Increased expression of perforin and granzyme B genes in patients with metastatic melanoma treated with recombinant interleukin-2

Epigenetic Modification of PD-1/PD-L1-Mediated Cancer Immunotherapy against Melanoma

Malignant melanoma is one of the representative skin cancers with unfavorable clinical behavior. Immunotherapy is currently used for the treatment, and it dramatically improves clinical outcomes in patients with advanced malignant melanoma. On the other hand, not all these patients can obtain therapeutic efficacy. To overcome this limitation of current immunotherapy, epigenetic modification is a highlighted issue for clinicians. Epigenetic modification is involved in various physiological and pathological conditions in the skin. Recent studies identified that skin cancer, especially malignant melanoma, has advantages in tumor development, indicating that epigenetic manipulation for regulation of gene expression in the tumor can be expected to result in additional therapeutic efficacy during immunotherapy. In this review, we focus on the detailed molecular mechanism of epigenetic modification in immunotherapy, especially anti-PD-1/PD-L1 antibody treatment for malignant melanoma.

IL-2 immunotherapy in chronically SIV-infected Rhesus macaques

Background

Despite inducing a sustained increase in CD4+ T cell counts, intermittent recombinant IL-2 (rIL-2) therapy did not confer a better clinical outcome in HIV-infected patients enrolled in large phase III clinical trials ESPRIT and SILCAAT. Several hypotheses were evoked to explain these discrepancies. Here, we investigated the impact of low and high doses of IL-2 in Rhesus macaques of Chinese origin infected with SIVmac251 in the absence of antiretroviral therapy (ART).

Results

We demonstrated that rIL-2 induced a dose dependent expansion of CD4+ and CD8+ T cells without affecting viral load. rIL-2 increased CD4 and CD8 Treg cells as defined by the expression of CD25^highFoxP3⁺CD127^low. We also showed that rIL-2 modulated spontaneous and Fas-mediated CD4⁺ and CD8⁺ T cell apoptosis. The higher dose exhibited a dramatic pro-apoptotic effect on both CD4⁺ and CD8⁺ T cell populations. Finally, all the animals treated with rIL-2 developed a wasting syndrome in the month following treatment simultaneously to a dramatic decrease of circulating effector T cells.

Conclusion

These data contribute to the understanding of the homeostatic and dosage effects of IL-2 in the context of SIV/HIV infection.

Expression of Cellular FLICE Inhibitory Protein, Caspase-8, and Protease Inhibitor-9 in Ewing Sarcoma and Implications for Susceptibility to Cytotoxic Pathways

PURPOSE

Ewing sarcoma is a common pediatric bone tumor with an unfavorable prognosis for metastatic or recurrent disease. Cellular immunotherapy may provide new treatment options and depends on the cytolytic death receptor and perforin/granzyme pathways. Expression of death receptor pathway inhibitor cellular FLICE inhibitory protein (cFLIP), initiator caspase-8, and granzyme B inhibitor protease inhibitor-9 (PI-9) have been reported to determine susceptibility to cell- and chemotherapy-mediated killing in several tumor types. Here, we have studied their in vitro and in vivo expression in Ewing sarcoma and the implications for susceptibility to cytotoxicity.

EXPERIMENTAL DESIGN

Ewing sarcoma cell lines (n = 8) were tested for cFLIP, PI-9, and caspase-8 expression. Functional significance was tested by anti-Fas antibody (death receptor pathway) or natural killer cell (perforin/granzyme pathway) treatment. Immunohistochemistry was done on 28 sections from 18 patients. In half of the cases, sequential material, including metastases, was available.

RESULTS

Although all tested Ewing sarcoma cell lines expressed cFLIP, resistance to CD95/Fas-mediated apoptosis was only observed in two cell lines lacking caspase-8 expression. PI-9 was expressed at low levels in four of eight Ewing sarcoma cell lines, but positive cell lines remained susceptible to perforin/granzyme-mediated killing. In primary Ewing sarcoma, including metastases, cFLIP was abundantly expressed in 18 of 18 patients. Caspase-8 was expressed in all patients but showed more intertumoral and intratumoral variation in both intensity and heterogeneity of staining. PI-9, in contrast, was undetectable.

