variability of androgen-related phenotypes in the shionogi ... · nucleus. however, relapse is...

[CANCER RESEARCH 45, 682-689, February 1985]

Variability of Androgen-related Phenotypes in the Shionogi Mammary

Carcinoma during Growth, Involution, Recurrence, and Progressionto Hormonal Independence1

Nicholas Bruchovsky,2 Paul S. Rennie, Navdeep Otal, Adrian Vanson, Martin Giles, and Hugh Pontifex

Departments of Cancer Endocrinology [N. B., P. S. R., N. O., M. G.], and Pathology [H. P.], Cancer Control Agency of British Columbia, Vancouver, British Columbia,V5Z 3J3, and Department of Medicine [A. V.Â¡,University of Alberta, Edmonton, Alberta, T6G 2G3, Canada

ABSTRACT

Several parameters of androgen action were measured inhormone-dependent Shionogi carcinoma cells during phases ofgrowth, regression, and recurrence. In the parental C1 line understeady state conditions, dihydrotestosterone is localized exclusively in the nucleus while testosterone is confined almost entirelyto the cytoplasm. After castration, the concentration of testosterone declines more rapidly than that of dihydrotestosterone.Spontaneous recurrent growth is not accompanied by significantelevation of the whole-tissue concentration of either androgen.Neither are changes observed in the concentration of cyto-

plasmic receptor or in the rate of uptake of androgens into thenucleus. However, relapse is associated with the appearance ofa glucose-6-phosphate dehydrogenase double-enzyme pheno-

type and a loss of responsiveness to androgen withdrawal. Theautonomous C3 variant line which is devoid of androgen-related

markers is characterized by a deficiency of androgen retentionby whole tissue and possibly a permeability defect of the plasmamembrane. This variant tends to express a glucose-6-phosphatedehydrogenase double-enzyme phenotype. In contrast, the autonomous C4 variant line retains the ability to concentrate modest levels of testosterone in whole tissue and high levels ofdihydrotestosterone in the nucleus. Although the number ofnuclear binding sites is the same as that observed in the parentalC1 line, the concentration of cytoplasmic receptor and the rateof nuclear uptake of androgens are relatively decreased. Expression of a glucose-6-phosphate dehydrogenase double-enzyme

phenotype is less frequent than in the autonomous C3 variantline. The above results suggest that a recurrent tumor maycontain hormone-sensitive cells which resume growth in anandrogen-depleted environment. They also imply that progression from the androgen-dependent to the autonomous conditioninvolves the selection and outgrowth of hormone-insensitive cells

of variable phenotype.

INTRODUCTION

The recurrence of advanced carcinomas of the breast, endo-

metrium, and prostate is inevitable in the majority of patientsdespite manifestations of encouraging responses to primarytherapy based on surgery and hormonal manipulations (9, 12,30, 37). On the one hand, uncontrollable recurrent growth canbe explained by the emergence of hormone-resistant cells from

1This research was supported by grants from the National Cancer Institute ofCanada and the Medical Research Council of Canada.

2To whom requests for reprints should be addressed, at Department of CancerEndocrinology, Cancer Control Agency of British Columbia, 2656 Heather Street,Vancouver, Canada V5Z 3J3.

Received June 18, 1984; accepted October 21, 1984.

an initial mixed population consisting of both hormone-sensitiveand -insensitive cells. On the other hand, progression of a singleclone of hormone-dependent cells might result in the outgrowth

of resistant cells through stepwise and irreversible changes incertain biological properties (3,17, 23,31 ). Although progressionmay give rise to considerable phenotypic heterogeneity, it is notclear whether the process is related to multicellular origin ofcarcinoma, somatic mutation, or adaptational change (11, 23).Evidence from animal experiments generally supports the ideathat the progression of carcinomas of the breast (36) and prostate (23, 24) is due to selection of hormone-resistant clones.

Since the androgen-related phenotypes of dependent and

autonomous variants of the Shionogi mammary carcinoma areconsiderably different (5), we assumed that sequential phenotypic screening of the tumor during states of growth, regression,and recurrence might yield chronological evidence concerningthe etiology of progression. Our results suggest that progressionis a complex process involving selection of resistant clones andadaptation of surviving hormone-sensitive cells.

MATERIALS AND METHODS

Shionogi Mammary Carcinoma. Variant lines of the SC-115 androgen-dependent mouse mammary carcinoma may be grouped into at

least 4 different classes as described in a previous report (5). Class 1(C1) tumors are androgen dependent and positive for cytoplasmic binding, nuclear uptake, and displaceable nuclear binding. The dependenttumors used in the present study have progressed to a stage which ischaracterized by a high percentage of recurrence following castration.Class 2 (C2) tumors are autonomous and positive for cytoplasmic bindingand nuclear uptake but have relatively less displaceable nuclear binding.Class 3 (C3) tumors are autonomous and devoid of any positive markersof hormone action. Class 4 (C4) tumors are also autonomous. They arenegative for cytoplasmic binding and nuclear uptake but positive fordisplaceable nuclear binding. Depending upon conditions of propagation,the tumor may be diploid or aneuploid but retains a karyotype that iscompatible with the presence of 2 X chromosomes (5). Abdominalcastrations were performed using ether. At appropriate times, experimental animals were anesthetized and killed by decapitation; the tumors(2 to 3 g or 1 to 7 days after castration) were dissected free of s.c. tissueand weighed.

Homogenization of Tissue. The tumors were minced with scissorsand forced through a stainless steel wire screen (30 mesh). Fragmentsof tissue were collected in 50 ml of ice-cold tissue culture medium andsuspended after centrifugation in 40 to 50 ml of 10 ITIM2-|[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]amino|sulfonic acid buffer (pH 7.0) contain

ing 0.5 mw mercaptoethanol, 1.5 rriM CaCk, 0.05 mw EDTA, 5 mw MgCk,and 0.25 M sucrose. The suspension was homogenized in a Dounceapparatus and further separated into cytosol (104,000 x g supernatant)and purified nuclear fractions as described elsewhere (5).

