Evaluation of the genotoxic effects of the boron neutron capture reaction in human melanoma cells using the cytokinesis block micronucleus assay

Abstract

The present work reports on the genotoxicity of the boron neutron capture (BNC) reaction in human metastatic melanoma cells (A2058) assessed by the cytokinesis block micronucleus assay (CBMN) using p-borono-L-phenylalanine (BPA) as the boron delivery agent. Different concentrations of BPA (0.48, 1.2 and 2.4 mM) and different fluences of thermal neutrons were studied. Substantial genotoxic potential of α and lithium particles generated inside or near the malignant cell by the BNC reaction was observed in a dose–response manner as measured by the frequency of micronucleated binucleated melanoma cells and by the number of micronuclei (MN) per binucleated cell. The distribution of the number of MN per micronucleated binucleated cell was also studied. The BNC reaction clearly modifies this distribution, increasing the frequency of micronucleated cells with 2 and, especially, ≥3 MN and conversely decreasing the frequency of micronucleated cells with 1 MN. A decrease in cell proliferation was also observed which correlated with MN formation. A discrete genotoxic and anti-proliferative contribution from both thermal neutron irradiation and BPA was observed and should be considered secondary. Additionally, V79 Chinese hamster cells (chromosomal aberrations assay) and human lymphocytes (CBMN assay) incubated with different concentrations of BPA alone did not show any evidence of genotoxicity. The presented results reinforce the usefulness of the CBMN assay as an alternative method for assessment of the deleterious effects induced by high LET radiation produced by the BNC reaction in human melanoma cells.

Introduction

The capture of thermal neutrons (n_th, low energy, average value 0.025 eV) by the minor stable isotope of boron (¹⁰B) releases α and lithium particles. This reaction is called the boron neutron capture (BNC) reaction [¹⁰B(n,α)⁷Li: ¹⁰B + n_th→[¹¹B]→⁴He (α) + ⁷Li + 2.3 MeV]. The propagation of α and ⁷Li particles in biological tissues is characterized by a short range (~9 and 5 μm, respectively) and a high linear energy transfer (LET) with a remarkable destructive power (reviewed in Hawthorne, 1993; Coderre and Morris, 1999). This reaction has thus been primarily proposed for the treatment of some types of malignant melanoma and glioma in a targeted binary radiation therapy called boron neutron capture therapy (BNCT).

The present work reports on the genotoxicity induced by the BNC reaction in human metastatic melanoma cells (A2058) assessed by the cytokinesis block micronucleus assay (CBMN assay) using p-borono-l-phenylalanine (BPA) (Figure 1) as the boron delivery agent. Different concentrations of BPA as well as different periods of irradiation with thermal neutrons were studied.

The CBMN assay is a standard method for assessing DNA damage and has been reported as a useful biomarker for evaluation of the genotoxic effects of ionizing radiation (Brooks et al., 1990; Fenech et al., 1990; Ono et al., 1996; Shibamoto et al., 1998; Streffer et al., 1998; Gil et al., 2000a). Formation of micronuclei (MN) is also known to be associated with mitosis-linked cell death (reviewed in Cohen-Jonathan et al., 1999).

The CBMN assay has been studied in different cell types (Fenech, 2000), either normal or tumoral, including melanoma cells (Champion et al., 1995; Courdi et al., 1995; Widel et al., 1997; Poma et al., 1999). Fenech (2000) has extensively reviewed the inherent advantages of the CBMN assay and a broad range of different applications.

The possible intrinsic genotoxic and anti-proliferative contributions of the two components of the BNC reaction, the thermal neutrons and the boron delivery agent BPA, were also studied. Neutron capture reactions involving other nuclides, such as hydrogen [¹H(n,γ)²H] and nitrogen [¹⁴N(n,p)¹⁴C], as well as proton recoils from the fast neutrons [¹H(n,n')¹H), are present in the final reaction (Gupta et al., 1994) and can be assessed by the thermal neutron irradiation of non-BPA incubated cells. The same is valid for the background of γ-rays from the reactor, which can be physically monitored. This low LET radiation increases the genotoxicity of the BNC final reaction, as it is a function of irradiation time, and its contribution can also be evaluated by comparison with genotoxicity curves of ⁶⁰Co γ-rays.

