DNA mismatch repair protein expression and microsatellite instability in primary mucosal melanomas of the head and neck.

AIMS: To examine the expression of DNA mismatch repair (MMR) proteins and the presence of microsatellite instability (MSI) in seven primary mucosal melanomas of the head and neck (MMHN). METHODS AND RESULTS: Haematoxylin and eosin staining and immunohistochemical analysis for routine diagnostic markers and for MMR proteins were performed. Six cases were examined for MSI. Four cases were monomorphous and three cases were pleomorphic type MMHN. Melanocytic markers were positive in all cases. Immunoreactivity for MMR proteins was weak in normal epithelium. The neoplastic tissue in six cases showed positivity for all MMR proteins with different percentages. One case showed weak positivity for hMSH2 and hMSH6 and no immunoreactivity for hMLH1 or hPMS2. Staining intensity was higher in tumour cells than in matched normal mucosa in three cases for hMSH2 and hMLH1 and in two cases for hPMS2. None of the examined cases showed MSI. CONCLUSIONS: Expression of hMSH2 and hMLH1 proteins was up-regulated in three cases, whereas in two cases that of hPMS2 was increased. hMSH6 expression was comparable to that of normal cells in all cases. The percentage of positive neoplastic cells and the intensity of staining seemed to be greater in pleomorphic melanomas. Six cases were MMR-proficient and microsatellite stable.

DNA repair enzymes and cytotoxic effects of temozolomide: comparative studies between tumor cells and normal cells of the immune system.

O6-alkylguanine-DNA alkyltransferase (OGAT) and the mismatch repair system (MRS) play a crucial role in the susceptibility of tumor cells to the cytotoxic effects of agents that generate O6-methylguanine in DNA, including the triazene compound temozolomide (TMZ). Studies performed with peripheral blood mononuclear cells (MNC) showed that TMZ was scarcely active on lymphocyte functions not dependent on cell proliferation (e.g. NK activity and cytokine-mediated induction of CD1b molecule in adherent MNC). In contrast, TMZ depressed proliferation and lymphokine activated killer (LAK) cell generation in response to IL-2. In this case, a reasonably good inverse relationship was found between OGAT levels of MNC and their susceptibility to TMZ. This study also analyzed the ratio of the toxic effect of TMZ on MNC and on tumor cells (i.e. "Tumor-Immune Function Toxicity Index", TIFTI). A particularly favorable TIFTI can be obtained when OGAT levels are extremely high in MNC and markedly low in tumor cells. This holds true for MRS-proficient neoplastic cells, but not for MRS-deficient tumors. In conclusion, strategies aimed at modulating OGAT and MRS may improve the clinical response to TMZ. However, the use of OGAT inhibitors to potentiate the antitumor activity of TMZ might result in a concomitant increase of the immunosuppressive effects of the drug, thus reducing the relative TIFTI.

Microsatellite instability in primary brain tumors.

Microsatellite instability has been reported in hereditary colorectal cancer syndrome and in various kinds of human sporadic tumors. It has also been shown in brain tumors, although with conflicting results. In the present study, DNA samples obtained from 20 primary brain tumors (10 glioblastomas, three astrocytomas, five meningiomas, one ependymoma, one hemangiopericytoma) were analyzed to detect microsatellite instability. Nine microsatellites, mono, di-, tri- and tetranucleotide repeat markers, located on nine different chromosomes, were used. Four of the 20 neoplasias (20%) showed microsatellite alterations in tumor DNA with respect to normal DNA. Two glioblastomas and one atypical meningioma (15%) showed additional bands or bands with shift of electrophoretic mobility, whereas allelic loss was observed in two glioblastomas (10%). In one glioblastoma, one allelic loss and one extra allele were observed at two different loci. These data indicate that in primary brain tumors there is not a high genetic instability. Although we used markers with inherently high levels of instability, only sporadic microsatellite alterations were found. Consequently, alterations in the mechanisms of DNA mismatch-repair, the most important cause of replication errors in hereditary and sporadic colorectal cancers, do not seem to play a major role in the oncogenesis of primary brain tumors.

