R-Peptide Cleavage Potentiates Fusion-Controlling Isomerization of the Intersubunit Disulfide in Moloney Murine Leukemia Virus Env

Abstract

Fusion of the membrane of the Moloney murine leukemia virus (Mo-MLV) Env protein is facilitated by cleavage of the R peptide from the cytoplasmic tail of its TM subunit, but the mechanism for this effect has remained obscure. The fusion is also controlled by the isomerization of the intersubunit disulfide of the Env SU-TM complex. In the present study, we used several R-peptide-cleavage-inhibited virus mutants to show that the R peptide suppresses the isomerization reaction in both in vitro and in vivo assays. Thus, the R peptide affects early steps in the activation pathway of murine leukemia virus Env.

During maturation of the Moloney murine leukemia virus (Mo-MLV), the viral protease cleaves a 16-residue-long peptide, the R peptide, from the cytoplasmic tail of the TM subunit of the membrane fusion protein Env (5, 6, 17). The cleavage potentiates the receptor-induced fusion activation in Env, but the mechanism for this effect has remained obscure (7, 8, 15, 16, 18). Similar regulation has been observed in other gammaretroviruses and in the Mason-Pfizer monkey betaretrovirus (2, 3). Recently, it was shown that the activation of MLV fusion is also controlled by receptor-induced isomerization of the intersubunit disulfide of the Env SU-TM subunit complex (14, 19). Here, we studied whether R-peptide cleavage facilitates the isomerization.

R-peptide cleavage site mutants, shown to inhibit cleavage in earlier studies, were created at the P1 (L649V, L649R, and L649I) and P1′ (V650I) positions by PCR mutagenesis, using Mo-MLV proviral DNA (4, 8, 16, 18). Corresponding particles were produced and labeled with 50 μCi/ml [³⁵S]Cys in calcium phosphate precipitate-transfected HEK 293T cells. Analysis by reducing sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of sedimentation-purified viruses that had been lysed in buffer containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, and 2 mM EDTA and immunoprecipitated with anti-MLV polyclonal antibody (pAb) (HE863; Viromed Biosafety Laboratories) showed that the mutant Env's, with the exception of V650I, incorporated with wild-type (wt) efficiency and were significantly inhibited in R-peptide cleavage, i.e., in cleavage of the Pr15E form of TM into the p15E form (Table 1). Incorporation of Env-V650I was reduced to 63%, and the R peptide was cleaved as in the wt. The infectivities of the mutants were analyzed in XC cells by using MLV p12 monoclonal antibody (Ab) 548 (B. W. Chesebro, Rocky Mountain Laboratories). Cell binding was augmented by centrifugation of the virus-cell sample at 850 × g for 1 h at 4°C in the presence of 8 μg/ml polybrene. The infectivities of the mutants were found to correlate with the biochemical findings (Table 1). This was also the case with their fusion efficiencies, which were tested in an XC cell-to-cell fusion-from-without assay (19). Thus, all mutants but V650I were inhibited in R-peptide cleavage and membrane fusion. Notably, the cleavage inhibition was only partial, and the L649V mutant contained a significant fraction (about 30%) of R-peptide-cleaved SU-p15E complexes. To differentiate between the alkylated isomerization-arrested SU-Pr15E and SU-p15E complexes (see below), we introduced an anti-R-peptide Ab (K. B. Andersen, The Danish University of Pharmaceutical Sciences). This precipitated only Pr15E from lysed samples of the wt and mutants and SU-Pr15E complexes from samples that were lysed in the presence of N-ethylmaleimide (NEM), which blocked the lysis-induced isomerization by modifying the active thiol in SU (Fig. 1) (13).

TABLE 1.

Features of R-peptide cleavage site mutants

Virus	Env incorporationa (%)	R-peptide cleavageb (%)	Infectivityc (%)	Fusiond (%)
wt	1.0	77 ± 6	100	100
L649V	0.93 ± 0.1	29 ± 3	24 ± 12	10
L649R	0.98 ± 0.1	15 ± 2	20 ± 3	3
L649I	0.86 ± 0.3	17 ± 4	12 ± 2	3
V650I	0.63 ± 0.1	78 ± 4	72 ± 16	100

^aIncorporation was quantified from a phosphorimage of a reduced SDS-PAGE gel of [35^S]Cys-labeled viral proteins immunoprecipitated from sedimentation-purified particles. The amount of viral capsid protein was used for normalization of the samples. Mean efficiencies are relative to that of the wt (± SD; n = 3).

^bCleavage was quantified from the analyses used for the estimation of incorporation. Mean efficiencies are given as percent p15E of total TM (p15E + Pr15E) (± SD; n = 3).

^cInfectivity was quantified in XC cell cultures that had been infected with corresponding numbers of particles. Infected cells were detected by immunofluorescence using anti-p12 monoclonal Ab. Infectivities (± SD; n = 3) were given as percentages of that of the wt.

^dFusion was quantified in XC cell cultures that were subjected to fusion from without with corresponding amounts of particles in several dilutions. Fusion efficiencies were estimated by matching wt and mutant dilutions with similar polykaryon formations and are given as percentages of that of the wt. Three experiments with similar results were performed.

