Search

Background: Cathepsin B (CB), a cysteine protease, is usually found in perinuclear lysosomes of epithelial cells of normal organs and non-malignant tumors, but is associated with the plasma membranes of many solid organ malignant tumors. Plasma membrane localized CB facilitates degradation of extracellular matrix proteins and progression of tumor cells from one biological compartment to another. The activities of CB and its subcellular distribution have not been investigated in malignant prostate. Our objective was to examine the subcellular distribution of CB by determining the activities of CB in lysosome and plasma membrane/endosome subcellular fractions and its subcellular localization by immunogold electron microscopy., Methods: Prostate tissue pieces obtained immediately after prostatectomy were homogenized and fractionated into subcellular components for determining biochemical activities of CB and cysteine protease inhibitors (CPIs). Distribution of CB was compared with that of prostate specific antigen (PSA, a serine protease), which is abundant in secretory vesicles and granules of normal prostate, benign prostatic hyperplasia (BPH) and malignant prostate cells. Localization of CB was investigated in resin embedded lysosomes and plasma membrane/endosome subcellular fractions and in prostate tissue sections by immunogold electron microscopy., Results: We have demonstrated the specificity of CB activity in human prostate homogenates by using a variety of inhibitors in our assay. We did not find any difference in the specific activity of CB based on protein or DNA content in homogenates of malignant prostate (Gleason histologic scores 5-7) and BPH (no histological evidence of cancer) whether it was measured by chromogenic or fluorogenic peptide substrate assay techniques. We found significantly higher activities of CB in the plasma membrane/endosome fractions of malignant prostate than in BPH. In contrast, CPI activity was increased relative to CB activity in plasma membrane/endosome fraction of BPH versus prostate cancer. Our data indicated a shift in the balance of enzyme to inhibitor that would favor increased activities of CB in prostate cancer. The immunogold microscopic study showed specific localization of CB in plasma membrane. They also showed localization of CB in lysosomes that were often adjacent to luminal and/or basal surfaces of malignant cells in contrast to the usual perinuclear distribution of lysosomes in hyperplastic prostate glands. PSA was localized in secretory granules and vesicles, including the plasma membranes and secretory blebs in malignant prostate cells. Occasional PSA positive secretory vesicles or membrane profiles were seen in the plasma membrane/endosomal and lysosomal fractions., Conclusions: The increased activity of CB in plasma membrane/endosomal fractions is associated with malignant prostate and not with BPH or normal prostate. Morphologic distribution CB is associated with the plasma membranes or lysosomes adjacent to apical and basal cell surfaces. This distribution is characteristic feature prostate cancer cells, but not in BPH or normal prostate cells. Subcellular distribution of PSA occurs in secretory vesicles and granules of the cytoplasm, but not in lysosomes. Our biochemical and morphological data could be used to distinguish malignant prostates from non-malignant tumors., (Copyright 2001 Wiley-Liss, Inc.)

Redistribution of lysosomes to the cell surface and secretion of lysosomal proteases appear to be general phenomena in cells that participate in local proteolysis. In the present study, we have determined whether malignant progression affects the intracellular distribution and secretion of the lysosomal protease cathepsin B in three model systems, each of which consists of cell pairs that differ in their degree of malignancy. The intracellular distribution of vesicles staining for cathepsin B was evaluated by immunofluorescent microscopy and the secretion of cathepsin B was evaluated by two complementary techniques: stopped assays of activity secreted into culture media; and continuous assays of activity secreted from viable (> or = 95%) cells growing on coverslips. We observed that the intracellular distribution of cathepsin B+ vesicles was more peripheral in the cells of higher malignancy in all three model systems and that active cathepsin B was secreted constitutively from these cells. Because an acidic pericellular pH has been shown to induce translocation of lysosomes in macrophages and fibroblasts, we evaluated the intracellular distribution of cathepsin B+ vesicles and secretion of cathepsin B in cell pairs incubated at slightly acidic pH. Acidic pericellular pH induced a redistribution of cathepsin B+ vesicles toward the cell periphery. In the more malignant cells, this resulted with time in reduced intracellular staining for cathepsin B and enhanced secretion of active cathepsin B. Translocation and secretion of cathepsin B were dependent on a functional microtubular system. Both the redistribution of cathepsin B+ vesicles toward the cell surface induced by acidic pH and the constitutive and acidic pH-induced secretion of active cathepsin B could be inhibited by microtubule poisons and stabilizers. We suggest that the redistribution of active cathepsin B to the surface of malignant cells and its secretion may facilitate invasion of these cells.

