Expression of Cathepsin D in Primary and Metastatic Human Melanoma and Dysplastic Nevi

Expression of Cathepsin D in Primary and Metastatic Human Melanoma and Dysplastic Nevi Osvaldo L. Podhajcer,* Laura Bover,* Alicia I. Bravo,t M. Fernanda Ledda,* Claudia Kairiyama,* Ignado Calb,$ Liliana Guerra,* Francoise Capony,1f and Jose Mordoh* *Instituto de lnvestigaciones Bioquimicas Luis F. Leloir "Fundacion Campomar" and Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires; tHospital Eva Peron, Buenos Aires; and ^Hospital Pinero, Buenos Aires, Argentina; and UlNSERM U 148, Hormones and Cancer, Montpellier, France

High levels of cytosolic cathepsin D expression have been associated w^ith poor prognosis in breast cancer node-negative patients. In this ivork, we provide evidence that three cell lines established from human metastatic melanomas—IIB-MEL-J, IIB-MEL-LES, and IIB-MEL-IAN—express high levels of procathepsin D mRNA. IIB-MEL-J cells secreted into the conditioned media about 30% of the newly synthesized protein, which was active at acidic pH. Melanoma tumors arising in nude mice after injection of the three different cell lines expressed high levels of procathepsin D mRNA. Moreover, 13 human metastatic melanomas expressed variable levels of procathepsin D mRNA. To study the possible association between cathepsin D expression and melanoma de-

velopment, samples corresponding to 10 primary tumors, 11 metastatic melanomas, 10 dysplastic nevi, 27 nevocellular nevi, and normal melanocytes were studied by immunohistochemistry for cathepsin D-specific staining. 'We found that cathepsin D was expressed in all of the dysplastic nevi and primary and metastatic melanomas tested but in only 18% of nevocellular nevi (five of 27), w^hereas normal melanocytes showed no cathepsin D expression. The overall data indicate that cathepsin D is expressed at a high level by melanoma cells, and because of its expression in preneoplastic lesions, it may be associated with melanoma development. / Invest Dermatol 104:340-344, 1995

T

to 60% of the inactive proenzyme as a result of altered processing of the 52-kD precursor at the level of its interaction with mannose-6 P-receptors [7]. Secreted procathepsin D, activated under mUd acidic conditions, is capable of digesting the extracellular matrix of endothelial bovine comeal cells, suggesting its potential role in tumor invasion [8]. Indeed, increased cathepsin D expression in human breast cancer has been associated with a worse prognosis [9]. A previous report showed cathepsin D expression in some human melanomas [10]. In the present study, we confirm and extend those results by showing that the great majority of primary and metastatic melanomas and dysplastic nevi express high levels of cathepsin D, whereas only five of 27 common acquired nevocellular nevi express cathepsin D and normal melanocytes express none. The present data suggest that the expression of this aspartyl protease might be associated with melanoma development.

he incidence of cutaneous melanoma is increasing rapidly throughout the world, and a sequence of events leading from nevocellular nevi to dysplastic nevi to melanoma has been proposed [1]. Human melanomas are characterized by their high metastatic capacity. Tumor invasion of the surrounding matrix is a crucial step in the multistage process leading to metastasis formation. Several types of extracellular-matrix-degrading enzymes have been impUcated in stroma invasion by malignant cells (see [2] and references therein). Some of them, like cathepsin B [3] and more recently the matrix metalloproteinases [4], have been associated with the invasive and metastatic capacity of melanoma cells. A 52-kD protein secreted in excess by cultured human breast cancer cells has been found to be similar to the aspartyl protease, cathepsin D [5,6]. This lysosomal endopeptidase is produced under normal conditions as a 52-kD inactive proenzyme, and is processed intracellularly into an intermediate 48-kD form and a 34-kD mature form [5]. Consistent with normal processing, normal mammary cells secrete in vitro less than 5% of the newly synthesized procathepsin D. In contrast, breast cancer cell lines may secrete up

MATERIALS AND METHODS Cell Lines and Proteinase Assay The IIB-MEL-J cell line was derived from a metastatic human melanoma [11]. The IIB-MEL-LES and IIB-MELIAN human melanoma cell lines, also derived from metastatic melanomas, were established in our laboratory and will be described elsewhere. Tbe estrogen-receptor (ER)-positive breast cancer cell line MCF-7 and the ER-negadve breast cancer cell lines MDA-MB-231 and IIB-BR-G have been described [12-14]. To assay the proteinase activity in conditioned media, IIB-MEL-J cells were cultured in the absence of fetal bovine serum for 18 b, after which tbe medium was recovered. The proteinase activity was measured using hemoglobin [15] or (metbyl-^'*C)metbemoglobin [5] as substrate. In experiments involving proteinase inhibitors, pepstatin A and

