DNA Oxidative Damage during Differentiation of HL-60 Human

Oct 18, 2001 - DNA oxidative damage was measured in human promyelocytic leukemia HL-60 cells, in the same cells committed to granulocytic differentiat...
0 downloads 0 Views 71KB Size
Download PDF
1492

Chem. Res. Toxicol. 2001, 14, 1492-1497

DNA Oxidative Damage during Differentiation of HL-60 Human Promyelocytic Leukemia Cells Valeria Covacci, Angela Torsello, Paola Palozza, Alessandro Sgambato, Giampiero Romano, Alma Boninsegna, Achille Cittadini, and Federica I. Wolf* Institute of General Pathology and Giovanni XXIII Cancer Research Center, Faculty of Medicine, Catholic University of Sacred Heart, Rome, Italy Received January 30, 2001

DNA oxidative damage was measured in human promyelocytic leukemia HL-60 cells, in the same cells committed to granulocytic differentiation with dimethyl sulfoxide (DMSO) or alltrans-retinoic acid (RA) and in mature human peripheral granulocytes (HPG). DNA damage was evaluated as single strand breaks and 8-OHdG adducts, measured by single cell electrophoresis or by monoclonal antibodies, respectively. The basal levels of either marker of DNA damage were higher in undifferentiated HL-60 cells than in HPG and DMSO- or RAdifferentiated cells. Treatment with H2O2 increased 8-OHdG formation in all cells, but the levels of DNA damage remained higher in undifferentiated cells as compared to the differentiated ones. Three lines of evidence suggested that the higher levels of DNA damage observed in undifferentiated cells were at least in part attributable to a reduced detoxification of reactive oxygen species (ROS). First, undifferentiated cells were shown to accumulate higher levels of dichlorodihydrofluorescein-detectable ROS than HPG and DMSO- or RA-differentiated cells. Second, undifferentiated HL-60 cells were characterized by reduced levels of GSH and lower GSH/GSSG ratios as compared to the differentiated cells. Third, pretreatment of undifferentiated HL-60 cells with antioxidants such as R-tocopherol or β-carotene suppressed the elevation of ROS and the formation of 8-OHdG induced by H2O2. Further evidence for the importance of the oxidant/antioxidant balance was obtained by modulating the iron-catalyzed decomposition of H2O2 to hydroxyl radicals in undifferentiated HL-60 cells. In fact, pretreatment with FeSO4 increased the formation of 8-OHdG induced by H2O2, whereas pretreatment with the iron chelator deferoxamine produced the opposite effect. These results illustrate correlations between the oxidant/antioxidant balance and DNA damage and suggest that the capability of a cell population to withstand oxidative stress and DNA damage may depend on its degree of differentiation.

Introduction The formation of reactive oxygen species (ROS)1 and the consequent induction of oxidative stress have been implicated as causative mechanisms in several pathologic processes (see refs 1-3 for reviews). Cells can counteract the damaging action of ROS by means of enzymatic or nonenzymatic scavengers such as superoxide dismutase (SOD), glutathione peroxidase (GSH-Px), catalase, and hydrophilic or lipophilic low molecular weight antioxidants. An imbalance between ROS formation and detoxification may result in oxidative deterioration of important cell constituents such as lipids, proteins, and nucleic acids (reviewed in ref 4). Oxidative damage to DNA is of paramount importance because it may be followed by cell death or mutation (5-7). The level of oxidatively modified DNA reflects multiple factors such as the severity of oxidative stress, the availability of protective mechanisms, and the efficiency * To whom correspondence should be addressed. Phone: +39 06 3016619. Fax: +39 06 3012753. E-mail: [email protected]. 1 Abbreviations: DAB, diaminobenzidine; H DCF-DA, dichlorodi2 hydrofluorescein-diacetate; DMSO, dimethyl sulfoxide; GSH-Px, glutathione peroxidase; HPG, human peripheral granulocytes; 8-OHdG, 8-hydroxydeoxy guanosine; RA, all-trans-retinoic acid; ROS, reactive oxygen species; SOD, superoxide dismutase; SSB, single strand breaks.

