Murine Mammary Gland Hormone-independent Cell

Secretion of a Cysteine Proteinase from a Hormone-independent Cell Population of Cultured Explants of Murine Mammary Gland Anneliese D. Recklies, Chantal White, Jane Mitchell, et al. Cancer Res 1985;45:2294-2300. Published online May 1, 1985.

Updated Version

Citing Articles

E-mail alerts Reprints and Subscriptions Permissions

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/45/5/2294

This article has been cited by 2 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/45/5/2294#related-urls

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].

Downloaded from cancerres.aacrjournals.org on July 11, 2011 Copyright © 1985 American Association for Cancer Research

[CANCER RESEARCH

45, 2294-2301,

May 1985]

Secretion of a Cysteine Proteinase from a Hormone-independent Population of Cultured Expiants of Murine Mammary Gland1

Cell

Anneliese D. Recklies,2 Chantal White, Jane Mitchell, and A. Robin Poole Department of Experimental Surgery, McGill University, and Joint Diseases Laboratory, Shriners Hospital for Crippled Children, Montreal, Quebec H3G 1A6, Canada

ABSTRACT Cultured expiants of mouse mammary gland were investigated for their capacity to secrete a cathepsin B-like cysteine proteinase. This enzyme had been shown previously to be secreted excessively from cultured expiants of human breast tumors and spontaneous mouse mammary carcinomas. We now show that secretion is also observed from cultured expiants of mammary gland. Lactating tissue and tissues obtained from mid- and latepregnant mice were found to secrete the cysteine proteinase at very high rates, but secretion was also detectable from expiants of virgin mammary glands and from tissues obtained from retired breeders. However, in all cases, it was found that secretion was greatest from expiants maintained in hormone-free medium and did not depend on the maintenance of normal mammary gland function. Secretion was greatly reduced in the presence of the lactogenic hormone combination of insulin, prolactin, and hydrocortisone, and this suppression was found to be due to hydrocortisone. Insulin and prolactin, while resulting in better tissue maintenance in culture, had no effect on the secretion. Enzyme release was reversibly inhibited by cycloheximide and required the presence of viable tissue, ruling out the possiblity that the accumulation of enzyme activity in the culture medium is due to dying cells. The presence of metabolically active cells in expiants cultured in the absence of hormones was also demonstrated by the incorporation of radiolabeled precursors into protein and DMA. DNA synthesis in cultured expiants of lactating tissue was not stimulated by the addition of hormones. Histological studies revealed that, while large areas of the expiants showed severe degeneration after culture without added hormones, clusters of cells persisted which displayed a relatively high mitotic activity and which showed a lack of normal epithelial organization. These observations suggest the presence of a hormoneindependent cell population in the mammary gland, which se cretes the stable cysteine proteinase. The secretion process itself, however, is inhibited by corticosteroids.

of mammary gland tissue to growth or lactation-promoting stimuli has been shown to parallel the in vivo morphological and bio chemical changes during pregnancy and lactation (1,10). Hence, this culture system provides a good model for the study of mammary gland biology. We have shown previously that both human adenocarcinomas of the breast and spontaneous mammary carcinomas in C3H mice secreted excessive amounts of a cysteine proteinase in organ culture (17, 20, 21). The proteinase is enzymically indistin guishable from the lysosomal enzyme cathepsin B, but it has a larger molecular weight and more acidic isoelectric focusing pattern compared to cathepsin B; it is also stable at neutral pH (16). The enzyme released from human tumors was shown to cross-react with an antiserum to lysosomal cathepsin B (19), suggesting that it might be a precursor form of the lysosomal enzyme. While cysteine proteinase secretion was elevated in malignant tumors as compared to normal, age-matched human breast tissue and benign fibroadenomas of the breast, no cor relation was observed with the staging of the human tumors. As human tumors are known to exhibit various degrees of mammary epithelial differentiation and can synthesize mammary glandspecific proteins such as casein and a-lactalbumin (9, 27), the question arose as to whether or not secretion of the cysteine proteinase was related to some aspect of mammary gland func tion, such as the processing and secretion of milk proteins. Alternatively, the increased secretion of the cathepsin B-like cysteine proteinase could be characteristic of mammary tumor cells and perhaps also of preneoplastic cells in the mammary epithelium. To investigate this question, the mouse mammary gland system was chosen, since the organ culture of mouse mammary gland has been well documented, and mammary carcinomas from C3H mice, like human breast tumors, also secrete the cysteine proteinase (20, 21).

MATERIALS AND METHODS Materials. A colony of BALB/c mice was established with a breeding pair obtained from The Jackson Laboratory (Bar Harbor, ME) and main tained by brother-sister mating. All mice used in this study were from

INTRODUCTION Mammary glands undergo continuous cycles of growth, differ entiation, and involution during the reproductive life of a mammal. In humans, as well as in some strains of mice, they are also prone to development of abnormal growth patterns and malig nant tumors. Cultured mammary glands from virgin and hor mone-primed mice have been used to study the hormonal control of growth and differentiation of the mammary epithelium and the development of secretory functions (30). The in vitro response 1This work was supported by the Shriners of North America and the National

this colony. Culture medium (DMEM)3 was purchased from Grand Island Biological Co. (Grand Island, NY). Porcine insulin, ovine prolactin, hydrocortisone, aldosterone, estradici, bovine serum albumin, and /3-galactoside-A/-methylumbelliferone were from Sigma Chemical Co. (St. Louis, MO). The cathepsin B substrate BAÃ‘A was obtained from Aldrich Chemicals (Montreal, Quebec, Canada). CBZ-Arg-Arg-NA was synthesized by Dr. J. Mort (this laboratory) according to the method of Knight (14) with starting materials obtained from Sigma. [3H]Thymidine and [3H]leucine were from New England Nuclear (Montreal, Quebec, Canada). All other 'The abbreviations used are: DMEM, Dulbecco's modified essential medium; BAÃ‘A, benzoyl-arginine-/3-naphthylamide; CBZ-, carbobenzyloxy; NA, 0-naphthylamine; TCA, trichloroacetic acid.

