Genetic Regulation of Long-Term Nonprogression in E-55+ Murine Leukemia Virus Infection in Mice

Abstract

Certain inbred mouse strains display progression to lymphoma development after infection with E-55+ murine leukemia virus (E-55+ MuLV), while others demonstrate long-term nonprogression. This difference in disease progression occurs despite the fact that E-55+ MuLV causes persistent infection in both immunocompetent BALB/c–H-2^k (BALB.K) progressor (P) and C57BL/10–H-2^k (B10.BR) long-term nonprogressor (LTNP) mice. In contrast to immunocompetent mice, immunosuppressed mice from both P and LTNP strains develop lymphomas about 2 months after infection, indicating that the LTNP phenotype is determined by the immune response of the infected mouse. In this study, we used bone marrow chimeras to demonstrate that the LTNP phenotype is associated with the genotype of donor bone marrow and not the recipient microenvironment. In addition, we have mapped a genetic locus that may be responsible for the LTNP trait. Microsatellite-based linkage analysis demonstrated that a non-major histocompatibility complex gene on chromosome 15 regulates long-term survival and is located in the same region as the Rfv3 gene. Rfv3 is involved in recovery from Friend virus-induced leukemia and has been demonstrated to regulate neutralizing virus antibody titers. In our studies, however, both P and LTNP strains produce similar titers of neutralizing and cytotoxic anti-E-55+ MuLV. Therefore, while it is possible that Rfv3 influences the course of E-55+ MuLV infection, it is more likely that the LTNP phenotype in E-55+ MuLV-infected mice is regulated by a different, closely linked gene.

E-55+ murine leukemia virus (E-55+ MuLV) is a chronic ecotropic murine leukemia virus that causes the development of thymic lymphoma about 5 months after infection of immunocompetent, adult progressor BALB/c H-2^k (BALB.K) mice (1, 31). This virus has a high degree of sequence homology with F-MuLV, the helper component of Friend murine leukemia virus (FV), an acute transforming retrovirus (32). In contrast to the high incidence of lymphomagenesis in E-55+ MuLV-infected BALB.K progressor mice, infection of immunocompetent adult long-term nonprogressor (LTNP) C57BL/10– H-2^k (B10.BR) mice fails to cause thymic lymphoma despite the fact that these mice develop a persistent infection in the same manner as progressor mice (1). Despite the difference in progression to disease between the infected BALB.K progressor and B10.BR nonprogressor mice, mice from both strains develop an effective immune response during the acute phase of infection that results in a dramatic decrease in the number of virus-infected cells (1, 2).

In contrast to immunocompetent B10.BR mice, immunosuppressed B10.BR mice develop E-55+ MuLV-induced lymphomas (1), indicating that the ability to generate an effective antivirus immune response plays an important role in determining the LTNP phenotype. Previous studies with other retroviruses have also determined that the genetic regulation of the antivirus immune response can determine whether or not animals are resistant to retrovirus-induced pathogenesis (10, 17). For example, FV is an acute transforming virus that is composed of a replication-defective spleen focus-forming virus and a replication-competent Friend murine leukemia helper virus (28, 29). FV induces rapid polyclonal proliferation of immature erythroblasts, leading to acute splenomegaly in adult mice within a few days after infection (12) as the result of a virus component, gp55, encoded by the defective spleen focus-forming virus that binds to the erythropoietin receptor (15, 21, 25). Resistance to FV is known to be regulated by alleles of two H-2-linked genes, Rfv1 and Rfv2 (6), and a third gene, Rfv3, not linked to H-2, which regulate the antivirus immune response (5, 7, 8). In contrast, since progressor BALB.K and LTNP B10.BR share the same H-2^k haplotype, the gene(s) regulating the LTNP phenotype with respect to E-55+ MuLV-induced pathogenesis does not appear to be linked to the major histocompatibility complex (MHC).

Most studies to date have concentrated on the genetic regulation of immune responses to acute transforming retroviruses, like FV (10, 17), rather than chronic retroviruses, such as E-55+ MuLV, which cause malignant transformation in susceptible mice after a long latent period characterized by persistent infection. Thus, E-55+ MuLV can be utilized to map and select candidate loci that regulate phenotypic differences between mice from strains which progress to develop chronic virus-induced disease and those which are LTNPs. In this present study, phenotypic ratios in backcross analysis suggest that perhaps two non-MHC genes are responsible for the LTNP phenotype in E-55+ MuLV-infected mice.

