- Email: [email protected]
Differences in basal and induced DNA single-strand breaks between human peripheral monocytes and lymphocytes
Fundamental and Molecular Mechanisms of Mutagenesis
ELSEVIER
Mutation
Research 332
(I 995)
55-62
Differences in basal and induced DNA single-strand breaks between human peripheral monocytes and lymphocytes Olaf Holz *, Rudolf Jijrres, Anja Kiistner, Helgo Magnussen Krcrntmhuus
Grosshansdorj
Zentnrmjir
Pneumologie
und Thoraxchirurgie.
Grocshansdorf; Received 20
LVA-Freie
und Hansestadt
Hamburg.
D-22927
German,v
January 1995; revised 6 June 1995; accepted 5 July 1995
Abstract The aim of this study was to compare the susceptibility of peripheral monocytes and lymphocytes to oxidant-induced DNA single-strand breaks (SSB). DNA damage was assessed by the alkaline single-cell gel electrophoresis (SCGE) assay. Total peripheral mononuclear leukocytes (PML), PML enriched in lymphocytes and PML enriched in monocytes were used. The basal rate of SSB was measured after in vitro incubation of cells for I h in phosphate-buffered saline, and the induced rate after incubation in 10 /*M or 50 PM H,O,. Incubation was performed at 4°C to limit the possible influence of DNA repair. Lymphocyte-enriched PML were obtained after adhesion of the monocytes to tissue-culture treated plastic, and monocyte-enriched PML by removal of monocytes from the plastic through trypsin. In all samples, cell differentiation was performed using an immunofluorescence technique with antibodies against T- and B-lymphocytes and cytospin preparations. The rate of SSB was determined by visual scoring according to 6 predefined categories of DNA damage and was expressed as mean score (range O-500) per 100 cells. There was a linear relationship between the percentage of lymphocytes in the samples and the basal rate of SSB ( p < 0.001, slope 0.67 score units per o/o). The same was true for induced DNA damage after incubation in IO PM H202 (p < 0.001, slope 3.80 score units per 7rc) or 50 PM Hz02 ( p < 0.001. slope 3.22 score units per %I. These regression analyses revealed a 2.9-fold greater rate of basal DNA damage in lymphocytes compared to monocytes and an I I .3-fold greater rate for the damage induced by 10 PM H,Oz. We conclude that there are marked differences in the rate of basal and induced SSB between lymphocytes and monocytes, suggesting differences in antioxidant capacity between the two cell populations. These findings indicate that the assessment of SSB for biomonitoring and genotoxicity testing using PML has to take into account possible changes in cellular composition. Keyvordst
Comet assay; Antioxidant
defense: DNA
single-strand
break; Peripheral
1. Introduction The
(SCGE)
alkaline
single-cell
gel
assay has been developed
* Corresponding
author.
Tel.: (+49)
electrophoresis
to detect DNA
4102/601-O:
Fax: (+49)
4102/601-‘45.
0027-5 107/95/$09.50 SSDI 0027-5
0
1995 Elsevier
107(95)00154-9
Science B.V.
All rights reserved
mononuclear
leukocyte
single-strand breaks (SSB), including alkali-labile sites and open excision repair sites. This technique appears to be especially suited for samples with limited cell numbers (Hanley et al., 1993) and allows the determination of DNA damage in individual cells, in contrast to other methods of SSB measurement (Singh et al., 1988; McKelvey-Martin et al.,
56
0. Ho12 et (11./Mutation
1993). This has led to the observation that peripheral mononuclear leukocytes (PML) show large intercellular variability in their SSB response after in vitro exposure to different agents which are capable of inducing DNA damage, for example, by generating oxygen radicals (Singh et al., 1988; Anderson et al., 1994). Differences between monocytes and lymphocytes, as the major subtypes of PML, have been described for ‘unscheduled’ DNA synthesis (Knudsen et al., 1992) and for the formation of DNA adducts after in vitro exposure to benzo[ alpyrene (Holz et al., 199 1). To our knowledge, this has not been demonstrated for the induction of SSB. Vijayalaxmi et al. (1993) compared SSB formation in granulocytes and lymphocytes after radiation exposure, using the SCGE assay. These authors could not detect a significant difference in SSB between the two cell types. Hongslo et al. (1993) studied the effect of paracetamol on PML using the alkaline filter elution technique and reported that monocytes were less sensitive than lymphocytes to the enhanced induction of SSB by paracetamol after UV exposure. As paracetamol is suspected to increase the oxidative stress, by the production of reactive intermediates (Binkova et al., 1990), these data support the hypothesis that monocytes and lymphocytes differ in their susceptibility to oxidative DNA damage. When SSB are measured in mixed cell populations, differences in susceptibility between cell types may cause changes in the mean rate of SSB when cellular composition changes. This would interfere with the detection of possible effects of genotoxic agents in biomonitoring and genotoxicity testing. The aim of our study was (1) to assess whether monocytes and lymphocytes, isolated from peripheral blood, show different susceptibility to a well defined oxidant agent (H202); (2) to determine the magnitude of these differences; and (3) to evaluate their effect on the overall DNA damage as detected by the SCGE assay.
2. Materials and methods
2.1. Materials Throughout saline (buffer)
the study only phosphate-buffered free of calcium and magnesium was
Reseurch 332 C1995) 55-62
used. Ficoll paque was purchased from Pharmacia (Uppsala, Sweden), RPM1 1640 and fetal calf serum (FCS) from Gibco Life Technologies (Gaithersburg, MD, USA). Trypsin, 4,6-Diamidino-2-phenylindol (287 18-90-3) and the Wright stain kit were obtained from Sigma (St. Louis, MO, USA), low-melting and normal-melting agarose from FMC BioProducts (Rockland, USA), H,O, from Merck (Darmstadt, Germany, 7722-84- I), tissue culture flasks (Falcon) and the Simultest kit from Becton Dickinson (Heidelberg, Germany) and microscopic slides from Menzel (Braunschweig, Germany). Lysis buffer contained 2.5 M NaCl (Merck, 7647-14-51, 100 mM Na,-EDTA (Sigma, 638 l-92-61, 10 mM Tris (Merck. 77-86-l) and 1% Triton X (Merck, 9002-93-l) and was titrated to pH 10. Electrophoresis buffer contained 1 mM Na,-EDTA and 300 mM NaOH (Merck, 13 1O-73-2) and had pH > 13. 2.2. Subjects We studied four healthy, non-smoking volunteers on eight occasions. In two subjects (males: age 19 and 32 years), blood was drawn on three different days within a period of 4 weeks, and in two subjects (females: age 25 and 38 years) on 1 day. Subjects were asked to avoid alcohol, medication and intake of vitamin preparations 24 h before blood sampling. The study had been approved by the Ethics Committee of the Chamber of Physicians of the County Schleswig-Holstein and the subjects gave their informed consent to participate in the study. 2.3. Blood sampling and isolation of cells For PML isolation, we followed the method of Boyum (1977). After venous puncture, 20 ml blood was drawn into heparinized syringes and split into two aliquots to be carefully layered over Ficoll paque in sterile centrifuge tubes. Samples were centrifuged at 400 X g for 35 min. Monocytes and lymphocytes on top of the Ficoll paque layer were collected. The first aliquot was washed twice in cold buffer and the remaining cell pellet was resuspended in 1 ml of buffer. Cells were counted and their numbers were adjusted to 2.5 X IO6 cells/ml. Afterwards, cells were immediately mixed with low-melting agarose and pipetted onto the slides.
