Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Temozolomide (TMZ) is a methylating imidazotetrazinone presently under phase II/III clinical investigation for the treatment of various neoplasias, including non-Hodgkin's lymphoma, melanoma, and high-grade glioma (Bleehen et al., 1995; Woll et al., 1995; Bower et al., 1997). In physiological solutions, TMZ readily decomposes into 5-(3-methyl-1-triazeno)imidazole-4-carboxamide (MTIC), the methylating derivative that is also generated through host metabolism of dacarbazine [DTIC; 5-(3,3-dimethyl-1-triazeno)imidazole-4-carboxamide; Stevens et al., 1987]. MTIC further decomposes into a methyl-diazonium ion that reacts with nitrogen bases in DNA, mainly at the N⁷ and O⁶ positions of guanine (O⁶-G), N³ position of adenine, and O⁴ position of thymine. Adduct formation at O⁶-G is considered to be one of the major determinants of TMZ cytotoxicity because, in most cases, tumor cell sensitivity to the drug is inversely correlated with the activity of the DNA repair protein O⁶-alkylguanine-DNA alkyltransferase (OGAT; Tisdale, 1987; D'Atri et al., 1995; Tentori et al., 1995, 1997; Liu et al., 1996).

Human OGAT is a 22-kDa protein that appears to be present in varying amounts in normal tissue (Myrnes et al., 1983; Gerson et al., 1986; Citron et al., 1991). The content of OGAT also varies among individuals (Myrnes et al., 1983; Pegg, 1990; Citron et al., 1991) and is highly heterogeneous among tumor lines (D'Incalci et al., 1988; Fornace et al., 1990; Pegg, 1990). The protein removes methyl and other short alkyl groups from O⁶-G and transfers them to an internal cysteine residue in a stoichiometric and autoinactivating reaction (reviewed in Pegg, 1990). The extent of O⁶-alkylguanine repair is therefore dependent on the initial cellular OGAT levels and on the rate at which de novo synthesis of the protein occurs.

Depletion of OGAT, in tumors proficient for this protein, is considered to be a useful strategy to increase clinical responses to O⁶-G-methylating agents (Dolan and Pegg, 1997). Indeed, depletion of OGAT activity by O⁶-benzylguanine (BG), which acts as a potent substrate for the protein (Dolan et al., 1990), has been shown to increase tumor cell sensitivity to TMZ both in vitro (Baer et al., 1993; Tentori et al., 1995, 1997; Liu et al., 1996; Wedge et al., 1996a) and in vivo (Wedge and Newlands, 1996b; Wedge, 1997).

The mechanism by which the persistence of O⁶-methyguanine (O⁶-MeG) in DNA leads to cell cytotoxicity involves recognition of O⁶-MeG:T and O⁶-MeG:C mispairs, occurring during cell duplication, by the mismatch repair system (reviewed in Jiricny, 1996; Modrich, 1997). Thus, independent of their OGAT activity, tumor cell lines harboring mutations in the constituent proteins of the mismatch repair system are resistant to the cytotoxic effects of TMZ (Liu et al., 1996; D'Atri et al., 1998) and other O⁶-G methylating agents (Karran and Bignami, 1992; Kat et al., 1993; Carethers et al., 1996).

A number of studies are available suggesting that stimulation of the host's immune system with immunomodulating agents or antitumor vaccines might be of benefit in the treatment of patients with cancer, particularly in the case of minimal residual disease (Ockert et al., 1999). Immunochemotherapy protocols based on DTIC plus interferon and/or interleukin (IL)-2 administration are already used in the treatment of metastatic melanoma (Cohen and Falkson, 1998). In these protocols, it is assumed that stimulation of the host's immune function results in better clinical responses.

Previous studies performed in the murine model have shown that triazene compounds not only possess antitumor activity but also are immunosuppressive. Indeed, humoral and cell-mediated immune responses are profoundly inhibited in mice exposed to DTIC (Giampietri and Bonmassar, 1978; Nardelli et al., 1984). Moreover, murine splenocytes treated in vitro with the drug show a marked impairment of antigen-dependent cell-mediated immunity and of proliferative responses to mitogens (Giampietri and Bonmassar, 1978; Nardelli et al., 1984). On the other hand, the effects of triazene compounds on immune function of human competent cells have not been extensively investigated (Bonmassar et al., 1994), and no data are presently available on the immunosuppressive activity of TMZ.