In vitro effect of hyperthermia on natural cell-mediated cytotoxicity.

It is well known that hyperthermia (HY), which is used for the treatment of cancer, depresses natural cell-mediated immunity in vitro. Experiments were performed to confirm the inhibitory effect of HY (42 degrees C for 1 hour) on natural killer (NK) activity and to evaluate the influence of HY on the generation and cytotoxic activity of interleukin-2 (IL-2)-activated NK cells. Additional experiments were also carried out to evaluate the effect of a simultaneous exposure of effector and target cells to HY. The results showed that HY profoundly reduced the lytic activity of NK cells and demonstrated that this inhibition was transient and not due to an apoptosis-induced reduction of the number of effector cells. Moreover, the exposure of mononuclear cells to HY before IL-2 stimulation did not affect the generation of IL-2-activated NK cells, whereas, the hyperthermic treatment of IL-2-activated NK cells produced a marked reduction of their cytotoxic activity. The results also showed that the simultaneous exposure of effector and target cells to HY, during the cytotoxicity assay, produced a marked reduction of lytic activity of NK and IL-2-activated NK cells, and that this impairment was specific for effector cells. In this context, heat-exposure of target cells alone, did not substantially modify their susceptibility to lysis induced by either NK or IL-2-activated NK cells. These results add further evidence of HY-induced inhibition of natural cell-mediated immunity, and suggest that, in the course of therapeutic HY, immune response could be significantly altered.

O(6)-benzylguanine enhances the in vitro immunotoxic activity of temozolomide on natural or antigen-dependent immunity.

Temozolomide (TMZ) is a new cytotoxic triazene compound of clinical interest that is able to generate methyl adducts at the O(6)-guanine of DNA, which can be repaired by O(6)-alkylguanine-DNA alkyltransferase (OGAT). It was previously found that triazene compounds are highly immunosuppressive in mice. In the present study, we investigate whether TMZ could affect immune functions of human competent cells and whether methylation of O(6)-guanine could be involved in the immunosuppressive activity of the drug. Mononuclear cells (MNCs) obtained from peripheral blood of healthy donors were tested for OGAT activity and treated with TMZ alone or combined with the OGAT inhibitor O(6)-benzylguanine. Control or drug-treated MNCs were then assayed for natural killer activity and for the ability to proliferate and to generate cytotoxic effector cells in response to interleukin-2 or allogeneic MT-2 tumor cells. The results show that TMZ inhibited both proliferation and induction of lytic activity in response to interleukin-2 or allogeneic MT-2 cells. Moreover, an inverse correlation was found between the OGAT activity of MNCs and their sensitivity to TMZ. The involvement of O(6)-guanine methylation in the immunosuppressive effects of TMZ was further confirmed by the finding that O(6)-benzylguanine increased the activity of the drug. On the other hand, the natural killer activity of MNCs was only moderately affected by TMZ, and no relationship was observed between OGAT levels and sensitivity to the drug. These data suggest that in patients with tumors who are undergoing TMZ treatment, the drug may impair immune responses involving cell proliferation, depending on OGAT levels of MNCs, and that O(6)-benzylguanine may potentiate this activity.

2'-2'-difluorodeoxycytidine: in vitro effects on cell-mediated immune response.

Little or no data are available on the immunotoxicity of the new deoxycytidine analogue, gemcitabine (2'-2'-Difluorodeoxycytidine, dFdC). The drug was tested on natural killer (NK), interleukin 2-activated killer (LAK) and antigen-dependent cytotoxic effector cells (CTL) activity. NK cells were treated for 16 hours and then tested against K562 cell line. LAK cells were pretreated for 16 hours before or after IL2 stimulation, and tested against DAUDI cells on day 4. CTL were pretreated for 16 hours on day-1 or on day 4 of coculture, and then tested against MT-2 on day 5. Cytotoxic activity was evaluated by a 4 hours 51Cr-release assay. The results indicate that dFdC inhibits markedly LAK or CTL generation, but does so less efficiently on "mature" LAK or CTL lymphocyte function and only slightly on NK cell activity. Therefore dFdC can be considered immunotoxic for either natural or antigen-dependent cell-mediated immunity.

