Functional domains of the pp60v-src protein as revealed by analysis of temperature-sensitive Rous sarcoma virus mutants.

Four temperature-sensitive (ts) Rous sarcoma virus src gene mutants with lesions in different parts of the gene represent three classes of alteration in pp60src. These classes are composed of mutants with (i) heat-labile protein kinase activities both in vitro and in vivo (tsLA27 and tsLA29), (ii) heat-labile kinases in vivo but not in vitro (tsLA33), and (iii) neither in vivo nor in vitro heat-labile kinases (tsLA32). The latter class indicates the existence of structural or functional pp60src domains that are required for transformation but do not grossly affect tyrosine kinase activity.     

Reconstitution of the Rous sarcoma virus transforming protein pp60v-src into phospholipid vesicles.

An artificial membrane system was developed to study the molecular basis for interaction of pp60v-src, the Rous sarcoma virus transforming protein, with lipid bilayers. pp60v-src was extracted from cell membranes by detergent solubilization and reincorporated into phospholipid vesicles. Reconstituted pp60v-src retained tyrosine kinase activity and was integrally associated with the liposome through a 10-kilodalton (kDa) amino-terminal domain. The same 10-kDa domain was shown to anchor pp60v-src to the plasma membrane of transformed cells. Reconstitution experiments performed with nonmyristylated pp60v-src proteins revealed that these polypeptides did not interact with phospholipid vesicles. In contrast, myristylated, soluble pp60v-src molecules (including a highly purified pp60v-src preparation) could be reconstituted into liposomes, but their interaction with the liposomal bilayer was not mediated by the 10-kDa amino-terminal domain. When membrane proteins were included during reconstitution of purified pp60v-src, binding through the 10-kDa anchor was restored. A model is presented to accommodate the different types of interactions of pp60v-src with liposomes; the model postulates the existence of an additional membrane component that anchors the pp60v-src polypeptide to the phospholipid bilayer.     

Forms of pp60v-src isolated from Rous sarcoma virus-transformed cells.

It has previously been shown that an electrophoretic variant form of the Rous sarcoma virus transforming protein, pp60v-src, exists in src-transformed cells. This variant, which was readily observed in vanadate-treated cells, was characterized as possessing extensive amino-terminal domain phosphotyrosine modification. Its appearance was further correlated with increased src-specific protein kinase activity. In this study, we used a src-specific monoclonal antibody (MAb) to resolve immunologic forms of pp60v-src. The MAb was able to distinguish between two populations of typical lower-band pp60v-src and was unreactive with the electrophoretic variant upper-band pp60v-src species. Using serial immunoprecipitations, we resolved four populations of pp60v-src: src protein either immunoreactive or unreactive with the MAb from both untreated and vanadate-treated transformed cells. The pp60v-src in each fraction displayed a distinct phosphoamino acid composition and tryptic phosphopeptide profile. However, analysis of their tyrosyl kinase specific activities showed that the immunologically resolved populations of pp60v-src from a given culture did not differ. Both pp60v-src fractions from vanadate-treated cells exhibited similar kinase specific activities, which were greatly enhanced over those of enzyme preparations from untreated cells. Since the MAb-reactive pp60v-src fraction from vanadate-treated cells lacked the electrophoretic variant upper-band pp60v-src species yet still possessed enhanced enzymatic specific activity, the initially stated correlation between the appearance of the electrophoretic variant src form and increased src kinase activity breaks down. These results suggest that yet to be defined modifications of the src protein may be involved in its functional regulation.     

Cooperative transformation studies with temperature-sensitive mutants of Rous sarcoma virus.

Stocks of Rous sarcoma virus Bryan strain were mutagenized using a bromodeoxyuridine treatment immediately after infection. Thirty temperature-sensitive (ts) mutants defective in transformation (td) were isolated by a replica plating technique. Twenty of these mutants were preliminarily characterized and found to be defective in late functions related to transformation. These mutants were used in experiments of cooperative transformation with four Prague strain td ts mutants of different co-transformation group. A small number of Bryan ts mutants were found to cooperate with some of the Prague mutants in transforming chicken embryo cells at the nonpermissive temperature. However, the amount of co-transformation observed was lower than that observed with cooperating Prague ts mutants and no clear-cut pattern of cotransformation was obtained in Prague and Bryan crosses. Indirect evidence indicates that cooperative transformation is the result of recombination events.     