CONCLUSIONS

The expression patterns of cFLIP, caspase-8, and the absence of PI-9 provide a rationale to preferentially exploit the perforin/granzyme pathway in cytotoxic therapies against Ewing sarcoma.

Isolation and characterization of novel single-chain Fv specific for human granzyme B

Granzyme B, a neutral serine protease, has been demonstrated to be a pivotal molecule for protective immunity against viral infection and cellular malignant transformation. To facilitate monitoring of granzyme B levels, we have recently applied phage display technology to produce single-chain Fv antibodies specific for granzyme B, as versatile alternatives and complementary reagents to currently available monoclonal antibodies. Through four rounds of panning on purified human granzyme B-coated on solid phase, three unique clones were isolated. Expressed soluble scFv antibodies demonstrated specific immunological applications including ELISA, Western blotting, immunoprecipitation and intracellular staining. Based on sequence analyses and structural modeling, one scFv, Fv17, may have overlapping antigen binding specificity with monoclonal antibodies 2C5/F5 and GB11. Owing to the availability of its DNA sequence and large scale production capability, Fv17 should be a superior reagent for monitoring granzyme B expression in natural killer cells and antigen specific CD8+ T cell immunity.

Cationic lipid gene transfer of an IL-2 transgene leads to activation of natural killer cells in a SCID mouse human tumor xenograft

Natural killer (NK) cells play an important role in combating infectious and malignant diseases and interleukin-2 (IL-2) has been shown to promote proliferation and activation of NK cells in vitro and in vivo. Here we investigate the effects of local cationic lipid-mediated IL-2 gene transfer on intratumoral accumulation and activation of NK cells in a SCID mouse tumor model. UM449 human melanoma tumors in SCID mice received intratumoral injections of DMRIE/DOPE admixed with VR1103, a DNA plasmid encoding the gene for human IL-2. Dissagregated tumor cells were tested for IL-2 secretion and were characterized using antibodies to asGM1, MAC-1, and F4/80 antigens. Granzyme A, a proteolytic serine esterase, was also measured in tumor cell lysates. IL-2 secretion from tumors injected with VR1103:DMRIE/DOPE peaked at 48 h after injection and fell to baseline levels on day 8. Intratumoral granzyme A activity was significantly increased in tumors injected with IL-2 plasmid:DMRIE/DOPE complexes, but not by an irrelevant plasmid DNA:DMRIE/DOPE control. Importantly, the growth of UM449 tumors was slowed in VR1103:DMRIE/DOPE-injected tumors. These results indicate that local cationic lipid-mediated gene transfer of IL-2 induces activation of intratumoral NK cells and slows tumor growth.

Acute activation of CD8+ T lymphocytes in interleukin-2-treated HIV-infected patients. ANRS-048 IL-2 Study Group. Agence Nationale de Recherches sur le SIDA

CD8+ T lymphocytes play a key role in the control of HIV infection, through both cytotoxic and noncytotoxic mechanisms. To study in vivo effects of interleukin-2 (IL-2) treatment on this cell compartment, the level of activation of CD8+ T lymphocytes was evaluated before and just after 5-day administration of IL-2 in 16 HIV-infected patients. The serum level of soluble CD25 and of soluble CD8 significantly increased following IL-2 administration. The number of mRNA molecules coding for perforin and granzyme B, two enzymes that are contained in granules of cytotoxic cells, also significantly increased in peripheral blood mononuclear cells and in purified CD8+ cells (p < .001). Variations of plasma HIV viremia and perforin gene expression following IL-2 administration were inversely correlated (p = .023), suggesting that IL-2-induced activation of CD8+ T lymphocytes contributes to limit HIV replication in vivo. In contrast to perforin and granzyme B gene expression, IL-2 administration did not increase the expression of macrophage inhibitory protein-1alpha (MIP-1alpha), MIP-1beta, and regulated-on-activation normal T-expressed and secreted (RANTES) genes. These findings indicate that CD8+ T lymphocytes in HIV-infected patients are acutely activated by IL-2 treatment, which may improve long-term control of HIV infection.

