Effect of murine host genotype on MCF virus expression, latency, and leukemia cell type of leukemias induced by Friend murine leukemia helper virus

VIROLOGY

128, 221-233

Effect

BRUCE Department

(1983)

of Murine Host and Leukemia Friend CHESEBRO,’

Genotype on MCF Virus Cell Type of Leukemias Murine Leukemia Helper

JOHN

L. PORTIS,

KATHY

Expression, Latency, Induced by Virus

WEHRLY,

AND

of Health and Human Senrices, Public Health Service, Natixmd Institutes Institute of Allergy and Igectiozls Diseases, Laboratory of Pereistent Viral Rocky Mountain Laboratories, Ham&m, Montana 59840 Received

Februurg

11, 1983; accepted

April

JANE

NISHIO

of Health, Diseases,

National

6, 1983

Leukemias induced by neonatal inoculations of several mouse strains with different strains of Friend murine leukemia helper virus (F-MuLV) were followed for time of disease onset, cytochemieal analysis of predominant cell types in leukemic organs, and expression of infectious mink cell focus-inducing (MCF) viruses detected by mink cell foci or MCF-specific monoclonal antibodies. Most BALB.B and IRW mice had a rapidly appearing, severe anemia and hepatosplenomegaly consisting of erythroid cells. MCF viruses were usually isolated from enlarged spleens of IRW mice. In contrast, C5’7BL/lO mice had a lower incidence of disease and much slower course. Splenomegaly and lymphadenopathy with mild anemia were seen, and the predominant cell types were either myeloid (chloroleukemia) or lymphoid. MCF viruses were never isolated from this mouse strain. (C5’7BL/lO X IRW)FI mice were intermediate in latency, but all mice had disease by 8 months. Myeloid, lymphoid, and some mixed leukemias with an erythroid component were observed, but in no case did we see the severe anemia or pure erythroid involvement typical of IRW and BALB.B mice. MCF viruses were, however, isolated from 22% of these mice regardless of leukemia cell type. DBA/L mice had a disease pattern similar to the (C57BL/lO X IRW)FI mice, and MCF viruses were isolated from three of six mice tested. Inoculation of IRW mice with the low virulence B3 strain of F-MuLV produced disease with a longer latency than F-MuLV 5’7, but similar cell types were transformed by both viruses. In vitro cell lines were derived from 14 mice, and most were tumorigenic in tivo. Three lines released infectious MCF virus, and three others expressed MCF-specific cell surface antigens but did not release virus. Eight lines expressed no MCF infectious virus or viral antigens. Several lines released infectious xenotropic viruses and/or expressed xenotropic MuLV cell surface antigens recognized by monoclonal antibodies reactive with xenotropic viruses. The lack of MCF expression in many primary leukemic tissues as well as in in vitro derived leukemia cell lines of C57BL/lO and (BLO X IRW)FI mice suggested that MCF virus generation and expression may not be required for leukemogenesis in some mouse strains or in some hemopoietic lineages.

observed to induce leukemia characterized by anemia, enlargement of the spleen, liver, and occasionally lymph nodes and thymus, with a latency of less than 10 weeks (Steeves et ab, 1971). Newborn Swiss, CXBL, and C3H mice were susceptible as were adult Swiss mice. Microscopically, the majority of cells in all these instances were thought to be lymphoblasts, although occasionally myeloid (chloroleukemia) cells were observed (Steeves et aL, 1971). Subsequently, Troxler and Scolnick (1978)

INTRODUCTION

Friend murine leukemia virus (F-MuLV), the helper virus of the Friend virus complex (Friend, 195’7), was initially separated from the defective spleen focus-forming virus (SFFV) component by passage in newborn rats or C57BL mice (Dawson et ah, 1966,1968). F-MuLV free of SFFV was ’ Author addressed.

to whom

requests

for reprints

should

be 221

0042-6822/83 Copyright All rights

$3.00

Cl 1983 by Aeademle Press. Inc. of reproduction in any form reserved.

222

CHESEBRO

noted that several cloned strains of FMuLV induced leukemia characterized by splenomegaly and anemia only 3 weeks after neonatal inoculation of BALB/c or NIH Swiss mice. These leukemia cells appeared to be erythroid because a high percentage contained spectrin and produced erythroid colonies when cultured with erythropoietin (MacDonald et ul, 1980). More recent studies have shown that all three types of leukemias (lymphoid, myeloid, and erythroid) can be induced after inoculation of F-MuLV (Shibuya and Mak, 1982; Wendling et I.& 1983). However, there is at present no explanation for the variability in types of leukemias observed after inoculation of this virus, and it remains unclear whether various strains of F-MuLV differ in their potential to induce transformation of these different cell types. A further complicating factor in FMuLV-induced leukemogenesis has been the isolation of recombinant dual-tropic mink cell focus-inducing (MCF) viruses from leukemic spleens (Troxler et cd, 1978; Ishimoto et al, 1981). Based both on the extensive sequence homology between these MCF viruses and SFFV and also on the strong correlation between appearance of similar MCF viruses and spontaneous leukemia in AKR mice (Hartley et cd., 1977), it was speculated that presence of recombinant MCF viruses in defective (e.g., SFFV) or nondefective forms might be required for all retrovirus-induced leukemogenesis in mice (Troxler and Scolnick, 1978). Furthermore, murine host genes were found to influence susceptibility to FMuLV-induced leukemia (Ruscetti et cd, 1981; Shibuya and Mak, 1982), and in the case of DBA/2 mice recent data suggested that this resistance might possibly be mediated by viral interference due to expression on normal cell surfaces of gp70 molecules antigenically related to MCF viruses (Bassin et cd, 1982). In an effort to understand some of the complex interactions occurring during FMuLV-induced leukemogenesis we have compared the latencies and types of cells transformed by various F-MuLV strains in several mouse strains. Furthermore, using conventional virus isolation techniques

ET

AL.