Manuscript received April 8, 1994; revised October 21, 1994; accepted for publication November 14, 1994. Liliana Guerra's present address: Fundacion CIMAE, Luis Viale 2831, Buenos Aires, Argentina. Reprint requests to: Dr. Jose Mordoh, Instituto de lnvestigaciones Bioquimicas Luis F. Leloir "Fundacion Campomar," Av. Patricias Argentinas 435, 1405 Buenos Aires, Argentina. Abbreviation: ER, estrogen receptor.

0022-202X/95/$09.50 • SSDI0022-202X(94)00354-A • Copyright © 1995 by The Society for Investigative Dermatology, Inc.

340

VOL. 104, NO. 3 MARCH 1995

leupeptin at the concentrations mentioned in the Results were added immediately before the assay. Immunoprecipitation Assay Exponentially growing IIB-MEL-J cells were plated, 10^ each, into 35-mm Petri dishes and laheled with [^^S]methionine. The medium and the cells were collected separately to perform an immunoprecipitation assay with the monoclonal antihody D8F5, which recognizes the 52-kD procathepsin D and the intermediate and mature forms of the enzyme [16]. The labeling and processing of the cells have heen descrihed previously [5].

Tumor Cell Dissociation, RNA Extraction, and Northern Analysis Total RNA was obtained either from exponentially growing cultures of melanoma cell lines or from metastatic tumors. Briefly, recently excised metastatic tumors were trimmed rapidly of fat and necrotic tissue and conserved in hquid nitrogen until use. Alternatively, metastatic tumors were cut in small fragments and dissociated mechanically hy pressing through a nylon mesh until a cell suspension was obtained. Isolated cells were cryopreserved until use [14]. Cell viahility was estimated with the trypan blue dye exclusion test. For RNA preparation, tumor fragments or cells were thawed, washed twice with phosphate-buffered saline, and resuspended in buffer containing sodium isothiocyanate. If necessary, cells were disrupted with a Polytron homogenizer. Total RNA was ohtained as descrihed [17], electrophoresed in 1% agarose gels in the presence of formaldehyde, transferred to nylon membranes (Hybond N, Amersham), and hybridized with random-primed cDNA prohes (10'' cpm/ml). RNA integrity and loaded amount were checked with methylene hlue. After 6-8 h of pre-hybridization, the memhranes were hyhridized overnight at 42°C with 50% formamide. Membranes were washed under high stringency conditions, twice v^dth 2 X sodium citrate/sodium chloride buffer, 0.1% sodium dodecylsulfate at room temperature, and four times with 0.1 X sodium citrate/sodium chloride buffer, 0.1% sodium dodecylsulfate at 55°C [18]. The human procathepsin D cDNA probe is a 2.0-kilobase-pair &oRI insert [5]. Human melanomas were induced in mice using exponentially growing melanoma cells, which were harvested by trypsinization, washed with culture medium without serum, and resuspended in phosphate-buffered saline. Five million cells in a volume of 0.1 ml were injected subcutaneously into the flanks of 6-8-week-old male nude mice (BALB/c-nu/nu). Tumors 1 cm in diameter were used for RNA extraction, as descrihed ahove. Immunochemical Determinations Paraffin-embedded tissue hlocks were used to detect cathepsin D by immunohistochemistry. Shces 5 jum thick were mounted on gelatin-coated slides by standard procedures. Melanin pigments were eliminated from hydrated sections as described [19]. Melanin removal was checked routinely by use of a control with phosphatebuffered saline instead of the first antibody. Cathepsin D was detected immunohistochemically with a commercial monoclonal antibody (Triton Diagnostics, Alameda). An avidin-biotin bridge was used following the manufacturer's instructions (Vectastain, Vector Laboratories). The reaction was developed vvath 0.02% diaminobenzidine and hydrogen peroxide. Staining intensity was scored as follows: 1+ = low intensity; 2+ = moderate intensity; 3+ = strong intensity. Cathepsin D expression in normal melanocytes corresponds to that in normal areas adjacent to henign pathologies and in well-preserved autopsies. An ER-immunocytochemical assay monoclonal antibody (Abbott Laboratories, North Chicago, IL) was used for ER determination on melanoma cells.

