 days in a humidified CO 2 incubator . Cells were fixed with 10% formaldehyde and stained with 0 .1 % crystal violet, and the plaques were counted . The extent of neutralization was expressed by the neutralization index (NI), which is the log of the ratio of the number of PFU of virus to that of control medium or antibody .

[3 H]Leucine-or [35 S]methionine-labeled TGEV was grown in Dulbecco's modified Eagle (DME) medium completely or partially depleted of leucine or methionine, respectively, in the presence of fetal calf serum dialyzed against PBS, and 20 µCi of [ 3 H]leucine (177 Ci/mmol, New England Nuclear) or 20 µCi of [35 S]methionine (800 Ci/mmol, SJ204, Amersham), respectively, per milliliter .

TGEV was purified through two sucrose gradients in 1 M NaCl, the first one containing 0 .2% Tween 20, as previously described .

Radioimmunoassay. Binding of the antibody to the virus was determined in a radioimmunoassay (RIA) in which TGEV (0 .25 µg of protein per well in 50 µl of PBS) was adsorbed to polyvinyl disposable flat-bottom plates (Titertek, Flow Laboratories) by overnight incubation at 37°. The subsequent steps of the assay were performed as described by Sanz et al . (1985) , using a rabbit antiserum to mouse immunoglobulins to amplify the reaction and 125 1-labeled protein A to develop the assay . The titer in the RIA was defined as the maximum antibody dilution that bound threefold the background radioactivity .

Polyclonal antiserum. TGEV-specific swine antiserum was induced by immunizing 6-week-old minipigs (haplotype cc) (Sachs et al., 1976) with three doses of 35µg of purified TGEV, in complete Freund's adjuvant, incomplete adjuvant, and saline solution . The serum had a titer by RIA of 78,125 and its NI was >7 .1 .

Monoclonal antibodies and "mar" mutants . MAbs and MAb-resistant (mar) mutants were produced and characterized as described previously (Jimenez et al ., 1986 ; Correa et al., 1988) . The properties of the MAbs used in this study are summarized in Table 1 . The mar mutants isolated were neutralized less than 10°' S -fold bythe homologous MAb, which neutralized the original virus more than 10 4°-fold .

Kinetics of the neutralization . Hybridoma culture supernatants were titrated in the RIA and diluted to the same titer (2000) . Samples (200 µl) of each hybridoma   TABLE 1   CHARACTERISTICS OF THE TGEU-SPECIFIC MAbs USED IN THIS STUDY   a The first and second capital letters indicate the virus protein and the antigenic site, respectively, and the small letter indicates the subsite (Correa et S., 1988) .° Not determined .

were equilibrated at 37° and a volume of the virus (5 x 10' PFU) was added to each MAb . At the indicated times, 20-µ1 samples were diluted 100-fold in PBS with 2% fetal calf serum at 4° and immediately titrated on ST cells . Binding of TGE virus to ST c ells. S T cells were grown to confluence in flat-bottom 96-well microplates (Falcon Plastics) . All remaining steps were performed at 4°. Nonspecific binding sites were blocked by incubating each well for 15 min in the presence of 200 µ1 of PBS containing 5 mg of bovine serum albumin (BSA) per milliliter . The blocking solution was removed, and 100 µI of 3 H-labeled purified TGEV (1 x 10 4 to 2 x 10 4 cpm, 1 .9 x 10 4 cpm/µg of protein, 1 X 10 1 PFU/µg) was added per well in DME medium (with no bicarbonate), supplemented with 2 TIM glutamine, 50 µg of gentami- cin/ml, 50 mM Hepes, and 10% fetal calf serum, pH 7 (binding medium) . The microplates were incubated for 120 min or the indicated times . Unbound virus was collected and the wells were washed three times with 200 Al of 2% fetal calf serum in PBS . Cell-bound radioactivity was solubilized with 100 pl of 1 % Nonidet P-40 in PBS per well, samples of this material and of the supernatants were precipitated with 5% trichloroacetic acid, and the radioactivity was evaluated by liquid scintillation techniques .