Cancer Institute of Canada. 2 To whom requests for reprints should be addressed. Received 8/15/84; revised 12/28/84; accepted 1/9/85.

CANCER RESEARCH

VOL. 45 MAY 1985

2294


CYSTEINE

PROTEINASE

SECRETION

materials were of the purest grade available. Mammary Gland Culture. Pseudopregnant virgin mice were produced by daily injections of estrogen and progesterone over a 9-day period as described by Ichinose and Nandi (10), and mammary glands were re moved on Day 10. Mammary gland tissue was also removed from timed pregnant or lactating animals. Generally, the second, third, and fourth pairs of glands were dissected aseptically, pooled, and cut into small pieces. Expiants were established on expanded metal grids in small Retri dishes as described for the culture of human and mouse tumor tissues (20, 21). The medium used was DMEM containing glucose (4.5 mg/ml) and supplemented with porcine insulin (1 Mg/ml) and antibiotics (50 units penicillin G and 50 ng streptomycin per ml, respectively). Individual hormones were added as indicated. Expiants were maintained in a humidified incubator at 37Â°C in an atmosphere of 95% air and 5% CO2 or in an atmosphere of 95% O2 and 5% CO2 in Mclntosh-Fildes jars. Growth medium was changed on Day 1 of the culture period and either every 24 or 48 h thereafter, as indicated. Culture medium was stored frozen until used for enzyme analysis. At the end of the culture period, expiants were collected and stored frozen for DNA determinations. Enzyme Assays. Cysteine proteinase activity was determined rou tinely at pH 5.5 in the presence of 2 mw cysteine, using BAÃ‘A as substrate as described (20). Incubation was overnight at 36Â°C. When higher sensitivity and shorter incubation times were required, CBZ-ArgArg-NA was used as substrate, since the secreted cysteine proteinase, like cathepsin B, has an approximate 20-fold-higher affinity for this substrate (16). Incubation with this substrate was at 40Â°C. Cysteine proteinase activity is expressed as nmol NA released per unit time. Cathepsin D activity was assayed at pH 3.8, using acid-denatured hemoglobin as substrate (18). Activity is expressed as nmol tyrosine equivalents released per unit time. J-Galactosidase was assayed fluorometrically with ÃŸ-galactoside-N-methylumbelliferone as substrate and incubation at room temperature (2). DNA Determination. Tissue DNA was determined according to the method of Kissane and Robbins (12), using the modifications described by Thomas and Farquhar (28). Tissue was prepared for DNA determi nation as described previously (21). Incorporation of [3H]Thymidine into Cultured Expiants. Mammary

FROM CULTURED MAMMARY

GLAND

culture system used for mouse mammary carcinomas (20), ca thepsin B-like cysteine proteinase activity could be detected in the medium (Chart 1). Activity accumulated in the culture medium in a relatively linear fashion. In these experiments, medium collected on Day 1 was not assayed for activity. Similar results were obtained when mammary glands from hormone-primed mice were cultured, but the amount of activity released per /Â¿g of tissue DNA was significantly less (Chart 1). The cysteine proteinase activity measured in the culture me dium was characterized as cathepsin B like and had similar physical properties to the stable cathepsin B-like cysteine pro teinase secreted from human and mouse mammary tumors (20, 21). This work is described in detail in an accompanying paper. To determine whether the cysteine proteinase secretion was a feature of lactating mammary gland, expiants from various stages of mammary growth and development were studied. The capacity to secrete this enzyme was present in mammary glands at all stages tested. The highest levels of activity were obtained from expiants of mice in early lactation and late pregnancy (Table 1). Lower levels were secreted from mammary glands from midpregnant mice (about 60% of the rate observed in latepregnant glands) and from glands late in lactation (about 40% of the secretion rate of early lactating glands). Cysteine proteinase was also secreted at low levels from mammary gland expiants of virgin mice. Secretion from glands obtained from retired breed ers was about twice as high as from virgin animals and was the same as that observed with glands from hormone-primed ani mals. To determine whether the secretion of this cysteine proteinase

