Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Treatment of malignant glioma often includes radiation therapy and/or chemotherapy, usually consisting of a DNA-damaging agent. Despite poor prognosis, escalation of dose for either therapy is not feasible due to increased toxicity. Investigators have included halopyrimidines, such as 5bromo-2'-deoxyuridine (BrdUrd) or 5-iodo-2'-deoxyuridine, in their treatment protocols in an attempt to improve response to radiation therapy (Hoshino, 1974; Russo et al., 1984; Phuphanich et al., 1984; Greenberg et al., 1988). These nucleosides, as analogs of thymidine, are converted to the triphosphate level by cellular enzymes and are substrates for the replicating polymerases. Once incorporated into DNA, either BrdUrd or 5-iodo-2'-deoxyuridine can sensitize cells to gamma radiation (Hakala, 1959; Djordjevic and Szybalski, 1960; Erikson and Szybalski, 1961; Mohler and Elkind, 1963).

DNA-damaging agents BCNU and cisplatin are among the chemotherapeutic agents used to treat malignant glioma (Stewart et al., 1982; Lehane et al., 1983; Greenberg et al., 1984). However, these agents are already administered at maximally tolerated levels and significant dose escalation is not possible without myelosuppression, neurotoxicity, and pulmonary toxicity. Finding an agent to enhance the cytotoxicity induced by the DNA-damaging agents may improve response. In a previous study, increased sensitivity to mechlorethamine in cultured murine mast cells was observed when bromouracil was incorporated into DNA (Schindler et al., 1966). In a separate investigation using Chinese hamster cells, a 16% replacement of thymine in DNA by iodouracil resulted in a 1.5-fold enhancement of cytotoxicity to melphalan and cisplatin (Russo et al., 1986). The present study investigates the use of BrdUrd as a chemosensitizer in human glioma cells.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cells. Human glioma cell line D54 was obtained from Dr. Darell D. Bigner of Duke University. Cultures were fed twice weekly with RPMI 1640 medium supplemented with 10% fetal bovine serum (GIBCOBRL, Grand Island, NY) and grown at 37°C in a humidified incubator with 5% CO₂. Cultures were found to be free of mycoplasma contamination using a biological culture method.

Cytotoxicity Studies. After BrdUrd exposure, medium was removed and replaced with fresh medium containing BCNU or cisplatin and incubated for 2 h. Cells grown as monolayer were removed from plates using trypsin/EDTA and subsequently counted using a Coulter counter (Coulter Corp., Hialeah, FL). In studies with plateau-phase cultures, cell number was determined to estimate fractional confluency after BrdUrd treatment.

Quantitation of Bromouracil in DNA. At the end of the treatment period, cells were collected as above. Cell lysis and the isolation and enzymatic hydrolysis of DNA were conducted as described previously (Maybaum et al., 1987). Briefly, cells were lysed in an SDS buffer after which the lysate was treated with proteinase K (Sigma Chemical Co., St. Louis, MO), extracted twice with chloroform, and DNA was precipitated with ethanol. DNA was hydrolyzed to the nucleoside level after treatment with deoxyribonuclease I, snake venom phosphodiesterase, and alkaline phosphatase (enzymes from Sigma Chemical Co.). BrdUrd and thymidine (dThd) were converted to their respective base by thymidine phosphorylase. After the addition of chlorouracil (internal standard) the bases were extracted into ethyl acetate using saturated ammonium sulfate, derivatized with bis(trimethylsilyl)trifluoroacetamide, and quantitated using gas chromatography/mass spectrometry (GC/MS) with selected ion monitoring as reported previously (Stetson et al., 1986).

Determination of DNA Interstrand Cross-Links. Cells were labeled for three doublings with either 8-[¹⁴C]deoxyadenosine (8-[¹⁴C]dAdo) or methyl-[³H]dThd (internal standard). After removal of label, cells were incubated for 24 h with or without BrdUrd. Media containing BrdUrd was removed before a 2-h exposure to BCNU. Cells, after overnight incubation in drug-free media, were removed by gentle scraping and treated without or with irradiation on ice using AECL Theratron 80 (⁶⁰Co) irradiator. Alkaline elution was performed essentially as described previously (Kohn et al., 1976; Kohn, 1991). DNA binding proteins were removed by treatment with proteinase K. Elutions in duplicate were collected up to 10 h.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