To determine the inhibition of binding of virus to ST cells induced by MAbs, the assay was performed as described, except that samples of 50 pl with the indicated amounts of each antibody were incubated at 4°f or 15 min with 1 vol of the labeled virus . Then, the mixture was incorporated into the wells . The percentage of specific binding, that is, the binding of labeled virus to the cells that was inhibited by unlabeled virus, obtained in the presence of an inhibitor MAb was calculated with the formula % binding = cpm°cpm°X 100 cpm, -cpm°c pm 0 and cpm, are the radioactivity associated with the cells in the presence and the absence of inhibitor MAb, respectively ; and cpm°is the radioactivity associated with the cells in the presence of a 40-fold excess of unlabeled virus . The radioactivity bound, within the different batches of virus, ranged from 10 to 40% of the total labeled virus added to the cells . This amount was taken as the reference against which to calculate the binding inhibition by different antibodies or competitor virus .

Internalization of TGEV in ST cells . Internalization of TGEV (or its binding to ST cells, when it was correlated with internalization) was studied on ST cell monolayers grown on 24-well microtest plates (Costar) . Mixtures of 35 S-labeled purified virus (1 .5 x 10 4 cpm, 1 .2 x 10 4 cpm/µg, 1 .4 x 10 8 PFU/µg) and antibody dilutions in 100 gl of binding medium were incubated at 4° for 15 min, added to each well, and incubated at 37° for 60 min (unless otherwise indicated), on a rocking platform, in a CO 2 incubator . The plates were cooled on ice, the supernatants were collected, the cells were washed once with binding medium, and the supernatants were combined . To determine the amount of cell-associated virus, the monolayers were rinsed twice more with binding medium . Cell-associated radioactivity was determined by solubilizing the monolayers with 1 % Nonidet P-40 in PBS . TCA-insoluble radioactivity was determined in samples from the supernatants as well as from the cell monolayers .

To determine the internalized virus, after it was bound to the cells, the monolayers were washed once TGE CORONAVIRUS NEUTRALIZATION MECHANISMS 5 6 1 with 40 mM citric acid, 10 mM KCI, 135 mM NaCl, pH 3, at 4° for 1 min and immediately washed a second time with binding medium (Highlander etal., 1987) , and the radioactivity was determined as indicated above .

TGEV neutralization by MAbs . The majority of TGEV was neutralized (Fig . 1 ) by individual murine MAbs in the first 2-3 min of the reaction, following an exponential decline in surviving virus, which was followed by a second-phase reaction with a decreased rate of neutralization, leaving in some cases a persistent fraction that remained constant after 2 hr of incubation (results not shown) . The extent of neutralization in the first minutes of the reaction was similar to that obtained ( Fig . 1 ) when the virus adsorption period was reduced to 5 min . Extent of virus neutralization was a function of the MAb (Fig . 1A ) and, for each MAb, it was related to its concentration ( Fig . 1 B) . Culture supernatants containing MAb I E .F9 neutralized virus infectivity 100-fold, while 10 -' or 10 -3 dilutions of this MAb reduced infectivity 10-fold or virtually not at all, respectively . Neutralization of TGEV by MAbs was reversed between 1 023 and 10 6 '-fold (Table 2 ) by diluting the virus-antibody complexes and incubating to allow their dissociation, which indicates that TGEV neutralization is a reversible process .

Characterization of TGEV and of 11 mar mutants derived from it, by RIA and neutralization using MAbs (Fig .  2) , showed that some MAbs (for example, MAb 1 B .C1) that bound and neutralized the original strain of TGEV also bound, but did not neutralize, the mar mutants, indicating that neutralization requires more than simple binding of the antibody to the critical epitope .

To determine if MAbs specific for the different antigenic subsites involved in TGEV neutralization cooperated in this process, inactivation by all possible binary combinations of saturating amounts of six MAbs, which bind to three antigenic subsites (subsite Aa : MAbs 1G .A6, 1C .C12, and 1B .C1 ; subsite Ab : MAb 1D .E8 ; subsite Ac : MAbs 1A .F10 and 6A .C3) was studied . The neutralization by any pair of MAbs tested resulted in a nonsignificant cooperation . In contrast, when site A-specific MAbs (1G .A7, 1B .C1, 1D .E8, 1A .F10, and 6A .C3) were combined with MAb 5G1 specific for a site different from site A (R . Woods, R . Wesley, and L . Enjuanes, unpublished results), a synergistic effect was observed, which produced an increase of 2 in the NJ above the NI obtained with the best neutralizing MAb individually used (results not shown) . was saturable, reaching a plateau 90-120 min afterthe reaction was started ( Fig . 3A ), when free virus was still available . The binding was specific, as it was inhibited up to 90% by a 40-fold excess of unlabeled virus, but°

