VIROLOGY
82, 221-231 (1977)
Pseudotypes of Vesicular Stomatitis Virus with Envelope Antigens Provided by Murine Mammary Tumor Virus J. ZAVADA,’ Imperial
C. DICKSON,2
Cancer Research Fund Laboratories,
AND
R. WEISS
P.O. Box 123, Lincoln’s England
Inn Fields,
London,
WC2.4 3PX.
Accepted May 26,1977 Infection of two mouse mammary carcinoma cell lines with vesicular stomatitis virus (VSV) resulted in the formation of at least two types of particles containing the VSV genome but expressing different envelope characteristics (VSV pseudotypes). One of these VSV pseudotypes was infectious for a cell line derived from normal mouse mammary epithelial cells and mouse embryo cells but noninfectious for 3T3 cells, mink lung cells, and Vero cells. If mouse mammary tumor cells were treated with dexamethasone some days prior to infection with VSV, the titer of this pseudotype was significantly increased. In contrast, the second pseudotype was infectious for mink cells, but not for the other cell lines tested, and the titer of this second pseudotype was unaffected by the presence of dexamethasone. The first pseudotype was found to be almost completely neutralized by anti-murine mammary tumor virus (MuMTV) serum whereas the second pseudotype was only partially neutralized at a higher antiserum concentration. Neither pseudotype showed the neutralization, host range, or interference properties of either ecotropic or xenotropic murine C-type viruses. These results suggest that the first pseudotype is VSV(MuMTV). The other pseudotype is less well defined but conceivably may represent a xenotropic MuMTV. In the course of these studies, a filterable agent was observed in GR mammary carcinoma cultures that reactivated the infectivity of VSV neutralized by antiserum. This agent was transmissible to mink cells. INTRODUCTION
a low efficiency (Vaidya et al., 1976; Lasfargues et al ., 1976). When cells producing oncornaviruses are infected with vesicular stomatitis virus (VSV), the resulting mixed infection yields a proportion of particles which contain the VSV genome and the envelope glycoproteins of the oncornaviruses (VSV pseudotypes). Consequently, the VSV pseudotype possesses several properties specified by the oncornavirus glycoprotein, such as neutralization, host range, and characteristics (Zgvada, interference 1972a, b; Huang et al., 1973; Krontiris et al., 1973; Love and Weiss, 1974; Boettiger et al., 1975, Ztivada, 1976). In this paper we describe the infection of two cell lines derived from mouse mammary carcinomas with VSV which result in the formation of at least two different pseudotypes. One pseudotype appears to be VSV(MuMTV), while the other may be
Carcinogenesis induced by mouse mammary tumor viruses (MuMTV), provides a model system for studying the interactions of genetic, hormonal, and viral functions in the progress of malignant disease (Bentvelzen, 1972; Nandi and McGrath, 1973). In regard to the viral function examinations of the species and tissue tropism and of neutralization by homologous or heterologous antisera have been retarded by the lack of an in vitro assay for MuMTV. Only recently has in vitro infection been successfully performed, albeit at ’ Permanent address: Institute of Virology, Slovak Academy of Sciences, Bratislava, Czechoslovakia. Recipient of a Fellowship under the British Royal Society-Czechoslovak Academy of Sciences Exchange Agreement. 2 Author to whom reprint requests should be addressed. 221 Copyright 0 1977 by Academic Press, Inc. All rights of reproduction in any form reserved.
