spontaneous malignant granulosa cell tumors in ovaries of...

[CANCER RESEARCH 45, 5575-5581, November 1985]

Spontaneous Malignant Granulosa Cell Tumors in Ovaries of Young SWR Mice1

Wesley G. Beamer, Peter C. Hoppe, and Wesley K. Whitten

The Jackson Laboratory, Bar Harbor, Maine 04609

ABSTRACT

Granulosa cell tumors (GCT) of the ovary appear spontaneously at 4-6 weeks of age in SWR/J and in SWR/Bm inbred

strain mice, with a maximum incidence reached by 10 weeks.Cancer was confirmed by metastasis to abdominal organs andby transplantability of primary tumors to histocompatible hosts.Results of genetic crosses snowed that GCT appear in SWR xSJL F, but not in SJL x SWR F, nor in other Ft females derivedfrom matings of SWR mice with A/HeJ, C57BL/6By, CBA/J, orDBA/2J mice. These findings suggest the maternal transmissionof GCT susceptibility. Recombinant inbred strains SWXJ wereproduced from a progenitor mating of a SWR female to a SJLmale. At F20, females in 3 of 14 SWXJ strains developed GCT,with one strain displaying a 5-fold increase in incidence. Embryotransfer studies with SWXJ-6 and -9 mice suggested that ma

ternal transmission was most likely via the fertilized egg ratherthan through milk or placenta-uterine contact. Analysis of meta-

phase chromosomes indicated that the modal number in tumorsand bone marrow was 40 (2n = 40) with 2 X chromosomes

present. Gross chromosomal aberrations were not detected. Aworking hypothesis proposes that interaction of a unique SWRfactor, perhaps cytoplasmic, with nuclear genomic material common to Swiss mouse stocks results in occurrence of GCT inyoung SWR and SWR-derived mice.

INTRODUCTION

The literature on spontaneous ovarian tumors in mice havebeen reviewed by several authors (1-5). These tumors areusually found in female rodents of advanced age (12-24 months),

and the most prevalent type is derived from the granulosa cell.Granulosa cell tumors are classified as sex cord-stromal tumors

(6), are the most common spontaneous ovarian tumor found inlaboratory and domestic species, and are the most commontype of experimentally induced ovarian tumor in rodents.

In human females, 10-15% of all ovarian tumors are sex cord-

stromal tumors, a substantial proportion of which are of theGCT2 type. As is the case with GCT incidence in laboratory

rodents, the majority of human GCT occur in later life (7, 8).However, in recent years several reports have accumulateddescribing granulosa cell and granulosa-theca cell tumors in very

young human females (9). Such tumors appear to be associatedwith precocious pubertal development, are typically unilateral,and suppress the contralateral ovary. Cytologically these tumorsconsist of diffuse or macrofollicular patterns of granulosa cellsand variable degrees of luteinization. Thecal involvement is variable within and between individual tumors. The etiology of juvenile GCT remains even more obscure than that of adult GCT,

1This work was supported by NIH Grants CA-24145 (W. G. B.) and HD-10381

(P. C. H.).zThe abbreviations used are: GCT, granulosa cell tumors; RI, recombinant

inbred.Received 4/24/85; revised 7/17/85; accepted 8/1/85.

since the low associations found between contraceptive use,parity, and ovarian tumor incidence in adults are not applicable(8, 10). In view of the early age of onset for juvenile GCTdevelopment, it is plausible that genetic factors are of greaterimportance for tumorigenesis than are the commonly cited environmental stimuli.

During the past several years we have been developing amodel of spontaneous GCT of the ovaries in young inbred SWRstrain mice. Data are reported here on the incidence, histopa-

thology, and initial genetic analyses of these neoplasms.

MATERIALS AND METHODS

Sources of Mice. Inbred SWR/J female mice of variable age andreproductive status (weanling: 21-24 days; nonproductive mating: 90-100 days; retired breeders: 180-210 days) were obtained from the

Animal Resource colonies of The Jackson Laboratory. These mice weremaintained under 12-h light, 12-h dark cycles and were given pasteurized

Wayne Lab Blox (crude protein, 24%; fat, 4.0%; fiber, 4.5%, ash, 5.9%;vitamin and mineral fortified; Allied Mills, Peoria, IL) and HCI acidifiedwater (pH 2.8-3.2) supplemented with sodium menadione bisulfate (0.35