BPA is an adequate ¹⁰B delivery agent for melanoma BNCT (Coderre et al., 1987,1991; Mishima et al., 1989a,b; ,Matalka et al.1993; Mishima, 1997) and is selectively accumulated in melanoma cells since there is an elevated amino acid transport rate at the cell membrane (Coderre and Morris, 1999). Moreover, BPA is currently being tested in various clinical trials not only for malignant melanoma but also for other types of cancer, namely glioblastoma multiforme. BPA, as previously shown through in vitro and in vivo experiments (Coderre et al.1992; ,Matalka et al.1994), is also selectively accumulated in other non-melanoma tumoral cells presently having a key role in BNCT. Concerning this point it seems relevant to evaluate the genotoxic effects of BPA not only in A2058 human melanoma cells (CBMN assay) and its contribution to the total effect of the BNC reaction in these cells, but also in non-tumoral cells, namely V79 Chinese hamster cells (chromosomal aberrations assay, CA assay) as well as in human lymphocytes (CBMN assay).

Materials and methods

Chemicals and culture medium

Fetal calf serum, RPMI medium, Ham's F-10 medium, cytochalasin B (Cyt-B), l-glutamine, penicillin and streptomycin were purchased from Sigma (St Louis, MO). 4-Borono-l-phenylalanine ¹⁰B enriched (>99%) was obtained from KatChem (Prague, Czech Republic). Acetic acid, methanol and Giemsa dye were obtained from Merck (Darmstadt, Germany). Phytohaemagglutinin (PHA; HA 15) was purchased from Murex (Dartford, UK) and reconstituted in 5 ml of sterile water. Heparin was obtained from Braun (Melsungen, Germany). Colchicine was purchased from Fluka (Buchs, Switzerland) and trypsin from Difco Laboratories (Detroit, MI).

Melanoma cell culture and BPA incubation

Human melanoma A2058 cells (~2.0×10⁵) were cultured in RPMI medium supplemented with 10% fetal calf serum, penicillin (100 IU/ml) and streptomycin (100 μg/ml) and incubated at 37°C under an atmosphere containing 5% CO₂. This cell line was previously established from a brain metastasis of a 43-year-old man (Todaro et al., 1980) and was kindly provided by Dr O.Csuka (Budapest).

For the BPA-treated cells medium supplemented with a BPA stock solution was prepared with a final BPA concentration of 2.4 mM (500 μg/ml, 24.0 p.p.m. ¹⁰B). Cells were seeded in 25 cm² tissue culture flasks (Greiner, Frickenhausen, Germany) and incubated either with 5 ml of BPA medium (0.48, 1.2 and 2.4 mM) or with BPA-free culture medium. The cells were grown as monolayers for 48 h and then irradiated with thermal neutrons.

Thermal neutron irradiation protocol

Irradiation of human melanoma cells took place at the vertical access of the thermal column of the Portuguese Research Reactor (RPI). Characterization of the radiation field and reduction of the background γ-radiation from the reactor were essential tasks for these radiobiological experiments and have been described elsewhere (Gonçalves et al., 1999). The neutron spectrum is basically thermal, with a low epithermal component and reduced γ-ray background. The thermal neutron flux (φ_th), epithermal neutron flux (φ_epi) and γ-ray dose rate in air (Dγ, air) are of the order of 5.7×10⁷ n_th/cm²/s¹, 2×10⁴ n_epi/cm²/s¹ and 0.2–0.3 Gy/h, respectively.

Three different periods of irradiation were studied, 30, 60 and 120 min, corresponding to average fluences (± SD) of 1.1 ± 0.06, 2.2 ± 0.01 and 4.3 ± 0.30×10¹¹ n_th/cm² as measured by individual dosimetry using gold foil activation. Two independent experiments were performed. Controls included cells irradiated with thermal neutrons without BPA incubation (neutron control cells), cells incubated with BPA without thermal neutron irradiation (BPA control cells) and cells unirradiated and without BPA incubation (background control cells). BPA control cells were used to assess the genotoxic potential of BPA as compared with background control cells using a two-tailed Student's t-test for statistical analysis.