The second aliquot was washed twice in cold RPM1 and the pellet was suspended in RPM1 containing 25% FCS. This aliquot was used to obtain cell fractions enriched in lymphocytes or monocytes by adherence in a cell culture flask. We followed the separation procedure described by Nielsen ( 1987). with minor modifications. The aliquot was transferred to a tissue-culture flask, which had been pretreated for one hour at 37°C with FCS. After incubation at 37°C for I h. the non-adherent cells were gently decanted and washed once in buffer. The adherent cells were carefully rinsed 3 times with RPM1 (2.5% FCS) and once with buffer. Trypsin (5 ml) was added for IO min at 37°C. Cells were harvested from the flask and washed once in buffer. The two fractions were counted. adjusted to 2.5 X 1Oh cells/ml. mixed with low-melting agarose and transferred to the slides. In some of the experiments. a third aliquot was taken to determine the possible effect of the medium used for cell enrichment and the I h time lag on SSB rates. The aliquot was washed twice in cold RPMI. the pellet was resuspended in I ml of RPMI. cells were counted. adjusted to 2.5 X lOh cells/ml. mixed with low-melting agarose and transferred onto the slides. Afterwards. slides were incubated for I h at 37°C in RPM1 containing 25% FCS. Small samples of all the aliquots and fractions. respectively. were taken just before transfer to the gels in order to determine the cellular composition. 2.4. Exposure to H,O, e1ei~tr~~phore.si.s~ssci~
cud alkalirle single cell gd
Single-strand breaks were measured according to the method of Singh et al. (1988) as modified by Gedik et al. (1991) (Holz et al., 1994). The frosted area of microscopic slides was covered with 80 ~1 of normal-melting agarose to enhance the attachment of subsequent layers. Isolated PML were diluted I : IO in I % low-melting agarose and 50 ~1 were placed onto the first agarose layer and immediately covered with a I8 X I8 mm No. I coverglas. After solidification and removal of the coverglas. another 40 ~1 of low-melting agarose were added and covered. Solidification was allowed again and the coverglas was removed. Incubation was performed in buffer, IO FM. or 50 PM H,O,. for I h at 4°C to inhibit DNA
repair. Afterwards, gels were shortly washed in cold buffer to remove the H,O, and then lysed for I h in cold (4°C) lysis buffer. DNA unwinding was achieved by incubation of the slides for 20 min in electrophoresis buffer at 4°C. Electrophoresis was conducted for 35 min at room temperature (25 V. 300 mA) in a horizontal electrophoresis tank. Subsequently. gels were neutralized with 0.5 M Tris-buffer (pH 7.5) and stained with 4 ~1 of 4.6-diamidino-2phenylindol. The handling of cells was performed under red light or in the dark. 2.5. Slide scoring To evaluate the rate of SSB. 300-500 cells in the center of the gel were evaluated at magnification 400 X (Labolux. filters BP 340-380 and LP 430. Leica, Germany). DNA damage was judged visually according to 6 predefined categories similar to those published by Anderson et al. (1994). A rank number ranging from 0 to 5 was assigned to each of the categories. The sum of rank numbers for 300-500 cells was taken and expressed as the mean rate of damage per 100 cells (score). Therefore. a score of 0 represented undamaged cells and a score of 500 the maximum level of damage. 2.6. Crll d#erentifrtiorl The percentages of lymphocytes and monocytes in the blood samples were determined by the Simultest kit. which allowed the simultaneous two colour enumeration of T- and B-lymphocytes by coupling tluorescin (FITC) and phycoerythrin (PE) labeled monoclonal antibodies to CD3 and CD19. respectively. Unstained (non-T/B) cells comprised monocytes. granulocytes and natural killer cells and these were categorized as the monocyte fraction for the purpose this study. In addition. cytospin preparations were made in order to compare the total numbers of lymphocytes and monocytes and to determine the number of granulocytes. Cytospin preparations were stained in an automated slide stainer using the Wright stain kit.
The susceptibility of monocytes and lymphocytes was derived from mixed samples enriched in one of
58
0.
Hob et al. /Mutation
these cell types. The observed overall score (S) comprised the rates of SSB from both cell types (called S, for monocytes and S, for lymphocytes), their respective relative contributions given by the percentages of monocytes (M) or lymphocytes (L). In the analysis, both percentages were assumed to add to 100%. After incubation in buffer, the observed overall score reflected basal scores, called B, for monocytes and B, for lymphocytes. Scores after incubation in H20Z comprised basal rates and in addition the effects of H,O,. These effects were characterized by the slopes of SSB rate vs. concentration (c) of H,Oz (called R, for monocytes and R, for lymphocytes). Slopes were considered as average slopes between 0 PM H,O, (buffer) and 10 PM H,O,. The scores observed after incubation in buffer were plotted against the percentage of lymphocytes and standard linear regression analysis was performed. B, was derived as the intercept (0% lymphocytes) and B, as the score estimated for 100% lymphocytes. A similar regression analysis was performed for the scores obtained after 10 PM H20,. The H,O1-dependent slopes for monocytes and lymphocytes could be obtained from this regression line and the regression line for buffer. The arguments are summarized in the following formulae: =M+L, =B,+c~R,andS,=B,+c~R,. = S, .(M/lOO) + S, .(L/lOO).
100 &I S
Therefore, the observed overall score (5) as a function of % lymphocytes (L) and concentration of H202 Cc) was S=[B,+C.R,]+(L/~OO)~(B,-B,) +c.(R,-R,)]. From the regression mated as follows: c = 0: c=
10:
L=O L= 100 L=O L= 100
lines, the parameters -+ + + +
S=B, S=B, S=B, S = B,
were esti-
+lO.R, +lO.R,
The different rates of basal DNA damage were expressed as the ratio BJB, and the different suscep-
Reseurch 332 11998155-62
tibility to HZO1 as the ratio of slopes RJR,. Mean values and standard deviations were calculated for all parameters, correlations were computed as Pearson’s linear correlation coefficients, and statistical comparisons performed by t-statistics. Statistical significance was assumed when p was less than 0.05.
3. Results 3.1. Percentages of monocytes and lymphocytes The percentages of T- and B-lymphocytes ranged between 56 and 85% in the total PML, 12 and 35% in the monocyte-enriched PML, and 74 and 90% in the lymphocyte-enriched PML. The numbers of lymphocytes obtained from the cytospin preparations were significantly correlated with the numbers assessed by the immunofluorescent method (r = 0.94. p < 0.001). Similarly, monocyte numbers from cytospin slides correlated with those of non-T/B cells from the immunofluorescent technique (r = 0.93, p < 0.001). The cytospin preparations revealed that the percentage of granulocytes was always below 5%. 3.2. Scores of DNA damage Incubation of PML in RPM1 containing FCS prior to their in vitro exposure did not result in significantly altered levels of SSB compared to the cells immediately exposed to buffer or HZOZ (first aliquot). Therefore, the time lag and medium used in the procedure of cell enrichment (second aliquot) did not cause distortions of basal or induced SSB scores. Mean ( + SD) scores of total PML after incubation in buffer,10 PM, or 50 PM HzO, were 84.4 k 2819. 3 18.5 f 25.2 and 384.6 + 26.2 units, respectively. The samples enriched in lymphocytes showed mean (k SD) scores of 86.6 f 21.6, 397.9 & 27.1 and 470.5 -t_8.2 score units after incubation in buffer, 10 PM or 50 PM H1O,, respectively. For the samples enriched in monocytes, corresponding mean (&SD> scores were 55.8 _+ 14.7, 160.9 f 37.1 and 265.9 & 51.7 units. There was a linear relationship between the percentages of lymphocytes in the samples and basal rates of SSB (p < O.OOl), the intercept being 35.7 score units and the slope 0.67 score units per %
59
0. Hok et al. /Mutation Research 332 CIYYSI 55-62 500
.
.
500
.
400
,,, 300
-
..
.
2 g ii
ii 200
-
100
. 0. 0 0
s.
. *..