The NI was determined either immediately after neutralizing the virus or after diluting 100-fold the virus-antibody mixture in PBS with 295 fetal calf serum and incubating for 3 hr at 4°.°

The decrease in the NI was determined by subtracting the NI of the virus-antibody mixture diluted 100-fold and incubated for 3 hr to allow dissociation of the complex, from the NI forthe mixture immediately titrated . ble 3) was best inhibited by site A-specific MAbs (78-96%), followed by site D (53-55%)-, site B (9-39%)-, and site C (24-26%)-specific MAbs . As expected, the M-and N-specific MAbs did not inhibit binding . A general direct correlation between the blocking activity of a MAb and its capability to neutralize the virus was observed (Table 3) . MAb 1 B .G4 has low neutralizing activity (NI 0 .3) and inhibited binding 65%, possibly because it is large (IgM isotype), and not as a consequence of the proximity of the epitope bound by the MAb to the determinant recognized by the cellular receptor . The effect of the antibody-to-virus ratio on TGEV neutralization and on inhibition of virus-to-cell binding was studied (Figs . 4A, B) . Interestingly, MAbs specific for two critical antigenic subsites of S protein did not inhibit at all the binding of TGEV to ST cells at 4°, at concentrations at which they neutralized the virus infectivity 10 5fold, indicating that the two activities of the MAbs (binding inhibition and neutralization) can be split .

Internalization after 1 -2 hr of incubation at 37° (Fig . 5A) . Virus internalization proceeded with 15 to 30 min of delay and reached the maximum 2 hr later . Virus internalization was temperature dependent (Fig . 5B) , being maximal at 37°. More than 90% of the virus was internalized through a specific receptor, as this process was efficiently inhibited by a 40-fold excess of unlabeled virus and by site A-specific MAbs (i .e ., 1C .C12), but not by MAbs specific for other sites (i .e ., 6A .A6) (Fig . 5B) .

TGEV neutralization at 37° was observed (Fig . 6 ) at low antibody concentrations (dilutions between 1 in 64 and 1 in 256) which only slightly prevented internalization of the virus and which increased virus binding to cells . At intermediate antibody concentrations (dilution of 1 in 32), the antibody inhibited mainly the internalization, with little blocking of the binding . At high antibody concentrations (dilutions lower than 1 in 4), both internalization and binding were inhibited and extensive (between 10 2 -and 10 6 -fold) neutralization was observed .

MAbs specific for three subsites of S glycoprotein still neutralized 75 to 97% of the virus after it was cell associated (Table 4) . Under these conditions about 10% of the virus was not exposed at the cell surface since it was not detachable with a low-pH buffer (see Materials and Methods) (Fig . 5B) , which may be the To determine if there was a correlation between inhibition of binding and internalization, and the neutralization of TGEV by polyclonal antibodies, the effect of different antiserum concentrations on these parameters was determined . The results (Fig . 7) showed that the inhibition of internalization and that of binding were the same, and that extensive (10 2-fold) neutralization was observed at antibody dilutions (i .e ., I in 80) for which no inhibition of either binding or internalization was observed . Furthermore, at higher dilutions (between 1 in 80 and 1 in 640), at which the antiserum produced an increase in binding and in internalization of the virus (Fig . 7) , significant neutralization was observed, Although dilutions between 1 in 80 and 1 in 640 of the control serum induced a small increase in the binding of the virus, no increase in its internalization and neutralization were noted . 

It has been shown that TGEV can be neutralized by antibodies of a single specificity, in a reversible reaction, which is followed by a second phase with a decreased neutralization rate and, in some cases, by the onset of a persistent fraction . In addition, the predominant mechanisms of TGEV neutralization have been determined . Neutralization of TGEV by MAbs occurs rapidly (2-3 min) . Adsorption of the virus was performed after diluting 100-fold the antibody-virus mixture, and it was shown (Table 2) that the neutralization of TGEV by the MAbs used could be reversed by incubating the diluted mixtures, and never increased . In fact, when the time of adsorption was reduced to 5 SUNS ET AL . min, no significant decrease in the extent of neutralization was observed .

