Cloning of two genes that are specifically expressed in activated cytotoxic T lymphocytes

Toward a Molecular Understanding of Adaptive Immunity: A Chronology, Part III

Early reports on T cell antigen receptor (TCR) signaling uncovered a rapid increase in intracellular calcium concentration and the activation of calcium-dependent protein kinase as necessary for T cell activation. Cytolytic T cell clones were instrumental in the discovery of intracellular cytolytic granules, and the isolation of the perforin and granzyme molecules as the molecular effectors of cell-mediated lysis of target cells via apoptosis. Cytolytic T cell clones and TCR cDNA clones were also instrumental for the generation of TCR transgenic animals, which provided definitive evidence for negative selection of self-reactive immature thymocytes. In addition, studies of TCR complex signaling of immature thymocytes compared with mature T cells were consistent with the interpretation that negative selection occurs as a consequence of the incapacity of immature cells to produce IL-2, resulting in cytokine deprivation apoptosis. By comparison, taking advantage of cloned TCRs derived from T cell clones reactive with male-specific molecules, using TCR transgenic mice it was possible to document positive selection of female thymocytes when the male-specific molecules were absent. Focusing on the molecular mechanisms of T cell "help" for the generation of antibody-forming cells following the path opened by the elucidation of the IL-2 molecule, several groups were successful in the identification, isolation, and characterization of three new interleukin molecules (IL-4, IL-5, and IL-6) that promote the proliferation and differentiation of B cells. In addition, the identification of a B cell surface molecule (CD40) that augmented B cell antigen receptor-stimulated proliferation and differentiation led to the discovery of a T cell activation surface molecule that proved to be the CD40-ligand, thus finally providing a molecular explanation for "linked or cognate" recognition when T cells and B cells interact physically. Accordingly, the decade after the generation of the first T cell clones saw the elucidation of the molecular mechanisms of T cell cytotoxicity and T cell help, thereby expanding the number of molecules responsible for adaptive T cell immunity.

Construction and characterization of novel, completely human serine protease therapeutics targeting Her2/neu

Immunotoxins containing bacterial or plant toxins have shown promise in cancer-targeted therapy, but their long-term clinical use may be hampered by vascular leak syndrome and immunogenicity of the toxin. We incorporated human granzyme B (GrB) as an effector and generated completely human chimeric fusion proteins containing the humanized anti-Her2/neu single-chain antibody 4D5 (designated GrB/4D5). Introduction of a pH-sensitive fusogenic peptide (designated GrB/4D5/26) resulted in comparatively greater specific cytotoxicity although both constructs showed similar affinity to Her2/neu-positive tumor cells. Compared with GrB/4D5, GrB/4D5/26 showed enhanced and long-lasting cellular uptake and improved delivery of GrB to the cytosol of target cells. Treatment with nanomolar concentrations of GrB/4D5/26 resulted in specific cytotoxicity, induction of apoptosis, and efficient downregulation of PI3K/Akt and Ras/ERK pathways. The endogenous presence of the GrB proteinase inhibitor 9 did not impact the response of cells to the fusion construct. Surprisingly, tumor cells resistant to lapatinib or Herceptin, and cells expressing MDR-1 resistant to chemotherapeutic agents showed no cross-resistance to the GrB-based fusion proteins. Administration (intravenous, tail vein) of GrB/4D5/26 to mice bearing BT474 M1 breast tumors resulted in significant tumor suppression. In addition, tumor tissue excised from GrB/4D5/26-treated mice showed excellent delivery of GrB to tumors and a dramatic induction of apoptosis compared with saline treatment. This study clearly showed that the completely human, functionalized GrB construct can effectively target Her2/neu-expressing cells and displays impressive in vitro and in vivo activity. This construct should be evaluated further for clinical use.

A quarter century of granzymes

Granzymes (Grs) were discovered just over a quarter century ago. They are produced by cytotoxic T cells and natural killer cells and are released upon interaction with target cells. Intensive biochemical, genetic, and biological studies have been performed in order to study their roles in immunity and inflammation. This review summarizes research on the family of Grs.

A novel lineage-specific hypersensitive site is essential for position independent granzyme B expression in transgenic mice

The granzyme B gene is activated upon cytotoxic T cell stimulation and the protein is a key inducer of apoptosis in target cells. Previous studies have identified important proximal regulatory regions but these proved insufficient to drive expression in vivo. We identified a DNase1 hypersensitive site (HS2) 3.9kb upstream of the transcription start site that was present in stimulated but not resting CD8+ cells. The CTL line CTLL R8 was stably transfected with GFP reporter constructs and showed consistently higher fluorescence values when HS2 was included. In transgenic mice the presence of the relevant region of DNA resulted in inducible, CTL-specific transcription of the transgene in all transgenic founder lines analyzed. Deletion of HS2 resulted in a 10-fold reduction in expression. This is the first report of a major distal regulatory element in the control of granzyme B transcription.

