key: cord-290231-4m9lj0uq
authors: Guirakhoo, Farshad; Bolin, Richard A.; Roehrig, John T.
title: The Murray Valley encephalitis virus prM protein confers acid resistance to virus particles and alters the expression of epitopes within the R2 domain of E glycoprotein
date: 1992-12-31
journal: Virology
DOI: 10.1016/0042-6822(92)90267-s
sha: 
doc_id: 290231
cord_uid: 4m9lj0uq

Abstract To study the role of the precursor to the membrane protein (prM) in flavivirus maturation, we inhibited the proteolytic processing of the Murray Valley encephalitis (MVE) virus prM to membrane protein in infected cells by adding the acidotropic agent ammonium chloride late in the virus replication cycle. Viruses purified from supernatants of ammonium chloride-treated cells contained prM protein and were unable to fuse C6/36 mosquito cells from without. When ammonium chloride was removed from the cells, both the processing of prM and the fusion activity of the purified viruses were partially restored. By using monoclonal antibodies (MAbs) specific for the envelope (E) glycoprotein of MVE virus, we found that at least three epitopes were less accessible to their corresponding antibodies in the prM-containing MVE virus particles. Amino-terminal sequencing of proteolytic fragments of the E protein which were reactive with sequence-specific peptide antisera or MAb enabled us to estimate the site of the E protein interacting with the prM to be within amino acids 200 to 327. Since prM-containing viruses were up to 400-fold more resistant to a low pH environment, we conclude that the E-prM interaction might be necessary to protect the E protein from irreversible conformational changes caused by maturation into the acidic vesicles of the exocytic pathway.

For enveloped viruses to enter into their target cells, virus-cell membrane fusion must take place to release the nucleocapsid into the cytoplasm. This occurs either directly at the plasma membrane, as with paramyxoviruses and herpesviruses, or in the endosomal vesicles after the viruses are taken up by receptor-mediated endocytosis, as with togaviruses and orthomyxoviruses (for review see Marsh and Helenius, 1989) . Generally the fusion-mediating envelope proteins are activated by proteolytic cleavage of the spike precursor proteins by a host protease late during virus assembly. Cleavage activation of the glycoproteins involved in fusion has been demonstrated with the gPE2 of murine coronaviruses (Sturman and Holmes, 1984; Sturman et al., 1985) the HA of influenza viruses (Wiley and Skehel, 1987 ) the gP160 of retroviruses (McCune et al., 1988; Willey et al., 1991) and the PE2 of alphaviruses (Lobigs and Garoff, 1990; Lobigs et a/., 1990; White et al., 1983) . The cleaved glycoprotein (e.g., HA) undergoes an irreversible conformational change in the acidic pH milieu of the endocytic vesicles which exposes a hydrophobic fusion site that mediates the virus-cell membrane fusion. This cleavage of the precursor protein is essential for virus infectivity but not particle formation. A Semliki Forest virus (SFV) mutant defective in PE2 cleavage expressed PE2 at the cell ' To whom reprint requests should be addressed.

surface but did not function in low pH-triggered cell-tocell fusion (Lobigs and Garoff, 1990; Lobigs et al., 1990) .

Murray Valley encephalitis (MVE) virus is a member of the Flaviviridae, a family of small, positive-strand RNA viruses. The mature flavivirions contain three structural proteins: capsid (C), membrane (M), and envelope (E), with molecular masses of 14-15, 7-8, and 50-60 kDa, respectively. The E protein is the only structural protein responsible for the induction of neutralizing, hemagglutination-inhibiting, and protective antibodies, it is the receptor-binding protein and is also involved in acid-catalyzed membrane fusion (Heinz and Roehrig, 1990; Guirakhoo et al., 1991) . Little is known about the function of the M protein. The M protein is cleaved from its precursor protein (prM) after the consensus sequence R-X-R/K-R shortly before or after virus release Westaway, 1987) . This cleavage is associated with a rearrangement of the oligomeric structures on the surface of the virion (Wengler and Wengler, 1989) . By using monoclonal antibodies (MAbs) and protease maps, we previously demonstrated that the E glycoprotein of tick-borne encephalitis (TBE) virus undergoes an irreversible conformational change, predominantly in the epitopes of domain A, at mildly acidic pH . Consistent with this, purified TBE virus demonstrated fusion from without (FFWO) of the mosquito cell line C6/36 only after exposure of cell-bound virus to low pH. The FFWO was abolished if the processing of prM to M was interrupted using the acidotropic agent ammonium chloride (Guirakhoo et a/., 1991) .