CATHEPSIN D EXPRESSION IN HUMAN MELANOMA

341

Cathepsin D is Secreted by Human Melanoma CeUs The pepstatin inhibition of proteinase activity at acidic pH prompted us to investigate whether an activated human melanoma cathepsin D could be responsible for this activity. For this purpose, we used a monoclonal antibody raised against the human breast cancer procathepsin D that recognizes both the inactive precursor and the mature forms [16]. An immunoprecipitable 52-kD band corresponding to the secreted procathepsin D was found in the IIBMEL-J—conditioned medium. Moreover, the 34-kD mature form of cathepsin D was immunoprecipitated from cell homogenates (Fig 1) Quantitative analysis also demonstrated that approximately 30% of the newly synthesized procathepsin D was secreted into the medium after a 16-h labeling period (data not shown). This secretion is not under estradiol regulation in the IIB-MEL-J cell line, as the addition of 10~^ M estradiol in the absence of phenol red for 48 h to IIB-MEL-J cells deprived of estradiol for 5 d did not affect the secretion of procathepsin D. In addition, IIB-MEL-J cells were found to express no ER by immunocytochemical analysis (data not shown). Cathepsin D mRNA is Expressed hy Established Human Melanoma Cell Lines The high levels of procathepsin D observed in the IIB-MEL-J-conditioned medium correlated with high levels of mRJSfA expression. Northern blot analysis has shovkoi that IIB-MEL-J ceUs as well as IIB-MEL-LES and IIB-MEL-IAN cells expressed high levels of a 2.2-kilobase-pair band corresponding to the expected size of procathepsin D mRNA. This expression was comparable to the levels observed vdth different ER-positive and ER-negative human breast cancer cell lines (Fig 2). To evaluate whether the expression of procathepsin D in human melanoma cells is the result of cell adaptation to growth in vitro, we also determined procathepsin D mRNA levels in melanomas induced in nude mice by injection of the three different cell lines. Figure 2 shows high levels of procathepsin D mRNA expression in all the tumors tested, suggesting that the constitutive expression of

A B C D -92K

.46K

RESULTS Human Melanoma-Cell-Conditioned Medium Contains Proteases Active at Acidic pH Initial experiments using hemoglobin as substrate indicated that conditioned medium obtained from human melanoma IIB-MEL-J cells contained proteinase activity with an optimum acidic pH. By using [methyl-^ "^Cj methemoglobin, we found that IIB-MEL-J-conditioned medium solubilized 8% ± 4% of total trichloroacetic-acid-precipitable radioactivity after 60 min incubation and 15% ± 3% after 120 min incubation (mean ± SD of three different experiments) at pH 4.0 [5]. Twenty-six percent of this activity was inhibited upon the addition of the aspartyl protease inhibitor pepstatin A at 10"*" M. That pepstatin A did not completely prevent the protease activity of IIB-MEL-J cells at acidic pH suggested the presence of additional proteases. Indeed, 32% inhibition was observed in the presence of the cysteine protease inhibitor leupeptin at 10"^ M, indicating the presence of active serine and/or cysteine proteases.

-30K

.18K Figure 1. Immunoprecipitation of the 52-kD procathepsin D and the 34-kD mature enzyme from the cellular extract and conditioned medium of IIB-MEL-J cells. Afrer labeling of cells with [^^Sjmethionine, the conditioned medium {A,B) and cellular extract {C,D) were ohtained from IIB-MEL-J cells and were immunoprecipitated as descrihed in Materials and Methods with the D8F5 monoclonal antibody {B,D) or with normal mouse immunoglobulins {A,C).

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

PODHAJCER ET AL

z

1

1

JJ

UJ

_i Ul

j

Ul

Ul

UJ

UJ

S

:s

2

2 m

t^

CD J_

S

m =

u 2

CD

AN

LES

o

->

O DC

LES

MELANOMA

BREAST

m

three (Fig 3). We found no correlation between mPJ^A procathepsin D expression and the origin of the metastasis or of the samples (from tumor fragments or isolated cells).