A molecular view of cytotoxic T lymphocyte induced killing

Cytotoxic T lymphocytes (CTLs) search out and destroy pathogenic cells, such as those infected with viruses. The biochemistry laboratory at the University of Alberta (Edmonton, Alta.) studies the molecular mechanisms used by these effectors, and this review covers research on this topic primarily from this group. Research there began with the discovery of the granzyme genes and the realization that granzyme B (GrB) had an unusual substrate specificity. Cleavage at aspartate residues gave us the clue that caspases, key regulators of apoptosis, were important substrates. However, it is now clear that mitochondria are also important in controlling granzyme-induced apoptosis. This led to the discovery that the proapoptotic member of the Bcl2 family, Bid, is also activated by GrB. Cleaved Bid then translocates to the mitochondria, resulting in the release of antagonists of inhibitors of apoptosis proteins. The evolution of our understanding of the molecular basis of CTL killing is presented.Key words: cytotoxic T Lymphocyte, granzyme, apoptosis, mitochondria, caspase.

Characterisation of two serine protease inhibitors expressed in the pituitary gland

Serine protease inhibitors (serpins) are a family of structurally related proteins that play key roles in the regulation of proteolytic homeostasis. We have isolated a novel intracellular serpin, termed raPIT5a, from the rat pituitary gland. Northern blot analysis indicated raPIT5a mRNA expression in a range of tissues, including the adrenal gland and the brain. In situ hybridisation histochemistry revealed raPIT5a mRNA expression in specific cell populations in the rat pituitary gland, adrenal gland, and pancreas. Based on sequence similarities to other intracellular serpins, we predicted raPIT5a may inhibit the pro-apoptotic serine protease granzyme B. We confirmed this experimentally by identification of a stable inhibitory complex between granzyme B and raPIT5a. To determine whether granzyme B or granzyme B-related enzymes were expressed in the rat pituitary gland, we performed PCR using primers predicted to amplify granzyme B and two other published granzyme sequences. We identified rat natural killer protease-1 (RNKP-1), the rat homologue of granzyme B, and a novel putative serine protease highly similar to granzyme-like protein III (GLP III), which we termed GLP IIIa. These data suggest raPIT5a may regulate apoptosis in the pituitary by inhibition of granzyme B or GLP IIIa, or members of the caspase enzyme family which have similar substrate specificity. We have also identified expression of a second serpin, called neuroserpin, in pituitary tissue and found that it alters the morphology of the AtT20 corticotrope cell line, presumably through changes in cell adhesion. These results identify new roles for serpins in pituitary cell function.

Granzyme B: a marker of risk for influenza in institutionalized older adults

Risk for influenza increases with age while cellular immune responses decline. This was a prospective study to determine the relationship between cytokine and granzyme B levels in peripheral blood mononuclear cells stimulated with live influenza virus, and subsequent influenza illness. Granzyme B levels were lower in the group who later developed symptomatic laboratory-confirmed influenza (n=10) compared to the group who did not (n=90) (ANOVA, P=0.024). In contrast, none of the cytokine levels were related to the development of influenza. Thus, granzyme B is a potential marker of influenza risk in older adults.

Noncaspase proteases in apoptosis

Biochemical and genetic analysis of apoptosis has determined that intracellular proteases are key effectors of cell death pathways. In particular, early studies have pointed to the primacy of caspase proteases as mediators of execution. More recently, however, evidence has accumulated that noncaspases, including cathepsins, calpains, granzymes, and the proteasome complex, also have roles in mediating and promoting cell death. An important goal is to understand the importance of distinct noncaspases in various forms of apoptosis, and to determine whether pathways mediated by noncaspase proteases intersect with those mediated by caspases. In this review the roles of noncaspase proteases in the biochemistry of apoptosis will be discussed.

Dual mechanisms of apoptosis induction by cytotoxic lymphocytes

Cytotoxic T lymphocytes and natural killer cells together comprise the means by which the immune system detects and rids higher organisms of virus-infected or transformed cells. Although differing considerably in the way they detect foreign or mutated antigens, these cells utilize highly analogous mechanisms for inducing target cell death. Both types of effector lymphocytes utilize two principal contact-dependent cytolytic mechanisms. The first of these, the granule exocytosis mechanism, depends on the synergy of a calcium-dependent pore-forming protein, perforin, and a battery of proteases (granzymes), and it results in penetration by effector molecules into the target cell cytoplasm and nucleus. The second, which requires binding of FasL (CD95L) on the effector cell with trimeric Fas (CD95) molecules on receptive target cells, is calcium independent and functions by generating a death signal at the inner leaflet of the target cell membrane. Exciting recent developments have indicated that both cytolytic mechanisms impinge on an endogenous signaling pathway that is strongly conserved in species as diverse as helminths and humans and dictates the death or survival of all cells.

Mechanisms and management of acute renal allograft rejection

We used RT-PCR for the molecular characterization of human renal graft rejection. The studies showed that intragraft display of mRNA encoding cytotoxic attack molecule granzyme B, and immunoregulatory cytokines IL-10 or IL-2 are correlates of acute rejection, and intrarenal expression of TGF-1 mRNA, of chronic rejection. The current immunosuppressive protocol involves the use of multiple drugs, each directed at a discrete site in the T-cell activation cascade and each with distinct side effects. The immunosuppressants can be classified as inhibitors of: transcription (CsA, tacrolimus); nucleotide synthesis (azathioprine, mycophenolate mofetil, and mizoribine); growth factor signal transduction (sirolimus); and differentiation (DSG). Polyclonal antibodies and monoclonal antibodies directed at cell surface proteins are quite effective as induction therapy or anti-rejection drugs.