.
f
40
0 80
60
LYnPHOCYTES
“““‘S...S
100
(X)
lymphocytes (Fig. 1). Similarly, there was a linear relationship after incubation in 10 ,uM H,O, ( p < 0.001) (Fig. 2). The intercept was 66.2 score units and the slope 3.80 score units per % lymphocytes which was significantly higher than the slope obtained for buffer ( p < 0.00 1). After incubation in 50 PM H,O,, the intercept was 182.4 score units and the slope 3.22 score units per % lymphocytes (p < 0.001) (Fig. 3). The slope obtained after 50 FM H,O, was significantly different from that obtained for buffer ( p < 0.001) and the slope observed after 10 PM HzOz ( p < 0.05). There were no significant differences between the four study subjects with 500
. 400
. .
w 300
.
ti
-0 .
-
0
~~.‘~~.“~“.-“~~~ 20
200
100
:
:
‘0.
Fig. 1. Overall scores of DNA damage versus percentage of lymphocytes in total PML samples and samples enriched in lymphocytes or monocytes. after incubation in phosphate-buffered saline (basal scores).
ii
-
. .
.
0.
.
. 200
.
.
=s
. . -a
.
400
-
w 300
.*..
SO/AM H,O,
PBS
l
a-.
.*
-
.
*
““-’ 20
40 LYMPHOCYTES
60
80
100
(%)
Fig. 3. Overall scores of DNA damage versus percentage of lymphocytes in total PML samples and samples enriched in lymphocytes or monocytes. after incubation in 50 PM H,Oz (induced scores).
respect to the slopes of the relationship percentage of lymphocytes. 3.3. Diflerential
susceptibility
of scores vs.
of cells
The fact that the slope after exposure to 50 PM H,O, was in the same range as the slope after 10 FM H,O, indicated a flattening of the concentration-response curve to H202. Therefore, the susceptibility to H,Oz was estimated from the values obtained for incubation in buffer and 10 PM H?O?. The standard linear regression analyses allowed to estimate the basal score as 35.7 units for monocytes and 102.7 units for lymphocytes. Basal damage in lymphocytes was therefore 2.9-fold higher than in monocytes. The total score after incubation in 10 PM H,O, was 66.2 units for monocytes and 446.2 units for lymphocytes. The increase of score per PM H,O, (slope) was estimated to be 3.05 for monocytes and 34.4 for lymphocytes. Therefore, the rate of H,O,-induced damage was 11.3-fold higher in lymphocytes compared to monocytes.
t
t
100
4. Discussion 0
u
20
40 LYtlPHOCYTES
60
80
100
tx)
Fig. 2. Overall scores of DNA damage versus percentage of lymphocytes in total PML samples and samples enriched in lymphocytes or monocytes. after incubation in 10 FM H202 (induced scores).
Our study demonstrated a close relationship between the cellular composition of PML samples expressed as percent lymphocytes and the overall rate of SSB. This was true for basal scores after in vitro incubation in buffer and for induced scores after
60
0. Hot
er al./Murarion
incubation in H?O?. The basal rate and the induced rates of SSB were significantly higher in lymphocytes than in monocytes. Heterogeneity of DNA damage between cells has been observed in several studies, which used the SCGE assay for the analysis of PML (Singh et al., 1988; Anderson et al., 1994). Our study demonstrated that lymphocytes showed higher basal rates of SSB than monocytes. To minimize the source of variation produced by repair mechanisms in our study, cells were maintained at 4°C as soon as they were transferred to the gel. DNA repair has been shown to be faster in monocytes than in lymphocytes (Knudsen et al., 1992). Therefore, we cannot exclude the possibility that part of the observed differences between cells in basal SSB could be due to different repair of lesions generated in vivo prior to isolation of cells or during the isolation and assay procedures before incubation in buffer at 4°C. These arguments are perhaps less important for the differential susceptibility of monocytes and lymphocytes against H,O,-induced SSB, because the oxidant-induced damage was much larger than basal rates and exposure was at 4°C. Therefore, the differences could be interpreted as differences in susceptibility to oxidants. These differences in susceptibility to H,O, could be due to differences in their permeability to H,O,, in radical scavenging capacity or other mechanisms resulting in a varying access of HZO, to the nucleus (Singh et al., 1988). which could be caused, for example, by differences in chromatin structure (Wlodek and Olive, 1992). H202 is able to enter the cell and to damage DNA within minutes (Cochrane, 1991). The mechanisms are not fully understood but it is likely that HZOZ reacts with endogenous superoxide anions and metal ions (Haber Weiss reaction) to produce highly reactive hydroxyl radicals. As ultimate damaging species, oxo-complexes of metal ions or products of lipid peroxidation have also been proposed (Epe et al., 1993). Hydroxyl radicals produce a specific ‘fingerprint’ of DNA damage by inducing similar amounts of SSB, apurinic sites and base modifications (Epe et al., 1993). The alkaline SCGE assay is able to detect SSB and also apurinic sites, because these are sensitive to alkali treatment (Eastman and Barry, 1992) and can be converted to SSB in electrophoresis buffer (pH > 13). Further-
Research
332 (IYY5)
55-62
more, apurinic sites potentially leading to SSB are produced by glycosylases during the repair of modified DNA bases (Friedberg, 1985). However, at 4°C. the activity of glycosytases is greatly reduced and probably only a minor part of SSB detected in our study could be attributed to these enzymes. This interpretation is consistent with the occurrence of increased rates of SSB after incubation of cells with HZ02 at 25°C (data not shown). Accordingly, increased rates of SSB at 37°C as compared to 0°C have been reported by Sandstrom (1991). though at higher H,Oz concentrations. The incubation temperature of 4°C which we have used not only lowered the activity of repair enzymes, but coutd also have influenced antioxidant enzymes. There are no data on the activity of superoxide dismutases and gtutathion peroxidases under these conditions. It has been shown that glutathion reductase displays only about IO-15% of its activity observed at 37°C (Beutler et al., 1977). In contrast. catalase exhibits only a minor dependence on temperature (Q,,,, 1.05 1.12) due to the low activation energy for the decomposition of H,O, (Aebi. 1983). Other antioxidant processes such as the scavenging of radicals or sequestration of metals probably show only minor dependence on temperature. In the cell enrichment procedure, cells were incubated in culture medium (RPM0 containing FCS. There was no change in SSB rates caused by the culture medium or the time lag introduced by the cell enrichment procedure. We did not determine the possible influence of trypsin on the enriched monocyte fraction. In fibroblasts it has been shown that trypsin could cause additional SSB (Singh et al., 199 1). However, in our study the monocyte-enriched samples, which were treated with trypsin, showed the lower rate of DNA damage compared to lymphocytes. The possible effect of trypsin on the rate of SSB would have caused underestimation of the difference in susceptibility. Using V79 lung fibroblast cells. Olive et at. (1992) demonstrated that methodological factors could, at least in part. increase the heterogeneity of rates of SSB between cells. The authors showed that, when gels were not rinsed after lysis, excessive salt concentrations hindered the expansion and movement of DNA during electrophoresis. In our study, we allowed for a sufficient unwinding period as described by Singh et al. (1988).
