Pericellular proteolysis by leukocytes and tumor cells on substrates: focal activation and the role of urokinase-type plasminogen activator

Previous studies have shown that the urokinase-type plasminogen activator receptor (uPAR) is localized to the adherence sites of leukocytes and tumor cells suggesting that pericellular proteolysis may accompany focal activation of adherence. To assess for focused pericellular proteolytic activity, we prepared two-dimensional substrates coated with FITC-casein or Bodipy FL-BSA. These molecules are poorly fluorescent, but become highly fluorescent after proteolytic degradation. Fluorescent peptide products were observed at adherence sites of stationary human neutrophils and at lamellipodia of polarized neutrophils. During cell migration, multiple regions of proteolysis appeared sequentially beneath the cell. Similarly, proteolytic action was restricted to adherence sites of resting HT1080 tumor cells but localized to the invadopodia of active cells. Using an extracellular fluorescence quenching method, we demonstrate that these fluorescent peptide products are extracellular. The uPA/uPAR system played an important role in the observed proteolytic activation. Plasminogen activator inhibitor-1 significantly reduced focal proteolysis. Sites of focal proteolysis matched the membrane distribution of uPAR. When uPA was dissociated from uPAR by acid washing, substantially reduced pericellular proteolysis was found. uPAR-negative T47D tumor cells did not express significant levels of substrate proteolysis. However, transfectant clones expressing uPAR (for example, T47D-26) displayed high levels of fluorescence indicating proteolysis at adherence sites. To provide further evidence for the role of the uPA/uPAR system in pericellular proteolysis, peritoneal macrophages from uPA knock-out (uPA–/–) and control (uPA+/+) mice were studied. Pericellular proteolysis was dramatically reduced in uPA-negative peritoneal macrophages. Thus, we have: (1) developed a novel methodology to detect pericellular proteolytic function, (2) demonstrated focused activation of proteolytic enzymatic activity in several cell types, (3) demonstrated its usefulness in real-time studies of cell migration, and (4) showed that the uPA/uPAR system is an important contributor to focal pericellular proteolysis.     

Complement deposition on immune complexes reduces the frequencies of metabolic, proteolytic, and superoxide oscillations of migrating neutrophils.

Neutrophils exhibit intrinsic sinusoidal metabolite concentration oscillations of 3 min in resting cells and an additional approximately 10- or 20-s oscillation in migrating/adhering cells. To better understand immune complex (IC)-mediated leukocyte activation, we have studied neutrophil metabolic oscillations in the presence of ICs either with or without fixed complement. Using a microscope photometer we quantitated NAD(P)H autofluorescence oscillations. Cells exposed to ICs exhibited metabolic oscillation periods of approximately 12 s in the absence of complement and approximately 22 s in the presence of complement opsonization. To determine if the effects could be associated with C3 deposition, we used ICs opsonized with only C3 or only C1 and C4. Untreated ICs, heat-inactivated complement-treated ICs, and C1,C4-treated ICs trigger rapid metabolic oscillations, as do fMLP and yeast; in contrast, ICs treated with full complement or C3 alone did not affect NAD(P)H oscillations in comparison to controls. The induction of higher frequency (approximately 10 s) NAD(P)H oscillations by ICs could be blocked by addition of anti-FcgammaRII, but not FcgammaRIII mAb fragments, suggesting the participation of FcgammaRII in cellular metabolic responses to ICs. Parallel changes in the frequencies of oxidant release and pericellular proteolysis were found for all of these stimuli. Thus, immune complex composition affects both intracellular metabolic signals and extracellular functional oscillations. We suggest that complement attenuates the phlogistic potential of ICs by reducing the frequency of cytoplasmic NAD(P)H oscillations.     

