Sergei A.
Snopov
*a,
Frank R.
de Gruijl
b,
Len
Roza
c and
Jan C.
van der Leun
d
aInstitute of Cytology of the Russian Academy of Sciences, Tikhoretsky ave 4, 194064 St Petersburg, Russia. E-mail: snopov@mail.cytspb.rssi.ru
bSylvius Laboratories, Department of Dermatology, Leiden University Medical Centre, Wassenaarseweg 72, NL-2333 AL Leiden, The Netherlands. E-mail: F.R.de_Gruijl@lumc.nl.
cDepartment of Nutritional Physiology, TNO Nutrition and Food Research, Utrechtseweg 48, NL-3704 HE Zeist, The Netherlands. E-mail: Roza@voeding.tno.nl
dEcofys, Kanaalweg 16G, NL-3526 KL Utrecht, The Netherlands. E-mail: j.vanderleun@ecofys.nl
First published on 19th August 2003
Studies of UV-induced skin cancers show that malignisation of skin cells, as well as alterations in anti-tumor immune control, are triggered by UV-induced lesions in cellular DNA. Such lesions can probably appear in the human mononuclear leukocytes (lymphocytes) during exposure of skin to sunlight. With the aim of studying the processing of UV-induced DNA lesions in these cells, we used flow cytometry and labelling of their partially denatured nuclei with the monoclonal antibody (H3) that binds cyclobutane pyrimidine dimers in single-stranded DNA. After the first few hours of cultivation of the irradiated cells, we found an increase in H3-specific fluorescence from cellular nuclei, while there was a decrease in the number of H3-positive sites in isolated DNA from these cells. We examined cells cultured under different conditions and concluded that the effect of enhancement of H3 labelling of nuclei did not result from changes in temperature and culture medium. Furthermore, we have found that this effect, as well as the decrease in H3 labelling in isolated DNA, are both prevented by pretreatment of the cells with Novobiocin, which we used as an inhibitor for the topoisomerase II-induced relaxation of supercoiled DNA prior to repair-specific incision. The inhibition by Novobiocin of the above-mentioned changes in H3 labelling in cellular nuclei and isolated DNA of the irradiated cells clearly indicate the association of both effects with an excision repair-related DNA modification. While the partial loss of H3-binding sites from isolated DNA is obviously a result of excision of some fraction of pyrimidine dimers, the enhancement of the H3 labelling of nuclei might be due to the formation of open structures at dipyrimidine-containing DNA fragments in preparation for incision. We suggest that formation of open structures predominates quantitatively over dual incision and excision of these fragments, and leads to enhanced exposure of the pyrimidine dimers in nuclei to H3 binding. Thus, unstimulated human lymphocytes appear to be capable of performing pre-incision steps for removal of these DNA lesions.
It is assumed that sunlight can directly affect not only human skin cells, but also peripheral blood mononuclear leukocytes (PBMLs).13 Ninety-nine percent of PBMLs are resting, unstimulated cells, mainly lymphocytes. The possible cellular and immune function consequences of induction, repair and ‘unrepair’ of CPDs in these cells have not been studied extensively. Whereas, the difference in processing of DNA lesions between unstimulated and stimulated PBMLs is obvious; it has been shown that unstimulated lymphocytes have a decreased capacity, compared to skin fibroblasts, keratinocytes and stimulated lymphocytes, to repair CPDs by their nucleotide excision repair system,14,15 due to a defective excision of CPDs16 and low deoxyribonucleotide pool sizes.17 It remains, however, unclear whether the pre-excision enzyme treatment of CPDs is deficient or not in unstimulated lymphocytes.
With the aim of studying the processing of CPDs in the nuclei of unstimulated PBMLs, we tried to quantify the kinetics of CPD removal using anti-CPD and FITC-conjugated antibodies and flow cytometry. This technique has been shown to be sensitive for the detection of these lesions in the nuclei of PBMLs after doses as low as 1–2 J m−2 of UVC.18 Contrary to our expectations of observing a decrease in specific CPD fluorescence from the nuclei of UV-irradiated cells due to a repair removal of CPDs, in fact, we observed an increase in this fluorescence signal during the first four hours of cultivation. In our subsequent experiments, we examined whether the cell culture conditions could affect anti-CPD antibody binding to DNA. Our results led us to conclude that the enhancement of specific CDP antibody binding to the nuclei of unstimulated PBMLs occurred due to DNA modification at the non-excised CPD-containing fragments after the pre-incision enzyme activity.