0. Hoi: et 01. /Mutation
We used the enriched fractions of lymphocytes and monocytes in addition to total PML in order to cover a wide range of fractional compositions of PML. Applying linear regression analysis of SSB scores on the percentage of lymphocytes, we were able to derive the susceptibility of pure lymphocyte and monocyte fractions. Using this approach, we aimed to reduce artifacts arising from prolonged time requirements which would have been associated with procedures capable to achieve a higher degree of ceil separation. Regression analysis showed linear reiationships between the percentages of lymphocytes and SSB scores for basal DNA damage as well as DNA damage in response to H,O,. These findings confirm the assumptions on which our analysis was based and support the validity of the ceil enrichment and quantification procedures. In contrast, there was no linear relationship between the concentration of H202 and SSB scores. This was demonstrated by the fact that the slope after 50 PM H202 was in the same range as the slope after 10 FM H20, indicating a flattening of the concentration-response curve. Therefore, the susceptibility to H20Z was derived from the data obtained for the IO PM concentration. Due to the fact that the concentration-response curve may already start to flatten below the concentration of IO PM H,O,, the ratio of lymphocyte to monocyte susceptibility estimated by us represents the average value for concentrations up to 10 FM H,O,. The ratio of DNA damage per /IM H20, could be even higher for lower concentrations of H?O,. The major subtypes of ceils found in PML are lymphocytes and monocytes. As the percentage of lymphocytes was de’termined by labeling T- and B-lymphocytes in our study, the fraction called monocytes, by definition, comprised monocytes, natural killer ceils as a subtype of lymphocytes and granuiocytes. It has been shown that granuiocytes do not differ in their rate of SSB from ceil fractions of PML mainly comprising lymphocytes (Vijayaiaxmi et al.. 1993). Natural killer ceils are a minor component of PML and there are no data suggesting that they differ in their SSB response from other iymphocytes. When quantifying the percentage of monocytes from cytospin preparations. ceil numbers were consistent with those determined for the fraction which was categorized as monocytes from the immunofluorescent labeling. Therefore, it is likely that
Research 332
f 199515542
61
our data on the differential susceptibility of iymphocytes and monocytes to oxidant-induced DNA damage are not severiy biased. Since the possible heterogeneity applies to the monocyte fraction, a bias would have led to an underestimation of the differences between monocytes and lymphocytes. PML are often used in biomonitoring or in studies which aim to determine the potential effects of genotoxic agents, and the SCGE assay could be suitable for this purpose if reproducibility is taken into account (e.g., Hoiz et al., 1994). Our data emphasize that cellular composition is important when mixed ceil populations such as PML are used. Given different rates of SSB for different types of ceils, varying overall levels of SSB could be produced as well as masked by changes in ceil composition, e.g., due to infections. This appears to be even more important when in vitro assays are used to determine the DNA damage induced by oxidants. Recently, direct iabeiing of ceils by coupling of magnetic beads has been proposed as an alternative procedure to handle mixtures of cells in the SCGE assay (Strauss et al., 1994). Possibly this method will be useful in biomonitoring to account for the differences between ceil types in induced DNA damage which our study has revealed. In conclusion, our data demonstrate different rates of SSB for lymphocytes and monocytes, under basal conditions as well as in response to H,O,. These findings are consistent with the hypothesis of intercellular differences in the susceptibility to oxidative our data indicate that, in agents. Furthermore, biomonitoring and genotoxicity testing using PML, the analysis of SSB has to account for possible changes in cellular composition.
Acknowledgements This study was supported by the Landesversicherungsanstait Freie und Hansestadt Hamburg. We would like to thank Dr. Nicoia Watson for her helpful comments on the manuscript.
References Aebi, HE. (1983) in: H.U. Bergmeyer (Ed.). Methods of Enzymatic Analysis. 3rd Edn., Vol. 3. Verlag Chemie, Deerfield Beach, FL. pp. 173-286.
62
0. Holy et al. /Mutation
Anderson, D., T.W. Yu, B.J. Phillips and P. Schmezer (1994) The effect of various antioxidants and other modifying agents on oxygen-radical-generated DNA damage in human lymphocytes in the comet assay, Mutation Res., 307, 261-271. Boyum, A. (1977) Separation of lymphocytes, lymphocyte subgroups and monocytes: A review, Lymphology, 10, 33-38. Beutler, E., K.G. Blume, J.C. Kaplan, G.W. Liihr, B. Ramot and W.N. Valentine (1977) International Committee for standardization in hematology: Recommended methods for red-cell enzyme analysis. Br. J. Hematol., 35, 331-340. Binkova, B., J. Topinka and R.J. Sram (1990) The effect of paracetamol on oxidative damage in human peripheral lymphocytes, Mutation Res., 244. 227-23 1. Cochrane, C.G. (1991) Cellular injury by oxidants. Am. J. Med., 91, 23%30s. Eastman. A. and M.A. Barry (1992) The origin of DNA breaks: A consequence of DNA damage, DNA repair, or apoptosis? Cancer Invest., 10, 229-40. Epe, B., M. Pflaum, M. I-Baring, 3. Hegler and H. Ridiger (1993) Use of repair endonucleases to characterize DNA damage induced by reactive oxygen species in cellular and cell-free systems, Toxicol. Lett., 67, 57-72. Friedberg, E.C. (1985) DNA Repair. W.H. Freeman and Co., New York. Gedik, CM., S.W.B. Ewen and A.R. Collins (1991) Single cell gel electrophoresis applied to the analysis of UV-C damage and its repair in human cells, Int. J. Radiat. Biol., 62, 313-320. Hanley, N.M., W.J. Kozumbo, D.L. Costa, R.R. Tice and M.C. Madden (1993) Induction of DNA single strand breaks in lung ceils by ozone exposure in vivo and in vitro; Abstract from the International Conference of the American Thoracic Society 1993, A670. Holz, 0.. T. Krause and H.W. Ridiger (1991) Differences in DNA adduct formation between monocytes and lymphocytes after in vitro incubation with benzo[a]pyrene, Carcinogenesis, 12, 2181-2183. Holz, 0.. R. JGrres, A. Kastner, T. Krause and H. Magnussen (1994) Reproducibility of basal and induced DNA single-strand breaks detected by the single cell gel electrophoresis assay in human peripheral mononuclear leukocytes, Int. Arch. Occup. Environ. Health, in press.
Research 332 (1995) 55-62 Hongslo, J.K., Brunborg, G., LL. Steffensen and J.A. Holme (1993) Paracetamol inhibits UV-induced DNA repair in resting human mononuclear blood cells in vitro, Mutagenesis, 8, 423-429. Knudsen, L.E., L.P. Ryder and K. Wassermann (1992) Induction of DNA repair synthesis in human monocytes/B-lymphocytes compared with T-lymphocytes after exposure to N-acetoxyN-acetylaminofluorene and dimethylsulfate in vitro, Carcinogenesis, 13, 1285-1287. McKelvey-Martin. V.J., M.H.L. Green, P. Schmezer, B.L. PoolZobel, M.P. DeMeo and A. Collins (1993) The single cell gelelectrophoresis assay (comet assay): A European review. Mutation Res., 288, 47-63. Nielsen, H. (1987) Isolation and functional activity of human blood monocytes after adherence to plastic surfaces: comparison of different detachment methods, Acta. Pathol. Microbial. Immunol. Stand. Sect. C, 95. 81-84. Olive, P.L., D. Wlodek. R.E. Durand and J.P. Banath (1992) Factors influencing DNA migration from individual cells subjected to gel electrophoresis, Exp. Cell Res., 198, 259-267. Sandstrom, B.E. (1991) Induction and rejoining of DNA singlestrand breaks in relation to cellular growth in human cells exposed to three hydroperoxides at 0°C and 37°C Free Rad. Res. Commun., 15, 79-89. Singh, N.P., M.T. McCoy, R.R. Tice and E.L. Schneider (1988) A simple technique for quantitation of low levels of DNA damage in individual cells, Exp. Cell Res., 175, 184 -191. Singh, N.P., R.R. Tice. R.E. Stephens and E.L. Schneider (1991) A microgel electrophoresis technique for the direct quantitation of DNA damage and repair in individual fibroblasts cultured on microscope slides, Mutation Res., 252, 289-296. Strauss, G.H.S., W.P. Peters and R.B. Everson (1994) Measuring DNA damage in individual cells of heterogeneous mixtures: a novel application of an immunological typing technique, Mutation Res., 304, 211-216. Vijayalaxmi, G.H.S. Strauss and R.R. Tice (1993) An analysis of y-ray-induced DNA damage in human blood leukocytes, lymphocytes and granulocytes, Mutation Res., 292, 123- 128. Wlodek, D. and P.L. Olive (1992) Neutral filter elution detects differences in chromatin organisation which can influence cellular radiosensitivity, Radiat. Res.. 132, 242-247.