NUDE MICE TUMORS

CELL LINES

CD

CD

IS

IAN

342

m

128 S CAT

[18 S

Figure 2. Human melanoma cell lines and melanoma tumors induced in nude mice express cathepsin D mRNA levels comparable to those expressed by human breast cancer cell lines. Twelve

micrograms of total RNA was loaded in each lane, electrophoresed, transferred to nylon membranes, and hybridized with a human cathepsin D probe. Cell lines used were two ER-negative (MDA-MB-231 and IIBBR-G) and one ER-positive (MCF-7) breast cancer cell lines and three human melanoma cell hnes (IIB-MEL-LES, IIB-MEL-J, and IIB-MELIAN). Nude mice tumors were human melanoma tumors obtained in nude mice as described in Materials and Methods. The positions of the 28 S and the 18 S ribosomal RNAs are indicated. Methylene blue staining of nylon membranes shows RNA integrity and comparable loading.

this aspartyl protease in human melanomas is not a result of growth in culture. Cathepsin D mRNA is Expressed by Human Metastatic Melanomas Because cultured cells are not necessarily representative of malignant cells in the intact tissues from which they were derived, we studied the expression of cathepsin D in human metastatic melanoma samples by Northern and immunohistochemical analyses. Tissue fragments or isolated cells corresponding to 24 cryopreserved human metastatic melanomas were used for total PJMA extraction. After checking RNA integrity by methylene blue staining (Fig 3), we selected 13 samples for Northern blot analysis. As with the established melanoma cell lines, only a 2.2-kilobasepair procathepsin D mRNA band was detected in all the tumor samples examined. High procathepsin D mRNA levels were observed in four samples, moderate levels in six, and low levels in

METASTATIC MELANOMAS 1

2

3

4

5

6

7

8

9

Immunohistochemical Detection of Cathepsin D in Primary and Metastatic Melanoma and Dysplastic Nevi Because previous reports have shown that cathepsin D can be expressed by both mahgnant and nonmalignant host-infiltrating cells [20], we performed an immunohistochemical analysis to evaluate cathepsin D distribution in melanomas. All of the 10 metastatic melanomas tested expressed intracytoplasmic cathepsin D (Table I, Fig 4a,b), although wide variabiUty was observed in the staining intensity and in the percentage of cells expressing cathepsin D (Table I). In those samples studied by both Northern and immunohistochemical analyses, we found that low expression of mPJSIA correlated with low to moderate staining intensity and with less than 40% positive cells. By immunohistochemical analysis, cathepsin D was found to be expressed preferentially by the malignant cells, although infiltrating macrophages also expressed it (Fig ^c,d). In six of 10 metastatic melanoma samples, as well as in all the primary melanomas tested (see below), cathepsin D expression was found to be associated with fine granules distributed throughout the cytoplasm (Fig 4a). However, in four metastatic melanomas, cathepsin D expression was associated with clusters of perinuclear granules with net borders (Fig 4b). No evidence of stromal reactivity or of staining of lymphocytes or muscle cells was observed. However, macrophages, sebaceous glands, sweat glands, and some epidermal keratinocytes showed staining (Fig 4c,d). To determine the significance of cathepsin D expression in melanomas, we performed further studies in primary melanomas, benign nevocellular nevi, dysplastic nevi, and normal melanocytes (see below). In these cases, cathepsin D expression was studied only by immunohistochemistty because of the small amount of material available. Intracytoplasmic staining was observed in all of the 11 primary melanomas tested and in approximately 90% of ceUs (Table I, Fig 4c). Nine primary melanomas showed intense intracytoplasmic staining, and the remaining two samples expressed cathepsin D with moderate intensity (Table I). All of the 10 dysplastic nevi analyzed by immunohistochemistry expressed cathepsin D with moderate (two of 10) to strong (eight of 10) intensity and with an overall mean cell positivity of 82% (range 55%-95% of malignant cells; see Table I and Fig 4eJ). Similarly, all three Spitz nevi analyzed showed strong intensity of cathepsin D expression (data not shown). However, only five of 27 nevocellular nevi showed low (two of five) to moderate (three of five) intracytoplasmic expression of cathepsin D (Table I). Only

Table I.

10 11 12 13

Immunohistochemical Detection of Cathepsin D Staining Intensity

;28 S CAT D •

18 S

Figure 3. Human metastatic melanomas express cathepsin D

mRNA. Eight micrograms of total RNA was loaded in each lane (for further details see Fig 2 and Materials and Methods). The sources of RNA were as follows: isolated cells, samples 1-4 and sample 6; tumor fragments, sample 5 and samples 7-13. The sites of origin of the metastases were as follows: lung, sample 1\ lymph nodes, samples 2, 3, 5, 7, 9, 10, and 12; derma, samples 4, 6, and 11\ brain, sample 8; viscera, sample 13. Sex of the patients was female, in samples 4, 7, and 11. The positions of the 28 S and 18 S ribosomal RNAs are indicated. Methylene blue staining of nylon membranes shows RNA integrity and comparable loading.

