Nancy A. Day

Inhibitory properties of low molecular mass cysteine proteinase inhibitors from human sarcoma.

Elevated activities of cysteine proteinases such as cathepsins B and L and cancer procoagulant have been linked to tumor malignancy. In the present study we examined the hypothesis that these elevated activities could be due to impaired regulation by the endogenous low molecular mass cysteine proteinase inhibitors (cystatins). Inhibitors from human sarcoma were compared to those from human liver, a normal tissue in which the inhibitors had been characterized previously. An extract of cystatins from sarcoma was less effective against papain and cathepsin B (liver or tumor) than was an extract from liver. This reduced inhibitory capacity in sarcoma was not due to a reduction in either the concentrations or specific activities of the cystatins or an absence of any family or isoform of cystatins. We purified two members of the cystatin superfamily (stefin A and stefin B) to homogeneity and determined their individual inhibitory properties. Stefins B from liver and sarcoma exhibited comparable inhibition of papain and cathepsin B. In contrast, stefin A from sarcoma exhibited a reduced ability to inhibit papain, human liver cathepsins B, H and L and human and murine tumor cathepsin B. The Ki for inhibition of liver cathepsin B by sarcoma stefin A was 10-fold higher than that for inhibition of liver cathepsin B by liver stefin A, reflecting a reduction in the rate constant for association and an increase in the rate constant for dissociation. Cancer is now the third pathologic condition reported to be associated with alterations in cystatins, the other two being amyloidosis and muscular dystrophy.

Tumor Cell Membrane Cathepsin B

The lysosomal cysteine peptidase cathepsin B was found to be associated with plasma membrane/endosomal fractions of murine B16 amelanotic melanoma cells. Confocal microscopy with three dimensional image analysis indicated that cathepsin B was associated with the external basal cell surface, which would be consistent with its proposed role in degradation of extracellular matrix proteins. We purified and partially characterized cathepsin B from homogenates of murine liver and B16 amelanotic melanoma cells and from lysosomal and membrane/endosomal fractions of the B16 tumor cells. By SDS-PAGE under reducing conditions, the purified cathepsin B from the tumor homogenates was resolved as a single protein band of Mr 31000, corresponding to the single chain form of cathepsin B. In contrast, cathepsin B from liver homogenates was resolved as two bands of Mr 31000 and 24000, corresponding to the single chain and the heavy chain of the double chain form, respectively. The tumor cathepsin B consisted of four isozymes with pIs of 5.64, 5.33, 5.2 and 5.1, whereas the liver cathepsin B consisted of five isozymes with pIs of 5.64, 5.5, 5.45, 5.35 and 5.3. The additional acidic isoforms of cathepsin B in the B16 tumor probably reflect altered glycosylation in tumors. The commonality of isoforms in the B16 plasma membrane/endosomal and lysosomal fractions suggests that retrograde trafficking of cathepsin B from the lysosome to the endosome and its exocytotic release result in the association of cathepsin B with the tumor cell membrane.

Cathepsin B-Like Cysteine Proteinases and Metastasis

During the process of metastasis tumor cells traverse several extracellular matrix barriers in order to gain entry to the vascular space at the site of the primary tumor and to the perivascular space at the sites of metastatic tumors. A number of in vitro model systems have been designed to study tumor cell invasion through extracellular matrices. Several of these model systems including systems to study the concomitant digestion of these matrices by tumor cells are discussed elsewhere in this volume (see Chapters 18–21). A focus of much recent research is the basement membrane which underlies endothelial cells. The physical location of the basement membrane suggests that it must be traversed for tumor cells to form hematogenous metastases. Since the basement membrane in contrast to other extracellular matrix barriers contains type IV collagen a number of investigators have hypothesized that tumor cells must be able to degrade type IV collagen in order to invade through the basement membrane (see also Chapter 19). Metastatic variants of the B16 melanoma have been shown to secrete a type IV collagenase (1). However, type IV collagenase is secreted by tumor cells in a latent form which requires activation (2) suggesting that type IV collagenase is not by itself sufficient for basement membrane invasion by tumor cells and that additional proteinases or a proteolytic cascade may participate in basement membrane invasion.