Search

Bortezomib is a proteasome inhibitor that has direct antitumor effects. We and others have previously demonstrated that bortezomib could also sensitize tumor cells to killing via the death ligand, TRAIL. NK cells represent a potent antitumor effector cell. Therefore, we investigated whether bortezomib could sensitize tumor cells to NK cell-mediated killing. Preincubation of tumor cells with bortezomib had no effect on short-term NK cell killing or purified granule killing assays. Using a 24-h lysis assay, increases in tumor killing was only observed using perforin-deficient NK cells, and this increased killing was found to be dependent on both TRAIL and FasL, correlating with an increase in tumor Fas and DR5 expression. Long-term tumor outgrowth assays allowed for the detection of this increased tumor killing by activated NK cells following bortezomib treatment of the tumor. In a tumor purging assay, in which tumor:bone marrow cell mixtures were placed into lethally irradiated mice, only treatment of these mixtures with a combination of NK cells with bortezomib resulted in significant tumor-free survival of the recipients. These results demonstrate that bortezomib treatment can sensitize tumor cells to cellular effector pathways. These results suggest that the combination of proteasome inhibition with immune therapy may result in increased antitumor efficacy.

Optimal differentiation of cytotoxic NK cells is important to provide protective innate immunity to patients after bone marrow transplantation. In vitro differentiation of CD56(+)CD3(-) NK cells takes weeks and is supported by several cytokines, including IL-2, IL-7, and IL-15, and thus can be useful for immunotherapy. However, IL-2 therapy is problematic in vivo, and NK cells differentiated in vitro with only IL-7 lack cytotoxicity. We assessed whether human NK cells initially differentiated in vitro from CD34(+)Lin(-) bone marrow cells with IL-7 could acquire cytotoxicity after exposure to additional cytokines and what changes promoted cytotoxicity. The cells cultured with IL-7 already had granzyme B as well as perforin, as previously reported, the proteins of cytotoxic granules. The cells also lacked LFA-1. After 1 wk of secondary culture with either IL-2 or IL-15, but not with IL-12 or IL-18, the IL-7-cultured cells acquired cytotoxicity. IL-2 or IL-15 also induced LFA-1. Ab to the LFA-1 subunits CD11a and CD18 blocked lysis by the NK cells, indicating that the new LFA-1 correlated with, and was essential for, the cytotoxic function of the in vitro generated cells. The LFA-1 also participated in target cell binding by the in vitro differentiated cells. In this study, we demonstrated a new function for IL-15, the induction of LFA-1 in NK progenitor cells, and that IL-15 does more than merely support NK progenitor cell proliferation. The efficacy after only 1 wk of IL-15 administration is a positive practical feature that may apply to human therapy.

The components within cytotoxic lymphocyte granules are responsible for a significant fraction of T and NK cell-mediated death. Perforin is stored in these granules together with calreticulin. Calreticulin has long been recognized as a chaperone protein of the endoplasmic reticulum (ER) and is the only resident ER protein to be found in the cytotoxic granules. Here we implicate a role for calreticulin in killing and report that it controls osmotic lysis mediated by purified perforin. Calreticulin, at a concentration of 2.2 x 10-7 M, completely blocked perforin-mediated lysis. Inhibition was stable and held over 5 h. Recombinant calreticulin, at a concentration of 8. 8 x 10-7 M, also blocked lysis, indicating the inhibition was due to calreticulin and not a copurifying protein in the native calreticulin preparations. Using calreticulin domain fragments (expressed as GST fusion proteins), we found inhibitory activity in the high-capacity calcium-binding C-domain, which does not bind perforin. The N- or P-domains, which can bind perforin, were unable to block lysis. The inhibition of lysis was independent of granzyme inactivation or the ability of calreticulin to sequester calcium. Our data indicate that calreticulin regulation of perforin-mediated lysis probably occurs without direct interaction with perforin. We propose a novel model in which calreticulin stabilizes membranes to prevent polyperforin pore formation.