The location of genes that determine the LTNP phenotype was investigated by microsatellite-based mapping with a large number of (B10.BR × BALB.K)F₁ × BALB.K backcross mice. Microsatellite markers were used to scan the genome to determine linkage with chromosomal regions with particular attention to regions close to immunologically relevant genes (e.g., interleukin 4 [IL-4], IL-10, and FasL, etc.). One region, on chromosome 15, is significantly linked to the LTNP trait (P = 0.0001). Studies using radiation bone marrow chimeras indicated that these genes affect the development of disease as the result of their expression in bone marrow-derived cells rather than in the stromal elements of the microenvironment of the mouse.

MATERIALS AND METHODS

Mice.

Adult C57BL/10–H-2^k (B10.BR) mice were purchased from the Jackson Laboratory (Bar Harbor, Maine). BALB/c–H-2^k (BALB.K) and backcross mice were bred in the Research Animal Facility at MCP Hahnemann University. BALB.K mice are congenic partners with BALB/c mice which express the H-2^d haplotype. B10.BR mice (H-2^k) are congenic partners with C57BL/10 (B10) mice which express the H-2^b haplotype.

Virus.

E-55+ MuLV was isolated from a leukemic spleen harvested from a BALB.K mouse that was injected with cell-free culture supernatant from a T-cell leukemia line (24). The virus used in these studies was passaged in vivo by intraperitoneal injections of immunosuppressed BALB.K. For the present experiments, each mouse was injected intraperitoneally with 2 × 10⁵ focus-forming units (FFU) of E-55+ MuLV as determined in the fluorescent focus assay (FFA) described below.

Antibodies.

Hybridoma cells producing the monoclonal antibody (Ab) m34 (specific for p15-gag) (9) used for the FFA were a gift from Bruce Chesebro. Polyclonal fluorescein isothiocyanate-conjugated anti-mouse immunoglobulin Ab used for the FFA was obtained from Southern Biotechnology Associates, Inc. (Birmingham, Ala.). Monoclonal Ab 145.2C11, used for in vitro depletion, is specific for CD3 epsilon and was obtained from Boehringer Mannheim (Indianapolis, Ind.).

Cells.

Mus dunni fibroblasts were maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum (FCS), 2 mM l-glutamine, 100 U of penicillin per ml, and 0.1 mg of streptomycin per ml.

FFA.

The FFA was performed as previously described (27). On day 1, M. dunni fibroblasts were plated at a concentration of 6,000 cells/well in a 24-well plate (Dulbecco’s modified Eagle’s medium supplemented with 10% FCS, 2 mM l-glutamine, 100 U of penicillin per ml, 0.1 mg of streptomycin per ml, and 10 μg of Polybrene per ml). On day 2, the M. dunni cells were infected with 100 μl of 10-fold serial dilutions of 10% spleen homogenate per well. Serial dilutions ranged from 10⁻¹ to 10⁻⁵. One hour after infection, the supernatant was aspirated, the wells were washed with phosphate-buffered saline (PBS), and fresh medium was added. On day 5, when cells had grown to confluency, the medium was aspirated, the wells were washed with PBS plus 2% FCS and incubated with Ab m34 for 1 h at 4°C, and the wells were washed again with PBS plus 2% FCS and incubated with fluorescein isothiocyanate-conjugated anti-mouse immunoglobulin Ab for 1 h at 4°C. Finally, the cells were washed three times with PBS plus 2% FCS, and fluorescent foci were counted on an inverted fluorescence microscope. Virus titers were expressed as FFU per milliliter of spleen homogenate.

Bone marrow cell transfers.

Bone marrow cells were aspirated from the femurs of donor mice with RPMI (supplemented with 100 U of penicillin per ml and 0.1 mg of streptomycin per ml). Cells were washed twice and resuspended at 2 × 10⁷ cells/ml of RPMI supplemented with 5% FCS, 2 mM l-glutamine, 100 U of penicillin per ml, and 0.1 mg of streptomycin per ml. To remove T cells from the bone marrow population, equal volumes of supernatant containing monoclonal Ab 145.2C11 (anti-CD3) and cells were incubated at 4°C for 1 h. Cells were then washed, resuspended in RPMI supplemented as described above and also containing 10% baby rabbit complement (Accurate Chemical, Westbury, N.Y.), and incubated at 37°C for 1 h. Finally, cells were washed twice and resuspended at 4 × 10⁷ cells/ml of RPMI containing 100 U of penicillin per ml and 0.1 mg of streptomycin per ml, and 0.2 ml of the cell suspension was injected intravenously into lethally irradiated (900 rads) recipients. The recipients were irradiated 14 to 24 h before cell transfer. Chimeric mice were reconstituted for 7 weeks before inoculation with E-55+ MuLV.