CBMN assay in A2058 human melanoma cells

For the CBMN assay 6 h after the irradiation with thermal neutrons the cell culture medium was removed, the cells washed and placed in BPA-free culture medium. Cyt-B was added at a final concentration of 6 μg/ml (Van Hummelen and Kirsh-Volders, 1990) and the cells were grown for a further 22 h for recovery of binucleated A2058 cells. The cells were then harvested by trypsinization, rinsed and submitted to mild hypotonic treatment as described elsewhere (Van Hummelen and Kirsh-Volders, 1990; Gil et al., 2000a,b). The centrifuged cells were placed on dry slides and smears were made. After air drying the slides were fixed with cold methanol (30 min). One day later the slides were stained with Giemsa (4% v/v in 0.01 M phosphate buffer, pH 6.8) for 10 min. For each experimental point 1000 binucleated (BN) A2058 cells with well-preserved cytoplasm were scored. MN were identified according to the criteria of Caria et al. (1995) using a 1250× magnification on a light microscope. Two indices were evaluated, number of micronuclei per binucleated cell (MN/BN), which represents the average number of MN per BN cell, and frequency of micronucleated binucleated melanoma cells (%MNBN), which represents the fraction of cytokinesis blocked (BN) cells with MN, regardless of the number of MN per BN cell.

The decrease in cell proliferation for the experiments described above was assessed using the frequency of BN cells (%BN). For this index 1000 human melanoma cells with well-preserved cytoplasm were analyzed according to number of nuclei using a 500× magnification.

Evaluation of BPA genotoxicity in human lymphocytes: CBMN assay

Aliquots of 500 μl of whole blood from four healthy donors were cultured in 4.5 ml of Ham's F-10 medium supplemented with 24% fetal calf serum, penicillin (100 IU/ml), streptomycin (100 μg/ml), 1% l-glutamine and 1% heparin (50 IU/ml). Lymphocytes were stimulated using 25 μl of PHA and incubated at 37°C. At 24 h culture the cells were exposed to different doses of BPA (0.12, 0.24, 0.36 and 0.48 mM) for 3 h. In these experiments BPA was not previously dissolved in the culture medium, as for the A2058 and V79 Chinese hamster cell experiments. Instead, BPA was prepared as a stock solution of 48.1 mM dissolved in 0.2 M HCl. Background control cells were incubated with the maximum quantity of 0.2 M HCl used for the 0.48 mM BPA concentration (this volume did not exceed 1%). After BPA incubation the cells were centrifuged and placed in fresh culture medium. At 44 h culture Cyt-B was added (6 μg/ml). At 72 h culture cells were harvested by centrifugation, treated twice with 5 ml of a mixture of RPMI 1640:deionized water 4:1 (pH 7.2) supplemented with 2% fetal calf serum and submitted to a mild hypotonic treatment as described above. Preparation of slides and scoring of MN was also performed as described for melanoma cells in 1000 BN lymphocytes and %MNBN, which represents the fraction of cytokinesis blocked (BN) cells with MN was assessed. In order to evaluate cell proliferation, 1000 human lymphocytes cells were analyzed using a 500× magnification.

Statistical analysis for evaluation of the genotoxic potential of BPA was carried out using a two-tailed Student's t-test.

Evaluation of BPA genotoxicity in V79 Chinese hamster cells: CA assay

Wild-type V79 Chinese hamster (MZ) cells were kindly provided by Prof. H.R.Glatt (Mainz and Postdam). These cells were cultured in the same culture medium used for human melanoma cells and incubated at 37°C under an atmosphere containing 5% CO₂. A BPA stock supplemented culture medium was prepared with a final BPA concentration of 2.4 mM (500 μg/ml, 24.0 p.p.m. ¹⁰B).