The critical epitopes must be in a certain structural context for an effective neutralization (Fig, 2) , since although certain epitopes were present in some mar mutants, as demonstrated by the binding of MAbs to these viruses (Fig . 2B) , they were not neutralized ( Fig .  2A) , in contrast to the neutralization of the wild-type virus . Probably the structure of the antigenic subsite is changed in the mutant . A similar phenomenon has been described for poliovirus, in which the loss of the neutralization function of an epitope does not necessarily result in loss of its antibody-binding capacity . Such epitopes binding MAbs that neutralize the wt virus, but which do not neutralize the mar mutants, exist naturally on Mahoney and Sabin 1 viruses (Blondel et at, 1986) .

The neutralization of TGEV by the MAbs was reversed by dilution (Table 2 ) . In contrast, the neutralization of poliovirus is apparently an irreversible process which is not reversed by dilution (Icenogle et at, 1983) . These results probably reflect the avidity of the antibodies used in each case, as the neutralization reaction is reversible in most viral systems (Mandel, 1979) .

Very little cooperation in TGEV neutralization was observed between site A-specific MAbs, even when MAbs reacting to different antigenic subsites were used . This result seems logical, as the binding of a site A-specific MAb prevented the binding of a second one specific for the same site . In contrast, a synergistic effect was observed between MAbs binding to different sites involved in neutralization (results not shown), as could be expected for MAbs which did not interfere in their binding to the virus .

Coronaviruses, such as mouse hepatitis virus and infectious bronchitis virus, attach to host cells through the S glycoprotein (Sturman and Holmes, 1983 ; Cavanagh and Davis, 1986 ; Holmes et at, 1989) . Similarly, this must also be the case for TGEV, as MAbs specific for the S glycoprotein, but not MAbs specific for the N or M protein, inhibited the binding to ST cells . MAbs specific for antigenic sites A and D were the best inhibitors of binding (Table 3 ), suggesting that the domain recognized by the cellular receptor on ST cells must be located spatially close to these sites .

To study the relationship between neutralization and binding inhibition by TGEV-specific MAbs (Table 3) , it was essential to show first that the binding of virus to ST cells was specific . Three types of results indicated that this was the case : (i) the binding was saturable (Figs . 3A and 5A) ; (ii) binding of 3 H-labeled or of 35 Slabeled virus was inhibited up to 90% by a 40-fold higher concentration of the unlabeled one, but not by the serologically unrelated mouse hepatitis coronavi- Fig . 5B ) .

The internalized virus was defined as the amount of radiolabeled virions that remained cell associated after incubation in the presence of a pH 3 buffer . Use of a low-pH buffer to differentiate between exposed and internalized virions has previously been described (Huang and Wagner, 1964 ; Highlander etal., 1987) . In the TGEV system this assay was valid on the following bases ( Fig . 5B) : (i) cell viability was not affected by incubation with the pH 3 buffer ; (ii) by incubating at 4°, the virions were bound but not internalized, because at this temperature, cell-bound virions are not detached from cells with PBS but are removed with pH 3 buffer ; (iii) after incubation at 37° the virions were internalized, because after this incubation, the virions were not detached by washing either with PBS or with the pH 3 buffer ; (iv) the amount of virus remaining cell associated after treatment with the pH 3 buffer was dependent on the temperature of incubation, being maximum at 37°C, as expected for internalized virions ; (v) the internalization defined by this procedure was a specific process, since it was inhibited by MAbs specific for site A but not for site C, and by unlabeled TGEV, but not by the serologically unrelated MHV .

