key: cord-0010817-qhc5i2ma authors: Gardner, Murray B.; Klement, Vaclav; Rongey, Robert R.; McConahey, Patricia; Estes, John D.; Huebner, Robert J. title: Type C Virus Expression in Lymphoma-Paralysis-Prone Wild Mice date: 1976-09-01 journal: J Natl Cancer Inst DOI: 10.1093/jnci/57.3.585 sha: 6765c3b70b6411e31565a70c95df859cf2c5ecbe doc_id: 10817 cord_uid: qhc5i2ma Wild mice trapped near Lake Casitas (LC) in southern California showed a high prevalence of infectious type C virus in the liver, spleen, and thymus within the first few weeks of life. By young adulthood about 80% of LC mice (including their genital tissues) were infected. Virus isolates from these mice cause lymphoma and lower limb paralysis under both natural and experimental conditions. Mice destined to develop paralysis showed higher levels of serum gs antigen early in life, whereas mice destined to develop lymphoma or remain free of these diseases could not be distinguished by this test. The individual variation in virus expression suggested that differences in virus type or in the immune or other host defense mechanisms greatly influenced susceptibility or resistance to indigenous type C virus-caused disease in LC wild mice. Wild mice (Mus musculus) from a trapping area In southern California near Lake Casitas (LC) show, in comparison to most other wild mice, a high level of indigenous type C virus gs antigen expression; with increased age, they are prone to spontaneous lymphoma, other cancers, and a neurogenic hind leg paralysis (1, 2) . We have presented evidence that the indigenous type C virus is the essential determinant of both lymphomc ous and paralytic diseases (2, 3) . In this phper, we indicate that LC mice generally have infectious virus early in life, and that the serum gs antigen level is elevated especially in individual mice destined to develop paralytic disease. These findings help to explain the pathogenesis. of type C virus-related diseases in wild mice. Live wild mice from the LC trapping site (1) were brought to the laboratory, weighed, and bled by retroorbital puncture. They were shown to be completely free of antibodies by hemagglutination inhibition and CF tests to a battery of 16 murine virus antigens. 8 Newly trapped nonpregnant LC mice were either promptly killed for collection of tissues or individually housed in Masonjars and allowed to live out their life-spans under laboratory observation (4) . Some LC mice, pregnant when trapped, were promptly killed to obtain fetal and maternal tissues; others delivered their progeny, which were individually housed in the laboratory and killed at various predetermined postnatal intervals from 1 to 42 days. Necropsies of all sick, paralyzed, or suspect tumor-bearing mice under long-term laboratory observation were made for microscopic and virologic study. Established procedures were used for histopathology, EM, and CF tests (5) . The routine EM screening procedure consisted of a thorough scan at 15,000X of three grids from each of two blocks of tissue. To be positive, more than one typical budding or extracellular type C particle had to be seen. For CF testing, individual whole embryos and separately pooled tissues (viscera, carcass, VOL. 57, NO.3, SEPTEMBER 1976 585 liver, spleen, thymus, muscle) from 3-4 fetal and postnatallittermates up to 21 days of age were collected; at 42 days of age tissues from individual mice were mostly tested by CF. Some fetal and newborn mice were rinsed with ether to prevent possible virus contamination from the maternal genital tract. Primary tissues and tissue culture cell packs were sonically disrupted, and 10% (wt/ vol) aqueous extracts were tested by CF at 1:2 and 1:4 antigen dilutions. In mice under 15 g, individual spleen and thymus extracts were, however, approximately 5% (wt/vol) strength because of the small size of these organs. MuSV rat serum pools #30, 322, 3490, and 4918 were used during this study. By comparison with guinea pig antisera to purified M uL V gs (p30) antigen (6), they were sensitive and specific at the dilution used (l :20) for murine gs (p30) antigen. For virus isolation, 10% tissue extracts frozen at -70 0 C for 1-6 months were thawed; 0.1 ml was assayed undiluted and, in most