FIG. 1.

Anti-R-peptide Ab captures alkylated isomerization-arrested SU-Pr15E complexes. Proteins of [³⁵S]Cys-labeled wt and mutant viruses that had been lysed in the absence or presence of NEM were immunoprecipitated with anti-Mo-MLV pAb (αMLV) or anti-R-peptide pAb (αR) and analyzed by reducing SDS-PAGE. Shown is the corresponding phosphorimage.

Isomerization of the intersubunit disulfide is mediated by the removal of suppressing Ca²⁺ ion(s) from Env and can therefore be triggered in vitro by subjecting the virus to Ca²⁺ depletion (19). Furthermore, an alkaline pH decreases the triggering threshold of the reaction by increasing the ionization of the isomerization-active CXXC-thiol in SU (9). Therefore, the effect of the R peptide on isomerization was studied by incubating wt and mutant viruses in Ca²⁺-free TN buffer (14 mM Tris, 12 mM HEPES, 150 mM NaCl) (pH 8.0) for 0 to 5 h at 37°C. After the incubation, NEM (20 mM) was added, the samples were lysed, and the viral proteins were immunoisolated with MLV pAb for isomerization analysis by SDS-PAGE at nonreducing conditions. NEM prevented lysis-induced isomerization and thus facilitated the measurement of the incubation-induced isomerization. The latter was measured by monitoring the release of subunits from covalently linked SU-TM complexes. As controls, we used samples that were incubated with NEM. The analyses showed that the kinetics of total Env isomerization was significantly slower in the cleavage-inhibited mutants L649R and L649I than in the wt and the cleavage-competent mutant V650I (Fig. 2A). The decrease was less evident in the less-cleavage-inhibited mutant L649V but was confirmed by quantification (Fig. 2B). To avoid the problem of partial cleavage, we measured the levels of isomerization separately for the SU-Pr15E and SU-p15E complexes by quantifying Pr15E and p15E, which were released from the corresponding covalently linked SU-TM complexes. The ratios of the SU-p15E to SU-Pr15E complexes in the original nonincubated samples were assumed to correspond to those of p15E to Pr15E in the fully isomerized samples incubated for 5 h (Fig. 2A, lanes 4, 8, 12, 16, and 20). The signal levels facilitated reliable quantification of SU-p15E isomerization in wt, L649V, and V650I and of SU-Pr15E isomerization in L649V, L649R, and L649I. The quantifications showed that the SU-p15E complexes of the wt and the mutants isomerized with mutually similar kinetics (Fig. 2C, column groups 1, 2, and 6). The SU-Pr15E complexes of the different preparations also isomerized with mutually similar kinetics but at a significantly slower rate than that of the SU-p15E complexes (Fig. 2C, column groups 3 to 5). Thus, after 1 h of incubation, almost all of the SU-p15E complexes but less than half of the SU-Pr15E complexes had isomerized. We concluded that the R peptide suppresses the Ca²⁺ depletion-triggered isomerization.

FIG. 2.

Suppression of Ca²⁺ depletion-induced isomerization in SU-Pr15E Env complexes. [³⁵S]Cys-labeled wt and mutant viruses were incubated in Ca²⁺-free buffer, pH 8.0, at 37°C for 1 to 5 h. The samples were lysed in the presence of the alkylator NEM, and viral proteins were immunoprecipitated with anti-MLV pAb for analysis by nonreducing SDS-PAGE. Control samples (0 h) were incubated in the presence of NEM. (A) Phosphorimage of the SDS-PAGE gel, with the TM part of the gel in higher contrast below. (B) Total SU-TM isomerization efficiencies (i.e., those of the SU-Pr15E and SU-p15E complexes together). (C) Specific isomerization values for the SU-Pr15E and SU-p15E complexes. All quantitations are given as means ± standard deviations (SD; n = 3). Note that in panel A the released SU migrated closely behind a weak but sharp Gag band.

