Vol. 20, No. 3
JOURNAL OF VIROLOGY, Dec. 1976, P. 545-554 Copyright © 1976 American Society for Microbiology
Printed in U.S.A.
Induction of GIX Antigen and Gross Cell Surface Antigen After Infection by Ecotropic and Xenotropic Murine Leukemia Viruses In Vitro P. V. O'DONNELL* AND E. STOCKERT Memorial Sloan-Kettering Cancer Center, New York, New York 10021
Received for publication 12 April 1976
A number of ecotropic and xenotropic murine leukemia viruses were examined for their ability to induce the Glx antigen and Gross cell surface antigen (GCSA) in tissue culture fibroblasts. G1x appears to be a constituent of murine leukemia virus gp7O; a molecular characterization of GCSA has not yet been reported. Antigen induction was measured by the ability of productively infected cells to absorb cytotoxic activity from the standard Gjx- and GCSA-typing antisera. Cells infected by ecotropic viruses displayed four distinct phenotypes: Glx:+/GCSA++, G1x-/GCSA++, G,X++/GCSA+, and G,X-/GCSA+; cells infected by xenotropic viruses were either G,X-/GCSA+ or G1x-/GCSA-. G1x induction appeared to be a type-specific property of some but not all Gross-AKR type ecotropic viruses. Differences in the degree of absorption of the GCSA antiserum by ecotropic virus- and xenotropic virus-infected cells indicated that GCSA may comprise multiple antigenic determinants.
The G1x antigen (42) and Gross cell surface antigen (GCSA) (33) are two serologically wellcharacterized cell surface antigens associated with naturally occurring (Gross-AKR type) murine leukemia virus (MuLV). Glx was first recognized as a differentiation alloantigen found on thymocytes of certain mouse strains (G1x+) often in the absence of complete viral expression, as in strain 129. Glx is absent in other strains (Glx-). Recent data (32, 46) indicate that G1x is a constituent of gp7O, the major MuLV envelope glycoprotein (6). Unlike G1x, the appearance of GCSA on lymphoid cells is always correlated with the production of GrossAKR type MuLV and has been localized in regions of the infected cell surface which are devoid of budding virions by means of immunoelectron microscopy (2). These antigens are defined by the cytotoxic reactivities of specific G1x-typing (rat) and GCSA-typing (mouse) antisera (33, 42). The G1x-typing serum, in particular, is known to be broadly reactive and contains antibodies to GCSA determinants (3, 15, 20), viral envelope antigenic determinants (4, 15), and internal viral structural proteins (14, 15), as well as to G1x (42). Specificity of the Glx-typing system is obtained by use of appropriate target cells in the cytotoxicity assay, strain 129 thymocytes, which restricts the assay to detection of G1x antigen (42). The GCSA-typing serum is a predominantly type-specific reagent when it is used with E &G2 target cells, a passage A Gross
virus-induced leukemia of strain C57BL/6 (3, 33). By absorption of the appropriate typing antiserum it was shown that Glx and GCSA antigens could be detected on tissue culture fibroblasts that were productively infected with Gross virus (12). To investigate further the relationship between G1x and GCSA expression and infection of cells by naturally occurring MULV, we examined a number of ecotropic and xenotropic MuLV isolates for their ability to induce these antigens in vitro. Induction of Glx expression was found to be a type-specific property of some but not all ecotropic MuLVXs which had previously been classified as Gross-AKR type on the basis of virus neutralization tests (17, 18, 22). The measurement of GCSA expression indicated that a system of multiple antigenic determinants was probably involved with characteristics of both type and group specificity.
MATERILS AND METHODS Virus. A summary of the viruses used and their histories is given in Table 1. The host range of all virus isolates was verified prior to use in these experiments. The host tropism of ecotropic viruses was determined by differential growth on NIH/Swiss mouse embryo (ME) cells and BALB/c-ME cells (28) and failure to replicate in heterologous mink lung fibroblast (CCL64) cells (19). Conversely, isolates of xenotropic virus were able to replicate in CCL64 cells but not in ME cells. 545