Microsatellite typing.

Genomic DNA was prepared as previously described (3). SSLP primers were purchased from Research Genetics (Huntsville, Ala.) and were used for PCR (11) in 96-well plates. After PCR, electrophoresis was performed on samples from each well by using a 6% polyacrylamide gel. Gels were lifted onto Whatman paper and exposed to film overnight. The autoradiographs were read by at least three persons.

Linkage analysis.

Each backcross mouse was designated with a C if its phenotype was similar to BALB.K or a B if its phenotype was similar to B10.BR. Backcross mice were phenotyped by using 100 microsatellite markers (Mouse Genome Database [17a]) as described above. The Map Manager computer software program was used for the qualitative linkage analysis (22). The significance of linkage was evaluated as described by Lander and Kruglyak (19): for suggestive linkage, P was ≤0.0034 and LOD was ≥1.9, and for significant linkage, P was ≤0.0001 and LOD was ≥3.3.

Statistics.

All P values (except for the linkage analysis) were calculated by using Student’s t test. Bayesian statistics (26) were used for the qualitative linkage analysis.

RESULTS

Long-term nonprogression in E-55+ MuLV infection is regulated by at least two non-MHC-linked genetic loci.

To determine the number of loci controlling the LTNP phenotype in B10.BR mice, (BALB.K × B10.BR)F₁ mice, which express the dominant LTNP trait (1) (Table 1), were backcrossed to BALB.K mice, which express the recessive progressor trait, to generate cohorts of backcross progeny in which a random assortment of genes occurred. Thirty-nine out of 151 backcross mice progressed to lymphomas (Table 1) between 5 and 12 months after E-55+ MuLV inoculation. The number of mice expressing each phenotype (i.e., progressor or LTNP) resulted in a 1:3 ratio among the backcross progeny, indicating that at least two loci determine the LTNP phenotype.

TABLE 1.

Incidence of progression to E-55+ MuLV-induced lymphoma^a

Mouse strain	Progressors/total (%)
BALB.K	20/20 (100)
B10.BR	0/40 (0)
(B10.BR × BALB.K)F1	0/15 (0)
F1 × BALB.K	39/151 (25.8)
BALB.K (550 rads)	10/10 (100)
B10.BR (550 rads)	10/10 (100)

^aAdult BALB.K, B10.BR, F₁, and backcross (F₁ × BALB.K) progeny were inoculated with 2 × 10⁵ FFU of E-55+ MuLV and monitored for progression to lymphoma. Groups of BALB.K and B10.BR mice were inoculated with E-55+ MuLV after sublethal irradiation (550 rads). 

Consistent with previous findings (1, 31), all (100%) B10.BR and BALB.K mice immunosuppressed by sublethal irradiation progressed to develop lymphoma 2 to 3 months after E-55+ MuLV infection (Table 1). This result indicated that the LTNP trait is determined by the genetic regulation of the ability to generate an effective antivirus immune response.

Genes that determine the LTNP phenotype are expressed in bone marrow-derived cells.

Since the genes that are involved in controlling LTNP appear to regulate the anti-E-55+ MuLV immune response, it became important to determine if these genes are expressed in bone marrow-derived cells that generate the cells that mediate immune responses or in the stromal microenvironment (e.g., thymus), which influences the repertoire of antigens recognized as nonself. We, therefore, investigated whether or not recipient progressor BALB.K mice expressed the LTNP phenotype as the result of transfer of donor B10.BR bone marrow cells. We also examined if the converse was true, i.e., if LTNP recipient B10.BR mice that received bone marrow from progressor BALB.K mice expressed the progressor phenotype.