Functional imaging of proteolysis: stromal and inflammatory cells increase tumor proteolysis

The underlying basement membrane is degraded during progression of breast and colon carcinoma. Thus, we imaged degradation of a quenched fluorescent derivative of basement membrane type IV collagen (DQ-collagen IV) by living human breast and colon tumor spheroids. Proteolysis of DQ-collagen IV by HCT 116 and HKh-2 human colon tumor spheroids was both intracellular and pericellular. In contrast, proteolysis of DQ-collagen IV by BT20 human breast tumor spheroids was pericellular. As stromal elements can contribute to proteolytic activities associated with tumors, we also examined degradation of DQ-collagen IV by human monocytes/macrophages and colon and breast fibroblasts. Fibroblasts themselves exhibited a modest amount of pericellular degradation. Degradation was increased 4-17-fold in cocultures of fibroblasts and tumor cells as compared to either cell type alone. Inhibitors of matrix metalloproteinases, plasmin, and the cysteine protease, cathepsin B, all reduced degradation in the cocultures. Monocytes did not degrade DQ-collagen IV; however, macrophages degraded DQ-collagen IV intracellularly. In coculture of tumor cells, fibroblasts, and macrophages, degradation of DQ-collagen IV was further increased. Imaging of living tumor and stromal cells has, thus, allowed us to establish that tumor proteolysis occurs pericellularly and intracellularly and that tumor, stromal, and inflammatory cells all contribute to degradative processes.     

Pericellular proteolysis by neutrophils in the presence of proteinase inhibitors: effects of substrate opsonization

Inflammatory cells are capable of degrading extracellular matrix macromolecules in vivo in the presence of proteinase inhibitors. We and others have hypothesized that such proteolysis is permitted in large part by mechanisms operative in the immediate pericellular environment, especially at zones of contact between inflammatory cells and insoluble matrix components. To further test this hypothesis in vitro, we have used a model system in which viable polymorphonuclear neutrophils (PMN) are allowed to contact a surface coated with proteinase-sensitive substrate, and in which PMN interaction with the surface can be modulated. We have evaluated proteolysis of the surface-bound protein in the presence and absence of proteinase inhibitors. Our results were: (a) In the presence (but not in the absence) of proteinase inhibitors, proteolysis was confined to sharply marginated zones subjacent to the cells; (b) opsonization of the surface enhanced spreading of the PMN, (c) opsonization diminished the effectiveness of alpha-1-proteinase inhibitor (alpha-1-PI) and alpha-2-macroglobulin as inhibitors of proteolysis of surface-bound protein; (d) anti-oxidants did not alter the effectiveness of alpha-1-PI in inhibiting proteolysis of opsonized substrate by PMN; and (e) PMN could restrict entry of alpha-1-PI into zones of contact with opsonized surfaces. We conclude that: (a) In the presence of proteinase inhibitors, PMN can express sharply marginated and exclusively pericellular proteolytic activity; (b) locally high proteinase concentrations and/or exclusion of proteinase inhibitors from pericellular microenvironments may be important mechanisms for pericellular matrix degradation by PMN; and (c) these observations may have general relevance to extracellular matrix remodeling by a variety of inflammatory and other cell types.     

Spontaneous and stimulated transients in cytoplasmic free Ca(2+) in normal human osteoblast-like cells: aspects of their regulation.

We characterize two patterns of transients in cytoplasmic free calcium ([Ca2+]i) in normal human osteoblast-like cells (hOB cells). Firstly, spontaneous oscillations in [Ca2+]i were found to be common. The [Ca2+]i oscillations were completely inhibited by thapsigargin, indicating that Ca2+ fluxes between intracellular Ca2+ pools and the cytosol contributed to the generation of the [Ca2+]i oscillations. Removing extracellular Ca2+ either attenuated or completely inhibited spontaneous [Ca2+]i oscillations. Gadolinium, an inhibitor of stretch activated cation channels (SA-cat channels), reduced the frequency of [Ca2+]i oscillations. Hence, entry of calcium from the extracellular space, possibly through SA-cat channels also seemed to be of importance in the regulation of these [Ca2+]i oscillations. The role of the observed spontaneous [Ca2+]i oscillations in hOB cell function is not clear. Secondly, a decrease in pericellular osmolality, which causes the plasma membrane to stretch, transiently increased [Ca2+]i in hOB cells. This effect was also observed in a Ca2+ free extracellular environment, suggesting that osmotic stimuli release Ca2+ from intracellular pools. This finding indicates a possible signaling pathway by which mechanical strain can promote anabolic effects on the human skeleton.     

Tumor cell invasion of model 3-dimensional matrices: demonstration of migratory pathways, collagen disruption, and intercellular cooperation.