ISSN 0042-6822
222
ZAVADA,
DICKSON,
related to MuMTV, on the basis of neutralization studies, but demonstrates a different host range. MATERIALS
AND
METHODS
Viruses and Cells VSV. A mutant tlB17 of VSV (Indiana
strain) which has a thermolabile glycoprotein was used for all experiments (Zavada, 197213). Stock virus was recloned and grown up in Vero cells. MuMTV. Two cell lines, GR and MmEiMT/cl (Mm5) derived from spontaneous mouse mammary tumors of the GR and C3H mouse strains, respectively, were used as a source of MuMTV. The GR cell line was described by Ringold et al., (1975) and was kindly supplied by Dr. K. Yamamoto, San Francisco Medical Center, and the MmSMT/cl is a clone of the line derived by Owens and Hackett (1972) and was kindly supplied by Dr. L. 0. Arthur, Frederick Cancer Research Center, Maryland. Production of MuMTV by both cell lines was monitored by reverse transcriptase assays using the exogenous template/primer poly(rC:dG),,-,s (Dickson et al., 1974). The enzyme activity showed a preference for magnesium as the divalent cation characteristic of MuMTV and other B-type viruses (Dion et al., 1974). Neither cell line produced detectable amounts of Ctype viruses as assayed by Mn2+-dependent reverse transcriptase, nor were C-type viral proteins detected in highly concentrated virus preparations by radiolabeling and concomitant analysis on polyacrylamide gels. Murine (ecotropic) leukemia virus. The Moloney strain of murine leukemia virus (MuLV-M) was grown as a chronic infection of NIH/3T3 cells. Murine xenotropic C-type virus (MuLVX) . Two different isolates of MuLV-X were
used: One was isolated from BALB/c mice (N. Teich, unpublished) and grown in rabbit SIRC cells (Leerhoy, 1965); the other was obtained by cocultivation of bone marrow cells from NZB mice with a line of mink cells (CCL 64) (Weiss and Wong, 1977). Normal mammary gland cell line (NMG). This cell line was isolated fi-om
AND
WEISS
the NAMRU mouse strain by Owens et al. (1974) and is maintained in culture as a continuous cell line. NIHi3T3 cells. These are a fibroblastoid cell line isolated from the NIH/Swiss mouse strain (Jainchill et al., 1969). BALBi3T3 cells. These are a similar line derived from BALB/c mouse embryos (Aaronson and Todaro, 1968). Mink lung cells (CCL 64). These were derived from mink embryo lung cell cultures (Henderson et al., 1974). The Vero cell line. This was established from an African green monkey kidney culture (Earley et al., 1967). NRK cells. These are a line derived from normal rat kidney (Due-Nguyen et al., 1966). Bat Tb-1 -Lu cells. These are a line (CCL 88) derived from an adult bat lung. PYS cells. These are a cell line of endodermal character derived from a teratocarcinema (Lehman et al., 1974). Primary cultures of mouse embryos. The BALB/c, C57BL, CBA, and TO strains were supplied by the cell production laboratory, Imperial Cancer Research Fund. Media
The normal and transformed mammary gland cell lines were maintained in Dulbecco’s modified Eagle’s medium containing antibiotics and supplemented with heat-inactivated (56” for 30 min) 10% fetal calf serum. All other cell lines were maintained in the same medium supplemented with heat-inactivated 10% calf serum. Antiserum
and y-Globulin
Preparations
Antiserum against VSV was prepared in sheep as previously described (Zavada, 1972b). Antisera against MuMTV and MuLV-M were prepared in rabbits using gradientpurified virus, which was disrupted with 0.5% Triton X-100 and 0.5% sodium deoxycholate. The MuMTV was prepared from the tissue culture fluids of primary spontaneous mammary tumors of the BR6 mouse strain grown in the presence of dexamethasone to stimulate virus production (Dickson et al., 1974). MuLV-M was grown in NIHl3T3 cells as described above. The rab- _
VSV(MuMTV)