Mg/ml). In our research colony, the subline SWR/JWtBm (hereafter calledSWR/Bm, currently F140) has been maintained continuously since 1972by sibling matings after separation from SWR/J at F105. Subline SJL/JDgWtBm (hereafter called SJL/Bm, currently F74) has been maintainedsince 1960 by sibling matings following separation from SJL/J at F16.These mice and all other strains (see below) present or brought into theresearch colony were maintained under the same lighting and dietaryconditions, and were provided chlorinated (10-20 ppm sodium hypo-

chlorite) water without sodium menadione supplementation.Recombinant inbred strains were developed from SWR/Bm and SJL/

Bm progenitor strains for genetic analyses. The Rl strains were initiatedwith a SWR/Bm female (her sibling had a GCT) mated to a SJL/Bmmale. Four pairs of F1 offspring were intercrossed to produce 20 pairsof F2 offspring that were subsequently propagated by sister x brothermatings. Fourteen Rl strains, designated SWXJ-1 to -14, survived to thepresent (1985; F25-F28). By F25, there is approximately a 99% proba

bility that each genetic locus will be homozygous (11).Genetics. Studies on the inheritance of GCT susceptibility were

carried out through determination of GCT incidence in 10-week-old hybrid

F( females, Rl strain females, and females derived from embryo transferstudies. Hybrid F, females from reciprocal matings between SWR/Bmand A/HeJ, C57BL/6By, CBA/J, DBA/2J, or SJL/Bm mice were examined for presence of GCT at autopsy. Rl strain females were evaluatedfor presence of GCT by laparotomy under general anesthesia [tribro-

moethanol, 0.25 mg/g body weight (12) or sodium pentobarbital (13)]using aseptic techniques.

Finally two special sublines were established by transfer of preimplan-tation embryos, (a) from GCT-free SWXJ-6 females into GCT-susceptibleSWXJ-9 host mothers; and (b) from GCT-susceptible SWXJ-9 femalesinto GCT-free C57BL/6J x SJL/J F, host mothers (14). Following normal

delivery and subsequent weaning by host mothers, sibling offspring weremated, as were their progeny, to produce substrains SWXJ-6e andSWXJ-9e. GCT incidence in females of these two sublines was deter

mined by laparotomy as described above.Numerous observations were required to establish the presence and

incidence of GCT under control and experimental conditions. In view of

CANCER RESEARCH VOL. 45 NOVEMBER 1985

5575

on May 27, 2019. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

GRANULOSA CELL TUMORS IN YOUNG SWR MICE

the initial findings that GCT incidence in SWR/Bm mice was approximately 2-3%, an epidemiological formula estimating sample size wasadopted where n = log desired a level/log % of normal population (15).

Assumption of 0.025 for a and 97.5% normal mice yielded a sample sizeof 154 females. Although additional data revealed that GCT incidencewas closer to 3%, continued use of the 154 observation sample size foreach group improved the Â«level to 0.01.

Metaphase chromosome preparations were made from cultured ovarian GCT cells (16). Cell cultures were prepared from viable tumor tissueinitially dissected free of necrotic tissue, blood-filled cysts, and connectivetissue in a large depression slide filled with 1 ml of Eagle's minimal

essential medium. The tumor tissue was then placed in a dish containing5 ml of trypsin:EDTA solution (8 g NaCI, 0.2 g KH2PO4, 0.2 g KCI, 1.15g Na2HPO4, 0.2 g EDTA disodium salt/liter H2O plus 10 ml of 2.5%trypsin), and diced into 1-mm pieces. Trypsinizing solution and tumortissue were transferred to 25-ml Erlenmeyer flasks in a 37Â°C shaker

water bath for 10 min. Solutions were then transferred to sterile conicalcentrifuge tubes and centrifugea 5 min at 200 x g in a clinical table topcentrifuge. The supernatant was discarded and the pellet was resus-

pended in 4 ml of the Eagle medium. Cells were cultured for 48 h with amedium change at 24 h; 1 h prior to harvesting cells were treated with7.5 nQ colchicine. Then the colchicine containing medium was poured offand replaced with 2 ml trypsin:EDTA solution to loosen cells from thedish and each other. After 10-15 min the mixture was transferred to a

conical centrifuge tube and centrifuged at 200 x g. After decanting thesupernatant, 5 ml of warm hypotonie KCI solution (0.075 M) was addedto the centrifuged pellet and allowed to stand for 10 min. Then the tubewas centrifuged again at low speed, the supernatant was discarded, andthe pellet was fixed in 4 ml of methanohglacial acetic acid (3:1). Followingrecentrifugation 0.25 ml of fresh fixative was used to resuspend thepellet, and then standard air-dried chromosome preparations were made.