One mechanism of killing by cytotoxic lymphocytes involves the exocytosis of specialized granules. The released granules contain perforin, which assembles into pores in the membranes of cells targeted for death. Serine proteases termed granzymes are present in the cytotoxic granules and include several chymases (with chymotrypsin-like specificity of cleavage). One chymase is selectively reactive with an inhibitor, Biotinyl-Aca-Aca-Phe-Leu-PheP(OPh)2, that blocks perforin lysis. We report the purification and characterization of this chymase, lymphocyte chymase I, from rat natural killer cell (RNK)-16 granules. Lymphocyte chymase I is 30 kDa with a pH 7.5 to 9 optimum and primary substrate preference for tryptophan, a preference distinct from rat mast cell chymases. This chymase also reacts with other selective serine protease inhibitors that block perforin pore formation. It elutes by Cu2+-immobilized metal affinity chromatography with other granzymes and has the N-terminal protein sequence conserved among granzymes. Chymase I reduces pore formation when preincubated with perforin at 37 degrees C. In contrast, addition of the chymase without preincubation had little effect on lysis. It should be noted that the perforin preparation contained sufficient residual chymase activity to support lysis. Thus, the reduction of lysis may represent an effect of excess prolytic chymase I or a means to limit perforin lysis of bystander cells. In contrast, other chymases and granzyme K were without effect when added to perforin during similar preincubation. Identification of the natural substrate of chymase I will help resolve how it regulates perforin-mediated pore formation.

Serine proteases (granzymes) in killer lymphocytes are required for lymphocyte cytotoxic granules to lyse target cells. Herein we report the development of a 3-step PCR cloning technique to amplify novel granzyme genes and two new rat granzymes are described. Degenerate oligonucleotide primers were designed based on sequence motifs selectively expressed in granzymes. These motifs flank "delta" regions that are unique for each granzyme. Total RNA of RNK-16 cells or activated splenocytes was amplified by reverse transcriptase-PCR to obtain cDNA fragments of several new granzymes. Gene-specific primers based on these delta regions were then used for 3'-RACE to obtain clones with the 3' gene ends. Reverse (antisense) delta-based or active site serine primers were used with a granzyme 5'-UTR primer to obtain clones extending to the 5' ends. Using this technique, two new cDNAs, RNKP-4 and RNKP-7, which encode granzymes of 248 and 241 amino acids, respectively, were cloned from activated lymphocytes. RNKP-4 is likely the rat equivalent of mouse granzyme C. RNKP-7 is most closely related to granzymes F and G. Modeling of the predicted proteins suggests large/polar P1 (Gln/Asn) specificity for RNKP-4 and large/hydrophobic P1 (e.g., Phe) specificity for RNKP-7. These specific protease activities were found in cytotoxic RNK-16 lymphocyte granules indicating that the two new genes may be translated and stored as active granzymes.

Cytotoxic NK and T lymphocytes kill virally infected cells within minutes without causing damage to themselves or bystander cells. One mechanism of killing involves exocytosis of granules containing serine proteases and perforin. Serine protease inhibitors block killing of target cells mediated by the cytotoxic lymphocytes. There are at least five different serine protease activities in cytolytic granules. Ten different serine protease sequences have been identified with the use of cDNA-specific clones. It is not known whether only one or several of these serine proteases are essential for cytolytic activity. In this study we show that an irreversible serine protease inhibitor, biotinyl-Aca-Aca-Phe-Leu-PheP(OPh)2, selectively inhibits a chymotrypsin-like (chymase) serine protease activity of rat RNK-16 granule extracts. Under the same conditions, only one 30-kDa (reduced) band was detected on protein blots. Furthermore, only one of three chymase peaks separated by hydrophobic interaction chromatography was inhibited. When this granzyme was inhibited, granule-mediated lysis of erythrocytes was diminished. NK cell killing was completely blocked when biotinyl-Aca-Aca-Phe-Leu-PheP(OPh)2 was added to cytotoxicity assays at 200 microM with rat splenocytes as effectors. By confocal fluorescence microscopy, we show that this inhibitor localizes to distinct regions within RNK-16 cells and rat NK cells. Inhibitor treatment of intact cells inactivated the chymase activity and reduced lysis found in their dense organelles. Together these data indicate that biotinyl-Aca-Aca-Phe-Leu-PheP(OPh)2 inhibits a granule chymase that is essential to cytolytic activity of NK cells.