of repair mechanisms such as base excision repair, mismatch repair, nucleotide excision repair, etc. (8-10). Whereas the cellular mechanisms of ROS formation and detoxification have been characterized in detail, less is known about the cause-effect relationship between the induction of oxidative stress and the level of damage to DNA. A major problem in this setting involves the method chosen for evaluating DNA damage. Whereas alkaline elution and single cell electrophoresis are sensitive methods for detecting low levels of double (11) or single (12) strand breaks (SSB), respectively, 8-OHdG or MDA-DNA adducts are selective markers of oxidative damage to specific DNA residues (reviewed in refs 13 and 14). Among the latter, 8-OHdG is most used for two reasons: first, the levels of 8-OHdG have been shown to correlate with mutagenesis and carcinogenesis processes in which oxidative damage was involved as causative mechanism (15-17); second, there are several techniques for detecting 8-OHdG, such as HPLC-mass spectroscopy, gas chromatography, 32P-post-labeling, and monoclonal antibodies (18). As mentioned, oxidative modification of DNA may occur during mutagenesis or carcinogenesis (2, 3, 5, 7). In the present study, we addressed the possibility that the susceptibility to undergo DNA oxidative damage

10.1021/tx010021m CCC: $20.00 © 2001 American Chemical Society Published on Web 10/18/2001

DNA Oxidative Damage in Differentiation

might change also during the process of cell differentiation. To this aim, we measured basal and H2O2-induced DNA damage in HL-60 human promyelocytic leukemia cells, in the same cells committed to granulocytic differentiation by DMSO or retinoic acid, and in mature peripheral human granulocytes (HPG). DNA damage was assessed by both single cell electrophoresis and immunochemical labeling of 8-OHdG. We demonstrate that the process of differentiation is accompanied by an increased resistance to both basal and H2O2-induced DNA oxidation and that such resistance may reflect, at least in part, an improved detoxification of the oxidizing agents.

Experimental Procedures Chemicals. tert-butyl-hydroperoxide was from Sigma Italia, Milano; desferal (deferoxamine metansulfonate) was from CibaGeigy, Basel, CH. R-Tocopherol and β-carotene (Fluka, Milan, Italy) were dissolved in THF. The amount of THF delivered to the cells never exceeded 0.5% (v/v). All other chemicals and reagents were of the highest analytical grade. Cells and Treatment Protocols. Human promyelocytic HL60 cells were grown as suspension cultures in RPMI 1640 (Sigma Italia, Milano) supplemented with 10% foetal bovine serum (FBS) (Life Technologies Italia, Milano) at 37 °C in 5% CO2/air atmosphere. HL-60 cells were induced to differentiate to neutrophyl granulocytes by culturing the cells in a medium containing 1.3% DMSO (Sigma Italia, Milano) for 5 days or 1 µM all-trans-retinoic acid (RA) (Sigma Italia, Milano) for 3 days. The agents were subsequently removed and cell cultures utilized after 24 h. Differentiation was assessed by morphologic changes and by evaluating growth arrest. In all cases cells were utilized for experiments only when viability, assessed by the trypan blue exclusion test, was >95%. Experimental incubations were carried out at 37 °C in cultures of 1 × 106cells/mL in RPMI 1640 + 10% FBS, in the presence of appropriate stimuli or cofactors. At the end of incubations, cell viability was checked by trypan blue exclusion test and shown to be always >90%. Human peripheral neutrophyl granulocytes were collected from 10 mL of blood of healthy volunteer donors. Heparinized blood was diluted 1:1 (v:v) with saline and mixed with 4 mL of 6% (w:v) dextran (Sigma Italia, Milano) (MW 500.000). The mixture was sedimented at room temperature for about 40 min. The supernatants (∼8 mL), containing plasma and leucocytes, were stratified on 4 mL of lymphoprep (sodium-metrizoate and Ficoll d ) 1077 g/mL) (Nycomed, Oslo, Norway) and centrifuged at room temperature for 15 min at 700g. After removal of the supernatant, the pellets were hemolyzed with 500 µL of cold deionized water for 30 s, and subsequently diluted with 15 µL of saline. Pelleted granulocytes were counted and suspended at the desired concentration. Assays for 8-OHdG and Single Strand Breaks. Cytospin samples were prepared as it follows: cells were diluted in sucrose buffer (0.25 M sucrose, 1.8 mM CaCl2, 25 mM KCl, 50 mM Tris, pH 7.5) at a density of about 3.5 × 106 cells/mL. A total of 50 µL was added to carbowax-ethanol buffer (carbowax stock: 77 mL of PEG-1000 in 50 mL of water, 1 mL of stock in 74 mL of 70% ethanol) (Sigma Italia, Milano) and mixed. Aliquots of 150 µL were placed into cytospin funnels and centrifuged at 300 rpm for 5 min. Slides were coated with aminopropyl-triethoxysilane (Kindler, Freiburg, Germany). Samples were air-dried for 10-30 min, fixed in 95% cold ethanol (-20 °C) for 10 min, and stored at -20 °C. Detection of 8-OHdG by immunohistochemistry coupled with DAB (Vector, Burlingham) was carried out essentially as described by Yarborough et al. (19). 1F7 monoclonal antibody for 8-OHdG was kindly provided by Dr. R. M. Santella, Columbia School of Public Health, NY. Semiquantitative evaluation of the staining was carried out by an optical microscope (Zeiss, Germany, 400×) connected to an Optimas 5 morphometer