as well as monoclonal antibodies specific for MCF viruses we have studied MCF virus expression in the primary leukemic tissues and in leukemia cell lines derived from these organs. Our results indicated that the mouse strains differed markedly in the types and latencies of leukemias induced. MCF virus expression also varied widely and appeared not to be required for leukemogenesis in all mouse strains. MATERIALS

AND

METHODS

Mice. All mice used for neonatal virus inoculation were bred at Rocky Mountain Laboratories. C57BL/lOSn (BlO) parental mice were purchased from the Jackson Laboratory, Bar Harbor, Massachusetts. BALB.B (H-2b-congenic with BALB/c) mice were originally obtained from Dr. F. Lilly, Albert Einstein School of Medicine, Bronx, New York. Inbred Rocky Mountain white (IRW) mice were derived from an outbred colony maintained at the Rocky Mountain Laboratories. The inbred line was started from a randomly selected pair and was inbred by brother-sister mating for 15 generations. F1 hybrid crosses were made using C57BL/lO females and BALB.B or IRW males. V&-uses. Molecularly cloned F-MuLV strain 57 was used for most experiments (Oliff et uZ., 1980). Limiting dilution cloned F-MuLV strains 67 and 25-57NB were described previously (Chesebro et uZ., 1983). F-MuLV strain I-5 (Mathieu-Mahul et al, 1982) was kindly provided by Dr. S. Gisselbrecht, Laboratoire d’Immunologie et Virologie des Tumeurs, Hopital Cochin, Paris, France. F-MuLV strain B3 (Linemeyer et uL, 1980) was obtained from Dr. L. Evans at the Rocky Mountain Laboratories. All F-MuLV strains were recloned by limiting dilution on SC-1 cells (Hartley and Rowe, 1975), and stocks for inoculation were prepared from supernatant fluids of these cell cultures. Ecotropic F-MuLV was assayed by SL- assay on D56 cells (Bassin et uL, 19’71) or by XC plaque assay on SC-1 cells. All F-MuLV stocks were negative for spleen focus-forming virus (SFFV) as they did not give rapid splenomegaly or spleen foci after inoculation of

FRIEND

MCF

VIRUS

susceptible adult mice (Axelrad and Steeves, 1964). Le~~~mogenesis experiments. Mice less than 24 hr old were inoculated intraperitoneally with 0.05 ml of undiluted tissue culture supernatant fluid containing FMuLV (500 to 5000 SL-PFU). Mice were followed for development of splenomegaly and lymphadenopathy by palpation under ether anesthesia. Mice with grossly enlarged spleens (>0.8 g) or lymph nodes were sacrificed for further analysis. Cytology. Single cell suspensions of enlarged organs were made in phosphatebuffered balanced salt solution (PBBS) (Chesebro and Wehrly, 1976). Two to 20 x lo6 cells were pelleted by centrifugation and resuspended in l-2 vol of PBBS. Cells were applied to scored microscope slides with a small paint brush and then air dried. Slides for Giemsa staining were fixed l-2 min in methanol, dried, and then stained 30 min with Giemsa stain, rinsed with water, dried, and mounted. Slides for Sudan black or nonspecific esterase staining were fixed 4 min in formaldehyde vapor (4 drops 37% Formalin (Fisher Scientific, Springfield, N. J.) on a cotton ball in a sealed Coplin jar containing the slides). After airdrying slides were either stained immediately or stored at -70”. Sudan black staining was done for 40 min (Lillie and Burtner, 1953; Lillie, 1954). No counterstain was used. Staining for nonspecific esterase in the presence and absence of NaF utilized cY-naphthyl butyrate as a substrate and was done for 1 hr (Yam et al., 19’71). Slides were counterstained with toluidine blue. Cell types of grossly enlarged spleens or lymph nodes were designated as erythroid if greater than 10% of cells were positive with cr-naphthyl butyrate esterase and were designated as myeloid if greater than 10% of cells were positive with Sudan black. If greater than 80% of cells were negative with both histochemical stains, cells were designated as lymphoid and morphology after Giemsa staining was consistent with this conclusion. Mice whose spleen and/or lymph node cells met two of these criteria were designated as mixed leukemias. Virus isolation. MCF viruses were iso-

EXPRESSION

223

lated by plating l-5 X 10” cells from enlarged organs or leukemia cell lines on tissue cultures seeded 1 day earlier with l-2 x lo5 mink lung cells (CCL64) or SC-1 cells (Hartley et ak, 1977; Ishimoto et cd, 1977). Mink cell cultures were examined for the presence of mink cell foci 5 and 7 days later. Negative cultures were passaged l3 times, and monolayers were examined for foci as cells reached confluency. Supernatants from cultures with positive foci were retested for confirmation of foci by cell-free passage on to new mink cell cultures in order to avoid the ambiguities caused by having large numbers of spleen cells present on the mink cell monolayers when reading foci. After three passages mink cells and SC-1 cells were also tested with monoclonal antibodies and ‘251-protein A to detect expression of cell surface gp70 antigens specific for F-MuLV (antibodies 48 and 307) or MCF viruses (antibodies 514 and 502) (Chesebro et aZ., 1981, 1983). Antibody 514 was particularly useful as it reacted with 39 out of 40 MCF viruses tested so far. Antibody 502 was also used because it reacted with about 75% of MCF viruses studied, including the 1 MCF virus negative with antibody 514. Neither antibody 514 or 502 had reactivity with ecotropic or xenotropic viruses tested previously (Chesebro et ah, 1983). Leukemia cell lines. Cells from enlarged spleens and lymph nodes of mice were placed in Linbro TC24 culture wells at twoto three-fold dilutions ranging from 4 X lo6 cells/ml to 1.2 X lo5 cells/ml in RPM1 1640 medium plus 5 X 10m5 M 2-mercaptoethanol, penicillin (200 U/ml), and 10% fetal calf serum, selected for ability to support growth of Friend erythroleukemia cell lines. Wells with cells showing evidence of proliferation were split 1 to 2 at appropriate times, and if such passages were successful, larger dilutions were made at weekly intervals. Cell lines established were examined cytochemically with Giemsa, Sudan black, and nonspecific esterase stains, and were tested by infectious center assay for production of F-MuLV, MCF viruses and xenotropic MuLV as described above for leukemic organs. Cells were also tested for expression of cell sur-