546
J. VIROL.
O'DONNELL AND STOCKERT
TABLE 1. Viruses studied Mouse origin
M. musculus inbred strains BALB/c
Virus"b
History-'
WN1802N CL B5
Ecotropic (N) virus, formerly referred to as BALB/c-S2N, isolated from normal spleen (18); NIH-ME (®); cloned and propagated in III6A cells (O'Donnell et al., Virology, in press) Ecotropic (B) virus, formerly referred to as BALB/c-S2B, isolated from same normal spleen as WN1802N (18); BALB/c-ME 21; cloned and propagated in II16A cells (O'Donnell et aT, Virology, in press) Ecotropic (B) virus isolated from BALB/c radiation-induced leukemia RL d 1 (36); cloned directly from a cell-free extract of tumor tissue and propagated in II16A cells (unpublished data) Xenotropic virus induced from a BALB/3T3 cell line after 5-bromodeoxyuridine treatment (7); propagated in CCL64 (mink lung fibroblast) cells after passage in CCL60 (rabbit cornea) cells Xenotropic virus isolated after cocultivation of BALB/c and NIH/Swiss splenocytes with CCL60 cells (38); propagated in CCL64 cells Ecotropic (N) virus (pool 1920) isolated from normal spleen (J. W. Hartley, personal communication); NIH-ME cloned and propagated in III6A cells Ecotropic (B) virus (pool 1909) isolated from same normal spleen as B6tN-(J. W. Hartley, personal communication); BALB/c-ME Q5); cloned and propagated in III6A cells. Ecotropic (N) virus isolated from spleen of a mouse bearing the EdG2 leukemia (33); cloned directly from a cell-free extract of tumor tissue and propagated in I116A cells (unpublished data) Ecotrop c (N) virus isolated from AKR leukemia (18); NIHME (m; cloned and propagated in III6A cells Xenotropic virus recovered from RD human tumor cells passaged in immunosuppressed NIH/Swiss mice (44); propagated in CCL64 cells after passage in RD cells Xenotropic virus isolated after spontaneous activation of NZB CL S2 cells (25); propagated in 64J1 cells Ecotropic (N) virus isolated after in vivo passage of C3Hf/Gs thymic lymphoma extract in C3Hf/Gs mice and subsequently in W/Fu rats (24); cloned free from sarcoma virus in NIH/3T3 cells (1); propagated in II16A cells Ecotropic (N) virus originally isolated from the spleen of a Swiss mouse bearing an Ehrlich ascites carcinoma (13); multiple in vivo passage in Ha/ICR, DBA/2 mice; NIH/3T3 propagated in III6A cells Ecotropic (B) virus recovered after Fs plus B/T-L virus (43) mixed infection of a BALB/c mouse in vivo (29); BALB/3T3 propagated in III6A cells Ecotropic (NB) virus originally isolated from a BALB/c mouse bearing an ascites tumor of Swiss mouse strain origin (35); multiple in vivo passage in BALB/c, NIH/Swiss mice; BALB/c-ME (, NIH-ME (©); cloned and propagated in III6A cells Ecotropic (NB) virus originally isolated from a BALB/c mouse bearing sarcoma 37 of A/LN mouse strain origin (30); cloned from Electro-Nucleonics preparation (lot no. 445-40-12A) and propagated in III6A cells Ecotropic (N) virus isolated after cocultivation of pooled M. musculus subsp. molossinus spleen and kidney cells with NIH/3T3 cells (26); propagated in I116A cells
WN1802B CL Dl
RL1 CL 2
S16CL10(I)
MLC 60
C57BL/6
B6N CL A3 B6B CL D3
EdG2 CL A6
AKR
AKR-L1 CL G12
NIH/Swiss
AT124
NZB
NZB
Uncertain
Kirsten
Friend, Fs
Friend, F, Rauscher CL 1
Moloney CL H6
M. musculus subsp. molossinus
Molossinus-N
(D; (D;
VOL. 20, 1976
MuLV-INDUCED G,x AND GCSA CELL SURFACE ANTIGENS
547
TABLE 1-Continued Mouse origin
Virus&b
Historycd
Xenotropic virus isolated after cocultivation of pooled M. musculus subsp. castaneus spleen and kidney cells with FCf2Th (dog thymus) cells (26); propagated in CCL64 cells M. caroli Caroli Xenotropic virus induced from M. caroli cells after 5-bromodedeoxyuridine treatment (27); propagated in CCL64 cells a The sources of viruses used were as follows: WN1802N, WN1802B, B6N, B6B, AKR-L1, and Rauscher MuLV's from J. W. Hartley and W. P. Rowe, National Institute of Allergy and Infectious Diseases; S16CL10(I), MLC 60, and AT124 from C. Sherr, National Cancer Institute; Molossinus-N and Castaneus-X from U. Rapp, National Cancer Institute; NZB from J. Levy, University of California; Kirsten MuLV and Caroli virus from J. Stephenson, National Cancer Institute; Friend Fs and FT from F. Lilly, Albert Einstein College of Medicine. b Ecotropic viruses were cloned prior to use in these experiments as described in the text and are indicated by clone (CL) designations. ' N-, B-, or NB-tropic host range of ecotropic viruses is indicated in parentheses after the ecotropic designation. d Circled numbers indicate the number of cell-free passages in vitro.