Chem. Res. Toxicol., Vol. 14, No. 11, 2001 1493 (Optimas Corporation). Nuclear staining was evaluated in approximately 100 cells of randomly chosen images by operators who were blind to the status of cell treatment, as recommended in ref 19. Data are reported as arbitrary units of optical density (AOD). Detection of SSB by single cell microgel electrophoresis was performed by the method of Singh (12) with minor modifications (20). Data are reported as the ratio between tail/nucleus areas evaluated by Optimas 5 morphometer. Assay for ROS. Intracellular ROS were detected in dichlorodihydrofluorescein-diacetate (H2DCF-DA)-loaded cells (Molecular Probe, Leiden, Netherlands), using a Cytofluor 2300/ 2350 (Millipore, Bedford, MA). Samples of 2 × 106 cells placed on Corning 6-wells plates were preincubated with 5 µM H2DCF -DA in PBS for 1 h at 37 °C. Plates were centrifuged at 1200 rpm for 10 min and the basal fluorescence of cells was read in the Cytofluor (excitation at 504 nm, emission at 526 nm). Cells were subsequently treated with H2O2, at the desired concentration, and the fluorescence was read after 15 min. Net values of H2O2-induced ROS formation was calculated after corrections for basal fluorescence (usually 70-80 arbitrary fluorescent units/A. F. U. in all cell populations). Assay for Glutathione. Total and oxidized glutathione were measured by HPLC with fluorimetric detection of glutathioneorthophthalaldehyde adduct according to ref 21. A total of 50 × 106 cells was homogenized in ice with HClO4 (1 M)-EDTA (2 mM) and centrifuged 10 min at 4 °C at 1.000g. The neutralized supernatants were extracted and analyzed by HPLC (PerkinElmer series 200) according to ref 21, using a 25 × 0.46 cm, 5 µm particle size LC-18 Supelcosil column (Supelco, Bellefonte, PA), equipped with a 2 × 0.46 cm 5 µm particle packing C18 Supelcosil precolumn. The mobile phase was 7.5% methanol/ 92.5% acetate buffer (0.15 M, pH 7.00). The flow rate was 1.5 mL/min. Peaks were detected fluorimetrically (excitation at 340 nm/emission at 420 nm). GSH and GSSG concentration in samples were derived comparing the derivative peaks area to a standard curve generated by derivatizing known amounts of GSH. 8-OHdG, GSH/GSSG, and ROS were assayed in parallel on the same cell preparation. Values are given as means ( SE, indicated by vertical bars; values without vertical bars have SE within the symbols. Statistical analyses were performed by unpaired Student’s t-test, and differences were considered significant when P < 0.05. Other details are given in the legends to figures and tables.

Results I. Oxidative DNA Damage in Undifferentiated and Differentiated HL-60 Cells. Figure 1 shows oxidative DNA damage measured as SSB or as 8-OHdG adducts in undifferentiated and DMSO-differentiated HL-60 cells. Panel A reports the extent of DNA oxidative damage evaluated on single cell by microgel electrophoresis; it can be seen that the basal level of DNA damage was 44% lower in DMSO-treated than in control HL-60 cells (0.5 vs 0.8 arbitrary units of tail/nucleus area). An acute oxidative stress, induced by 15 min incubation with 50-300 µM H2O2, increased the levels of DNA damage in both undifferentiated and DMSOdifferentiated cells; however, DNA damage remained higher in undifferentiated cells (see also Figure 1A). A similar pattern of basal or H2O2-induced DNA damage was observed by assaying for 8-OHdG (Figure 1B). To assess whether the level of DNA oxidative damage was influenced also by the cell proliferation rate, which is much higher in undifferentiated cells as compared with differentiated cells, HL-60 cells were incubated for 4 days in a medium in which FBS was lowered from 10% to 0.1%. Under such conditions of serum starvation and

1494

Chem. Res. Toxicol., Vol. 14, No. 11, 2001

Covacci et al.