Metaphase chromosomes were stained with Giemsa (Harleco; No. 620)and inspected by light microscopy under a x100 oil immersion lens. Thechromosome number of at least 20 metaphase plates was recorded,with special attention paid to identifying the X chromosomes and tonoting any gross chromosomal abnormalities.

Tumor Detection and Preservation. Identification of mice with GCTwas made at autopsy or by inspection of ovaries exteriorized through aflank laparotomy incision. When a tumor was found, it was measured inthree dimensions and either removed and fixed for later microscopicexamination; or, in anesthetized mice, it was returned to the abdominalcavity. Muscle incisions were sutured with 4-0 chromic gut and the skin

incision was closed with wound clips. A similarly treated littermate orage-match control was retained, along with tumor bearing mice, forfuture studies. GCT specimens fixed in Bouin's solution were prepared

for light microscopy by standard paraffin embedding procedures to obtain6- to 8-Mm-thick serial or representative sections stained with hematox-

ylin and eosin.Transplantability. The growth potential of these spontaneous granu

losa cell tumors was tested by transplantation to SWR x SJL Fi hosts.Four tumors were harvested at autopsy from SWR females, rapidlystripped of excess connective tissue, and placed in a sterile depressionslide. Solid tissue was minced in a few drops of sterile 0.85% NaCIsolution and aliquots were placed s.c. by trocar over the right thorax of6-8 ovariectomized SWR x SJL Ft hosts. Host mice were examined

weekly by palpation and progressive growth of transplanted tumor, ifany, was recorded. Subsequent serial transplants were carried out at 4-to 5-month intervals with additional groups of 6-8 F, females.

RESULTS

Tumor Incidence in SWR Sublines. Over a period of approximately 4.5 years, 5033 SWR/J females from nonproductivematings and 1462 SWR/Bm virgin females were examined forthe presence of ovarian granulosa cell tumors. The incidences

of GCT in these mice are presented in Table 1. Both strainsshowed similar rates of just over 3% as adults of 14-20 weeksof age. The SWR/Bm mice were examined initially at 5-6 weeksof age and then again at 19-20 weeks of age. Two-thirds of theGCT were present at the first inspection, with 16 additionaltumors arising thereafter. Five mice developed a GCT in theopposite ovary after having been identified as carrying a unilateraltumor at 5-6 weeks of age. No GCT were found at autopsies ofthree hundred SWR/Bm females 21-26 days of age. Autopsieson retired breeder SWR/J females (6-7 months) revealed a GCTincidence of 0.44%.

Pathology. The ovarian GCT were characterized by rapidgrowth. At 5-6 weeks, the size of 24 tumors from SWR/Bm

females averaged approximately 6x5x5 mm in dimension,within a range of 3x3x2 to 12x10x10 mm. The averagevolume of 24 tumors, calculated on the assumption that tumorsresembled oblate spheroids;

v = 4/37ra2o

was 143 Â±37 (SD) mm. At 20 weeks, a separate group of 20GCT averaged approximately 11 x 8 x 6 mm, within a range of5x5x3 to 21 x 15 x 11 mm and had average volumes of 667Â±184 mm. By 10 weeks of age, approximately one-third of the

tumors encountered during laparotomy were too large to exteriorize for measurement; such tumors typically weighed in excessof 2 g. Females bearing large GCT could be identified by theirpoor pelage, abdominal distension, protrusion of perianal tissues,and pale ear color, presumably due to the GCT-induced anemia[hematocrits (n = 9 pairs): GCT females 49.0 Â± 1.3; normal

littermates: 53.5 Â±1.1]. Since SWR females are fractious, palpation was not dependable as a detection method until a GCTapproximated the size of a kidney. There were 222 unilateral(127 right ovary, 95 left ovary) and 27 bilateral tumors among249 GCT found in SWR/J and SWR/Bm mice. The apparentpreference of tumors for the right ovary was not statisticallysignificant (x2 = 2.32; d.f. = 1).