ELSEVIER
Mutation
Research 332
(I 995)
55-62
Differences in basal and induced DNA single-strand breaks between human peripheral monocytes and lymphocytes Olaf Holz *, Rudolf Jijrres, Anja Kiistner, Helgo Magnussen Krcrntmhuus
Grosshansdorj
Zentnrmjir
Pneumologie
und Thoraxchirurgie.
Grocshansdorf; Received 20
LVA-Freie
und Hansestadt
Hamburg.
D-22927
German,v
January 1995; revised 6 June 1995; accepted 5 July 1995
Abstract The aim of this study was to compare the susceptibility of peripheral monocytes and lymphocytes to oxidant-induced DNA single-strand breaks (SSB). DNA damage was assessed by the alkaline single-cell gel electrophoresis (SCGE) assay. Total peripheral mononuclear leukocytes (PML), PML enriched in lymphocytes and PML enriched in monocytes were used. The basal rate of SSB was measured after in vitro incubation of cells for I h in phosphate-buffered saline, and the induced rate after incubation in 10 /*M or 50 PM H,O,. Incubation was performed at 4°C to limit the possible influence of DNA repair. Lymphocyte-enriched PML were obtained after adhesion of the monocytes to tissue-culture treated plastic, and monocyte-enriched PML by removal of monocytes from the plastic through trypsin. In all samples, cell differentiation was performed using an immunofluorescence technique with antibodies against T- and B-lymphocytes and cytospin preparations. The rate of SSB was determined by visual scoring according to 6 predefined categories of DNA damage and was expressed as mean score (range O-500) per 100 cells. There was a linear relationship between the percentage of lymphocytes in the samples and the basal rate of SSB ( p < 0.001, slope 0.67 score units per o/o). The same was true for induced DNA damage after incubation in IO PM H202 (p < 0.001, slope 3.80 score units per 7rc) or 50 PM Hz02 ( p < 0.001. slope 3.22 score units per %I. These regression analyses revealed a 2.9-fold greater rate of basal DNA damage in lymphocytes compared to monocytes and an I I .3-fold greater rate for the damage induced by 10 PM H,Oz. We conclude that there are marked differences in the rate of basal and induced SSB between lymphocytes and monocytes, suggesting differences in antioxidant capacity between the two cell populations. These findings indicate that the assessment of SSB for biomonitoring and genotoxicity testing using PML has to take into account possible changes in cellular composition. Keyvordst
Comet assay; Antioxidant
defense: DNA
single-strand
break; Peripheral
1. Introduction The
(SCGE)
alkaline
single-cell
gel
assay has been developed
* Corresponding
author.
Tel.: (+49)
electrophoresis
to detect DNA
4102/601-O:
Fax: (+49)
4102/601-‘45.
0027-5 107/95/$09.50 SSDI 0027-5
0
1995 Elsevier
107(95)00154-9
Science B.V.
All rights reserved
mononuclear
leukocyte
single-strand breaks (SSB), including alkali-labile sites and open excision repair sites. This technique appears to be especially suited for samples with limited cell numbers (Hanley et al., 1993) and allows the determination of DNA damage in individual cells, in contrast to other methods of SSB measurement (Singh et al., 1988; McKelvey-Martin et al.,
56
0. Ho12 et (11./Mutation
1993). This has led to the observation that peripheral mononuclear leukocytes (PML) show large intercellular variability in their SSB response after in vitro exposure to different agents which are capable of inducing DNA damage, for example, by generating oxygen radicals (Singh et al., 1988; Anderson et al., 1994). Differences between monocytes and lymphocytes, as the major subtypes of PML, have been described for ‘unscheduled’ DNA synthesis (Knudsen et al., 1992) and for the formation of DNA adducts after in vitro exposure to benzo[ alpyrene (Holz et al., 199 1). To our knowledge, this has not been demonstrated for the induction of SSB. Vijayalaxmi et al. (1993) compared SSB formation in granulocytes and lymphocytes after radiation exposure, using the SCGE assay. These authors could not detect a significant difference in SSB between the two cell types. Hongslo et al. (1993) studied the effect of paracetamol on PML using the alkaline filter elution technique and reported that monocytes were less sensitive than lymphocytes to the enhanced induction of SSB by paracetamol after UV exposure. As paracetamol is suspected to increase the oxidative stress, by the production of reactive intermediates (Binkova et al., 1990), these data support the hypothesis that monocytes and lymphocytes differ in their susceptibility to oxidative DNA damage. When SSB are measured in mixed cell populations, differences in susceptibility between cell types may cause changes in the mean rate of SSB when cellular composition changes. This would interfere with the detection of possible effects of genotoxic agents in biomonitoring and genotoxicity testing. The aim of our study was (1) to assess whether monocytes and lymphocytes, isolated from peripheral blood, show different susceptibility to a well defined oxidant agent (H202); (2) to determine the magnitude of these differences; and (3) to evaluate their effect on the overall DNA damage as detected by the SCGE assay.
2. Materials and methods
2.1. Materials Throughout saline (buffer)
the study only phosphate-buffered free of calcium and magnesium was
Reseurch 332 C1995) 55-62
used. Ficoll paque was purchased from Pharmacia (Uppsala, Sweden), RPM1 1640 and fetal calf serum (FCS) from Gibco Life Technologies (Gaithersburg, MD, USA). Trypsin, 4,6-Diamidino-2-phenylindol (287 18-90-3) and the Wright stain kit were obtained from Sigma (St. Louis, MO, USA), low-melting and normal-melting agarose from FMC BioProducts (Rockland, USA), H,O, from Merck (Darmstadt, Germany, 7722-84- I), tissue culture flasks (Falcon) and the Simultest kit from Becton Dickinson (Heidelberg, Germany) and microscopic slides from Menzel (Braunschweig, Germany). Lysis buffer contained 2.5 M NaCl (Merck, 7647-14-51, 100 mM Na,-EDTA (Sigma, 638 l-92-61, 10 mM Tris (Merck. 77-86-l) and 1% Triton X (Merck, 9002-93-l) and was titrated to pH 10. Electrophoresis buffer contained 1 mM Na,-EDTA and 300 mM NaOH (Merck, 13 1O-73-2) and had pH > 13. 2.2. Subjects We studied four healthy, non-smoking volunteers on eight occasions. In two subjects (males: age 19 and 32 years), blood was drawn on three different days within a period of 4 weeks, and in two subjects (females: age 25 and 38 years) on 1 day. Subjects were asked to avoid alcohol, medication and intake of vitamin preparations 24 h before blood sampling. The study had been approved by the Ethics Committee of the Chamber of Physicians of the County Schleswig-Holstein and the subjects gave their informed consent to participate in the study. 2.3. Blood sampling and isolation of cells For PML isolation, we followed the method of Boyum (1977). After venous puncture, 20 ml blood was drawn into heparinized syringes and split into two aliquots to be carefully layered over Ficoll paque in sterile centrifuge tubes. Samples were centrifuged at 400 X g for 35 min. Monocytes and lymphocytes on top of the Ficoll paque layer were collected. The first aliquot was washed twice in cold buffer and the remaining cell pellet was resuspended in 1 ml of buffer. Cells were counted and their numbers were adjusted to 2.5 X IO6 cells/ml. Afterwards, cells were immediately mixed with low-melting agarose and pipetted onto the slides.