Type of Lesion

Number of Cases

0

1+

Primary tumors

11

Metastases

10

1 (15)

Dysplastic nevi

10

2 (55, 50)

Nevocellular nevi

27

22

2 (7,5)

2+

3+

2 (65, 70)

9 [80-100] 94.4 ± 8.3 8 [35-100] 63.7 ± 29.4 8 [76-95] 88.3 ± 6.8

1 (60)

3 [10-60] 28.3 ± 22.5

" Staining intensity was sco.ed as described in Materials and Methods, Data are presented as the number of positive cases, percentage of positive cells in each case (in parentheses), range of positive cells in percentage (in brackets), and mean ± SD percentage of positive cells.

VOL. 104, NO. 3 MARCH 1995

Figure 4. Immunohistochemical detection of cathepsin D in melanoma and dysplastic nevi. Sections obtained from paraffin-embedded tissues were stained by the immunoperoxidase method as described in Materials and Methods. a,b) Metastatic melanomas; c) primary melanoma; d) compound nevocellular nevus; ej) nevi with a dysplastic component, a) Peritoneal metastasis; intense staining is observed in the cytoplasm of malignant cells (arotvheads). b) Perinuclear cathepsin D is observed (arrowhead), c) nodular fusocellular primary melanoma. Intense staining is observed in neoplastic melanocytes (Mel) and macrophages (A). Lymphocytes (Ly) do not stain, d) Moderate staining is observed in intraepidermal nests of type A nevus cells disposed along the dermoepidermal junction (A) and in type B nevus cells (B). e) Dysplastic and dermal nevus. Intense staining is observed in the atypical melanocytic hyperplasia at the dermoepidermal junction (arrowhead). See also the negative dermal nevus ( D N ) . ^ Dysplastic nevus shovvdng the granular intracytoplasmic staining in cells of the junctional nests (arrowhead). Bars: a,b, 2.3 /xm; c, 4 jLtm; d, 23 jLtm; e, 15 jam; /, 3.1

two of the 27 samples were compound nevi, and in both cases the junctional component consisting of type A nevus cells and intradermal type B nevus cells expressed cathepsin D (Fig 4d). The remaining 25 cases corresponded to dermal-nevus-containing type B cells. On the other hand, normal intraepidermal and follicular melanocytes showed no expression of cathepsin D (not shown). DISCUSSION This study confirms and extends a previous work [10] by showing that all the primary and metastatic human melanomas and dysplastic nevi tested expressed, in most cases strongly, the aspartyl protease cathepsin D, suggesting a role for this protease in the development