224

CHESEBRO

face viral antigens using membrane immunofluorescence or ‘%I-protein A assays and monoclonal antibodies (Chesebro et oJ, 1981, 1983). To determine whether cells from in vitro lines were tumorigenic in vivo l-2 X lo6 cells were inoculated intravenously into adult syngenic mice. Mice were followed for organ enlargement by palpation. RESULTS

Incidence and Latency of Disease Induced by F-MuL V Strains We inoculated several strains of neonatal mice with various F-MuLV isolates to determine the types of cells transformed and extent of expression of recombinant MCF viruses in each instance. Splenomegaly and lymphadenopathy were followed by abdominal palpation under ether anesthesia. BALB.B, IRW, BlO, (BlO X BALB.B)S, and (BlO X IRW)F1 mice all had a significant incidence of splenomegaly after inoculation of F-MuLV-5’7. However, the latency was strikingly different among these strains. Most BALB.B and IRW mice developed splenomegaly from l-3 months postinoculation (Fig. l), whereas BlO, (BlO X BALB.B)Fi, and (BlO X IRW)F1 mice had no signs of disease at this time. (BlO x 1RW)Fi and (BlO X BALB.B)Fi mice began to develop organ enlargement 3 months after inoculation, and 100% of these mice were affected by 8-11 months. In contrast, BlO mice developed disease at an even slower rate, and by 10 months only 40% had splenomegaly. Organ distribution of disease also appeared to differ among the mouse strains since IRW mice only rarely manifested lymphadenopathy whereas greater than 50% of the mice of other strains had palpable lymph node enlargement (data not shown). IRW mice and (BlO X IRW)F1 mice were inoculated with three other F-MuLV strains (67, Ib5, and NB) and gave results similar to those seen after F-MuLV-57 (data not shown). However, IRW mice inoculated with F-MuLV-B3 strain differed in that fewer mice became clinically ill and

ET

AL.

the latent period was prolonged (Fig. 1). Therefore, the F-MuLVB3 strain appeared to have a reduced virulence in these mice. We have not yet tested this virus in BlO or (BlO X IRW)Fi mice. Cell Types Present in Leukemic

Organs

Previous data indicated that leukemias of different hemopoietic lineages (erythroid, myeloid, and lymphoid) could be induced by F-MuLV (Shibuya and Mak, 1982; Wendling et aL, 1983). In the present study, we wished to determine whether disease involving a particular cell type was correlated with the differences in disease latency noted among the various mouse strains inoculated. Cells from enlarged spleens or lymph nodes of leukemic mice were examined cytochemically to determine the predominant cell type(s) contributing to the organomegaly. In preliminary studies we were able to distinguish erythroid, myeloid, and lymphoid cells using two cytochemical stains. Erythroid cells were strongly positive for cu-naphthyl butyrate esterase, which could be partially inhibited by fluoride ions, (Kass, 1977; Fioritoni et al, 1980) and were negative with Sudan black. Myeloid cells contained Sudan black-positive lysosomal granules and were negative with a-naphthyl butyrate esterase. Lymphoid cells were negative with both Sudan black and esterase stains. Examples of these procedures as applied to in vitro passaged leukemia cell lines and to leukemic spleen populations are shown in Fig. 2. When mice were compared as to predominant cell types present in leukemic spleens or lymph nodes, striking differences were seen among the mouse strains (Table 1). Of 28 IRW mice, 16 had predominantly erythroid cells and 8 had combined erythroid plus myeloid cells in their spleens. Three had mostly myeloid cells and only one had combined erythroid and lymphoid cells. In contrast, none of the BlO mice had prominent erythroid involvement. Of 14 mice studied, 12 had myeloid cells, 5 had lymphoid cells, and 2 had mixed myeloid and lymphoid cells. (BlO X 1RW)FI mice were similar to BlO mice in that none

FRIEND

MCF VIRUS

225

EXPRESSION

MONTHS

FIG. 1. Percentage of leukemic mice (splenomegaly, lymphadenopathy, or dead) at various times after neonatal inoculation with F-MuLV strain 5’7 (---) or strain B3 (-----). The number of mice was as follows: strain 5’7-BALB.B (21 mice), IRW (35 mice), BlO X IRW (24 mice), BlO X BALB.B (14 mice), and BlO (35 mice). Strain B3-IRW (22 mice).

had solely erythroid proliferation. However, 7 of 31 had combined erythroid plus myeloid cells which was not observed in BlO mice. The rest were evenly divided among myeloid, lymphoid, and mixed myeloid and lymphoid categories. So far, only 6 DBA/Z mice have been studied. One had erythroid, four myeloid, and one mixed lymphoid plus myeloid leukemia. The pattern of organ enlargement was of relatively little assistance in predicting the predominant cell types in these leukemic mice. Mice with splenomegaly in the absence of lymphadenopathy were found to have erythroid, myeloid, lymphoid, and erythroid plus myeloid patterns, and mice with both ~enlarged spleens and lymph nodes were found to have myeloid, lymphoid, myeloid plus lymphoid, erythroid plus myeloid, and erythroid plus lymphoid patterns (data not shown). Furthermore, the individual F-MuLV strains inoculated did not appear to influence the predominant cell types seen in IRW and (BlO

X IRW)F1 mice where several different ruses were tested (data not shown). Isolation of Recombhad Leukemic Organs

vi-

MCF Viruses from

To study whether MCF virus expression was associated with particular leukemia cell types in different mouse strains, cells from leukemic organs were plated as infectious centers in vitro on SC-l cells and mink lung cells. MCF viruses were detected either by ability to produce typical mink cell foci or by induction of “MCF-specific” cell surface antigens detected by monoclonal antibodies (514 and 502). Of the 84 mice tested, 25 were MCF positive by both tests (Table 2). Ten mice were MCF antigen positive, but no mink cell foci were observed. MCF antigens induced by viruses from these 10 mice were detected in SC-1 cells only, and were not successfully transferred to mink cells, thus accounting for the lack of mink cell foci. The inability to

226

CHESEBRO

ET

AL.