M. musculus subsp. castaneus
Castaneus-X
Cells. Clone III6A of the feral ME line WM1511 and XC cells were provided by J. W. Hartley. The III6A cell line, which is also called SC-1, does not exhibit Fv-1 restriction of naturally occurring MuLV and is sensitive to both N- and B-tropic viruses (16). These cells were recloned prior to use in these experiments. ME cells were prepared from embryos in day 14 to 16 of gestation (17) and stored prior to use at - 196°C. Pregnant NIH/Swiss and BALB/c mice were kindly provided by W. P. Rowe. Strain 129, 129-Glx-, and C57BL/6 mice were from our own breeding colony. A W/Fu cell line was constructed from W/Fu rat embryo cells according to the 3T3 protocol (45). Pregnant W/Fu rats were obtained from our own breeding colony. NRK cells were provided by S. Aaronson, National Cancer Institute, CCL64 and 64J1 cells by C. Sherr, and JLSV9 cells by T. Pincus, Memorial Sloan-Kettering Cancer Center. For virus producer cell lines, the virus released is indicated in parentheses after the cell line designation. Cells were grown in Dulbecco modified Eagle minimal essential medium (EMEM) supplemented with 250 U of penicillin per ml, 250 ,ug of streptomycin per ml, and 10% heat-inactivated fetal calf serum (Microbiological Associates, Bethesda, Md.). Cell cultures were assayed for the presence of mycoplasma (5) by D. Armstrong, Memorial Sloan-Kettering Cancer Center. XC plaque and infectious center assays. Ecotropic virus XC plaque titrations and infectious center determinations of the percentage of productively infected cells per culture were performed on II16A indicator cells at 37°C (34). Microtiter XC plaque assay. The wells of Falcon Microtest II plates were planted with 2.5 x 103 I6A cells in 0.1 ml of Dulbecco growth medium containing 10 ug of polybrene per ml (Aldrich Chemical Co., Milwaukee, Wis.). On the next day the microcultures were infected by transferring approximately 0.02 ml of culture fluid from the wells of a master plate by means of a replica plating device
(Cooke Laboratory Products, Alexandria, Va.). After 4 days of incubation at 37°C the culture fluids were removed, and the cells were UV irradiated overlaid with 104 XC cells, and developed as described above in the standard XC plaque assay. Virus cloning. MuLV was cloned at limiting dilution using a modified microtiter method (39; O'Donnell et al., Virology, in press). Approximately 1 x 105 to 2 x 105 III6A cells were planted in 60-mm petri dishes (Falcon), infected with virus 18 to 24 h later at a multiplicity of infection of 2.5 x 10-4 PFU/ cell, dispersed with 0.75% trypsin, and planted at a density of 200 cells/well in Falcon Microtest II plates such that there would be an average of 0.05 infectious centers/well. A number of experiments showed that this resulted in a Poisson distribution of infectious centers in the microtiter wells. After 2 weeks of incubation at 37°C the microtiter fluids were assayed for the presence of virus by replica plating in a microtiter XC plaque assay. After determining the distribution of virus-positive cultures, the master plate that had been stored frozen at -80°C was thawed and the culture fluids from a positive well were used to infect fresh cultures of III6A cells. The resulting cultures of chronically infected III6A cells were then used to absorb the anti-G1x and antiGCSA sera as described. From the distribution of positive cultures it was possible to calculate the probability that such a culture was derived from a single infectious center. For most clonal isolations this statistic was greater than 90%. Phenotypically mixed virus. A xenotropic virus pseudotype of WN1802N virus was constructed by treating JLSV9 cells that were productively infected by WN1802N virus (105 PFU/ml of culture fluid) with 5-iododeoxyuridine (IUdR; Calbiochem, San Diego, Calif.), which is known to induce endogenous xenotropic virus from these cells (9). Cells were treated with 10 ,tg of IUdR per ml for 24 h, washed, and refed. Culture fluids were harvested at 3 days after the addition of IUdR, the time at which maximum yields of endogenous virus were obtained from
548
O'DONNELL AND STOCKERT
uninfected JLSV9 cells. These fluids were used as the stock of ecotropic/xenotropic, phenotypically mixed virus. RDDP assay. Culture fluids were clarified at 1,500 x g for 15 min at 4°C and then concentrated approximately 30-fold by centrifugation at 30,000 x g for 90 min. Virus pellets were resuspended (37) and assayed for RNA-directed DNA polymerase (RDDP) activity as described by Stephenson et al. (39), using poly(rA) oligo(dT),2,, (5:1 complex, PL Biochemicals, Milwaukee, Wis.) as template primer and [3H]TTP (2.4 Ci/mmol, Schwarz/Mann, Orangeburg, N.Y.) as substrate. Assays were performed under conditions in which the incorporation of [3H]TTP was linearly proportional to the concentration of enzyme in the reaction mixture. Activity is expressed as total picomoles of TTP incorporated in culture fluids per 106 cells. Immunofluorescence. The percentage of antigenpositive cells per culture was determined by indirect immunofluorescence (21), using rabbit antiRauscher p30 provided by W. D. Hardy, Jr., Memorial Sloan-Kettering Cancer Center (23). Fluorescein-conjugated goat anti-rabbit immunoglobulin was purchased from Hyland Laboratories, Inc., Los Angeles, Calif., and used at 1:50 dilution in phosphate-buffered saline. Cytotoxicity assay. For the cytotoxicity assay (42), serial dilutions (50 ,l) of antiserum were made in medium 199 and mixed with equal volumes of target cells (5 x 106 cells/ml) and appropriately diluted complement. After incubation for 45 min at 37°C, viability counts were made in the presence of trypan blue. Glx-typing system. For the Glx-typing system (42), the standard G,x antiserum produced in rats was used: (W/Fu x BN)F, anti-W/Fu leukemia (C58NT)D, abbreviated "anti-NTD." The target cells were thymocytes from strain 129 mice and the complement source was absorbed rabbit serum diluted 1:10. GCSA-typing system. For the GCSA-typing system (33), the standard GCSA antiserum was used: C57BL/6 anti-AKR spontaneous leukemia K36, abbreviated "B6 anti-K36." The target cells were E dG2 leukemia cells (transplanted C57BL/6 leukemia induced by passage A Gross virus) and the complement source was pooled guinea pig serum diluted 1:2. Test for absorption of Glx and GCSA antibody by cells. Antisera were appropriately diluted, mixed with washed, packed cells, and incubated for 30 min at 4°C to test for the absorption of Glx and GCSA antibody by cells (42). After centrifugation at 900 x g for 10 min the supernatant was tested in the standard cytotoxicity assay. For G1x tests anti-NTD serum was diluted 1:100 or 1:150 (approximately three twofold dilutions below the end point of 50% cytotoxicity of the serum with 129 thymocytes). The proportion of diluted serum to packed cells was 2:1. For GCSA tests B6 anti-K36 serum was diluted 1:10 or 1:15 (approximately two twofold dilutions below the end point of 50% cytotoxicity of the serum with E d G2 cells). The proportion of diluted serum to packed cells was 1:1.