BrdUrd Sensitization of BCNU Cytotoxicity. To determine whether incorporation of BrdUrd into DNA is a prerequisite for increased cytotoxicity, cultures were grown at varying cell density such that during exposure to BrdUrd they ranged from exponential to plateau growth. To minimize nonspecific effects of BrdUrd, after exposure, cultures were replaced with medium without BrdUrd before the addition of BCNU for 2 h with subsequent viability determination. As shown in Fig. 1, BCNU exerts a similar effect regardless of growth phase. The lowest fractional survival of cultures treated with BrdUrd followed by BCNU occurred during exponential growth, suggesting that incorporation of BrdUrd in DNA is required for enhanced cytotoxicity. Moreover, the level of enhanced lethality was mitigated as cultures approached plateau growth presumably due to less BrdUrd incorporation into DNA.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Cytotoxicity of BCNU alone and with BrdUrd administered during log and plateau growth. Cultures were seeded at varying cell density to ensure a differential phase of growth during a 24-h exposure to 14 µM BrdUrd. Then media (without BrdUrd) containing 5 µM BCNU was added and cultures were incubated 2 h before viability determination. Conditions are represented by average of triplicate cultures. Fr, fractional.

BrdUrd Uptake into DNA. To elucidate the relationship between cell density and extent of DNA incorporation, cultures of varying cell number were exposed to BrdUrd and the level of thymine replacement by bromouracil in DNA was determined by GC/MS analysis. Figure 2 depicts the rise of BrdUrd in DNA as the concentration in medium is increased in exponentially growing D54 cells after a 20-h exposure. Figure 3 demonstrates the decline in DNA incorporation due to cell density after a 24-h exposure of 40 µM BrdUrd. Low-level incorporation was also observed when BrdUrd was added to cultures at plateau growth.

View larger version (37K): [in this window] [in a new window]   Fig. 2.   BrdUrd uptake into DNA. Exponentially growing cultures were grown in the presence of varying concentrations of BrdUrd for 20 h before GC/MS analysis as described in Materials and Methods.

View larger version (68K): [in this window] [in a new window]   Fig. 3.   Effect of cell density on incorporation of BrdUrd into DNA. Level of thymine replacement by bromouracil in DNA was determined by GC/MS analysis after a 24-h exposure of 40 µM BrdUrd to cells at various levels of growth.

View larger version (38K): [in this window] [in a new window]   Fig. 4.   Modulation by FdUrd of BrdUrd DNA uptake during log and plateau growth. Cultures in log or plateau growth were exposed to 14 µM BrdUrd in the absence or presence of 5 nM FdUrd for 24 h before GC/MS analysis. Fr, fractional.

Extent of Chemosensitization of BCNU and Cisplatin by BrdUrd. The effect of a fixed level of BrdUrd in DNA on the cytotoxicity of BCNU and cisplatin was investigated. At a thymine replacement by bromouracil of 28.6 ± 0.064%, a 2-h exposure of 5 and 10 µM BCNU resulted in a respective 3.1- and 4.1-fold greater cell loss compared with identical cultures grown in the absence of BrdUrd (Fig. 5). For cisplatin, at 2 and 4 µM, the increased cell loss was 3.2- and 3.5-fold, respectively. Logit analysis of the alkylating agent alone or in combination with BrdUrd revealed a shift in IC₅₀ from 9.0 to 2.3 µM for BCNU and 4.6 to 1.1 µM for cisplatin (r² values 0.953, 0.919, 0.684, and 0.936, respectively).

View larger version (39K): [in this window] [in a new window]   Fig. 5.   BrdUrd sensitization of BCNU or cisplatin cytotoxicity in human glioma cells. Log-phase cells were grown in the absence or presence of 14 µM BrdUrd for 24 h before a 2-h treatment with BCNU or cisplatin and subsequent cell survival assessment. Conditions are represented by average of triplicate cultures. GC/MS analysis of DNA revealed a thymine replacement by bromouracil of 28.6% ± 0.064.

View larger version (68K): [in this window] [in a new window]   Fig. 6.   Effect of BCNU concentration on cytotoxicity enhancement at a fixed level of BrdUrd incorporation. Exponentially growing cells, previously exposed to either saline or 2.5 µM BrdUrd for 24 h, were treated with various concentrations of BCNU for 2 h before viability determination. Value for enhancement of cell lethality is described by ratio of plating efficiency of treated and untreated cells as described in Materials and Methods. Each condition is represented by average of triplicate cultures.

View larger version (75K): [in this window] [in a new window]   Fig. 7.   Relationship between BrdUrd incorporation in DNA and sensitization of BCNU. Data are from four separate experiments, each of which had different levels, as measured by GC/MS, of BrdUrd incorporated into DNA at time of a 2-h treatment with either 5 µM () or 10 µM () BCNU. Value for enhancement of cell lethality is described by ratio of plating efficiency of treated and untreated cells as described in Materials and Methods. Highest value for percentage of thymine replacement was taken from experiment represented in Fig. 5. Three additional experiments were performed in the same manner (not previously shown).