Distinguishable transformation-defective phenotypes among temperature-sensitive mutants of Rous sarcoma virus.

Eight transformation-defective, temperature-sensitive (ts) mutants of the Prague strain of Rous sarcoma virus, subgroup A, have been isolated after mutagenesis with 5-bromodeoxyuridine followed by selection on the basis of focus tests. Five of these mutants, ts GI201, GI202, GI203, GI204, and GI205, exhibit properties like most previously reported isolates in that they show a temperature-sensitive response to each of a variety of transformation-specific parameters tested. Interestingly, GI201, in addition to the temperature-sensitive defect, carries a lesion that was observed as a nonconditional loss of expression of plasminogen activator protease. Three mutants, ts GI251, GI252, and GI253 have been disignated partial transformation-defective (PTD) mutants since they behave as ts mutants according to some tests for transformation and as wild type according to others. These three mutants fail to form foci at the nonpermissive temperature (41 degrees C) and art nontumorigenic in 3-week-old chickens (body temperature, 42 degrees C). The agglutinability by concanavalin A of cells infected with these mutants shows a definite temperature sensitivity, as do the rate of 2-deoxyglucose uptake and the disappearance of the 250, 000-dalton normal cell glycoprotein (large, external, transformation sensitive [LETS]). Although the PTD mutant-infected cells, unlike cells infected with other transformation mutants, exhibit a cell-bound plasminogen activator protease at the nonpermissive temperature, this activator is not detectable as a free protease in the medium, as it is with wild-type, virus-infected cells. The PTD mutants behave like the wild-type parent in their ability to induce transformed growth properties in the infected cells, i.e., growth beyond normal cell saturation density with or without serum-supplemented medium and growth leading to colony formation in soft-agar- or methyl cellulose-containing suspension media.     

Transformation parameters and pp60src localization in cells infected with partial transformation mutants of Rous sarcoma virus.

Rous sarcoma virus (RSV)-induced transformation is mediated by the action of the viral src gene product pp60src. This transforming protein is found at several cytoplasmic locations, including the adhesion plaques of RSV-transformed cells. In these studies, we have focused on the adhesion plaque location of pp60src and determined whether any of the induced transformation parameters correlate with the presence of pp60src in the adhesion plaques. A series of partial transformation mutants of RSV that induce distinct transformation phenotypes were used, and infected chicken embryo cells were examined for (i) intracellular pp60src location, (ii) vinculin localization, (iii) abundance of phosphotyrosine on vinculin, (iv) integrity of stress fibers, and (v) expression of cell surface fibronectin. The results indicate that, among the limited number of mutants studied here, the presence of pp60src in adhesion plaques is independent of growth in soft agar and the increased phosphorylation of vinculin on tyrosine, but it does correlate with the loss of cell surface fibronectin. An elevated abundance of phosphotyrosine on vinculin is insufficient to cause stress fiber dissolution and is independent of the loss of fibronectin from the extracellular matrix. However, the increased relative amount of phosphotyrosine on vinculin is related to the ability of the cells to grow in soft agar. The adhesion plaque binding and tyrosine-specific kinase activities seem to represent two independent functions of pp60src.     

Synthesis and processing of viral glycoproteins in two nonconditional mutants of Rous sarcoma virus.

We have studied the pattern of glycoprotein synthesis in two nonconditional mutants of Rous sarcoma virus. One mutant, SE33, produces no viral particles but synthesizes Pr92env, which is cleaved intracellularly to mature glycoproteins. The second mutant, SE521, encodes a gPr92env which is not cleaved to gp85 or gp37 and therefore produces virions with the phenotype of Bryan RSV(-) or NY8. Neither of these mutants have detectable genomic deletions. The study of these mutants has led to the following conclusions. (i) In the absence of particle production or p15 synthesis, gPr92env can be cleaved to the mature glycoprotein which is found on the cell surface. (ii) Noncleaved gPr92env is not packaged into virions but is found on the cell surface. (iii) gPr92env alone can account for subgroup specific viral interference. (iv) gPr92env is probably transported to the cell surface before additional glycosylation or cleavage to mature virion glycoprotein. The nonprocessed precursor of SE521 appears to be glycosylated normally, and thus far we have been unable to determine the basis for the defect in this mutant.     

Structurally and functionally modified forms of pp60v-src in Rous sarcoma virus-transformed cell lysates.