J. VIROL.
The cells used for absorption were confluent cul-
tures of uninfected control cells or productively infected cells. Cultures were washed once with Ca and
Mg-free phosphate-buffered saline and dispersed by incubation with 0.05% EDTA for 5 min at 37°C. Cells
were harvested by centrifugation at 500 x g and washed two times in EMEM. After a final wash in medium 199, the cells were packed by slowly increasing the speed of centrifugation to 900 x g over a
period of 10 min.
In some experiments the transplanted leukemias EL4 and ERLD, which are carried in C57BL/6 mice (33), were used as additional controls. The cells were washed two times in EMEM and one time in medium 199 as described above. Quantitative absorption test of Glx. For the quantitative absorption test of Glx (42; see Fig. 2), different numbers of cells, washed and packed as described above, were mixed with 60 pAl of anti-NTD serum diluted 1:600 (approximately one twofold dilution below the end point of 50% cytotoxicity of the serum with 129 thymocytes) and incubated for 30 min at 4CC. After centrifugation, the absorbed antibody samples were tested in the cytotoxicity assay with 129 thymocytes as target cells as described above.
RESULTS
Induction of Glx and GCSA antigens by ecotropic MuLV. Induction ofthe cell surface antigens GI, and GCSA after exogenous infection of cultured cells by different MuLV isolates was measured by the ability of such cells to absorb cytotoxic antibodies from the standard typing sera. Figure 1 shows the pattern of antigen expression for N- and B-tropic MuLV of BALB/c origin. The cytotoxic reactivity of the G,,- and GCSA-typing sera was completely absorbed by III6A(WN1802N) cells, indicating the expression of both antigens on these cells. II16A(WN1802B) cells completely absorbed GCSA but not G,x activity and thus appear to express only GCSA on their cell surface. The G1x-negative (G,x-) phenotype of III6A cells infected by the B-tropic isolate was confirmed by quantitative absorption tests shown in Fig. 2 in which anti-NTD serum was absorbed with varying numbers of infected cells and assayed for residual cytotoxicity. The cytotoxic reactivity of the GI, antiserum was completely removed by absorption with 7 x 105 II16A(WN1802N) cells, whereas no absorption was observed with up to 4 x 10"; II16A(WN1802B) cells. Since the producer lines release equivalent amounts of progeny virus, it is unlikely that this result can be attributed to a quantitative difference in the level of viral gene expression, but probably represents a qualitative difference between the gp7O molecules of these two viruses. This same analysis was extended to a number of other ecotropic MuLV isolates. Produc-
t _A 1=nA_ ~
VOL. 20, 1976
MuLV-INDUCED AX AND GCSA CELL SURFACE ANTIGENS
F~~~ m " 00 Qs)
~
v
100
200
d400
15
w
(n
I
800
§I 1.
I
E Q) v
ok
6A(WN1802B)
5En
j _ / I _G .
QU
549
16A(WN1802N)
GC_
50 3
-i
1-
vs
30
0
60
1/Antiserum dilution
FIG. 1. Absorption of G,x and GCSA cytotoxic antibodies by III6A cells infected by WN1802N (0) and WN1802B (S) viruses. For G1x, anti-NTD serum (1:100 dilution) and, for GCSA, B6 anti-K36 serum (1:15 dilution) were absorbed with washed, packed cells and assayed for residual cytotoxicity with the standard target cells as described in the text. Controls included unabsorbed antiserum (-) and antiserum absorbed with uninfected III6A cells (A) and EL4 leukemia cells (x), whose phenotype is Glx-IGCSA-. The levels of virus production in the cultures used for absorption are given in Table 2.
tively infected HII6A cell lines were used to absorb the appropriate typing sera. The results are given in Table 2, which includes measurements of the level of virus production at the time the cells were harvested for absorption. Complete absorption (+ +) represents removal of cytotoxic antibodies as shown in Fig. 1 for WN1802N-infected cells. Failure to absorb cytotoxicity was scored as negative (-) as observed for the GIX antiserum with WN1802B-infected cells (Fig. 1). Partial absorption (+) was observed only in the case of GCSA typing, where the titer of the absorbed antiserum was lowered as compared with the negative absorption of uninfected control cells (Fig. 3). Of the GrossAKR type viruses only the B-tropic virus of C57BL/6 mice (B6B) showed partial absorption of GCSA antibody, although this was a common characteristic of the Friend-Moloney-Rauscher (FMR) subgroup viruses. However, cells infected by B6B virus completely absorbed Glx activity. To determine whether the partial absorption of anti-GCSA activity was due to limited GCSA cross-reactivity or to quantitative differences in the level of GCSA expression on infected cells, we performed double absorption experiments shown in Fig. 4. B6 anti-K36 serum was absorbed twice with approximately 40 x 106 to 60 x 106 infected cells. Residual cytotoxicity was measured after each absorption. After the second absorption with cells infected by B6B and
JOURNAL OF VIROLOGY, Dec. 1976, P. 545-554 Copyright © 1976 American Society for Microbiology
Printed in U.S.A.