Nuclear transport of granzyme B (fragmentin-2). Dependence of perforin in vivo and cytosolic factors in vitro

Cytotoxic T and natural killer cells are able to kill their target cells through synergistic action of the pore-forming protein perforin and the serine protease granzyme B, resulting in very distinctive nuclear changes typical of apoptosis. Whereas perforin acts at the membrane, granzyme B appears to be both capable of entering the cytoplasm of target cells and accumulating in isolated nuclei. In this study we examine nuclear transport of fluoresceinated granzyme B both in vivo in intact cells in the presence of perforin and in vitro in semi-permeabilized cells using confocal laser scanning microscopy. Granzyme B alone was observed to enter the cytoplasm of intact cells but did not accumulate in nuclei. In the presence of sublytic concentrations of perforin, however, it accumulated strongly in intact cell nuclei to levels maximally about 1.5 times those in the cytoplasm after about 2.5 h. In vitro nuclear transport assays showed maximal levels of nuclear and nucleolar accumulation of granzyme B of about 2.5- and 3-fold those in the cytoplasm. In contrast to signal-dependent nuclear accumulation of SV40 large tumor antigen (T-Ag) fusion proteins in vitro, nuclear/nucleolar import of granzyme B was independent of ATP and not inhibitable by the non-hydrolyzable GTP analog GTPgammaS (guanosine 5'-O-(3-thiotriphosphate)). Similar to T-Ag fusion proteins, however, granzyme B nuclear and nucleolar accumulation was dependent on exogenously added cytosol. Specific inhibitors of granzyme B protease activity had no effect on nuclear/nucleolar accumulation, implying that proteolytic activity was not essential for nuclear targeting. The results imply that granzyme B (32 kDa) may be transported from the cytoplasm to the nucleus through passive diffusion and accumulate by binding to nuclear/nucleolar factors in a cytosolic factor-mediated process. Active and passive nuclear transport properties were normal in the presence of unlabeled granzyme B, implying that the nuclear envelope and pore complex are not granzyme B substrates.

Functional roles for granzymes in murine epidermal gamma(delta) T-cell-mediated killing of tumor targets

Granzymes, a family of serine proteases contained in cytoplasmic granules of cytotoxic T lymphocytes and natural killer cells, play a critical role in killing tumor targets by triggering rapid breakdown of DNA and subsequent apoptosis. We have reported previously that dendritic epidermal T cells, which are skin-specific members of the tissue-type gamma(delta) T-cell family in mice, are capable of killing selected tumor cell lines. Here we report that short-term cultured dendritic epidermal T-cell lines contain significant N-alpha-benzyloxycarbonyl-L-Lys-thiobenzyl esterase activity, produce granzyme A protein, and express constitutively mRNA for granzymes A and B. Messenger RNA expression for granzyme B was also confirmed in freshly procured Thy-1+ epidermal cells (i.e., dendritic epidermal T cells). Finally, preincubation of dendritic epidermal T cell lines with a granzyme inhibitor, dichloroisocoumarin, but not with a cysteine protease inhibitor, E-64, abrogated completely their capacity to trigger DNA breakdown in YAC-1 target cells. These results reinforce the concept that dendritic epidermal T cells represent skin-resident killer cells that share several functional properties with conventional killer leukocytes, thereby playing a local immunosurveillance role against tumor development.

Cleavage of CPP32 by granzyme B represents a critical role for granzyme B in the induction of target cell DNA fragmentation

Cytotoxic T lymphocytes (CTLs) are able to recognize and destroy target cells bearing foreign antigen using one of two distinct mechanisms: granule- or Fas-mediated cytotoxicity. The exact mechanisms involved in the induction of apoptotic cell death remain elusive; however, it seems likely that a family of cysteine proteases related to interleukin-1beta converting enzyme are involved. One family member, CPP32, has been identified as an intracellular substrate for granzyme B, a CTL-specific serine protease responsible for the early induction of target cell DNA fragmentation. Here we use cytolytic cells from granzyme B-deficient mice to confirm that cleavage and activation of CPP32 represents a nonredundant role for granzyme B and that this activation plays a role in the induction of DNA fragmentation in target cells, a signature event for apoptotic cell death. A peptide inhibitor of CPP32-like proteases confirmed the function of these enzymes in fragmentation. 51Cr release was not suppressed under these conditions, suggesting that granzyme B cleavage of CPP32 is primarily involved in the induction of DNA fragmentation and not membrane damage during CTL-induced apoptosis.

Cloning and expression of the recombinant mouse natural killer cell granzyme Met-ase-1