Unirradiated cells showed a background autofluorescence, recorded in the green channels of the flow cytometer, which was independent of cultivation period [Fig. 1(a), lowest line]. In the UV-irradiated cells fixed directly after irradiation, the fluorescence level was clearly higher due to a specific signal from anti-CPD antibodies bound to the nuclei and showed an apparent dependency on UV dose [Fig. 1(a), three upper lines], confirming the sensitivity achieved in our previous study.18
Fig. 1 Intensities of anti-CPD (a, b) and anti-ssDNA (c, d) specific fluorescence from partially denatured nuclei of mononuclear leukocytes exposed to UVC radiation: (a, c) cells were irradiated in ice-cold PBS and then cultivated in supplemented RPMI medium at 37 °C; (b, d) cells were preincubated over a period of 1 h at room temperature in PBS with or without Novobiocin (0.175 mM), then irradiated at room temperature and cultivated in PBS at 37 °C. Doses of UVC and Novobiocin (NB) in the culture medium are shown alongside the corresponding lines. Abscissa: cultivation time/h; ordinate: scaled geometrical mean of the specific fluorescence (channel number average) from 104 nuclei measured with the flow cytometer. Bars show the standard deviation of the mean. |
In the first few hours of cultivation of the UV-irradiated cells, the intensity of CPD fluorescence became unexpectedly higher in 13 of 16 blood samples examined. Fig. 1(a) (upper three lines) shows the typical dynamic of the CPD signal after UV irradiation and cultivation of the cells. The CPD fluorescence increased in intensity up to 4 h of cultivation and then remained practically constant during the whole period of observation. Next, we attempted to ascertain the reason for this increase in the CPD signal via the following experiments.
First, we found that incubation of PBMLs at different temperatures (0, 20 and 37 °C) for 1 h before fixation led to considerable changes in ssDNA-fluorescence emission from the prepared cell nuclei; these changes were completely reversible after cultivation of the cells at 37 °C for 18 h (Fig. 2). This result indicates that the strandedness of the nuclear DNA of the fixed cells is dependent on the temperature conditions of the preceding cell cultivation. To avoid any possible influence of temperature changes on the DNA strands in the following experiments, we no longer cooled the cells down to 0 °C and kept them at room temperature (about 20 °C) from the moment of drawing the blood sample, before and during UV irradiation and cultivation.
Fig. 2 Histograms of anti-ssDNA-specific fluorescence from partially denatured nuclei of mononuclear leukocytes stored for 1 h at different temperatures (shown above the corresponding peaks) and then transferred and cultivated in supplemented RPMI medium at 37 °C for 0 (A) or 17 h (B). Abscissa: geometric mean of the specific ssDNA fluorescence; ordinate: relative cell count. |
Next, we found that the cultivation of PBMLs in the supplemented RPMI medium used here could also influence the DNA strands. This was manifested in changes in ssDNA fluorescence from the cellular nuclei of unirradiated cells. This fluorescence rose noticeably after 1 h cultivation and then gradually returned to the initial level [Fig. 1(c), lowest line]. In the UV-irradiated cells, the initial rapid rise in ssDNA fluorescence was followed by a further increase up to 21 h of observation. This second enhancement of the ssDNA signal appeared to be dependent on the UV dosage [Fig. 1(c), upper three lines].
To avoid any influence of culture medium exchange on the DNA strandedness in our further experiments we kept the cells in PBS during the entire cultivation period. Under these conditions, the unirradiated and UV-irradiated cells showed a somewhat lower increase in ssDNA fluorescence [Fig. 1(d), lines for 0 and 10 J m−2]. Despite this, the cultivation of UV-irradiated cells still led to an increase in CPD fluorescence [Fig. 1(b), top line]. The changes in specific ssDNA and CPD fluorescences in the nuclei of all the studied cell samples during cultivation reveal conformity to some extent, but their absolute means varied greatly between the different experiments and cell samples, which resulted in insignificant overall correlation between them.