As already occurring for DITC, it is possible that TMZ could be used in association with immunomodulating agents in the treatment of melanoma. Therefore, studies have been performed to investigate whether the molecular mechanisms underlying the antitumor activity of TMZ may also be responsible for immunosuppressive effects. This was done to obtain useful information for a rational approach to cancer treatment based on TMZ associated with biological response modifiers. The present report illustrates the in vitro effects of TMZ, alone or in combination with BG, on natural or antigen-dependent cell-mediated immunity of human peripheral blood mononuclear cells (MNCs).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Lines and Culture Conditions. Erythroleukemia, K562, and Burkitt lymphoma, Daudi, cell lines were obtained from American Type Culture Collection (Rockville, MD). The human MT-2 cell line human T-cell leukemia virus type I-transformed cord blood T lymphocytes; Miyoshi et al., 1981) was kindly provided by Dr. B. Macchi (University of Tor Vergata, Rome, Italy). This cell line expresses class I and class II HLA antigens and B7 molecule and was used to elicit allogeneic antigen-dependent immune responses in human MNCs.

Chemical and Biological Agents. TMZ was kindly provided by Schering-Plow Research Institute (Kenilworth, NJ). Because the drug readily decomposes in aqueous solution into MTIC (Stevens et al., 1987), solutions were always prepared fresh by dissolving the drug in RPMI 1640. The solutions were protected from light and used immediately after preparation. The OGAT inhibitor BG was synthesized and generously donated by Prof. L. Lassiani (Institute of Pharmacological Chemistry, University of Trieste, Italy). Mitomycin C (Kiowa Hakko Kogyo Co. Ltd., Tokyo, Japan) was diluted in water and stored at 80°C. Human recombinant IL-2 (kindly provided by Hoffman-La Roche, Basel, Switzerland) was diluted in RPMI 1640 medium and stored at 20°C.

Preparation of MNCs and In Vitro Treatment with TMZ and BG. MNCs were separated from the peripheral blood of healthy donors on a Ficoll-Hypaque gradient and washed twice in PBS. MNCs were then incubated (at 37°C for 1 h) in 50-ml tubes (Falcon; Becton and Dickinson Labware, Franklin Lakes, NJ; 2 × 10⁶ cells/ml, 10 ml/tube) in RPMI 1640 supplemented with 2% FCS, 2 mM L-glutamine, and antibiotics [hereafter referred to as complete medium (CM)] or in CM containing BG at the final concentration of 2 µM. Suspensions of control or BG-treated MNCs were then admixed with equal volumes of CM alone or CM containing TMZ to give final drug concentrations ranging between 62.5 and 500 µM and incubated at 37°C for 4 h. Because MNCs were not washed before the addition of TMZ, BG was always present during the 4-h treatment with the triazene compound. Thereafter, MNCs were washed in PBS and tested immediately for cytotoxic activity against K562 target cells [i.e., natural killer (NK) activity]. MNCs were also stimulated with IL-2 or allogeneic MT-2 tumor cells either in CM or in CM containing 2 µM BG. MNC responses to IL-2 or MT-2 cells were evaluated in terms of proliferation and generation of lymphokine-activated killer (LAK) cells or antigen-dependent cytotoxic T lymphocytes (CTL; see below).

Evaluation of MNC Proliferative Responses. Control or drug-treated MNCs (1 × 10⁶cells/ml in CM) were plated onto 96-well microtiter plates (0.2 ml/well) and stimulated with IL-2 (500 IU/ml) or with mitomycin-inactivated (80 µg/ml for 1 h) allogeneic MT-2 tumor cells (5 × 10³ cells/well) for 4 or 5 days, respectively. Stimulation of MNCs that had been preexposed to BG or TMZ plus BG was performed either in CM alone or in CM containing BG 2 µM. [methyl-³H]Thymine deoxyriboside (³H-TdR; Amersham International Plc, Amersham, UK) was added to the cultures (5 µCi/ml) 18 h before the end of incubation, and radioactivity incorporation was determined by conventional liquid scintillation counting with a TRI-CARB 1900 counter (Packard Instrument Co., Meriden, CT).