instances, also at a 10-1 dilution for induction of gs antigen detected by CF on secondary NIH Swiss embryo cells after 21 days (COMUL) (7) . The same MuSV rat sera were used. Some of the cultures that were COMUL-negative at 21 days were passaged and retested after 42 days. Since little difference was noted in gs antigen or infectious virus detection whether undiluted or lO-l -diluted extracts were used and whether or not fetuses and newborns were rinsed with ether before dissection, these data are combined in the "Results" section. Fibroblast cultures initiated from minced whole embryos with or without initial ether rinse were fed Eagle's minimum essential medium with 10% heated fetal calf serum, 2 mM glutamine, 100 U penicillin/ml, and 100 JLg streptomycin/m!. Cultures were grown in 75-cm 2 Falcon flasks incubated at 37° C in 5% CO 2 and subcultured once weekly with 0.1 % trypsin in 0.02% Versene in Hanks' balanced salt solution. Supernatants (0.2 ml) were harvested in primary or secondary subpassage and assayed undiluted for infectious virus by COMUL on secondary NIH Swiss embryo cells. Sera were analyzed for MuL V gs (p30) antigen in Dr. Frank Dixon's laboratory (Scripps Clinic and Research Foundation) by a double antibody RIA that is a modification of the methods described by Oroszlan et a!. (8) and Parks and Scolnick (9) . The Scripps' leukemia virus p30 was labeled with 125 1 (10) ; guinea pig antisera to M uL V p30 from Dr. R. Gilden (Flow Laboratories, Rockville, Md.) was used, and saturated ammonium sulfate was the precipitating reagent for antigen-antibody complexes. The specificity of the RIA for MuLV p30 antigen was demonstrated by the fact that this antigen was completely precipitable by antibody prepared against MuL V p30 but not by antibody against fetal calf serum or MuLV gp69/71. Specificity of the p30 RIA was further supported by the fact that lysates of murine celt lines shown to be non producers of type C virus by other methods were negative by RIA for p30 antigen. Virus expression as measured by liver and spleen gs antigen assay by CF and virus isolation from frozen spleen extracts in newly trapped, apparently healthy, LC mice according to weight is shown in table 1. An equally high prevalence (70-80%) and titer (2:: 1:4) of gs antigen were detected in the extracts of liver in each age group. Due to the small size of the spleens «60 mg) in the mice having lower weights, the extracts of this organ were only 5%, which is probably reflected in the lower amount of antigen detected in this group. In the mice having heavier weights, the prevalence and amount of gs antigen closely approximated that in the corresponding liver. In about 10% of skeletal muscle extracts, gs antigen was detected by CF. Virus was recovered in vitro from nearly 80% of spleen extracts in each weight group. Tissues from laboratory-born LC mice of known age that were the offspring of mothers caught in the wild were tested for gs antigen by CF, for type C particles by EM, and for infectious virus by COMUL (table 2) . Infectious virus and gs antigen were almost completely undetected in fetal tissues. By EM, however, extracellular and budding type C particles were seen in association with hematopoietic cells in 4 of 24 fetal livers. In newborn and 3-day-old mice, although gs antigen and type C particles were almost entirely undetected, virus could be isolated from 3 of 10 pooled liver and spleen extracts. detectable by CF in liver and spleen at low titer (1 :2) was at 3 days of age. By 7-10 days of age, the frequency of low-titered gs antigen was 23% in liver and spleen and 48% in thymus. At 2-3 weeks of age, gs antigen was found by CF in higher titer (2:: 1 :4) in liver, spleen, and thymus in 60-67% of LC mice and at 42 days in 100% of individual liver and pooled liver, spleen, and thymus extracts. In LC mice 6 weeks of age and older, the titer of gs antigen in liver and spleen often reached end points of 1:8-1:32. From 14 to 42 days of age, virus was recovered from 34 of 52 (65%) of the total assays done on liver, spleen, thymus, and muscle extracts; during this period type C particles were found by EM in 4 of 9 spleens and 6 of 10 thymuses. After 14 days of age a strong positive correlation