CELL LINES Erythroid

Diff. Myeloid

Undiff. Myeloid

Lymphoid

Esterase

1” SPLEENS Erythroid

Myeloid

Mixed E/M

Lymphoid

FIG. 2. (Top) Comparison of four types of in vitro-grown leukemia cell lines using Giemsa, cu-naphthyl butyrate esterase, and Sudan black staining procedures. Cell lines used were as follows: erythroid, Y57 clone 2C (Collins and Chesebro, 1980); differentiated myeloid, 7320, undifferentiated myeloid, 2349; and lymphoid, 1593. (Bottom) Comparison of primary (1”) leukemic spleens of mice with four different pattern of leukemia using Giemsa, oi-naphthyl butyrate esterase, and Sudan black stains. Cells reactive with esterase were dark red (erythroid cell line and spleen, and mixed leukemic spleen), whereas nonreactive cells were stained pale blue. This obvious color difference is not so readily perceived in the black and white photomierograph.

infect mink cells with these viruses appeared to be due to the heavy coinfection of the SC-1 cells by ecotropic F-MuLV

which may have caused most MCF virus to be released as an ecotropic pseudotype and thus be unable to infect mink cells.

FRIEND

MCF VIRUS TABLE

DISTRIBUTION

EXPRESSION

227

1

OF PREDOMINANT CELL TYPES IN ENLARGED SPLEENS AND LYMPH OF MICE INOCULATED AT BIRTH WITH F-MuLV

NODES

Mouse strain Cell type” Erythroid Myeloid Lymphoid Erythroid plus myeloid’ Lymphoid plus myeloid” Erythroid plus lymphoid” Total

IRW

BlO

(BlO X IRW)Fl

DBAA

16 3 0 8 0 1

0 12 5 0 2 0

0 12 6 7 6 0

1 4b 0 0 1 0

28

19

31

6

’ Cells from enlarged spleens and lymph nodes of mice inoculated at birth with F-MuLV were studied with Sudan black and nonspecific e&erase stains to determine the predominant cell types as described in the Methods section. Data were pooled for mice inoculated with F-MuLV strains 57, B3, 6’7, NB, and I-‘. b Lymph node and spleen cells from two of these DBA/Z mice were weakly positive with cY-naphthyl butyrate esterase inhibitable by NaF, and were strongly Sudan black positive. Thus, they fit criteria of both myeloid and monocytic cells and could best be described as myelomonocytic. ‘Some mice had mixed leukemias in the same organ, but in other individuals the predominant cell types were different in enlarged lymph nodes and spleen.

Cells from the different mouse strains varied in expression of MCF viruses (Table 3). Twenty-five of 28 IRW mice were MCF positive, whereas none of the BlO mice were positive. The (BlO X IRW)Fi mice were intermediate in that 7 of 31 mice were MCF positive. Data were available on six DBA/ 2 mice, and three of these six were MCF positive. Thus, there was no obligatory association of MCF virus release with presence of leukemia in most of these mice. The association of MCF expression with a particular predominant cell type in enlarged organs was difficult to confirm. Erythroid involvement, whether alone or in combination with another cell type, appeared strongly correlated with MCF expression in IRW mice. However, when IRW and (BlO X IRW)F1 mice with the same types of leukemia were compared, it was clear that the incidence of MCF virus detection was much higher in the IRW mice regardless of leukemia cell types. An&&s

of Leukemia

Cell Lines Derived with F-MuLV

from Mice Inoculated

Cell populations primary leukemic

obtained directly from organs are heteroge-

neous and contain many nonleukemic cells. In order to characterize the leukemic cells further, cells from enlarged organs of all mice presented in Table 3 were cultured in vitro to attempt to grow tissue culture lines. Cells from successfully derived lines were inoculated into syngeneic mice to see whether tumors developed. Fourteen lines were successfully grown in vitro, and most grew in and killed recipient mice within 3 weeks of inoculation (Table 4). Of the 14 lines, 11 were myeloid, 2 lymphoid, and only 1 erythroid. This skewed distribution differed markedly from that seen in the primary enlarged organs. In particular, only one erythroid line was derived in spite of the fact that many mice had a predominance of erythroid cells in their spleens. This could indicate either that our culture conditions selected against erythroleukemia cells or that many of the cases of prominent erythroid proliferation in vivo were not true neoplastic states. This did not appear to be the case for the myeloid cells, as many lines were derived, and all appeared neoplastic in character. These cell lines were also analyzed for release of infectious viruses and for mem-

228

CHESEBRO TABLE

2

brane expression of viral antigens. All released infectious F-MuLV, but only three (2800,2349, and 7119) released MCF virus. Three cell lines (5513, 7235, and 7229) reacted with MCF-specific antibodies (502, 514, and 508), but did not release detectable infectious MCF viruses (Table 4). Several of the lines did release mink cell-tropic virus which did not cause mink cell foci or induce MCF-specific antigens. These appeared to be xenotropic viruses and reacted with monoclonal antibodies (516 and 518) known to recognize some viruses of this

CORRELATION OF MCF VIRUS DETECTION AS MEASURED BY MINK CELL FOCI AND REACTIVITY WITH MCFSPECIFIC MONOCLINAL ANTIBODIES

Mink cell focib MCF antigens detected”

+

+ -

25 0

10 49

“SC-1 cells or mink lung cells were infected by cocultivation with l-5 X lo6 cells from enlarged spleens or lymph nodes. Cells were passed three to four times and then were assayed for cell surface reactivity with MCF-specific monoclonal antibodies 514 and 502 using the =I-protein A method. b MCF virus detected by infectious center assay of cells from enlarged spleens or lymph nodes from mice inoculated at birth with F-MuLV. Assays were done both directly on mink lung cells and on SC-l cells, which were uv irradiated on Day 3 and then overlaid with mink lung cells. Supernatant fluid from all cultures with foci or suspicious cytopathic effect was retested by cell-free infections at lo-fold dilutions on mink lung cells for confirmation of mink cell focusinducing activity.