Tumorous ovaries appeared pinkish white, exhibited variablysized hemorrhagic cysts, were supplied by large dilated bloodvessels, and usually were encapsulated by the ovarian bursa.Serial sections of small GCT found Â¡nmice at 5-6 weeks of age

showed single to multiple foci of rapidly proliferating granulosacells within follicles that did not contain oocytes (see Fig. 1).Follicles in various stages of development were displaced towardthe periphery and presumably were destined for elimination bymechanical compression. The GCT cellular arrangements variedfrom solid masses and occasional luteinized patches (Fig. 2a),follicular-like arrangements around giant mononuclear cells (per-

Table 1Incidence of granulosa cell tumors in SWRmice

SublineSWR/JSWR/J

SWR/BmSWR/BmReproductive

statusNonproductive

matingRetired breedersVirgin8

VirginAge

(wk)14-1626-30

5-619-20TotalExamined5033457

14621462Normal4866455

14271411No.

Oftumors1672

35"51Incidence

(%)3.320.44

2.393.49

* Same mice reexamined at 19-20 weeks for presence of additional GCT.6 Fiveof these miceadvancedfrom unilateralto bilateraltumors at 19-20 weeks.


5576



â€¢vÅ“sv^


haps macrophages; Fig. 2b), or cystadenomatous patterns (Fig.2c), to hemorrhagic cysts and areas of small to massive necrosis.The primary tumor cells were characterized by their immatureappearance, i.e., small size with minimal cytoplasm, ovoid nuclei,and abundant mitotic figures that in some areas exceeded 12per x400 field (Fig. 2d). Thecal and stromal cells were typicallynot involved with these tumors. The contralateral ovary, if it didnot contain a neoplasm, was so atrophie as often to be difficultto locate. Histological sections of these atrophie ovaries revealedoocytes within small follicles, but no large antral follicles; corporalutea, or corpora albicans.

The malignant potential of the spontaneous GCT was demonstrated by the following two lines of evidence: first, experimental attempts to establish transplantable tumor lines from fourseparate SWR/Bm GCT were partially successful. That is, afteran initial period of 6-12 months, one failed to become established

in any host, two were not recovered during the fourth transplanteffort, and one continues after 17 transplant generations. Transplants of tissue from this latter GCT line grew equally well inovariectomized and in intact SJL x SWR FT hosts. Attempts topassage this transplantable tumor by cell-free extracts or by cell

homogenates failed. Second, when females bearing GCTachieved more advanced ages of 250-350 days, metastatic

nodules were observed in some of them in one or more sites,including abdominal lymph nodes, kidneys, pancreas, liver, anddiaphragm (Fig. 3). Such findings clearly demonstrate that theSWR GCT are carcinomatous.

Uteri of GCT-bearing mice were markedly thickened, and

vaginal cytology was generally acyclic but did show cornified andnucleated epithelial cells mixed with leukocytes. The pubic ligament was broadened and the cervix was enlarged and solid.

Heritability and Genetic Studies. In the course of separateinvestigations of reproductive biology, two young adult SWR/Bm x SJL/Bm FI females were found to have unilateral GCT.Since such tumors had never been observed in SJL/Bm, andsince both SWR and SJL strains originated from Swiss stockimported to the Jackson Laboratory at different times, it waspossible that a genetic basis for occurrence of GCT dependedon a gene with low penetrance, or perhaps on more than onegene. The possible existence of a dominant trait with lowpenetrance was explored through reciprocal matings of SWR/Bm with A/HeJ, C57BL/6By, CBA/J, DBA/2J, and SJL/Bm. Thedata gathered on 8- to 10-week-old F, mice are presented in

Table 2.The GCT were again observed among SWR/Bm x SJL/Bm F! mice but were conspicuously absent from SJL/Bm xSWR/BM F, and all other F! females. In addition, preliminarydata show that these tumors are present in SWR/Bm x ST/bJF, mice (11 of 213) but not in ST/bJ females (0 of 171 ), a Danishstrain also likely to be derived from Swiss stock. These findings(a) eliminated unqualified hypotheses of a simple dominant orrecessive condition with low penetrance, (b) outlined a case forthe more powerful Rl strains as alternative tools for geneticanalyses, and (c) suggested an important role for the SWRmaternal organism in transmission of GCT susceptibility.