The poor prognosis of human malignant gliomas is due to their invasion and recurrence, the molecular mechanisms of which remain poorly characterized. We have accumulated substantial evidence implicating the cysteine protease cathepsin B in human glioma malignancy. Increases in cathepsin B expression were observed throughout progression. In primary brain tumor tissue, transcript abundance (Northern blot analysis) increased in low-grade astrocytoma to high-grade glioblastoma from 3- to 6-fold, respectively, above normal brain levels. This increase correlated with increases in protein abundance (from + to ) as measured by immunohistochemistry. Furthermore, in glioblastoma cell lines increases in transcript abundance (ranging from 3- to 12-fold) were accompanied by increases in enzyme activity (44-133 nmol/min x mg protein). Altered subcellular localization was observed both immunohistochemically and by indirect immunofluorescence confocal microscopy and was found to correlate with increased grade. In addition, this increase in cathepsin B expression and altered subcellular localization correlated with histomorphological invasion and clinical evidence of invasion as detected by magnetic resonance imaging. These data support the hypothesis that cathepsin B plays a role in human glioma progression and invasion.

Gelatinase A (MMP-2) and cathepsin B are proteinases which have been proposed to participate in human tumor invasion and metastasis. Precise quantitation of the activity of these enzymes in invading tumors has not been previously described. We utilized a novel tissue microdissection technique to determine levels of enzyme activity in specific microscopic areas of invasive human colon cancer. Tissue specimens smaller than one high power field can be extracted from the samples and analyzed. Increased levels of pro-enzyme and active enzyme forms of gelatinase A (MMP-2) and increased cathepsin B activity were localized in regions of tumor invasion as compared with a matched number of normal epithelial cells from the same patient. Levels of progelatinase B (MMP-9) were also increased in the tumors; however, we did not observe activation of this enzyme. To investigate the mechanism of gelatinase A activation, we amplified DNA of specific microdissected tumor cell populations using polymerase chain reaction. We did not detect a mutation in the activation locus of the enzyme in any of the tumors studied, which suggests that activation may be due to up-regulation of a tumor-associated gelatinase A activating species. Microdissection of frozen tissue sections may prove valuable in the study of proteinases in human tumor invasion as well as in the detection of genetic alterations in human cancers.