TABLE FREQUENCY

OF

MCF VIRUS ISOLATION OF MICE

INOCULATED

ET AL.

group. DISCUSSION

The present results suggested that the patterns of F-MuLV-induced diseases in different mouse strains were more diverse than previously realized. There was also considerable heterogeneity within each mouse strain, and in some cases leukemias of differing lineages even appeared to coexist in the same individual (Table 1). Although our cytochemical approaches to characterization of the cells in enlarged organs were useful, it was not possible to 3

FROM ENLARGED SPLEENS OR LYMPH AT BIRTH WITH F-MuLV

NODES

Mouse strain Cell type’ Erythroid Myeloid Lymphoid Erythroid plus myeloid Lymphoid plus myeloid Erythroid plus lymphoid Total

IRW

BlO

(BlO

x 1RW)Fi

DBA/2

15/16b 2/3 o/o 713 o/o l/l

o/o o/12 o/5 o/o o/2 o/o

o/o 2/12 l/6 2/7 2/6 o/o

l/l 2/4 o/o o/o

25/23”

o/19

7/31

3/6

O/l

o/o

a Predominant cell types in enlarged spleens and lymph nodes were determined as described in the Methods section. bNumber of mice with MCF viruses isolated from enlarged spleens or lymph nodes as detected by mink cell foci or reactivity with MCF-specific monoclonal antibodies (514 or 502) divided by total number of mice studied in each category. ‘Six IRW mice inoculated with B3 virus were included in this column. MCF viruses were isolated from three of four mice with predominant erythroid cell type, and from two of two mice with predominant myeloid cell type.

BlO

B10 BlO BlO BlO BlO B10 BlO

BlO BlO

DBA/2

1593

7320 5513 7233 7202 7201 7235 7229

7301 7303

7557

strain

IRW IRW IRW IRW IRW IRW IRW

57

57 57

67 67 NB NB NB 1-6 r6

57

33

67 NB

myeloid

myeloid lymphoid

myeloid myeloid myeloid myeloid myeloid myeloid myeloid

lymphoid

myeloid

erythroid myeloid

Cell Web

+

+ +

F-MuLV,

+

F-MuLV

F-MuLV, F-MuLV,

F-MuLV F-MuLV F-MyLV, F-MuLV, F-MuLV F-MuLV F-MuLV,

F-MuLV,

F-MuLV, F-MuLV,

Viruses isolatedd

+

+ +

In viva growthc

xeno xeno

xeno

xeno xeno

xeno

MCF

MCF MCF

+

+ +

+ + + + + + +

Id

+

+ +

307

LEUKEMIA CELLLINESDERIVEDFROMMICEINOCULATEDWITH

4

+ -

+ + -

+

+ +

514

Viral

-

+

+ +

502

F-MuLV

-

-

-

+ +

+ -

-

+ + -

+ -

+ + -

-

+ +

+ -

-

518

516

508

antigense

’ F-MuLV strain inoculated; NB = 25-57NB. ‘Cell type determined cytochemically. ’ 3-10 X lo6 cells were inoculated intravenously into syngeneic adult recipient mice which were then followed by palpation for appearance of splenomegaly or tumor masses. Cells which were tumorigenic in viva usually produced signs or symptoms in recipient mice in lo-30 days. d MuLV isolated by infectious center assay either directly on mink lung cells or on SC-1 cells. Duplicate SC-1 were uv irradiated and overlaid with mink lung cells. Mink cells and SC-1 cells were analyzed for expression of cell surface viral antigens using monoclonal antibodies specific for F-MuLV or MCF viruses. Mink cells were also observed for appearance of typical mink cell foci. Viruses spreading in mink cells which failed to give mink cell foci but were detected with fluorescein-labeled goat anti-MuLV serum were assumed to be xenotropic viruses. A assays using monoclonal antibodies. ’ Cell surface viral antigens on leukemia cell surfaces were detected by membrane immunofluorescence of Y-protein Antibody 307 is specific for Friend or Rauscher MuLV gp70 (Chesebro et al., 1981). Antibodies 502, 514, and 508 react with a variety of MCF viruses but not with any ecotropic, xenotropic, or amphotropic viruses tested; antibody 516 reacts with some MCF viruses and some xenotropic viruses; antibody 518 reacts with some xenotropic viruses) Chesebro et al, 1983). ’ Ig: Cell line 1593 is a B-lymphocyte line which expresses mouse immunoglobulin detectable by [%]methionine labeling or by cell surface reactivity with anti-mouse immunoglobulin serum. Therefore, this cell line could not be typed for cell surface viral antigens by these methods.

x X x x x X x

X BALB.B

IRW

7119

Mouse

BALB.B BALB.B

line

2800 2849

Cell

Virus inoculated”