Differential Cytotoxicity of Cord Blood and Bone Marrow–Derived Natural Killer Cells

Allogeneic cord blood is now being widely used as a source of stem cells for hematologic reconstitution after myeloablative therapy, with reported significantly lower levels of graft-versus-host disease (GVHD) compared with the use of allogeneic bone marrow (BM). This study was undertaken to investigate biologic aspects of natural killer (NK) cell activity, as recognized effector cells of the GVHD and graft-versus-leukemia (GVL) response, from cord blood and conventional BM. NK-cell activity levels of freshly isolated cells from cord blood and BM against K562 targets were comparable. Lymphokine activated killer (LAK) cells from both hematopoietic cell sources were compared for their ability to kill target cells by necrotic or apoptotic mechanisms using specific target cell lines. Cord blood cells had significantly higher necrosis-mediated cytotoxic activity against Daudi target cells compared with BM-derived cells. Cord blood LAK cells had relatively high levels of apoptotic-mediated cytotoxicity against YAC-1 target cells, whereas BM-derived LAK cells were unable to induce apoptosis in these cells. Interleukin-2 (IL-2) induced significant granzyme B activity in cord cells in contrast to BM cells, in which very little activity was measured. Western blotting confirmed these findings, with IL-2 inducing granzyme B protein expression in cord cells but not detectable levels in BM cells. BM cells had significantly lower cell surface expression of IL-2R and prolonged culture in IL-2 was only partially able to restore their deficient apoptotic cytotoxic activity. Thus, major differences exist between cord blood-derived and BM-derived mononuclear cells with respect to their NK-cell–associated cytotoxic behavior. This could have important implications for stem cell transplantation phenomena, because it suggests that cord blood may have increased potential for a GVL effect.

Differential Cytotoxicity of Cord Blood and Bone Marrow–Derived Natural Killer Cells

Allogeneic cord blood is now being widely used as a source of stem cells for hematologic reconstitution after myeloablative therapy, with reported significantly lower levels of graft-versus-host disease (GVHD) compared with the use of allogeneic bone marrow (BM). This study was undertaken to investigate biologic aspects of natural killer (NK) cell activity, as recognized effector cells of the GVHD and graft-versus-leukemia (GVL) response, from cord blood and conventional BM. NK-cell activity levels of freshly isolated cells from cord blood and BM against K562 targets were comparable. Lymphokine activated killer (LAK) cells from both hematopoietic cell sources were compared for their ability to kill target cells by necrotic or apoptotic mechanisms using specific target cell lines. Cord blood cells had significantly higher necrosis-mediated cytotoxic activity against Daudi target cells compared with BM-derived cells. Cord blood LAK cells had relatively high levels of apoptotic-mediated cytotoxicity against YAC-1 target cells, whereas BM-derived LAK cells were unable to induce apoptosis in these cells. Interleukin-2 (IL-2) induced significant granzyme B activity in cord cells in contrast to BM cells, in which very little activity was measured. Western blotting confirmed these findings, with IL-2 inducing granzyme B protein expression in cord cells but not detectable levels in BM cells. BM cells had significantly lower cell surface expression of IL-2R and prolonged culture in IL-2 was only partially able to restore their deficient apoptotic cytotoxic activity. Thus, major differences exist between cord blood-derived and BM-derived mononuclear cells with respect to their NK-cell–associated cytotoxic behavior. This could have important implications for stem cell transplantation phenomena, because it suggests that cord blood may have increased potential for a GVL effect.

Limited T-cell repertoire in renal allograft and allogeneic melanoma transmitted by the graft

In a patient with metastatic melanoma transmitted by the renal allograft, HLA serves as an alloantigen per se and is associated with tumor antigens at the same time. The influence of this antigeneic pattern on the Vbeta T-cell repertoire in an allogeneic melanoma, allograft, and peripheral blood mononuclear cells (PBMC) was assessed by polymerase chain reaction. Vbeta13.1 and 19 were found in both the melanoma and the graft. Vbeta14 was detected only in the melanoma and Vbeta6 was detected only in the kidney. PBMC revealed an unrestricted Vbeta pattern. Markers for cytotoxic activity of T cells--granzyme B and perforin--were not expressed during immunosuppressive therapy as clinically reflected in a nonrejecting allograft and in a progressing melanoma. In vitro PBMC proliferated to recombinant interleukin-2, whereas recombinant interferon-gamma did not augment this response. Initiation of immune therapy, in addition to discontinuation of immunosuppression, might support the rejection of the allogeneic tumor by dominant Vbeta T cells.