In Vitro Generation of LAK Cells and CTL. To generate LAK cells, control or drug-treated MNCs (1 × 10⁶/ml in CM) were dispensed (3 ml/well) in 6-well plates and incubated with 500 IU/ml IL-2 at 37°C for 4 days. To generate allosensitized CTL, MNCs were plated as above and cocultured with mitomycin-inactivated MT-2 tumor cells (7.5 × 10⁴/well) at 37°C for 5 days. Stimulation of MNCs pretreated with BG or TMZ plus BG was performed either in the absence or the presence of BG 2 µM. At the end of incubation period, effector cells were recovered from the cultures and tested for cytotoxic activity against Daudi (LAK cells) or against MT-2 (CTL) target cells with the use of a conventional ⁵¹Cr-release assay (see below).

Cytotoxicity Assay. Aliquots of K562, Daudi, or MT-2 cells were centrifuged, resuspended in 0.1 ml of FCS, and incubated with 100 µCi of Na₂⁵¹CrO₄ (Amersham) at 37°C for 1 h. After incubation, the cells were extensively washed in RPMI 1640 medium, suspended in RPMI 1640 containing 10% FCS and 2 mM L-glutamine, and used in the cytotoxicity assay.

Evaluation of Cytotoxic Activity of Effector Cells. The percentage of specific target cell lysis was calculated as follows:
where test cpm is the mean cpm relative to ⁵¹Cr released in presence of effector cells, autologous cpm is the mean cpm relative to ⁵¹Cr released by target cells incubated with unlabeled autologous cells in place of effector cells, and total cpm is the mean cpm of 2 × 10^{3 51}Cr-labeled target cells (i.e., total radioactivity incorporated by target cells).

Evaluation of MNC Sensitivity to TMZ. MNC sensitivity to TMZ was expressed in terms of the 50% inhibitory drug concentration (IC₅₀, or the concentration of drug capable of inhibiting MNC proliferative responses or cytotoxic activity by 50%) calculated on the regression line in which cpm (for proliferative responses) or KC(10⁶) or KC(culture; for cytotoxic activity) values were plotted against the logarithm of drug concentration.

OGAT Assay. MNCs, either untreated or exposed to BG, were washed twice with PBS and stored as pellets at 80°C until used. OGAT activity was determined by measuring the transfer of ³H-methyl groups from a DNA substrate to the OGAT protein (Morten and Margison, 1988). Briefly, cell pellets were thawed, resuspended in 1 ml of lysis buffer (0.5% 3-[(3-cholamidopropyl)dimethylammonio]propanesulfonate, 50 mM Tris · HCl pH 8.0, 1 mM EDTA, 3 mM dithiothreitol, 100 mM NaCl, 10% glycerol, 2 mg/ml leupeptine), and incubated for 30 min at 4°C. Cell lysates were then centrifuged at 15,000 rpm for 10 min. Aliquots of supernatants were then diluted in 50 mM Tris · HCl buffer, pH 8.3, containing 1 mM EDTA and 3 mM dithiothreitol and incubated with 10 µg of ³H-methylated DNA at 37°C for 1 h. DNA was then hydrolyzed by heating samples at 75°C for 45 min in the presence of 1 N perchloric acid and protein precipitated with 1 mg of BSA as carrier. Pellets were washed with 1 N perchloric acid and resuspended in 0.01 N NaOH, and radioactivity counted in a liquid scintillation counter (TRI-CARB 1900) after the addition of scintillation liquid (Ultima Gold; Packard Instruments, Groningen, the Netherlands). Protein concentration in supernatants was evaluated according to the method of Bradford (1976) using the Bio-Rad (Hercules, CA) Dye Solution and BSA as standard. OGAT activity was expressed in terms of fmol ³H-methyl groups transferred/mg of protein in cell extract.

Statistical Analysis. Differences in proliferative responses of control or drug-treated MNCs were evaluated with the use of the Student's t test. Differences in cytotoxic activity of control or drug-treated MNCs were evaluated taking into account the percentage of specific cytotoxicity at all E/T ratios; therefore, p values were calculated using ANCOVA performed on the regression of the percentage of specific ⁵¹Cr release over the logarithm of the number of effector cells/well. All the data relative to cell-mediated cytolysis were expressed in terms of KC(10⁶) or KC(culture) values, without conventional S.E. or S.D. values. Actually, no statistical analysis can be performed using these parameters, which are not suitable for ANCOVA of regression lines.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

OGAT Activity of MNCs Obtained from Peripheral Blood of Healthy Donors. The level of OGAT activity was evaluated in MNCs obtained from 22 different healthy donors. The results (Fig. 1) show that a wide range of enzyme activity (370-1500 fmol/mg protein) was detectable in MNCs. However, the majority of the samples showed OGAT levels ranging from 800 to 1100 fmol/mg protein; only 3 of 22 samples showed enzyme activity lower than 500 fmol/mg protein. These MNCs were considered to be endowed with low OGAT levels.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   OGAT activity of MNCs obtained from peripheral blood of healthy donors. OGAT activity is expressed in terms of fmol of 3H-methyl groups transferred/mg of protein in cell extract. Each value represents the arithmetic mean value of at least three independent experiments. Bars, S.E.