CANCER RESEARCH VOL. 45 FEBRUARY 1985

682

on July 23, 2020. © 1985 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

ANDROGEN INDEPENDENCE IN SHIONOGI MAMMARY CARCINOMA

Radioimmunoassay. Methods for extracting and assaying the concentrations of testosterone and dihydrotestosterone in whole-tissue

homogenates and samples of purified nuclei have been reported previously (10, 14).

Quantitation of Cytoplasmic Receptor. Tumor tissue was mincedwith scissors and suspended in a final volume of 3 ml of ice-coldphosphate-buffered saline (6.3 mm Na2HPOÂ«:0.8HIM KH2PO4:0.154 MNaCI, pH 7.4). [1,2-3H]Testosterone (40 Ci/mmol) was added to a final

concentration of 500 nM; duplicate samples were prepared which contained a 100-fold excess of nonradioactive testosterone. After warmingfor 2 min in a 37Â°water bath, the samples were washed with phosphate-

buffered saline and homogenized. Cytoplasmic receptor was recoveredfrom the cytosol by precipitation with 80% ammonium sulfate; the finalsample containing 5 to 20 mg protein was then analyzed by SephadexG-25:G-200 dual-column chromatography. Specific binding was calcu

lated as the difference between the amount of radioactivity in the receptorpeak in the sample without competitor and the one with competitor.Details of this procedure are given in Ref. 5.

Quantitation of Displaceable Nuclear Binding. Purified nuclei wereresuspended in 0.5 ml of 2-|[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]

aminolsulfonic acid buffer containing 50 rnw NaCI and sonicated withthree 5-sec pulses of a Bronwill Biosonik III sonicator at an intensity

setting of 50. After addition of an equal volume of buffer containing 1.15M NaCI, the resultant solution was sonicated with three 5-sec pulses.

Centrifugation of this solution at 17,000 x g for 30 min yielded asupernatant fraction of solubilized chromatin (6). One-half of the samplewas incubated with 50 nw [1,2-3H]dihydrotestosterone, while the otherhalf was incubated with both [1,2-3H]dihydrostosterone and a 1,000-fold

excess of nonradioactive dihydrotestosterone. Following incubation for16 hr at 20Â°, the samples containing 5 to 15 AaÂ»units (absorption at

260 nm in 1 N NaOH) were analyzed by Sephadex G-25:G-200 dual-

column chromatography. From the difference in the amount of androgenbinding between the sample without competitor and the one with competitor, an estimate of displaceable nuclear binding was obtained. Themean K,, of the binding reaction with dihydrotestosterone as ligand was4 ran.

Measurement of Nuclear Uptake of Androgens. Tumor-bearing micewere given injections of [1,2-3H]testosterone; (200 nC\ i.V.); this dose

maintains the rate of uptake of androgens into the nucleus at a maximumfor at least 30 min (6). After this interval, the mice were killed and thetumors were removed. Samples of purified nuclei recovered from homogenized tissue were assayed for radioactivity, and the results wereexpressed as dpm/nucleus.

Glucose-6-Phosphate Dehydrogenase Purification. Glucose-6-

phosphate dehydrogenase was recovered from the cytosol fraction asfollows. The final 104,000 x g supernatant was subjected to a fractionalprecipitation with ammonium sulfate which was added slowly over aperiod of 50 min while the temperature was controlled at 0Â°.The 30 to

50% fraction containing glucose-6-phosphate dehydrogenase (13) was

centrifugad at 18,000 x g for 20 min to yield a protein precipitate whichwas frozen at -20Â°. At the time of analysis, the sample was thawed and

resuspended in 50 mw NaCI and then desalted by gel filtration chromatography using PD-10 Sephadex G-25. The desalted samples containingabout 150 tig protein were layered on disc gels made of 7.5% acrylam-

ide:0.18% bis, pH 8.9 and run at 4.0 mamp/gel for about 2 hr (18). Thegels were stained with an enzyme-specific blue formazan mixture as

described by Hunter (22). Scanning of gels was performed in a Gilfordspectrophotometer at 550 nm.

Histology. Tumors were fixed in 10% formalin and embedded inparaffin. Sections were stained with hematoxylin and eosin and examinedby light microscopy.

Radioactive Materials. [1,2-3H]Dihydrotestosterone (40 Ci/mmol) and[1,2-3H]testosterone (40 Ci/mmol) were purchased from New England

Nuclear (Boston, MA). Purity was checked by thin-layer chromatography,

and the steroids were considered acceptable only if they were 95 to100% pure. Solutions for experimental incubations and injections were

prepared as follows. Radioactive steroid in ethanol:benzene was driedunder N2 and dissolved in a small volume of ethanol. Distilled H2Ocontaining 5% (v/v) polyoxethylene sorbitan monopalmitate was thenadded to bring the steroid solution to the desired concentration. For invivo administration, 250 /Â¿Iof such a solution were injected into the tailvein of each mouse.

Radioactivity was counted in a solution containing 6 g of diphenylox-azole:75 ml of water:232 g of Bio-solv (BBS-3; Beckman Instruments,Fullerton, CA) per liter of toluene. The efficiency of counting 3H with this

preparation was 33%.Other Analytical Procedures. DMA was measured by the method of

Burton (8), and protein was measured by the method of Bradford (1).For purposes of calculation, the following recoveries were used: nuclei,

7.5 x 107 Â±0.6 x 107 (S.E.)/g tumor (n = 52); DNA content of whole

tissue, 14.6 Â±0.8 mg/g tumor (n = 29); DNA content of nucleus, 15.8Â±2.8 pg/nucleus (n = 12); AaÂ»,5.4 x 10~7 Â±0.4 x 10~7 units/nucleus(n = 31); protein content, 6.3 x 10"e mg/cell (5).