Cytokine-induced cytotoxic function expressed by lymphocytes of the innate immune system: distinguishing characteristics of NK and LAK based on functional and molecular markers

Several molecular events are now identifiable during the activation, recognition, and killing by natural killer (NK) cells that are distinct from those differentiated in response to cytokines during the generation of lymphokine-activated killer (LAK) cells or during lymphocyte proliferation. Because LAK and MHC-unrestricted killing activities also include the prototypic NK targets as part of their broad recognition spectra, accurate identification of the complete function being studied is critical. In many publications, past and present, only NK-sensitive target cells were used (K562, Molt-4, others), and, therefore, the results do not necessarily indicate whether the effectors are NK or have differentiated into LAK cells. Such a consideration becomes critical when the effectors are grown in interleukin-2 (IL-2), and an attempt is made to define receptor recognition, signal transduction pathways, and specificity at the molecular level. In some instances, effector cells are likely to have stopped, therefore merely expressing NK activity, and have also acquired LAK function. The identified receptors may not have been unique to NK cells or NK function. Not until the targets employed are also confirmed to be NK sensitive, and the effectors do not kill NK-resistant targets can the results of molecular studies be proposed to represent aspects unique to NK. Reports of the use of IL-2-expanded NK clones are most likely providing data concerning the biology of LAK and not of classic NK. The classic NK activity surveying our blood apparently performs an important function, including the ability to respond rapidly to certain cytokines and to acquire additional functions and receptors for use in destroying a vast array of target cells. It is critical for scientists to appreciate the functional distinctions and to identify the molecules and pathways unique to each of these curious cytolytic systems.

Role of naive and memory T cells in tumor cell lysis mediated by bi-specific antibodies

Bispecific monoclonal antibodies (Bi-mAb) with specificity for a tumor associated antigen and the CD3 or CD28 antigen on T lymphocytes, respectively, induce activation of resting T lymphocytes and target-specific tumor cell lysis. Former studies had confirmed that T cells expressing the CD45RO "memory" antigen at high levels were the most potent effectors of Bi-mAb-mediated cytotoxicity when compared to their "naive" counterparts expressing the CD45RA antigen. Further analysis of the T cell subpopulations revealed that within the memory T cell pool, CD8+ T cells were the effector cell, population with strongest cytolytic activity. The cytolytic activity was correlated with the expression level of perforin and granzymes B mRNA. Ca2+ complexing agents, which abrogate perforin activity, reduced necrosis, while inhibition of granzyme activity in effector or target-cells had a similar effect on apoptosis. These results confirm the crucial role perforin and granzymes play in target-cell lysis and explain why CD8+CD45RO+ T cells activated by combined CD3 and CD28 antigen triggering represent the T cell pool with highest cytolytic potential.

Intratumoral activation of CD8-positive cytotoxic lymphocytes in acquired immunodeficiency syndrome lymphomas

The incidence of lymphomas is unusually high in human immunodeficiency virus (HIV)-infected patients. Because cytotoxic T lymphocytes (CTL) represent a major mechanism of the antitumoral immune response in immunocompetent individuals, we asked whether intratumoral activation of CTL was impaired in acquired immune deficiency syndrome (AIDS) lymphomas. Immunohistochemical experiments showed that in AIDS lymphomas intratumoral CD8-positive T lymphocytes accumulated and expressed the TIA-1 antigen, a marker of cytotoxic cells. Flow cytometry studies and in situ hybridization of lymphomatous tissue confirmed the differentiation of CD8-positive cells in cytotoxic cells and their activation, as assessed by their expression of CD38 and human leukocyte antigen (HLA) DR markers as well as the perforin and granzyme B genes, which code for two molecules involved in target cell killing. On average, perforin-producing cells were as numerous in AIDS lymphomas (5,647 +/- 2,655 cells/cm2) as in lymphomas from immunocompetent individuals (3,294 +/- 1,544 cells/cm2). The density of activated CD8-positive cells in the 22 AIDS lymphomas tested was not correlated with peripheral CD4-positive cell counts. These results suggest that in AIDS lymphomas the steps of differentiation and activation of cytotoxic CD8-positive cells are not altered by immune deficiency and that they can take place through pathways relatively independent of CD4-positive T lymphocytes. Thus, other mechanisms of immune deficiency should account for the increased frequency of lymphomas in patients with AIDS.