Relationship between OGAT Activity of MNCs and TMZ-Mediated Impairment of Proliferative Responses Induced by IL-2 or by Allosensitization with MT-2 Cells. The effects of TMZ on MNC proliferative responses induced by IL-2 or inactivated allogeneic MT-2 tumor cells were evaluated in samples endowed with different OGAT levels. The relationship between OGAT activity and MNC sensitivity to TMZ is illustrated in Fig. 2. The results show that TMZ inhibited MNC proliferation elicited by IL-2 (Fig. 2A) or by allostimulation (Fig. 2B) to comparable extents, with IC₅₀ values ranging between 134 and 515 µM. In both cases, MNC sensitivity to the drug was inversely correlated with OGAT activity.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Relationship between OGAT activity of MNCs and TMZ-mediated impairment of proliferative responses induced by IL-2 or by allogeneic MT-2 cells. MNCs were exposed to graded concentrations of TMZ (62.5-500 µM) for 4 h, washed, and incubated with IL-2 (500 IU/ml) for 4 days (A) or cocultured with mitomycin-inactivated allogeneic MT-2 cells for 5 days (B). Proliferative responses of control and drug-treated MNCs were evaluated in terms of 3H-TdR incorporation. MNCs sensitivity to TMZ is expressed in terms of IC50 (i.e., the concentration of drug capable of inhibiting MNCs proliferative responses by 50%, calculated on the regression line in which cpm values were plotted against the logarithm of drug concentration). OGAT activity is expressed in terms of fmol of 3H-methyl groups transferred/mg of protein in cell extract. Correlation coefficients were 0.92 and 0.87 for A and B, respectively.

Relationship between OGAT Activity of MNCs and TMZ-Induced Impairment of NK Function and of LAK Cell and CTL Generation. The effect of TMZ on NK function was evaluated in three MNC samples endowed with markedly different enzyme activity. The results illustrated in Table 1 shows that TMZ inhibited moderately the NK function, with the IC₅₀ of the agent being about 400 µM. However, these inhibitory effects were not clearly dependent on the level of OGAT activity, being similar for all the three MNC preparations used.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Relationship between OGAT activity of MNCs and TMZ-induced impairment of LAK cell and CTL generation. MNCs were exposed to graded concentration of TMZ (62.5-500 µM) for 4 h, washed, and incubated with IL-2 (500 IU/ml) for 4 days (A) or cocultured with mitomycin-inactivated allogeneic MT-2 cells for 5 days (B). At the end of incubation period, effector LAK cells or CTL were recovered and tested for cytotoxic activity against Daudi cells or MT-2 cells, respectively, using a 51Cr-release assay (see Materials and Methods). MNC sensitivity to TMZ is expressed in terms of IC50 calculated on the regression lines obtained by plotting KC(106) () or KC(culture) () values against the logarithm of drug concentration. Correlation coefficients were 0.92 () and 0.98 () for A and 0.94 () and 0.99 () for B.