We report a novel 3-dimensional model for visualizing tumor cell migration across a nylon mesh-supported gelatin matrix. To visualize migration across these model barriers, cell proteolytic activity of the pericellular matrix was detected using Bodipy-BSA (fluorescent upon proteolysis) and DQ collagen (fluorescent upon collagenase activity). For 3-dimensional image reconstruction, multiple optical images at sequential z axis positions were deconvoluted by computer analysis. Specificity was indicated using well-known inhibitors. Using these fluorescent proteolysis markers and imaging methods, we have directly demonstrated proteolytic and collagenolytic activity during tumor cell invasion. Moreover, it is possible to visualize migratory pathways followed by tumor cells during matrix invasion. Using cells of differing invasive potentials (uPAR-negative T-47D wild-type and uPAR-positive T-47D A2--1 cells), we show that the presence of the T-47D-A2--1 cells facilitates the entry of T-47D wild-type cells into the matrix. In some cases, wild-type cells follow T-47D A2--1 cells into the matrix whereas other T-47D-wild-type cells appear to enter without the direct intervention of T-47D A2--1 cells. Thus, we have developed a new 3-dimensional model of tumor cell invasion, demonstrated protein and collagen disruption, mapped the pathways followed by tumor cells during migration through an extracellular matrix, and illustrated cross-talk among tumor cell populations during invasion.     

Functional imaging of pericellular proteolysis in cancer cell invasion

Proteolytic interactions between cells and extracellular matrix (ECM) are involved in many physiological and pathological processes, such as embryogenesis, wound healing, immune response, and cancer. The visualization of cell-mediated proteolysis towards ECM is thus required to understand basic mechanisms of tissue formation and repair, such as the breakdown and structural remodelling of ECM, inflammatory changes of tissue integrity, and the formation of proteolytic trails by moving cells. A panel of synergistic techniques for the visualization of pericellular proteolysis in live and fixed samples allow monitoring the of proteolytic tumor cell invasion in three-dimensional (3D) fibrillar collagen matrices in vitro. These include the quantification of collagenolysis by measuring the release of collagen fragments, the detection of protease expression and local activity by dequenching of fluorogenic substrate, and the staining of cleavage-associated neoepitopes together with changes in matrix structure. In combination, these approaches allow the high-resolution mapping of pericellular proteolysis towards ECM substrata including individual focal cleavage sites and the interplay between cell dynamics and alterations in the tissue architecture.     

Regulation of pericellular proteolysis by hepatocyte growth factor activator inhibitor type 1 (HAI-1) in trophoblast cells

Hepatocyte growth factor activator inhibitor type 1/serine protease inhibitor Kunitz type 1 (HAI-1/SPINT1) is a membrane-bound Kunitz-type serine protease inhibitor that is abundantly expressed on the surface of cytotrophoblasts, and is critically required for the formation of the placenta labyrinth in mice. HAI-1/SPINT1 regulates several membrane-associated cell surface serine proteases, with matriptase being the most cognate target. Matriptase degrades extracellular matrix protein such as laminin and activates other cell surface proteases including prostasin. This study aimed to analyze the role of HAI-1/SPINT1 in pericellular proteolysis of trophoblasts. In HAI-1/SPINT1-deficient mouse placenta, laminin immunoreactivity around trophoblasts was irregular and occasionally showed an intense punctate pattern, which differed significantly from the linear distribution along the basement membrane observed in wild-type placenta. To explore the molecular mechanism underlying this observation, we analyzed the effect of HAI-1/SPINT1 knock down (KD) on pericellular proteolysis in the human trophoblast cell line, BeWo. HAI-1/SPINT1-KD BeWo cells had increased amounts of cellular laminin protein and decreased laminin degradation activity in the culture supernatant. Subsequent analysis indicated that cell-associated matriptase was significantly decreased in KD cells whereas its mRNA level was not altered, suggesting an enhanced release and/or dislocation of matriptase in the absence of HAI-1/SPINT1. Moreover, prostasin activation and pericellular total serine protease activities were significantly suppressed by HAI-1/SPINT1 KD. These observations suggest that HAI-1/SPINT1 is critically required for the cell surface localization of matriptase in trophoblasts, and, in the absence of HAI-1/SPINT1, physiological activation of prostasin and other protease(s) initiated by cell surface matriptase may be impaired.     