RESULTS

Growth, Regression, and Recurrence of Androgen-depen-

dent Shionogi Carcinoma Cells. Following the injection of 2 x107 Shionogi carcinoma cells into adult male mice of the DOS

strain, the tumor becomes palpable on the tenth day and growswith a doubling time of approximately 24 hr as shown in Chart1A Withdrawal of androgens brought about by castration induces regression of the tumor such that over a period of 5 to 10days the bulk of tissue disappears and is not detectable bypalpation. About 30 to 35 days after the transplant, recurrent

o

1 6

I

castration testosterone/

10 20 30 40 50

Time After Transplant (days)

Chart 1. Growth, regression, and recurrence of hormone-dependentShionogicarcinomacells. Two groups of mateDOS mice were given implants of about 2 x107tumor cells. After 14 days, when the tumors had reached a mean weight of 4

g, the animalswere castrated. The first group (A)consisting of 20 animals receivedno further treatment while the 5 animals in the second group (B) were each giveninjections of testosterone (200 Â¿ig)on the 27th and 28th days after the implant.Tumor diameters were measuredwith calipers. The formula (length x width*)^ =

mass (mg)was used to estimate the mean weight of the tumors (35).


683




growth is observed with a much slower doubling time of 72 hr.Earlier reactivation of the tumor can be elicited by the injectionof testosterone as shown by the results in Chart IB. The latentperiod before recurrent growth resumes is markedly foreshortened, and the doubling time is abbreviated 3-fold to 24 hr or

less.The histology of the Shionogi carcinoma during the 3 phases

of response is shown in Figs. 1 to 6. During the growth phase(Fig. 1), the tumor has the appearance of an undifferentiatedmedullary mammary carcinoma. The cells have poorly definedborders and large pleomorphic nuclei often containing severalsmall nucleoli. Mitotic figures are common. The phase of regression, illustrated in Fig. 2 is characterized by large numbers ofcells undergoing lytic degeneration. The nuclei are separatedfrom one another and take on an ill-defined smudged appear

ance. Mitotic figures are absent. Growth of recurrent tumorstends to be concentrated around blood vessels as shown in Fig.3. On either side of the capillary in the center of the field viabletumor cells are seen, but further away from the vessel, the cellsbecome spindle shaped. These cells are shown under highermagnification in Fig. 4. Being somewhat removed from the bloodsupply, it is possible that such cells owe their unusual appearance to hypoxia rather than to autonomous growth (27). Thearchitecture of autonomous variant lines C3 and C4 is shown forcomparison in Figs. 5 and 6, respectively. These tumors arepoorly differentiated, with ill-defined cellular borders and large

pleomorphic nuclei with prominent and sometimes multiple nucleoli. Mitoses are numerous in keeping with the rapid doublingtime of the tumors. The autonomous variant lines thus resembleboth the androgen-dependent C1 parental line depicted in Fig. 1

and the perivascular cells of the recurrent tumor shown in Fig.3.

Concentration of Androgens. Since the results in Chart 18indicate that a significant number of hormone-dependent cells

survive regression, it is possible that the recurrent growth described in Chart 1/1 might simply represent renewed androgen-

dependent proliferation, with the adrenal glands providing thesource of hormone. To measure the effects of hormone withdrawal by castration, the concentrations of testosterone anddihydrotestosterone were measured during the 3 phases ofgrowth of the Shionogi carcinoma illustrated in Chart 1/4; theresults are presented in Table 1. The mean concentration oftestosterone in whole tissue is 1.9 pmol/mg DMA, and surprisingly, the nuclear compartment is virtually devoid of testosterone.The whole-tissue concentration is equivalent to 28 nw, virtually

identical to the 10 to 20 nw plasma concentration of testosterone

in male mice of the DOS3 and other strains (15, 29). This

correspondence of values suggests that either the tumor is highlyvascular or that testosterone diffuses freely into the cytoplasmof the malignant cells. The mean concentration of dihydrotestosterone in whole tissue is 1.3 pmol/mg DMA, nearly identical tothe mean concentration in the nucleus, 1.4 pmol/mg DMA (f test,p > 0.05). These results indicate that all of the intracellulardihydrotestosterone is concentrated in the nucleus and that thecytoplasm is devoid of dihydrotestosterone. Castration bringsabout a reduction in the whole tissue concentration of testosterone to trace levels in the regressing tumors, and the concentration remains low in the recurrent tumors as well. Owing to thenuclear localization of dihydrotestosterone, the whole-tissue con

centration was not measured in regressing or recurrent tumors.However, the concentration of dihydrotestosterone in the nucleus also falls after castration but does not decline as rapidlyas does the whole tissue concentration of testosterone; in fact,regressing tumors are characterized by a relatively high concentration of dihydrotestosterone. The concentration of dihydrotestosterone in recurrent tumors is only about 15% of that observedin the initial growing tumors (f test, p < 0.05).

The exclusive nuclear localization of dihydrotestosterone intumors was an unexpected finding since the results of previouswork based on tracer studies with [3H]testosterone indicated

that in short-term pulse studies very little conversion of testos

terone takes place. More than 85% of the intranuclear androgenis in the form of unmetabolized testosterone and only 3% in theform of dihydrotestosterone (4). In contrast, the data in Table 1implies that under physiological conditions, the nucleus selectively concentrates dihydrotestosterone.

Concentration of Binding Sites. The concentrations of binding sites were next measured in cytoplasmic and nuclear fractions of the Shionogi carcinoma in the different proliferativestates. The results of sequential measurements of the cellularcontent of androgen receptors are listed in Table 2. Whereas theconcentration of binding sites in the cytoplasm does not changeduring growth regression and recurrence of the Shionogi carcinoma (f test, p > 0.05), there is a 50 to 70% reduction in themean number of nuclear binding sites for dihydrotestosterone inthe regressing and recurrent tumors (t test, p < 0.01 ).