was observed between the three assays when done on the same tissues. However, virus was sometimes recovered from skeletal muscle despite the absence of gs antigen detectable by CF in the corresponding tissue extract. Liver and spleen pools from each of 5 littermates from 8 newly trapped pregnant LC wild mice were tested by CF at 2 weeks of age for gs antigen and infectious virus (table 3) . A considerable litter-to-litter variation was noted in the prevalence of detectable gs antigen and infectious virus in liver and spleen. In the different litters, antigen was found in 0,1,2,3,4, or all 5littermates. Virus was isolated from the corresponding gs-positive tissues and occasionally from gs-negative tissues. In at least one litter (#4), virus expression appeared significantly less than in the other litters. In several litters (# 1, #3 and #5), virus expression seemed exceptionally high in that CF assays revealed that vir- , Each assay represents tissues pooled from 3-6 littermates, each littermate pool from a different mother. The single newborn yielding infectious virus was prerinsed with ether. ! Tissues were pooled from 5 mice in each of 8 litters from different mothers. D Individual tissues were from nonlittermates and were different mice than those used for tissue pools. h Pooled tissues were from nonlittermates. tually each mouse had high-titered (2:: 1 :4) gs antigen and infectious virus in liver and spleen. Genital tissues from random samples of newly trapped healthy LC mice (including pregnant mice whose fetal tissues were included in table 2) were exam-VOL. 57, NO.3, SEPTEMBER 1976 ined for gs antigen by CF, for type C particles by EM, and for infectious virus by COMUL (table 4). In 36 of 52 extracts (70%) of ovary and uterus, and in 5 of 5 (100%) epididymis extracts, gs antigen was detected; it was detected in only 1 of 26 placenta extracts. Numerous type C particles, including budding and extracellular virions, were seen in ovary and uterus specimens but not in placenta or testis. Virus was isolated from 23 of 38 (61 %) extracts of ovary, uterus, and placenta, from 12 of 12 (100%) extracts of amniotic fluid, and from 9 of 10 (90%) extracts of testis and epididymis. Virus was recovered from 10 placenta and testis extracts that were negative by CF or EM. Fibroblast cultures established from 18 whole embryos (12) (13) (14) (15) (16) (17) (18) (19) days estimated gestation) from 12 newly trapped LC mice were examined in primary or secondary subpassage for infectious virus by COMUL on secondary NIH Swiss embryo cells. In primary culture, virus was isolated from 1 of 4 embryos rinsed in ether and 3 of 6 embryos not rinsed in ether; in secondary subculture of different embryos, virus was isolated from 1 of 2 em- bryos rinsed in ether and 4 of 6 embryos not rinsed in ether. Serum p30 antigen titers (ng/ml) were determined by RIA on the same individual 45 LC mice at trapping and 1-2 years later at death (text- fig. 1 ). The mice weighed 10-20 g at trapping and were assumed to be young adults at that time. Of these mice, 16 ultimately developed paralysis, 15 developed lymphoma, and 14 died from other causes. The average observation time in the laboratory was 10 months for the paralyzed mice, 15 months for the lymphomatous mice, and 24 months for mice remaining free of these diseases. Considerable individual variation was noted in p30 levels (text- fig. 1 ). However, mice destined to develop paralysis generally had a higher p30 titer at trapping than did those later developing lymphoma or remaining free of paralysis and lymphoma. At trapping, 11 of 16 (69%) mice eventually developing paralysis had serum p30 titers greater than 100 ng/ml, and only 4 of 15 (27%) developing lymphoma and 2 of 14 (14%) remaining free of these diseases had these serum p30 titers. When the paralysis group was compared with the combined lymphoma and "normal" groups by the Mann-Whitney U test (11), this difference was statistically significant (P=0.004). Mice dying with lymphoma showed a slightly, but not statistically significant, higher geometric mean titer of serum p30 at trapping than did mice that did not develop lymphoma. At death, markedly elevated serum p30 titers (>200 ng/ml) were found in