Effect of BrdUrd Substituted DNA on BCNUInduced DNA Cross-Links. Because it is widely held that bifunctional alkylating agents exert cytotoxicity by covalently linking complementary strands of DNA, it was anticipated that the observed enhanced cytotoxicity by the drug combination might be due to increased levels of cross-links. Alkaline elution (Kohn et al., 1976; Kohn, 1991) was used to assess the level of interstrand cross-links in DNA. However, as shown in Table 1, the cross-link index was not increased in cultures with BrdUrd-substituted DNA. Figures 8 and 9, respectively, depict a typical single elution experiment and a plot against control cells (internal standard) using the average of duplicates.

View larger version (76K): [in this window] [in a new window]   Fig. 8.   Assessment of DNA damage by alkaline elution. Exponentially growing cells were incubated for 3 days with [14C]dAdo to uniformly label DNA. Label was removed and cultures were grown in the absence or presence of 14 µM BrdU for 24 h followed by a 2-h treatment of 80 µM BCNU. Cultures were incubated overnight before alkaline elution. Untreated cultures grown in the presence of [3H]dThd were used as internal standard. Untreated [3H] () and treated 14C () cells were irradiated with 200 rad before loading onto filters. Cell lysis, proteinase K, and subsequent steps were performed as described previously (Kohn, 1991). Experimental conditions were: A, saline; B, BrdU; C, BCNU; and D, BrdU+BCNU. Ordinate, fraction retained; abscissa, fraction (2 h) eluted. Plots are from one elution only and are meant to illustrate typical profile. Complete data are represented in Table 1.

View larger version (93K): [in this window] [in a new window]   Fig. 9.   Elution plot against internal standard. In an attempt to minimize experimental variability, an internal standard was included with each elution. Alkaline elution [14C]DNA retained plotted against 3H internal standard DNA retained. The value of 14C retained at 0.5 3H internal standard was used to calculate cross-link factor (Table 1). Plots are average of duplicate elutions for each condition: A, saline; B, BrdU; C, BCNU; and D, BrdU+BCNU.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The DNA-damaging agents BCNU and cisplatin are used in the treatment of patients with high-grade glioma (Stewart et al., 1982; Lehane et al., 1983; Greenberg et al., 1984). Both agents have been used in combination with radiotherapy and administered systemically and via the internal carotid artery (Stewart et al., 1982; Lehane et al., 1983; Feun et al., 1983; Greenberg et al., 1984). Despite current use of these agents at maximum tolerated dose, tumor growth often continues. Agent combinations that augment the cytotoxic action of alkylating agents may extend treatment response. Previous reports in murine (Schindler et al., 1966) and Chinese hamster cells (Russo et al., 1986) have demonstrated increased sensitivity to alkylating agents by halogenated pyrimidines.

In the present investigation the enhancement of alkylating agent-induced cell lethality was equivalent to a 3- to 4-fold lower IC₅₀ for both BCNU and cisplatin in replicating D54 human glioma cells previously exposed to BrdUrd compared with cultures grown in the absence of BrdUrd (Fig. 5). Although the level of thymine replacement by bromouracil was quite high at 28.6% compared with the labeling indices reported (Hoshino et al., 1986) in patients with glioblastoma multiforme (range 2 to 26%, median 9.3%), the concentration of BCNU used (Fig. 5) is lower than clinically achievable (Hassenbusch et al., 1996). Similar enhancement may be achieved at lower levels of BrdUrd incorporation when cells are exposed to higher concentrations of alkylating agent (Fig. 7). The findings suggest the approach of combining an alkylating agent with a sensitizing agent, which requires replicating cells for active metabolite formation, may minimize toxicity to normal nondividing brain tissue while at the same time possibly providing greater cell loss to the growing areas of the tumor.

The mechanism of enhanced cell lethality induced by BrdUrd/BCNU treatment is not known. The cytotoxicity of BCNU is believed due primarily to alkylation of DNA, with the N⁷ and O⁶ position of guanine (G) being favored sites of adduct formation. These DNA lesions, if not repaired, result in altered DNA integrity with subsequent misincorporation and strand breaks as well as intra- and interstrand cross-links, the latter of which has been correlated with cytotoxicity (Colvin et al., 1976; Roberts and Thomson, 1979; Erickson et al., 1980).