When analyzed from transformed cell lysates, pp60v-src, the product of the Rous sarcoma virus src gene, typically appears as a single polypeptide of 60,000 molecular weight, phosphorylated at two major sites, an amino-terminal region serine residue and carboxy-terminal region tyrosine residue. We describe here the identification of variant forms of pp60v-src present in transformed cell lysates that exhibited an altered electrophoretic mobility in sodium dodecyl sulfate-polyacrylamide gels. This change in migration appeared to be the result of some alteration in the amino-terminal portion of the molecule and paralleled the appearance of extensive amino-terminal region tyrosine phosphorylation on the pp60v-src molecule. These structural modifications were further correlated with a dramatic increase in the protein kinase-specific activity of pp60v-src. The detection of these variant forms of pp60v-src depended on the prior treatment of the transformed cell cultures with vanadium ions or the inclusion in the cell disruption buffer of Mg2+ or ATP-Mg2+. The implications is that modified, highly active forms of the pp60v-src protein exist in transformed cells, but are transient and rapidly converted to stable forms, possibly by specific dephosphorylation. We suggest that amino-terminal region tyrosine phosphorylation of pp60v-src, presumably the result of autophosphorylation, serves to greatly enhance src protein enzymatic activity, but that much of the regulation of this transforming protein's function may involve a phosphotyrosyl protein phosphatase.     

Highly specific antibody to Rous sarcoma virus src gene product recognizes a novel population of pp60v-src and pp60c-src molecules

Antiserum to the Rous sarcoma virus (RSV)-transforming protein, pp60v-src, was produced in rabbits immunized with p60 expressed in Escherichia coli. alpha p60 serum immunoprecipitated quantitatively more pp60v-src than did tumor-bearing rabbit (TBR) sera. When RSV-transformed cell lysates were preadsorbed with TBR serum, the remaining lysate contained additional pp60v-src, which was recognized only by reimmunoprecipitation with alpha p60 serum and not by TBR serum. In subcellular fractions of RSV-infected chicken embryo fibroblasts (RSV-CEFs) and field vole cells probed with TBR serum, the majority of the pp60v-src was associated with the plasma membrane-enriched P100 fraction. However, alpha p60 serum revealed equal distribution of pp60v-src and its kinase activity between the P1 (nuclear) and P100 fractions. The same results were obtained for pp60c-src in uninfected CEFs. On discontinuous sucrose gradients nearly 50% of the P1-pp60v-src sedimented with nuclei, in fractions where no plasma membrane was detected. Indirect immunofluorescence microscopy of RSV-CEFs with alpha p60 serum revealed a distinct pattern of perinuclear fluorescence, in addition to staining at the cell periphery. Thus the use of a highly specific antibody reveals that enzymatically active pp60v-src and pp60c-src molecules are present in other intracellular structures, probably juxtareticular nuclear membranes, in addition to the plasma membrane in normal, uninfected, and wild-type RSV-infected cells.     

Study of pp60v-src protein kinase activity in synchronized chicken embryo fibroblasts infected with Rous sarcoma virus

Chicken embryo fibroblast (CEF) cultures, synchronized by the addition of serum to stationary cells, were exposed to Schmidt-Ruppin strain of Rous Sarcoma Virus (SR-RSV) and the appearance of pp60v-src protein kinase activity was examined through the cell cycle. In cells infected either at the beginning or at the end of G1, the onset of pp60v-src protein kinase activity was coincidental, closely following mitosis, with a delay between the infection of cells with SR-RSV and the appearance of protein kinase activity of about 20 and 16 h, respectively. In cells infected during the S phase this delay was 16 h, as observed for late G1 cells. These experiments show that the activity of pp60v-src protein kinase, which cannot be detected before the first mitosis following infection does not depend on G1. The aphidicolin prevented protein kinase activity if added before or at the beginning of S phase, but not if added later, which is presumably related to the inhibition of S phase, required for provirus integration. The use of colcemid, which suppresses cell division, did not inhibit but delayed the appearance of protein kinase activity. These results show that the synthesis of an active oncogene product, such as pp60v-src protein kinase, depends on both S phase and mitosis.     