CATHEPSIN D EXPRESSION IN HUMAN MELANOMA

343

of melanoma. High levels of cytosolic cathepsin D protein have been suggested as a marker of poor prognosis in primary breast cancer [9,21,22]. High levels of this protein also were associated with shorter recurrence-free survival and overall survival in breast cancer patients, leading to the suggestion that cathepsin D expression by malignant cells may be associated with the capacity of tumor cells to invade [9]. It has also been shown that cathepsin D levels correlate with an increased risk for developing breast cancer in highly proHferative benign mastopathies [10]. ;, More extensive studies with clinical follow-up are w^arranted to confirm that cathepsin D expression can be usefiil as a predictive marker in melanoma development. The expression of cathepsin D by all of the melanoma samples and proliferating precursor lesions, but only in part of the nevocellular nevi analyzed and not at all in normal melanocytes, argues in favor of this possibility. Our results show that 100% of the dysplastic nevi samples studied express high levels of cathepsin D. Although some controversy persists concerning the diagnosis of dysplastic nevi, this lesion is widely considered a precursor of melanoma, especially for individuals in melanomaprone families, as well as a marker of high risk for melanoma. Indeed, 17% to 30% of sporadic melanomas have been associated with dysplastic nevi [23]. The strong staining we observed in the three Spitz nevi studied and the expression of cathepsin D in part of the acquired nevi should be investigated fiirther. Spitz nevi are the result of spindle cell proliferation downward from the dermoepidermal junction. Histologically, cell nuclei appear normal, but as in dysplastic nevi, atypical mitotic figures usually are found. Occasional documentation has suggested that this lesion may metastasize [1]. On the other hand, in a recent histologic review, it was found that 15.4% of melanomas were associated with contiguous benign nevi, usually arising from the intraepidermal or junctional component [1]. One of the most controversial aspects concerning this aspartyl protease is the possible involvement of the tumor-cell-derived cathepsin D in cancer invasion. The studies linking high levels of cathepsin D expression to a Avorse prognosis in breast cancer were based on the immunologic detection of cathepsin D and did not differentiate between the precursors and active forms or between the origin of the enzyme from tumor or host-infiltrating cells [9,21]. In conflict with these findings, a recent immunohistochemical study reported that a high level of intracytoplasmic cathepsin D expressed by the malignant cells, rather than the cathepsin D level in tissue homogenates, is a marker of better prognosis [23]. This is in agreement with what one would expect for a gene initially described in breast cancer to be under estrogen regulation [24], although recent data have shown, in agreement with our results, that ER-negative breast cancer cell lines constitutively express cathepsin D [8]. This conflicting evidence concerning the predictive value of cathepsin D expression in human breast cancer, and a more recent article showing that secreted cathepsin D does not contribute to the in vitro invasive capacity of MCF-7 cells [25], raised the possibility that macrophage-derived cathepsin D also might be involved in breast cancer invasion. The data presented here demonstrate that cathepsin D is expressed constitutively by malignant melanoma cells. The IIB-MEL-J cell line does not contain ER; nor was an ER-independent pathway of cathepsin D regulation observed upon estrogen addition in the absence of phenol red. Immunohistologic studies of primary and metastatic melanomas identified a comparatively low level of host-infiltrating cells expressing cathepsin D, which suggests that cathepsin D is contributed mostly by tumor cells. Whatever role cathepsin D could have as a prognostic marker, it is still difficult to understand its function as an extracellular protease when it is virtually inactive at pH 7.0. Some clues concerning the mechanism by which it might participate directly in matrix degradation have emerged recently. In vitro studies have demonstrated that procathepsin D secreted by breast cancer cells and autoactivated at acidic pH may degrade the extracellular matrix of endotheUal bovine comeal cells [8]. More recently, it was shown that cathepsin D localized in intracytoplasmic acidic vesicles may par-

344

PODHAJCER ET AL

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

ticipate in cancer invasion either at the intracellular level, by degrading endocytosed extracellular matrix, or at the extracellular level, by releasing the vesicles' content, transiently creating an extracellular acidic environment [26], as has been reported during bone resorption [27]. By histologic analysis, we have found that in all primary melanomas and some metastases, cathepsin D is associated with fine granules, most probably corresponding to lysosomal or endosomal vesicles distributed throughout the cytoplasm. However, in some metastatic tumors (but not in primary tumors), cathepsin D staining was found to be associated with perinuclear large vesicles several micrometers in size, resembling the large acidic vesicles found previously in human breast cancer samples [26]. It remains to be established whether these differences in the structure and localization of cathepsin D-containing vesicles correspond to differences in lysosomal enzyme targets. An indirect role for cathepsin D in extracellular matrix degradation arose from recent data linking cathepsin D with the cysteine protease cathepsin B, a lysosomal protease that has been associated with the metastatic capacity of B16 mouse melanoma cells [3,28]. The inhibition of procathepsin B maturation is eliminated by the aspartyl protease inhibitor pepstatin A, which strongly suggests a role for cathepsin D in this process [29], Evidence was presented that only cathepsin D can inactivate the specific cysteine protease inhibitor cystatin, leading to increased activity of cysteine proteases like cathepsin B [30]. The activation of cathepsin B in turn may result in the transformation of procollagenase [31] and pro—urokinase plasminogen activator [32] to their active forms, initiating the process of extracellular matrix degradation [4]. It is interesting that only part of the proteinase activity released by melanoma cells into conditioned media appears to be aspartic proteases, as the activity is not inhibited completely in the presence of pepstatin A, a strong inhibitor of aspartyl protease activity at acidic pH. This is strikingly different fi-om what was reported previously for MCF-7 breast cancer cells, i.e., that pepstatin A completely inhibited the protease activity of MCF-7-conditioned medium at acidic pH [8]. Moreover, the serine and cysteine protease inhibitor leupeptin, which was almost inactive on MCF7—conditioned medium [8], partially inhibited the protease activity of IIB-MEL-J cells at acidic pH, suggesting that melanoma cells secrete into the extracellular media cysteine proteases like cathepsin B, which may interact with cathepsin D for their mutual processing and activation. Recent data shovi^ing that B16 mouse melanoma cells express cathepsins D and B mRNAs, among other cathepsin mRNAs [28], support this hypothesis.