Additional studies on the mode of inheritance of the GCT traitwere pursued with the newly developed SWXJ Rl strains. First,as each SWXJ strain achieved inbred status at F20, matingswere expanded and resultant female offspring were examinedby laparotomy to determine the incidence of GCT at 10 weeksof age. Table 3 presents the results obtained with the two

Table 2

Occurrence of ovarian tumors in F, offspring of SWR/Bm mice outcrossed to fivedifferent inbred strains

MatingSWR/Bm

xSJL/BmSWR/BmxA/HeJSWR/BmxC57BL/6BySWR/BmxCBA/JSWR/BmxDBA/2JSJL/Bm

xSWR/BmA/HeJxSWR/BmC57BL/6By

xSWR/BmCBA/JxSWR/BmDBA/Jx SWR/BmTumor

bearing/totalfemales3/1490/1670/1740/2340/1680/1740/1580/2130/2990/174

Table3

Incidence of ovarian tumors in progenitor (SWR/Bm and SJL/Bm) recombinantinbred strains at generation F20 to F27

StrainsProgenitorSWRSJL/BmRecombinantSWXJ-1SWXJ-2SWXJ-3SWXJ-4SWXJ-5SWXJ-6SWXJ-7RecombinantSWXJ-8SWXJ-9SWXJ-1

0SWXJ-11SWXJ-12SWXJ-13SWXJ-1

4Tumor

bearing/totalfemales25/10020/5803/163Â°0/1660/1712/1260/1710/1670/1690/16626/170Â°0/1740/1740/1600/1670/167Incidence(%)2.5001.84001.596000015.2900000

" Minimum sample size must exceed 154 females based on assumption of 2%

incidence in SWR/Bm progenitor strain.b Tentative incidence based upon incomplete data.c Overall incidence to the autumn of 1984: 120/950 (12.63%).

progenitor and 14 Rl strains. No SJL/Bm females producedtumors, while SWR/Bm had an incidence of 2.50%. Females of11 Rl strains showed no tumors and thus appear similar to SJL/Bm. On the other hand, females of strains SWXJ-1, -4, and -9produced GCT at rates of 1.84,1.59, and 15.29%, respectively.The 15% incidence in SWXJ-9 mice represents a 5-fold increaseover that observed in mice of the SWR sublines. When SWXJ-9females were examined for GCT incidence at ages of 6,16, and32-40 (retired breeders) weeks, the rates were 5.3% (5 of 95),

14.3% (15 of 105), and 1.89% (2 of 106), respectively. Theretired breeders, of course, represent a selected populationbecause females that have a GCT rarely produce litters. Datafrom retired breeders indicate that few GCT appear after 10weeks of age. Since Rl strain SWXJ-4 mice reproduce at marginal levels, the incidence data are not based on 154 observations. Nevertheless it appears that SWXJ-4 females have a SWR/Bm progenitor-like GCT incidence.

Second, two Rl strains were utilized to develop special stocksby embryo transfer techniques (SWXJ-6e and SWXJ-9e) to gain

insight into the mode of maternal transmission of the GCT trait.The results showed 0% (0 of 166) GCT in SWXJ-6e females and11.3% (23 of 203) GCT in SWXJ-9e females; however, the latterincidence was not significantly different (x = 1.28; d.f. = 1) from


5577




the expected incidence of 15.29% found in Table 3. These dataindicated that factor(s) essential to GCT expression reside withinthe fertilized egg and are unlikely to be transferred via milk orplacental contact.

Metaphase chromosome preparations were obtained fromtumors of four SWR/J and two SWR/Bm mice and from bonemarrows of these same mice. The majority of metaphases fromcultured tumor and marrow cells were found to possess thenormal 20 pairs of chromosomes, including 2 X chromosomes.No indication of gross alterations in chromosome morphologywere observed in metaphases from either tumor cells or frommarrow cells. Our level of analysis, however, does not precludethe existence of more subtle changes such as small inversions,reciprocal translocations, or deletions.

Effect of Housing Conditions on GCT Incidence. The incidence of GCT in monogamously mated (breeding) mice wascompared with that in mice held at 2 females per cage (cycling)and with that of mice caged in groups of 20 (cycling suppressed).The results showed that regardless of the housing conditionsmanipulating reproductive cycles, the GCT incidences varied ina nonsignificant fashion from 1.32 to 2.01%. This contrasts withthe observations of Muhlbock (17) that the incidence of mammarytumors was affected by caging conditions.