bits were inoculated intramuscularly at four sites, initially with 0.5 mg of viral protein in Freund’s complete adjuvant and subsequently at intervals of 2 weeks with the same disrupted virus in Freund’s incomplete adjuvant. Rabbits were bled 1 week following each booster injection and the serum was stored at -70”. The serum obtained before immunization with MuMTV was used as a supply of normal rabbit serum. y-Globulin (IgG) fractions from the above normal and immune sera were prepared by ammonium sulfate precipitation, once with 50% saturated and twice with 40% saturated solutions. The precipitated globulins were dissolved in phosphatebuffered saline, pH 7.4 (PBS), to a volume equivalent to the original serum, dialyzed against the same buffered saline, and filtered through 0.22-pm Millipore filters. Preparation
of VSV Pseudotypes
Initially, VSV pseudotype stocks were prepared by growing VSV tlB17 in the oncornavirus-producing cell lines at 32” and neutralizing the parent-like VSV with antiserum (Zavada, 1972a) or by heating the mutant at 45” for 1 hr as previously described (Zavada, 1972b). These preliminary experiments revealed that the VSV pseudotypes obtained in the mammary tumor cell lines were easily inactivated nonspecifically. Consequently several precautions were taken to protect the integrity of the VSV pseudotypes formed. Unidentified entities in serum were found to inhibit pseudotype stability, and therefore the medium was always supplemented with heat-inactivated fetal calf serum. Maximum titers of the pseudotypes were obtained when cultures were infected at an m.o.i. of 1-5 PFU per cell and culture media were harvested after 12-16 hr incubation at 32”. The harvested fluids were carefully adjusted with 1% acetic acid to pH 7.2-7.4. Following pH adjustment the infectious culture fluids were clarified by centrifugation at 4000 r-pm for 10 min and stored at -70”. Infectivity of the VSV pseudotypes was lost in a few days at 4”, but stored well at -70”. Neutralization of the parent VSV was car-
223
PSEUDOTYPES
ried out overnight at 4”. In experiments where specific neutralizations were performed, normal and various immune IgGs (anti-MuLV or anti-MuMTV) at several dilutions were also incorporated with the anti-VSV IgG. Dilution of the pseudotype stocks was performed in complete PBS containing heat-inactivated 1% fetal calf serum. The titer of VSV pseudotypes was determined in a plaque assay using several cell lines grown to confluence in 30mm-diameter petri dishes (Nunc). The virus was allowed to adsorb at room temperature for 2 hr in the presence of Polybrene (20 pglml). After adsorption the residual inoculum was removed and the cells were rinsed with PBS and overlaid with agar medium. In some experiments the harvest fluids at a 1:lO dilution in complete PBS were heated at 45” for 1 hr by immersion of a vial in a water bath. This treatment completely inactivates the thermolabile envelope antigen of VSV tlB17 so that only pseudotype particles survive as plaqueforming units (Zavada, 197213).The presence of pseudotypes was then assayed as described above, except that the cells were not rinsed with PBS following incubation at room temperature for 2 hr. RESULTS
Host Range of VSV Pseudotypes Obtained from Mouse Mammary Tumor Cells and Effect of Dexamethasone
When VSV tlB17 was grown through control or oncornavirus-producing cells, a titer of about log PFU/ml of virus was obtained, except for the Mm5 cell line which yielded about 10’ PFU/ml (Fig. 1). The plaque titers were similar when assayed on Vero, mink, or NMG cell lines but showed a lo-fold lower efficiency on NIH/3T3 cells. The VSV stocks were also treated with neutralizing antiserum or IgG specific to VSV and assayed for their ability to form plaques on the test cell lines. After antiserum treatment, the virus stocks derived from uninfected cell lines showed no residual infectivity, whereas those grown through oncornavirus-producing cells all showed a residual
224
ZAVADA, GR
DICKSON, GR+DEX
AND
Mm5
WEISS Mm5 + DEX
SIRC MuLV-X
d
NIH/3T3 MuLV-M
1 :d
FIG. 1. Host range specificities of VSV pseudotypes obtained from cells producing MuMTV, MuLV-M, and MuLV-X. VSV was propagated in control and oncornavirus-infected cells. The mammary tumor lines were maintained in media with or without dexamethasone(DEX1 as indicated. Infectivity of the harvested VSV was determined directly (open columns) and after treatment with anti-V?%’ IgG (shaded columns) at a 150 dilution overnight at 4”. The plaque assays were performed on (a) mink cells, (b) Vero cells, (cl NMG cells, and (d) NIH/3T3 cells.