glands were dissected from lactating mice 5 days postpartum and incorporation of [3H]thymidine was determined using either freshly estab lished expiants (fresh tissue) or expiants cultured for 4 days in the following media: (a) growth medium, containing insulin (1 Mg/ml), prolactin (1 Mg/ml), hydrocortisone (1 Mg/ml), estrogen (1 ng/ml), and progesterone (1 Mg/ml); (o) lactogenic medium with insulin, hydrocortisone, and prolac tin at the above concentrations; and (c) regression medium containing insulin and aldosterone (1 Mg/ml). Control medium contained only insulin. The volume of medium was 1.5 ml/dish. Medium was changed on Day 2 of the culture period and on Day 4 prior to addition of labeled precursors. For determination of nucleic acid synthesis, 5 nC'\ of [3H]thymidine were added to triplicate dishes, and after 4 h of incubation, the medium was removed, the expiants were washed several times in cold phosphate-buffered saline, and the tissue was homogenized in H2O, using a Tissumizer homogenizer, and precipitated with 10% cold TCA. Washed TCA precipitates were hydrolyzed in 0.6 N perchloric acid, and incorporated radioactivity and total DNA content were determined in aliquots of the hydrolysate. Histological Examination. For histological examination, expiants were established for culture on Millipore filters supported by stainless steel grids to maintain tissue orientation during fixation. At the indicated times, expiants were fixed for 30 min in acidified Zenker's fixative, followed by nonacidified Zenker's fixative for 16 h. Tissue was embedded, cut into 5-Mm-thick sections, and stained with hematoxylin and eosin.

RESULTS When expiants of mammary glands from lactating mice were cultured in medium supplemented with insulin only, similar to the CANCER RESEARCH

DAYS

IN CULTURE

Chart 1. Accumulation of cysteine proteinase activity in culture medium from expiants of lactating (O) and hormone-primed (â€¢)mammary gland tissue. Virgin BALB/c mice were primed by daily injections of estrogen and progesterone for 9 days. Mammary glands were dissected on Day 10 from these mice and 5 days postpartum from lactating mice, and tissue from the second and third pair of glands was pooled. Expiants were cultured in DMEM supplemented with insulin (1 ^g/ml). Medium was changed on Day 1 of the culture period and every 2 days thereafter. Cysteine proteinase activity was determined using BAÃ‘A as substrate. Points, averages from 8 expiants; bars, SD. Activity is expressed as nmol NA released per figDNAper 16h.

VOL. 45 MAY 1985

2295


CYSTEINE

PROTEINASE

SECRETION


Table 2 DNA content of cultured expiants of lactating mammary gland

Table 1 Secret/on rafes oÃ-cysteine prore/rase from BALB/c mammary gland expiants at various stages ot growth Secretion of cysteine proteinase is shown from mammary gland expiants cul tured in the presence of insulin (5 tig/mi). Medium was changed every 48 h after Day 1 of the culture period, and cysteine proteinase activity was determined using BAÃ‘A as substrate. The number of animals used for each determination (n) is given. For each animal, 16 dishes from pooled mammary gland tissue were established. Activity is expressed as nmol substrate hydrolyzed Â¡n16 h. Cysteine proteinase se creted in 48 h/jig DNA

Mammary gland status Lactating (n = 4) (2-5 days postpartum) Lactating (n = 3) (25 days postpartum) Late pregnant (n = 2) (16 and 18 days)

Expiants were established from mammary gland tissue dissected from lactating BALB/c mice 4 days postpartum. Organ culture was for 9 days, after which time, tissue was harvested, weighed, and stored frozen for DNA analysis. Hormones were added from Day 0 of the culture period. Growth medium contained insulin (1 iig/ml), prolactin (1 jjg/ml), aldosterone (1 /ig/ml), estradici (1 ng/ml), and proges terone (1 /jg/ml). Lactogenic medium contained insulin (1 ng/m\), prolactin (1 /*g/ ml), and hydrocortisone (1 /Â¿g/ml).Regression medium contained insulin (1 (Â¿g/ml) and aldosterone (1 /*g/ml). DNA content of uncultured tissue was determined by randomly selecting 4 expiant dishes in each experiment.

14.54 Â±3.05a

Hormone primed (n = 10)

3.05 Â±1.14

Virgin (n = 4)

1.59 Â±0.62

Retiredbreeders(n = 5)

3.00 Â±0.94

17.9 Â±2.4Â°

Control medium Growth medium3 Lactogenic medium3 Regression medium0 Insulin (1 /ig/ml)c Insulin and prolactin (1 /ig/ml)c

13.81 Â±1.99 9.80 Â±2.48

DNA content (jjg/expiant)

Culture conditions

6.54 Â±1.45

Midpregnant (n = 2) (10 and 12 days)

GLAND

47.4 Â±7.9 35.6 Â±2.3 28.3 Â±5.2 22.4 Â±4.3 22.7 Â±4.9 41.1 Â±8.1 44.4 Â±6.4 71.1 Â±10.2

Insulin and hydrocortisone (0.1 ng/m\f Insulin and hydrocortisone (1 >ig/ml)c Uncultured tissuec

" Data are from 4 separate experiments. In each experiment, 4 dishes were set

a Average Â±SD of secretion rates from the activities observed on Days 3, 5, and 7 of the culture period.

up for each culture condition. 6 Mean Â±SD. c Data are from 2 separate experiments with 4 dishes set up for each culture condition.