Met-ase-1 is a 30 000 Mr serine protease (granzyme) that was first isolated in the cytolytic granules of rat CD3(-) large granular lymphocytes. We screened a mouse genomic library with rat Met-ase-1 cDNA, and obtained bacteriophage clones that contained the mouse Met-ase-1 gene. The mouse Met-ase-1 gene comprises five exons spanning approximately 5.2 kilobases (kb) and exhibits a similar structural organization to its rat homologue and a family of neutrophil elastase-like serine proteases. Mouse Met-ase-1 mRNA was only detected in total cellular and poly A mRNA of mouse CD3(-) GM1(+) large granular lymphocytes derived from splenocytes stimulated with IL-2 and the mouse NK1.1(+) cell line 4 - 16. Spleen T-cell populations generated by Concanavalin A stimulation and a number of mouse pre-NK and T cell lines did not express mouse Met-ase-1 mRNA. The 5' flanking region of the mouse Met-ase-1 gene also shares considerable regions of identity with the 5' flanking region of the rat Met-ase-1 gene. A 3.3 kb segment of 5' sequence flanking the mouse Met-ase-1 gene was inserted upstream of the chloramphenicol acetyltransferase reporter gene and this construct transiently transfected into a variety of mouse and rat large granular lymphocyte leukemia and T-cell lines. The transcriptional activity of the mouse Met-ase-1 5' flanking region was significant in the RNK-16 large granular lymphocyte leukemia, strongest in the 4 - 16 mouse NK1.1(+) cell line, and weak in several mouse pre-NK cell lines. Reverse transcriptase polymerase chain reaction of mouse large granular lymphocyte mRNA was used to derive the full-length coding sequence for mouse Met-ase-1. The predicted hexapropeptide of mouse Met-ase-1 (Asn-6 to Gln-1), was deleted by polymerase chain reaction mutagenesis to enable expression of active mouse Met-ase-1 in mammalian COS-7 cells. Northern blot analysis and protease assays of transfected COS cell lysates against a panel of thiobenzyl ester substrates formally demonstrated that the mouse Met-ase-1 gene encodes a serine proteinase that hydrolyzes substrates containing a long narrow hydrophobic amino acids like methionine, norleucine, and leucine in the P1.

Anti-CD3-activated killer T cells: interferon-gamma and interleukin-10 cross-regulate granzyme B expression and the induction of major histocompatibility complex-unrestricted cytotoxicity

We have investigated the effect of interferon-gamma (IFN-gamma) and interleukin (IL)-10 on granzyme B expression and the induction of major histocompatibility complex (MHC)-unrestricted cytotoxic activity in mouse T cell cultures following activation with anti-CD3 monoclonal antibody (mAb). First, metabolic inhibitors of granule-dependent and granule-independent cytolytic pathways were used to show that anti-CD3-activated killer T (AK-T) cells kill allogeneic P815 mastocytoma target cells primarily by the granule-dependent granzyme/perforin pathway. In comparison to control AK-T cells, lower levels of cytolytic activity were evident when AK-T cells were generated in the presence of anti-IFN-gamma neutralizing mAb or exogenous IL-10, whereas enhanced cytotoxicity was observed when AK-T cell cultures contained anti-IL-10 neutralizing mAb or exogenous IFN-gamma. In addition, granzyme B mRNA expression by AK-T cells was diminished when IFN-gamma bioactivity was neutralized or exogenous IL-10 was present in AK-T cell-cultures, whereas neutralization of IL-10 bioactivity or the addition of exogenous IFN-gamma resulted in increased expression of granzyme B mRNA. Similar results were obtained when granzyme B enzymatic activity in AK-T cell lysates was quantified using a colorimetric granzyme B assay. Altered cytotoxic potential, granzyme B mRNA expression, and granzyme B enzymatic activity following T cell activation in the presence of anti-IFN-gamma or anti-IL-10 neutralizing mAb or exogenous IFN-gamma or IL-10 could not be attributed to gross changes in T cell activation status or to altered percentages of CD4+ and CD8+ T cells in AK-T cell cultures. We conclude that IFN-gamma and IL-10 cross-regulate the induction of the granule-dependent cytolytic machinery of AK-T cells.

Perforin and lymphocyte-mediated cytolysis

We have discussed in the previous sections the recent progress made toward elucidating the regulatory mechanism of perforin gene transcription and the domain structure of the perforin molecule. It appears that the expression of perforin is, at least partially, controlled at the transcription level through the interaction between killer cell-specific cis- and trans- acting factors. One of such cognate pairs, NF-P motif (an EBS-homologous motif) and NF-P2 (a killer cell-specific DNA-binding protein), has been described. The regulatory mechanism of gene transcription, however, is likely to involve multiple factors which act in a coordinated fashion to bring about the most efficient expression of perforin limited strictly to activated killer lymphocytes. Through studies using synthetic peptides and recombinant perforins, it has been suggested that the N-terminal region of the perforin molecule is an important, though not the only, domain responsible for the lytic activity. Further studies are warranted to elucidate the role(s) of other potential amphiphilic structures located in the central portion of the perforin molecule in the overall pore-forming activity. The molecular basis underlying the resistance of killer lymphocytes to perforin-mediated lysis still remains an open question. Preliminary results, however, suggest that the surface protein(s) restricted to killer cells may account for their self-protection against perforin. Based on recent studies using perforin-deficient mice, the involvement of perforin in lymphocyte-mediated cytolysis both in vivo and in vitro has been confirmed. Two functional roles, a direct (lytic) and an indirect (endocytosis enhancer; conduit), both of which may contribute critically to the cell-killing event can be attributed to perforin. The fact that lymphocytes may also employ perforin-independent killing mechanism(s), e.g. Fas-dependent pathway, is beyond the scope of this review. There is, nevertheless, no doubt that these alternative cytolytic mechanisms may also play important roles in immune effector and/or regulatory responses associated with killer lymphocytes. Obviously, we are still a long way from concluding on the functional relevance of each individual cytolytic mechanism seen in different physiopathological situations. Suffice it to say, however, that a wealth of information on lymphocyte-mediated killing has already emerged through the multidisciplinary efforts conducted in our and other laboratories that promise to further dissect this complicated event in the years to come.