Nuclear Uptake of Androgens. Since the rate of uptake intothe nucleus varies as much as 47-fold between dependent and

autonomous variants of the Shionogi carcinoma (6), we assumedthat a major decrease in this parameter might identify an autonomous condition of a recurrent tumor. Accordingly, tumor-bear-

3N. Bruchovsky, unpublisheddata.

Table 1Concentrationof androgens in hormone-dependentShionogi carcinoma cells

Tumors weighing approximately 2 g were harvested during stages of growth, regression, and recurrence.The concentrationsof testosterone and dihydrotestosterone in homogenstesof whole tissue and in preparationof purified nuclei were measured by radioimmunoassay.Results are normalized on the basis of DMAcontentof the fractions.

Testosterone (pmol/mgDMA)ConditionGrowth

RegressionRecurrenceWhole

tissue1.9 Â±0.9a(13)"

<0.1(6)0.1Â±0.1 (3)Nucleus0.1

Â±0.1 (6)<0.1 (8)

0.2 Â±0.1 (3)Dihydrotestosterone

(pmol/mgDNA)Whole

tissue1.3

Â±0.5(13)Nucleus1.4Â±0.3(6)0.9 Â±0.2 (8)0.2 Â±0.1 (3)

1Mean Â±S.E.' Numbers in parentheses,number of tumors examined.


684




Table 2Concentration of binding sites in hormone-dependent Shionogi carcinoma cells

Cytoplasmic (cytosol) protein and nuclear extracts were labeled with 3H-andro-gen as described in "Materials and Methods." Samples were analyzed for binding

by Sephadex G-25:G-200 dual-column chromatography. Specific binding wascalculated as the difference between the amount of radioactivity in the receptorpeak in the sample without and the one with competitor.

Binding sites

ConditionGrowth

RegressionRecurrenceCytoplasm

(dpm/mg protein)3500

Â±900a (3)6

4000 Â±1000 (3)2900 Â±500 (3)Nucleus

(dpm/A2Â«o)1500Â±

100(4)500 Â±200 (4)700 Â±200 (4)

" Mean Â±S.E." Numbers in parentheses, number of tumors analyzed.

ing animals were given injections of [1,2-3H]testosterone (200

fiC\), and 30 min later, the amount of radioactivity incorporatedinto purified nuclei was determined. The rate of uptake of andro-gens expressed as dpm per 30 min per nucleus is 3.2 x 10~4 Â±0.6 x 10-4 (n = 7), 7.2 x 1Q-4 Â±1.5 x 1Q-4 (n = 8), and 12.3 x10â„¢4Â±2.1 x 10~4 (n = 8) during tumor growth, regression, and

recurrence, respectively. By comparison, the autonomous C3and C4 lines described later in this report demonstrate meanrates of uptake of only 0.5 to 0.6 x 10~4 dpm/30 min/nucleus.

Thus, recurrent tumors are characterized by a rate of uptakewhich is more typical of androgen-dependent cells than of auton

omous ones.Expression of Glucose-6-Phosphate Dehydrogenase. Since

the Shionogi carcinoma is derived from female mammary tissue,the transplantable malignant cells contain 2 X chromosomes(26). This affords the possibility of examining the tumor formosaicism of the X-linked marker, glucose-6-phosphate dehy-drogenase. A double-enzyme phenotype might be expected in

tumors of multicellular origin although other interpretations wouldhave to be considered (16, 21 ). The results presented in Charts2 and 3 and Table 3 indicate that growing and regressing parentalC1 cells are characterized almost exclusively by a single-enzymephenotype. In contrast, a double-enzyme phenotype occurs with

increasing frequency in the recurrent tumor and in the parentalC1 line after an interval of progression spanning 10 transplantgenerations. However, as can be seen from the tracing in Chart2Â£,the fast-moving form of glucose-6-phosphate dehydrogenase

in the progressed cells is present only in a small amount compared to the quantity detected in recurrent tumors growing ineither male (Chart 2D) or female hosts (Chart 3A). The expressionof the fast-moving form is reduced after the recurrent tumor has

been passaged in a male host (Chart 36). Castration of the malehost tends to amplify the amount of this component slightly(Chart 3C) even though the passaged tumor fails to involute(data not shown). It is to be noted that no evidence of glucose-6-phosphate dehydrogenase polymorphism is observed in thecontrol liver tissue which is characterized by a single-enzyme

phenotype in both female and male animals (Chart 2A). Thus,only in recurrent or progressed tumors is the double-enzyme

phenotype expressed with regularity. Also, manipulation of theandrogen environment appears to have a greater effect on theexpression of the fast-moving variant.

Autonomous C3 Shionogi Carcinoma Cells. The concentration of testosterone in whole tissue of the autonomous C3 tumorin male animals is 0.1 the concentration in cells of the androgen-

dependent parental C1 line (Table 4); only trace amounts of

Chart 2. Expression of glucose-6-phosphate dehydrogenase isoenzymes inShionogi carcinoma cells. Fractional precipitation of cytosol with ammonium sulphate yielded a 30 to 50% fraction containing glucose-6-phosphate dehydrogenase.The protein was pelleted by centrifugation at 18,000 x g for 20 min and frozen at-20Â°. Samples were later thawed, suspended in 50 mw NaCI, and desalted by gel

filtration chromatography using PD-10 Sephadex G-25. About 150 ng of salt-freeprotein were applied to each disc gel made of 7.5% acrylamide:0.18% bis, pH 8.9,and run at 4.0 mA/gel for 2 hr. The gels were stained with an enzyme-specific blueformazan mixture. Scanning of gels was performed in a Gilford spectrophotometerat 550 nm. Glucose-6-phosphate dehydrogenase isoenzymes in: A, normal liver;B, growing tumor; C, regressing tumor; D, recurrent tumor; and Â£,parental C1tumor after 10 transplant generations.