as would be expected if pseudotypes were present. Figure 1 also shows that the pseudotypes obtained from cells producing ecotropic and xenotropic MuLV possessed the expected host range. The ecotropic VSV(MuLV-M) formed plaques on both mouse cell types used in the plaque assay, namely, NIH/3T3 cells and NMG cells, but did not form plaques on mink or Vero cells; the xenotropic VSV(muLV-X) formed plaques on the mink and Vero cells but not on mouse cells. The pseudotypes obtained from both of the mouse mammary tumor cell lines showed a different host range from either of the VSV(MuLV) pseudotype stocks, in that they plated on NMG cells and mink cells but not on NIH/3T3 cells or Vero cells. Previous studies have shown that, following dexamethasone stimulation, these mammary tumor cell lines produced greatly enhanced amounts of MuMTV (McGrath, 1971; Parks et al., 1974; Dickson et al., 1974; Fine et al., 1974; Ringold et al., 1975). Consistent with these reports, cultures grown in the presence of dexa-
infectivity,
methasone yielded a lo- to 20-fold higher titer of VSV pseudotypes when assayed on NMG cells (Fig. 1). However, dexamethasone had no effect on the pseudotype titer when the same stock was assayed on mink cells. This finding indicates that the VSV pseudotype stock obtained from either GR or Mm5 cells may be comprised of two different pseudotypes; one plates on NMG cells and is stimulated by the presence of dexamethasone in the cultures during VSV infection Yecotropic”), and the other plates on mink cells and is unaffected by dexamethasone (“xenotropic”). The use of these terms “ecotropic” and “xenotropic” refers to the ability of the derived envelopes of the VSV pseudotypes to adsorb to and penetrate the cells mentioned and does not imply that the virus providing the envelope can itself infect and replicate in these same cells. Further evidence to support the idea of two different pseudotypes is that the xenotropic component appears to be more labile than the ecotropic component when stored at 4” and that the former is more sensitive to inhibitors in nonimmune rabbit serum (Table 1).
VSV(MuMTV)
225
PSEUDOTYPES
Several other cell lines and secondary ble to infection. The pseudotypes showed a mouse embryo cell cultures were tested for lower plating efficiency on embryo cell cultheir susceptibility to infection by the VSV tures of the C57BL and TO strains than pseudotypes obtained from the mammary on cells of the BALB/c and CBA strains. tumor cell lines (Table 2). Apart from the The titers of the pseudotype were enmink cell line, all the nonmurine cell hanced by dexamethasone in a similar lines, bat, rat, monkey, and human, and manner to that found for pseudotypes plating on NMG cells, thus indicating that the mouse fibroblastic cell lines, BALB/ 3T3 and NIH/3T3, were resistant to infec- only the ecotropic MuMTV pseudotype tion. In contrast a cell line derived from a was infectious in all these murine culmouse teratocarcinoma of endodermal tures . character (PYS cells) and secondary Specificity of VSV Pseudomouse embryo cell cultures were suscepti- Neutralization twes TABLE 1 EFFECT OF UNHEATED NONIMMUNE RABBIT SERUM AND STORAGE AT 4” ON THE INFECTIVITY OF VSV PSEUD~TYPES OBTAINED FROM MOUSE MAMMARY TUMOR CELLS (GR) AND PLAQUE ASSAYED IN MOUSE (NMG) AND MINK CELLS
Infectivity (log PFU/ml) of pseudotypes= assayed on
Treatment
Untreated control Normal rabbit serum (1:16) Harvest stored at 4” for 0 Days 3 Days
NMG
Mink
3.82 3.28
3.72 2.18
3.36 3.09
3.88 2.37
(2Pseudotypes determined as infectivity to anti-VSV IgG 1:30.
resistant
To characterize the VSV pseudotypes derived from cultures infected with MuLV-M and MuLV-X further, and to compare them with the VSV pseudotypes derived from the mammary tumor cell lines, the pseudotype stocks were treated with control or immune IgG prepared against MuMTV or MuLV-M as described in Materials and Methods (Table 3). Antibodies to MuLV-M are very effective at neutralizing the VSV(MuLV-Ml, but do not affect either of the pseudotypes derived from the mammary tumor cell lines. By contrast, the antibodies to MuMTV effectively neutralize the ecotropic component of the mammary tumor cell line and, to a lesser extent, the xenotropic component.
TABLE 2 Host RANGE SPECIFICITY OF THE PUTATIVE VSV(MuMTV) PSEUD~TYPES OBTAINED FROM MAMMARY TUMOR CELL LINES
Assay cells (Species: Line or Strain) Monkey:Vero Mink:CCL 64 Rat:NRK Bat:Tb .l. lu Human:HeLa Mouse:NMG PYS NIH/3T3 BALB/3T3 C57BLb CBAb TO* BALB/c*
Host cell for VSV propagation: Progeny assayed with (+) and without (-1 anti-VSV IgG GR plus Dex
GR
Mm5 plus Dex
Mm5
-
+
-
+
-
+
-
+
8.25” 8.69 7.15 7.07 7.00 8.38 8.42 6.85 7.12 6.30 6.80 7.03 7.25
Cl.00 2.95 Cl.00 Cl.00 Cl.00 2.76 2.54 Cl.00
82, 221-231 (1977)
Pseudotypes of Vesicular Stomatitis Virus with Envelope Antigens Provided by Murine Mammary Tumor Virus J. ZAVADA,’ Imperial
C. DICKSON,2
Cancer Research Fund Laboratories,
AND
R. WEISS
P.O. Box 123, Lincoln’s England
Inn Fields,
London,
WC2.4 3PX.