The influence of the Novobiocin on the CPD labelling was already obvious at the 0 h cultivation time point, i.e. at the time of fixation of the UV-irradiated cells, which was about 18 min after the irradiation. (This period was required to collect the cells of all samples in centrifuge tubes and to sediment them and remove the supernatant). By this time, the CPD luminescence of DNA isolated from PBMLs pretreated with Novobiocin was discernibly higher than that for cells not treated with Novobiocin (Fig. 3, compare line 4 to line 3). This difference indicates that Novobiocin is indeed effective in inhibiting rapid excision of a considerable bulk of CPD that occurred during that short period following UV irradiation in PBMLs untreated with Novobiocin. After 6.5 h cultivation, the CPD luminiscence of DNA isolated from UV-irradiated cells not treated with Novobiocin decreased to a slight extent (Fig. 3, line 7) from the pre-cultivation level (Fig. 3, line 3), indicating that a further amount of CPDs was excised during the hours of cultivation. This latter excision seemed not to be greatly influenced by the pretreatment of cells with Novobiocin: the DNA isolated from these cells showed a similar slight decrease in CPD luminescence (Fig. 3, line 8) from its pre-cultivation level (Fig. 3, line 4).
Fig. 3 Photograph of the developed film exposed to CPD-specific luminescence from heat-denatured isolated DNA from UVC-irradiated mononuclear leukocytes. Cells were preincubated for a period of 1 h at room temperature in PBS, with or without 0.200 mM Novobiocin (NB), and then irradiated at room temperature. UVC and NB doses in the culture medium are shown next to the corresponding lines. The four upper lines correspond to DNA from irradiated cells fixed without cultivation (0 h, i.e. ∼18 min after UV irradiation), while the four lower lines were obtained for DNA from cells cultivated in PBS at 37 °C for 6.5 h. The amounts of DNA used to obtain the lines in columns 1, 2 and 3 were 900, 300 and 100 ng, respectively. |
In contrast to isolated DNA, the nuclei of the Novobiocin-pretreated PBMLs showed a lower CPD fluorescence at the moment of fixation after UV irradiation than the nuclei of cells that were not pretreated with Novobiocin [Fig. 1(b), 0 h]. Furthermore, after 6.5 h cultivation with Novobiocin, there was hardly any increase in CPD labelling of PBML nuclei, compared to a clear increase for PBMLs cultivated without Novobiocin [Fig. 1(b), 6.5 h]. Consequently, Novobiocin prevented the increase in CPD labelling of nuclei, despite the fact that DNA isolated from Novobiocin-pretreated cells contained significantly more anti-CPD antibody binding sites then their untreated counterparts.
Since the H3 antibody used here binds to CPD in ssDNA,20 we decided to check first whether the ssDNA increase could be due to the conditions employed to culture the cells. Using an antibody against ssDNA, we determined that the cultivation, in fact, resulted in an increase in ssDNA-specific fluorescence intensity [Fig. 1(c)]. Because modifications of chromatin structure and stress protein synthesis have been reported to be associated with changes in temperature from 22 to 37 °C and from 37 to 20 °C28,29 we checked whether DNA strandedness might be changed in our experiments as a result of the transfer of cells from cold PBS to a warm humidified supplemented RPMI. We did indeed detect changes in ssDNA fluorescence after storage of cells at 0, 20 and 37 °C; and these changes were reversible when the cells were cultivated at a constant temperature [Fig. 2(a) and (b)]. We speculated that these temperature-related changes in DNA strandedness occurring in the first 4 h of cell cultivation might contribute to the increase in the ssDNA signal for both unirradiated and UV-irradiated cells over this period. However, when we kept the cells at constant room temperature during irradiation and cultivation they still showed changes in DNA strandedness and increased CPD labelling of UV-irradiated cell nuclei.
Furthermore, our experiments showed that UV-irradiation definitely evoked specific formation of anti-ssDNA antibody binding sites, since the ssDNA fluorescence signal did not decline after the initial increase, as for the unirradiated cells, but continued to rise in a dose-dependent manner from 4 to 21 h of cultivation [Fig. 1(c)]. However, these late changes in DNA strandedness were not accompanied by any detectable increase in CPD labelling of cellular nuclei.