In vitro antitumor activity of 3'-desamino-3'(2-methoxy-4-morpholinyl) doxorubicin on human melanoma cells sensitive or resistant to triazene compounds.

A new methoxymorpholinyl derivative of Adriamycin (ADR), FCE 23762 (MRD), has recently been selected for phase I clinical trials for its reduced cardiotoxicity and for its cytotoxic activity against a broad spectrum of solid tumors and leukemias that are sensitive or resistant to ADR. The purpose of the present study was to compare the in vitro antitumor activity of MRD and ADR on human melanoma lines with different chemosensitivity to triazene compounds, among which dacarbazine remains a reference drug in the treatment of melanoma. Both MRD and ADR were tested in vitro on three melanoma lines, MI13443-MEL, SK-MEL-28, and M14, previously screened for their chemosensitivity to the triazene compound p-(3-methyl-1-triazeno) benzoic acid, potassium salt (MTBA). The three lines were also analyzed for P-170 expression, total glutathione (GSH) content, and GSH-related enzyme activity. All melanomas, whether sensitive or resistant to MTBA, were susceptible to anthracycline treatment. The cytotoxic activity of MRD was comparable with that of ADR, and no substantial difference was found in cell growth inhibition between the two drugs. When the relative chemosensitivity of the three lines was considered, SK-MEL-28 was found to be slightly less sensitive to MRD treatment than the other tumors. This finding seems to correlate with the higher GSH-peroxidase activity of this melanoma relative to that of the MI13443 and M14 lines. These results show a homogeneous response of melanoma lines to MRD treatment in vitro, suggesting that phase I clinical trials concerning this drug, which in vivo appears to be activated to a more cytotoxic metabolite, could be extended to metastatic melanomas, including those completely resistant to triazene compounds.

In vitro tumour cell growth inhibition: a comparative study between allosensitized cytotoxic T lymphocytes and lymphokine activated killer cells.

There is general agreement that several distinct subpopulation of lymphocytes, including major histocompatibility complex (MHC)-restricted T lymphocytes and non-restricted natural killer, or lymphokine-activated killer (LAK), cells are active in lysing neoplastic cells. In this study experiments were designed to compare the inhibitory effects of LAK cells and allosensitized cytotoxic T lymphocytes (CTL) on in vitro growth of an Epstein-Barr virus-transformed B-cell line (BSM) and of a HTLV-I producer T-cell line (MT-2). It was found that allosensitized CTL are more efficient at inducing BSM, or MT-2, cell growth inhibition than LAK cells. These results are consistent with the hypothesis that MHC-restricted T effector cells could mediate higher tumour suppressive effects than non-MHC restricted LAK cells.

Differential interference of alpha, beta or gamma interferons with HTLV-I integration and expression in vitro.

Alpha, beta and gamma interferons (IFNs) can exert a powerful and direct antiviral activity against HTLV-I and can also modulate positively some cell-mediated immune functions of the host cell. These multiple effects of IFNs can induce a long-lasting inhibition of viral infection in recipient cells, probably by "priming" the host cell to an active antiviral competence. It has to be underlined that each IFN was active differentially on the replicative cycle of HTLV-I, thus suggesting the possibility of a complementary action of IFNs in inhibiting HTLV-I infection. This might be relevant to a possible therapeutical approach for prevention of HTLV-I related diseases.

Selection of HTLV-I positive clones is prevented by prostaglandin A in infected cord blood cultures.