The second aliquot was washed twice in cold RPM1 and the pellet was suspended in RPM1 containing 25% FCS. This aliquot was used to obtain cell fractions enriched in lymphocytes or monocytes by adherence in a cell culture flask. We followed the separation procedure described by Nielsen ( 1987). with minor modifications. The aliquot was transferred to a tissue-culture flask, which had been pretreated for one hour at 37°C with FCS. After incubation at 37°C for I h. the non-adherent cells were gently decanted and washed once in buffer. The adherent cells were carefully rinsed 3 times with RPM1 (2.5% FCS) and once with buffer. Trypsin (5 ml) was added for IO min at 37°C. Cells were harvested from the flask and washed once in buffer. The two fractions were counted. adjusted to 2.5 X 1Oh cells/ml. mixed with low-melting agarose and transferred to the slides. In some of the experiments. a third aliquot was taken to determine the possible effect of the medium used for cell enrichment and the I h time lag on SSB rates. The aliquot was washed twice in cold RPMI. the pellet was resuspended in I ml of RPMI. cells were counted. adjusted to 2.5 X lOh cells/ml. mixed with low-melting agarose and transferred onto the slides. Afterwards. slides were incubated for I h at 37°C in RPM1 containing 25% FCS. Small samples of all the aliquots and fractions. respectively. were taken just before transfer to the gels in order to determine the cellular composition. 2.4. Exposure to H,O, e1ei~tr~~phore.si.s~ssci~
cud alkalirle single cell gd
Single-strand breaks were measured according to the method of Singh et al. (1988) as modified by Gedik et al. (1991) (Holz et al., 1994). The frosted area of microscopic slides was covered with 80 ~1 of normal-melting agarose to enhance the attachment of subsequent layers. Isolated PML were diluted I : IO in I % low-melting agarose and 50 ~1 were placed onto the first agarose layer and immediately covered with a I8 X I8 mm No. I coverglas. After solidification and removal of the coverglas. another 40 ~1 of low-melting agarose were added and covered. Solidification was allowed again and the coverglas was removed. Incubation was performed in buffer, IO FM. or 50 PM H,O,. for I h at 4°C to inhibit DNA
repair. Afterwards, gels were shortly washed in cold buffer to remove the H,O, and then lysed for I h in cold (4°C) lysis buffer. DNA unwinding was achieved by incubation of the slides for 20 min in electrophoresis buffer at 4°C. Electrophoresis was conducted for 35 min at room temperature (25 V. 300 mA) in a horizontal electrophoresis tank. Subsequently. gels were neutralized with 0.5 M Tris-buffer (pH 7.5) and stained with 4 ~1 of 4.6-diamidino-2phenylindol. The handling of cells was performed under red light or in the dark. 2.5. Slide scoring To evaluate the rate of SSB. 300-500 cells in the center of the gel were evaluated at magnification 400 X (Labolux. filters BP 340-380 and LP 430. Leica, Germany). DNA damage was judged visually according to 6 predefined categories similar to those published by Anderson et al. (1994). A rank number ranging from 0 to 5 was assigned to each of the categories. The sum of rank numbers for 300-500 cells was taken and expressed as the mean rate of damage per 100 cells (score). Therefore. a score of 0 represented undamaged cells and a score of 500 the maximum level of damage. 2.6. Crll d#erentifrtiorl The percentages of lymphocytes and monocytes in the blood samples were determined by the Simultest kit. which allowed the simultaneous two colour enumeration of T- and B-lymphocytes by coupling tluorescin (FITC) and phycoerythrin (PE) labeled monoclonal antibodies to CD3 and CD19. respectively. Unstained (non-T/B) cells comprised monocytes. granulocytes and natural killer cells and these were categorized as the monocyte fraction for the purpose this study. In addition. cytospin preparations were made in order to compare the total numbers of lymphocytes and monocytes and to determine the number of granulocytes. Cytospin preparations were stained in an automated slide stainer using the Wright stain kit.
The susceptibility of monocytes and lymphocytes was derived from mixed samples enriched in one of
58
0.
Hob et al. /Mutation
these cell types. The observed overall score (S) comprised the rates of SSB from both cell types (called S, for monocytes and S, for lymphocytes), their respective relative contributions given by the percentages of monocytes (M) or lymphocytes (L). In the analysis, both percentages were assumed to add to 100%. After incubation in buffer, the observed overall score reflected basal scores, called B, for monocytes and B, for lymphocytes. Scores after incubation in H20Z comprised basal rates and in addition the effects of H,O,. These effects were characterized by the slopes of SSB rate vs. concentration (c) of H,Oz (called R, for monocytes and R, for lymphocytes). Slopes were considered as average slopes between 0 PM H,O, (buffer) and 10 PM H,O,. The scores observed after incubation in buffer were plotted against the percentage of lymphocytes and standard linear regression analysis was performed. B, was derived as the intercept (0% lymphocytes) and B, as the score estimated for 100% lymphocytes. A similar regression analysis was performed for the scores obtained after 10 PM H20,. The H,O1-dependent slopes for monocytes and lymphocytes could be obtained from this regression line and the regression line for buffer. The arguments are summarized in the following formulae: =M+L, =B,+c~R,andS,=B,+c~R,. = S, .(M/lOO) + S, .(L/lOO).
100 &I S
Therefore, the observed overall score (5) as a function of % lymphocytes (L) and concentration of H202 Cc) was S=[B,+C.R,]+(L/~OO)~(B,-B,) +c.(R,-R,)]. From the regression mated as follows: c = 0: c=
10:
L=O L= 100 L=O L= 100
lines, the parameters -+ + + +
S=B, S=B, S=B, S = B,
were esti-
+lO.R, +lO.R,
The different rates of basal DNA damage were expressed as the ratio BJB, and the different suscep-
Reseurch 332 11998155-62
tibility to HZO1 as the ratio of slopes RJR,. Mean values and standard deviations were calculated for all parameters, correlations were computed as Pearson’s linear correlation coefficients, and statistical comparisons performed by t-statistics. Statistical significance was assumed when p was less than 0.05.
3. Results 3.1. Percentages of monocytes and lymphocytes The percentages of T- and B-lymphocytes ranged between 56 and 85% in the total PML, 12 and 35% in the monocyte-enriched PML, and 74 and 90% in the lymphocyte-enriched PML. The numbers of lymphocytes obtained from the cytospin preparations were significantly correlated with the numbers assessed by the immunofluorescent method (r = 0.94. p < 0.001). Similarly, monocyte numbers from cytospin slides correlated with those of non-T/B cells from the immunofluorescent technique (r = 0.93, p < 0.001). The cytospin preparations revealed that the percentage of granulocytes was always below 5%. 3.2. Scores of DNA damage Incubation of PML in RPM1 containing FCS prior to their in vitro exposure did not result in significantly altered levels of SSB compared to the cells immediately exposed to buffer or HZOZ (first aliquot). Therefore, the time lag and medium used in the procedure of cell enrichment (second aliquot) did not cause distortions of basal or induced SSB scores. Mean ( + SD) scores of total PML after incubation in buffer,10 PM, or 50 PM HzO, were 84.4 k 2819. 3 18.5 f 25.2 and 384.6 + 26.2 units, respectively. The samples enriched in lymphocytes showed mean (k SD) scores of 86.6 f 21.6, 397.9 & 27.1 and 470.5 -t_8.2 score units after incubation in buffer, 10 PM or 50 PM H1O,, respectively. For the samples enriched in monocytes, corresponding mean (&SD> scores were 55.8 _+ 14.7, 160.9 f 37.1 and 265.9 & 51.7 units. There was a linear relationship between the percentages of lymphocytes in the samples and basal rates of SSB (p < O.OOl), the intercept being 35.7 score units and the slope 0.67 score units per %
59
0. Hok et al. /Mutation Research 332 CIYYSI 55-62 500
.
.
500
.
400
,,, 300
-
..
.
2 g ii
ii 200
-
100
. 0. 0 0
s.
. *..