â€¢A /K

Origin

Chart 3. Expression of glucose-6-phosphate dehydrogenase isoenzymes in autonomous Shionogi carcinoma cells. The conditions for this analysis were asdescribed in Chart 2. Glucose-6-phosphate dehydrogenase isoenzymes in: A,recurrent tumor transplanted to female host; B, recurrent tumor after severaltransplant generations in male host; C, recurrent tumor after castration followingpassage in male host; and D, autonomous C3 and C4 variant lines.

testosterone are found in the nuclear compartment. The whole-tissue and nuclear preparations obtained from tumors growingin female animals are virtually devoid of testosterone. The extremely low concentration of testosterone in this class of tumorssuggests either that the tumor is relatively avascular or that thereis deficient transport of testosterone across the plasma membrane. The concentration of dihydrotestosterone in autonomousC3 cells is at least as high if not higher than the concentrationof testosterone, and this androgen appears to be concentratedentirely in the nuclear compartment. In view of the low concentration of testosterone in the whole tissue, especially in that


685




Tab)e3Frequency of single- and double-enzyme phenotypes of glucose-6-phosphate

dehydrogenase in Shionogi carcinoma cells

Frequency

Tissue Condition

Single- Double-No, enzyme enzymeana- pheno- pheno-

lyzed type type Chart

UverParentalCIlineParentalC1lineRecurrent

tumorParentalC1lineRecurrent

tumorAutonomousC3lineAutonomous

C4lineGrowth

phaseRegressionphaseMalehostAfter

10generationsFemalehost7611712410107610210150015114952A282C202Â£3/13030

Table 4

Concentration of androgens in autonomous C3 cells

Tumors weighing approximately 2 g were harvested from mate and female miceand analyzed for whole-tissue and nuclear concentrations of testosterone anddihydrotestosterone by radioimmunoassay. Results are normalized on the basis ofDMA content of the fractions.

Testosterone (pmol/mg DMA)Dihydrotestosterone (pmol/mg

DMA)

Host Whole tissue Nucleus Whole tissue Nucleus

MaleFemale0.2

Â±0.1a (11)"

<0.1 (6)<0.1(6)

<0.1 (6)0.3Â±0.1 (11)

0.2 Â±0.1 (11)0.4Â±0.1 (6)

0.4 Â±0.1 (6)" Mean Â±S.E.b Numbers in parentheses, number of tumors examined.

Tables

Concentration of androgens in autonomous C4 cells

Tumors weighing approximately 2 g were harvested from male and female miceand analyzed for whole-tissue and nuclear concentrations of testosterone anddihydrotestosterone by radioimmunoassay. Results are normalized on the basis ofDNA content of the fractions.

Testosterone (pmol/mg DNA)Dihydrotestosterone (pmol/mg

DNA)

Host Whole tissue Nucleus Whole tissue Nucleus

MaleFemale0.7

Â±0.3" (5)6

<0.1 (4)<0.1(4)

<0.1 (4)3.7Â±0.6 (5)

0.7 Â±0.4 (4)2.8Â±1.0 (4)

0.3 Â±0.1 (4)" Mean Â±S.E.6 Numbers in parentheses, number of tumors examined.

obtained from female animals, it is interesting to note that thenuclear concentration of dihydrotestosterone is equivalent intumors from animals of both sexes.

A limited number of experiments was also carried out toconfirm the results of previous studies on the concentration ofbinding sites and nuclear uptake of androgens in autonomouscells. Analysis of C3 tumors in female hosts yielded a value of400 Â±100 (n = 4) dpm/mg protein for the quantity of cytoplasmicreceptor and a rate of uptake of androgens of 0.6 x 10~4 Â±0.1x 10~4 (n = 5) dpm/30 min/nucleus. Both parameters are con

siderably lower than the corresponding indices in the dependenttumors during phases of growth, regression, and recurrence.

As shown in Chart 3D and Table 3, glucose-6-phosphatedehydrogenase is expressed as a double-enzyme phenotype in

90% of the tumors.Autonomous C4 Shionogi Carcinoma Cells. The concentra

tion of testosterone Â¡nwhole tissue of the autonomous C4 tumoris 50% less than that in tumors of the androgen-dependent

parental C1 line (Table 5). Again, very little testosterone is

recovered Â¡nthe nuclear compartment and in the tumors removedfrom female hosts. A striking accumulation of dihydrotestosterone is observed in the autonomous C4 tumors growing in malemice. In this instance, the nuclei concentrate at least as muchdihydrotestosterone as do those from the parental androgen-dependent C1 tumors, with no statistical difference in levels (ftest, p > 0.05). In contrast, the concentration of dihydrotestosterone in autonomous C4 cells in female mice resembles that ofthe autonomous C3 cells.

Consistent with the elevated concentration of dihydrotestosterone in the nuclear compartment of autonomous C4 tumors inmale hosts, the concentration of nuclear binding sites at 1900 Â±500 (n = 3) is identical (f test, p > 0.05) to the value observed inthe androgen-dependent parental C1 line (Table 2).

The quantity of cytoplasmic receptor in tumors propagated infemale hosts, 400 Â±100 (n = 8) dpm/mg protein is not different

than the amount in autonomous C3 cells. Neither is the rate ofuptake of androgens at 0.5 x 10~4 Â±0.1 x 10~4 (n = 10) dpm/

30 min/nucleus significantly different (f test, p > 0.05). Thesefindings confirm that both autonomous C3 and C4 tumors aredeficient in cytoplasmic binding and nuclear uptake.

Glucose-6-phosphate dehydrogenase is expressed as a double-enzyme phenotype in only 50% of the autonomous C4 tu

mors (Chart 3D; Table 3).