6.

7.

8.

9.

10.

11.

12.

13. 14.

15. 16.

17.

18.

19.

Manual of Histologic Staining Methods of the Armed Forces Institute of Pathology. The

20.

21.

22.

23. This work was supported in part by a Special Grant from the Consejo Nacional de Investigaciones Cientificas y Tecnicas (CONICET, Argentina) and the INSERM (France). It was also supported by grants from the Fundacion para la Investigacion y Prevencion del Cancer (FUCA, Argentina), the Fundacion Sales (Argentina), and the International Centre for Genetic Engineering and Biotechnology (Italy). OLP, LB, MFL, LG, andJM are members of the CONICET, and FC is a member of the INSERM. We acknowledge Dr. Carlos Quintans, National Commission of Atomic Energy, for providing us with the nude mice.

24. 25. 26.

27.

REFERENCES Bamhill RL, Mihm MC Jr: Hlstopathology of malignant melanoma and its precursor lesions. In: Balch CM, Houghton AN, Milton GW, Sober AJ, Soong S-J (eds.). Cutaneous Melanoma. JB Lippincott, Philadelphia, 1992, pp 234-263 Nakajima M, Chop AM: Tumor invasion and extracellular matrix degradative enzymes: regulation of activity by organ factors. Semin Cancer Biol 2:115-127,1991 Sloane B, Dum JR, Honn KV: Lysosomal cathepsLn B: correlation with metastatic potential. Saem-p 212:1151-1153, 1981 Montgomery AMP, De Clerck YA, Langley KE, Reisfeld RA, Mueller BM: Melanoma mediated dissolution of extracellular matrix: contribution of urokinase-dependent and metalloproteinase-dependent proteolytic pathways. Cancer Res 53:693-700, 1993 Capony F, Morisset M, Barrett AJ, Capony JP, Broquet P, Vignon F, Chambon M, Ix)iusot P, Rochefort H: Phosphorylation, glycosylation, and proteolytic activity of the 52 kD estrogen-induced protein secreted by MCF-7 cells. J Cell Biol 104:253-262, 1987

Augereau P, Garcia M, Mattei MG, CavaiUes V, Depadova F, Derocq D, Capony F, Ferrara P, Rochefort H: Cloning and sequencing of the 52 K cathepsin D cDNA of MCF-7 breast cancer cells and mapping on chromosome 11. Mol Endocrinol 2:186-192, 1988 Capony F, Rougeot C, Montcourrier P, Cavailles V, Salazar G, Rochefort H: Increased secretion, altered processing, and glycosylation of pro-cathepsin D in human mammary cancer cells. Cancer Res 49:3904-3909, 1989 Briozzo P, Morisset M, Capony F, Rougeot C, Rochefort H: In vitro degradation of extracellular matrix with Mr 52,000 cathepsin D secreted by breast cancer cells. Cancer i?e5 48:3688-3692, 1988 Thorpe SM, Rochefort H, Garcia M, Freiss G, Christensen IJ, Khalaf S, Paolucci F, Pau B, Rasmussen BB, Rose C: Association between high concentrations of Mr. 52,000 cathepsin D and poor prognosis in primary human breast cancer. Cancer Res 49:6008-6014, 1989 Garcia M, Salazar-Retana G, Pages A, Richer G, Domergue J, Pages AM, CavaUe G, Martin JM, Lamarque JL, Pau B, Pujol H, Rochefort H: Distribution of the Mr 52,000 estrogen-regulated protein in benign breast diseases and other tissues by immunohistochemistry. Cancer Res 46:3734—3738, 1986 Guerra L, Mordoh J, Slavutsky I, Larripa I, Medrano EE: Characterization of IIB-MEL-J: a new and highly heterogeneous human melanoma cell line. Pigment Cell Res 2:504-509, 1989 Soule HD, Vazquez J, Long A, Albert S, Brennan MA: A human cell line from a pleural effusion derived from a breast carcinoma. J Natl Cancer Inst 51:14091413, 1973 Cailleau R, Young R, Olive M, Reeves WJ: Breast tumor cell lines fi-om pleural effusions. J Nari Cancer Inst 53:661-674, 1974 Bover L, Barrio M, Slavutsky I, Bravo Al, Quintans C, Bagnati A, Lema B, Schiaffi J, Yomha R, Mordoh J: Description of a new human breast cancer cell line, IIB-BR-G, established from a primary undifferentiated tumor. Breast Cancer Res Treat 19:47-56, 1991 Takahashi T, Tang J: Cathepsin D from porcine and bovine spleen. Methods Enzymol 80:565-581, 1981 Garcia M, Capony F, Derocq D, Simon D, Pau B, Rochefort H: Characterization of monoclonal antibodies to the estrogen-regulated Mr 52,000 glycoprotein and their use in MCF-7 cells. Cancer Res 45:709-716, 1985 Chomczynski P, Sacchi N: Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162: 156-159, 1987 Basset P, Bellocq JP, Wolf C, StoU I, Hutin P, Limacher JM, Podhajcer OL, Chenard MP, Rio MC, Chambon P: A novel metalloproteinase gene specifically expressed in stromal cells of breast carcinoma. Nature 348:699—704, 1990 Luna LG: Routine staining procedures; melanin pigment. In: Luna LG (ed.).