Figure 2. Intracellular ROS levels in undifferentiated and DMSO-differentiated HL-60 cells upon treatment with H2O2. The two cell populations were loaded with 5 µM H2DCF-DA for 1 h, treated for 15 min with the indicated concentrations of H2O2, and then analyzed in a Cytofluor apparatus. Intracellular ROS, expressed as arbitrary fluorescence units (A. F. U.), are means ( SE of 10 separate experiments. See Materials and Methods for technical details. (*) P < 0.005 vs DMSO differentiated cells.

Figure 1. Basal and H2O2-induced DNA damage in undifferentiated and DMSO-differentiated HL-60 cells. (A) DNA oxidative damage as SSB, evaluated by single cell electrophoresis, was assayed in the two cell populations before and after 15 min treatment with the indicated concentrations of H2O2. Data are expressed as the ratio between the tail and nucleus area and represent means ( SE of three separate experiments. (B) DNA oxidative damage was evaluated as 8-OHdG before and after 15 min treatments with the indicated concentrations of H2O2. Data are expressed as arbitrary optical density units (A. O. D.) and represent the means ( SE of nine separate experiments. (*) P < 0.01 and (**) P < 0.05 vs DMSO differentiated cells.

growth arrest the basal levels of immunodetectable 8-OHdG were similar to those detected in proliferating controls (432 ( 40 vs 447 ( 37 A. O. D., respectively). II. Intracellular ROS and DNA Damage. Intracellular levels of ROS were measured in control and DMSOtreated cells loaded with H2DCF-DA and then incubated with increasing concentrations of H2O2. As shown in Figure 2, the levels of ROS were significantly higher in undifferentiated than in DMSO-differentiated HL-60 cells. The very low yield of ROS in DMSO-differentiated cells was not attributable to an interference of DMSO with the methods chosen for inducing and detecting oxidative stress, as DMSO had been removed from the cell culture 24 h prior to H2O2 treatment. Moreover, DMSO-differentiated cells accumulated 30% less ROS than undifferentiated cells also when oxidative stress was induced by replacing H2O2 with 100 µM tert-butylhydroperoxide (not shown). Finally, we performed experiments in which HL-60 cells were differentiated with alltrans-retinoic acid (RA) in place of DMSO and subse-

quently exposed to H2O2. As shown in Figure 3A, RAdifferentiated cells responded to H2O2 by accumulating significantly less ROS than undifferentiated cells. Reduced accumulation of ROS, therefore, was a consistent feature of differentiated cells, regardless of the differentiating and oxidizing agents used in different experiments. While accumulating less ROS, RA-differentiated HL-60 cells also exhibited lower levels of 8-OHdG, both basally and after treatment with H2O2 (Figure 3B). To obtain further evidence that differentiated cells were less prone to DNA damage than undifferentiated cells, we measured basal and H2O2-induced levels of 8-OHdG in HPG and compared them to the corresponding levels in undifferentiated and DMSO- or RA-differentiated HL-60 cells. As shown in Table 1, undifferentiated HL-60 cells and HPG exhibited the highest or lowest levels of basal and H2O2-induced 8-OHdG, respectively, whereas DMSO- or RA-differentiated cells exhibited intermediate levels. This trend was confirmed by measuring single strand breaks (see also Table 1). Importantly, H2O2-treatment was accompanied by formation of cellular levels of ROS which reproduced those of DNA damage, i.e., undifferentiated HL-60 cells > RAor DMSO-differentiated cells > HPG. While anticipating possible relationships between ROS formation and DNA oxidation, these results suggested that differentiated cells might have been able to minimize DNA damage by scavenging ROS more efficiently than undifferentiated cells. To obtain information in this setting, undifferentiated and differentiated HL-60 cells were assayed for their content in glutathione, an essential component of ROS detoxification. As shown in Table 2, the levels of reduced glutathionesthat is the form available for either enzymatic or nonenzymatic detoxification of ROSswere indeed lower in undifferentiated HL-60 cells than in RAor DMSO-differentiated cells (1.35 ( 0.08 vs 1.47 ( 0.04 or 1.87 ( 0.11 nM/106cells, respectively; P < 0.05). The GSH/GSSG ratio similarly increased from undifferentiated to RA- or DMSO-differentiated HL-60 cells (see also Table 2), providing good correlation with the fact that undifferendiated cells accumulated more ROS than dif-