In this paper we identify those epitopes on the E protein which are covered or dissociated by the presence of prM, demonstrate that prM-containing viruses are more resistant to low pH, localize the site on the E protein which is involved in the E-prM interaction, and finally hypothesize that the association of E-prM is necessary for the authentic expression of the E protein.

Production and analysis of the immature virus particles Aedes albopictus cells (clone C6/36) were grown on microcarriers (Cytodex 2, Pharmacia Fine Chemicals, Upsala, Sweden)' in Dulbecco's minimal essential medium (D-MEM) buffered with 25 mNI HEPES and supplemented with 10% fetal calf serum (FCS, Hyclone Laboratories, Inc., Logan, UT). Cells were infected with MVE virus (strain Ord River) at a multiplicity of infection (m.o.i.) of 10 in D-MEM with 29/o FCS. Ammonium chloride (25 m/M) was added at 48 hr postinfection (p.i.) and the viruses were harvested on Day 6. At this time, ammonium chloride-containing medium was replaced with D-MEM, 2% FCS, without ammonium chloride, and the viruses were reharvested on Day 9 p.i. Viruses were precipitated with 8% polyethylene glycol 8000 (Fisher Scientific, Fair Lawn, NJ), purified by two cycles of glycerol-tartrate gradients (Obijeski et al., 1976) , and resuspended in 0.3 ml of D-MEM. Protein concentrations were determined by the Bio-Rad protein assay (Bio-Rad Laboratories, Richmond, CA). Untreated MVE viruses (without ammonium chloride) were propagated and purified as described above and were used as positive controls in all experiments. For the acid resistance experiments (see below) MVE viruses were also grown in SW1 3 mammalian cells in the presence or absence of ammonium chloride. To quantify the amount of viral structural proteins in each propagation, the purified viruses were subjected to SDS-PAGE on 12.5% acryamide gels (Laemmli, 1970) and proteins were stained using a rapid silver stain kit (ICN, Cleveland, OH) according to the manufacturer's protocols. Gels were scanned by a microprocessor-controled spectrophotometer (Gilford System 2600) and the peaks corresponding to each viral protein were identified and integrated to determine areas under the peaks, and the area ratios of prM to E were calculated.

Purified viruses were diluted geometrically (from 200 to 6.25 pg/mI) in D-MEM containing 1% BSA. Thirty microliters of each dilution was added to prechilled C6/ 36 cells and the fusion assay was performed at pH 5.0. The number of nuclei and the number of cells in five microscopic fields (magnification 1 OO-fold) were counted and the fusion index [l -(number of cells/ number of nuclei)] was calculated.

Capture enzyme-linked immunosorbent assay (ELISA)

Microtiter plates were coated overnight at 4" with a 1:500 dilution of rabbit anti-MVE hyperimmune serum. This dilution of capture antibody was determined by box-titration against M-and prM-containing virus seeds, maximizing and standardizing the concentration of captured antigens. Wells were blocked with 3% goat serum in PBS, a 1:5 dilution of tissue cultures containing prM-or M-MVE virus was added, and plates were incubated for 1 hr at 37". Ascites of the various MAbs were then titrated on both viruses for 1 hr at 37" starting at 1: 100 or 1: 1000 dilution depending on antibody potency. Bounds MAbs were detected by goat anti-mouse alkaline phosphatase conjugate. Quantitation of bound conjugate was determined by adding Sigma 104 substrate and reading absorbance at 405 nm in a Titertek MC spectrophotometer after 30 min. The specificity and the biological activity of all MAbs except 6B4A-10 have been published elsewhere (Hawkes et a/., 1988) . MAb 6B4A-10 was prepared against Japanese encephalitis virus, strain Nakayama, using standard hybridoma techniques.

This MAb reacted with all members of the MVE virus serocomplex, blocked MVE virus-mediated hemagglutination, and neutralized MVE virus infectivity in vitro. The E glycoprotein epitope defined by 6B4A-10 has been designated E-8.