To perform this study, bone marrow depleted of CD3⁺ cells from donor B10.BR mice (LTNP) was transferred into lethally irradiated (900 rads) BALB.K recipient mice and vice versa. In all chimera studies, reconstitution was complete as determined by microsatellite analysis (data not shown). BALB.K mice that were reconstituted with B10.BR bone marrow cells displayed the LTNP phenotype, similar to the untreated B10.BR mice and B10.BR mice that were reconstituted with CD3-depleted syngeneic bone marrow cells (Table 2). Conversely, recipient B10.BR mice that received BALB.K donor bone marrow depleted of CD3⁺ cells expressed the progressor phenotype in the same manner as normal BALB.K mice (Table 2). Both lethally irradiated BALB.K recipient mice reconstituted with syngeneic BALB.K bone marrow cells and untreated BALB.K control mice progressed to lymphomas after inoculation with E-55+ MuLV. All lymphomas expressed antigens encoded by E-55+ MuLV but not endogenous retroviruses, indicating that lymphomagenesis did not develop as the result of endogenous retrovirus expression (data not shown).

TABLE 2.

Phenotypes of chimeric mice infected with E-55+ MuLV

BM transfersa	Progressors/total	Spleen size (g)	FFU/mlb	Mean ± SD/105
BALB.K→B10.BR	7/7	0.19–0.86	1 × 105–5 × 105	2.7 ± 1.6c
BALB.K→BALB.K	3/3	0.36–0.92	2 × 105–7 × 105	4.7 ± 2.0d
B10.BR→BALB.K	0/4	0.07–0.12	0	0
B10.BR→B10.BR	0/4	0.05–0.11	0	0
None (BALB.K)	5/5	0.43–1.0	2 × 105–6 × 105	3.2 ± 1.2
None (B10.BR)	0/5	0.09–0.11	0	0

^aBone marrow (BM) cells depleted of CD3⁺ cells were transferred into lethally irradiated recipient mice as described in Materials and Methods. Recipient mice were inoculated with E-55+ MuLV and observed for progression to lymphomas for 11 months. 

^bFFU of E-55+ MuLV per milliliter of spleen homogenate determined by FFA. 

^cP value of 0.0006, compared to B10.BR, and P value of 0.19, compared to BALB.K→B10.BR. 

^dP value of 0.05, compared to BALB.K. 

In conclusion, radiation bone marrow chimeras made reciprocally between LTNP (B10.BR) and progressor (BALB.K) strains consistently expressed the phenotype of the bone marrow donor, suggesting that the genes that regulate the LTNP phenotype are expressed in bone marrow cells and not in the microenvironment of the mouse.

Linkage analysis by using the (B10.BR × BALB.K) × BALB.K backcross mice.

Twenty progressor and 20 LTNP (B10.BR × BALB.K) × BALB.K backcross mice (all H-2^k) were tested for alleles at microsatellite markers to determine linkage with the genes that determined the LTNP phenotype. A total of 100 microsatellite markers were used to scan all 19 autosomes to an average resolution of 18 centimorgans (cM) (range of intermarker distance, 0.5 to 49 cM). Additional microsatellite markers were tested on the backcross mice either because they were closely linked to genes with immune function (e.g., Il4 and Ifng) or were chosen randomly to scan the remainder of the genome. In addition, a region of chromosome 15 was tested because of a previous study that mapped Rfv3 to this position (16). Rfv3 appeared a likely candidate gene since this gene has been determined to regulate the antivirus immune response against FV (7).

Qualitative analysis (Table 3) revealed significant linkage to chromosome 15 with a high χ² value of 12.1 (P = 0.0005). This result indicated that Rfv3 or a closely linked gene plays an important role in determining the LTNP phenotype. Since no other significant linkages were revealed in this analysis, the location of the putative second gene, whose existence was suggested by the ratio of progressor/LTNP mice in the backcross analysis, remains unknown.

TABLE 3.

Linkage of lymphoma incidence in backcross mice^a

Locus	cM	LOD	Lymphoma-positive mice (progressors)		Normal mice (LTNP)		χ2	P
			c/c	b/c	c/c	b/c
D15Mit187	44	1.8	15	5	6	14	8.1	0.004
D15Mit28	44	1.8	15	5	6	14	8.1	0.004
D15Mit68	46	1.8	15	5	6	14	8.1	0.004
D15Mit32	48	2.3	15	5	5	15	9.1	0.003
D15Mit159	52	2.3	15	5	5	15	9.1	0.003
D15Mit77	56	2.8	15	5	4	16	12.1	0.0005b
D15Mit79	66	1.9	14	5	5	14	8.1	0.004

^aNumbers of mice with or without evidence of lymphoma are given, grouped by their genotypes at the indicated locus. Mice were genotyped as c/c or b/c depending on whether the alleles detected at the indicated locus (e.g., D15Mit187) were derived from BALB.K (c) or B10.BR (b). χ² and P values were derived from a 2-by-2 contingency test. 