Influence of BG on TMZ-Induced Impairment of Cell-Mediated Immune Responses of MNCs. To further confirm the relationship between OGAT activity and MNC sensitivity to TMZ, the influence of BG on TMZ-mediated impairment of immune responses induced by IL-2 or by allogeneic MT-2 cells was investigated. Studies were performed to establish whether exposure to BG before and during TMZ treatment was sufficient to increase MNC sensitivity to the triazene compound. It was also tested whether additional exposure to the OGAT inhibitor after TMZ treatment (i.e., during the stimulation period with IL-2 or MT-2 cells) was required for amplification of the suppressive effects of the triazene compound on MNC function.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Influence of BG on TMZ-mediated impairment of proliferative responses of MNCs endowed with high OGAT activity. MNCs obtained from donor 21 (OGAT activity, 903 ± 77 fmol/mg protein) were incubated in CM alone or in CM containing BG (2 µM) for 1 h. Control and BG-treated MNCs were then exposed to CM alone () or to CM containing 62.5 µM TMZ (), 125 µM TMZ (), 250 µM TMZ (), or 500 µM TMZ (), without removing the OGAT inhibitor, for 4 h. Thereafter, cells were washed and tested for their ability to proliferate in response to IL-2 (A) or allogeneic MT-2 cells (B). Stimulation of MNCs pretreated with BG or TMZ plus BG was performed either in CM alone or in CM containing 2 µM BG. MNCs proliferation was evaluated by the 3H-TdR incorporation assay and expressed in terms of mean cpm calculated on four replicates. Bars, S.E. The IC50 values and their confidential limits for TMZ relative to MNCs proliferative response to IL-2 (A) were TMZ alone, 216 (196-237) and TMZ plus BG, 149 (118-180). The IC50 values and their confidential limits for TMZ relative to MNCs proliferative responses to allostimulation (B) were TMZ alone, 201 (175-232) and TMZ plus BG, 98 (72-134). Because no overlapping occurred between confidential limits relative to IC50, in all cases differences between IC50 values relative to TMZ and IC50 values relative to TMZ plus BG were statistically significant (p < .05). When stimulation of MNCs pretreated with TMZ plus BG was performed without the OGAT inhibitor, no increment was observed in the suppressive effect of the triazene compound (data not shown).

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Influence of BG on TMZ-mediated impairment of LAK cell and CTL generation of MNCs endowed with high OGAT activity. MNCs obtained from donor 21 were incubated in CM alone or in CM containing BG (2 µM) for 1 h. Control and BG-treated MNCs were then exposed to CM alone () or to CM containing 62.5 µM TMZ () , 125 µM TMZ (), 250 µM TMZ (), or 500 µM TMZ (), without removing the OGAT inhibitor, for 4 h. Thereafter, cells were washed and incubated in the presence of IL-2 (500 IU/ml) for 4 days (A) or in the presence of allogeneic inactivated MT-2 cells for 5 days (B). Stimulation of MNCs pretreated with BG or TMZ plus BG was performed either in CM alone or in CM containing 2 µM BG. At the end of incubation period, effector cells were recovered and tested for cytotoxic activity against Daudi cells (LAK cells; A) or MT-2 cells (CTL; B), using a 51Cr-release assay (see Materials and Methods). Cytotoxic activity (y-axis) is expressed in terms of KC(culture) (i.e., the number of target cells sacrificed by the total number of effector cells present in each culture at the end of stimulation period, see Materials and Methods), calculated on four replicates. The IC50 values and their confidential limits for TMZ relative to LAK cell cytotoxic activity (A) were TMZ alone, 252 (232-278) and TMZ plus BG, 144 (118-180). The IC50 values and their confidential limits for TMZ relative to CTL cytotoxic activity (B) were TMZ alone, 221 (154-319) and TMZ plus BG, 82 (98-62). Because no overlapping occurred between confidential limits relative to IC50, in all cases, differences between IC50 values relative to TMZ and IC50 values relative to TMZ plus BG were statistically significant (p < .05). When stimulation of MNCs pretreated with TMZ plus BG was performed without the OGAT inhibitor, no increment in the suppressive effect of the triazene compound was observed (data not shown).

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Influence of BG on TMZ-mediated impairment of cell-mediated immune responses to allogeneic MT-2 of MNCs endowed with low OGAT activity. MNCs obtained from donor 19 (OGAT activity, 375 ± 30 fmol/mg protein) were incubated in CM alone or in CM containing BG (2 µM) for 1 h. Control and BG-treated MNCs were then exposed to CM alone () or to CM containing 62.5 µM TMZ (), 125 µM TMZ (), 250 µM TMZ (), or 500 µM TMZ (), without removing the OGAT inhibitor, for 4 h. Thereafter, cells were washed and incubated in the presence of inactivated allogeneic MT-2 cells for 5 days to test their ability to proliferate (A) and to generate CTL (B). Stimulation of MNCs pretreated with BG or TMZ plus BG was performed either in CM alone or in CM containing 2 µM BG. A, at the end of the incubation period, MNC proliferation was evaluated by the 3H-TdR incorporation assay and expressed in terms of mean cpm calculated on four replicates. Bars, S.E. Statistical analysis of the differences between groups treated with TMZ alone or with TMZ plus BG was performed with Student's t test. *p < .05, **p < .01. The IC50 values and their confidential limits for TMZ relative to MNC proliferative responses to allostimulation were TMZ alone, 154 (124-190) and TMZ plus BG, 87 (57-129). B, the cytotoxic activity of CTL was tested against MT-2 cells using a 51Cr-release assay (see Materials and Methods) and expressed in terms of KC(culture) (see legend for Fig. 5) calculated on four replicates. Statistical evaluation of the differences between groups treated with TMZ alone or with TMZ plus BG was performed by ANCOVA. **p < .01. The IC50 values and their confidential limits for TMZ relative to CTL cytotoxic activity were TMZ alone, 118 (98-144) and TMZ plus BG, 77 (51-129). When stimulation of MNCs pretreated with TMZ plus BG was performed without the OGAT inhibitor, no increment in the suppressive effect of the triazene compound was observed (data not shown).