Inter- and intramolecular interactions of highly purified Rous sarcoma virus-transforming protein, pp60v-src

We have purified intact pp60v-src, the product of the Rous sarcoma virus src gene, over 2400-fold, based on the phosphorylation of tumor-bearing rabbit IgG. The purification procedure involved detergent extraction of the particulate fraction of the cells and sequential chromatography on hydroxylapatite, butyl agarose, DEAE-Sephacel, ADP-agarose, and Sephacryl S-200. Analysis of the preparation by sodium dodecyl sulfate-polyacrylamide gel electrophoresis revealed a single silver-stained band with an apparent molecular weight of 60,000. Our results show that the activities of this preparation were qualitatively similar to those described previously for partially purified pp60v-src. Upon analysis by two-dimensional gel electrophoresis, the purified pp60v-src yielded one major species which migrated to the same position as the least acidic of the three major species detectable in cellular lysates, suggesting that the pp60v-src had been dephosphorylated during the purification procedure. We found that pp60v-src was very prone to aggregation; to maintain it as a monomer both Nonidet P-40 and KCl were required. Under conditions which maintained pp60v-src as a monomer, the rate of autophosphorylation was independent of its concentration and thus proceeded via an intramolecular process. Preincubation of pp60v-src with ATP or GTP as well as nonphosphorylating analogs of ATP or GTP preserved its phosphorylating activity toward alpha-casein whereas its activity was reduced 80% upon preincubation in the absence of nucleotides. We suggest that protection with nucleotides rather than autophosphorylation accounts for the apparent increase in the activity of pp60v-src after incubation of the enzyme with ATP.     

Glutamic acid decarboxylase activity is stimulated in quail retina neuronal cells transformed by Rous sarcoma virus and is regulated by pp60v-src.

Rous sarcoma virus (RSV) stimulates in quail embryo neuro-retina (NR) cultures the specific activity of glutamic acid decarboxylase (GAD), the enzyme responsible for the synthesis of gamma-aminobutyric acid, a major inhibitory neurotransmitter in NR and in central nervous system. In quail embryo NR cultures transformed by ts NY-68, a thermodependent transformation-defective mutant of RSV, stimulation of GAD activity is regulated by pp60v-src, the product of the src gene of RSV. Fibroblasts and myoblasts have a very low GAD activity that is not stimulated after transformation by RSV. Neuronal clones, previously derived from ts NY-68-transformed established NR cell lines, have a high GAD activity which is regulated by pp60v-src, while other clones have a low GAD activity apparently not regulated by pp60v-src. These data indicate that pp60v-src selectively activates the expression of GAD in distinct neuronal cells of quail embryo NR cultures transformed by RSV. GAD activity is also stimulated in NR cells infected with viruses containing v-mil.     

The specific interaction of the Rous sarcoma virus transforming protein, pp60src, with two cellular proteins

Sera from rabbits bearing tumors induced by Rous sarcoma virus (RSV) were previously found to contain antibody to the RSV transforming protein, pp60^src. Two additional transformation-specific phosphoproteins from RSV-transformed avian cells are immunoprecipitated with these sera. These proteins, having molecular weights of 90,000 (pp90) and 50,000 (pp50), are not precipitated from uninfected or transformation-defective virus-infected cells and are not related to any RSV structural proteins. Neither pp50 nor pp90 shares any partial or complete proteolytic cleavage peptides with pp60^src, suggesting that pp90 and pp50 do not represent either a precursor or a cleavage product of pp60^src. Sedimentation analysis of RSV-transformed cell lysates on glycerol gradients revealed that the RSV pp60^src protein is present as two forms, one of which represents the majority (95%) of pp60^src and sediments as a monomer, 60,000 molecular weight protein and the other of which sediments with pp90 and pp50 as an apparent 200,000 molecular weight complex. Lysates from cells transformed by viruses containing a temperature-sensitive defect in the src gene contain a greater percentage of pp60^src associated with pp90 and pp50 under both permissive (35°C) and nonpermissive (41°C) conditions compared to wild-type virus-infected cell lysates. Phosphoserine and phosphotyrosine were found associated with pp60^src molecules that sedimented as a monomer, whereas pp60^src molecules that are complexed with pp90 and pp50 contain phosphoserine and greatly reduced amounts of phosphotyrosine. Only the monomer form of pp60^src is capable of phosphorylating IgG in the immune complex phosphotransferase reaction. Normal uninfected chicken cells contain a protein that shares identical partial proteolytic cleavage peptides with the pp90 protein immunoprecipitated from RSV-transformed cells. This pp90 protein is one of the major cytoplasmic proteins in uninfected cells. Antibody directed against pp90 also immunoprecipitates pp60^src and pp50 from lysates of RSV-transformed chicken cells.     