varies between the different glands, cultures from each of the second, third, and fourth pairs of glands (second and third thoracic glands and first lumbar) from lactating mice (5 days postpartum) were established separately. No significant differ ences in the secretion rates were observed. Thus, for all further experiments, these glands were pooled. In the culture system used, tissue expiants are maintained at the medium-air-CO2 interface. When mammary gland expiants were cultured submerged in medium, no enzyme activity was detected in the medium even after several days of culture. Thus, as is generally found for successful culture of mammary glands (10), the tissue has to be kept at the air-medium interface. Culture of expiants in 95% O2-5% CO2 did not affect the proteinase secretion or its modulation by hormones. Demonstration of Active Secretion of Cysteine Proteinase and Effects of Hormones. Since the culture of mammary glands in the absence of those hormones normally required for the maintenance of their structure and function results in severe tissue regression and cell death during culture (10), it was necessary to establish whether the proteinase was actively secreted from the tissue and did not simply leak from dying or dead cells. Similar to results reported for human breast carci noma (21) and mouse mammary carcinoma expiants (20), accu mulation of activity in the tissue culture medium required the presence of viable tissue and was inhibited by cycloheximide. Thus, while the culture conditions used result in regression of the expiants, the observed cysteine proteinase release is a true secretion that requires the continued presence of a viable and synthetically active cell population. To determine whether secretion was hormone dependent, expiants were cultured in the presence of combinations of hor mones generally used to promote growth (insulin, prolactin, hydrocortisone, estrogen, and progesterone), lactation (insulin, prolactin, and hydrocortisone), or regression (insulin and aldos terone) of cultured mammary gland (1,10). In analyzing secretion data, a major problem is presented by the varying degrees of tissue survival and, thus, the DNA content of tissues cultured under these various hormone conditions (Table 2). Lactating CANCER RESEARCH

:

Contro! GrowlÂ»

11

Day 3

Day 5

Day 3

Day 5

Chart 2. Effects of various hormone combinations on cysteine proteinase se cretion from expiants of lactating mammary gland. The following hormone additions were made to the culture medium from Day 0: control with DMEM alone; growth medium [insulin (1 ^g/ml). prolactin (1 Â¿ig/ml),hydrocortisone (1 jiQ/ml), progester one (1 jig/ml), and estradiol (1 ng/ml)j; lactogenic medium [insulin (1 ng/m\), prolactin (1 Â»ig/ml),and hydrocortisone (1 ijg/ml)]; regression medium [insulin (1 jig/ml) and aldosterone (1 jig/ml)]. For each hormone combination, 4 expiant dishes were used. Medium was changed on Day 1 and every other day thereafter. Cysteine proteinase activity was measured with BAÃ‘A as substrate. Data are expressed as the percentage of the cysteine proteinase secreted by control expiants between Days 1 and 3 of the culture period. A. activity secreted in 2-day intervals per expiant; B, same data normalized for the tissue DNA content at the end of the culture period. Columns, averages from 3 experiments; bars, SD.

mammary gland tissue cultured without any hormones (control) lost about 70% of its initial DNA content during a 9-day culture period, while expiants cultured in the presence of the growth medium-hormone combination or in the presence of insulin and hydrocortisone retained at least one-half of their initial DNA content. This makes comparisons of secretion rates based on tissue DNA difficult, since a possible stimulatory effect of a hormone on a given cell population would appear to be reduced, because of less loss of tissue DNA. The effects of various hormone combinations are shown in Chart 2, where secretion rates are compared to those observed for control cultures (DMEM only) on Day 3 and are shown as activity secreted per expiant (Chart 2A) and per ^g tissue DNA (Chart 2B). Hormones at the indicated concentrations were added at the beginning of the culture period. While the activity secreted from expiants cultured in DMEM alone (control) in creased somewhat during the later culture period, secretion was VOL. 45 MAY 1985

2296


CYSTEINE

PROTEINASE

SECRETION

FROM CULTURED

MAMMARY

GLAND

reduced with all hormone combinations tested. Reduced secre-

Table 3

tion/Mg tissue DMA could partially be accounted for by the reduced loss of tissue DMA on hormone addition, but the amount of total activity secreted from each expiant was also reduced in all cases as compared to controls. By Day 7, expiants cultured in the presence of hormones secreted cysteine proteinase at less than 10% of the control cultures. Thus, the effect of the hormone combinations used appears to be not only to maintain in part the tissue DNA levels but also to suppress actual secretion of the cysteine proteinase. When the effect of individual hormones was investigated, only hydrocortisone was found to suppress cysteine proteinase se cretion (Chart 3) at concentrations of 1 and 0.1 Â¿tg/ml.Insulin and prolactin had no effect. Expiants of similar initial size secreted essentially the same amounts of enzyme in the presence of these 2 hormones as those cultured in DMEM alone (Chart 3A). If data are normalized to the tissue DNA content at the end of the culture period, secretion rates are lower for expiants cultured in the presence of insulin or prolactin as compared to controls without hormones (Chart 3B). However, this is due to the de creased loss of tissue DNA in the presence of insulin or prolactin. The suppressive effect of hydrocortisone is seen by both meth ods of data analysis. Thus, hydrocortisone does not only have an apparent suppressive effect on proteinase secretion by allow ing increased survival of the mammary epithelial cells but also seems to affect proteinase secretion in a more direct fashion. The low level of activity observed in the culture medium collected on Day 1 (Chart 3) could be due to either low levels of enzyme protein or the presence of a cysteine proteinase inhibitor, which is released from the tissue. When Day 1 medium was tested for inhibitory activity, excess inhibitor was indeed found to be present (Table 3). Preliminary studies of the inhibitor by gel filtration indicate a molecular weight in the range of 10,000 to 15,000, similar to cysteine proteinase inhibitors reported in artic ular cartilage (23), epidermis (11), and the cystatin family of