Effect of Multiple TMZ Treatments on Cell-Mediated Immune Function of MNCs. Pharmacokinetics studies have shown that patients receiving TMZ on a repeated dose schedule (200 mg/m² × 5 days) attain a maximum plasma concentration of about 50 µM after each TMZ administration (Newlands et al., 1992). We therefore decided to verify whether an in vitro TMZ treatment of MNCs resembling the clinical condition could determine an impairment in cell-mediated immunity.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Effect of multiple TMZ treatments on cell-mediated immune function of MNCs. MNCs obtained from donor 21 were incubated in CM alone () or CM containing 2 µM BG () for 1 h. Then, either untreated () or BG-treated MNCs () were exposed daily to 50 µM TMZ for 5 days. The 5-day TMZ treatment of MNCs preexposed to BG was performed in CM containing 2 µM BG. Four hours after the last TMZ treatment, MNCs were tested for their ability to respond to IL-2 and allogeneic MT-2 tumor cells in terms of proliferation (A) or generation of cytotoxic effector cells (B). A, MNC proliferation was evaluated by the 3H-TdR-incorporation assay and expressed in terms of mean cpm calculated on four replicates. Bars, S.E. Statistical analysis of the differences between groups treated with TMZ alone or with TMZ plus BG was performed with Student's t test. **p < .01. B, the cytotoxic activity of LAK cells and CTL was tested against Daudi or MT-2 cells, respectively, using a 51Cr-release assay (see Materials and Methods) and expressed in terms of KC(culture) (see legend for Fig. 5), calculated on four replicates. Statistical evaluation of the differences between groups treated with TMZ alone or with TMZ plus BG was performed by ANCOVA. **p < .01.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

TMZ belongs to a class of antitumor drugs that show high immunosuppressive activity in murine models (Giampietri and Bonmassar, 1978; Nardelli et al., 1984). However, no studies have been performed up to now to investigate whether the main biochemical lesion underlying the antitumor properties of TMZ (i.e., methylation of O⁶-G) might also be responsible for immunosuppressive effects. The results illustrated in the present report show that TMZ inhibits proliferation-dependent immune functions of human MNCs in vitro. Moreover, for the first time it has been demonstrated that methylation of O⁶-G is, at least in great part, responsible for the immunosuppressive effects of the drug.

Clonal expansion of lymphocytes is central to the induction of both humoral and cellular immune function. It is well documented that a number of antitumor agents exert immunosuppressive activity through a mechanism that is related, at least in part, to the impairment of lymphocyte proliferation induced by several stimuli. The results illustrated in Fig. 2 show that this also occurs in the case of TMZ. The drug markedly inhibited MNC proliferation induced by IL-2 or by allogeneic tumor cells, thus preventing expansion of effector cells. In agreement with previous studies performed on tumor cells (Tisdale, 1987; Baer et al., 1993; D'Atri et al., 1995; Tentori et al., 1995, 1997; Liu et al., 1996; Wedge et al., 1996a), the antiproliferative effects of TMZ were strictly dependent on methylation of O⁶-G because MNCs sensitivity to the drug was inversely correlated with their OGAT activity.