DNA Oxidative Damage in Differentiation

Chem. Res. Toxicol., Vol. 14, No. 11, 2001 1495 Table 2. Glutathione Content in Undifferentiated and Differentiated HL-60 Cells nmol/106 cells HL-60 + RA + DMSO

GSH

GSSG

GSH/GSSG

1.35 ( 0.08a 1.47 ( 0.04b 1.87 ( 0.11c

0.35 ( 0.04a 0.32 ( 0.04a 0.26 ( 0.02b

4.1 ( 0.5a 4.7 ( 0.6a 7.5 ( 0.8b

a Within a column, values not sharing the same letter were significantly different (P < 0.05).

Table 3. Effect of Antioxidants on Basal and H2O2-Induced ROS and 8-OHdG Content in HL-60 Cells ROSa HL-60 + R-tocopherol + β-carotene

8-OHdGb

+H2O2

basal

+H2O2 c

518 ( 88 384 ( 38 408 ( 46

399 ( 19 405 ( 15 387 ( 10

495 ( 46d 420 ( 18 370 ( 39

a A.F.U. b A.O.D. c 100 µM H O ; 50 µM R-tocopherol; 10 µM 2 2 β-carotene. d P < 0.05 vs basal 8-OHdG.

Figure 3. Effect of H2O2 treatment on intracellular ROS and 8-OHdG in HL-60 cells differentiated by retinoic acid. HL-60 cells were differentiated by treatment with 1 µM RA for 72 h and subsequently incubated for 15 min with the indicated concentration of H2O2. ROS (panel A) and 8-OHdG (panel B) were assayed at the end of treatment. Data of RA-treated cells are means ( SE of three separate experiments. (*) P < 0.05 and (**) P ) 0.0001 vs RA-differentiated cells. Table 1. Basal and H2O2-Induced DNA Oxidative Damage and ROS in HPG Compared to Undifferentiated and Differentiated HL-60 Cells 8-OHdGa basal HL-60 + RA + DMSO HPG

399 ( 19 305 ( 12 249 ( 26 164 ( 10

+ H 2O 2

SSBb d

basal

ROSc + H 2O 2

+ H2O2

495 ( 46f 0.88 ( 0.16 6.02 ( 0.50 518 ( 88 389 ( 5 nde nd 396 ( 5 323 ( 11 0.50 ( 0.10 3.60 ( 0.15 111 ( 1 288 ( 6 0.40 ( 0.09 3.60 ( 0.14 90 ( 5

a A. O. D. b Single strand breaks (tail/nucleus area). c Corrected for the basal level of ROS (80 ( 3.5 A. F. U.). d 100 µM H2O2. e Not detected. f P < 0.05 vs basal 8-OHdG.

ferentiated cells after exposure to H2O2 (cf. Table 1, last column). III. Effects of Antioxidants on DNA Damage. To further probe the role of ROS in 8-OHdG formation, we evaluated the effects of antioxidants such as R-tocopherol and β-carotene. As shown in Table 3, pretreatment of HL60 cells with either antioxidant clearly reduced ROS elevation induced by subsequent exposure to H2O2. Such an effect was accompanied by a complete suppression of H2O2-induced formation of 8-OHdG, providing further evidence for the relationship between ROS elevation and DNA damage.

Figure 4. Effect of FeSO4 on ROS and 8-OHdG levels in HL60 cells stimulated with H2O2. HL-60 cells were preincubated 15 min with 100 µM FeSO4 and subsequently treated for 15 min with increasing concentration of H2O2. (A) ROS (A. U. F.), (B) 8-OHdG (A. O. D.). Data are means ( SE of three separate experiments.

IV. Effects of Iron on DNA Damage. 8-OHdG is formed primarily by hydroxyl radicals attacking guanine (22). Because hydroxyl radicals are produced through an Fe(II)-catalyzed decomposition of H2O2, we characterized how HL-60 cells responded to manipulations of iron levels. As shown in Figure 4, 15 min preincubation of HL-60 cells with 100 µM FeSO4 modified the response pattern usually observed upon treatment with H2O2, decreasing the levels of H2DCF-detectable ROS (panel

1496

Chem. Res. Toxicol., Vol. 14, No. 11, 2001

Figure 5. Effect of desferal on ROS and 8-OHdG levels in HL60 cells stimulated with H2O2. HL-60 cells were preincubated for 1 h with 1 mM desferal and subsequently treated for 15 min with increasing concentration of H2O2. (A) ROS; (B) 8-OHdG levels. Data are mean ( SE of three separate experiments.