The process of tumor cell invasion of the basement membrane is proposed to consist of three steps: attachment, local proteolysis and migration. 12-(S)-HETE, a 12-lipoxygenase metabolite of arachidonic acid, upregulates surface expression of integrin cytoadhesins and an autocrine motility factor receptor, suggesting that this metabolite may play an important regulatory function in tumor cell invasion. In the present study, we determined whether 12-(S)-HETE affects surface expression and/or release of cathepsin B, a cysteine protease that has been implicated in focal degradation of basement membrane. Secretion and distribution of cathepsin B was evaluated in two model systems for various stages of neoplastic progression: (i) murine B16 melanoma lines of low (B16-F1) and high (B16a) lung colonization potential, and (ii) immortalized and ras-transfected MCF-10 human breast epithelial cells that differ in their invasive capacities in vitro. In the B16a cells, 12-(S)-HETE induced release of native and latent cathepsin B activity and concomitantly reduced cell-associated cathepsin B immunoreactivity. In contrast, 12-(S)-HETE did not induce the release of cathepsin B from B16-F1 cells, suggesting that there may be an enhanced response to 12-(S)-HETE in more malignant cells. This was confirmed in the MCF-10 system, in which 12-(S)-HETE was able to induce the release of cathepsin B from the ras-transfected cells, but not from the immortal cells. A simultaneous reduction in staining for cathepsin B was observed in the ras-transfected cells, but not in their immortal counterparts. The release of cathepsin B may be mediated by PKC as pretreatment of B16a cells with the selective PKC inhibitor calphostin C, but not with the PKA inhibitor H8, prevented the stimulated release of cathepsin B. In B16a cells, the release of cathepsin B was accompanied by a translocation toward the cell periphery of vesicles staining for cathepsin B, resulting in focal areas of accumulation of cathepsin B. After 12-(S)-HETE stimulation of the ras-transfected MCF-10 cells, cathepsin B was distributed homogeneously on the apical surface. Thus, 12-(S)-HETE can upregulate the surface expression on tumor cells of proteins able to mediate each of the three steps of tumor cell invasion: adhesion, degradation, and migration.

Cathepsins B and L and their endogenous inhibitors were evaluated in rat embryo fibroblast lines which have been developed as a model system for the study of malignant progression and metastatic capability. Three groups of lines were analyzed: 1) immortalized/non-tumorigenic, 2) tumorigenic/metastatic lines transfected with c-Ha-ras, and 3) metastatic revertants transfected with c-Ha-ras+the E1A region of adenovirus type 2. The metastatic revertants are tumorigenic, but non-metastatic. No correlation was seen between tumorigenicity and metastatic potential and the level of expression of cathepsin B or the subcellular distribution of cathepsins B and L. However, cathepsin L activity was increased 2-fold in the 4R metastatic line. Although transfection of aneuploid 3T3 fibroblasts with ras has been shown to increase the expression of cathepsin L and cathepsin B, transfection of the diploid rat embryo fibroblasts with ras did not correlate with increased expression of cathepsin L or cathepsin B. However, ras transfection of the rat embryo fibroblasts was associated with a significant (4-15-fold) decrease in the activity of heat-stable cysteine endopeptidase inhibitors. Thus, in tumorigenic rat embryo fibroblast lines, regulation of the activities of cysteine endopeptidases by their endogenous inhibitors may be compromised, resulting in increased effective activities of the cysteine endopeptidases.

Expression, membrane association and secretion of both cathepsin B and cathepsin L have been associated with the malignant phenotype of murine B16 melanomas. Only native forms of the two enzymes were found in the lysosomal fractions, whereas both native and latent forms were found in the membrane fractions. Brief exposure to acid pH induced secretion of only native forms of both cathepsin B and cathepsin L and a concomitant reduction in membrane-associated activities. Thus, the pericellular acidification associated with malignant tumors may provide optimal conditions for proteolysis by the cysteine proteinases in terms of their enhanced stability and their induced release in native forms not requiring proteolytic activation.