Immunonanoparticles--an effective tool to impair harmful proteolysis in invasive breast tumor cells

Breast cancer cells exhibit excessive proteolysis, which is responsible for extensive extracellular matrix degradation, invasion and metastasis. Besides other proteases, lysosomal cysteine protease cathepsin B has been implicated in these processes and the impairment of its intracellular activity was suggested to reduce harmful proteolysis and hence diminish progression of breast tumors. Here, we present an effective system composed of poly(D,L-lactide-coglycolide) nanoparticles, a specific anti-cytokeratin monoclonal IgG and cystatin, a potent protease inhibitor, that can neutralize the excessive intracellular proteolytic activity as well as invasive potential of breast tumor cells. The delivery system distinguishes between breast and other cells due to the monoclonal antibody specifically recognizing cytokeratines on the membrane of breast tumor cells. Bound nanoparticles are rapidly internalized by means of endocytosis releasing the inhibitor cargo within the lysosomes. This enables intracellular cathepsin B proteolytic activity to be inhibited, reducing the invasive and metastatic potential of tumor cells without affecting proteolytic functions in normal cells and processes. This approach may be applied for treatment of breast and other tumors in which intracellular proteolytic activity is a part of the process of malignant progression.     

Urokinase-type plasminogen activator and its receptor synergize to promote pathogenic proteolysis

Urokinase-type plasminogen activator (uPA) is a potent catalyst of extracellular proteolysis, which also binds to a high-affinity plasma membrane receptor (uPAR). Binding of uPA may influence pericellular proteolysis and/or activate intracellular signal transduction. Transgenic mice overexpressing either uPA or uPAR in basal epidermis and hair follicles had no detectable cutaneous alterations. In contrast, bi-transgenic mice overexpressing both uPA and uPAR, obtained by crossing the two transgenic lines, developed extensive alopecia induced by involution of hair follicles, epidermal thickening and sub-epidermal blisters. The phenotype was due to uPA catalytic activity since combined overexpression of uPAR and uPAR-binding but catalytically inactive uPA in the same tissue was not detrimental in another bi-transgenic line. It was accompanied by increased plasmin-generating capacity, up-regulation and activation of matrix metalloproteinases type-2 and -9, and cleavage of uPAR. Thus, combined overexpression of uPA and uPAR acts in synergy to promote pathogenic extracellular proteolysis.     

Cyclic AMP and the control of aggregative phase gene expression in Dictyostelium discoideum.

The induction of aggregative phase functions and the acceleration of the onset of aggregation competence by nanomolar pulses of cyclic AMP can be mimicked by exposing developing cells to a high extracellular concentration of either cyclic AMP or cyclic GMP (5 × 10^?4M) during the first 1–2 hr of development. Pulses of cyclic AMP have previously been shown to result in oscillations of intracellular cyclic AMP concentration; we show that high extracellular concentrations of cyclic AMP and cyclic GMP cause intracellular cyclic AMP levels to increase. We describe a mutant, HM11, which has elevated levels of intracellular cyclic AMP from the beginning of development and which begins to accumulate cell-associated phosphodiesterase, an aggregative phase enzyme, within an hour of starvation. Our data suggest that the expression of aggregative phase functions is controlled by an elevation of intracellular cyclic AMP which may be either continuous or periodic.     

A subpopulation of peritoneal macrophages form capillarylike lumens and branching patterns in vitro

OBJECTIVE: We have previously shown that monocytes/macrophages (MC/Mph) influence neovascularization by extracellular matrix degradation, and by direct incorporation into growing microvessels. To date, neither the phenotype of these cells, nor the stages of their capillary-like conversion were sufficiently characterized. METHODS: We isolated mouse peritoneal Mph from transgenic mice expressing fluorescent proteins either ubiquitously, or specifically in the myelocytic lineage. These Mph were embedded in Matrigel which contained fluorescent protease substrates, exposed to an MCP-1 chemotactic gradient, and then examined by confocal microscopy after various intervals. RESULTS: Within 3 hrs after gel embedding, we detected TIMP-1 and MMP-12 dependent proteolysis of the matrix surrounding Mph, mostly in the direction of high concentrations of MCP-1. After 2 days, Mph developed intracellular vacuoles containing degradation product. At 5 days these vacuoles were enlarged and/or fused to generate trans-cellular lumens in approximately 10% of cells or more (depending on animal's genetic background). At this stage, Mph became tubular, and occasionally organized in three-dimensional structures resembling branched microvessels. CONCLUSION: Isolated mouse peritoneal Mph penetrate Matrigel and form tunnels via a metalloprotease-driven proteolysis and phagocytosis. Following a morphological adjustment driven by occurrence, enlargement and/or fusion process of intracellular vacuoles, similar to that described in bona fide endothelium, a subpopulation of these cells end up by lining a capillary-like lumen in vitro. Thus we show that adult Mph, not only the more primitive 'endothelial progenitors', have functional properties until now considered defining of the endothelial phenotype.     