A) while increasing those of 8-OHdG (panel B). Such an apparent inconsistency between decreased ROS and increased 8-OHdG is explained by keeping in mind that H2O2 accounts for a major fraction of H2DCF-derived fluorescence (23). It is therefore conceivable that FeSO4 supplementation, while increasing the intracellular levels of Fe(II), favored decomposition of H2DCF-detectable H2O2 to hydroxyl radicals and consequently increased oxidative damage to DNA. In support to this interpretation, an opposite pattern of increased ROS detection vs decreased 8-OHdG formation was observed when H2O2 treatment was preceded by exposure to desferal, a chelator which diverts Fe(II) from decomposing H2O2 to hydroxyl radicals (Figure 5, panels A and B).

Discussion In this study, DNA oxidative damage in HL-60 cells was evaluated as DNA SSB or formation of 8-OHdG, measured by single cell electrophoresis or by immunohistochemistry coupled with DAB detection, respectively (cf. Figure 1, panels A and B). Both methods proved to be sensitive and gave comparable results. According to previous reports (15-17), the assay for 8-OHdG has the potential advantage of providing a quantitative correlation between oxidative damage to DNA and mutagenesis or carcinogenesis. The steady-state levels of 8-OHdG represent a balance between the formation of ROS, the availability of anti-

Covacci et al.

oxidants, and the efficiency of DNA repair mechanisms (4, 10). Here we have shown that the levels of 8-OHdG increase anytime HL-60 cells are exposed to H2O2, precursor to the hydroxyl radicals which form the adduct, or decrease anytime the cells are supplemented with antioxidants, like R-tocopherol or β-carotene, which improve their capability to withstand an oxidative stress (24) (Table 3). In all such conditions, the changes in 8-OHdG formation tend to correlate with consistent changes in the levels of H2DCF-detectable ROS, although potential pitfalls and limitations in H2DCF-based measurements recommend qualitative rather than quantitative assessments of such correlation (25). In addition, we have shown that the levels of 8-OHdG correlate also with the availability of Fe(II) for reaction with H2O2 and formation of hydroxyl radicals, as evidenced by the experiments in which cellular iron was manipulated through supplementation with FeSO4 or chelation with desferal (Figures 4 and 5). Taken as a whole, these results indicate that the extent of DNA oxidative damage in HL-60 cells is largely influenced by modifications of the oxidant-antioxidant balance. Measurements of 8-OHdG or SSB demonstrate that undifferentiated HL-60 cells have higher basal levels of DNA oxidative damage than the same cells differentiated with DMSO. As mentioned (cf. Results, section I), this feature may not be ascribed to the higher proliferation rate of undifferentiated vs differentiated cells, as serum starvation did not reduce the levels of detectable 8-OHdG in HL-60 cells. Differentiated cells displayed a significant decrease in the basal levels of DNA damage also when DMSO was replaced with RA as an in vitro differentiating agent, or when HL-60 cells were compared with HPG that represent the corresponding, naturally differentiated terminal cells (Table 1). Thus, a reduction in the basal levels of DNA oxidative damage seems to be a general characteristic of differentiated cells as opposed to undifferentiated counterparts. In comparison to undifferentiated HL-60 cells, both the naturally differentiated HPG and DMSO- or RA-differentiated cells are characterized also by an improved capability to detoxify ROS after H2O2 treatment (cf. Figure 1 and Table 1). This factor may help to explain how differentiated cells accumulate less 8-OHdG than undifferentiated cells also after an acute oxidative stress (cf. Figures 1 and 3, and Table 1). The observation that DMSO- or RA-differentiated HL-60 cells are characterized also by higher levels of GSH and GSH/ GSSG ratios is in good agreement with their improved ability in detoxifying ROS formation after treatment with H2O2 (cf. Tables 1 and 2). These lines of evidence suggest that, among other possible factors, a greater availability in antioxidant defenses may contribute to render differentiated cells less prone to the accumulation of 8-OHdG. There is evidence in the literature for a differential expression and activity of ROS-detoxifying enzymes during the process of differentiation; however, the results are not always consistent. After initial reports of higher catalase levels in HL-60 cells than in DMSO-differentiated cells, more recent studies have shown that catalase availability in its cytosolic form may be higher in the differentiated cells (26). A differentiation-related increase in catalase levels has been documented also in murine epidermal keratinocytes (27). In agreement with this latter report, other investigators have shown that DMSOinduced differentiation of myeloid leukemia cells is