Transcription of granzyme A and B genes is differentially regulated during lymphoid ontogeny

During development, thymocytes express a number of genes typical for activated peripheral T lymphocytes, including granzymes. We have now analyzed by reverse transcription-polymerase chain reaction (RT-PCR), immunohistochemistry, and cytochemistry fetal liver cells and thymocytes at various developmental stages for the expression of granzyme A-G genes. At days 13-17 of gestation, only granzyme B but none of the other granzymes is expressed in fetal liver. In the most immature, Pgp-1+IL2R alpha-, thymocyte subpopulation mRNAs for granzymes A-C but not for granzymes D-G are detectable. Upon further differentiation via Pgp-1-IL-2R alpha + into more mature Pgp-1-IL-2R alpha- thymocytes the level of expression of granzymes A, B, and C gradually declines reaching its lowest level at the CD4+ 8+ double positive stage. In fetal thymic lobes depleted of lymphoid cells by treatment with deoxyguanosine, no transcripts for granzymes A, B, and C were found indicating that the PCR signals are derived exclusively from early precursor T/natural killer (NK) lineage cells rather than from residual stromal elements. In mature CD4+CD8- and CD4-CD8+ thymocytes, granzyme B mRNA is found at similar levels in both subsets whereas granzyme A mRNA is expressed selectively in the CD4-CD8+ subset. Enzymatic activity of granzyme A was only seen in a fraction of CD4-CD8+ thymocytes negative for heat stable antigen (HSA) but not in the more immature HSA+ fraction of CD4-CD8+ thymocytes. The data suggest that (a) granzyme B is a pro-thymocyte marker for all T/NK lineage cells; (b) granzyme A transcripts are associated with thymocytes with the potential to develop into the CD8+ lineage; and (c) granzyme A enzymatic activity is only expressed in the most mature CD4-CD8+ stage, suggesting that granzyme proteins are not involved in early stages of thymocyte development.

Conversion of the substrate specificity of mouse proteinase granzyme B

Mouse granzyme B is the prototypic member of a subfamily of serine proteinases expressed in cytolytic lymphocytes. Molecular modelling of granzyme B indicated that the side chain of Arg 208 partially fills the specificity pocket, thus predicting the preference of this enzyme for substrates containing acidic side chains, a feature unique among eukaryotic serine proteinases. Replacement of Arg 208 with glycine results in an enzyme lacking this activity, but which is able to hydrolyze hydrophobic substrates. These results demonstrate unequivocally that the substrate preference of granzyme B is determined by a positive charge in the specificity pocket and also represent one of the few examples of rational and efficient alteration of serine proteinase substrate-specificity following a single amino acid substitution.

Identification and promoter activity of DNase I hypersensitive sites in the region of the Hox-1.3 gene

The Hox genes encode transcriptional regulatory proteins that play a critical role in rostrocaudal specification in the developing embryo. The genes lie in four clusters in the mouse and human genome and are arranged such that a colinear relation exists between a gene's position in the cluster and the time of activation of the gene's expression. We have analyzed the Hox-1.3 region within the Hox-1 gene cluster for DNase I hypersensitive sites to identify putative regulatory sequences. Fragments identified in this way were then analyzed for transcriptional activity using gene transfer experiments in embryonal carcinoma (EC) cells. Three DNase I hypersensitive sites were identified, one of which includes the Hox-1.3 promoter and another, located 550 bp upstream, which enhances the Hox-1.3 promoter activity. The third occurs in the intron and may represent a Hox binding site. Significantly, the DNase I hypersensitive site pattern of this region of the Hox-1 cluster is not altered when F9 stem cells are differentiated with retinoic acid, suggesting that sequential activation of Hox genes by retinoic acid is not due to a sequential opening of the chromatin structure in the Hox gene region.

Regulation of perforin gene expression in a T cell hybrid with inducible cytolytic activity

A mouse x rat T cell hybrid (PC60) that does not require interleukin (IL)-2 for proliferation, was used as a model to study regulation of perforin gene expression. Perforin mRNA is barely detectable in non-induced PC60 cells; however, a 30-fold induction is observed after stimulation with IL-1 alone. Peak perforin mRNA levels were reached after 10 h of induction with IL-1, and these levels were maintained for as long as the stimulus was present. IL-2 by itself has no detectable effect. However, in combination with IL-1 it shows the same induction kinetics as IL-1 alone for the first 10 h, subsequently there is synergism (100-fold induction) between IL-1 and IL-2. The induction response was mainly due to increased transcriptional rates of the perforin gene, and require newly synthesized proteins. The half-life of perforin mRNA in this system is about 5 h. In addition, we confirm the existence of two types of mouse perforin mRNA that differ in their 5' untranslated regions, and show evidence that both mRNA are translated in vivo with similar efficiencies.