N-fatty acyl compounds inhibit myristoyl acylation of pp60v-src and reduce tumorigenicity of Rous sarcoma virus-infected cells.

Prevention of NH2-terminal myristoylation of pp60v-src was determined with N-fatty acyl glycinal derivatives. Of all the compounds tested, N-myristoyl and N-lauroyl glycinal diethylacetal, N-myristoyl glycyl glycinal diethylacetal, and N-myristoyl-4-aminobutyl aldehyde diethylacetal strongly inhibited myristoylation of pp60v-src; but N-myristoyl diglycyl, N-myristoyl triglycyl, N-decanoyl glycinal diethylacetal, and N-palmitoyl glycinal diethylacetal did not. N-Myristoyl glycinal diethylacetal (25 or 50 microM) suppressed both morphological transformation and colony formation of Rous sarcoma virus-infected chick embryo fibroblasts.     

Truncated gag-related proteins are produced by large deletion mutants of Rous sarcoma virus and form virus particles.

Large deletion (LD) mutants of Prague strain Rous sarcoma virus subgroup B (PrB), derived by serial undiluted passage through chicken (C/E) cells, contain two deletions relative to wild-type virus. One of these joins gag sequences in the p12 coding region to env sequences in region encoding gp37; the other deletion spans the src region. Analysis of the viral proteins of QT6 cell clones containing only LD proviruses by sodium dodecyl sulfate-polyacrylamide gel electrophoresis revealed a major truncated gag-related phosphoprotein of 60,000 to 66,000 daltons (P63LD). P63LD was stable, but could be cleaved in vitro to the predicted products by p15gag. A second gag-related LD protein of about 68,000 to 74,000 molecular weight (P70LD) was also found which often reacted with an anti-gp37 serum. P70LD was unstable and may represent a short-lived gag-gp37 fusion protein. Finally, immunoprecipitation indicated that particles containing P63LD were shed from QT6-LD clones. Thin section preparations of these clones viewed in an electron microscope showed enveloped budding particles of "immature" morphology. Thus, the synthesis and release of particles from infected cells does not require cleavage of the gag precursor, nor does it require the presence of p15 or (most of) p12.     

A 90,000-dalton binding protein common to both steroid receptors and the Rous sarcoma virus transforming protein, pp60v-src

Previous studies have shown that the avian progesterone receptor, when in the nontransformed 8 S state, is complexed to another cellular protein having a molecular weight of 90,000. In this report, we show that this receptor-binding protein is indistinguishable from the 90,000-dalton protein which associates in a complex with the Rous sarcoma virus transforming protein, pp60v-src. This identity was established by the following criteria. 1) Monoclonal antibodies directed against the pp60v-src-associated 90-kDa protein recognized the 90-kDa progesterone receptor binding protein in an immunoblot assay. Conversely, monoclonal antibodies that recognize the progesterone receptor binding protein bind to the 90-kDa protein which complexes with pp60v-src. 2) Peptide maps prepared from the 90-kDa proteins immunoprecipitated from chicken cells with monoclonal antibodies directed against either the 90-kDa receptor binding protein or the 90-kDa pp60v-src-associated protein were indistinguishable. 3) Preincubation of the progesterone receptor complex with monoclonal antibodies prepared against the pp60v-src-associated protein caused a shift in the sedimentation of the progesterone receptor. Previous studies have established that the pp60v-src-associated protein is indistinguishable from one of the major heat shock proteins which are induced under a variety of stress conditions in eukaryotic cells. These present studies implicate a new role for this 90-kDa protein in the action of steroid hormones.     