DNA Oxidative Damage in Differentiation

accompanied by increased expression and activity of GSH-Px (28), whereas sodium butyrate-induced differentiation of K562 erythroleukemia cells is accompanied by elevation of Cu-Zn-SOD (29). Finally, in Caco-2 gastrointestinal cell lines, it has been reported that the complete pattern of the major antioxidant enzymes displays an increased activity and expression during differentiation (30). On balance, there seems to be growing evidence that antioxidant defenses increase during differentiation, a pattern consistent with our observations that differentiated cells are characterized by an improved detoxification of ROS and by a significant reduction of DNA oxidative damage. Altogether, the data presented confirm that, in comparison to differentiated cells, undifferentiated proliferating HL-60 cells exhibit higher basal levels of DNA oxidative damage, assessed by either single cell electrophoresis or 8-OHdG formation. This characteristic has important implications for the genetic instability of neoplastic cells, which is the basis of tumor progression (5, 7). Our data also show that the reduced levels of 8-OHdG in differentiated cells may at least in part reflect an improved capability of these cells to resist an oxidative challenge. Measuring DNA oxidative damage by assessing the levels of 8-OHdG, immunohistochemically or by other available techniques, provides a reliable and sensitive tool to characterize the susceptibility of a given cell population to undergo oxidative damage and may help to predict the risk of cell death, mutagenesis, or carcinogenesis.

Acknowledgment. Work co-financed by MURST 60% and MURST Project 9706247467. We are indebted to Dr. R. M. Santella from Columbia School of Public Health, New York, for providing us with monoclonal antibody 1F7 against 8-OHdG.

References (1) Ames, N. B., Shigenaga, M. K., and Hagen, T. M. (1993) Oxidants, antioxidants, and the degenerative diseases of aging. Proc. Natl. Acad. Sci. U.S.A. 90, 7915-7922. (2) Cerutti, P. A. (1994) Oxy-radicals and cancer. Lancet 344, 862863. (3) Dreher, D., and Junod, A. F. (1996) Role of oxygen radicals in cancer development. Eur. J. Cancer 32A, 30-38. (4) Freeman, B. A., and Crapo, J. D. (1982) Biology of diseases. Free radicals and tissue injury. Lab. Invest. 47, 421-426. (5) Feig, D. I., Reid, T. M., and Loeb, L. A. (1994) Reactive oxygen species in tumorigenesis. Cancer Res. 54, 1890s-1894s. (6) Johnson, T. M., Yu, Z. X., Ferrans, V. J., Lowenstein, R. A., and Finkel, T. (1996) Reactive oxygen species are downstream mediators of p53-dependent apoptosis. Proc. Natl. Acad. Sci. U.S.A. 93, 11848-11852. (7) Ottender, M., and Lutz, W. K. (1999) Correlation of DNA adduct levels with tumor incidence: carcinogenic potency of DNA adducts. Mutat. Res. 424, 237-247. (8) Lindahl, T., and Wood, R. D. (1999) Quality control by DNA repair. Science 286, 1897-1905. (9) De Boer, J., and Hoeijmakers, J. H. J. (2000) Nucleotide excision repair and human syndromes. Carcinogenesis 21, 453-460. (10) Klungland, A., Rosewell, I., Hollenbach, S., Larsen, E., Daly, G., Epe, B., Seebeg, E., Lindahl, T., and Barnes, D. (1999) Accumulation of premutagenic DNA lesions in mice defective in removal of oxidative base damage. Proc. Natl. Acad. Sci. U.S.A. 96, 1330013305.