^bSuggestive linkage of marker to lymphoma incidence; not significant. 

DISCUSSION

Previous studies have shown that BALB/c–H-2^k (BALB.K) mice progress to develop T-cell lymphomas at 4 to 5 months after infection with E-55+ MuLV (31), whereas LTNP C57BL/10–H-2^k (B10.BR) mice fail to develop disease despite the fact that they are persistently infected (1). This difference between these strains of mice occurs despite the fact that both strains express the H-2^k haplotype, indicating that the LTNP phenotype is regulated by non-MHC genetic loci. The purpose of the present study was to map the genes that control differences in resistance to E-55+ MuLV-induced disease between mice from strains that either progress to develop virus-induced disease or exhibit long-term nonprogression and fail to develop disease.

In this study, phenotypic analysis of (B10.BR × BALB.K)F₁ × BALB.K backcross mice infected with E-55+ MuLV indicated that at least two non-MHC genetic loci influence the LTNP phenotype (Table 1). These loci are expressed in bone marrow-derived cells and not in the microenvironment (e.g., thymus) of the mouse, as demonstrated by studies using radiation bone marrow chimeras which consistently expressed the phenotype of the bone marrow donor (Table 2).

To map the genes that control the LTNP phenotype in E-55+ MuLV infection, an analysis of (B10.BR × BALB.K)F₁ × BALB.K backcross mice was performed. This experiment identified a region of chromosome 15 that is significantly associated (P = 0.0001) with the LTNP trait (Table 3), indicating that this region harbors a gene that controls resistance to E-55+ MuLV-induced disease. This gene has been named Rev1 (resistance to E-55+ MuLV 1). No other significant linkages were found, although another region on chromosome 8 showed suggestive linkage when latency (defined as the period from virus infection to the development of disease) was used as a quantitative trait (data not shown).

Previous studies (16) have determined that this region of chromosome 15 is the location of another gene, Rfv3, that regulates the immune response-mediated resistance to another retrovirus, FV (7). Based on the fine mapping performed in our studies, it is possible that Rfv3 and Rev1 are identical genes. Rfv3 has been found to regulate the production of anti-FV neutralizing antibodies; BALB mice express the Rfv3^s allele and fail to produce FV neutralizing antibodies, whereas C57BL mice express the Rfv3^r allele and produce high titers (8). However, both BALB and C57BL mice produce the same titers of antibody against E-55+ MuLV (1). Thus, if Rfv3 and Rev1 are identical, it appears that these genes may exert different effects on the antivirus immune response, depending on the virus used for inoculation. This difference in phenotypic expression occurs despite the fact that the helper component of FV has a high degree of sequence homology with E-55+ MuLV (32). This phenotypic difference may be related to the fact that FV is an acute transforming virus that causes the rapid proliferation of erythroid precursor cells, whereas E-55+ MuLV is a chronic transforming virus that causes a slowly emerging T-cell lymphoma. However, based on the difference in phenotypes determined by Rfv3 and Rev1 in BALB and C57BL mice, it is likely that these genes are different but very closely linked.

A number of genes have already been assigned to the mouse chromosome 15 at the region of Rev1. Four of these genes, Ly6 (20), IL2rb (4), and Il3rb1 and IL3rb2 (14), are associated with functions of the immune system. Ly6 represents a complex of genes, one (or more) of which encodes a lymphocyte antigen of unknown function (18). IL2rb encodes the IL-2 receptor beta chain, a component of the IL-2 receptor and also a component of the IL-15 receptor (13, 30). Binding of IL-2 to the IL-2 receptor results in activation of T cells (23). IL3rb1 encodes the beta chain of the receptor for IL-3, granulocyte-macrophage colony-stimulating factor, and IL-5 (14), all of which can affect immune response. While it is tempting to speculate whether any of these genes is Rev1, it is quite possible that Rev1 is unrelated to any of the genes previously mapped on this region of chromosome 15 and that its identification will await the assignment of additional genes or cloning.

In summary, we suggest that Rev1, mapped within a 20-cM region of chromosome 15, is a non-H-2 gene that influences the cellular immune response responsible for the difference in progression versus long-term nonprogression in BALB.K and B10.BR mice during E-55+ MuLV infection. Elucidation of the mechanism that leads to progression or long-term nonprogression in these strains is of major importance, since it may provide information that can be applied to the mechanisms that control progression versus long-term nonprogression during retrovirus infections in humans (human T-cell leukemia virus and human immunodeficiency virus).