almost all mice regardless of diagnosis, but mice dying with lymphoma or paralysis had slightly higher (not statistically significant) geometric mean titers (305 ng/ml) than did those dying from other causes (275 ng/ml). These findings confirm and extend our earlier observations that a markedly elevated indigenous type C virus activity is present in LC wild mice. In addition to a high prevalence and amount of spleen gs antigen and type C J NATL CANCER INST particles in adult LC mice (l), we now show that gs antigen and infectious virus are first detectable in liver and spleen in the perinatal period and can be found thereafter in a more widespread tissue distribution, including the genital tissues. The positive correlation between detection of gs antigen at higher titer (~I:4), type C particles, and isola table virus in about 80% of LC mice after several weeks of age and the rapidly increasing incidence of high-titered gs antigen with age indicates that, by 1-2 months of postnatal life, most LC mice are extensively replicating infectious virus. Viremia probably accounts for the occasional virus isolations from skeletal muscles, placenta, and other gs antigen-negative tissues as well as for the elevated serum p30 levels. In inbred mouse strains, serum p30 levels greater than 100 p,g/ml were invariably associated with recovery of infectious virus from spleen cultures (12) . Furthermore, in recent studies (13, 14) , we demonstrated that about 80% of adult LC mice have free infectious virus and high levels of particulate reverse transcriptase activity in their sera. Virus tally in milk (15) and saliva (16) . In this regard susceptible laboratory mice have been readily infected with the LC virus by foster nursing on LC females (Gardner MB, Klement V: Unpublished observations). Virus isolation from embryo cultures shows that the recovery of infectious virus was about twice as common in embryos that were not prerinsed with ether than in prerinsed ones. This suggests that external virus contamination from the maternal reproductive tract probably does account for some virus isolations, but it is not the only explanation for infectious virus recovered from the cultured embryos. The litter variation in virus expression also indicates a maternal influence on virus expression but does not establish whether this is a positive effect from epigenetic virus transmission or a negative effect from passively transferred antivirus neutralizing antibody. However, the former possibility seems most likely, since LC mice appear to have an ineffective humoral immune reactivity to their indigenous type C virus (13) . The cumulative disease incidence curves in aging LC mice indicate that, after a hiatus of 5-6 months, only 4% of the surviving mice develop either lymphoma or paralysis during each subsequent month of laboratory observation (4) . We showed that little difference exists in prevalence and titer of spleen CF gs antigen at necropsy in tumor-bearing or paralyzed LC mice as compared with age-matched non-tumor-bearing and non-paralyzed LC'mice (4). A similar conclusion was reached in a comparis6p of spleen p30 and gp69/71 antigen titers by RIA in lymphomatous, paralyzed, and normal LC wild mice (August jT: Personal communication). Little difference was also found between lymphomatous, paralyzed, and age-matched LG mice free of these diseases in the serum p30 antigen titers,by RIA at death. However, a comparison of serum p30 -antigen titers at trapping suggests that LC mice with the highest titer viremia early in life are those most at risk to the paralytic disease. Possibly only in those LC mice has vir-us infection been established in the central nervous system at a critically early age. On the other hand, LC mice destined to develop lymphoma could not be distinguished at trapping by the serum p30 titer from LC mice remaining lymphoma-free for a similar time. A more accurate index of virus load early in life (e.g., serum or tail infectious virus titers at 3-6 weeks of age) might better show in LC mice as in AKR X BALB/c crosses (17) a positive correlation in individual mice between virus titer early in life and relative risk to lymphoma. These findings indicate that, although virus is the