V79 cells (~1.0×10⁵) were seeded in 25 cm² tissue culture flasks (Greiner) and incubated with either 5 ml of BPA medium (0.48, 1.2 and 2.4 mM) or BPA-free culture medium. The cells were grown as monolayers in this medium for 64 h. The medium was then removed and colchicine added in BPA-free culture medium at a final concentration of 0.6 μg/ml. Cells were grown for a further 2.5 h and then harvested by trypsinization. After a 3 min hypotonic treatment with 75 mM KCl at 37°C the cells were fixed with methanol/acetic acid (3:1) and slides prepared and stained with Giemsa (4% v/v in 0.01 M phosphate buffer, pH 6.8) for 10 min. Six independent experiments were carried out for each BPA concentration as well as for untreated cultures and 100 metaphases were observed using a 1250× magnification and a light microscope. Scoring of the different types of aberrations followed the criteria described by Rueff et al. (1993). Mitotic index (MI) was estimated by scoring 1000 V79 cells using a 500× magnification under a light microscope. Statistical analysis for evaluation of the genotoxic potential of BPA was carried out using a two-tailed Student's t-test.

Results

Figure 2 shows the genotoxic effects of the BNC reaction in human melanoma cells preincubated for 48 h with three different concentrations of BPA (0.48, 1.2 and 2.4 mM) and irradiated with thermal neutrons for three periods of time, 30, 60 and 120 min, corresponding to average fluences of 1.1, 2.2 and 4.3×10¹¹ n_th/cm². In this figure two important indices concerning the CBMN assay are represented, MN/BN (average ± SD) (Figure 2A) and %MNBN (average ± SD) (Figure 2B).

The human melanoma cell line used (A2058) has a high background frequency of MN compared with normal cells (>10-fold), namely human lymphocytes (Gil et al.2000b). This inherent characteristic, which reflects a high level of spontaneous DNA damage, has been reported for other human melanoma cell lines (Champion et al., 1995; Poma et al., 1999).

Regarding both indices, MN/BN and %MNBN, a dose–response relationship was observed for both BPA concentration and thermal neutron fluence. Some differences may be observed, namely in the shapes presented by the curves. The MN/BN curves present a linear dependence on fluence whereas the %MNBN curves suggest some kind of saturation for the higher fluence studied, especially for the two highest BPA concentrations.

The number of MN per micronucleated BN cell is also an important indicator of the degree of lesion induced by the BNC reaction. In Figure 3 the relative frequency of each type of micronucleated cell classified by number of MN (1, 2, 3, 4, >4) is represented for the three fluences studied as well as for micronucleated BPA control cells (2.4 mM). These micronucleated cells have essentially 1 or 2 MN per BN cell (~80 and 15%, respectively), a pattern that is also present in the human melanoma control cells (~84 and ~11%). However, thermal neutron irradiation of BPA-incubated cells distinctly modified the number of MN per micronucleated cell. For the higher thermal neutron irradiation period the frequency of micronucleated cells with 1 MN decreased to values of ~45%, whereas the frequency of micronucleated cells with 2 and 3 MN increased (~25 and 15%, respectively). A clear dose-dependent relationship was found for the frequency of micronucleated cells with 3 MN, which seems to be a good indicator of the extent of DNA damage. In addition, heavily damaged cells containing >3 MN were observed, especially at the higher irradiation level (~15%).

A dose-dependent decrease in %BN was observed for human melanoma cells incubated with the two highest concentrations of BPA (1.2 and 2.4 mM) and irradiated with thermal neutrons (Figure 4). Regarding each thermal neutron fluence studied, a dose-dependent decrease in %BN was also observed with increasing BPA concentrations. For the higher fluence studied and for the higher BPA concentration this proliferation index decreased to ~50% of control cells. In addition, a very good correlation (r = –0.98) was found comparing the decrease in %BN with the increase in %MNBN for irradiated melanoma cells incubated with 2.4 mM BPA.

Thermal neutron irradiation in the absence of BPA incubation resulted in a slight increase in both MN/BN and %MNBN (Figure 2A and B). For the highest fluence studied these indices presented approximately double the values (2.1-fold) compared with background melanoma cell controls. The number of MN per micronucleated BN cell shows a pattern similar to background cells and BPA control cells. The micronucleated cells usually have 1 or 2 MN and the fluence of thermal neutrons did not significantly affect this index. Thermal neutron irradiation itself reduced %BN by ~10% but not in a dose-dependent manner (Figure 4).