DISCUSSION

The sublines of SWR and SWR-derived mice reported here

represent a valuable addition to animal models available forstudies of ovarian tumorigenesis. Within the WHO histologicalclassification system (6), these tumors are of the granulosa celltype in the sex cord-stromal grouping, and are the type of ovarian

tumor most often observed among laboratory rodents (5). Webelieve that distinctive features of the GCT found in SWR miceoffer an excellent opportunity to investigate this kind of tumor.

The most unusual characteristic of the GCT in SWR mice istheir occurrence in pubertÃ ! females. We believe that tumordevelopment is initiated during the fourth to fifth week of age,and that by 8-10 weeks, the majority of females destined todevelop ovarian tumors will have done so. This early onset is incontrast to literature reports that spontaneous rodent GCT typically have been found in females of advanced age (2, 4). In ananalysis of spontaneous lesions in SWR/J mice, Rabstein ef a/.(18) observed 2 hemangiomas of the ovary among 402 femalesin a life history study, but did not find ovarian GCT. Other groupsof workers have noted GCT in young adult SWR/J mice (19-20)

and in older randomly bred Swiss mice (21).A second distinctive feature of GCT in young mice is that

these tumors can arise from multiple foci in ovaries that arepopulated with growing oocytes and follicles. Thus it is notnecessary for an ovary to be devoid of germ cells prior to theinitiation of GCT growth. This situation is markedly different fromexperimental induction [i.e., irradiation (22); carcinogens (23);ovarian grafts to spleens (24); or mutant gene-induced germ cell

deletion, i.e., W alÃelesof dominant spotting (25) and SI alÃelesofSteel (26)] models where oocyte loss is accepted as a fundamental early step in ovarian tumorigenesis. GCT in older rodentsare correlated with cessation of ovarian activity, and in somestrains with substantial or total depletion of oocytes (27, 28).With respect to SWR ovaries, we are not certain whether those

follicles that begin rapid hyperplasia of granulosa cells and ultimately become tumors even possessed oocytes; it only can bepresumed that they did.

A third important feature of GCT in young SWR mice is theexperimental and spontaneous evidence of GCT neoplastic potential. Successful transplantation was accomplished in three offour attempts. One transplantable line continues to flourish after17 passages, growing equally well in either ovariectomized orintact hosts for 3-5 months before killing the host mice. Althougha systematic study of the incidence of mÃ©tastasesamong GCT-bearing females has not been carried out, more than 15 examplesof mice SWR/J, SWR/Bm, SWR x SJL F,, and SWXJ-9 havebeen recorded with small to large nodules scattered around theorgans of the abdominal cavity. The relevance of these latterobservations is emphasized by the scarcity of mÃ©tastasesfoundin reports on spontaneous ovarian tumors in mice. Only Slye eial. (29) have noted that 4 of 46 ovarian neoplasms in Slye strainmice were associated with metastatic nodules. Experimentallyinduced ovarian tumors have not been reported to be spontaneously metastatic. However, Furth (30) demonstrated that approximately one-half of the ovarian tumors induced in X-irradiated

mice could be successfully transplanted.The genetic analysis of the incidence of GCT in SWR mice is

a formidable task, and any substantive conclusion about thecharacteristics and number of genes involved is hampered atthis time by lack of F2 and backcross generation data. However,the historical relationship between SWR and SJL, the preliminarydata on GCT incidence in SWR x STb/J F, mice, the repeatedobservation of GCT in SWR x SJL F, but not in SJL x SWR F,mice, and the appearance of GCT susceptibility in the SWXJ Rlstrains suggest a familial basis for this tumor. In addition, thefinding of an unchanged high GCT incidence in SWXJ-9e and acontinued "0" GCT incidence in SWXJ-6e suggests that GCT

susceptibility is not readily influenced by factor(s) transmittedthrough milk or through placenta-uterine contact. This interpre

tation of the embryo transfer results, particularly for the negativedata from SWXJ-6e, must be tempered by the fact that we

cannot state at this time whether or not this strain is uniquelyresistant to GCT formation. Our hypothesized explanation ofthese data is that SWR, SJL, and probably ST/b are related andshare essential genomic material that interacts with a uniqueSWR factor (cytoplasmic?) to result in GCT expression. Theshared genomic element(s) function like a gene fixed in theseSwiss stocks and the rate of GCT depends upon the degree ofinteraction between the putative gene and cytoplasmic elements.The increased incidence in SWXJ-9 females may be a mutational