The acid sensitivity of the MVE virus epitopes was determined using a modification of the standard ELISA protocol. Two hundred and fifty microliters of purified MVE virus (100 pg total protein) in 0.2 M Tris-HCI, pH 8.0, was adjusted to pH 6.0 using 120 ~1 of 0.2 M HCI and incubated at 25" for 30 min. Both acid-treated virus and untreated virus (pH 8.0) samples were diluted to 5 ml with carbonate buffer, pH 9.5. lmmulon II microtiter plates were coated with 100 ~1 virus per well and incubated overnight at 4". The plates were blocked for 1 hr at 37" with 3% BSA in PBS. Protein

MAbs were diluted geometrically (from 10 pg/well total protein) and added to each well. After 2 hr at 25", the plates were washed and further incubated with a 1:500 dilution of goat antimouse alkaline phosphatase conjugate. Bound antibody was detected by addition of Sigma 104 substrate, and absorbance was measured at 405 nm after 5 min. Although each antibody was tested in a dilution series, only the results with 1.25 gg antibody are reported.

Analysis of the MAb reactivities with reduced and nonreduced MVE virus PAGE sample buffer (Laemmli, 1970) with or without the reducing agent, 2-mercaptoethanol, was added to 10 pg of purified MVE virus. Samples were boiled for 3 min at 100" and separated on 12.5% SDS gels. Proteins were then transferred to nitrocellulose sheet (Towbin et a/., 1979) and blocked with 3% goat serum. The reactivity of MAbs defining the epitopes E-4b, E-6, or E-8 with reduced or nonreduced forms of MVE virus E glycoprotein was assessed using goat anti-mouse alkaline phosphatase conjugate (Jackson lmmuno Research, West Grove, PA).

Proteolytic digestion of the native MVE virus was performed in D-MEM, pH 7.4, using predetermined concentrations of either trypsin or V8 protease (Sigma Chemical Co.) and the fragments reacting with MAb E-8 were identified by using goat anti-mouse alkaline phosphatase conjugate (see below). For the proteolytic digestion of the predenatured virus, 60 pg of purified MVE virus in D-MEM was precipitated in 60% TCA (end concentration 12%) on an ice bath for 30 min, washed with cold acetone, and resuspended in 50 ~1 0.125 M Tris, pH 6.8, containing 0.5% SDS and 10% glycerol. After 24 hr incubation at 37" samples were boiled for 3 min and subjected to proteolytic digestion using either chymotrypsin or V-8 proteases (Sigma Chemical Co.) at the protein to enzyme ratios of 4:l and 30: 1, respectively, at 37" for 30 min. After addition of SDS and 2-mercaptoethanol (final concentrations 4%) the samples were boiled and separated on 12.5% SDS gels. Fragments were transferred to nitrocellulose membrane by electroblotting (Towbin et al., 1979) . Nitrocellulose was then blocked by 3% goat serum in PBS, pH 7.4, for 1 hr. Strips were cut and incubated with either MAb (defining epitope E-8) or mouse sera which were immunized with E-specific synthetic peptides as follows: MVE 02, amino acids 35-50; MVE 04, amino acids 122-l 41; MVE 06, amino acids 230-251; and MVE 17, amino acids 356-376. The sequence and characteristics of these antipeptides have been published previously (Roehrig eta/., 1989) . After 1 hr at 37", goat anti-mouse alkaline phosphatase conjugate (Jackson lmmuno Research) was added and the immune fragments were visualized by adding BCIP/NBT substrate (KPI, Inc., Gaithersburg, MD). For the amino-terminal sequencing, fragments were prepared as described above except that electroblotting was performed on polyvinylidene difluoride (PVDF) membrane (Immobilon-P, Millipore, Bedford, MA) using CAPS transfer buffer (10 mM CAPS, 0.5 mM dithiothreitol, DTT, and 10% methanol adjusted to pH 1 1 .O with NaCH). Fragments were stained with 0.1% Coomassie blue (Baker G-250) in 50% methanol and 10% acetic acid. The amino-terminal sequences were determined either on an Applied Biosystems 473A (Foster City, CA) or on a Porton Instrument 2090 (Tarzana, CA) protein sequencer using standard Edman Chemistry.