Several lysosomal proteinases including the cysteine proteinase cathepsin B have been implicated in malignant progression of tumors. Many investigators have demonstrated correlations between increased activity of cathepsin B and increased metastatic capability of animal tumors or malignancy of human tumors. Such increases in cathepsin B activity in malignant tumors may reflect alterations in synthesis, in activation and processing, and/or in intracellular trafficking and delivery as well as in the endogenous inhibitors of cathepsin B. Increases in mRNA transcripts for cathepsin B have been observed in both murine and human tumors and multiple transcripts for cathepsin B have been identified, but an association of multiple transcripts with malignancy has not been confirmed. Cathepsin B precursors found in human malignant ascites fluid do not possess mannose-rich carbohydrates suggesting that a defect in the post translational processing of carbohydrate moieties on tumor cathepsin B may be responsible for the release of cathepsin B observed in many tumor systems. However, the intracellular trafficking of cathepsin B responsible for its association with plasma membrane/endosomal systems and for its release will require further study as both latent, precursor forms of cathepsin B and native forms of cathepsin B are involved. We speculate that malignant tumor cells adherent to basement membrane are capable of forming a digestive microenvironment in which lysosomal proteinases such as cathepsin B function optimally, a microenvironment similar to that formed between adherent osteoclasts and bone. One of the endogenous cysteine proteinase inhibitors, stefin A, also is affected by malignancy. Reduced expression (mRNA and protein) of stefin A is found as well as a reduction in its inhibitory capacity against cysteine proteinases. The data to date at both the molecular and protein levels supporting a functional role(s) for cathepsin B and its endogenous inhibitors in cancer progression are only correlative. Experimental approaches utilizing well-defined model systems in conjunction with genetic manipulation of cathepsin B and its endogenous inhibitors are needed to provide convincing evidence that cathepsin B has an important role in cancer.

Our laboratories have previously demonstrated that the malignancy of human and animal tumors is associated with increases in cathepsin B activity, due in part to increases in cathepsin B-specific RNA transcripts and in part to decreased regulation by the endogenous low molecular weight cysteine proteinase inhibitors (CPIs). In this study we have extended these observations to tumor cell subpopulations of B16 amelanotic melanoma (B16a) and Lewis lung carcinoma (3LL) isolated by centrifugal elutriation. B16a subpopulations exhibited a 10-fold differential in lung colonization potential, whereas 3LL subpopulations exhibited no differential. In the B16a subpopulations, cathepsin B activities, total cellular and plasma membrane-associated, corresponded positively (4- and 10-fold increase, respectively) with their lung colonization potentials. CPI activities, total cellular and plasma membrane-associated, corresponded inversely (2- and 5-fold decrease, respectively) with the lung colonization potential of the B16a subpopulations. In the 3LL subpopulations, neither cathepsin B nor CPI activities changed. In the plasma membrane fractions of all 3LL subpopulations the ratio of cathepsin B activity to CPI activity was less than 1, whereas in the plasma membrane fractions of all B16a subpopulations the ratio was 1 or greater. In the plasma membrane fractions of the B16a subpopulations of higher lung colonization potential the ratios were 2.5 and 7, indicating that the levels of endogenous CPIs in these fractions may not be sufficient to regulate cathepsin B activity. Cathepsin B mRNA levels were not increased in the B16a subpopulations expressing increased cathepsin B activity. Thus increased cathepsin B activity in these subpopulations was apparently due not to increased synthesis but to decreased regulation by the endogenous CPIs. These results suggest that membrane-associated cathepsin B and CPIs may both play a role in the expression of the experimental metastatic phenotype.

Activities of a cathepsin B-like cysteine proteinase have previously been observed to correlate with the malignancy of several animal and human tumors. Plasma membrane fractions of some of these tumors have been found to be enriched in cathepsin B-like activity. We have determined the subcellular distribution of this enzyme and three additional lysosomal hydrolases (cathepsin H, beta-hexosaminidase, and beta-glucuronidase) in normal murine liver and six metastatic variants of the B16 melanoma. The tissues were fractionated initially by differential centrifugation followed by Percoll density gradient centrifugation of the light mitochondrial fraction. Two fractions were obtained: an L-2 fraction enriched in all four lysosomal hydrolases; and an L-1 fraction enriched in a marker enzyme for the plasma membrane. Cathepsin B-like and beta-hexosaminidase activities, but not the other hydrolase activities, were also found to be enriched in the L-1 fractions of the metastatic B16 tumors. We explored the nature of the association of the cathepsin B-like activity with the plasma membrane using fractions from the spontaneously metastatic B16 amelanotic melanoma. Activity could not be dissociated from the plasma membrane fraction by washing with a physiological salt solution suggesting that it was not adsorbed to this fraction nonspecifically, nor could it be displaced by mannose 6-phosphate or other sugars which compete for binding to the known lysosomal receptors. High salt concentrations, low concentrations of the mild detergent saponin, mild acidification, or phosphatidylinositol-specific phospholipase C did not elute the cathepsin B-like activity. However, activity was eluted by exposure to 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate, a detergent used in the purification of integral membrane proteins. The B16 amelanotic melanoma plasma membrane-associated cathepsin B-like activity had a slightly higher pH optimum and was resistant to inactivation by neutral pH and to inhibition by three low molecular weight inhibitors of cysteine proteinases. The Ki values for inhibition by leupeptin and stefin A were 20-fold higher. The presence of a cathepsin B-like cysteine proteinase at the surface of metastatic tumor cells, particularly in a form which can retain activity at physiological pH and retain activity in the presence of extracellular proteinase inhibitors, may contribute to the focal dissolution of the extracellular matrix observed at sites of contact with invading tumor cells.