event associated with either of these entities or with some asyet unidentified third factor (gene?) that facilitates the interactionof the putative nuclear gene with the cytoplasmic factor. Theoretically all of the Rl strains should be expressing GCT formation,since each was derived from a single SWR/Bm female. However,since 11 of 14 Rl strains are negative for occurrence of GCT,this implies (a) the existence of some factor essential to GCTinitiation that is easily lost or is readily mutable to a wild-type

form, or (b) that the single gene concept needs to be replacedby a more complex system such as a two-gene model. The

involvement of virus with the induction of these tumors is anattractive idea for which we have little evidence other than failureto find obvious viral particles in electron micrographs of severaldifferent GCT specimens. Regardless of the etiology of GCT in


5578




SWR mice, this mouse model presents opportunities to analyzethe genetic, environmental, and host aspects of a spontaneousmalignant tumor.

ACKNOWLEDGMENTS

The authors gratefully acknowledge the dedicated technical assistanceof Stetson Carter and Kathy Shultz, and thank Dr. Janan Eppig for carrying out thechromosome analyses.The Jackson Laboratory is fully accredited by the AmericanAssociation for Accreditation of Laboratory Animal Care.

REFERENCES

1. Dickie, M. M. The usageof F, hybrid and backcross generationsto study newand uncommon tumor types. J. Nati. Cancer Inst., 75: 791-799,1954.

2. Carter, R. L. Pathology of ovarian neoplasms in rats and mice. Eur. J. Cancer,3; 537-543,1968.

3. Marchant, J. Animal models for tumors of the ovary and uterus. In: AncelBlaustein (ed.), Pathologyof the FemaleGenital Tract, Chapt. 38. New York:Springer-Verlag, 1977.

4. Liebelt, A. G. Ovarian tumors: mice. /: P. Altman and D. Katz (eds.), Inbredand GeneticallyDefinedStrains of Laboratory Animals, Part I, Mouse and Rat,pp. 193-206. Bethesda, MD: Federationof AmericanSociety for ExperimentalBiology, 1979.

5. Frith, C. H., Zuna, R. E., and Morgan, K. A morphologic classification andincidence of spontaneous ovarian neoplasms in three inbred strains of mice.J. Nati. Cancer. Inst., 67: 693-702, 1981.

6. Serov, S. F., Scully, R. E., and Sobin, L. H. Histological typing of ovariantumors. Geneva:World Health Organization, 1973.

7. Scully, R. E. Ovarian tumors, a review. Am. J. Pathol.,83: 686-720,1977.8. Weiss, N. S. Epidemiology of ovarian cancer. In: E. D. Murphy and W. G.

Beamer (eds.). Biology of Ovarian Neoplasia,Vol. 50, UICCTechnical ReportSeries, pp. 34-49. Geneva: InternationalUnionAgainst Cancer, 1908.

9. Fox, H. Sex cord-stromal tumours of the ovary. J. Pathol., 745: 127-148,1985.

10. Miller, A. B. An overview of hormone-associated cancers. Cancer Res., 38:3985-3990,1978.

11. Bailey, D. W. Sources of subline divergence and their relative importance forsublinesof six major inbred strains of mice. In: H. C. Morse III (ed.),OriginsofInbred Mice, pp. 197-215. New York: Academic Press, Inc., 1978.

12. Lumb W. V. Small Animal Anesthesia, p. 202. Philadelphia:Lea & Febiger,1963.

13. Barnes, C. D., and Eltherington, L. G. Drug dosage in laboratory animals.Berkeley: Universityof CaliforniaPress, 1964.

14. Hoppe, P. C., and Coman, D. R. Reduced survival in utero from transferredmouse blastocysts compared with morulae. Gamete Res., 7: 161-167,1983.

15. Schwabe, C. W., Riemann,H. P., and Franti, C. E. Epidemiologyin VeterinaryPractice, pp. 66-97. Philadelphia:Lea & Febiger, 1977.

16. Eppig, J. New Approaches to the Linkage Map of the Mouse. Ph.D. Dissertation, Universityof Maine at Orono, 1982.

17. Muhlbock, 0. The hormonal genesis of mammary cancer. J. Endocrinol., 77:vii-xv, 1958.

18. Rabstein, L. S., Peters, R. L., and Spann, G. J. Spontaneous tumors andpathologic lesions in SWR/J mice. J. Nati. Cancer. Inst., 50: 751-758,1973.