The low density lipoprotein receptor-related protein LRP is regulated by membrane type-1 matrix metalloproteinase (MT1-MMP) proteolysis in malignant cells

We demonstrate that the presentation of LRP and the subsequent uptake of its ligands by malignant cells are both strongly regulated by MT1-MMP. Because LRP is essential for the clearance of multiple ligands, these findings have important implications for many pathophysiological processes including the pericellular proteolysis in neoplastic cells as well as the fate of the soluble matrix-degrading proteases such as MMP-2. MT1-MMP is a key protease in cell invasion and a physiological activator of MMP-2. Cellular LRP consists of a non-covalently associated 515-kDa extracellular alpha-chain (LRP-515) and an 85-kDa membrane-spanning beta-chain, and plays a dual role as a multifunctional endocytic receptor and a signaling molecule. Through the capture and uptake of several soluble proteases, LRP is involved in the regulation of matrix proteolysis. LRP-515 associates with the MT1-MMP catalytic domain and is highly susceptible to MT1-MMP proteolysis in vitro. Similar to MT1-MMP, the metalloproteinases MT2-MMP, MT3-MMP and MT4-MMP also degrade LRP. The N-terminal and C-terminal parts of the LRP-515 subunit are resistant and susceptible, respectively, to MT1-MMP proteolysis. In cells co-expressing LRP and MT1-MMP, the proteolytically competent protease decreases the levels of cellular LRP and releases its N-terminal portion in the extracellular milieu while the catalytically inert protease co-precipitates with LRP. These events implicate MT1-MMP, not only in the activation of MMP-2, but also in the mechanisms that control the subsequent fate of MMP-2 in cells and tissues.     

Differences in domain structure between pericellular matrix and plasma fibronectins as revealed by domain-specific antibodies combined with limited proteolysis and S-cyanylation: a preliminary note

Differences in domain structure between human fibronectins obtained from pericellular matrix and plasma have been revealed by limited proteolysis and S-cyanylation, followed by identification of each domain with domain-specific antibodies. Although the overall domain structure is similar between pericellular and plasma fibronectins, the fragments derived from the COOH-terminal region of these fibronectins, which were defined by specific antibodies, exhibited clear differences in their molecular weights and protease susceptibility, suggesting that the structure near the COOH-terminal region is significantly different between these two proteins.     

Secretion processing and activation of Erwinia chrysanthemi proteases

E chrysanthemi, a phytopathogenic enterobacterium, secretes several enzymes into the medium such as pectinases cellulases and proteases. It also produces 3 distinct and antigenically related extracellular proteases. The proteases secretion pathway seems to be distinct from that of the other extracellular enzymes since pleiotropic mutants impaired in cellulase and pectinase secretion are unimpaired in protease secretion. E chrysanthemi proteases B and C secretion occurs without an N-terminal signal peptide and is dependent upon specific secretion functions which are encoded by genes adjacent to the protease structural genes. This secretion pathway might be analogous to the alpha-hemolysin secretion pathway in E coli. Protection against intracellular proteolytic activity is achieved by 2 distinct mechanisms: the proteases are synthesized as inactive precursors with an N-terminal extension of 15 aminoacids (protease B) and 17 aminoacids (protease C) absent in the mature active extracellular enzymes; an intracellular specific protease inhibitor is produced by some E chrysanthemi strains.     

Inhibition of intracellular proteolysis by insulin in isolated rat hepatocytes. Possible role of internalized hormone