A practical synthesis of 3'-phosphoadenosine 5'-phosphosulfate (IV) in yields of 68-72% from adenosine 2',3'-cyclic phosphate 5'-phosphate (II) is described. Reaction of II with triethylamine-N-sulfonic acid affords adenosine 2',3'-cyclic phosphate 5'-phosphosulfate (III) which, on treatment with ribonuclease-T2, provides IV. Spleen phosphodiesterase, on the other hand, converts III to 2'-phosphoadenosine 5'-phosphosulfate (V). The biological activity of IV, measured by sulfate transfer to [6,7-3H2]estrone as mediated by bovine adrenal estrone sulfotransferase (3'-phosphoadenylyl-sulfate:estrone 3-sulfotransferase, EC 2.8.2.4), is identical with that obtained with a sample of IV prepared by an established biochemical procedure. By contrast, V exhibits approximately one-third the activity of the natural isomer.

Analogs of adenosine 3',5'-diphosphate are described which aid in the characterization of the inhibition of estrone sulfurylation by 3'-phosphoadenosine 5'-phosphosulfate as mediated by bovine adrenal estrogen sulfotransferase (3'-5'-phosphosulfate:estrone 3-sulphotransferase, EC 2.8.2.4). The facile conversion of ribonucleosides to 2',3'-cyclic phosphate 5'-phosphate in neat pyrophosphoryl chloride is utilized to provide a reliable route to the requisite intermediates for enzymatic regiospecific conversion to ribonucleoside 3',5'- and 2',5'-diphosphates. The importance of the 3'-phosphate ester to inhibition of estrone sulfurylation is confirmed by Ki measurements. Replacement of the 6-amino group by hydrogen or oxygen leads to considerable loss in affinity for the enzyme as does also dimethylation of the exocylic amino group. Alterations in the pyrimidine ring are not well tolerated by the sulfotransferase but modifications in the imidazole ring as in tubercidin (7 -deazaadenosine) and 8-bromoadenosine 3',5'-diphosphate lead to an enhanced affinity. The latter findings are discussed in terms of an hypothesis of stacking of the aromatic ring of the estrogen substrate and the purine moiety and its analogs.

The ability of tumor cells to invade into and through normal tissue during the metastatic cascade has been attributed to tumor-associated degradative enzymes including proteinases of the metallo, serine and cysteine classes. Work from several laboratories has established that the cysteine proteinases cathepsins L and B are released from tumor cells, primarily as latent precursor forms. In addition, a cathepsin B-like cysteine proteinase has been shown to be associated with the plasma membrane fraction of several animal and human tumors. This form of the enzyme retains activity under physiologic (or pathologic) conditions including at neutral pH and in the presence of low Mr inhibitors. Since we have established that cathepsin B can degrade the basement membrane attachment glycoprotein laminin, we speculate that plasma membrane-associated cathepsin B may participate in focal dissolution of the basement membrane during tumor cell extravasation.