19. Bauminger, S., Webb, C., and Okon, E. Characterization of a spontaneousgranulosacell tumor in SWR mice. Cancer Treat. Rep., 63: 1178,1979.

20. Diwan, B. A., and Meier, H. Strain- and age-dependent transplacental carci-nogenesisby 1-ethyl-1-nitrosoureain inbred strains of mice. Cancer Res., 34:764-770,1974.

21. Shimizu, H., Nagel, D., and Toth, B. Ethylhydrazinehydrochloride as a tumorinducer in mice. Int. J. Cancer, Ã•3:500-505, 1974.

22. Guthrie, M. J. Tumorigenesis in ovaries of mice after X-irradiation. Cancer(Phila.),11: 1226-1235, 1958.

23. Marchant, J. The developmentof ovarian tumors in ovaries grafted from micepretreated with dimethylbenzanthracene.Inhibition by the presence of normalovarian tissue. Br. J. Cancer, 14: 514-518, 1960.

24. Biskind, M. S., and Biskind,G. R. Developmentof tumors in the rat ovary aftertransplantationinto the spleen.Proc. Soc. Exp. Biol. Med., 55:176-179,1944.

25. Murphy, E. D. Hyperplasticand early neoplasticchanges in the ovariesof micewith geniedeletion of germ cells. J. Nati. Cancer Inst., 48:1283-1295,1972.

26. Terada, N., Kitamura, Y., Namiki, M., Kuroda, H., Ito, M., Takasuku, D., andMatsumoto, K. Production of androgens and estrogens by tubular adenomaswhich developed in ovaries of mutant mice of S//S/' genotype. Cancer Res.,44: 1827-1830, 1984.

27. Jones, E. C., and Krohn, P. L. The relationship between age, numbers ofoocytes, and fertility in virgin and multiparous mice. J. Endocrinol., 27: 469-495,1961.

28. Murphy, E. D. Characteristic tumors. In: E. L. Green (ed.). Biology of theLaboratory Mouse, pp. 536-579. New York: Dover Publications,1966.

29. Slye, M., Holmes,H. F., and Wells, H. G. Primary spontaneous tumors of theovary in mice. Studies on the incidenceand heritabilityof spontaneoustumorsin mice. J. Cancer Res., 5: 205-226,1920.

30. Furiti. J. Transplantabilityof induced granulosa cell tumors and of luteoma inmice. Secondary effects of those growths. Proc. Soc. Exp. Biol. Med., 67:212-214, 1946.

FIG. IFig. 1. Example of GCT in ovary of a 43-day-old SWR/Bm mouse. The ovary shows normal follicles (nf ) with oocytes crowded to the periphery, cysts (c), and

granulosa cell tumor (GCT)foci, x 32.


5579




**.\ 'SÂ»,:j* Ve

Fl8.2a

. - - - -

i '"-,.. 'V..

> *

,,

Fig.2. Examplesof GCT morphology observed in ovaries of young adult SWR female,a, section through GCT with areas of high cell density and luteinizedgranulosacells (x. 128). b, section displaying follicular pattern (x 128). c, section through GCT with cystadenomatous pattern (x 80). d, examples of proliferating GCT cells withmitotic figures (arrows; x 320).


5580




>O'

irin

Gmn

>â€¢<>â€¢â€¢''A- ' â€¢

H mn

mn

m

Fig.3. Metastatic invasion of abdominal organs by GCT. a, kidney glomeruli (G) adjaÅ“nt to GCT metastatic nodule (mn; x 200). b, exocrine (E) and endocrinepancreas (islet of Langerhans;/) with metastatic nodule (mn; x 80). c, hepatic (H)tissue with lymphocytic infiltration(arrows)adjacent to a metastatic nodule (mn; x 200).d, diaphragmatic muscle (Dm)and invading metastatic neoplasm(mn; x 80).


5581



1985;45:5575-5581. Cancer Res Wesley G. Beamer, Peter C. Hoppe and Wesley K. Whitten Young SWR MiceSpontaneous Malignant Granulosa Cell Tumors in Ovaries of

Updated version

http://cancerres.aacrjournals.org/content/45/11_Part_2/5575

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/45/11_Part_2/5575To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



spontaneous malignant granulosa cell tumors in ovaries of...

Documents