Induction of GIX Antigen and Gross Cell Surface Antigen After Infection by Ecotropic and Xenotropic Murine Leukemia Viruses In Vitro P. V. O'DONNELL* AND E. STOCKERT Memorial Sloan-Kettering Cancer Center, New York, New York 10021
Received for publication 12 April 1976
A number of ecotropic and xenotropic murine leukemia viruses were examined for their ability to induce the Glx antigen and Gross cell surface antigen (GCSA) in tissue culture fibroblasts. G1x appears to be a constituent of murine leukemia virus gp7O; a molecular characterization of GCSA has not yet been reported. Antigen induction was measured by the ability of productively infected cells to absorb cytotoxic activity from the standard Gjx- and GCSA-typing antisera. Cells infected by ecotropic viruses displayed four distinct phenotypes: Glx:+/GCSA++, G1x-/GCSA++, G,X++/GCSA+, and G,X-/GCSA+; cells infected by xenotropic viruses were either G,X-/GCSA+ or G1x-/GCSA-. G1x induction appeared to be a type-specific property of some but not all Gross-AKR type ecotropic viruses. Differences in the degree of absorption of the GCSA antiserum by ecotropic virus- and xenotropic virus-infected cells indicated that GCSA may comprise multiple antigenic determinants.
The G1x antigen (42) and Gross cell surface antigen (GCSA) (33) are two serologically wellcharacterized cell surface antigens associated with naturally occurring (Gross-AKR type) murine leukemia virus (MuLV). Glx was first recognized as a differentiation alloantigen found on thymocytes of certain mouse strains (G1x+) often in the absence of complete viral expression, as in strain 129. Glx is absent in other strains (Glx-). Recent data (32, 46) indicate that G1x is a constituent of gp7O, the major MuLV envelope glycoprotein (6). Unlike G1x, the appearance of GCSA on lymphoid cells is always correlated with the production of GrossAKR type MuLV and has been localized in regions of the infected cell surface which are devoid of budding virions by means of immunoelectron microscopy (2). These antigens are defined by the cytotoxic reactivities of specific G1x-typing (rat) and GCSA-typing (mouse) antisera (33, 42). The G1x-typing serum, in particular, is known to be broadly reactive and contains antibodies to GCSA determinants (3, 15, 20), viral envelope antigenic determinants (4, 15), and internal viral structural proteins (14, 15), as well as to G1x (42). Specificity of the Glx-typing system is obtained by use of appropriate target cells in the cytotoxicity assay, strain 129 thymocytes, which restricts the assay to detection of G1x antigen (42). The GCSA-typing serum is a predominantly type-specific reagent when it is used with E &G2 target cells, a passage A Gross
virus-induced leukemia of strain C57BL/6 (3, 33). By absorption of the appropriate typing antiserum it was shown that Glx and GCSA antigens could be detected on tissue culture fibroblasts that were productively infected with Gross virus (12). To investigate further the relationship between G1x and GCSA expression and infection of cells by naturally occurring MULV, we examined a number of ecotropic and xenotropic MuLV isolates for their ability to induce these antigens in vitro. Induction of Glx expression was found to be a type-specific property of some but not all ecotropic MuLVXs which had previously been classified as Gross-AKR type on the basis of virus neutralization tests (17, 18, 22). The measurement of GCSA expression indicated that a system of multiple antigenic determinants was probably involved with characteristics of both type and group specificity.
MATERILS AND METHODS Virus. A summary of the viruses used and their histories is given in Table 1. The host range of all virus isolates was verified prior to use in these experiments. The host tropism of ecotropic viruses was determined by differential growth on NIH/Swiss mouse embryo (ME) cells and BALB/c-ME cells (28) and failure to replicate in heterologous mink lung fibroblast (CCL64) cells (19). Conversely, isolates of xenotropic virus were able to replicate in CCL64 cells but not in ME cells. 545