A correlation has been established between radiolabeled and infectious virions on the following basis . A decrease of 10 5 -fold in virion infectivity was detected in the absence of antibody, and normally was around 30-40% of the total radioactivity added . (Fig . 4) , indicating that 99 .999% of the virions were neutralized, while the binding of radiolabeled virions to the cells was inhibited 0% . The simplest explanation is that neutralization under those conditions was not due to blocking of the binding . The possibility that only the binding of the infectious virus was inhibited, which would represent a minor proportion, cannot be excluded . Nevertheless, it seems unlikely since the noninfectious virus was rendered noninfectious by different defects . A large proportion (up to 40%) of the virions were specifically bound to the cells . Obviously, the infectious virions were included among them . These results strongly suggest that both infectious and noninfectious virions have a functional spike protein mediating the binding of the virus to the cells . The limitations derived from the inhability to follow isolated infectious virions have been accepted in all the studies analyzing the interaction between viruses and cells, by electron microscopy or using radiolabeled virions (Fuller and Spear, 1985 ; FulleretaL, 1989 ; Highlander et al., 1987 ; Miller and Hutt-Fletcher, 1988) . In all cases, the observed virions, or most of the radioactivity, corresponds to noninfectious particles .

The number of antibody molecules interacting with each virus particle during neutralization can be estimated assuming that 1% of the coronavirus particles are infectious . This value was estimated to be between 4 x 10 3 and 1 x 10 4 immunoglobulin molecules of the MAbs 6A .C3 and 1 C .C12, pervirion, for high antibody- Kinetics of TGEV binding and internalization, optimum temperature for these processes, and specific inhibition . (A) Binding and internalization of 35 S-labeled purified TGEV were studied on ST cell monolayers grown on 24-well plates, by adding the virus (see Materials and Methods), incubating at 37°, immediately cooling the plates on ice at the indicated times, and removing the unbound virus by washing with binding medium to determine the cell-associated virus (U) or with pH 3 buffer to determine the internalized virus (•) . (B) Internalization of 35 S-labeled purified TGEV on ST cell monolayers grown on 24-well plates was determined at the indicated temperatures, in the presence of undiluted hybridoma supernatants containing the TGEV S protein site C-specific MAb 6A .A6 (shaded columns), the site A-specific MAb 1 C .C 12 (hatched columns), or 40-fold excess of unlabeled TGEV (blank columns) . The extent of internalization was not affected by the presence of a control coronavirus (MHV) or by the absence of a MAb in the medium (results not shown) . Maximum radioactivity bound or internalized at 37° in the absence of antibody and competing virus was set as 100% binding or internalization, respectively, and usually was 50% or 37%, respectively, of the total radioactivity added .

to-virus ratios (undiluted supernatants) . Assuming that TGEV has 10 3 spike glycoproteins per virion [the same number as influenza virus (Dimmock, 1984) ], at high antibody :virus ratios there will be 4 or 10 molecules of the MAb 6A .C3 or 1C .C12 per glycoprotein, which could interfere with virus-to-cell binding .

To analyze the neutralization of TGEV by MAbs, three situations have to be considered with respect to the antibody-to-virus ratio . At high ratios there was a SUNS ET AL . good correlation between the neutralizing activity of a MAb and its ability to inhibit the binding of TGEV to ST cells, as mast MAbs fit this correlation (Table 3) , indicating that at high antibody concentrations the binding of virus to the cellular receptor was prevented and a pseudoneutralization resulting from steric hindrance probably occurred (Mandel, 1979) . This observation was confirmed by studying the effect of different MAb concentrations on binding, both at 4° (Fig . 4) and at 37° (  Fig . 6) . As expected, neutralization was slightly more efficient at 37°. This is a phenomenon observed with TGEV on a regular basis .

At intermediate antibody-to-virus ratios, the predominant cause of virus neutralization was, probably, the inhibition of internalization, as the virus bound to and accumulated in the cell surface ( Fig . 6 and results not shown) .

At low antibody-to-virus ratios (Fig . 4) , TGEV was neutralized 105 -fold, and 100% of the virus was still cell associated, indicating that the step inhibited at this dilution ought to be different from the binding . At these concentrations, the binding of virus to ST cells was clearly increased, probably as the result of some aggre- Relationship between neutralization of TGEV by MAbs and inhibition of its binding and internalization on ST cells . 3°S-labeled purified TGEV was incubated at 4° for 15 min with the indicated dilutions of MAb 1C .C12, and the virus-MAb mixtures were added to ST cell monolayers grown on 24-well plates, which were incubated at 37° for 1 .5 hr. Cell-associated (•) or internalized virus (•) was determined after washing the monolayers either with binding medium or with a pH 3 buffer, respectively . Residual infectivity (A) was determined on the same virus preparation by performing dilutions after the incubation at 4°, which were immediately used to infect cell monolayers grown on 24-well plates . The results shown represent the average values from three experiments, which, individually considered, gave the same pattern . , respectively .° Neutralizing activity of the MAbs was determined by two procedures : (i) by first adsorbing different numbers of PFU to the cells, incubating 60 min at 4°, adding the MAb, incubating 60 min at 4°, removing the excess MAb, washing with medium, adding the culture medium, and incubating at 37°for plaque development ; (ii) by mixing the MAb with virus, incubating at 4°for 15 min, adding the mixture to the cells, incubating different numbers of PFU with the MAb for 60 min at 4°on the cell monolayer, washing with medium, and proceeding as above .