Inhibition of cysteine proteinase activity by medium from mammary gland expiants

1000

â€”r

800-

â€¢

600-

â€¢

400-

200-

â€¢

13579 DAYS IN CULTURE Chart 3. Variation of cysteine proteinase secretion rates during culture of lac tating mammary gland expiants. Mammary gland tissue was dissected from lactat ing mice 5 days postpartum and established in organ culture. Hormones were added from Day 0 of the culture period as follows: control (DMEM only) (O); insulin (1 Â»ig/ml)(â€¢);insulin and prolactin (1 >Â¡g/mleach) P); insulin and hydrocortisone (1 /jg/ml each) (A); and insulin (1 >ig/ml) and hydrocortisone (0.1 /ig/ml) (A). Medium was changed every 24 h, and cysteine proteinase activity was determined using CBZ-Arg-Arg-NA as substrate. Activity is expressed as nmol NA released per h. A, mean secretion rates per expiant with 4 expiants for each group; B, secretion rates per Â¡igtissue DNA, as determined at the end of the culture period.

CANCER RESEARCH

Day 5 medium from expiants cultured in DMEM only was used as a source of cysteine proteinase activity; 20 nl were mixed with the indicated amounts of medium, preincubated for 30 min at room temperature, and then assayed for cysteine proteinase activity with CBZ-Arg-Arg-NA as substrate. The initial activity of the Day 5 medium used was 31.4 nmol NA released per 2 h. Culture (h)2413fresh period

(Â¿il)1020202020%

inhibition446064620

mediumVolume

CVSTEINE PROTEINASE

Day 3

Day 5

Day 3

Day Ã•

Chart 4. Secretion of cathepsin D and cysteine proteinase from lactating mam mary gland explants. Explants were cultured in DMEM only until Day 3, at which time fresh medium with the following hormone supplements was added: H, hydrocortisone; H + I, hydrocortisone and insulin; H + Prl, hydrocortisone and prolactin; H + E, hydrocortisone and estradici (1 ng/ml). All other hormone concentrations were at 1 /jg/ml. Medium was collected every 2 days and assayed for cysteine proteinase activity with BAÃ‘A and for cathepsin D activity with acid- denatured hemoglobin as substrates. Secretion rates are expressed relative to the amount of activity measured on Day 3 for each expiant, such that each expiant serves as its own control. Columns, average from 6 experiments; bare, SD.

cysteine proteinase inhibitors (3). Inhibitory activity was also detected in medium after short-term incubation of lactating mam mary gland (1 and 3 h), which resulted in the release of a large part of the luminal contents, indicating that this inhibitor might be a milk component. No evidence for excess inhibitory activity was found in media from expiants cultured in the presence of hydrocortisone. Thus, the effect of hydrocortisone appears to be on the secretion of the enzyme protein rather than on the release of a cysteine proteinase inhibitor. The suppressive effect of corticosteroids on cysteine protein ase secretion was also observed, when they were added to expiants after an initial 3-day culture period in the presence of insulin only (Chart 4, right). Under these conditions, a large proportion of the mammary epithelium has already regressed, suggesting that the cell population secreting the proteinase is directly responsive to corticosteroids. Insulin, prolactin, and estradiol did not reverse or increase the suppressive effect of hydrocortisone. Data shown demonstrate the effect of lactogenic and mam mary growth hormones on cysteine proteinase secretion from lactating mammary gland. However, similar results were obtained with mammary glands from pregnant, hormone-primed, or virgin mice (data not shown). Secretion of Lysosomal Enzymes. Since the secreted cys teine proteinase was shown to be related to the lysosomal enzyme cathepsin B (see accompanying paper), the accumula tion of other lysosomal enzymes, namely cathepsin D and ÃŸgalactosidase, was studied to determine whether the release of these enzymes was affected similarly by hormones. For these VOL. 45 MAY 1985

2297


CYSTEINE

PROTEINASE

SECRETION

experiments, expiants from lactating mammary glands (5 days postpartum) were cultured for 3 days in hormone-free medium. Hormones were added on Day 3 of the culture period as indicated in Chart 4. While cysteine proteinase secretion was suppressed by 40 to 50% after 2 days of culture in the presence of hydrocortisone and by about 80% on Day 9, no such supression was observed for cathepsin D (Chart 4). On Day 9, cathepsin D activity released into the culture medium in the presence of hydrocortisone and insulin appeared to be significantly lower compared to controls, hydrocortisone alone, or hydrocortisone plus prolactin or estrogen. This effect is probably due to better maintenance of the mammary epithelium under these culture conditions as compared to the other hormone combinations used. Since cathepsin D release from mammary gland culture was not found to be greatly influenced by cycloheximide (data not shown), it is concluded that its release occurs mainly from dying cells and is not actively secreted. A similar lack of modu lation by hydrocortisone was found for the release of the lysosomal glycosidase, /3-galactosidase(Chart 5). The greatest re lease of /3-galactosidaseoccurred during the initial 2 days of the culture period, suggesting dying cells are the primary source of 1.5-T