Accepted May 26,1977 Infection of two mouse mammary carcinoma cell lines with vesicular stomatitis virus (VSV) resulted in the formation of at least two types of particles containing the VSV genome but expressing different envelope characteristics (VSV pseudotypes). One of these VSV pseudotypes was infectious for a cell line derived from normal mouse mammary epithelial cells and mouse embryo cells but noninfectious for 3T3 cells, mink lung cells, and Vero cells. If mouse mammary tumor cells were treated with dexamethasone some days prior to infection with VSV, the titer of this pseudotype was significantly increased. In contrast, the second pseudotype was infectious for mink cells, but not for the other cell lines tested, and the titer of this second pseudotype was unaffected by the presence of dexamethasone. The first pseudotype was found to be almost completely neutralized by anti-murine mammary tumor virus (MuMTV) serum whereas the second pseudotype was only partially neutralized at a higher antiserum concentration. Neither pseudotype showed the neutralization, host range, or interference properties of either ecotropic or xenotropic murine C-type viruses. These results suggest that the first pseudotype is VSV(MuMTV). The other pseudotype is less well defined but conceivably may represent a xenotropic MuMTV. In the course of these studies, a filterable agent was observed in GR mammary carcinoma cultures that reactivated the infectivity of VSV neutralized by antiserum. This agent was transmissible to mink cells. INTRODUCTION
a low efficiency (Vaidya et al., 1976; Lasfargues et al ., 1976). When cells producing oncornaviruses are infected with vesicular stomatitis virus (VSV), the resulting mixed infection yields a proportion of particles which contain the VSV genome and the envelope glycoproteins of the oncornaviruses (VSV pseudotypes). Consequently, the VSV pseudotype possesses several properties specified by the oncornavirus glycoprotein, such as neutralization, host range, and characteristics (Zgvada, interference 1972a, b; Huang et al., 1973; Krontiris et al., 1973; Love and Weiss, 1974; Boettiger et al., 1975, Ztivada, 1976). In this paper we describe the infection of two cell lines derived from mouse mammary carcinomas with VSV which result in the formation of at least two different pseudotypes. One pseudotype appears to be VSV(MuMTV), while the other may be
Carcinogenesis induced by mouse mammary tumor viruses (MuMTV), provides a model system for studying the interactions of genetic, hormonal, and viral functions in the progress of malignant disease (Bentvelzen, 1972; Nandi and McGrath, 1973). In regard to the viral function examinations of the species and tissue tropism and of neutralization by homologous or heterologous antisera have been retarded by the lack of an in vitro assay for MuMTV. Only recently has in vitro infection been successfully performed, albeit at ’ Permanent address: Institute of Virology, Slovak Academy of Sciences, Bratislava, Czechoslovakia. Recipient of a Fellowship under the British Royal Society-Czechoslovak Academy of Sciences Exchange Agreement. 2 Author to whom reprint requests should be addressed. 221 Copyright 0 1977 by Academic Press, Inc. All rights of reproduction in any form reserved.