of TGEV at low antibody concentrations was not due to the blocking of the binding, once the virus was bound, it could still be neutralized . In fact, 75 to 95% of TGEV already bound to ST cells, was neutralized by MAbs (Table 4 ) . Under these conditions (4°C), about 10% of the virus has been protected from cell dissociation at pH 3 ( Fig . 5B ) (Highlander at al ., 1987) , and was not accessible to MAb, which prevented the extensive neutralization observed when the virus interacted first with the antibody (Table 4) .

The results at low antibody-to-virus ratios (Fig . 6 ) suggest that all neutralization is not due to inhibition of virus internalization . The neutralization detected at these ratios could in part be the result of the aggregation of a portion (10-20%) of the virus caused by MAbs 1 C .C 12 and 6A .C3, as determined by electron microscopy (results not shown) . Also, the alterations observed in virus structure could be responsible for the inhibition of internalization and other steps taking place after this . In other viral systems such as adenovirus, neutralization was due to the blocking of a replication cycle step that occurs after internalization . In fact, adenovirions already attached to HeLa cells could be neutralized with anti-hexon or anti-penton base antisera, which did not prevent its internalization (Wohlfart, 1988 ; Wohlfart et al., 1985) . Similarly, influenza virus neutralized by saturating amounts of monoclonal IgG directed against the hemagglutinin enters the cells and undergoes primary but not secondary uncoating in vivo (Rigg et al., 1989) . In the herpes simplex virus system, MAbs have been described with potent neutralizing activity not associated with the inhibition of virus adsorption but with the inhibition of internalization (Fuller and Spear, 1985 ; Fulleretal., 1989) .

Virus neutralization and binding assays were conducted with the same virus and MAb stocks and under very similar conditions . One difference between assays was that for neutralization, the virus-antibody mixtures were diluted and, under those conditions, some neutralized virus might have been dissociated from the antibody, recovering its infectivity (Table 2 ) . These results imply that the neutralization values presented are minimum, and that the correlation between neutralization and binding inhibition is even lowerthan the one shown in Figs . 4, 6, and 7, strengthening the conclusion that direct inhibition of binding is not the mechanism of viral neutralization at low antibody-to-virus ratios .

In summary, neutralization of TGEV, and probably of other coronaviruses, may be the consequence of the inhibition of several steps of its replication cycle . These viruses may be neutralized by a few antibody molecules bound per virion, because of the small aggregation detected, the slight inhibition of internalization, and the blocking of a step that would take place after internalization . At intermediate antibody :virus ratios, inhibition of internalization is the predominant mechanism of neutralization . When the studies are performed at high antibody concentrations, the binding of virus is prevented, no further step takes place, and binding inhibition appears to be the mechanism of neutralization . The study was performed as indicated in the legend to Fig . 6 . Circles, squares, and triangles indicate cell-associated virus, virus internalized, and residual infectivity, respectively, after treatment with TGEVspecific antiserum (full symbols) or with the serum from nonimmune animals (blank symbols) .

The study of the inhibition of virus-to-cell binding and of the internalization, using MAbs, permitted the selective inhibition of each of these steps of the virus cycle ( Fig . 6) , while, with polyvalent antibodies (Fig . 7) , this differentiation was not possible . Murine (results not shown) or porcine polyvalent antisera which neutralized TGEV also inhibited the binding of virus to ST cells at high concentrations . In contrast, at lower antibodyto-virus ratios the virus could be neutralized by polyclonal antiserum without inducing inhibition of binding or internalization . These results appear at variance with those of Nguyen et al . (1986) , who reported that neutralization of TGEV by polyvalent antibodies is due to inhibition of internalization .

