Purification and characterization of a vimentin-specific protease in mouse myeloid leukemia cells. Regulation during differentiation and identity with cathepsin G.

Strong vimentin-degrading activity was found in a mouse myelomonocytic leukemic cell line, M1. When M1 cells were induced to differentiate into macrophage-like cells, this degrading activity decreased, while expression of the vimentin gene increased as reported previously [Tsuru, A., Nakamura, N., Takayama, E., Suzuki, Y., Hirayoshi, K. and Nagata, K. (1990) J. Cell Biol. 110, 1655-1664]. This activity was not due to calpain, which was reported to degrade vimentin, because it was independent of the presence or absence of Ca2+. This activity was revealed to be strongly associated with membranes by differential-centrifugation experiments. To identify this protease, purification of the degradation enzyme was performed. A membrane fraction was prepared and extracted with a buffer containing Triton X-100, then subjected to column chromatography using carboxymethyl-Sepharose and heparin-Sepharose. Quantitative analysis using the purified protease revealed that the specificity of this protease was more than 1000-fold higher for vimentin than for bovine serum albumin, ovalbumin and actin. Four protein bands expressing the activity were finally identified by SDS/PAGE. Amino-terminal sequences of these four proteins were identical, suggesting lower-molecular-mass proteins were degradative products. Furthermore, it was revealed that the sequence had the highest similarity with that of human cathepsin G. This result was consistent with the cathpsin-G-like properties of the purified protease, such as the optimum pH and the specificities for inhibitors. The purified protease degraded a synthetic substrate for cathespin G, succinyl-alanyl-alanyl-prolyl-phenylalanyl-p-nitroanilide, with a comparable specific activity to human cathespin G and was specifically detected with anti-(human cathepsin G) serum in immunoblot analysis. The purified protease thus belongs to the 'cathepsin G family', and perhaps is a mouse homologue of human cathepsin G.