Plaque assay of prM-and M-containing MVE viruses after exposure to different pH buffers Equal amounts of plaque forming units, PFUs (6 x 1 O'/mI for viruses grown in C6/36 cells and 4.5 X 1 O5 for viruses grown in SW13 cells), from the supernatants of ammonium chloride-treated and untreated MVE virus-infected cells were incubated in different pH buffers, ranging from 7.5 to 4.5, using 1 M HEPES or 1 M MES containing D-MEM. After a 15 min-incubation at room temperature, the inoculum was exponentially diluted in BA-1 diluent (Hunt eta/., 1991) and examined for infectivity in Vero cells using the standard plaque assay.

The relationship between prM processing and the fusion activity

To understand the role of prM protein in virus maturation and its interaction with the E glycoprotein, we investigated the effect that ammonium chloride had on MVE viruses grown in C6/36 mosquito cells. Ammonium chloride has been shown to accumulate in the post-Golgi acidic vesicles and interrupt the processing of prM to M protein (Randolph eta/., 1990) . The unprocessed prM protein was incorporated into the released viruses grown in the presence of ammonium chloride at 6 days p.i. (Fig. 1 A, lane 2) . When ammonium chloride was removed, the cells partially recovered, and the normal processing of the prM to M resumed (Fig. lA,  lane 3) . Since some unprocessed prM proteins were also found in MVE-infected cells in the absence of ammonium chloride ( Fig. 1 A, lane l) , we attempted to quantify the amount of prM in each sample. This was achieved by scanning the stained viral proteins following separation by SDS-PAGE. By using nonradioactive viruses it was not always possible to clearly demonstrate the presence of the small M protein in minigels. The fact that the ratios of the viral structural proteins in the absence of ammonium chloride remain constant enabled us to evaluate the amount of prM processing indirectly by calculating its ratio to other structural proteins (E or C) which are more abundant in virions. The ratio of the area of each peak corresponding to C or prM proteins to E as well as prM to C protein was calculated. The ratio of C to E protein remained constant in the presence or absence of ammonium chloride (data not shown) whereas the ratio of prM to E protein was significantly increased in the presence of ammonium chloride (from 0.15 to 0.86) and reduced to 0.55 following removal of ammonium chloride (Fig. 1A) . Similar results were obtained when the ratio of prM to C was calculated (data not shown). These data confirm that partial recovery of prM processing has occurred once the ammonium chloride has been removed.

To determine the fusion activity of the viruses obtained from the supernatants of virus-infected ammonium chloride-treated cells, the FFWO assay was performed using monolayers of C6/36 cells as described previously (Guirakhoo et a/., 199 1) . No polykaryocyte formation was observed (up to 200 pg/ml of purified virus) using viruses harvested in the presence of ammonium chloride at 6 days p.i.; however, when ammonium chloride was removed, the fusion activity of puri-fied viruses obtained on Day 9 was significantly restored ( Fig. 1 B) .

These results are in agreement with our previous finding with TBE virus which demonstrated a direct relationship between processing of prM and fusion activity of E glycoprotein (Guirakhoo eta/., 1991) and further manifest that both phenomena are reversible.

Epitope analysis of the E-proteins of prM-and Mcontaining viruses A capture ELISA was performed to analyze the effect that blocking the prM cleavage had on the expression of E glycoprotein epitopes in native virions. The reactivities of MAbs defining nine distinct epitopes on the MVE E glycoprotein were compared on M-and prMcontaining viruses using supernatants of ammonium chloride-treated or untreated virus-infected C6/36 cells. The antigenicity of the prM-and M-containing virus supernatants was equivalent in end-point ELISA antigen titration using a 1:500 dilution of polyclonal rabbit anti-MVE virus capture sera and anti-MVE virus mouse hyperimmune ascitic fluid (HIAF) as detector. Therefore, a 1:5 dilution of each supernatant was used in subsequent epitope mapping with MAbs. The equivalent antigenicity was confirmed by end-point ELISA titration of the MVE HIAF in the standardized capture assay (Fig. 2B) . Even though these virus supernatants had equal antigenicity, the prM-containing virus supernatant was 1 O-fold less infectious than the M-containing virus supernatant (3.9 X 1 O* vs 2.3 X 10' PFU/ml). The reactivities of three E glycoprotein epitopes were significantly reduced in prM-containing viruses (Fig.  2C) . Two of these epitopes were subcomplex-and complex-reactive (E-6 and E-8) and one was groupreactive (E-4b), being conserved in all flaviviruses (Hawkes et al., 1988) . The reactivities of other MVE virus-specific MAbs were identical with both viruses (Fig. 2A) .