DISCUSSION

The results of this investigation provide new information aboutthe androgen dependence of the Shionogi carcinoma. The proliferation of the tumor is stimulated by both testosterone anddihydrotestosterone (2, 20), and on the basis of these and otherstudies (25, 28, 34, 39), it has been inferred that dihydrotestosterone is responsible for the mitotic activity induced by testosterone. However, the Shionogi carcinoma contains little 5a-reductase, and consequently, in experiments in which the metabolism of exogenous radioactive testosterone has been studied either in vivo (2, 4, 39) or in vitro (20), the formation ofdihydrotestosterone has been slight. The data reported in Table1 indicate that under steady state conditions, as opposed toacute short-term experiments, selective accumulation of dihydro

testosterone by the nucleus takes place. The exclusive nuclearlocalization of dihydrotestosterone Â¡ndependent Shionogi carcinoma cells directly confirms that dihydrotestosterone is theactive intracellular androgen in this target tissue (7).

Progression of the Shionogi carcinoma through successivephases of growth, regression, and relapsing disease promotesthe emergence of a recurrent tumor the phenotype of whichdoes not differ markedly from that of the parental tumor exceptfor the acquirement of a fast-moving electrophoretic variant ofglucose-6-phosphate dehydrogenase. Although neither testos

terone nor dihydrotestosterone reaccumulate in the recurrenttumor to normal levels, several phenotypic indices of hormonalresponsiveness persist. These include the presence of cytoplasmic androgen receptor, a reduced but nevertheless detectable concentration of binding sites in the nucleus, and the abilityto incorporate androgens into the nucleus at a rapid rate. Comparison of the phenotype of recurrent tumors to those of severalvariant lines of autonomous Shionogi carcinoma cells (5) suggests that the recurrent growth corresponds to the C2 variantline of autonomous tumors. The chief alteration observed in this


686




variant is a reduction in the concentration of nuclear bindingsites. We have not determined whether this is a reversiblechange, and it remains possible that the reduction in the amountof nuclear binding is simply a reflection of androgen withdrawal.This would be consistent with the observation that the recurrenttumor is composed in part of cells which respond to androgenicstimulation, as the growth rate of this tumor is accelerated 3- to4-fold with androgen replacement (Chart IB).

The phenotype of the recurrent tumor differs markedly fromthe phenotypes of the C3 and C4 autonomous tumors. C3autonomous cells, in addition to lacking cytoplasmic receptor,nuclear binding sites, and ability to take up androgens into thenucleus, are characterized by a deficiency of androgen retentionby whole tissue. This would be consistent with a change inplasma membrane composition affecting permeability to steroids(19). Another unusual feature of the C3 variant is that it almostinvariably displays the glucose-6-phosphate dehydrogenase double-enzyme phenotype.

The phenotype of the C4 autonomous tumor is different fromthat of the recurrent tumor and the C3 autonomous tumor. TheC4 variant retains the ability to concentrate modest levels oftestosterone in whole tissue; in male animals, the concentrationof testosterone appears to be sufficient to afford a relatively highconcentration of dihydrotestosterone in the nucleus. Parallelingthe latter, the concentration of nuclear binding sites is the sameas that observed in the parental androgen-dependent tumor. On

the other hand, the concentration of cytoplasmic receptor andthe rate of nuclear uptake of androgens are noticeably decreased. The C4 autonomous tumor also differs from the C3variant in that the glucose-6-phosphate dehydrogenase double-

enzyme phenotype is not expressed as frequently.Multiple forms of glucose-6-phosphate dehydrogenase have

been observed in normal, preneoplastic, and carcinomatousmammary tissue of mice by Hilf er al. (21). These investigatorsnoted that pregnancy and lactation caused a shift in the electro-phoretic pattern with increased formation of a fast-moving variantat the expense of a slow-moving variant. Whether the variant

forms in the Shionogi carcinoma correspond to any of the formsdescribed by Hilf ef a/. (21) is difficult to judge since the hormonaland physiological conditions are considerably different. TheShionogi carcinoma is dependent on androgens rather than onestrogens, and the multiple-enzyme pattern is not seen in theunprogressed parental C1 line. It is of interest that glucose-6-

phosphate dehydrogenase activity can be inhibited by androgenspossessing a 5Â«-androstan configuration (32, 33), but owing to

the large K, values, i.e., in the /Â¿Mrange (32), it is doubtful thatthe relatively small intracellular concentration of dihydrotestosterone could directly influence glucose-6-phosphate dehydrogen

ase activity in the Shionogi carcinoma.Shifts in the pattern of molecular forms of glucose-6-phosphate

dehydrogenase may occur in some tissues owing to alterationsin the concentration of glutathione (38). However, glutathionereducÃase activity per se does not appear to determine themolecular form of glucose-6-phosphate dehydrogenase in a

changing hormonal environment (21). Furthermore, since thefast-moving variant of the Shionogi carcinoma appears during

progression (after 10 transplant generations) of the parental C1line (Table 3) in the absence of any hormonal change, it is unlikelythat our results can be explained by hormone-dependent fluctua

tions in glutathione levels.

Expression of the fast-moving variant of glucose-6-phosphate

dehydrogenase occurs more often under conditions which mightbe expected to favor the enrichment of androgen-insensitive

cells. This effect is observed in tumors that recur in castratedmale hosts (Chart 2D), in recurrent tumors passaged in femalehosts (Chart 3/4), and in parental C1 tumors after several transplant generations (Chart 2E). Expression of the fast-moving

variant is also a constant feature of the autonomous C3 line butis less certain in the autonomous C4 line (Table 3). The latterdifference is harmonious with the idea that the autonomous linesmay contain variable numbers of at least 2 resistant cell populations, one characterized by the slow-moving variant of glucose-6-phosphate dehydrogenase, the other by the fast-moving variant. The possibility that the fast-moving variant is a marker for

resistant cells is also supported by the observation that underconditions which would favor the growth of androgen-stimulated

cells, a relative reduction in the amount of the variant enzymetakes place (Chart 38).