PROPERTIESOF~~ Vitro

TABLE

!z

2

ii iz G

230

CHESEBRO

distinguish between true neoplastic leukemia cells and normal hemopoietic cells appearing in unusual numbers in the spleen as a compensatory response to anemia or leukopenia. One report (Oliff et al, 1981) suggested that transfusion treatment of the anemia associated with early erythroid proliferation in NIH Swiss mice resulted in prolonged survival and occasional recovery from splenomegaly, thus shedding further doubt on the leukemic status of the proliferating erythroid cells early in the disease. In the present study, we addressed this issue by attempting to derive in vitro cell lines from most of the mice. In most cases the cells were tumorigenic in vivo, which indicated that at least some of the cells from the primary organs used to derive these lines were truly neoplastic. However, it might not be valid to extend this interpretation to the many mice whose organs failed to yield in vitro lines. This was particularly a problem with the erythroid and lymphoid cell types where only one and two lines, respectively, were obtained (Table 4). It seemed likely that our conditions for derivation of cell cultures of these lineages might not be optimal. The association and transition from anemia and erythroid leukemia to myeloid leukemia has also been noted in humans, and is referred to as Di Guglielmo’s syndrome (Di Guglielmo, 1917). Mixed proliferative states are frequently seen and the leukemic status of the erythroid cells in particular is difficult to determine. This syndrome has been considered similar to the erythroleukemia induced in adult mice by Friend virus complex (Dameshek, 1969). However, it would appear to be more analogous to the severe anemia and combined erythroid and myeloid leukemias induced by F-MuLV. The disease patterns seen in mice inoculated with F-MuLV differ from the human cases because lymphoid leukemias are frequently induced in some mouse strains and this is not part of the human syndrome. In both man and mouse it remains unclear whether the heterogeneity of leukemia cell types observed represents separate transformation of different hemopoietic cell types or transformation of a single stem cell type with

ET

AL.

potential for differentiation along separate lines (Fioritoni et al, 1980; Marie et d, 1981). The ambiquity concerning the target cells for transformation following F-MuLV inoculation is further compounded because recombinant MCF viruses have frequently been isolated from F-MuLV-inoculated mice (Troxler et aL, 1978; Ishimoto et cd, 1981). It is uncertain whether different viruses might be involved in transformation of different hemopoietic cell types. Therefore, one of our main interests in carrying out this study was to investigate whether there was an association between presence of recombinant MCF viruses and appearance of leukemias of differing cell types in various mouse strains. The results indicated that there were striking differences among the mouse strains in MCF virus expression. Most IRW mice were positive, and all BlO mice were negative. (BlO X IRW)F1 hybrid mice and DBA/B mice were intermediate. Since erythroid leukemias were seen predominantly in IRW mice, it appeared that MCF virus expression was associated mainly with erythroid leukemia. In support of this possibility, BlO and (BlO X IRW)F1 mice never had this disease pattern. However, when mice of different strains with similar disease types (myeloid or mixed erythroid plus myeloid) were compared (Table 3), it seemed clear that the mouse strain itself also had a strong influence on MCF virus expression. It was surprising that no MCF viruses could be isolated from leukemias in BlO mice. The reasons for this are still unclear. It is possible that MCF viruses were never generated because the appropriate endogeneous viral sequences were not available in cells infected by the ecotropic F-MuLV in this strain. It is also possible that recombinant MCF virus might be generated but could not spread due to some host restriction. Replication defective MCF viruses might also be present in BlO mice; however, no MCF-specific viral antigens were detected by our monoclonal antibodies on in vitro cell lines derived from BlO mice (Table 4). In BlO mice, erythroid leukemias were never observed even late after F-MuLV inoculation. Friend MCF viruses

FRIEND

MCF

VIRUS

have been implicated in induction of the erythroid disease seen in NIH Swiss and other mouse strains (Troxler and Scolnick, 1978; Ruscetti et al, 1982). If Friend MCF viruses are necessary for leukemia induction in the erythroid lineage, it is possible that absence of MCF virus generation could explain the lack of erythroid proliferative states in BlO mice. In all other mouse strains tested, some individuals expressed MCF viruses and also showed some form of erythroid proliferation. However, there was resistance to early, but not late, erythroid disease in three instances: DBA/2 and (BlO X IRW)F1 inoculated with F-MuLV 57 and IRW mice inoculated with F-MuLV B3. It is possible that the reasons for resistance to early disease were different in each of these situations. Wendling et al. (1983) observed late erythroleukemia in about 20% of DBA/2 mice. In our study of DBA/B we saw one individual with pure erythroleukemia. This mouse expressed MCF virus as did two others with myeloid leukemia. Thus, the longer latency in this strain was not due to inability to generate MCF viruses, but rather was more likely due to slower MCF virus spread as has been postulated based on interference data (Ruscetti et a,!., 1981; Bassin et a,L, 1982). IRW mice inoculated with the B3 strain of F-MuLV also had a prolonged latent period for leukemia induction. Predominant erythroid leukemias were seen in a number of individuals, and these mice generated MCF viruses (Table 3). There appeared to be no general inhibition of spread of MCF viruses in this mouse strain because MCF viruses generated after inoculation of IRW mice with F-MuLV 57 were detected in large numbers of spleen cells within 1 month (data not shown). However, previous studies suggested that the B3 virus strain replicated poorly in tivo (Linemeyer et al, 1980). Thus, poor spread of the ecotropic virus might account for the long latency in this case. The B3 and 5’7 F-MuLV strains had very similar genetic structures which have indistinguishable oligonucleotide fingerprints (L. Evans, personal communication). It should be most interesting to identify the location(s) of any genetic

EXPRESSION

231

differences between these viruses because differing regions are likely to be critical for induction of the early erythroid disease. (BlO X IRW)F1 mice inoculated with FMuLV 5’7 also failed to develop early erythroid disease. These mice differed from DBA/B and IRW mice in that no “pure” erythroleukemias were observed at any time. Instead erythroid proliferation was always associated with myeloid leukemia, and it was difficult to be sure that the erythroid cells were actually leukemic since no erythroid lines were derived from these mice. Thus, it is possible that these Fi mice might be similar to the B10 parental mice in being resistant to both the early and late forms of erythroleukemia. This would suggest a resistance mechanism quite different from that of DBA/2 mice. Furthermore, this resistance would not be explained simply by the lack of MCF virus generation since 22% of these F1 mice expressed MCF viruses. Recent studies have implied that recombinant MCF viruses may play an important role in leukemogenesis in mice (Hartley et al., 1977; Fischinger et aL, 1977,1978, Van Griensven and Vogt, 1980)). In contrast, our results suggested that in some mouse strains and cell types, expression of MCF viruses might not be a required event in the process of leukemogenesis. Thus, different mechanisms of leukemogenesis might exist in different cells or mice, and “acceleration” of disease appearance by MCF generation might only be effective in some instances (Cloyd et aZ., 1980; O’Donnell et al, 1981). It also remains a distinct possibility that in some situations MCF expression is an eflect rather than a cause of leukemogenesis. MCF viruses might be more easily generated or propagated in rapidly dividing transformed cells. This could account for the inconsistent association of MCF virus expression in mice with myeloid and lymphoid leukemias and in cell lines derived from these mice. Lack of MCF virus expression in F-MuLV-induced myeloid lines was also observed by S. Gisselbrecht (personal communication) in cells derived from DBA/Z and (B6 x BALB/c)F, mice (Heard et aL, 1981; Fichelson et al., 1982). Thus, it would appear likely that in