1.0-

FROM CULTURED

MAMMARY

GLAND

this enzyme. Also, if cycloheximide was added, there was a transient increase in ÃŸ-galactosidasesecretion over control cul tures. These data indicate that the secretion of the cysteine protein ase is regulated independently to the release of lysosomal en zymes from cultured explants, as had been shown for the secretion of a similar enzyme from human breast tumors (21). Morphologyof Cultured Mammary Gland Expiants. Sinceit is well known that culture of mammary gland tissue in the absence of the hormones required for mammary gland growth and differentiation results in regression of the tissue (24), explants were examined histologically to study the appearance of the cell populations which might be responsible for secretion of cysteine proteinases. Tissue was fixed after 5, 8, and 12 days of culture in DMEM and insulin, and 5-u.msections were stained with hematoxylin and eosin. While most of the mammary epithe lium appeared dead even after 5 days of culture, areas persisted which not only showed intact cells but also appeared to be mitotically active; mitotic cells could easily be identified in these areas (Fig. 1). These cells generally appeared enlarged with pale nuclei. Some lobular organization was present, especially at the early stages (Fig. 1a). Cells often appeared to grow around regressing alveoli. Such viable areas were most often found around the edges of the expiants exposed to the gaseous phase of the culture system (Fig. 1, a and b). At 8 and 12 days, when the original structure of the mammary epithelium had largely collapsed, these cell populations still persisted (Fig. 1, c and d) and were often dispersed throughout the expiant. Incorporation of [3H]thymidine into expiants after 4 days of culture in the presence of various hormone combinations also suggested the presence of mitotically active cells. At this stage, DNA synthesis was not stimulated by lactogenic or growth-promoting hormones (Table 4). These observations suggest the presence of a cell population in these cultured expiants which survives the adverse culture conditions and which appears to be capable of hormoneindependent growth.

|o..J

Â§

DISCUSSION

l l l l l l l

1234567 DAYS

IN CULTURE

Chart 5. Secretion of /i-galactosidase from cultured expiants of lactating mam mary gland. Culture media assayed for (f-galactosidase activity were the same as those described in Chart 3 for expiants cultured in the presence of insulin (1 ng/ ml) (â€¢)and hydrocortisone (1 ng/ml) (O). Cycloheximide (5 jig/ml) was added to expiants cultured in DMEM and insulin for 1 day and replaced with fresh medium without inhibitor on Day 3 (â€¢).fi-Galactosidase was assayed as described in "Materials and Methods." Activity is expressed as nmol methylumbelliferone re leased per min. Points, average from 4 expiants; bars, SO. Table 4 DNA synthesis in cultured expiants from lactating mammary glands Expiants and culture conditions were as outlined in the legend of Table 1. Medium containing the appropriate hormones was replaced on Days 2 and 4 just prior to the addition of [3H]-thymidine (5 nCi/dish). For fresh tissue, expiants were cultured immediately after dissection for 4 h in the presence of insulin only. [3H]Thymidine incorporation Culture conditions Fresh tissue Control medium Growth medium Lactogenic medium Regression medium

(cpm/^g DNA) 2216 4916 3110 2913 4071

Â±258 Â±437 Â±396 Â±351 Â±1063

The data presented indicate that expiants of mouse mammary glands secrete a cysteine proteinase in a similar fashion to the secretion observed with expiants of malignant human breast tumors (21) and spontaneous mammary carcinomas in C3H mice (20). However, the absence of increased secretion in the pres ence of lactogenic hormones and its inhibition by hydrocortisone suggest that this process is not related to the normal secretory processes of the mammary gland. This is further supported by the finding that mammary glands at all stages of growth and differentiation have the capacity to secrete the cysteine protein ase. As secretion of the cysteine proteinase is highest in the absence of any additional hormones or in the presence of insulin only, conditions which favor regression and cell death of a large part of the mammary epithelium especially in lactating tissue, it could be argued that the enzyme leaks out of dying cells. However, the fact that cycloheximide reversibly inhibits secretion would argue against this. In addition, when mammary gland homogenates were assayed for the presence of a stable cysteine proteinase, only very low levels were detectable,4 which could 4 A. D. Recklies, unpublished observation.