Although it is tempting to attribute the foregoing observationsto X-linked mosaicism of the glucose-6-phosphate dehydrogen

ase marker, this interpretation is complicated by the absence ofenzyme polymorphism in female liver cells (Chart 2A) and thefact that no autonomous variant lines expressing only the fast-

moving variant have been isolated. As discussed by Fialkow(16), the interpretation of double-enzyme phenotypes would have

to take into consideration the possibility of normal cell admixture,the ratio of isoenzymes in the normal tissue from which theneoplasm originates, and activity of both X chromosomes in asingle cell. Generally speaking, our results would also be compatible with the latter possibility assuming that the activity of oneof the X chromosomes is normally suppressed in the presenceof androgens during the initial growth phase of the tumor. Whenproliferation of the recurrent tumor resumes in the absence ofandrogens, the activity of both X chromosomes would becomeapparent. Our failure to detect the double-enzyme phenotype inandrogen-withdrawn regressing cells might be explained by the

rapid decline in synthesis of scarce and abundant classes ofmRNA sequences during tumor involution.4 Whatever the basis

for the expression of the fast-moving variant of glucose-6-phos-

phate dehydrogenase, it is clear that the isoenzyme is stronglyassociated with the emergence of androgen-insensitive cells.

Thus, early progression of the androgen-dependent Shionogi

carcinoma cells is indicated by a drop in the number of nuclearandrogen-binding sites, the appearance of the fast-moving variant of glucose-6-phosphate dehydrogenase, and the loss of

responsiveness to androgen withdrawal. Such changes wouldbe consistent with a multicellular origin of the carcinoma andclonal selection, or alternatively, the irreversible adaptation ofandrogen-dependent cells associated with the expression of both

X chromosomes.

ACKNOWLEDGMENTS

We thank Cynthia Wells for assistance in the preparationof the manuscript.

REFERENCES

1. Bradford, M. A rapid and sensitive method for the quantitation of microgramquantities of protein utilizing the principle of protein-dye binding. Anal.

4M. Giles,N. Bruchovsky, and P. S. Rennie.Changesingene expressionrelatedto regression of the Shionogi mouse mammary carcinoma, submitted for publication.


687




Biochem., 72: 248-254,1976.2. Bruchovsky, N. The metabolism of testosterone and dihydrotestosterone in an

androgen-dependent tumour. A possible correlation between dihydrotestosterone and tumour growth in vivo. Biochem. J., 727: 561-575, 1972.

3. Bruchovsky, N., and Goldie, J. H. Basis for the use of drug and hormonecombinations in the treatment of endocrine-related cancer. In: N. Bruchovskyand J. H. Goldie (eds.), Drug and Hormone Resistance in Neoplasia, ClinicalConcepts, Vol. 2, pp. 129-162. Boca Raton, FL: CRC Press, 1983.

4. Bruchovsky, N., and Meakin, J. W. The metabolism and binding of testosteronein androgen-dependent and autonomous transplantable mouse mammarytumors. Cancer Res., 33: 1689-1695, 1973.

5. Bruchovsky, N., and Rennie, P. S. Classification of dependent and autonomousvariants of Shionogi mammary carcinoma based on heterogeneous patternsof androgen binding. Cell, 13: 273-280,1978.

6. Bruchovsky, N., Sutherland, D. J., Meakin, J. W., and Minesita, T. Androgenreceptors: relationship to growth response and to intracellular androgen transplant in nine variant lines of the Shionogi mouse mammary carcinoma. Biochim.Biophys. Acta, 387: 61-71, 1975.

7 Bruchovsky, N., and Wilson, J. D. The conversion of testosterone to 5a-androstan-170-ol-3-one by rat prostate in vivo and in vitro. J. Biol. Chem.,243:2012-2021, 1968.

8. Burton, K. A study of the conditions and mechanism of the diphenylaminereaction for the colorimetrie estimation of deoxyribonucleic acid. Biochem. J.,62:315-323, 1956.

9. Bush, R. S. Malignancies of the Ovary, Uterus and Cervix, pp. 101-141.London: Edward Arnold Publishers, 1979.

10. Callaway, T. W., Bruchovsky, N., Rennie, P. S., and Comeau, T. Mechanismsof action of androgens and antiandrogens on translocation of cytoplasmicandrogen receptor and nuclear abundance of dihydrotestosterone. The Prostate. 3: 599-610, 1982.

11. Coffey, D. S., and Isaacs, J. T. Experimental concepts in the design of newtreatments for human prostatic cancer. In: D. S. Coffey and J. T. Isaacs (eds.),Prostate Cancer, pp. 233-259. Geneva: International Union Against Cancer,

1979.12. Coppin, C. M. L. The treatment of advanced prostatic cancer with drugs and

hormones. In: N. Bruchovsky and J. H. Goldie (eds.). Drug and HormoneResistance in Neoplasia, Clinical Concepts, Vol. 2, pp. 57-93. Boca Raton,

FL: CRC Press, 1983.13. Dao, M. L., Watson, J., Delaney, R., and Johnson, B. C. Purification of a new

high activity form of glucose-6-phosphate dehydrogenase from rat liver andeffect of enzyme inactivation on its immunochemical reactivity. J. Biol. Chem.,254:9441-9447, 1979.

14. De Larminat, M.-A., Rennie, P. S., and Bruchovsky, N. Radioimmunoassaymeasurements of nuclear dihydrotestosterone in rat prostate: relationship toandrogen receptors and androgen regulated responses. Biochem. J., 200:465-474,1981.

15. Falvo, R. E., and Nalbandov, A. V. Radioimmunoassay of peripheral plasmatestosterone in males from eight species using a specific antibody withoutchromatography. Endocrinology, 95. 1466-1468,1974.

16. Fialkow. P. J. Clonal origin of human tumors. Biochim. Biophys. Acta, 458:283-321,1976.

17. Foulds, L. Neoplasie Development 1, pp. 69-76. New York: Academic Press,

Inc., 1969.18. Gabriel, O. Analytical disc gel electrophoresis. Methods Enzymol., 22: 565-

578, 1971.Giorgi, E. P. The transport of steroid hormones into animal cells. Int. Rev.19.

20.