Chem. Res. Toxicol., Vol. 14, No. 11, 2001 1497 (11) Kohn, K. W. (1991) Principles and practice of DNA filter elution. Pharmacol. Ther. 49, 55-77. (12) Singh, N. P., McCoy, M. T., Tice, R. R., and Schneider, E. L. (1988) A simple technique for quantitation of low levels of DNA damage in individual cells. Exp. Cell. Res. 175, 184-191. (13) Marnett, L. J. (2000) Oxyradicals and DNA damage. Carcinogenesis 21, 361-370. (14) Poirier, M. C., Santella, R. M., and Weston, A. (2000) Carcinogen macromolecular adducts and their measurement. Carcinogenesis 21, 353-359. (15) Musarrat, J., Areziana-Wilson, J., and Wani, A. A. (1996) Prognostic and aetiological relevance of 8-hydroxyguanosine in human brest carcinogenesis. Eur. J. Cancer 32A, 1209-1214. (16) Xu, A., Wu, L.-J, Santella, R. M., and Hei, T. K. (1999) Role of oxyradicals in mutagenicity and DNA damage induced by crocidolite asbestos in mammalian cells. Cancer Res. 59, 5922-5926. (17) Romano, G., Sgambato, A., Mancini, R., Capelli, G., Giovagnoli, M. R., Flamini, G., Boninsegna, A., Vecchione, A., and Cittadini, A. (2000) 8-Hydroxy-2′-deoxyguanosine in cervical cells: correlation with grade of displasia and human papillomavirus infection. Carcinogenesis 21, 1143-1147. (18) Collins, A., Cadet, J., Epe, B., and Gedik, C. (1997) Problems in the measurement of 8-oxoguanine in human DNA. Carcinogenesis 18, 1833-1836 (Report of a workshop, DNA oxidation, held in Aberdeen, U.K., January 19-21, 1997). (19) Yarborough, A., Zhang, Y. J., Hsu, T. M., and Santella, R. M. (1996) Immunoperoxidase detection of 8-Hydroxydeoxyguanosine in aflatoxin B1-treated rat liver and human oral mucosal cells. Cancer Res. 56, 683-688. (20) Guidarelli, A., Sestili, P., and Cantoni, O. (1994) Opposite effects of nitric oxide on DNA strand scission and toxicity caused by tertbutyl-hydroperoxide in U937 cells. Br. J. Pharmacol. 123, 13111316. (21) Neuschwander-Tetri, B. A., and Roll, F. J. (1989) Glutathione measurement by high-performance liquid chromatography separation and fluorimetric detection of the glutathione-orthophthalaldehyde adduct. Anal. Biochem. 179, 236-241. (22) Takeuchi, T., Nakajiama, M., and Morimoto, K. (1996) Relationship between the intracellular reactive oxygen species and the induction of oxidative DNA damage in human neutrophyl-like cells. Carcinogenesis 17, 1543-1548. (23) Carter, W. O., Narayanan, P. K., and Robinson, J. P. (1994) Intracellular hydrogen peroxide and superoxide anion detected in endothelial cells. J. Leukocyte Biol. 55, 253-258. (24) Di Mascio, P., Murphy, M. E., and Sies, H. (1991) Antioxidant defense systems: the role of carotenoids, tocopherols, and thiols. Am. J. Clin. Nutr. 53, 194S-200S. (25) Rota, C., Fann, Y. C., and Mason, R. P. (1999) Phenoxyl free radical formation during the oxidation of the fluorescent dye 2′,7′dichloroflourescein by horseradish peroxidase. Possible consequences for oxidative stress measurements. J. Biol. Chem. 274, 28161-28168. (26) Ballinger, C. A., Mendis-Handagama, S. M. L. C., Kalmar, J. R., Arnold, R. R., and Kinkade, J. M., Jr. (1994) Changes in the localization of catalase during differentiation of neutrophilic granulocytes. Blood 83, 2654-2668. (27) Reiners, J. J., Jr., Thai, G., Pavone, A., Rupp, T., and Kodari, E. (1990) Modulation of catalase activities in murine epidermal cells as a function of differentiation and exposure to 12-O-tetradecanoylphorbol-13-acetate. Carcinogenesis 11, 957-963. (28) Shen, Q., Chada, S., Whitney, C., and Newburger, P. (1994) Regulation of the human cellular glutathione peroxidase gene during in vitro myeloid and monocytic differentiation. Blood 84, 3902-3908. (29) Steinkuhler, C., Sapora, O., Carri, M. T., Nagel, W., Marcocci, L., Circolo, M. R., Weser, U., and Rotilio, G. (1991) Increase of Cu, Zn-superoxide dismutase activity during differentiation of human K562 cells involves activation by copper of a constantly expressed copper-deficient protein. J. Biol. Chem. 266, 2458024587. (30) Baker, S. S., and Baker, R. D., Jr. (1992) Antioxidant enzymes in the differentiated Caco-2 cell line. In Vitro Cell Dev. Biol. 28A, 643-647.

TX010021M