Temperature-sensitive transformation by Rous sarcoma virus and temperature-sensitive protein kinase activity

The transforming protein of Rous sarcoma virus, p60src, has associated with it a protein kinase activity. We examined whether a correlation exists between the cellular concentration of enzymatically active p60src and the degree to which chick cells are transformed by mutants of Rous sarcoma virus which are temperature-sensitive for transformation. Such a correlation does exist, but cells infected with some mutants could be shown to contain, at the nonpermissive temperature, an amount of protein kinase activity equal to 30 to 40% of that in a wild-type transformed cell. We quantified the amount of virus-induced protein kinase activity by precipitation of p60src with an excess of antitumor antiserum. Our initial measurements of activity were serious underestimates, due to the lability of the protein kinase activity associated with p60src of at least four temperature-sensitive mutants. In fact, no activity at all was associated with p60src of tsLA90 when immunoprecipitation was performed by standard means. However, when immunoprecipitation was performed with procedures which minimize inactivation, it became apparent both that cells transformed by tsLA90 contained protein kinase activity and that cells infected with either NY68 or BK5 contained at the nonpermissive temperature, one-third to one-half as much activity as wild-type transformed cells. This level of activity was much more than that arising from p60sarc in uninfected cells. In uninfected cells we found an amount of protein kinase activity which varied from 3 to 5% as much as that in a virally transformed cell. The lability of the protein kinase activity of each of these mutants is a further demonstration that this activity is essential for the transformation of cells by Rous sarcoma virus. So as to explain the high protein kinase levels in cells infected with NY68 and BK5 at the nonpermissive temperature, the idea that transformation may be a response to a small quantitative change in the total activity of p60src and the possibility that there may be more than one viral function which is essential for transformation are discussed.     

The use of Rous sarcoma virus transformation mutants with differing tyrosine kinase activities to study the relationships between vinculin phosphorylation, pp60v-src location and adhesion plaque integrity

Tyrosine-specific phosphorylation of cellular proteins has been implicated in the neoplastic transformation of cells by Rous sarcoma virus (RSV). One of the putative substrates for the src gene product (pp60v-src) of RSV is the cytoskeletal protein vinculin, giving rise to the hypothesis that tyrosine-specific phosphorylation of vinculin disrupts adhesion plaque integrity, leading to the characteristic rounded morphology of RSV-transformed cells. We have investigated this hypothesis by analysing the properties of fibroblasts transformed by conditional and non-conditional mutants of RSV which confer different morphologies on infected cells, with respect to formation of microfilament bundles, formation of vinculin-containing adhesion plaques, the deposition of a fibronectin-containing extracellular matrix, the localization of pp60v-src and the tyrosine-specific phosphorylation of vinculin. Cells transformed by the temperature-sensitive (ts) RSV mutant LA32 cultured at 41 degrees C were morphologically normal, and contained prominent microfilament bundles and well-developed adhesion plaques. However, these cells had a fully active pp60v-src kinase, had pp60v-src concentrated in their adhesion plaques and contained vinculin which was heavily phosphorylated on tyrosine residues. Cells transformed by a recovered avian sarcoma virus, rASV 2234.3 exhibited a markedly fusiform morphology with pp60v-src concentrated in well-developed adhesion plaques and an elevation of the phosphotyrosine content of vinculin. Cells transformed by LA32 at restrictive temperature comprise morphologically normal cells, indistinguishable from untransformed CEF, yet which contain tyrosine-phosphorylated vinculin and suggest that neither tyrosine-specific phosphorylation of vinculin nor pp60v-src concentration in adhesion plaques is sufficient for the rounded morphology of RSV-transformed cells.     

Cell fusion for genetic analysis of two nonconditional Rous sarcoma virus replication mutants.

Procedures for characterizing replication-defective viruses in nonpermissive mammalian cells were developed and applied to three nonvirogenic Rous sarcoma virus (RSV)-transformed mammalian cell lines--B4, a line of Bryan virus-transformed hamster cells, and two SRD-RSV transformed rat cell lines, LR3/1 and LR3/2. Cell fusion was used to study virus complementation. The three cell lines (i) fused with helper virus-infected chicken cells and the host range of the rescued virus examined, (ii) tested for complementation by fusion with chicken cells exhibiting various patterns of endogenous virus expression, (iii) fused with chicken cells infected with the temperature-sensitive replication mutant LA334 and assayed for complementation at permissive and nonpermissive temperatures, and (iv) tested for complementation of defective viruses in other RSV-transformed mammalian cell lines by fusing pairs of nonvirogenic cell lines and permissive chicken cells. Based upon these complementation studies, we concluded that B4 virus is defective only in the env gene, LR3/) virus is an absolute mutant in the gag and/or pol genes, and LR3/2 virus is a leaky env mutant. Clones of LR3/1 and LR3/2 virus-infected chicken cells were established, and the results obtained from the characterization of these viruses in permissive avian cells substantiates the conclusions reached in the fusion-rescue studies.