Cytol., 65, 49-115,1980.Gordon, J., Smith, J. A., and King, R. J. B. Metabolism and binding of

androgens by mouse mammary tumour cells in culture. Mol. Cell. Endocr. 7-259-270, 1974.

21. Hilf, R., Ickowicz, R., Bartley, J. C., and Abraham, S. Multiple molecular formsof glucose-6-phosphate dehydrogenase in normal, preneoplastic, and neoplas-tic mammary tissues of mice. Cancer Res., 35: 2109-2116, 1975.

22. Hunter, L. Glucose-6-phosphate dehydrogenase isoenzymes in cultured human cell lines: separation by isoelectric focussing. Anal. Biochem., 707: 78-87, 1980.

23. Isaacs, J. T. Hormonally responsive versus unresponsive progression of prostatic cancer to antiandrogen therapy as studied with the Dunning R-3327-ATand -G rat adenocarcinomas. Cancer Res., 42: 5010-5014,1982.

24. Isaacs, J. T. Cellular factors in the development of resistance to hormonaltherapy. In: N. Bruchovsky and J. H. Goldie (eds.). Drug and HormoneResistance in Neoplasia, Basic Concepts, Vol. 1, pp. 139-156. Boca Raton,FL: CRC Press, 1982.

25. King, R. J. B. Studies on the regulation of cell proliferation in culture bysteroids. In: J. Dumont and J. Nunez (eds.). Hormones and Cell Regulation 2,pp. 15-36. Amsterdam: Elsevier/North-Holland Biomedicai Press, 1978.

26. Kitamura, Y., Okamoto, S., Hayata, I., Uchida, N., Yamaguchi, K., and Matsu-moto, K. Development of androgen-dependent spindle cell tumors from androgen-dependent medullary Shionogi carcinoma 115 in androgen-depleted nudemice. Cancer Res., 39: 4713-4719,1979.

27. Kitamura, Y., Okamoto, S., Uchida, N., Yamaguchi, K., and Matsumoto, K.Effect on androgen depletion on growth and androgen dependency of Shionogicarcinoma 115. Cancer Res., 38: 4711-4716,1978.

28. Kitamura, Y., Uchida, N., Odaguchi, K., Yamaguchi, K., Okamoto, S., andMatsumoto, K. Insignificance of pituitary for growth of androgen-dependentmouse mammary tumor. J. Steroid Biochem., 73: 333-337, 1980.

29. Lucas, L. A., and Abraham, G. E. Radioimmunoassay of testosterone in murineplasma. Anal. Lett., 5:773-783,1972.

30. Menon, M., and Walsh, P.C., Hormonal therapy for prostatic cancer. In: G. P.Murphy (ed.), Prostatic Cancer, pp. 175-200. Littleton, MA: PSG PublishingCo., Inc., 1979.

31. Noble, R. L. Hormonal control of growth and progression in tumors of Nb ratsand a theory of action. Cancer Res., 37: 82-94,1977.

32. Raineri, R., and Levy, H. R. On the specificity of steroid interaction withmammary glucose-6-phosphate dehydrogenase. Biochemistry, 9: 2233-22431970.

33. Shreve, D. S., and Levy, H. R. Glucose-6-phosphate dehydrogenase fromlactating rat mammary gland and R3230AC adenocarcinoma. Enzyme (Basel),24:48-53,1979.

34. Sica, G., Longo, P., Ranelletti, F. O., and lacobelli, S. Androgen receptorspecificity and growth response in Shionogi carcinoma cells. J. SteroidBiochem., 74: 23-28, 1981.

35. Simpson-Herren, L., and Lloyd, H. H. Kinetic parameters and growth curvesfor experimental tumor systems. Cancer Chemother. Rep., 54:143-174,1970.

36. Sluyser, M. The emergence of hormone-independent cells in hormone-dependent breast cancer. In: C. M. McGrath, M. G. Brennan, and M. A. Rich (eds.),Cell Biology of Cancer, pp. 173-187. New York: Academic Press, Inc., 1980.

37. Stoll, B. A. The basis of endocrine therapy. In: B. A. Stoll (ed.). EndocrineTherapy in Malignant Disease, pp. 111-234. London: W. B. Saunders Co.,1972.

38. Watanabe, A., Taketa, K., and Kosaka, K. Glutathione-dependent interconversion of microheterogeneous forms of glucose-6-phosphate dehydrogenase inrat liver. J. Biochem. (Tokyo), 72: 695-701,1972.

39. Yamaguchi, K., Minesita, T., Kasai, H., Kurachi, K., and Matsumoto K. Detection of 17/3-hydroxy-5Â«-3-one-3H following administration of testosterone-3H inthe androgen-dependent tumor, Shionogi carcinoma 115. Steroids, 77: 345-355, 1971.

Fig. 1. Histology of androgen-dependent Shionogi carcinoma during growth phase. H & E, x 900.

Fig. 2. Histology of androgen-dependent Shionogi carcinoma during regression phase. H & E, x 900.

Fig. 3. Histology of recurrent Shionogi carcinoma. Cells close to central blood vessel show typical appearance. More distal cells are spindle shaped. H & E, x 350.

Fig. 4. Spindle-shaped cells under higher magnification, x 900.

Fig. 5. Histology of autonomous C3 variant of Shionogi carcinoma. H & E, x 900.

Fig. 6. Histology of autonomous C4 variant of Shionogi carcinoma. H & E, x 900.


688




1' >

.

^


689



1985;45:682-689. Cancer Res Nicholas Bruchovsky, Paul S. Rennie, Navdeep Otal, et al. and Progression to Hormonal IndependenceMammary Carcinoma during Growth, Involution, Recurrence, Variability of Androgen-related Phenotypes in the Shionogi

Updated version

http://cancerres.aacrjournals.org/content/45/2/682

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/45/2/682To request permission to re-use all or part of this article, use this link



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

mailto:[email protected]



variability of androgen-related phenotypes in the shionogi ... · nucleus. however, relapse is...

Documents