We have investigated the effect of age, a high-fat diet, sodium deoxycholate, and the ornithine analogue alpha-difluoromethylornithine on ornithine decarboxylase (ODC) activity in the rat colon. The relative levels of ODC activity were also determined in normal mucosa and tumor tissue from rat and human colon. The colonic ODC activity induced by intrarectal instillation of sodium deoxycholate in male Sprague-Dawley rats was highest in young animals, and it decreased with increasing age. A high level of dietary fat caused both an increased in basal colonic ODC activity and enhanced ODC induction by deoxycholate. alpha-Difluoromethylornithine given in drinking water inhibited, in a dose-dependent fashion, deoxycholate-induced ODC activity. The frequency of azoxymethane-induced intestinal tumors was also significantly reduced by alpha-difluoromethylornithine. Since colonic ODC activity is increased in carcinogenesis by known promoting agents and decreased by tumor inhibitors, this short-term assay may provide a useful system for identifying colon tumor promoters and inhibitors. The ODC activity in colon tumors of Sprague-Dawley rats was found to be significantly higher than in normal-appearing mucosa in the same animals. Similarly, ODC activity in human colon cancer was found to be higher than that of the normal-appearing mucosa in the same specimen. These results strengthen the utilization of the rat model for studies, the results of which may apply to the human situation.

Subcellular distribution of cathepsin L, the major protein released by transformed or ras transfected fibroblasts, was examined in murine liver, murine B16 amelanotic melanoma and human A2058 melanoma after sequential differential and Percoll density gradient centrifugation. In both murine and human melanomas, cathepsin L activity was found to be enriched in plasma membrane fractions; cathepsin L in these fractions was in both native and acid activatable forms. Plasma membrane fractions from B16 melanoma subpopulations of "low" and "high" metastatic potential were assayed for activity of cathepsin L and of heat stable endogenous inhibitors. The relative specific activity of cathepsin L was 7-fold greater in the subpopulation of "high" metastatic potential, whereas cysteine proteinase inhibitory activity was 5-fold less. Since cathepsin L can degrade intact basement membrane, this membrane-associated cathepsin L may well contribute to metastatic spread of melanomas.

4-Nitroestrone 3-methyl ether has been shown to be an effective growth inhibitor of certain dimethylbenz(a)anthracene-induced rat mammary tumors in intact or ovariectomized rats. When administered at optimum levels (24 mg/kg daily), this A-ring-substituted estrone displayed no toxicity, slight estrogenicity, and an antitumor activity which was comparable to that of tamoxifen and nafoxidine and was surpassed only by ovariectomy or pharmacological doses of 17 beta-estradiol 3-benzoate. In addition, the appearance of mammary tumors was prevented when this estrogen derivative was administered to rats just prior to or after dimethylbenz(a)anthracene intubation. Unique to the action of the methyl ether of 4-nitroestrone on mammary tumors was the destruction of adenocarcinomas while permitting the appearance of fibroadenomas. Systemically, 4-nitroestrone 3-methyl ether brought about focal atrophy within the pituitary and ovaries while causing moderate hypertrophy of the uterus. Plasma prolactin was unaffected.

Relative amounts of mRNA for cathepsin B were measured in normal murine liver and three murine tumors, an invasive liver tumor (hepatoma, Hepa cl 9) and two melanoma variants (B16-F1 and B16 amelanotic melanoma, B16a). Using a human cDNA to the cathepsin B coding region as a hybridization probe, we detected two species of cathepsin B specific RNA transcripts (2.2 and 4.1 kb) in total RNA preparations of all four tissues. The concentrations of the 2.2 and 4.1 kb species were 3.6 and 2.7-fold greater in the highly metastatic B16a melanoma than in normal liver. The concentration of the 2.2 kb species in the invasive hepatoma was 1.7-fold greater than in normal liver. The increased levels of the 2.2 kb message were reflected in increases in activity of cathepsin B in both Hepa cl 9 and B16a.