Type A prostaglandins (PGA1 and 16,16-dimethyl-PGA2-methyl ester) were found to block the proliferation of HTLV-I infected cord blood lymphocytes (CBL) in vitro, thus preventing the clonal immortalisation that is considered as a predisposing condition to HTLV-I positive leukaemia. PGA1 and di-M-PGA2 did not affect the long-term survival of normal non-infected CBL, whereas they suppressed the proliferation of an established cord-blood derived HTLV-I positive cell line, MT-2. As shown by the number of HTLV-I infected p19+ cells, the block of the selection of immortalised, infected clones by PGAs did not appear to be due to an inhibition of early stages of HTLV-I infection. The possibility that the effect of PGAs could be mediated by an action on the immune response was also examined. PGAs regulated the cell-mediated cytotoxic function of CBL to a different extent when normal non-infected or HTLV-I exposed CBL were compared. In fact, PGAs down-regulated the natural killing and macrophage/lymphocyte cytotoxic response of normal CBL, whereas they did not modify the already depressed immune response of CBL challenged with HTLV-I. These results suggest that the protective effect of PGAs against HTLV-I infection in vitro is mostly related to the direct suppression of the clonal expansion of virus-infected cells, rather than to the anti-viral activity or modulation of the cell-mediated immunity.

Increase of natural killer activity of mouse lymphocytes following in vitro and in vivo treatment with lithium.

The in vivo and in vitro influence of lithium lactate on mouse natural killer activity was investigated. In vitro exposure of effector-target mixture to graded concentrations of lithium did not substantially modify the natural killer activity of mouse splenocytes, untreated or pretreated with cyclophosphamide. However in vitro treatment of effector splenocytes increased the frequency of NK-percursor cells. The in vivo treatment with lithium lactate greatly increased the natural killer activity in intact mice, whereas it did not improve this cytotoxic function in host immunodepressed by cyclophosphamide. These data suggest that lithium salts produce a modulation of natural killer activity of mouse spleen cells, probably through a mechanism involving the increase of the number of NK-precursors in hosts not subjected to cytotoxic chemotherapy.

Low concentrations of suramin can reduce in vitro infection of human cord blood lymphocytes with HTLV-I during long-term culture.

In vitro infection of human cord blood lymphocytes (CBL) with human T-cell leukemia/lymphoma virus type I (HTLV-I) was found to be reduced by suramin treatment at a concentration ranging from 10-100 micrograms/ml. At higher concentrations (500 micrograms/ml) suramin was toxic to the cells and even resulted in an increased percentage of cells positive for the p19 viral core protein. Suramin treatment at the onset of the CBL coculture with a lethally irradiated HTLV-I donor cell line (MT-2) reduced virus transmission, evaluated as number of p19+ cells, and the consequent amount of integrated provirus in the host genome. The amount of viral RNA transcripts was not reduced in CBL cocultures. On the other hand, suramin affected HTLV-I replication in infected MT-2 cells, when used at a concentration of 50 micrograms/ml, and this might contribute to the reduced infectivity of suramin-treated MT-2 cells. In addition to its antiviral effects, suramin exerted a modest positive regulation on the natural killing activity of CBL and their early proliferative response in mixed lymphocyte/tumor cell culture.

Natural cell-mediated cytotoxicity: a micro assay suitable for clinical tests.

A micro version (microtest, MIT) of the 51Cr release assay for detecting Natural Killer activity (NK) has been developed. The test retains the sensitivity and the efficiency of conventional macroassay (macrotest, MAT) and provides a 5-fold reduction in the number of effector and target cells employed. In experiments performed with peripheral blood mononuclear cells (MNC), untreated or treated with interferon (IFN) or with hydrocortisone (Hy), comparable values of the percentage of specific lysis and of the number of lytic units were obtained using both MAT and MIT methods. Therefore MIT appears to be useful in monitoring the NK function of patients characterized by low MNC counts.