Structural properties of the epitopes corresponding to these MAbs were assessed by comparing the reactivities of MAbs after incubation of the purified viruses at low pH to those of untreated viruses (Fig. 3) . As can be seen, the antigenic reactivities of all MAbs and the HIAF were significantly reduced upon low pH treatment of the virus.

of the E-8 MAb binding site on the E glycoprotein Figure 4A demonstrates the reactivities of the three MAbs defining the epitopes E-4b, E-6, and E-8, with the MVE virus E glycoprotein under reducing and nonreducing conditions. As can be seen neither the E-4b nor E-6 MAbs recognized the E protein in its re- duced form (treated with 2-mercaptoethanol), whereas E-8 Mab was still reactive after protein reduction. This reactivity was further sustained after the reduction and alkylation of the cysteine residues of the E glycoprotein using DlT/SDS followed by iodoacetamide treatment (Johnson et a/., 1990) (data not shown). Since the reactivity of the E-8 MAb was reduced with prM-containing viruses and the fact that it could recognize the E protein in immunoblots in its linear conformation, we attempted to obtain more information about the site involved in prM-E interaction by identifying the E-8 binding site. The proteolytic digestion of the native E protein with trypsin (which cleaves after K or R) or V8 endoproteinase (which cleaves after E or D) did not produce fragments smaller than 30 kDa that were reactive with E-8 MAb (Fig. 4B) . In order to produce smaller fragment reactive with the E-8 MAb, the MVE virus was denatured by SDS prior to protease digestion. Using chymotrypsin (which cleaves after F, T, or Y) or V8 protease, 20-and 22-kDa fragments, Fr 20 and Fr 22, were identified, respectively. These two fragments, which reacted with E-8 MAb, were also recog-nized by mouse antiserum immunized with synthetic peptide MVE 06 (Figs. 4C and 4D). A 23-kDa fragment, Fr 23, and a 19-kDa fragment, Fr 19, were produced using V8 or chymotrypsin, respectively, which were recognized by MVE 04 antibodies but not by MVE 06 antibodies or E-8 MAb. Antisera prepared against MVE 17 also recognized a V8 fragment of 16 kDa Fr 16 (Fig.  4D ). Fr 23, 22, 16, and 14 were sequenced at their amino-terminal ends and the results are shown in Ta (Dalgarno et a/., 1986) predict only one potential Nlinked glycosylation site on the E protein at position N-l 54 and the fact that this site might contain carbohydrate of the complex type (endo H resistant) (Winkler Because it has been shown with West Nile virus that the E and prM glycoproteins remain as heterodimers during transport and maturation through acidic compartments of the exocytic route (Wengler and Wengler, 1989) , we compared the infectivity of the prM-containing MVE virus with that of the M-containing MVE virus following virus exposure to acidic conditions. Equal PFUs of the prM-and M-virus containing supernatants from C6/36 or SW1 3 cells were subjected to environments of different pH values and their resultant infectivity in a plaque assay was examined (Fig. 6) . No significant difference in infectivity of either virus preparation was observed after the incubation of viruses at pH values between 7.5 and 6.5. Upon incubation at pH 6.0, the infectivity of both M-containing viruses (grown in C6/36 or SW1 3 cells) was drastically abolished; however, this loss of infectivity was less dramatic for prMcontaining viruses than for M-containing viruses (150fold loss versus 1360-fold loss for viruses grown in C6/ 36 cells, Fig. 6A tion 199) or Fr 16 (starting at position 327), we predict its binding site to be between amino acids 200 and 327 on the E glycoprotein (Fig. 5) .

Low pH sensitivity of prM-and M-containing MVE viruses (Table 1) have been marked by arrowheads. 

a Fragments were generated by V-8 digestion of SDS-denatured purified MVE virus, separated on 12.5% SDS gels, and transferred to PVDF membranes by Western blotting. Membranes were stained with Coomassie blue and subjected to amino-terminal sequence analysis.

