Proteolytic enzymes, implicated to play a potential role in cell-mediated cytotoxicity, have been focused upon in many investigations of cytolytic T cells, Natural Killer (NK) cells and lymphokine activated killer (LAK) cells. The observation that large granular lymphocytes (LGL) account for NK cell activity and that LGL can be enriched to a high degree of purity have allowed for the use of enriched populations of NK cells for biochemical studies. The granzymes are the most widely examined of these killer cell-associated proteolytic enzymes. This collection of neutral, serine proteases are found in the cytolytic granules of cloned, murine cytolytic T-cells. Proteases have also been discovered in the lytic granules of human, cloned cytotoxic T-cells, rat NK cells lines and LAK cells from both human and murine sources. The biochemical properties of A-NKP-1 and A-NKP-2 coupled with their high molecular weight and substrate specificities led us to consider that this protease complex was related to multicatalytic proteinase complex.

The biochemical properties of heretofore uncharacterized extracellular proteases of rat interleukin-2-activated natural killer (A-NK) cells were investigated. Following p-aminobenzamidine- and D-phenylalanine-agarose affinity chromatography both trypsin- and chymotrypsin-like protease activities were found. Zymography revealed "trypsin-like" activities, with apparent molecular weights of approximately 42 kDa to 20.5 kDa, and "chymotrypsin-like" activities of approximately 27 kDa, to 22 kDa. These proteases might contribute to diverse NK cell functions, and the characteristics of these proteolytic activities are compared with those of granzymes of lytic granules and the proteasome of A-NK cells.

We have recently described nongranular, cytosolic, high‐molecular‐weight trypsin‐like (A‐NKP 1) and chymotrypsin‐like (A‐NKP 2) proteases of interleukin‐2–activated rat natural killer (A‐NK) cells. A functional correlation between the inactivation of A‐NKP 2 and the inhibition of rat A‐NK cell‐mediated cytotoxicity was found. Herein we describe the 6,000‐fold purification of A‐NKP 2 to apparent homogeneity following: isopycnic sucrose gradient fractionation of postnuclear supernatants, molecular sieve chromatography, and heparin‐Sepharose® chromatography. We also report the novel finding that A‐NKP 2 as well as A‐NKP 1, derived from either rat A‐NK cells or the rat NK leukemic cell line CRNK‐16, are constituents of the multicatalytic proteinase (MCP/proteasome) complexes of these cells. Characteristic biochemical, biophysical, and electron microscopic/ultrastructural similarity to the rat liver proteasome was observed. However, Western blot analysis using polyclonal antibodies to the rat liver proteasome clearly indicated differences in the rat hepatic proteasome and the CRNK‐16–derived proteasomal subunits. The identification, characterization, and purification of A‐NKP 1 and A‐NKP 2, described herein, now allow for further investigation of the potential role of these proteasome components in NK cell function. Moreover, the proteasome of NK and A‐NK cells can now be compared and contrasted to the granzymes of lytic granules with respect to their role in cell‐mediated cytotoxicity. © 1994 Wiley‐Liss, Inc.

Our investigations indicate that a variety of neutral serine proteases exist in highly purified, IL-2-activated rat NK (A-NK) cells. These enzymatic activities are not restricted to only cytolysin-containing granules and are not defined by only the assay of N-alpha-benzyloxycarbonyl-L-lysine thiobenzylesterase activity. These activities, which we term A-NKP 1, A-NKP 2, A-NKP 3, and A-NKP 4, cleave, respectively, the following fluorogenic peptide substrates: Boc-Phe-Ser-Arg-7-amino-4-methylcoumarin (AMC, trypsin-like); Suc-Ala-Ala-Phe AMC (chymotrypsin-like); Suc-Gly-Pro-Leu-Gly-Pro AMC (collagenase-like), and Z-Phe-Arg AMC (another trypsin-like enzyme). The proteases A-NKP 1, A-NKP 2, and A-NKP 3 are not cell surface-associated and appear to be cytosolic as defined by isopycnic sucrose density gradient centrifugation. In contrast, A-NKP 4 appears to be located in lysosomes. Treatment of rat A-NK cells with protease inhibitors that inhibit A-NKP 2 and A-NKP 3 also substantially inhibit A-NK cell-mediated cytotoxicity against both NK-sensitive and -resistant targets (YAC-1 and P815, respectively). These results indicate that A-NKP 2 and A-NKP 3 may play a role in IL-2-activated NK cell-mediated cytotoxicity. A variety of proteolytic enzymes, in addition to granzymes, therefore exist in A-NK cells. Our studies indicate that a prerequisite to a thorough understanding of the role of proteases in killer cell function is the investigation of several classes of enzymes in addition to granzymes contained in lytic granules.

While close contact between lymphokine-activated killer (LAK)/adherent, lymphokine-activated killer (A-LAK) cells and tumor cells is believed to be a prerequisite for initiating the events leading to tumor cell lysis, clear evidence for the ability of these effector cells to infiltrate tumors or tumor metastases in vivo still has to be obtained. In the present study, we report that a significant fraction of adoptively transferred A-LAK cells, labeled with fluorochromes for identification, accumulates in lung and liver metastases of the B16 melanoma, the MCA 102 sarcoma and the Lewis lung carcinoma lines. Thus, 5- to 10-fold higher numbers of A-LAK cells were found in the malignant lesions compared to the surrounding normal tissue. The infiltration seemed very heterogeneous after intravenous injection of moderate numbers of A-LAK cells (15 x 10(6)). However, after adoptive transfer of 45 million A-LAK cells, an A-LAK cell/tumor cell ratio higher than 1:1 in most metastases was observed. Surprisingly, approximately 5% of the lung metastases seemed totally resistant to infiltration even though neighboring metastases were highly infiltrated.While substantial infiltration of lung metastases was seen after iv. injection, significant infiltration of liver metastases was seen only after intraportal injection of the A-LAK cells indicating impaired traffic of intravenous injected A-LAK cells through the lung capillaries.These results present direct evidence that A-LAK cells, upon a proper route of administration, have the potential to migrate to and heavily infiltrate metastases from murine tumors of different origin.