ISSN 0042-6822
222
ZAVADA,
DICKSON,
related to MuMTV, on the basis of neutralization studies, but demonstrates a different host range. MATERIALS
AND
METHODS
Viruses and Cells VSV. A mutant tlB17 of VSV (Indiana
strain) which has a thermolabile glycoprotein was used for all experiments (Zavada, 197213). Stock virus was recloned and grown up in Vero cells. MuMTV. Two cell lines, GR and MmEiMT/cl (Mm5) derived from spontaneous mouse mammary tumors of the GR and C3H mouse strains, respectively, were used as a source of MuMTV. The GR cell line was described by Ringold et al., (1975) and was kindly supplied by Dr. K. Yamamoto, San Francisco Medical Center, and the MmSMT/cl is a clone of the line derived by Owens and Hackett (1972) and was kindly supplied by Dr. L. 0. Arthur, Frederick Cancer Research Center, Maryland. Production of MuMTV by both cell lines was monitored by reverse transcriptase assays using the exogenous template/primer poly(rC:dG),,-,s (Dickson et al., 1974). The enzyme activity showed a preference for magnesium as the divalent cation characteristic of MuMTV and other B-type viruses (Dion et al., 1974). Neither cell line produced detectable amounts of Ctype viruses as assayed by Mn2+-dependent reverse transcriptase, nor were C-type viral proteins detected in highly concentrated virus preparations by radiolabeling and concomitant analysis on polyacrylamide gels. Murine (ecotropic) leukemia virus. The Moloney strain of murine leukemia virus (MuLV-M) was grown as a chronic infection of NIH/3T3 cells. Murine xenotropic C-type virus (MuLVX) . Two different isolates of MuLV-X were
used: One was isolated from BALB/c mice (N. Teich, unpublished) and grown in rabbit SIRC cells (Leerhoy, 1965); the other was obtained by cocultivation of bone marrow cells from NZB mice with a line of mink cells (CCL 64) (Weiss and Wong, 1977). Normal mammary gland cell line (NMG). This cell line was isolated fi-om
AND
WEISS
the NAMRU mouse strain by Owens et al. (1974) and is maintained in culture as a continuous cell line. NIHi3T3 cells. These are a fibroblastoid cell line isolated from the NIH/Swiss mouse strain (Jainchill et al., 1969). BALBi3T3 cells. These are a similar line derived from BALB/c mouse embryos (Aaronson and Todaro, 1968). Mink lung cells (CCL 64). These were derived from mink embryo lung cell cultures (Henderson et al., 1974). The Vero cell line. This was established from an African green monkey kidney culture (Earley et al., 1967). NRK cells. These are a line derived from normal rat kidney (Due-Nguyen et al., 1966). Bat Tb-1 -Lu cells. These are a line (CCL 88) derived from an adult bat lung. PYS cells. These are a cell line of endodermal character derived from a teratocarcinema (Lehman et al., 1974). Primary cultures of mouse embryos. The BALB/c, C57BL, CBA, and TO strains were supplied by the cell production laboratory, Imperial Cancer Research Fund. Media
The normal and transformed mammary gland cell lines were maintained in Dulbecco’s modified Eagle’s medium containing antibiotics and supplemented with heat-inactivated (56” for 30 min) 10% fetal calf serum. All other cell lines were maintained in the same medium supplemented with heat-inactivated 10% calf serum. Antiserum
and y-Globulin
Preparations
Antiserum against VSV was prepared in sheep as previously described (Zavada, 1972b). Antisera against MuMTV and MuLV-M were prepared in rabbits using gradientpurified virus, which was disrupted with 0.5% Triton X-100 and 0.5% sodium deoxycholate. The MuMTV was prepared from the tissue culture fluids of primary spontaneous mammary tumors of the BR6 mouse strain grown in the presence of dexamethasone to stimulate virus production (Dickson et al., 1974). MuLV-M was grown in NIHl3T3 cells as described above. The rab- _
VSV(MuMTV)