' Fragments were generated as described above, transferred to nitrocellulose, and subjected to immune detection using MVE antipeptrde antibodies (P2. P4, P6. P17; for sequence and length see Materials and Methods) or Mab E-8 as described under Materials and Methods, for viruses grown in SW1 3 cells, Fig. 6B ). Maximal loss of infectivity was achieved upon incubation at pH 4.5; nevertheless, the prM-containing viruses were still 1 Ofold (those grown in C6/36 cells) and 400-fold (those grown in SW1 3 cells) more resistant to the low pH environment than the M-containing viruses.

The proteolytic cleavage of prM occurs after a dibasic amino acid motif, presumably by a trypsin-like host enzyme similar to those responsible for normal processing of prohormones (Fischer and Scheller, 1988; Steiner e2 al., 1984) . Recently, an enzyme designated paired basic amino acid cleaving enzyme (PACE) has been cloned and purified from a human liver cell line (Wise et al., 1990) . A similar endoprotease was also isolated from the trans-Golgi membranes of rat liver cells, which was shown to cleave in vitro the F, protein of New Castle disease virus after a dibasic residue motif (Sakaguchi et al., 199 1) . We believe that the late maturational processing of prM at the R-X-R/K-R cleavage motif is the mechanism used to activate the flavivirus fusion protein (E glycoprotein). The infectivity of prM-containing viruses (grown in C6/36 cells) for mammalian cells was lo-fold lower than that of the mature MVE (M-containing) viruses, although the amount of antigen detected by polyclonal antibodies was similar for both viruses. These results indicate that prM processing is required for full virus infectivity. A similar observation has been made with West Nile (WN) virus in which the ratio of physical particles to PFU was about 600 for the fully infectious extracellular (M-containing) viruses and was about 60-fold higher for the cell-associated (prM-containing) viruses (Wengler and Wengler, 1989) .

Upon subsequent infection and endocytosis, the E glycoprotein probably undergoes a conformational change in the acidic environment of endocytic vesicles, which results in membrane fusion and release of the nucleocapsid into the cytoplasm. Consistent with this hypothesis are our previous findings that the E glycoprotein of TBE virus undergoes irreversible conformational changes at pH less than 6.2. These changes, which predominantly occur in the A-domain epitopes and the i2 site Mandl et al., 1989) , are probably necessary to trigger the fusion activity of the E glycoprotein. It has been shown that the fusion activity of the TBE virus can be blocked effectively by three MAbs, two of which were specific for epitopes of the A-domain, and one recognized the i2 epitope (Guirakhoo et al., 1991) . The A-domain contains an amino acid sequence (G L F G) that is conserved in the fusion sequences of other enveloped viruses. We have also observed with dengue virus that the region containing the conservative fusion sequence is more accessible to antipeptide antibodies following low pH treatment . By using MAbs specific for MVE virus E glycoprotein in a capture ELISA, we demonstrated that if the processing of prM was interrupted at least three of these epitopes on the E glycoprotein were less accessible on the virus particle. The observation that all of the epitopes were acid sensitive indicates that they are probably located in the A-domain. The sensitivity of the HIAF to low pH also suggests that these epitopes are the predominant epitopes recognized following immunization with intact virus. It may be significant that these three epitopes are associated with hemagglutination of red blood cells. At least one of these epitopes (E-8) also elicits virus-neutralizing antibodies. Since other functionally important epitopes which induce neutralizing, hemagglutinationinhibiting, and protective antibodies (Hawkes et a/., 1988) are still present in the prM-containing viruses (Fig. 2 A) , and the fact that these viruses are still capable of binding to and infecting their host cells, we did not expect that the fusion peptide or the receptorbinding sites would be directly involved in the prM-E interactions. This was confirmed when the immunereactive fragments generated by V8 protease were sequenced. It was shown that predominantly three sites on the E glycoprotein were exposed to the proteolytic attack of the V8 enzyme. This cleavage, which occurred on positions 138, 199, and 326, generated four fragments reactive with sequence-specific antipeptide antibody or E-8 MAb. Similar segments on the E pro-tein of WN virus, which are located in Ll and R3, have also been shown to be exposed to proteolytic attack (Wengler er a/., 1987) . Combining results of the reactivity pattern of the antipeptide antibodies with proteolytic fragments and also information obtained from N-terminal sequence analysis of the immune fragments, we were able to map the binding site of the E-8 MAb which was covered or dissociated by the presence of prM. This site which reside between amino acids 200 and 327 on the E protein includes more than 80% of the R2 domain proposed for WN (Nowak and Wengier, 1987) and MVE (Roehrig et a/., 1989) viruses. No actual biological function of the R2 domain had been previously identified. It is noteworthy that the synthetic peptide MVE 06, which consists of 21 amino acids within the R2 domain (230-251), was cross-reactive with other members of the flaviviruses (Roehrig et al., 1989) , indicating the biological importance of this region.