Purification of three cytotoxic lymphocyte granule serine proteases that induce apoptosis through distinct substrate and target cell interactions

We recently reported the purification of a lymphocyte granule protein called "fragmentin," which was identified as a serine protease with the ability to induce oligonucleosomal DNA fragmentation and apoptosis (Shi, L., R. P. Kraut, R. Aebersold, and A. H. Greenberg. 1992. J. Exp. Med. 175:553). We have now purified two additional proteases with fragmentin activity from lymphocyte granules. The three proteases are of two types; one has the unusual ability to cleave a tripeptide thiobenzyl ester substrate after aspartic acid, similar to murine cytotoxic cell protease I/granzyme B, while two are tryptase-like, preferentially hydrolyzing after arginine, and bear some homology to human T cell granule tryptases, granzyme 3, and Hanukah factor/granzyme A. Using tripeptide chloromethyl ketones, the pattern of inhibition of DNA fragmentation corresponded to the inhibition of peptide hydrolysis. The Asp-ase fragmentin was blocked by aspartic acid-containing tripeptide chloromethyl ketones, while the tryptase fragmentins were inhibited by arginine-containing chloromethyl ketones. The two tryptase fragmentins were slow acting and were partly suppressed by blocking proteins synthesis with cycloheximide in the YAC-1 target cell. In contrast, the Asp-ase fragmentin was fast acting and produced DNA damage in the absence of protein synthesis. Using a panel of unrelated target cells of lymphoma, thymoma, and melanoma origin, distinct patterns of sensitivity to the three fragmentins were observed. Thus, these three granule proteases make up a family of fragmentins that activate DNA fragmentation and apoptosis by acting on unique substrates in different target cells.

Differential expression of granzyme A and granzyme B proteases and their secretion by fresh rat natural killer cells (NK) and lymphokine-activated killer cells with NK phenotype (LAK-NK)

Granzymes A-G are a family of serine proteases localized in the cytoplasmic granules of cytotoxic T lymphocytes (Masson, D. et al. Cell 1987. 49: 679) and granzyme A is secreted by T lymphocytes in response to antigenic stimulation. Granzyme A is also expressed by natural killer (NK) and lymphokine-activated killer (LAK) cells. Here we show that fresh rat NK cells constitutively express granzyme B and that granzyme A and granzyme B are differentially regulated in unstimulated NK cells vs. LAK cells with NK phenotype (LAK-NK cells). We also show that both granzymes A and B are secreted in a calcium-dependent manner, by NK and LAK-NK cells in response to stimuli which trigger NK cell functions.

A natural killer cell granule protein that induces DNA fragmentation and apoptosis

We report the purification from a rat natural killer (RNK) large granular lymphocyte leukemia of a 32-kD granule protein that induces rapid DNA fragmentation and apoptosis. The protein, which we have called "fragmentin," was capable of causing DNA from intact YAC-1 cells to be cleaved into oligonucleosomal-sized fragments and producing severe chromatin condensation within 1 h. Amino acid sequence of tryptic peptides indicated that fragmentin was highly homologous to the NK and T cell granule serine proteases RNK protease 1 and mouse cytotoxic T cell protease I (CCPI)/granzyme B. Preincubation with the serine esterase inhibitor 3,4-dichloroisocoumarin blocked fragmentin-induced DNA damage, but had no effect on cytolysin. Fragmentin activity against four lymphoma target cells was completely dependent on the presence of cytolysin. Fragmentin produced rapid membrane damage as well as DNA fragmentation at nonlytic cytolysin doses, suggesting that fragmentin activity was not limited to its effects on the nucleus. Fragmentin and cytolysin activity were completely inhibited by EGTA, indicating the process was Ca2+ dependent. A role for cytolysin in endocytosis of fragmentin was suggested by the observation that treatment of YAC-1 with cytochalasin B or sodium azide and 2-deoxyglucose blocked DNA fragmentation but not cytolysin activity. A 30-kD N alpha-CBZ-L-lysine thiobenzyl esterase, which copurified with fragmentin, was inactive on its own but was able to synergistically amplify the DNA damage induced by fragmentin in the presence of cytolysin. Fragmentin activity was not dependent on protein synthesis, as cycloheximide treatment of YAC-1 cells did not prevent DNA damage. We postulate that fragmentin is the molecular mediator of NK cell-mediated DNA fragmentation and apoptosis.

Transcription of two cytotoxic cell protease genes is under the control of different regulatory elements

Precursor cytotoxic T lymphocytes are activated upon interaction with antigen and interleukin 2. The development of the mature killer cell phenotype is achieved by the transcription of a number of function related genes including those encoding a family of cytotoxic cell proteases (CCP). Two of these proteases, CCP1 and CCP2 encoded by C11 and B10, are shown in this report to contain cell-specific transcriptional regulatory elements within their 5'-flanking regions. Two positive regulatory sequences were mapped for C11 (-682 to -427 and -243 to -112) and one for B10 (-279 to -189). In addition each flanking region contains a region of DNA (B10 -1617 to -1049 and C11 -427 to -243) that has a negative influence on transcription. The positive regions do not appear to correspond to any previously characterized regulatory elements but do map to the same region as DNase I hypersensitive sites. When ligated to heterologous promoters these elements can still stimulate transcription but cell-specificity of expression is lost. In addition the conbination of positive regulatory region and promoter is important for the stimulatory effect. The ability of these regulatory sequences to function and to determine cell-specific transcription does not appear to be an intrinsic property but also depends upon the context.