In vitro treatment with TMZ also impaired the generation in cytotoxic effector cells in MNCs incubated with IL-2 (i.e., LAK cells, Fig. 3A) or cocultivated with allogeneic tumor cells (i.e., CTL, Fig. 3B). As observed for proliferative responses induced by the same stimuli, sensitivity to TMZ was inversely correlated with OGAT levels of MNCs. This finding, together with the observation that the IC₅₀ values relative to TMZ calculated on KC(culture) were lower than those calculated on KC(10⁶), suggests that inhibition of LAK cell and CTL generation is largely dependent on the antiproliferative effects of TMZ. This hypothesis is also supported by the observation that NK cell activity, which does not involve MNC expansion, was only moderately affected by TMZ, with no obvious relationship with OGAT function (Table 1). In addition, TMZ-mediated impairment of NK activity was not influenced by MNC preexposure to BG (data not shown).

The involvement of O⁶-G methylation in the immunosuppressive effects of TMZ is further confirmed by the finding that continuous exposure to BG increased the sensitivity to TMZ of MNCs endowed with high OGAT levels. It must be emphasized that continuous exposure to BG ensured persistent abrogation of OGAT activity of MNCs, as extensively described in Results. Both proliferative responses and generation of cytotoxic effectors induced by IL-2 or allogeneic tumor cells were impaired to a higher extent in MNCs treated with TMZ plus BG with respect to MNCs treated with TMZ alone (Figs. 4 and 5). In MNCs possessing low OGAT levels, the effect of BG on sensitivity to TMZ was limited, and the decrease in the IC₅₀ values, although present, did not reach statistical significance (Fig. 6). However, if the influence of BG was analyzed on the basis of each single TMZ concentration, a potentiating effect of the inhibitor could be demonstrated at the lowest drug concentrations (i.e., 62.5 and 125 µM). On the other hand, when 250 or 500 µM of TMZ was used, the drug alone was highly effective in abrogating immune responses, and therefore exposure of MNCs to BG did not result in a further increase in immunosuppression. Similar patterns of differential effects of BG on the susceptibility to triazenes and other O⁶-G alkylating agents have been found in human tumor cells (Dolan et al., 1991; Baer et al., 1993; Wedge et al., 1996a).

Presently, the treatment schedule recommended for TMZ is 200 mg/m² × 5 days. Pharmacokinetics studies have shown that patients receiving TMZ with this schedule attain a maximum plasma concentration of about 50 µM after each TMZ administration (Newlands et al., 1992). The results of Figs. 4, 5, and 6 show that a single treatment with 62.5 µM TMZ (concentration comparable to that obtainable in the clinic) did not impair the immune functions of MNCs endowed with either high or low OGAT activity. However, when MNCs with high enzyme levels were exposed to five subsequent treatments with 50 µM TMZ, a significant inhibition was observed of both proliferative response and cytotoxic effector generation induced by IL-2 or allogeneic tumor cells (Fig. 7). Also in this case, the combined treatment of TMZ plus BG was more effective than TMZ alone in inhibiting MNC immune functions.

Taken together, these data demonstrate that TMZ down-regulates both natural and antigen-dependent cell-mediated immune functions in vitro and suggest that this could also occur in patients with cancer treated with the agent. Indeed, it has already been shown that in patients with melanoma receiving TMZ (150 mg/m²) on a 5-day schedule, a significant and progressive depletion of OGAT activity occurred in MNCs (Lee et al., 1994). If a complete loss of OGAT activity occurs in MNCs, an excess of the toxic O⁶-MeG lesions and, hence, immunodepression might be expected to take place. Complete suppression of OGAT activity of MNCs might occur preferentially in MNCs endowed with low levels of the protein, as already described in patients with melanoma treated with the methylating triazene compound CB10-277 (Sio et al., 1992).

On the basis of the results obtained in vitro, it is reasonable to predict that the NK function should not be substantially depressed by TMZ shortly after treatment in vivo. On the contrary, the ability of MNCs to generate LAK cells and antigen-dependent CTL is expected to be reduced in patients undergoing TMZ treatment. Therefore, the determination of individual OGAT levels in normal lymphocytes could provide valuable information into the susceptibility of the host to TMZ-induced impairment of the immune system.

The use of BG to potentiate the antitumor effects of TMZ could also result in a further increase of TMZ-induced immunodepression. However, a continuous exposure of MNCs to the inhibitor is required to enhance the immunotoxicity of TMZ. These findings should be taken into consideration in designing the protocols of immunochemotherapy. Actually, treatment with triazenes combined with BG could interfere with the immunostimulating activity of cytokines such as IL-2 or active immunization of patients against tumor-associated antigen.