essential determinant of lymphoma and paralysis, factors other than virus titer alone are obviously involved in the pathogenesis of these diseases. The type C virus population in these wild mice is complex, consisting of a mixture of mouse-tropic virus and a new class of type C virus with a broad (amphotropic) host range (18) (19) (20) . Both classes of virus are N -tropic for mouse cells and thus would be detected by the virus isolation technique used in this study. Preliminary in vivo transmission experiments with limiting dilution-cloned or fluorescent focus-purified viruses suggest that the mouse-tropic strains can induce both rhe paralytic disease and lym-VOL. 57, NO. 3, SEPTEMBER 1976 phoma, whereas the amphotropic strains induce only lymphoma [ (20) ; Parker jC, Hartley jW: Personal communication]. Therefore, the differences in serum p30 levels detected early in life could represent different ratios of these two classes of virus. Although both classes are apparently subject to Fv-1 b gene restriction, it is not yet determined if other genetic loci known to control susceptibility or resistance to viral leukemogenesis in laboratory mice (21) are also operative against each class of virus in wild mice. It will be of interest to relate lymphoma-paralysis resistance or susceptibility in individual LC wild mice with such genetic markers and with various measures of humoral and cellular immunity that may regulate the expression of each type of virus. Unusually high incidence of spontaneous lymphomas in wild house mice A spontaneous lower motor neuron disease apparent!y caused by indigenous type-C RNA virus in wild mice Prevention of lower limb paralysis by neutralization of type-C RNA virus in wild mice The epidemiology and virology of C-type virus-associated hematologic malignancies and related diseases in wild mice Spontaneous tumors of aging wild house mice. Incidence, pathology, and Ctype virus expression Coexistence ofintraspecies and interspecies specific antigenic determinants in the major structural polypeptide of mammalian type C viruses Complement-fixation and tissue culture assays for mouse leukemia viruses A rapid direct radioimmunoassay for type C group specific antigen and antibody Radioimmunoassay of mammalian type C viral proteins: Interspecies antigenic reactivities of the major internal polypeptide A method of trace iodination of proteins for immunologic studies Nonparametric Statistics for the Behavioral Sciences Transmission of murine leukemia virus (Scripps) from parent to progeny mice as determined by p30 antigenemia Inefficient humoral immune response of lymphoma-prone wild mice to persistent leukemia virus infection.] Nat! Cancer Inst Assay for type C virus in mouse sera based on particulate reverse transcriptase activity Types Band C RNA viruses in breast tissue and milk of wild mice Mammary tumor virus particles in the submaxillary gland, seminal vesicle, and nonmammary tumors of wild mice Correlation of early murine leukemia virus titer and H-2 type with spontaneous leukemia in mice of the BALB/c x AKR cross: A genetic analysis Amphotropic host range of naturally occurring wild mouse leukemia viruses Naturally occurring murine leukemia viruses in wild mice: Characterization of a new "amphotropic" class Clonal heterogeneity in wild mouse leukemia viruses: Host range and antigenicity Genetic control of murine viralleukemogenesis ABBREVIATIONS USED: LC=Lake Casitas; CF=complement fixation; EM=electron microscopy, MuSV=murine sarcoma virus; MuLV=murine leukemia virus; COMUL=CF test for MuLV; RIA = radioimmunoassay. 7 We thank Miss Mary Dougherty and Mrs. Teresa Zavala for technical assistance, Mr. john Casagrande for analysis of data, and Miss Ann Dawson for preparation of the manuscript. Animals for this research were maintained in animal care facilities fully accredited by the American Association of Laboratory Animal Care. 8 Lymphocytic choriomeningitis, adenovirus, polyoma, hepatitis virus, reovirus-3, K virus, Theiler's CD7 virus, Kilham rat virus, syncytial virus-5, Sendai virus, encephalomyelitis virus, SV-5, MVM, pneumonia, ectromelia, and rat coronavirus. These tests were done by Dr. j. Parker, Microbiological Associates, Inc., Bethesda, Maryland.