Age-related decrement in cytotoxic T lymphocyte (CTL) activity is associated with decreased levels of mRNA encoded by two CTL-associated serine esterase genes and the perforin gene in mice

The age-related decline in cytotoxic T lymphocyte (CTL) activity has been recognized for many years. Age-related alterations in several immunologic events have been suggested to be partly or completely responsible for this decline. We had previously demonstrated (Bloom et al., Cell. Immunol. 1988. 144: 440) in mice that a deterioration in the lytic mechanism may be at least in part responsible for the decline in CTL activity. We now report that this decline correlates with an age-related decrease in serine esterase activity released into the supernatant medium in the process of generating CTL. Northern analyses were then used to examine the effect of age on expression of genes encoding for perforin and two CTL-associated serine esterases. The products of all three of these genes have all been postulated to play roles in CTL-mediated lysis. We show that the expression of all three of these genes appears to decline with age in the process of generating allogeneic CTL. These alterations in gene expression correlated both with diminished cytolytic and released esterase activities generated by mixed leukocyte culture in spleen cells of old mice compared to young. The age-related decline in gene expression could not be attributed to shifts in T cell subsets, but CD8+ cells generated by allogeneic stimulation of nylon wool-passed spleen cells from old mice expressed significantly less cytolytic activity than those from young. This report is the first demonstration of an age-related decrease in expression of a functionally related group of genes. In addition, these findings are compatible with the suggested roles for perforin and serine esterase release in CTL-mediated target cell lysis.

A cluster of hematopoietic serine protease genes is found on the same chromosomal band as the human alpha/delta T-cell receptor locus

The chymotrypsin-like family of serine protease genes includes several members that are expressed exclusively in subsets of hematopoietic cells. For example, human neutrophil elastase and cathepsin G are expressed only in myelomonocytic precursors, and cytotoxic-T-cell serine proteases are found only in cytotoxic lymphocytes. We have used a cathepsin G cDNA probe to clone two cathepsin G-like genes (designated CGL-1 and CGL-2) from a human genomic library. We have determined that CGL-1 is identical to a previously identified gene (known as CCPI, CTLA I, or cytotoxic serine protease B) that is expressed only in activated cytotoxic T lymphocytes. We show here that cathepsin G, CGL-1, and CGL-2 are linked on an approximately 50-kilobase locus found on human chromosome 14 at band q11.2. This gene cluster maps to the same chromosomal band as the alpha and delta T-cell receptor genes; this region is involved in most chromosomal translocations and inversions that are specifically associated with T-cell malignancies.

Transcriptional regulation of two cytotoxic T lymphocyte-specific serine protease genes

The expression of two serine proteases is induced by antigenic stimulation in cytotoxic T lymphocytes. Using nuclear run-on analysis the increase in steady state mRNA level has been shown to correspond to transcriptional activation. However, the two genes appear to be sequentially rather than coordinately induced. Both genes were shown to be more sensitive to DNase I digestion than a beta-globin gene in cytotoxic T cells. In addition, for the cytotoxic cell protease 1 gene the 5' region of the gene was more sensitive than the 3' end. Two DNaseI hypersensitive sites were seen in the 5' flanking sequences of both genes. The DNA sequences of the upstream regions of both genes were determined and compared. Although the two flanking sequences are overall quite dissimilar, there are short regions which are shared between the two CTL-protease genes. A number of these have been implicated in regulating the expression of other T cell genes.

Complexity of the primary genetic response to mitogenic activation of human T cells

We describe the isolation and characterization of more than 60 novel cDNA clones that constitute part of the immediate genetic response to resting human peripheral blood T cells after mitogen activation. This primary response was highly complex, both in the absolute number of inducible genes and in the diversity of regulation. Although most of the genes expressed in activated T cells were shared with the activation response of normal human fibroblasts, a significant number were more restricted in tissue specificity and thus likely encode or effect the differentiated functions of activated T cells. The activatable genes could be further differentiated on the basis of kinetics of induction, response to cycloheximide, and sensitivity to the immunosuppressive drug cyclosporin A. It is of note that cyclosporin A inhibited the expression of more than 10 inducible genes, which suggests that this drug has a broad genetic mechanism of action.

Murine lymphocytes with natural killer activity express CTL-derived serine protease genes

Serine protease genes (C11, B10 and HF) derived from activated cytolytic T lymphocytes have been shown to be important in CTL-mediated cytotoxicity. In this study, we examined the expression of these genes in fresh natural killer (NK) cells from severe combined immunodeficiency (SCID) and athymic nude mice, as well as in T-cell lines with NK activity. All of these serine protease genes were expressed in NK cells freshly isolated from SCID and athymic nude mice. In addition, all lines showed similar strong levels of expression of C11 and B10 genes, but not the HF gene. However, levels of expression in the T-cell lines did not correlate with levels of NK-like cytotoxicity. These results suggest that C11, B10 and HF serine protease genes are necessary but not sufficient for NK-like cytotoxicity.