28.

29. 30.

31.

32.

Blakiston Division, McGraw-Hill Book Company, New York, 1960, p 42 Henry JA, McCarthy AL, Angus B, Westley BR, May FEB, Nicholson S, Cairns J, Harris AL, Home CHW: Prognostic significance of the estrogen-regulated protein, cathepsin D, in breast cancer. An immunohistochemical study. Cancer 65:265-271, 1990 Kute TE, Shao ZM, Sugg NK, Long RT, Russell GB, Case LD: Cathepsin D as a prognostic indicator for node-negative breast cancer patients using both immunoassays and enzymatic assays. Cancer Res 52:5198—5203, 1992 Maudelonde T, Khalaf S, Garcia M, Freiss G, Duporte J, Benatia M, Rogier H, Paolucci F, Simony J, Pujol H, Pau B, Rochefort H: Immunoenzymadc assay of Mr 52,000 cathepsin D in 182 breast cancer cytosols: low correlation with other prognostic parameters. Cancer Res 48:462—466, 1988 Albert LS, Sober AJ: The dysplastic nevus as precursor and marker of increased risk for melanoma. In: Balch CM, Houghton AN, Milton GW, Sober AJ, Soong S-J (eds.). Cutaneous Melanoma. JB Lippincott, Philadelphia, 1992, pp 59-69 Westley B, Rochefort H: A secreted glycoprotein induced by estrogen in human breast cancer cell lines. Cell 20:352-362, 1980 Johnson MD, TorriJA, Lippman ME, Dickson RB: The role of Cathepsin D in the invasiveness of human breast cancer cells. Cancer Res 53:873—877, 1993 Montcourrier PH, Mangeat PH, Salazar G, Morisset M, Sahuquet A, Rochefort H: Cathepsin D in breast cancer cells can digest extracellular matrix in large acidic vesicles. Cancer Res 50:6045-6054, 1990 Baron R, Neff L, Louvard D, Courtoy PJ: Cell-mediated extracellular acidification and bone resorption: evidence for a low pH in resorbing lacunae and localization of a lOO-kD lysosomal membrane protein at the osteoclast ruffled border. J Ce// Biol 101:2210-2222, 1985 Qian F, Bajkowski AS, Steiner DF, Chan SJ, Frankfater A: Expression of five cathepsins in murine melanomas of varying metastatic potential and normal tissues. Cancer Res 49:4870-4875, 1989 Nishimura Y, Kato K: Intracellular transport and processing of lysosomal cathepsin B. Biochem Biophys Res Commun 148:254-259, 1987 Lenarcic B, Kos J, Dolenc 1, Lucovnik P, Krisaj I, Turk V: Cathepsin D inactivates cysteine proteinase inhibitors cystatins. Biochem Biophys Res Commun 154:765-772, 1988 Eeckhout Y, Vaes G: Further studies on the activation of procollagenase, the latent precursor of bone collagenase. Effects of lysosomal cathepsin B, plasmin and kallikrein, and spontaneous activation. Biochem J 166:21—31, 1977 Kobayashi H, Schmitt M, Goretski L, Chucholowski N, Calvete J, Kramer M, Gunzler WA, Janicke F, Graeft H: Cathepsin B efficiently activates the soluble and the tumor cell receptor-bound form of the proenzyme urokinase-type plasminogen activator (Pro-LrPA).y BiW Chem 266:5147-5152, 1991