bits were inoculated intramuscularly at four sites, initially with 0.5 mg of viral protein in Freund’s complete adjuvant and subsequently at intervals of 2 weeks with the same disrupted virus in Freund’s incomplete adjuvant. Rabbits were bled 1 week following each booster injection and the serum was stored at -70”. The serum obtained before immunization with MuMTV was used as a supply of normal rabbit serum. y-Globulin (IgG) fractions from the above normal and immune sera were prepared by ammonium sulfate precipitation, once with 50% saturated and twice with 40% saturated solutions. The precipitated globulins were dissolved in phosphatebuffered saline, pH 7.4 (PBS), to a volume equivalent to the original serum, dialyzed against the same buffered saline, and filtered through 0.22-pm Millipore filters. Preparation
of VSV Pseudotypes
Initially, VSV pseudotype stocks were prepared by growing VSV tlB17 in the oncornavirus-producing cell lines at 32” and neutralizing the parent-like VSV with antiserum (Zavada, 1972a) or by heating the mutant at 45” for 1 hr as previously described (Zavada, 1972b). These preliminary experiments revealed that the VSV pseudotypes obtained in the mammary tumor cell lines were easily inactivated nonspecifically. Consequently several precautions were taken to protect the integrity of the VSV pseudotypes formed. Unidentified entities in serum were found to inhibit pseudotype stability, and therefore the medium was always supplemented with heat-inactivated fetal calf serum. Maximum titers of the pseudotypes were obtained when cultures were infected at an m.o.i. of 1-5 PFU per cell and culture media were harvested after 12-16 hr incubation at 32”. The harvested fluids were carefully adjusted with 1% acetic acid to pH 7.2-7.4. Following pH adjustment the infectious culture fluids were clarified by centrifugation at 4000 r-pm for 10 min and stored at -70”. Infectivity of the VSV pseudotypes was lost in a few days at 4”, but stored well at -70”. Neutralization of the parent VSV was car-
223
PSEUDOTYPES
ried out overnight at 4”. In experiments where specific neutralizations were performed, normal and various immune IgGs (anti-MuLV or anti-MuMTV) at several dilutions were also incorporated with the anti-VSV IgG. Dilution of the pseudotype stocks was performed in complete PBS containing heat-inactivated 1% fetal calf serum. The titer of VSV pseudotypes was determined in a plaque assay using several cell lines grown to confluence in 30mm-diameter petri dishes (Nunc). The virus was allowed to adsorb at room temperature for 2 hr in the presence of Polybrene (20 pglml). After adsorption the residual inoculum was removed and the cells were rinsed with PBS and overlaid with agar medium. In some experiments the harvest fluids at a 1:lO dilution in complete PBS were heated at 45” for 1 hr by immersion of a vial in a water bath. This treatment completely inactivates the thermolabile envelope antigen of VSV tlB17 so that only pseudotype particles survive as plaqueforming units (Zavada, 197213).The presence of pseudotypes was then assayed as described above, except that the cells were not rinsed with PBS following incubation at room temperature for 2 hr. RESULTS
Host Range of VSV Pseudotypes Obtained from Mouse Mammary Tumor Cells and Effect of Dexamethasone
When VSV tlB17 was grown through control or oncornavirus-producing cells, a titer of about log PFU/ml of virus was obtained, except for the Mm5 cell line which yielded about 10’ PFU/ml (Fig. 1). The plaque titers were similar when assayed on Vero, mink, or NMG cell lines but showed a lo-fold lower efficiency on NIH/3T3 cells. The VSV stocks were also treated with neutralizing antiserum or IgG specific to VSV and assayed for their ability to form plaques on the test cell lines. After antiserum treatment, the virus stocks derived from uninfected cell lines showed no residual infectivity, whereas those grown through oncornavirus-producing cells all showed a residual
224
ZAVADA, GR
DICKSON, GR+DEX
AND
Mm5
WEISS Mm5 + DEX
SIRC MuLV-X
d
NIH/3T3 MuLV-M
1 :d
FIG. 1. Host range specificities of VSV pseudotypes obtained from cells producing MuMTV, MuLV-M, and MuLV-X. VSV was propagated in control and oncornavirus-infected cells. The mammary tumor lines were maintained in media with or without dexamethasone(DEX1 as indicated. Infectivity of the harvested VSV was determined directly (open columns) and after treatment with anti-V?%’ IgG (shaded columns) at a 150 dilution overnight at 4”. The plaque assays were performed on (a) mink cells, (b) Vero cells, (cl NMG cells, and (d) NIH/3T3 cells.