Several studies have indicated that when influenza A virus and Semliki Forest virus were incubated at low pH they became irreversibly inactivated (Doms et a/., 1985; Kielian and Helenius 1985; Stegman eta/., 1989; White et a/., 1982) . This is probably due to exposure and self-aggregation of the hydrophobic fusion peptide which, therefore, can no longer interact with target membranes (Nir et al., 1990) . If the flavivirus E glycoprotein were not protected during transport to the cell surface, the acidic milieu of the post-Golgi vesicles could inactivate it. Because it has been shown with WN virus that the E and prM proteins remain as heterodimers during exocytosis (Wengler and Wengler, 1989) we reasoned that the function of the prM is to prevent conformational changes in the E glycoprotein at low pH, thus assuring safe transport of these proteins through post-Golgi acidic vesicles during maturation until the virions are released. A similar role has also been proposed for the PE2 of alphaviruses, which forms a stable dimer with the El protein (critical for infection and fusion) in the acidic compartment of the exocytic route (Lobigs and Garoff, 1990; Presley et a/., 1991; Wahlberg et al., 1989) . Different approaches might be utilized by other viruses to secure the authentic expression of the envelope proteins on the infected cells. It has been shown that the M2 protein of influ-enza virus raises the pH of the post-Golgi vesicles by about 0.8 unit, thus preventing the expression of the low pH conformation of the HA molecules on the plasma membranes of the infected cells (Ciampor et a/., 1992) . The high resistance of the prM-containing viruses to low pH might indicate that the prM could shield some segments of the E protein which are otherwise protonated in a low pH environment, leading to the exposure of the fusion peptide and consequent loss of infectivity. Concomitant with the low pH sensitivity of the M-containing viruses was a loss of epitope reactivity in ELISA following incubation at pH 6.0 or below. Because prM-containing viruses are still infectious, apparently the prM protein can be cleaved after the virus has undergone endocytosis. Using chymotrypsin, trypsin, and thermolysin it was demonstrated that the prM protein of the WN virus is highly sensitive to proteases . In our laboratory we also observed that by using a low concentration of trypsin, the prM protein of the MVE virus is cleaved prior to the E protein (data not shown).

Our results predict that the expression of authentic E glycoprotein requires the coexpression of prM. This would explain why flavivirus vaccine candidates produced from the expression of only E glycoprotein were partially protective against homotypic challenge, whereas those prepared from the expression of both prM and E glycoproteins gave a significant level of protection (Bray and Lai, 1991) . Similar results were observed with Japanese encephalitis virus, where high neutralizing and hemagglutination-inhibiting antibodies which correlated with a high level of protection were found in mice inoculated with recombinant vaccine viruses expressing both prM and E but not E pro-tein alone (Konishi et al., 1991; Yasuda et a/., 1990) .

We are currently performing experiments to define further the role of prM in virus infectivity and the events occurring after prM-containing viruses enter their target cells. We are also attempting to determine if a similar area on the E protein of other flaviviruses is involved in the prM interaction.