CANCER RESEARCH VOL. 45 MAY 1985 2298


CYSTEINE

PROTEINASE

SECRETION


GLAND

not account for the amounts of enzyme accumulating in the lesions, and their growth upon transplantation appears to be culture medium over a 7- or 9-day culture period. Also, ÃŸ- repressed in the presence of normal mammary epithelium (5, 6). galactosidase and cathepsin D, used as markers for lysosomal Their growth might be promoted in an in vitro environment which enzymes, showed quite different patterns of release, suggesting removes possible repressive effects of the normal mammary that a large part of these activities was released from dying cells. epithelium. Schaefer ef al. (25) have shown that exposure of mammary While release of the cysteine proteinase inhibitor during the initial culture period obscures the measurement of activity, it is unlikely glands to low levels of carcinogens produces a maximal number that no enzyme is secreted during the first 24 h of culture. of lesions, and these are not oncogenic upon transplantation. Preliminary investigations show that the inhibitor can be inacti They postulate the requirement for the completion of a multistep vated by pepsin treatment of the medium, resulting in the un process leading to oncogenic transformation. As our data sug gest that normal mammary gland tissue contains a hormonemasking of cysteine proteinase activity. This would suggest that some enzyme protein is secreted, even during the initial culture independent cell population, it is possible that these cells are the period, and that this process might occur in vivo, if the hormonal primary target for further transformation by chemical carcino gens. This process might be more efficient if expiants are ex environment is favorable. When the total amount of cysteine proteinase secreted from posed to carcinogens in the absence of exogenous hormones, expiants of approximately similar initial size is compared, it is removing any repressive effects of the normally differentiated evident that essentially similar amounts are released, independ mammary epithelium. This suggestion is supported by the finding that tumor growth from carcinogen-induced nodules in rat mam ent of the total number of surviving cells (i.e., culture in DMEM compared to culture in insulin-supplemented medium). This mary glands is increased, if the tissue is cultured in hormonewould suggest the existence of a cell population within the deficient medium before transplantation (22). mammary gland tissue which does not require hormones for If the cysteine proteinase is indeed secreted from a cell pop survival in culture and which secretes this enzyme. DNA synthe ulation of the mammary gland which is susceptible to oncogenic transformation, it might serve as a useful marker for the presence sis in expiants from lactating mice 5 days postpartum was not stimulated by hormones either at Day 0, as has been reported of hyperplastic lesions in this tissue. Such lesions are thought to by other investigators (13, 31), or after 4 days of culture. Thus, be morphological precursors of mammary tumors in mice (4, 5), the persistence of a dividing cell population in these expiants is and it has been suggested that this is also the case in the independent of any externally added hormones. Thymidine in development of human breast tumors (32). Since expiants from corporation per fig of tissue DNA is higher in the control medium mammary carcinomas also contain cell populations capable of compared to medium containing lactogenic or growth-promoting secreting a stable cysteine proteinase (20), it appears that this hormones. However, part of this is accounted for by the in capacity is acquired very early in the transformation process and creased survival of epithelial cells in the presence of hormones, retained by at least some cells during further malignant transfor resulting in higher final tissue DNA levels. While the data pre mation. sented here do not directly demonstrate that the mitotically active cell population is the source of the secreted cysteine proteinase, ACKNOWLEDGMENTS the requirement for viable cells and the continued secretion in The authors wish to thank M. Turner and M. Lepik for their help in preparing culture would suggest that this is the case. this manuscript, and J. Gilpin for performing the histology. The functional in vitro differentiation of mammary gland is greatly influenced by hormones, and abnormal courses of differ entiation in their absence have been reported (29). Thus, the REFERENCES development of squamous metaplasia as an alternative differ 1. Banerjee, M. R. Responses of mammary cells to hormones. Int. Rev. Cytol., entiation pathway for the mammary epithelium is greatly en 47:1-97,1976. 2. Barrett, A. J., and Heath, M. F. Lysosomal enzymes. In: J. T. Dingle (ed.), hanced in the presence of insulin and factors which increase Lysosomes, a Laboratory Handbook. Ed. 2, pp. 19-127. Amsterdam: North intracellular cyclic AMP levels (26, 29). The morphological ap Holland, 1977. pearance of the expiants reported in this study also suggests 3. Brzin, J., Kopitar, M., Turk, V., and Machleidt, W. I. Isolation and characteri zation of stefin, a cytosolic protein inhibitor of cysteine proteinases from human abnormal growth patterns which are not observed in the properly polymorphonuclear granulocytes. Hoppe-Seyler's Z. Physiol. Chem., 364. differentiated mammary gland, but further investigation is needed 1475-1480. 1983. to determine the exact nature of the surviving cell population. 4. De Ome, K. B., Faulkin. L. J., Bern, H. A., and Blair, P. B. Development of mammary tumors from hyperplastic alveolar nodules transplanted into the Dulbecco ef al. (8) have shown in the rat that the mammary gland-free fat pad of C3H mice. Cancer Res., 19: 515-520,1959. epithelium contains various proliferative cell pools, which appear 5. De Ome, K. B., Miyamoto, M. J., Osborne, R. C., Guzman, R. C., and Lum, K. Detection of inapparent nodule-transformed cells in the mammary gland tissues to respond to different factors and hormones during the estrous of virgin female BALB/cf C3H mice. Cancer Res., 38: 2103-2111,1978. cycle, and suggest a sequence of cell lineage from a stem cell 6. De Ome, K. B., Osborne, R. C., Miyamoto, M. J., and Guzman, R. C. A survey population based on cytoskeleton and membrane markers (7). It of mouse mammary tissues from eight strains and two hybrids for the presence of nodule-transformed cells. In: C. McGrath, M. J. Brennan. M. A. Rich (eds.). is possible that our culture system allows the differential survival Cell Biology of Breast Cancer, pp. 79-91. New York: Academic Press, Inc., of one of these cell populations and that the lack of hormones 1980. 7. Dulbecco, R., Unger, M., Armstrong, B., Bowman. M.. and Syka, P. Epithelial promotes an abnormal differentiation pathway. cell types and their evolution in the rat mammary gland determined by immuAlternatively, one could be dealing with a population of inapnological markers. Proc. Nati. Acad. Sci. USA, 80. 1033-1037, 1983. 8. Dulbecco, R., Henahan, M., and Armstrong, B. Cell types and morphogenesis in the mammary gland. Proc. Nati. Acad. Sci. USA, 79: 7346-7350,1982. 9. Hendrick, I.e., and Franchimont, P. Radioimmunoassay of casein in the serum of normal subjects and of patients with various malignancies. Eur. J. Cancer, 70: 725-730, 1974.

parent nodule transformed cells, as have been described by De Ome ef a/. (6) in various strains of parous and virgin mice. These nodules appear to be present in morphologically normal mam mary glands before the appearance of hyperplastic alveolar CANCER RESEARCH