Cartilage degradation, a hallmark of rheumatoid arthritis, is attributed to serine and metalloproteases secreted by neutrophils, synovial lining cells, macrophages, and chondrocytes. A large proportion of synovial fluid lymphocytes contains the granule-associated serine proteases granzymes A and B. We report that lysates of IL-2-stimulated lymphocytes contain an enzymatic activity (ECMase; cartilage extracellular matrix 35S release assay; extracellular matrix degrading activity) that solubilizes matrix synthesized by chondrocyte monolayers. ECMase activity is inactivated by the serine protease inhibitor diisopropylfluorophosphate, is stored in dense granules and cleaves aggrecan proteoglycans but not free glycosaminoglycans, hyaluronic acid, or type II collagen. ECMase is mediated by a cationic protein with biochemical properties identical to granzyme B, inasmuch as it preferentially hydrolyzes the substrate Boc-Ala-Ala-Asp-SBzl, immunochemically cross-reacts with an antibody that binds to a conserved amino-terminal region of lymphoid-myeloid serine proteases, and has amino-terminal sequence identity with human Q31 granzyme B. Using an agarose gel electrophoresis technique to assess cleavage of the rat sarcoma aggrecan, the catalytic efficiency of granzyme B for the digestion of aggrecan (catalytic efficiency = 1.7 x 10(7) M-1 s-1) was 425-fold faster than the catalytic efficiency reported for human stromelysin-1 at pH 7.5 (catalytic efficiency 4000 M-1 s-1) and 3200-fold faster than granzyme A. Based on these observations, we propose that granzyme B, secreted from cytotoxic lymphocytes within the rheumatoid joint, may contribute to cartilage loss by degrading resident aggrecan.

To kill other cells, lymphocytes can exocytose granules that contain serine proteases and pore-forming proteins (perforins). We report that mechanism-based isocoumarin inhibitors inhibited the proteases and inactivated lysis. When inhibited proteases were restored, lysis was also restored, indicating that the proteases were essential for lysis. We found three new lymphocyte protease activities, "Asp-ase,""Met-ase," and "Ser-ase," which in addition to ly-tryptase and ly-chymase, comprise five different protease activities in rat RNK-16 granules. The general serine protease inhibitor 3,4-dichloroisocoumarin (DCI) inhibited all five protease activities. Essentially all protease molecules were inactivated by DCI before lysis was reduced, as determined from DCI's second order inhibition rate constants for the proteases, the DCI concentrations, and the times of pretreatment needed to block lysis. The pH favoring DCI inhibition of lysis was the pH optimum for protease activity. Isocoumarin reagents acylate, and may sometimes secondarily alkylate, serine protease active sites. Granule proteases, inhibited by DCI acylation, were deacylated with hydroxylamine, restoring both the protease and lytic activities. Hydroxylamine does not restore alkylated proteases and did not restore the lytic activities after inhibition with 4-chloro-7-guanidino-3-(2-phenylethoxy)-isocoumarin, a more alkylating mechanism-based inhibitor designed to react with tryptases. It is improbable that isocoumarin reagents directly inactivated pore-forming proteins because 1) these reagents require protease activation, 2) their nonspecific effects are alkylating, and 3) alkylated proteins are not restored by hydroxylamine. We conclude that serine proteases participate in lysis when lysis is mediated by the complete assembly of granule proteins.

From an RNK-16 lambda-gt11 library, we have isolated and sequenced a novel cDNA rat NK cell protease 1 (RNKP-1) that has characteristics unique to serine proteases. The cDNA clone is 1102 bp and contains a complete open reading frame with an AUG start codon and a TAA stop codon. The open reading frame translates into a protein of 248 amino acids that has one glycosylation site. The characteristic N-terminal Ile-Ile-Gly-Gly and the His, Asp, and Ser amino acid residues that form the catalytic triad of serine proteases are present. The nucleotide and amino acid sequences have 87 and 80% identity, respectively, with the murine CTL-specific serine protease CCPI. However, there are extensive differences in the substrate binding regions of these proteases. Comparison of hydropathic profiles and amino acid sequences of other proteases indicate that RNKP-1 is distinct and belongs to the subfamily of serine proteases of bone marrow origin. Northern blot analysis of poly A+ RNA from rat splenocytes cultured with Con A showed 1000 and 1400 nucleotide mRNA are detected with RNKP-1 after 1 day of Con A-stimulation. The expression of the two mRNA bands continues through day 5 of culture with the lectin and may represent RNKP-1 mRNA plus related sequences due to cross-hybridization. RNKP-1 is also expressed in RNK-16 cells, but is not expressed in freshly isolated rat splenocytes, brain, lung, or lymph node tissues. The induction of RNKP-1 expression in the Con A-cultured spleen cells is accompanied by increases in both NK and lymphokine-activated killer lymphocyte activities. These data indicate that RNKP-1 is a unique serine protease that may be preferentially expressed in NK cells.