A functional and phenotypic comparison of murine natural killer (NK) cells and lymphokine-activated killer (LAK) cells

Lymphokine-activated killer (LAK) cells have been defined as interleukin 2 (IL-2)-activated cytolytic effector cells exhibiting non-MHC restricted killing against a wide range of NK-sensitive and NK-resistant tumor cells. There has been considerable debate as to whether LAK cells are derived from NK cells or from a unique precursor population. In the present study, we compare LAK cells derived from T-cell-depleted nylon-wool-non-adherent spleen cells with endogenous NK cells and NK cells activated with the interferon inducer polyl.C., in terms of their phenotype and functional characteristics. The predominant splenic LAK precursor in the mouse was found to be a nylon-wool-non-adherent, thy1-, MICG-, J11d.2-, asialo GM1+ cell. This phenotype is shared by endogenous NK cells. A significant number of activated NK cells express macromolecular insoluble cold globulin (MICG) in addition to asialo GM1. Neither endogenous nor activated NK cells express a heat-stable antigen found on bone-marrow cells, immature T cells, and most B cells and defined by monoclonal antibody (MAb)J11d.2. However, J11d.2 is expressed on some LAK precursor and effector cells. The asialo GM1 marker is common to all LAK effector cells, while many are also thy1+, and/or MICG+. LAK effector cells are therefore a heterogeneous population sharing some phenotypic characteristics in common with NK cells. In addition, there is a positive correlation between LAK and NK activity in "high NK" and "low NK" mouse strains, suggesting that NK cells and LAK cells share a common lineage. Monoclonal antibodies to the alpha and beta chains of LFA-I inhibit LAK and activated NK function, while endogenous NK and CTL killing is affected only by anti-alpha chain antibodies. LAK cells, like MHC-restricted and non-restricted CTL clones, express mRNA transcripts of the C11 serine protease gene. We conclude that LAK cells share several features in common with cells of the NK lineage and may therefore represent NK cells in a unique state of activation. LAK cells appear to employ cytolytic machinery common to other lytic cell types.

Mechanisms of lymphocyte-mediated lysis

Cytotoxic T lymphocytes (CTLs) and natural killer (NK) cells use multiple mechanisms to destroy their target cells. Pore formation resulting in osmotic lysis of the target is one mechanism; the pore-forming protein (perforin) responsible for this activity has been purified. Antigenically and functionally it resembles proteins of the membrane attack complex of complement. The other known mediators of cytotoxicity appear to be closely interrelated. Tumor necrosis factor (TNF), lymphotoxin (LT), and leukalexin are the three members of this group that have been purified, although their mechanisms of action are still unknown. CTLs fragment the DNA of target cells, as do TNF, LT, and leukalexin; this may be one of the mechanisms of action of these mediators. CTLs and NK cells do not self lyse. The basis of this phenomenon is unclear, although recent advances have shed some light on the problem.

Complexity of the primary genetic response to mitogenic activation of human T cells

We describe the isolation and characterization of more than 60 novel cDNA clones that constitute part of the immediate genetic response to resting human peripheral blood T cells after mitogen activation. This primary response was highly complex, both in the absolute number of inducible genes and in the diversity of regulation. Although most of the genes expressed in activated T cells were shared with the activation response of normal human fibroblasts, a significant number were more restricted in tissue specificity and thus likely encode or effect the differentiated functions of activated T cells. The activatable genes could be further differentiated on the basis of kinetics of induction, response to cycloheximide, and sensitivity to the immunosuppressive drug cyclosporin A. It is of note that cyclosporin A inhibited the expression of more than 10 inducible genes, which suggests that this drug has a broad genetic mechanism of action.

Molecular cloning of an inducible serine esterase gene from human cytotoxic lymphocytes

A cDNA clone encoding a human serine esterase gene was isolated from a library constructed from poly(A)+ RNA of allogeneically stimulated, interleukin 2-expanded peripheral blood mononuclear cells. The clone, designated HSE26.1, represents a full-length copy of a 0.9-kilobase mRNA present in human cytotoxic cells but absent from a wide variety of noncytotoxic cell lines. Clone HSE26.1 contains an 892-base-pair sequence, including a single 741-base-pair open reading frame encoding a putative 247-residue polypeptide. The first 20 amino acids of the polypeptide form a leader sequence. The mature protein is predicted to have an unglycosylated Mr of approximately equal to 26,000 and contains a single potential site for N-linked glycosylation. The nucleotide and predicted amino acid sequences of clone HSE26.1 are homologous with all murine and human serine esterases cloned thus far but are most similar to mouse granzyme B (70% nucleotide and 68% amino acid identity). HSE26.1 protein is expressed weakly in unstimulated peripheral blood mononuclear cells but is strongly induced within 6-hr incubation in medium containing phytohemagglutinin. The data suggest that the protein encoded by HSE26.1 plays a role in cell-mediated cytotoxicity.

A molecular-genetic analysis of cytotoxic T lymphocyte function

Two genes that are specifically expressed in T cells with cytolytic activity were isolated from a CTL cDNA library by differential screening. Both appear to encode serine proteases, thus suggesting a cascade mechanism, similar to complement, in activated CTL. Both CTL-specific proteases have a number of unusual structural features that suggest that they will have novel substrate specificities. One of the proteins (CCPI) has been oriented to the granules found in the cytoplasm of CTL. Taken together, these data strongly suggest that these molecules play an important role in target-cell lysis by CTL. Furthermore, we believe that the detailed molecular knowledge being accumulated through these studies may lead to the development of innovative forms of immunotherapy.

Isolation of two cDNA sequences which encode cytotoxic cell proteases

Two cDNAs which cross-hybridized with cytotoxic cell protease genes were identified in a library generated from a cytotoxic T cell line. Sequence analysis revealed that the two new members of the family contained the three catalytic triad residues which characterize the active sites of serine proteases. A comparison of the protein sequences revealed not only a high degree of homology but also the conservation of some unusual structural features. These include the lack of a disulphide bond which spans the active site serine, the presence of a signal sequence and the inference of a dipeptide activation sequence.