We are grateful to J . A . Garcia for critical review of the manuscript, and to C . Sanchez and J . Paladin for their excellent technical assistance .

QS ., G .1 ., I .C ., M .J .B ., F .G ., and C .S . hold fellowships from the Ministry of Education and Science . This investigation was funded by grants from the Consejo Superior de Investigaciones Cientificas, Comision Interministerial de Ciencia y Tecnologia, the European 100 a f I SUNS ET AL .

Economic Community (Project BAP .0464 .E gR)), and Fundacion Ramon Areces .

Mutations conferring resistance to neutralization with monoclonal antibodies in type 1 poliovirus can be located outside or inside the antibody-binding site

Induction of transmissible gastroenteritis coronavirus neutralizing antibodies "in vitro" by virus-specific T helper cell hybridomas

Coronavirus IBV : Removal of spike glycopolypeptide 31 by urea abolishes infectivity and haemagglutination but not attachment to cells

Localization of antigenic sites of the E2 glycoprotein of transmissible gastroenteritis coronavirus

Antigenic structure of the E2 glycoprotein from transmissible gastroenteritis coronavirus

Antigenic structure of transmissible gastroenteritis virus . II . Domains in the peplomer glycoprotein

Mechanisms of neutralization of animal viruses

Multiple mechanisms of neutralization of animal viruses

Location of antigenic sites of the S-glycoprotein of transmissible gastroenteritis virus and their conservation in coronaviruses

Pathogenicity of antigenic variants of marine coronavirus JHM selected with monoclonal antibodies

Neutralizing antibodies specific for glycoprotein H of herpes simplex virus permit viral attachment to cells but prevent penetration

Specificities of monoclonal and polyclonal antibodies that inhibit adsorption of herpes simplex virus to cells and lack of inhibition by potent neutralizing antibodies

Antigenicity of structural components from porcine transmissible gastroenteritis virus

Isolation of subviral components from transmissible gastroenteritis virus

Neutralizing monoclonal antibodies specific for herpes simplex virus glycoprotein inhibit virus penetration

Coronavirus receptors

Penetration of herpes simplex virus into human epidermoid cells

Neutralization of poliovirus by a monoclonal antibody : Kinetics and stoichiometry

Critical epitopes in transmissible gastroenteritis virus neutralization

Sequence analysis of the porcine transmissible gastroenteritis coronavirus nucleocapsid protein gene

Antigenic structure of transmissible gastroenteritis virus . I . Properties of monoclonal antibodies directed against virion proteins . )

Sequence and Nterminal processing of the transmembrane protein El of the coronavirus transmissible gastroenteritis virus

Interactions of viruses with neutralizing antibodies

Studies on transmissible gastroenteritis of swine . II . Selected characteristics of a cytopathogenic virus common to five isolates from transmissible gastroenteritis

A monoclonal antibody to glycoprotein gp85 inhibits fusion but not attachment of Epstein-Barrvirus

Monoclonal antibodies to the SI spike and membrane proteins of avian infectious bronchitis coronavirus strain Massachusetts M41

Neutralizing secretory IgA and IgG do not inhibit attachment of transmissible gastroenteritis virus

Linear neutralizing epitopes on the peplomer protein of coronaviruses

The predicted primary structure of the peplomer protein E2 of porcine coronavirus transmissible gastroenteritis virus

IgG-neutralizing influenza virus undergoes primary, but not secondary uncoating in vivo

Transplantation in miniature swine . I . Fixation of the major histocompatibilrty complex

Transmissible gastroenteritis

Antigenic homology among coronaviruses related to transmissible gastroenteritis virus

Monoclonal antibodies specific for African swine fever virus proteins

The biology of coronaviruses

The molecular biology of coronaviruses

Topographical mapping of epitopes on the glycoproteins of murine hepatitis virus-4 (Strain JHM) : Correlation with biological activities

Hybridoma antibodies to the murine coronavirus JHM : Characterization of epitopes on the peplomer protein (E2)

Neutralization of adenoviruses : Kinetics, stoichiometry, and mechanisms

Interaction between HeLa cells and adenovirus type 2 virions neutralized by different antisera

Complementdependent neutralization of transmissible gastroenteritis virus by monoclonal antibodies