546
J. VIROL.
O'DONNELL AND STOCKERT
TABLE 1. Viruses studied Mouse origin
M. musculus inbred strains BALB/c
Virus"b
History-'
WN1802N CL B5
Ecotropic (N) virus, formerly referred to as BALB/c-S2N, isolated from normal spleen (18); NIH-ME (®); cloned and propagated in III6A cells (O'Donnell et al., Virology, in press) Ecotropic (B) virus, formerly referred to as BALB/c-S2B, isolated from same normal spleen as WN1802N (18); BALB/c-ME 21; cloned and propagated in II16A cells (O'Donnell et aT, Virology, in press) Ecotropic (B) virus isolated from BALB/c radiation-induced leukemia RL d 1 (36); cloned directly from a cell-free extract of tumor tissue and propagated in II16A cells (unpublished data) Xenotropic virus induced from a BALB/3T3 cell line after 5-bromodeoxyuridine treatment (7); propagated in CCL64 (mink lung fibroblast) cells after passage in CCL60 (rabbit cornea) cells Xenotropic virus isolated after cocultivation of BALB/c and NIH/Swiss splenocytes with CCL60 cells (38); propagated in CCL64 cells Ecotropic (N) virus (pool 1920) isolated from normal spleen (J. W. Hartley, personal communication); NIH-ME cloned and propagated in III6A cells Ecotropic (B) virus (pool 1909) isolated from same normal spleen as B6tN-(J. W. Hartley, personal communication); BALB/c-ME Q5); cloned and propagated in III6A cells. Ecotropic (N) virus isolated from spleen of a mouse bearing the EdG2 leukemia (33); cloned directly from a cell-free extract of tumor tissue and propagated in I116A cells (unpublished data) Ecotrop c (N) virus isolated from AKR leukemia (18); NIHME (m; cloned and propagated in III6A cells Xenotropic virus recovered from RD human tumor cells passaged in immunosuppressed NIH/Swiss mice (44); propagated in CCL64 cells after passage in RD cells Xenotropic virus isolated after spontaneous activation of NZB CL S2 cells (25); propagated in 64J1 cells Ecotropic (N) virus isolated after in vivo passage of C3Hf/Gs thymic lymphoma extract in C3Hf/Gs mice and subsequently in W/Fu rats (24); cloned free from sarcoma virus in NIH/3T3 cells (1); propagated in II16A cells Ecotropic (N) virus originally isolated from the spleen of a Swiss mouse bearing an Ehrlich ascites carcinoma (13); multiple in vivo passage in Ha/ICR, DBA/2 mice; NIH/3T3 propagated in III6A cells Ecotropic (B) virus recovered after Fs plus B/T-L virus (43) mixed infection of a BALB/c mouse in vivo (29); BALB/3T3 propagated in III6A cells Ecotropic (NB) virus originally isolated from a BALB/c mouse bearing an ascites tumor of Swiss mouse strain origin (35); multiple in vivo passage in BALB/c, NIH/Swiss mice; BALB/c-ME (, NIH-ME (©); cloned and propagated in III6A cells Ecotropic (NB) virus originally isolated from a BALB/c mouse bearing sarcoma 37 of A/LN mouse strain origin (30); cloned from Electro-Nucleonics preparation (lot no. 445-40-12A) and propagated in III6A cells Ecotropic (N) virus isolated after cocultivation of pooled M. musculus subsp. molossinus spleen and kidney cells with NIH/3T3 cells (26); propagated in I116A cells
WN1802B CL Dl
RL1 CL 2
S16CL10(I)
MLC 60
C57BL/6
B6N CL A3 B6B CL D3
EdG2 CL A6
AKR
AKR-L1 CL G12
NIH/Swiss
AT124
NZB
NZB
Uncertain
Kirsten
Friend, Fs
Friend, F, Rauscher CL 1
Moloney CL H6
M. musculus subsp. molossinus
Molossinus-N
(D; (D;
VOL. 20, 1976
MuLV-INDUCED G,x AND GCSA CELL SURFACE ANTIGENS
547
TABLE 1-Continued Mouse origin
Virus&b
Historycd
Xenotropic virus isolated after cocultivation of pooled M. musculus subsp. castaneus spleen and kidney cells with FCf2Th (dog thymus) cells (26); propagated in CCL64 cells M. caroli Caroli Xenotropic virus induced from M. caroli cells after 5-bromodedeoxyuridine treatment (27); propagated in CCL64 cells a The sources of viruses used were as follows: WN1802N, WN1802B, B6N, B6B, AKR-L1, and Rauscher MuLV's from J. W. Hartley and W. P. Rowe, National Institute of Allergy and Infectious Diseases; S16CL10(I), MLC 60, and AT124 from C. Sherr, National Cancer Institute; Molossinus-N and Castaneus-X from U. Rapp, National Cancer Institute; NZB from J. Levy, University of California; Kirsten MuLV and Caroli virus from J. Stephenson, National Cancer Institute; Friend Fs and FT from F. Lilly, Albert Einstein College of Medicine. b Ecotropic viruses were cloned prior to use in these experiments as described in the text and are indicated by clone (CL) designations. ' N-, B-, or NB-tropic host range of ecotropic viruses is indicated in parentheses after the ecotropic designation. d Circled numbers indicate the number of cell-free passages in vitro.