The interstrand cross-link created after exposure of cells to BCNU is a covalent link between G and C of cDNA strands (Tong et al., 1982). Formation involves alkylation at position O⁶ of guanine, intramolecular rearrangement to form cyclization with N¹ of guanine with a subsequent reaction at N³ of cytosine on the opposite strand to form the cross-link (Tong et al., 1982). The O⁶ adduct can be repaired by the DNA repair enzyme, O⁶-alkylguanine-DNA alkyltransferase (AGT), which removes the alkyl moiety from O⁶-alkylguanine-DNA (Brent, 1985; Ludlum, 1990). It has been shown that AGT transfers the alkyl group onto itself, resulting in inactivation of the enzyme (Lindahl et al., 1988; Pegg, 1990).

We hypothesized that the enhanced cell kill with the BrdUrd/BCNU combination might be due to elevated levels of DNA interstrand cross-links. However, alkaline elution revealed no increase in cross-linking number in BrdUrd/BCNU-treated cultures compared with those treated with BCNU alone. This was unexpected because the level of cell lethality was increased with the drug combination. The explanation is not known but possibilities include limitations in the elution technique or alteration of preferred sites of alkylation in Brdurd-substituted DNA.

The limitations of alkaline elution have been reported when dealing with agents that cause strand breaks in DNA (Kohn, 1991). Neutral elution revealed no double-strand break formation (data not shown). However, both BrdUrd and BCNU cause single-strand breaks making it difficult to assess the levels of interstrand cross-links. Also, BrdUrd is a known radiosensitizer and additional strand breaks are formed by exposure to gamma radiation. In the present study a low radiation dose was used to minimize BrdUrd radiosensitization effects (Fig. 8). Strand-break contributions have been corrected as outlined previously (Kohn, 1991). The findings qualitatively reveal little change in the production of interstrand cross-links after a 24-h exposure to BrdUrd at 14 and 40 µM and that cross-link levels may be reduced in cells containing BrdUrd- substituted DNA (Table 1). These data suggest that BrdUrd in DNA may interfere (see below) with interstrand cross-link formation.

An alternative explanation is that synergy in strand break formation, which could not be corrected, occurred with BrdUrd/BCNU-treated cultures after radiation exposure. If greater than additive breaks occurred with the drug combination during radiation exposure, this might explain the observed reduction in interstrand cross-link number compared with BCNU alone.

Another possibility is that the lower level of interstrand cross-links in DNA with dThd replaced with BrdUrd may be due to local electronic effects increasing favorability of nucleophilic attack on guanine or bromouracil. For example, the N⁷ position may be slightly more reactive in those sequences containing BrdUrd. This would result in more DNA damage without necessarily increasing the level of interstrand cross-links. Electronic effects have been implicated to account for base sequence specific alkylation for N⁷ and O⁶ positions (Pullman and Pullman, 1981; Bubienko et al., 1983; Mattes et al., 1986; Briscoe et al., 1990). Any neighboring base that increases the electronegativity at N⁷, making the site more susceptible to nucleophilic attack, would be expected to contribute to the overall reactivity of this site. As neighboring bases to G, thymine has been shown to decrease adduct formation at N⁷ (Hartley et al., 1986; Briscoe and Duarte, 1988; Briscoe et al., 1990). Although it is not known what effect bromouracil would have under similar conditions, the electronegative bromine atom in place of the methyl group at position 5 of thymine may change the local electronic distribution, creating more favorable sites of alkylation on guanine in sequences with adjacent bromouracil.

The same reasoning can be applied to include bromouracil as a site of alkylation. In general, natural pyrimidines are less susceptible to alkylation than purines but it has been shown both in vitro (Dolan et al., 1984) and in vivo (Preston et al., 1986; O'Toole et al., 1993) that adduct formation at the O⁴ position of thymine can occur with certain carcinogens. Although addition of a methyl or ethyl group at O⁴-alkylthymine is very low compared with O⁶-alkylguanine, the former lesion is repaired less efficiently by mammalian AGT (Richardson et al., 1985; Pegg et al., 1985). Furthermore, alkylation at position O⁴ of thymine is not known to be involved with interstrand cross-links. Certain sites on bromouracil, including the O⁴ position, may be more susceptible to nucleophilic attack due to the electronegative bromine atom. The presence of such an adduct would be consistent with the elution results. The new lesion would damage DNA without increasing interstrand cross-links.

The findings indicate that DNA interstrand cross-linking is not increased when BrdUrd is in DNA and that an additional factor is probably involved in the sensitization of BCNU cytotoxicity. This further suggests that the level of O⁶-alkylguanine may not be the primary target of the BrdUrd-mediated effect and that the levels of AGT alone may not predict a sensitization response.