From these results, we conclude that two types of ssDNA increase take place in nuclei during cultivation of UV-irradiated cells. The first increase recorded during the first 4 h of cultivation is transient, independent of UV dose and attributable to changes in culture conditions. The second increase gradually develops over 21 h, is not reversible and is most likely caused by UV irradiation, since it clearly showed dependency on the UV dosage. During the first 4 h of cultivation, both types of ssDNA increase probably occur simultaneously. In spite of the fact that we were not able to determine a significant correlation between ssDNA- and CPD-specific fluorescence, we cannot rule out the possibility of such a correlation during the first 4 h. In any case, the observed UVC-dose dependent ssDNA rise in cellular nuclei indicated the specific character of the ongoing DNA modifications, which could be associated with enzymatic repair of DNA.
Direct confirmation of the ongoing excision repair in UV-irradiated cells was provided by further experiments using CPD labelling of isolated DNA. Using isolated DNA, we clearly determined an obvious decrease in CPD-containing sites after 6.5 h of cultivation (Fig. 3). This decrease is not particularly great, and it has been reported that even after 24 h cultivation of UV-irradiated unstimulated peripheral lymphocytes, the number of CPDs detected with a dimer-specific endonuclease remains as high as 85% of their initial count.14 However, unlike this study, that investigation reported no detectable removal of CPDs within 6 h after irradiation. Another study found evidence for incision–excision in this time period after UV irradiation in unstimulated blood leukocytes.30 In our experiments, removal of CPDs not only over 6.5 h, but also at shorter cultivation times has been observed. We conclude this from the experiments with PBMLs pretreated with Novobiocin in non-cytotoxic concentrations. Novobiocin inhibits topoisomerase II, which relaxes supercoiled DNA in the stages preceding repair-specific DNA incision.23–27 We have found that Novobiocin pretreatment leads to higher CPD labelling of DNA isolated from UV-irradiated PBMCs in a period from 18 min to 6.5 h after irradiation of cells (Fig. 3). This higher CPD labelling appears to be attributable to inhibition by Novobiocin of the CPD removal by excision repair in the cells. Detectable removal of CDPs in the first 18 min after irradiation in cells not treated with Novobiocin coincides well with immediate activation of nucleotide excision repair after DNA damage and substantial incision–excision actions within the first 30–60 min. The fact that, along with rapid excision of CPDs from isolated DNA, there was a considerable enhancement in the CPD-specific labelling of partially denatured nuclei [Fig. 1(b)] supports our assumption that this latter effect is associated with DNA modification in the early phase of repair, inhibited by Novobiocin.
It is to be noted that Novobiocin did not seem to inhibit the eventual removal of some CPD sites from isolated DNA after 6.5 h of cultivation. This removal was observed to a comparable extent with both Novobiocin-treated and untreated cells (Fig. 3). Hence, the later CPD removal was not inhibited by Novobiocin and, remarkably, was not accompanied by any detectable changes in CPD labelling of nuclei, despite the considerable increase in the ssDNA signal emitted by Novobiocin-treated cells during this period [Fig. 1(c) and (d)]. This fact is in favour of the suggestion that the increase in anti-CPD antibody binding of nuclei is due to a DNA modification which takes place in the initial phases of nucleotide excision repair.
It is assumed that for efficient DNA lesion detection by the global nucleotide excision repair system, the chromatin has to be relaxed. The UV-induced global chromatin relaxation is extended over the whole nucleus and this process requires the tumour-suppressor protein p53.31 Nucleotide excision repair of lesions in human cells begins with ATP-dependent formation of an open DNA structure of approximately 25 nucletides around the DNA adduct, which is brought about by the action of topoisomerase II relaxing the supercoiled DNA structure. This opening phase is followed by dual incision of that fragment executed by two endonucleases, XPG on the 3′ side of the lesion and ERCC1-XPF on the 5′ side, with 3′ cleavage occurring first.32,33 Unstimulated human lymphocytes have been reported to show defective global repair of CPDs due to an insufficiency in excision, rather than low deoxyribonucleotide pool sizes.16 Our findings suggest that unstimulated PBMLs perform substantially more pre-incision enzymatic opening of CPD-containing DNA fragments than the subsequent dual incision and excision of these fragments. Dominance of such DNA opening over complete excision of CPD leads to enhanced anti-CPD antibody binding to DNA in nuclear structures resistant to the pepsin–HCl denaturing treatment. The PBMLs appear to be capable of pre-incision preparation phases for the enzymatic repair of damaged DNA, while showing a diminished capacity to excise the damage.
This journal is © The Royal Society of Chemistry and Owner Societies 2004 |