Concerning the genotoxic and anti-proliferative effects of BPA in human melanoma cells, BPA incubation of non-irradiated cells (BPA control cells) mildly increased both indices (MN/BN and %MNBN) and decreased %BN as compared with the background values. These results are presented in detail in Figure 5A. For the highest BPA concentration studied (2.4 mM) the MN/BN index was 0.126 ± 0.031 for BPA control cells, whereas 0.097 ± 0.016 MN/BN was found as the background value (NS, P = 0.08). %MNBN was 9.9 ± 1.8 for BPA control cells versus 7.6 ± 1.3 for melanoma background cells (P < 0.05). A value of P = 0.05 was found for 0.48 mM BPA (%MNBN of 9.5 ± 1.7).

V79 Chinese hamster cells (CA assay) and human lymphocytes (CBMN assay) incubated with different concentrations of BPA alone showed no evidence of genotoxicity. No significant increases in the frequencies of chromosomal aberrant cells excluding gaps (%CAEG) were observed in V79 Chinese hamster cells for the same concentrations of BPA (0.48, 1.2 and 2.4 mM) as used in melanoma cells and with a long incubation period of >60 h (Figure 5B). The average %CAEG presented by negative control cultures (1.5 ± 0.5%) was within the normal values presented by this cell line in previous experiments (Oliveira et al., 1997; Alves et al., 2000). BPA incubation increased this background by only 0.8% at the maximum level (0.48 mM BPA; NS). No evidence of any anti-proliferative effects due to BPA incubation was observed.

In addition, human lymphocytes incubated with BPA (0.12–0.48 mM) for a 3 h period (G₁ phase) did not show evidence of genotoxicity using the CBMN assay, as shown in Figure 5C. The %MNBN values presented by negative control cultures (0.58 ± 0.29%) were within the normal range found in healthy donors (Caria et al., 1995; Gil et al., 2000b; Oliveira et al., 2000). No evidence of any anti-proliferative effects due to BPA incubation was observed.

Discussion

The promising usefulness of BNCT in the treatment of some types of malignant melanoma (Mishima et al., 1989a,b; Busse et al., 1997; Mishima, 1997) has prompted a multi-disciplinary effort in order to improve its efficacy. The cytotoxic effects of the BNC reaction have been extensively evaluated using clonogenic assays (percent tumor cell survival) either using in vitro or in vivo approaches (reviewed in Coderre and Morris, 1999). There are also some reports using the MN assay as the end-point for study of this reaction. These reports have been mainly performed using an experimental mouse squamous cell carcinoma model (SCCVII cells) in C3H/He mice (Ono et al.1996; ,Masunaga et al.1999). Additional data on the pattern of genotoxicity induced by this high LET radiation towards a highly invasive human melanoma cell line (A2058), using the CBMN assay, is thus important to help in understanding the mechanisms underlying cell death/damage.

The MN/BN index roughly represents the average number of chromosomal lesions present in a BN human melanoma cell and %MNBN represents the frequency of DNA damaged cells regardless of the number of lesions per cell. Both indices are a measure of the genotoxic burden caused by the BNC reaction and are complementary. Although %MNBN increases in a typical dose-dependent manner, the genotoxicity observed in terms of MN/BN is higher. The different shapes of the two curves can be explained by the increase in the number of α-particles produced per cell, with no parallel increase in the number of cells damaged. The increase in multi-micronucleated cells, which enhances the MN/BN index, is also in agreement with a similar increase in cell cycle arrest (as measured by the BN index), especially for the two highest BPA concentrations tested.