The mechanism of insulin's action upon intracellular proteolysis in isolated hepatocytes was studied. At 37 degrees C insulin inhibited intracellular degradation of intracellular proteins in a dose-dependent manner. A maximal 40% inhibition of intracellular proteolysis was achieved at an insulin concentration of 500 ng/ml with a half-maximal inhibition observed at 2.5 ng/ml of insulin. Insulin inhibited intracellular proteolysis both in the presence and in the absence of amino acids in the incubation mixture. Low concentrations of trypsin (10 micrograms/ml) mimicked insulin's effect upon glucose incorporation into glycogen, but not on intracellular proteolysis. Four protease inhibitors (phenylmethylsulfonyl fluoride (0.5 mM), p-nitrophenyl-p-guanidinobenzoate (0.25 mM), p-tosyl-L-arginine methyl ester (1 mM), and N alpha-p-tosyl-L-lysine chloromethyl ketone (1 mM) blocked the stimulatory effect of insulin upon [14C]glucose incorporation into glycogen, but did not affect the inhibitory action of insulin upon intracellular proteolysis. These results suggest that the mechanism of insulin's action upon intracellular proteolysis differs from that involved in stimulation of glycogenesis. Low temperature (15 degrees C) and short time exposure (10 min) of the hepatocytes to insulin eliminated the inhibitory effect of insulin on intracellular proteolysis. Similarly, insulin's effect on intracellular proteolysis was eliminated by dansylcadaverine, a transglutaminase inhibitor that blocked insulin internalization. In contrast, dansylcadaverine had no effect on insulin's ability to stimulate [14C]glucose incorporation into glycogen. These experiments strongly suggest the necessity of insulin internalization for its inhibitory effect on endogenous protein degradation.     

Activation and silencing of matrix metalloproteinases

Matrix metalloproteinases (MMPs) were first described as proteases that act on protein components of the extracellular matrix. However, subsequent studies of MMP function in vivo have revealed that these proteinases also cleave numerous non-ECM protein substrates. Because their substrates are diverse in functions, MMPs are involved in variety of homeostatic functions, such as tissue repair and immunity, as well as pathological processes, including cancer, fibroses and inflammation. Essential steps in regulating MMP proteolysis are conversion of the zymogen into an active proteinase and subsequent inactivation. A number of mechanisms including proteolysis, allosteric interactions, oxidative modification, pericellular compartmentalization, interaction with tissue inhibitor of metalloproteinases (TIMPs), endocytosis, and more have been proposed to control the activation and inactivation of MMPs. In this paper, we discuss these and other mechanisms, and their relevance to in vivo control of MMP-mediated functions.     

Kinetic characterization and fed-batch fermentation for maximal simultaneous production of esterase and protease from Lysinibacillus fusiformis AU01

The simultaneous production of intracellular esterase and extracellular protease from the strain Lysinibacillus fusiformis AU01 was studied in detail. The production was performed both under batch and fed-batch modes. The maximum yield of intracellular esterase and protease was obtained under full oxygen saturation at the beginning of the fermentation. The data were fitted to the Luedeking–Piret model and it was shown that the enzyme (both esterase and protease) production was growth associated. A decrease in intracellular esterase and increase in the extracellular esterase were observed during late stationary phase. The appearance of intracellular proteins in extracellular media and decrease in viable cell count and biomass during late stationary phase confirmed that the presence of extracellular esterase is due to cell lysis. Even though the fed-batch fermentation with different feeding strategies showed improved productivity, feeding yeast extract under DO-stat fermentation conditions showed highest intracellular esterase and protease production. Under DO-stat fed-batch cultivation, maximum intracellular esterase activity of 820?×?10³ U/L and extracellular protease activity of 172?×?10³ U/L were obtained at the 16th?hr. Intracellular esterase and extracellular protease production were increased fivefold and fourfold, respectively, when compared to batch fermentation performed under shake flask conditions.     

Fibrinolytic activity in a human fibrosarcoma cell line and evidence for the induction of plasminogen activator secretion during tumor formation.

Seven clones were isolated from the HT1080 human fibrosarcoma cell line using a fibrinagarose overlay technique. Three of these clones induced lysis of the fibrin overlay, whereas four did not. The extracellular and intracellular levels of protease were then measured using 125I-fibrin plates incubated with acid-treated human serum. The extracellular protease can be directly assayed in the medium from cells incubated with 10% fetal calf serum. Although there were large differences in the amounts of protease secreted by these two sets of clones, the intracellular levels of protease were similar. No significant differences were found between the abilities of the cells to grow in soft agar or as tumors in immunosuppressed hamsters. However, cells grown from tumors derived from all the low secretors of protease showed an increase in the amount of protease secreted. It appeared, therefore, that the secretion of protease might be selected for or induced during tumor growth. Further detailed studies with one of the low secreting clones (clone E) suggested an inductive rather than a selective mechanism for this increase in extracellular plasminogen activator.