as would be expected if pseudotypes were present. Figure 1 also shows that the pseudotypes obtained from cells producing ecotropic and xenotropic MuLV possessed the expected host range. The ecotropic VSV(MuLV-M) formed plaques on both mouse cell types used in the plaque assay, namely, NIH/3T3 cells and NMG cells, but did not form plaques on mink or Vero cells; the xenotropic VSV(muLV-X) formed plaques on the mink and Vero cells but not on mouse cells. The pseudotypes obtained from both of the mouse mammary tumor cell lines showed a different host range from either of the VSV(MuLV) pseudotype stocks, in that they plated on NMG cells and mink cells but not on NIH/3T3 cells or Vero cells. Previous studies have shown that, following dexamethasone stimulation, these mammary tumor cell lines produced greatly enhanced amounts of MuMTV (McGrath, 1971; Parks et al., 1974; Dickson et al., 1974; Fine et al., 1974; Ringold et al., 1975). Consistent with these reports, cultures grown in the presence of dexa-
infectivity,
methasone yielded a lo- to 20-fold higher titer of VSV pseudotypes when assayed on NMG cells (Fig. 1). However, dexamethasone had no effect on the pseudotype titer when the same stock was assayed on mink cells. This finding indicates that the VSV pseudotype stock obtained from either GR or Mm5 cells may be comprised of two different pseudotypes; one plates on NMG cells and is stimulated by the presence of dexamethasone in the cultures during VSV infection Yecotropic”), and the other plates on mink cells and is unaffected by dexamethasone (“xenotropic”). The use of these terms “ecotropic” and “xenotropic” refers to the ability of the derived envelopes of the VSV pseudotypes to adsorb to and penetrate the cells mentioned and does not imply that the virus providing the envelope can itself infect and replicate in these same cells. Further evidence to support the idea of two different pseudotypes is that the xenotropic component appears to be more labile than the ecotropic component when stored at 4” and that the former is more sensitive to inhibitors in nonimmune rabbit serum (Table 1).
VSV(MuMTV)
225
PSEUDOTYPES
Several other cell lines and secondary ble to infection. The pseudotypes showed a mouse embryo cell cultures were tested for lower plating efficiency on embryo cell cultheir susceptibility to infection by the VSV tures of the C57BL and TO strains than pseudotypes obtained from the mammary on cells of the BALB/c and CBA strains. tumor cell lines (Table 2). Apart from the The titers of the pseudotype were enmink cell line, all the nonmurine cell hanced by dexamethasone in a similar lines, bat, rat, monkey, and human, and manner to that found for pseudotypes plating on NMG cells, thus indicating that the mouse fibroblastic cell lines, BALB/ 3T3 and NIH/3T3, were resistant to infec- only the ecotropic MuMTV pseudotype tion. In contrast a cell line derived from a was infectious in all these murine culmouse teratocarcinoma of endodermal tures . character (PYS cells) and secondary Specificity of VSV Pseudomouse embryo cell cultures were suscepti- Neutralization twes TABLE 1 EFFECT OF UNHEATED NONIMMUNE RABBIT SERUM AND STORAGE AT 4” ON THE INFECTIVITY OF VSV PSEUD~TYPES OBTAINED FROM MOUSE MAMMARY TUMOR CELLS (GR) AND PLAQUE ASSAYED IN MOUSE (NMG) AND MINK CELLS
Infectivity (log PFU/ml) of pseudotypes= assayed on
Treatment
Untreated control Normal rabbit serum (1:16) Harvest stored at 4” for 0 Days 3 Days
NMG
Mink
3.82 3.28
3.72 2.18
3.36 3.09
3.88 2.37
(2Pseudotypes determined as infectivity to anti-VSV IgG 1:30.
resistant
To characterize the VSV pseudotypes derived from cultures infected with MuLV-M and MuLV-X further, and to compare them with the VSV pseudotypes derived from the mammary tumor cell lines, the pseudotype stocks were treated with control or immune IgG prepared against MuMTV or MuLV-M as described in Materials and Methods (Table 3). Antibodies to MuLV-M are very effective at neutralizing the VSV(MuLV-Ml, but do not affect either of the pseudotypes derived from the mammary tumor cell lines. By contrast, the antibodies to MuMTV effectively neutralize the ecotropic component of the mammary tumor cell line and, to a lesser extent, the xenotropic component.
TABLE 2 Host RANGE SPECIFICITY OF THE PUTATIVE VSV(MuMTV) PSEUD~TYPES OBTAINED FROM MAMMARY TUMOR CELL LINES
Assay cells (Species: Line or Strain) Monkey:Vero Mink:CCL 64 Rat:NRK Bat:Tb .l. lu Human:HeLa Mouse:NMG PYS NIH/3T3 BALB/3T3 C57BLb CBAb TO* BALB/c*
Host cell for VSV propagation: Progeny assayed with (+) and without (-1 anti-VSV IgG GR plus Dex
GR
Mm5 plus Dex
Mm5
-
+
-
+
-
+
-
+
8.25” 8.69 7.15 7.07 7.00 8.38 8.42 6.85 7.12 6.30 6.80 7.03 7.25
Cl.00 2.95 Cl.00 Cl.00 Cl.00 2.76 2.54 Cl.00