Mechanistic knowledge on DNA and cell damage induced by α-particles remains limited (National Research Council, 1999). Due to the high LET nature of this radiation, traversal of cells by α-particles is considered to be lethal. The insult from α-particles is concentrated in a relatively small densely ionizing track that crosses the cell in <10^–12 s and deposits a large localized energy (~10–50 cGy). In addition, individual damage is mainly due to direct ionization more than to hydroxyl radical reactions (National Research Council, 1999). Clonogenic lethality for α-particles has been extensively described and it has been well established that high LET radiation produces more biological damage per unit absorbed dose than sparsely ionizing radiation, such as X- and γ-rays (National Research Council, 1999). Dose-dependent chromosomal damage has also been studied and associated with the deleterious effects of α-radiation. In fact, α-particles induce CAs (Edwards et al., 1980; ,Lloyd et al.1988; Bauchinger and Schmid, 1998; ,Pohl-Rülling et al.2000), MN (Bilbao et al.1989; ,Brooks et al.1990; ,Ono et al.1996; ,Belyakov et al.1999) and sister chromatid exchanges (Aghamohammadi et al.1988; Nagasawa and Little, 1992; Lehnert and Goodwin, 1997). Recent in vitro studies with α-particle emitters have shown that up to 80% of cells traversed by one α-particle survive damage, yet sustain a considerable increase in mutation frequency (Hei et al.1997). Traversal of cells by up to four α-particles further increases mutation frequency, while having a moderate cytotoxic effect. Additionally, other studies have shown that the biological effects of α-radiation are not limited to cells actually traversed. Enhanced frequencies of mutagenic effects have been observed in non-irradiated `bystander' cells (Nagasawa and Little, 1992; Zhou et al., 2000), including chromosomal instability (Lorimore et al., 1998; ,Kadhim et al.1994), possibly through reactive oxygen species (Narayanan et al.1997; Lehnert and Goodwin, 1997; ,Wu et al.1999). Concerning this point, we have measured the concentration of malondialdehyde in the supernatants of melanoma cell cultures exposed to the BNC reaction and have found no difference from unirradiated controls (data not shown).

Monte Carlo calculations for the BNC reaction suggested that ~10% of the total dose is due to external ¹⁰B, 45% from ¹⁰B present in the cytoplasm and 45% from the nucleus (reviewed in Fairchild et al.1990). Accordingly, the results presented here could also be the result of genotoxicity due to α-particles generated in the cytoplasm or even in the extracellular compartment.

The BNC reaction has also proven to interfere with cell cycle progression as a function of BPA concentration and thermal neutron fluence. Delays in cell cycle progression, with the purpose of providing time for DNA repair and also for induction of transcription of genes that could enhance DNA repair (reviewed in Elledge, 1996), have been unequivocally related to checkpoint disturbances. Low LET ionizing radiation has been shown to induce cell cycle arrest in G₁ before DNA replication, in S phase and also in G₂ before mitosis. Inhibition of cdk–cyclin complexes, which regulate both the G₁/S and G₂/M checkpoints, mediate these arrests (reviewed in Cohen-Jonathan et al., 1999). High LET radiation such as α-particles induces more extensive delays in the S or G₂ phases as compared with low LET radiation, although Azzam et al. (2000) have shown the existence of a p53-mediated G₁ arrest in human diploid fibroblasts exposed to α-particles.

Thermal neutron irradiation of non-BPA-incubated cells slightly increased the yields of MN/BN and %MNBN as a function of fluence. This effect, which approximately duplicates the values of both indices for the fluence of 4.3×10¹¹ n_th/cm², is mainly due to the γ-ray background from the reactor and other reactions following the interaction of thermal neutrons with biological tissues. In a 2 h irradiation period, which corresponds to the highest fluence studied, the cells were exposed to ~0.4–0.6 Gy of γ-rays. The CBMN assay was performed under the same experimental conditions as described for the BNC reaction but with ⁶⁰Co γ-rays (data not shown) and the linear fitting parameters of the MN/BN and MNBN (as decimal fraction) curves were 0.120 ± 0.007 and 0.080 ± 0.005 per Gy, respectively (γ-ray dose rate 31.01 mGy/min, dose range 0.25–3 Gy). The γ-ray background dose of 0.4–0.6 Gy thus results, by extrapolation, in values of 0.05–0.07 for MN/BN and 3.2–5.0 for %MNBN.

The BPA contribution to the genotoxic burden of the BNC reaction was also assessed and shown to be secondary, although we cannot exclude that BPA alone is genotoxic to human melanoma cells. Nevertheless, no dose–response dependence for BPA genotoxicity nor a substantial increase in %MNBN values was observed. The results obtained with V79 Chinese hamster cells (CA) and with human lymphocytes (CBMN) do not show any kind of genotoxicity and/or anti-proliferative effects even at high doses (120 μM–2.4 mM). These two cell types, with their inherent sensitivities and different genetic end-points, were used in order to achieve a maximal likelihood of properly ascertaining a possible genotoxic activity of BPA alone. The results are in agreement with the idea that boron compounds are not positive in the usual mutagenicity tests performed (reviewed in Commission of the European Community, 1993; World Health Organization, 1998), although only a few studies of this nature are available, mainly with inorganic boron compounds.