.
f
40
0 80
60
LYnPHOCYTES
“““‘S...S
100
(X)
lymphocytes (Fig. 1). Similarly, there was a linear relationship after incubation in 10 ,uM H,O, ( p < 0.001) (Fig. 2). The intercept was 66.2 score units and the slope 3.80 score units per % lymphocytes which was significantly higher than the slope obtained for buffer ( p < 0.00 1). After incubation in 50 PM H,O,, the intercept was 182.4 score units and the slope 3.22 score units per % lymphocytes (p < 0.001) (Fig. 3). The slope obtained after 50 FM H,O, was significantly different from that obtained for buffer ( p < 0.001) and the slope observed after 10 PM HzOz ( p < 0.05). There were no significant differences between the four study subjects with 500
. 400
. .
w 300
.
ti
-0 .
-
0
~~.‘~~.“~“.-“~~~ 20
200
100
:
:
‘0.
Fig. 1. Overall scores of DNA damage versus percentage of lymphocytes in total PML samples and samples enriched in lymphocytes or monocytes. after incubation in phosphate-buffered saline (basal scores).
ii
-
. .
.
0.
.
. 200
.
.
=s
. . -a
.
400
-
w 300
.*..
SO/AM H,O,
PBS
l
a-.
.*
-
.
*
““-’ 20
40 LYMPHOCYTES
60
80
100
(%)
Fig. 3. Overall scores of DNA damage versus percentage of lymphocytes in total PML samples and samples enriched in lymphocytes or monocytes. after incubation in 50 PM H,Oz (induced scores).
respect to the slopes of the relationship percentage of lymphocytes. 3.3. Diflerential
susceptibility
of scores vs.
of cells
The fact that the slope after exposure to 50 PM H,O, was in the same range as the slope after 10 FM H,O, indicated a flattening of the concentration-response curve to H202. Therefore, the susceptibility to H,Oz was estimated from the values obtained for incubation in buffer and 10 PM H?O?. The standard linear regression analyses allowed to estimate the basal score as 35.7 units for monocytes and 102.7 units for lymphocytes. Basal damage in lymphocytes was therefore 2.9-fold higher than in monocytes. The total score after incubation in 10 PM H,O, was 66.2 units for monocytes and 446.2 units for lymphocytes. The increase of score per PM H,O, (slope) was estimated to be 3.05 for monocytes and 34.4 for lymphocytes. Therefore, the rate of H,O,-induced damage was 11.3-fold higher in lymphocytes compared to monocytes.
t
t
100
4. Discussion 0
u
20
40 LYtlPHOCYTES
60
80
100
tx)
Fig. 2. Overall scores of DNA damage versus percentage of lymphocytes in total PML samples and samples enriched in lymphocytes or monocytes. after incubation in 10 FM H202 (induced scores).
Our study demonstrated a close relationship between the cellular composition of PML samples expressed as percent lymphocytes and the overall rate of SSB. This was true for basal scores after in vitro incubation in buffer and for induced scores after
60
0. Hot
er al./Murarion
incubation in H?O?. The basal rate and the induced rates of SSB were significantly higher in lymphocytes than in monocytes. Heterogeneity of DNA damage between cells has been observed in several studies, which used the SCGE assay for the analysis of PML (Singh et al., 1988; Anderson et al., 1994). Our study demonstrated that lymphocytes showed higher basal rates of SSB than monocytes. To minimize the source of variation produced by repair mechanisms in our study, cells were maintained at 4°C as soon as they were transferred to the gel. DNA repair has been shown to be faster in monocytes than in lymphocytes (Knudsen et al., 1992). Therefore, we cannot exclude the possibility that part of the observed differences between cells in basal SSB could be due to different repair of lesions generated in vivo prior to isolation of cells or during the isolation and assay procedures before incubation in buffer at 4°C. These arguments are perhaps less important for the differential susceptibility of monocytes and lymphocytes against H,O,-induced SSB, because the oxidant-induced damage was much larger than basal rates and exposure was at 4°C. Therefore, the differences could be interpreted as differences in susceptibility to oxidants. These differences in susceptibility to H,O, could be due to differences in their permeability to H,O,, in radical scavenging capacity or other mechanisms resulting in a varying access of HZO, to the nucleus (Singh et al., 1988). which could be caused, for example, by differences in chromatin structure (Wlodek and Olive, 1992). H202 is able to enter the cell and to damage DNA within minutes (Cochrane, 1991). The mechanisms are not fully understood but it is likely that HZOZ reacts with endogenous superoxide anions and metal ions (Haber Weiss reaction) to produce highly reactive hydroxyl radicals. As ultimate damaging species, oxo-complexes of metal ions or products of lipid peroxidation have also been proposed (Epe et al., 1993). Hydroxyl radicals produce a specific ‘fingerprint’ of DNA damage by inducing similar amounts of SSB, apurinic sites and base modifications (Epe et al., 1993). The alkaline SCGE assay is able to detect SSB and also apurinic sites, because these are sensitive to alkali treatment (Eastman and Barry, 1992) and can be converted to SSB in electrophoresis buffer (pH > 13). Further-
Research
332 (IYY5)
55-62
more, apurinic sites potentially leading to SSB are produced by glycosylases during the repair of modified DNA bases (Friedberg, 1985). However, at 4°C. the activity of glycosytases is greatly reduced and probably only a minor part of SSB detected in our study could be attributed to these enzymes. This interpretation is consistent with the occurrence of increased rates of SSB after incubation of cells with HZ02 at 25°C (data not shown). Accordingly, increased rates of SSB at 37°C as compared to 0°C have been reported by Sandstrom (1991). though at higher H,Oz concentrations. The incubation temperature of 4°C which we have used not only lowered the activity of repair enzymes, but coutd also have influenced antioxidant enzymes. There are no data on the activity of superoxide dismutases and gtutathion peroxidases under these conditions. It has been shown that glutathion reductase displays only about IO-15% of its activity observed at 37°C (Beutler et al., 1977). In contrast. catalase exhibits only a minor dependence on temperature (Q,,,, 1.05 1.12) due to the low activation energy for the decomposition of H,O, (Aebi. 1983). Other antioxidant processes such as the scavenging of radicals or sequestration of metals probably show only minor dependence on temperature. In the cell enrichment procedure, cells were incubated in culture medium (RPM0 containing FCS. There was no change in SSB rates caused by the culture medium or the time lag introduced by the cell enrichment procedure. We did not determine the possible influence of trypsin on the enriched monocyte fraction. In fibroblasts it has been shown that trypsin could cause additional SSB (Singh et al., 199 1). However, in our study the monocyte-enriched samples, which were treated with trypsin, showed the lower rate of DNA damage compared to lymphocytes. The possible effect of trypsin on the rate of SSB would have caused underestimation of the difference in susceptibility. Using V79 lung fibroblast cells. Olive et at. (1992) demonstrated that methodological factors could, at least in part. increase the heterogeneity of rates of SSB between cells. The authors showed that, when gels were not rinsed after lysis, excessive salt concentrations hindered the expansion and movement of DNA during electrophoresis. In our study, we allowed for a sufficient unwinding period as described by Singh et al. (1988).