Receptor-mediated triggering of the isomerization reaction in Env with uncleaved R peptide was studied by following the efficiency by which Env was converted into a CXXC-thiol alkylatable intermediate. After receptor binding and the removal of Ca²⁺, Env obtains an open conformation, which exposes the CXXC-thiol for modification (19, 20). Thus, if Env is activated in the presence of an alkylator, the CXXC-thiol will be modified before it attacks the intersubunit disulfide. As a consequence, isomerization will be blocked, further activation will be arrested, and Env will accumulate as an intermediate (21). Therefore, the mutant and wt viruses were bound by centrifugation to the receptor-positive XC cells on ice and then incubated at 37°C in Dulbecco's phosphate-buffered saline (pH 7.5) for 30 min in the presence of a membrane-impermeable alkylator, M135 (2 mM; Toronto Research Chemicals). After this process, the alkylator was washed off, the virus-cell samples were lysed in the absence of an alkylator, and the viral proteins were immunocaptured for analyses of alkylated, covalently linked SU-TM complexes and free subunits by SDS-PAGE at nonreducing conditions. Because lysis in the absence of an alkylator will induce isomerization in all Env's that have not already undergone receptor-mediated triggering and subsequent alkylation, the amount of covalently linked SU-TM complexes will correspond to the amount of receptor-triggered Env. As controls, we used cell-bound viruses that were kept on ice with M135 for 30 min before being washed and lysed. The control samples showed only free SU and TM, i.e., Pr15E and p15E, but no SU-TM complexes (Fig. 3A, lanes 1, 3, 5, 7, and 9). This indicated that no Env's had been induced by the receptor and subsequently isomerization blocked by the alkylator at these conditions. However, isomerization-resistant SU-TM complexes, i.e., alkylated Env intermediates, appeared in the samples that had been incubated at 37°C in the presence of an alkylator (Fig. 3A, lanes 2, 4, 6, 8, and 10). Notably, the amounts of intermediates were very much reduced in the cleavage-inhibited mutants L649R and L649I and significantly reduced in the less-cleavage-inhibited mutant L649V compared to those of the wt and the cleavage-competent mutant V650I. This finding was confirmed by quantification of the fraction of TM subunits present in alkylated SU-TM complexes (Fig. 3B). To get an estimate of how efficiently the Pr15E-SU and p15E-SU complexes, respectively, were triggered and blocked as intermediates, we measured the amounts of Pr15E and p15E that had been released from the corresponding SU complexes by lysis-induced isomerization (Fig. 3A, lanes 2, 4, 6, 8, and 10) and compared them to the total amounts of Pr15E and p15E in the corresponding nonincubated samples, which had been totally isomerized (Fig. 3A, lanes 1, 3, 5, 7, and 9). The signal levels facilitated reliable quantifications of Pr15E release in L649V, L649R, and L649I and of p15E release in the wt, L649V, and V650I. We found that the lysis-releasable p15E fraction was only 50% to 70%, in contrast to the lysis-releasable Pr15E fraction, which was about 95% (Fig. 3C). This indicated that the SU-p15E complexes were preferentially triggered by the receptor and converted into the alkylated intermediate, whereas the SU-Pr15E complexes were inhibited in this reaction. This was confirmed by an analysis of the alkylated samples with the R-peptide pAb, which reacted only with the R-peptide-containing, lysis-released Pr15E and alkylated Pr15E-SU intermediates (Fig. 3D). Quantification showed that only about 5% to 6% of the Pr15E-SU complexes of the mutants was triggered by the receptor to form the alkylated intermediate.

FIG. 3.

Inhibition of receptor-mediated triggering of isomerization in SU-Pr15E Env complexes. [³⁵S]Cys-labeled wt and mutant viruses were bound to XC cells and then incubated at 37°C for 30 min in the presence of the alkylator M135. Control samples (0 min) were kept on ice in the presence of the alkylator. After the incubation, the alkylator was removed by several washes, and the virus-cell samples were lysed in the absence of the alkylator. (A) Viral proteins were captured with anti-MLV or anti-R-peptide pAb and analyzed by nonreducing SDS-PAGE (the TM part is shown in higher contrast below). (B) The total amount of lysis-resistant SU-TM complexes, i.e., the amount of alkylated Env intermediates (± SD; n = 3), is given for each 30-min sample. (C) Lysis-induced isomerization of SU-Pr15E and SU-p15E Env that had not been receptor triggered and blocked in isomerization (± SD; n = 3). Their reciprocal values indicate the efficiencies by which Pr15E-SU and p15E-SU complexes, respectively, are triggered by the receptor and blocked as Env intermediates. (D) Phosphorimage of the viral proteins captured with the anti-R-peptide pAb from the 30-min samples. At the bottom of the panel are the percentages of Pr15E in complex with SU, i.e., in alkylated-Env (Alkyl. Env) intermediates.

Altogether, we demonstrated that the R peptide suppressed the fusion-controlling SU-TM disulfide isomerase contained within the CXXC-motif of the SU subunit. The R peptide might interfere with the activation of the isomerization-active CXXC-thiol or any preceding step of Env activation, including receptor binding. However, the fact that suppression was also observed in the receptor-bypassing in vitro reaction using Ca²⁺ depletion suggests that another early step or steps are involved. Earlier, it was found that cell-associated Env with the R peptide is unable to support the membrane stalk and pore formation required for membrane fusion, suggesting that the step(s) before TM hairpin completion are affected (12). Several studies have suggested that the R peptide stabilizes the structure of the native Env oligomer, e.g., by participating in coiled-coil formation of the TM cytoplasmic tail or interacting with a cellular factor and that cleavage potentiates activation through its destabilization. Thus, structural differences have been observed in the ectodomain of Env with and without the R peptide (1). Mutations of residues predicted to support R-peptide coiled-coil formation have been shown to rescue fusion of R-peptide-containing Env (11, 18, 23). Finally, the R peptide has been shown to suppress the activation of other viral fusion proteins when fused to their cytoplasmic tails (10, 22). Our present results are also consistent with the stabilization model.