232

CHESEBRO

lymphoid and myeloid cells of some mouse strains the F-MuLV ecotropic virus itself might be leukemogenic. This would be analogous to the situation in chickens where Iymphatic leukemia virus appears to be directly oncogenic (Hayward et al, 1981). According to this hypothesis, the heterogeneity of disease patterns observed would be increased in mice where the rapid MCF virus-associated erythroleukemia was delayed by various mechanisms to the point where slower ecotropic virus-mediated transformation events had a chance to be expressed. ACKNOWLEDGMENTS The authors thank Dr. Faramarz Naeim, Department of Pathology, UCLA School of Medicine, for his help in the initial cytochemical screening of leukemia cell lines. We also thank Dr. Leonard Evans for helpful discussions and Mrs. Helen Blahnik for typing the manuscript. REFERENCES AXELRAD, A. A., and STEEVES, R. A. (1964). Assay for Friend leukemia virus: rapid quantitative method based on enumeration of macroscopic spleen foci in mice. virdogy 24, 513-518. BASSIN, R. H., TUT~L.E, N., and FISCHINGER,P. J. (1971). Rapid cell culture assay technique for murine leukemia viruses. Nature (London) 229, 564-566. BASSIN, R. H., RUSCETTI, S., ALI, I., HAAPALA, D. K., and REIN, A. (1982). Normal DBA/Z mouse cells synthesize a glycoprotein which interferes with MCF virus infection. virology 123, 139-151. CHESEBRO,B., and WEHRLY, K. (1976). Studies on the role of the host immune response in recovery from Friend virus leukemia. I. Anti-viral and anti-leukemia cell antibodies. J. Exp. Me& 143,73-84. CHESEBRO, B., WEHRLY, K., CLOYD, M., BRA, W., PORTIS, J., COLLINS, J., and NISHIO, J. (1981). Characterization of mouse monoclonal antibodies specific for Friend murine leukemia virus-induced erythroleukemia cells: Friend-specific and FMR-specific antigens. fir112, 131-144. CHESEBRO.B., BRIM, W., EVANS, L., WEHRLY, K., NISHIO, J., and CLOYD, M. (1983). Characterization of monoclonal antibodies reactive with murine leukemia viruses: Use in analysis of strains of Friend MCF and Friend ecotropic murine leukemia virus. virdosy

127,

134-138.

CLOYD,M. W., HARTLEY, J. W., and Rowa, W. P. (1986). Lymphomagenicity of recombinant mink cell focusinducing murine leukemia viruses. J. Exp. Med 151, 542-552.

ET AL. COLLINS,J. K., and CHESEBRO,B. (1980). Spontaneous cessation of Friend murine leukemia virus production by leukemia cell line Y57: Overgrowth by nonproducer cells. J. Nut. Cancer In& 64, 11531159. DAYESHEK, W. (1969). The DiGuglielmo syndrome revisited. Blood 34, 56’7-572. DAWSON, P. J., ROSE, W. M., and FIELDSTEEL, A. H. (1966). Lymphatic leukemia in rats and mice inoculated with Friend virus. Brit J. Cmcer 20,114121. DAWSON, P. J., TACKE, R. B., and FIELDSTEEL, A. H. (1968). Relationship between Friend virus and an associated lymphatic leukemia virus. Brit. J. Cancer 22, 569-576. DI GUGLIELMO, G. (1917). Un case di eritroleucemia. Folia

Med

3,386-396.

FICHELSON, S., WENDLING, F., HEARD, J. M., TAMBOURIN,P., and VARET, B. (1982). Establishment of in vitro myelomonocytic permanent cell lines from in viva virus-induced chloroleukemias in mice. J. Cell. Biochm

6,240.

FIORITONI, G., BERTOLINI, L., TORLONTANO. G., and REVOLTELLA, R. (1986). Cytochemical characteristics of leukopoietic differentiation in murine erythroleukemic (Friend) cells. Cancer Res. 40, 866-872. FISCHINCER, P. J., IHLE, J. N., DENORHONHA, F., and BOLOGNESI, D. P. (1977). Oncogenic and immunogenic potential of cloned HIX viruses in mice and cats. Med Micra Immund 164, 119-129. FISCHINGER,P. J., FRANKEL, A. E., ELDER, J. H., LERNER, R. A., IHLE, J. N., and BOLOGNESI,D. P. (1978). Biological, immunological, and biochemical evidence that HIX virus is a recombinant between Moloney leukemia virus and a murine xenotropic C-type virus. virology 90, 241-254. FRIEND, C. (1957). Cell-free transmission in adult Swiss mice of a disease having the character of a leukemia. .I. Exp. Mea! 105,307-318. FRIEND, C., SCHER, W., HOLLAND, J. G., and SATO, T. (1971). Hemoglobin synthesis in murine virus-induced leukemic cells in vitro: Stimulation of erythroid differentiation by dimethyl sulfoxide. Proc. Nat. Acad Sci USA 68,378-382. HARTLEY, J. W., and ROWE, W. P. (1975). Clonal cell lines from a feral mouse embryo which lack host range restrictions for murine leukemia viruses. V& robgg 65, 128-134. HARTLEY, J. W., WOLFORD,N. K., OLD, L. J., and ROWE, W. P. (1977). A new class of leukemia virus associated with development of spontaneous lymphoma. Proc Nat. Aed Sci USA 74, 789-792. HAYWARD, W. S., NEEL, B. G., and ASTIN, S. M. (1981). Activation of a cellular oru: gene by promotor insertion in ALV-induced lymphoid leukosis. Nature (London) 290, 475-480. HEARD, J. M., FICHELSON, S., GISSELBRECHT, S., and VARET, B. (1981). Comparative effects of polycy-