0. Hoi: et 01. /Mutation
We used the enriched fractions of lymphocytes and monocytes in addition to total PML in order to cover a wide range of fractional compositions of PML. Applying linear regression analysis of SSB scores on the percentage of lymphocytes, we were able to derive the susceptibility of pure lymphocyte and monocyte fractions. Using this approach, we aimed to reduce artifacts arising from prolonged time requirements which would have been associated with procedures capable to achieve a higher degree of ceil separation. Regression analysis showed linear reiationships between the percentages of lymphocytes and SSB scores for basal DNA damage as well as DNA damage in response to H,O,. These findings confirm the assumptions on which our analysis was based and support the validity of the ceil enrichment and quantification procedures. In contrast, there was no linear relationship between the concentration of H202 and SSB scores. This was demonstrated by the fact that the slope after 50 PM H202 was in the same range as the slope after 10 FM H20, indicating a flattening of the concentration-response curve. Therefore, the susceptibility to H20Z was derived from the data obtained for the IO PM concentration. Due to the fact that the concentration-response curve may already start to flatten below the concentration of IO PM H,O,, the ratio of lymphocyte to monocyte susceptibility estimated by us represents the average value for concentrations up to 10 FM H,O,. The ratio of DNA damage per /IM H20, could be even higher for lower concentrations of H?O,. The major subtypes of ceils found in PML are lymphocytes and monocytes. As the percentage of lymphocytes was de’termined by labeling T- and B-lymphocytes in our study, the fraction called monocytes, by definition, comprised monocytes, natural killer ceils as a subtype of lymphocytes and granuiocytes. It has been shown that granuiocytes do not differ in their rate of SSB from ceil fractions of PML mainly comprising lymphocytes (Vijayaiaxmi et al.. 1993). Natural killer ceils are a minor component of PML and there are no data suggesting that they differ in their SSB response from other iymphocytes. When quantifying the percentage of monocytes from cytospin preparations. ceil numbers were consistent with those determined for the fraction which was categorized as monocytes from the immunofluorescent labeling. Therefore, it is likely that
Research 332
f 199515542
61
our data on the differential susceptibility of iymphocytes and monocytes to oxidant-induced DNA damage are not severiy biased. Since the possible heterogeneity applies to the monocyte fraction, a bias would have led to an underestimation of the differences between monocytes and lymphocytes. PML are often used in biomonitoring or in studies which aim to determine the potential effects of genotoxic agents, and the SCGE assay could be suitable for this purpose if reproducibility is taken into account (e.g., Hoiz et al., 1994). Our data emphasize that cellular composition is important when mixed ceil populations such as PML are used. Given different rates of SSB for different types of ceils, varying overall levels of SSB could be produced as well as masked by changes in ceil composition, e.g., due to infections. This appears to be even more important when in vitro assays are used to determine the DNA damage induced by oxidants. Recently, direct iabeiing of ceils by coupling of magnetic beads has been proposed as an alternative procedure to handle mixtures of cells in the SCGE assay (Strauss et al., 1994). Possibly this method will be useful in biomonitoring to account for the differences between ceil types in induced DNA damage which our study has revealed. In conclusion, our data demonstrate different rates of SSB for lymphocytes and monocytes, under basal conditions as well as in response to H,O,. These findings are consistent with the hypothesis of intercellular differences in the susceptibility to oxidative our data indicate that, in agents. Furthermore, biomonitoring and genotoxicity testing using PML, the analysis of SSB has to account for possible changes in cellular composition.
Acknowledgements This study was supported by the Landesversicherungsanstait Freie und Hansestadt Hamburg. We would like to thank Dr. Nicoia Watson for her helpful comments on the manuscript.
References Aebi, HE. (1983) in: H.U. Bergmeyer (Ed.). Methods of Enzymatic Analysis. 3rd Edn., Vol. 3. Verlag Chemie, Deerfield Beach, FL. pp. 173-286.
62
0. Holy et al. /Mutation
Anderson, D., T.W. Yu, B.J. Phillips and P. Schmezer (1994) The effect of various antioxidants and other modifying agents on oxygen-radical-generated DNA damage in human lymphocytes in the comet assay, Mutation Res., 307, 261-271. Boyum, A. (1977) Separation of lymphocytes, lymphocyte subgroups and monocytes: A review, Lymphology, 10, 33-38. Beutler, E., K.G. Blume, J.C. Kaplan, G.W. Liihr, B. Ramot and W.N. Valentine (1977) International Committee for standardization in hematology: Recommended methods for red-cell enzyme analysis. Br. J. Hematol., 35, 331-340. Binkova, B., J. Topinka and R.J. Sram (1990) The effect of paracetamol on oxidative damage in human peripheral lymphocytes, Mutation Res., 244. 227-23 1. Cochrane, C.G. (1991) Cellular injury by oxidants. Am. J. Med., 91, 23%30s. Eastman. A. and M.A. Barry (1992) The origin of DNA breaks: A consequence of DNA damage, DNA repair, or apoptosis? Cancer Invest., 10, 229-40. Epe, B., M. Pflaum, M. I-Baring, 3. Hegler and H. Ridiger (1993) Use of repair endonucleases to characterize DNA damage induced by reactive oxygen species in cellular and cell-free systems, Toxicol. Lett., 67, 57-72. Friedberg, E.C. (1985) DNA Repair. W.H. Freeman and Co., New York. Gedik, CM., S.W.B. Ewen and A.R. Collins (1991) Single cell gel electrophoresis applied to the analysis of UV-C damage and its repair in human cells, Int. J. Radiat. Biol., 62, 313-320. Hanley, N.M., W.J. Kozumbo, D.L. Costa, R.R. Tice and M.C. Madden (1993) Induction of DNA single strand breaks in lung ceils by ozone exposure in vivo and in vitro; Abstract from the International Conference of the American Thoracic Society 1993, A670. Holz, 0.. T. Krause and H.W. Ridiger (1991) Differences in DNA adduct formation between monocytes and lymphocytes after in vitro incubation with benzo[a]pyrene, Carcinogenesis, 12, 2181-2183. Holz, 0.. R. JGrres, A. Kastner, T. Krause and H. Magnussen (1994) Reproducibility of basal and induced DNA single-strand breaks detected by the single cell gel electrophoresis assay in human peripheral mononuclear leukocytes, Int. Arch. Occup. Environ. Health, in press.
Research 332 (1995) 55-62 Hongslo, J.K., Brunborg, G., LL. Steffensen and J.A. Holme (1993) Paracetamol inhibits UV-induced DNA repair in resting human mononuclear blood cells in vitro, Mutagenesis, 8, 423-429. Knudsen, L.E., L.P. Ryder and K. Wassermann (1992) Induction of DNA repair synthesis in human monocytes/B-lymphocytes compared with T-lymphocytes after exposure to N-acetoxyN-acetylaminofluorene and dimethylsulfate in vitro, Carcinogenesis, 13, 1285-1287. McKelvey-Martin. V.J., M.H.L. Green, P. Schmezer, B.L. PoolZobel, M.P. DeMeo and A. Collins (1993) The single cell gelelectrophoresis assay (comet assay): A European review. Mutation Res., 288, 47-63. Nielsen, H. (1987) Isolation and functional activity of human blood monocytes after adherence to plastic surfaces: comparison of different detachment methods, Acta. Pathol. Microbial. Immunol. Stand. Sect. C, 95. 81-84. Olive, P.L., D. Wlodek. R.E. Durand and J.P. Banath (1992) Factors influencing DNA migration from individual cells subjected to gel electrophoresis, Exp. Cell Res., 198, 259-267. Sandstrom, B.E. (1991) Induction and rejoining of DNA singlestrand breaks in relation to cellular growth in human cells exposed to three hydroperoxides at 0°C and 37°C Free Rad. Res. Commun., 15, 79-89. Singh, N.P., M.T. McCoy, R.R. Tice and E.L. Schneider (1988) A simple technique for quantitation of low levels of DNA damage in individual cells, Exp. Cell Res., 175, 184 -191. Singh, N.P., R.R. Tice. R.E. Stephens and E.L. Schneider (1991) A microgel electrophoresis technique for the direct quantitation of DNA damage and repair in individual fibroblasts cultured on microscope slides, Mutation Res., 252, 289-296. Strauss, G.H.S., W.P. Peters and R.B. Everson (1994) Measuring DNA damage in individual cells of heterogeneous mixtures: a novel application of an immunological typing technique, Mutation Res., 304, 211-216. Vijayalaxmi, G.H.S. Strauss and R.R. Tice (1993) An analysis of y-ray-induced DNA damage in human blood leukocytes, lymphocytes and granulocytes, Mutation Res., 292, 123- 128. Wlodek, D. and P.L. Olive (1992) Neutral filter elution detects differences in chromatin organisation which can influence cellular radiosensitivity, Radiat. Res.. 132, 242-247.