VOL. 45 MAY 1985

2299


CYSTEINE

PROTEINASE

SECRETION

10. Ichinose, R. R., and Mandi, S. Influence of hormones on lobuloalveolar differ entiation in mouse mammary tissues in vitro. J. Endocrino!., 35: 331-340, 1966. 11. Jarvinen, M. Purification and some characteristics of the human epidermal SHprotease inhibitor. J. Invest. Dermatol., 72: 114-118,1978. 12. Kissane, J. M., and Robbins, E. The fluorometric measurement of deoxyribcnucleic acid in animal tissues with special reference to the central nervous system. J. Biol. Chem., 233:184-188, 1958. 13. Knight, C. H., and Peaker, M. Mammary cell proliferation in mice during pregnancy and lactation in relation to milk yield. Quart. J. Exp. Physiol., 67: 165-177,1982. 14. Knight, G. C. Human cathepsin B. Biochem. J., 789:447-453,1980. 15. Medina, D. Preneoplastic lesions in murine mammary cancer. Cancer Res., 36: 2589-2595,1976. 16. Mort, J. S., Recklies, A. D., and Poole, A. R. Characterization of thtol proteinase secreted by malignant human breast tumours. Biochim. Biophys. Acta, 674: 134-143,1980. 17. Poole, A. R., Tiltman, K. J., Recklies, A. D., and Stoker, T. A. M. Differences in the secretion of the proteinase cathepsin B from the invading edges of mammary carcinomas and fibroadenomas. Nature (Lond.), 273: 545-547, 1978. 18. Poole, A. R. The locations of cathepsin activity and tf-glucuronidase in the Guerin T8 tumor. Biochem. J., 778: 543-549,1970. 19. Recklies, A. D., Poole, A. R., and Mort, J. S. A cysteine proteinase secreted from human malignant breast tumors is immunologically related to cathepsin B. Biochem. J., 207: 633-636, 1982. 20. Recklies, A. D., Mort, J. S., and Poole, A. R. The secretion of a thiol proteinase from mouse mammary carcinomas and its characterization. Cancer Res., 42: 1026-1032, 1982. 21. Recklies, A. D., Tiltman, K. J., Stoker, T. A. M., and Poole, A. R. Secretion of proteinases from malignant and non-malignant human breast tissue. Cancer

CANCER RESEARCH

FROM CULTURED

MAMMARY

GLAND

Res., 40: 550-556, 1980. 22. Rivera, E. M., Hill, S. D., and Taylor, M. Organ culture passage enhances the oncogenicity of carcinogen-induced hyperplastic mammary nodules. In Vitro (Rockville), 77: 159-166,1981. 23. Roughley, P. J., Murphy, G., and Barrett, A. J. Proteinase inhibitors of bovine nasal cartilage. Biochem. J., 769: 721-724,1978. 24. Salazar, H., and Tobon, H. Morphological changes of the mammary gland during development, pregnancy, and lactation. In: J. B. Josimovich (ed.), Lactogenic Hormones, Fetal Nutrition, and Lactation, pp. 221-277. New York: John Wiley & Sons, 1974. 25. Schaefer, F. V., Tonelli, Q. J., Dickens, M. S., Custer, R. P., and Sorof, S. Nononcogenic hormone-independent alveoli produced by carcinogens in cul tured mouse mammary glands. Cancer Res., 43: 3310-3315,1983. 26. Schaefer, F. V., Custer, R. P., and Sorof, S. Induction of squamous metaplasia: requirement for cell multiplication, and competition with lobuloalveolar devel opment in cultured mammary glands. Differentiation, 25: 185-192,1983. 27. Schultz, G. S., and Ebner, K. E. a-Lactalbumin levels in human mammary tumors, sera, and mammary cell culture lines. Cancer Res., 37: 4489-4492, 1977. 28. Thomas, P. S., and Farquhar, M. N. Specific measurement of DNA in nuclei and nucleic acids using diaminobenzoic acid. Anal. Biochem., 89:35-44,1978. 29. Tonelli, Q. J., and Sorof, S. Expression of a phenotype of normal differentiation in cultured mammary glands is promoted by epidermal growth factor and blocked by cyclic adenine nucleotide and prostaglandins. Differentiation, 20: 253-259,1981. 30. Topper, Y. J., and Freeman, C. S. Multiple hormone interactions in the developmental biology of the mammary gland. Physiol. Rev., 60: 1049-1106, 1980. 31. Traurig, N. H. Cell proliferation in the mammary gland during late pregnancy and lactation. Anatom. Ree., 757: 489-503,1967. 32. Wellings, S. R. A hypothesis of the origin of human breast cancer from the terminal ductal lobular unit. Pathd. Res. Pract., 768: 515-535, 1980.

VOL. 45 MAY 1985

2300


CYSTEINE

PROTEINASE

SECRETION

FROM

CULTURED

MAMMARY

GLAND

Fig. 1. Light micrographs of cultured mammary gland explants. Explants from lactating mammary gland isolated 6 days postpartum were cultured in DMEM containing insulin (1 /ig/ml) for the indicated times, a. expiant fixed after 5 days of culture. The viable area is surrounded by degenerated alveoli; some cells still contain large lipid droplets, b and c, expiants fixed after 8 days of culture. Mitotic figures are still frequent; cells do not appear to contain large amounts of secretory material, d, expiant fixed after 12 days of culture, x 450 (a); x 350 (b to d). Photomicrographs were taken with a Zeiss Photomicroscope III.

CANCER RESEARCH

VOL. 45 MAY 1985

2301