The rationale for the use of BPA in BNCT has been the assumption that this compound belongs to the class of melanin precursors as it is an analog of phenylalanine. This argument is still a matter of debate (Hawthorne, 1993) and other mechanisms could be involved, namely chemical complex formation of BPA with melanin monomers (Mishima, 1997). In addition, active transport of this compound through the melanoma cell membrane and that of other non-melanoma tumors has been observed (reviewed in Coderre and Morris, 1999) and recent data on the underlying mechanisms of this transport published (Wittig et al., 2000). These observations suggest utilization of the compound in other metabolic processes in tumoral cells but the reason why non-melanoma cells selectively accumulate BPA remains unknown (Wittig et al., 2000).

The slight increase in %MNBN and decrease in %BN on BPA treatment herein observed for human melanoma cells could be partially explained by the chemical resemblance to phenylalanine. In fact, the cytotoxicity and genotoxicity of melanin synthesis precursors have long been known for melanoma cells (reviewed in Herlyn and Houghton, 1992) and some reports have been published concerning this matter (Miranda et al., 1997; Poma et al., 1999). However, other mechanisms could also be involved, since in other non-melanoma tumors, e.g. SCCVII cells from mice treated with BPA, the MN frequencies were also slightly higher (~2-fold) than those observed in the control group (Masunaga et al.1999).

An overall analysis of the presented results reveals a substantial genotoxic potential associated with the BNC reaction due to the generation of α and lithium particles inside or near the malignant cell. Discrete genotoxic contributions from both thermal neutron irradiation alone and the boron delivery agent BPA alone were observed and should be considered secondary.

Finally, it should be pointed out that the interest of this reaction is far beyond its application in malignant melanoma and glioma. Other tumors are significantly destroyed by this reaction and the reaction could be used to treat rheumatoid arthritis in so-called boron neutron capture synovectomy (reviewed in Hawthorne, 1998). Moreover, one can look at the BNC reaction as a useful tool in understanding the mechanisms of high LET α-particle radiation-induced DNA damage in tumoral and normal cells.

Fig. 1.

Chemical structure of the boron delivery agent BPA.

Fig. 2.

Induction of micronuclei in human melanoma cells incubated with BPA (0.48, 1.2 and 2.4 mM) and irradiated with thermal neutrons. (A) MN/BN. (B) %MNBN. Results are expressed as means ± SD from two independent experiments.

Fig. 3.

Distribution of the number of micronuclei per micronucleated binucleated human melanoma cell incubated with BPA (2.4 mM) and irradiated with thermal neutrons (average values of two independent experiments).

Fig. 4.

Decrease in the frequency (%) of binucleated human melanoma cells incubated with BPA (0.48, 1.2 and 2.4 mM) and irradiated with thermal neutrons. Results are expressed as means ± SD from two independent experiments.

Fig. 5.

Genotoxic and anti-proliferative effects of BPA without thermal neutron irradiation in different cell types. (A) Human melanoma cells (CBMN assay, six independent experiments). (B) V79 Chinese hamster cells (CA assay, six independent experiments). (C) Human lymphocytes (CBMN assay, four healthy donors). Results are means ± SD. CAEG, chromosomal aberrant cells excluding gaps; BN, binucleated cells, MNBN, micronucleated binucleated cell; MI, mitotic index. *P < 0.05.

6

To whom correspondence should be addressed. Fax: [phone]; Email:

We warmly thank our colleague Prof. A.Laires for his many helpful suggestions, fruitful observations and pertinent criticisms. Our appreciation is extended to Dr A.Ferro de Carvalho for the ⁶⁰Co γ-irradiation as well as Dr A.Ramalho and Eng. F.Cardeira from the RPI. We also gratefully acknowledge our colleague O.Monteiro Gil for technical assistance.