FRIEND

MCF VIRUS EXPRESSION

themia-inducing Friend leukemia virus and Mohey leukemia virus on hematopoietic progenitors in murine long-term bone marrow cultures. CUnxer Res. 41,3260-3265.

ISHIMOTO,A., HARTLEY, J. W., and ROWE, W. P. (1977). Detection and quantitation of phenotypically mixed viruses: mixing of ecotropic and xenotropic murine leukemic viruses. Virology 81, 263-269. ISHIMOTO, A., ADACHI, A., SAKAI, K., YORIFUGI, T., and TSURUTA, S. (1981). Rapid emergence of mink cell focus-forming (MCF) virus in various mice infected with NB-tropic Friend virus. Virdogy 113, 644-655. KASS, L. (1977). E&erase activity in erythroleukemia. Amer.

J. Clin

PathoL

67,368-370.

LILLIE, R. D, (1954). “Histopathologic Technique,” 391-392. McGraw-Hill (Blakiston), Philadelphia. LILLIE, R. D., and BURTNER, H. J. (1953). Stable danophilia of human neutrophil leukocytes in lation to peroxidase and oxidase. .I Histochem tochem.

pp. sureCy-

1, 8-26.

LINEMEYER, D. L., RUSCE~I, S. K., MENKE, J. G., and SCOLNICK,E. M. (1980). Recovery of biologically active spleen focus-forming virus from molecularly cloned spleen focus-forming virus-pBR.322 circular DNA by cotransfection with infectious type C retroviral DNA. J. ViroL 35, 710-721. MACDONALD, M. E., MAK, T. W., and BERNSTEIN, A. (1980). Erythroleukemia induction by replicationcompetent type C viruses cloned from the anemiaand polycthemia-inducing isolates of Friend leukemia virus. J. Exp. Med 151, 1493-1503. MARIE, J. P., IZAGUIRRE, C. A., CIVIN, C. I., MIRRO, J., and MCC~LL~~H, E. A. (1981). The presence within single K-562 cells of erythropoietic and granulopoietie differentiation markers. Blood 58, 708-711. MATHIEU-MAHIJL, D., HEARD, J. M., FICHELSON, S., GISSELBRECHT,S., S~LA, B., and LARSEN, C. J. (1982). Viral expression in two myelomonocytic cell lines obtained from long-term bone marrow culture infected with the Friend polycthemia-inducing virus (FV-P). Virology 119, 59-67. O’DONNELL, P. V., STOCKERT, E., OBATA, Y., and OLD, L. J. (1981). Leukemogenic properties of AKR dualtropic (MCF) viruses: Amplification of murine ieukemia virus-related antigens on thymocytes and acceleration of leukemia development in AKR mice. Virology

112, 548-563.

233

OLIFF, A. I., HAGER, G. L., CHANGE, E. H., SCOLNICK, E. M., CHAN, H. W., and LOWY, D. R. (1989). Transfection of molecularly cloned Friend murine leukemia virus DNA yields a highly leukemogenic helper-independent type C virus. J. vird 33. 475486. OLIFF, A., RUSCET~I, S., DOUGLASS, E. C., and SCOLNICK, E. (1981). Isolation of transplantable erythroleukemia cells from mice infected with helperindependent Friend murine leukemia virus. Blood 58.244454. RUSCE~I, S., DAVIS, L., FE~LD, J., and OLIFF, A. (1981). Friend murine leukemia virus-induced leukemia is associated with the formation of mink cell focusinducing viruses and is blocked in mice expressing endogenous mink cell focus-inducing xenotropic viral envelope genes. J. Exp. Med 154,907-920. SHIBUYA, T., and MAK, T. (1982). Host control of SUBceptibility to erythroleukemia and to the types of leukemia induced by Friend murine leukemia virus: initial and late stages. CeU 31,483-493. STEEVES, R. A., ECKNER, R. J., BENNETT, M., MIRAND, E. A., and TRUDEL, P. J. (1971). Isolation and characterization of a lymphatic leukemia virus in the Friend virus complex. J. Nat. Cancer Inst. 46,12991217. TROXLER, D. H., and SCOLNICK, E. M. (1978). Rapid leukemia induced by cloned strain of replicating murine type C virus: Association with induction of xenotropic-related RNA sequences contained in spleen focus-forming virus. virdosy 85, 17-27. TROXLER, D. H., YUAN, E., LINEMEYER, D., RUSCETTI, S., and SCOLNICK,E. M. (1978). Helper-independent mink cell focus-inducing strains of Friend murine type-C virus: Potential relationship to the origin of replication-defective spleen focus-forming virus. J. Exp.

Med

148,639-653.

VAN GRIENSVEN, L. J. L. D., and VOGT, M. (1980). Rauscher “mink cell focus-inducing” (MCF) virus causes erythroleukemia in mice: Its isolation and properties. Virology 101, 376-388. WENDLING, F., HEARD, J. -M., and TAMBOURIN, P. (1983). Phenotypic heterogeneity of leukemias induced by ecotropic replication-competent virus isolated from Friend virus complex. Blood, in press. YAM, L. T., LI, C. Y., and CROSBY,W. H. (1971). Cytochemical identification of monocytes and granulocytes. Amer. J. C!Zin, Path& 55.283-290.