M. musculus subsp. castaneus
Castaneus-X
Cells. Clone III6A of the feral ME line WM1511 and XC cells were provided by J. W. Hartley. The III6A cell line, which is also called SC-1, does not exhibit Fv-1 restriction of naturally occurring MuLV and is sensitive to both N- and B-tropic viruses (16). These cells were recloned prior to use in these experiments. ME cells were prepared from embryos in day 14 to 16 of gestation (17) and stored prior to use at - 196°C. Pregnant NIH/Swiss and BALB/c mice were kindly provided by W. P. Rowe. Strain 129, 129-Glx-, and C57BL/6 mice were from our own breeding colony. A W/Fu cell line was constructed from W/Fu rat embryo cells according to the 3T3 protocol (45). Pregnant W/Fu rats were obtained from our own breeding colony. NRK cells were provided by S. Aaronson, National Cancer Institute, CCL64 and 64J1 cells by C. Sherr, and JLSV9 cells by T. Pincus, Memorial Sloan-Kettering Cancer Center. For virus producer cell lines, the virus released is indicated in parentheses after the cell line designation. Cells were grown in Dulbecco modified Eagle minimal essential medium (EMEM) supplemented with 250 U of penicillin per ml, 250 ,ug of streptomycin per ml, and 10% heat-inactivated fetal calf serum (Microbiological Associates, Bethesda, Md.). Cell cultures were assayed for the presence of mycoplasma (5) by D. Armstrong, Memorial Sloan-Kettering Cancer Center. XC plaque and infectious center assays. Ecotropic virus XC plaque titrations and infectious center determinations of the percentage of productively infected cells per culture were performed on II16A indicator cells at 37°C (34). Microtiter XC plaque assay. The wells of Falcon Microtest II plates were planted with 2.5 x 103 I6A cells in 0.1 ml of Dulbecco growth medium containing 10 ug of polybrene per ml (Aldrich Chemical Co., Milwaukee, Wis.). On the next day the microcultures were infected by transferring approximately 0.02 ml of culture fluid from the wells of a master plate by means of a replica plating device
(Cooke Laboratory Products, Alexandria, Va.). After 4 days of incubation at 37°C the culture fluids were removed, and the cells were UV irradiated overlaid with 104 XC cells, and developed as described above in the standard XC plaque assay. Virus cloning. MuLV was cloned at limiting dilution using a modified microtiter method (39; O'Donnell et al., Virology, in press). Approximately 1 x 105 to 2 x 105 III6A cells were planted in 60-mm petri dishes (Falcon), infected with virus 18 to 24 h later at a multiplicity of infection of 2.5 x 10-4 PFU/ cell, dispersed with 0.75% trypsin, and planted at a density of 200 cells/well in Falcon Microtest II plates such that there would be an average of 0.05 infectious centers/well. A number of experiments showed that this resulted in a Poisson distribution of infectious centers in the microtiter wells. After 2 weeks of incubation at 37°C the microtiter fluids were assayed for the presence of virus by replica plating in a microtiter XC plaque assay. After determining the distribution of virus-positive cultures, the master plate that had been stored frozen at -80°C was thawed and the culture fluids from a positive well were used to infect fresh cultures of III6A cells. The resulting cultures of chronically infected III6A cells were then used to absorb the anti-G1x and antiGCSA sera as described. From the distribution of positive cultures it was possible to calculate the probability that such a culture was derived from a single infectious center. For most clonal isolations this statistic was greater than 90%. Phenotypically mixed virus. A xenotropic virus pseudotype of WN1802N virus was constructed by treating JLSV9 cells that were productively infected by WN1802N virus (105 PFU/ml of culture fluid) with 5-iododeoxyuridine (IUdR; Calbiochem, San Diego, Calif.), which is known to induce endogenous xenotropic virus from these cells (9). Cells were treated with 10 ,tg of IUdR per ml for 24 h, washed, and refed. Culture fluids were harvested at 3 days after the addition of IUdR, the time at which maximum yields of endogenous virus were obtained from
548
O'DONNELL AND STOCKERT
uninfected JLSV9 cells. These fluids were used as the stock of ecotropic/xenotropic, phenotypically mixed virus. RDDP assay. Culture fluids were clarified at 1,500 x g for 15 min at 4°C and then concentrated approximately 30-fold by centrifugation at 30,000 x g for 90 min. Virus pellets were resuspended (37) and assayed for RNA-directed DNA polymerase (RDDP) activity as described by Stephenson et al. (39), using poly(rA) oligo(dT),2,, (5:1 complex, PL Biochemicals, Milwaukee, Wis.) as template primer and [3H]TTP (2.4 Ci/mmol, Schwarz/Mann, Orangeburg, N.Y.) as substrate. Assays were performed under conditions in which the incorporation of [3H]TTP was linearly proportional to the concentration of enzyme in the reaction mixture. Activity is expressed as total picomoles of TTP incorporated in culture fluids per 106 cells. Immunofluorescence. The percentage of antigenpositive cells per culture was determined by indirect immunofluorescence (21), using rabbit antiRauscher p30 provided by W. D. Hardy, Jr., Memorial Sloan-Kettering Cancer Center (23). Fluorescein-conjugated goat anti-rabbit immunoglobulin was purchased from Hyland Laboratories, Inc., Los Angeles, Calif., and used at 1:50 dilution in phosphate-buffered saline. Cytotoxicity assay. For the cytotoxicity assay (42), serial dilutions (50 ,l) of antiserum were made in medium 199 and mixed with equal volumes of target cells (5 x 106 cells/ml) and appropriately diluted complement. After incubation for 45 min at 37°C, viability counts were made in the presence of trypan blue. Glx-typing system. For the Glx-typing system (42), the standard G,x antiserum produced in rats was used: (W/Fu x BN)F, anti-W/Fu leukemia (C58NT)D, abbreviated "anti-NTD." The target cells were thymocytes from strain 129 mice and the complement source was absorbed rabbit serum diluted 1:10. GCSA-typing system. For the GCSA-typing system (33), the standard GCSA antiserum was used: C57BL/6 anti-AKR spontaneous leukemia K36, abbreviated "B6 anti-K36." The target cells were E dG2 leukemia cells (transplanted C57BL/6 leukemia induced by passage A Gross virus) and the complement source was pooled guinea pig serum diluted 1:2. Test for absorption of Glx and GCSA antibody by cells. Antisera were appropriately diluted, mixed with washed, packed cells, and incubated for 30 min at 4°C to test for the absorption of Glx and GCSA antibody by cells (42). After centrifugation at 900 x g for 10 min the supernatant was tested in the standard cytotoxicity assay. For G1x tests anti-NTD serum was diluted 1:100 or 1:150 (approximately three twofold dilutions below the end point of 50% cytotoxicity of the serum with 129 thymocytes). The proportion of diluted serum to packed cells was 2:1. For GCSA tests B6 anti-K36 serum was diluted 1:10 or 1:15 (approximately two twofold dilutions below the end point of 50% cytotoxicity of the serum with E d G2 cells). The proportion of diluted serum to packed cells was 1:1.