Human natural cell-mediated cytotoxicity is inhibited by macromolecular protease inhibitors. Human plasma alpha 1 antiproteases are more effective than the plant antiproteases lima bean trypsin inhibitor and soybean trypsin inhibitor for reduction of cytotoxicity to the "slow" targets T24 human bladder carcinoma and NKI-1 melanoma. This inhibition of natural cytotoxicity is more readily demonstrable in serum-free medium containing crystalline bovine serum albumin than in medium containing fetal calf serum. Although electrophoretically homogeneous plasma alpha 1 antitrypsin inhibits natural cytotoxicity, partially purified alpha 1 antitrypsin preparations that contain several apha 1 proteins are more inhibitory at equivalent trypsin inhibitory capacities. Partially purified alpha 1 antichymotrypsin with no antitrypsin activity is an extremely potent inhibitor. Thus, it seems likely that several of the plasma antiproteases, including alpha 1 antitrypsin and alpha 1 antichymotrypsin, are capable of influencing natural cytotoxicity. These data indicate that serine-dependent proteases hae a critical role in triggering and/or effecting cell-mediated cytolysis. Furthermore, since alpha 1 antichymotrypsin and alpha 1 antitrypsin are acute phase proteins, the increase in plasma concentration or turnover rates of the proteins could influence natural killer cell activity in vivo.

K-76 COONa is a 440 m.w. fungal product that can inhibit complement activity of C5 and Factor I. K-76 COONa abrogated both human natural killer (NK) cell activity and antibody-dependent cell-mediated cytotoxicity (ADCC) (ID50 approximately 1.5 mM). To be effective, K-76 COONa had to be present during the assay, because pretreatment of lymphocytes with highly inhibitory concentrations of K-76 COONa did not inhibit cytolysis. The monocarboxylic K-76 derivative was more inhibitory to NK than the dicarboxylic derivative. This relative efficacy is similar to that observed for inhibition of complement lysis. K-76 COONa inhibited NK when added before NK conjugate formation, but had little effect when added after conjugate formation. The compound also inhibited the formation of conjugates by NK and K cells. Therefore, this reagent selectively affected events that occurred between the initial effector-target cell interactions and the formation of stable conjugates. It had little influence on the post-binding "lethal hit" stage of cytolysis. These data imply a) that if any molecules similar to C5 are activated during the "lethal hit" stage of cytolysis, then they are inaccessible to K-76 COONa , and b) that C3bi-like molecules may be involved in lymphocyte binding.

We used reagents specific for serine-dependent proteinases to verify that a proteinase of this class is necessary for natural cytotoxicity (NK). NK was inhibited by phenylmethylsulfonylfluoride (PMSF), by diisopropylfluorophosphate (DFP), and by the plasma antiproteinase alpha-1-antichymotrypsin (alpha-1-X), all of which are specific for serine-dependent proteinases. Substrate specificity was then determined on the basis of the specificity of the plasma and fungal anti-proteinases and synthetic alternate substrates that affected NK. alpha-1-X, which inhibits only serine proteinases with aromatic amino acid specificity, blocked NK. Chymostatin, but not other fungal inhibitors, also blocked NK activity. Furthermore, the only synthetic substrates that effectively reduced NK were those derived from aromatic amino acids. The ester derivatives of these substrates inhibited NK better than the amides. NK inhibition with these alternate substrates was also stereospecific, with the L forms twofold more active than the D forms. These reagents did not block initial lymphocyte-target cell binding. Therefore we propose that the "NK-proteinase" is involved in either the initiation of cytolysis, perhaps as part of stimulus and secretion of cytolytic molecules, or in the cascade of events that may lead to the formation of final lytic substance.