Dengue virus premembrane and membrane proteins elicit a protective immune response. L/iro/ogy 185

Regulation of pH by the M2 protein of influenza A viruses

Partial nucleotide sequence of the Murray Valley encephalitis virus genome: Comparison of the encoded polypeptides with Yellow Fever virus structural and non-structural proteins

Membrane fusion activity of the influenza virus hemagglutinin: The low pH-induced conformational change

Prohormone processing and the secretory pathway

Epitope model of tick-borne encephalitis virus envelope glycoprotein E: Analysis of structural properties, role of carbohydrate side chain, and conformational changes occurring at acidic pH

Fusion activity of flaviviruses: Comparison of mature and immature (prM-containing) tick-borne encephalitis virions

Antigenic structure of the Murray Valley encephalitis virus E glycoprotein

Flaviviruses: The basis for serodiagnosis and vaccines

Synthetic peptides of the E2 glycoprotein of Venezuelan equine encephalomyelitis virus. II. Antibody to the amino terminus protects animals by limiting viral replication

Variants of Venezuelan equine encephalitis virus that resist neutralization define a domain of the E2 glycoprotein

pH-Induced alterations in the fusogenic spike protein of Semliki Forest virus. /. Cell f?iol

Comparison of protective immunityelicited by recombinant vaccinia viruses that synthesize E or NSl of Japanese encephalitis virus

Cleavage of structural proteins during the assembly of the head of bacteriophage T4

Fusion function of the Semliki Forest virus spike is activated by proteolytlc cleavage of the envelope glycoprotein precursor ~62

Spike protein ollgomenzation control of Semliki Forest virus fusion

Antigenic structure of the flavivirus envelope protein E at the molecular level, using tick-borne encephalitis virus as a model

Japanese encephalitis virus-vaccinia recombinants produce particulate forms of the structural membrane proteins and induce high levels of protection against lethal JEV infection

Endoproteolytic cleavage of gpl60 is required for the activation of human immunodeficiency virus

Fusion of enveloped viruses with cells and liposomes: Activity and Inactivation. ln

Analysis of disulfides present In the membrane proteins of the West Nile flavivirus

Structural proteins of La Crosse virus

Proteolytic processing of the Sindbis virus membrane protein precursor PE2 is nonessential for growth in vertebrate cells but is required for efficient growth in Invertebrate cells

Acidotropic amines inhibit proteolytic processing of flavivirus prM protein. Viral-ogy174

Structure of the flavivirus genome

Synthetic peptides derived from the deduced amino acid sequence of the E-glycoprotein of Murray Valley encephalitis virus elicit antiviral antibody

Antibodies to dengue 2 virus E-glycoprotein synthetic peptides identify antigenic conformation

ldentiflcatton of endoprotease activity in the rrans Golgi membranes of rat liver cells that specifically processes in vitro the fusion glycoprotein precursor of virulent Newcastle disease virus

Membrane fusion activity of influenza virus: Effects of gangliosides and negatively charged phospholipids in target liposomes

Golgiigranule processing of peptide hormone and neuropeptide precursors: A minireview

Proteolytic cleavage of peplomeric glycoprotein E2 of MHV yields two 90K subunits and activates cell fusion

Proteolytic cleavage of the E2 glycoprotein of murine coronavirus: Activation of cell-fusing activity of virions by trypsin and separation of two different 90K cleavage fragments

Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications

The heterodimerit association between the membrane proteins of Semliki Forest virus changes its sensitivity to low pH during virus maturation

Cell-associated West Nile flavivirus is covered with E + Pre-M protein heterodimers which are destroyed and reorganized by proteolytic cleavage during virus release

Analysis of the influence of proteolytic cleavage on the structural organization of the surface of the West Nile flavivirus leads to the isolation of a protease-resistant E protein oligomer from the viral surface

Flavivirus replication strategy

Hemagglutinin of influenza virus expressed from a cloned gene promotes membrane fusion

Membrane fusion proteins of enveloped animal viruses

The structure and function of the hemagglutinin membrane glycoprotein of influenza virus

Mutations within the human immunodeficiency virus type 1 gpl60 envelope glycoprotein alter its intracellular transport and processing

Studies on the glycosylation of flavivirus E proteins and the role of carbohydrate in antigenic structure

Expression of a human proprotein processing enzyme: Correct cleavage of the von Wilebrand factor precursor at a paired basic amino acid site

Induction of protective immunity in animals vaccinated with recombinant vaccinia viruses that express PreM and E glycoproteins of Japanese encephalitis virus

We thank Craig Miles and Cindy Bozic of Macromolecular Resources, Colorado State University, and Danny Jue and Tamara Hine of the Biotechnology Core Facility, Centers for Disease Control, for amino-terminal sequencing of the enzymatically derived E glycoprotein fragments.