J. VIROL.
The cells used for absorption were confluent cul-
tures of uninfected control cells or productively infected cells. Cultures were washed once with Ca and
Mg-free phosphate-buffered saline and dispersed by incubation with 0.05% EDTA for 5 min at 37°C. Cells
were harvested by centrifugation at 500 x g and washed two times in EMEM. After a final wash in medium 199, the cells were packed by slowly increasing the speed of centrifugation to 900 x g over a
period of 10 min.
In some experiments the transplanted leukemias EL4 and ERLD, which are carried in C57BL/6 mice (33), were used as additional controls. The cells were washed two times in EMEM and one time in medium 199 as described above. Quantitative absorption test of Glx. For the quantitative absorption test of Glx (42; see Fig. 2), different numbers of cells, washed and packed as described above, were mixed with 60 pAl of anti-NTD serum diluted 1:600 (approximately one twofold dilution below the end point of 50% cytotoxicity of the serum with 129 thymocytes) and incubated for 30 min at 4CC. After centrifugation, the absorbed antibody samples were tested in the cytotoxicity assay with 129 thymocytes as target cells as described above.
RESULTS
Induction of Glx and GCSA antigens by ecotropic MuLV. Induction ofthe cell surface antigens GI, and GCSA after exogenous infection of cultured cells by different MuLV isolates was measured by the ability of such cells to absorb cytotoxic antibodies from the standard typing sera. Figure 1 shows the pattern of antigen expression for N- and B-tropic MuLV of BALB/c origin. The cytotoxic reactivity of the G,,- and GCSA-typing sera was completely absorbed by III6A(WN1802N) cells, indicating the expression of both antigens on these cells. II16A(WN1802B) cells completely absorbed GCSA but not G,x activity and thus appear to express only GCSA on their cell surface. The G1x-negative (G,x-) phenotype of III6A cells infected by the B-tropic isolate was confirmed by quantitative absorption tests shown in Fig. 2 in which anti-NTD serum was absorbed with varying numbers of infected cells and assayed for residual cytotoxicity. The cytotoxic reactivity of the GI, antiserum was completely removed by absorption with 7 x 105 II16A(WN1802N) cells, whereas no absorption was observed with up to 4 x 10"; II16A(WN1802B) cells. Since the producer lines release equivalent amounts of progeny virus, it is unlikely that this result can be attributed to a quantitative difference in the level of viral gene expression, but probably represents a qualitative difference between the gp7O molecules of these two viruses. This same analysis was extended to a number of other ecotropic MuLV isolates. Produc-
t _A 1=nA_ ~
VOL. 20, 1976
MuLV-INDUCED AX AND GCSA CELL SURFACE ANTIGENS
F~~~ m " 00 Qs)
~
v
100
200
d400
15
w
(n
I
800
§I 1.
I
E Q) v
ok
6A(WN1802B)
5En
j _ / I _G .
QU
549
16A(WN1802N)
GC_
50 3
-i
1-
vs
30
0
60
1/Antiserum dilution
FIG. 1. Absorption of G,x and GCSA cytotoxic antibodies by III6A cells infected by WN1802N (0) and WN1802B (S) viruses. For G1x, anti-NTD serum (1:100 dilution) and, for GCSA, B6 anti-K36 serum (1:15 dilution) were absorbed with washed, packed cells and assayed for residual cytotoxicity with the standard target cells as described in the text. Controls included unabsorbed antiserum (-) and antiserum absorbed with uninfected III6A cells (A) and EL4 leukemia cells (x), whose phenotype is Glx-IGCSA-. The levels of virus production in the cultures used for absorption are given in Table 2.
tively infected HII6A cell lines were used to absorb the appropriate typing sera. The results are given in Table 2, which includes measurements of the level of virus production at the time the cells were harvested for absorption. Complete absorption (+ +) represents removal of cytotoxic antibodies as shown in Fig. 1 for WN1802N-infected cells. Failure to absorb cytotoxicity was scored as negative (-) as observed for the GIX antiserum with WN1802B-infected cells (Fig. 1). Partial absorption (+) was observed only in the case of GCSA typing, where the titer of the absorbed antiserum was lowered as compared with the negative absorption of uninfected control cells (Fig. 3). Of the GrossAKR type viruses only the B-tropic virus of C57BL/6 mice (B6B) showed partial absorption of GCSA antibody, although this was a common characteristic of the Friend-Moloney-Rauscher (FMR) subgroup viruses. However, cells infected by B6B virus completely absorbed Glx activity. To determine whether the partial absorption of anti-GCSA activity was due to limited GCSA cross-reactivity or to quantitative differences in the level of GCSA expression on infected cells, we performed double absorption experiments shown in Fig. 4. B6 anti-K36 serum was absorbed twice with approximately 40 x 106 to 60 x 106 infected cells. Residual cytotoxicity was measured after each absorption. After the second absorption with cells infected by B6B and
