Role of Nucleotide Excision Repair and p53 in Zidovudine (AZT)-Induced Centrosomal Deregulation

Abstract

The nucleoside reverse transcriptase inhibitor zidovudine (AZT) induces genotoxic damage that includes centrosomal amplification (CA>2 centrosomes/cell) and micronucleus (MN) formation. Here we explored these end points in mice deficient in DNA repair and tumor suppressor function to evaluate their effect on AZT-induced DNA damage. We used mesenchymal-derived fibroblasts cultured from C57BL/6J mice that were null and wild type (WT) for Xpa, and WT, haploinsufficient and null for p53 (6 different genotypes). Dose-responses for CA formation, in cells exposed to 0, 10, and 100 μM AZT for 24 hr, were observed in all genotypes except the Xpa^(+/+)p53^(+/−) cells, which had very low levels of CA, and the Xpa^(−/−)p53^(−/−) cells, which had very high levels of CA. For CA there was a significant three-way interaction between Xpa, p53, and AZT concentration, and Xpa^(−/−) cells had significantly higher levels of CA than Xpa^(+/+) cells, only for p53^(+/−) cells. In contrast, the MN and MN + chromosomes (MN + C) data showed a lack of AZT dose response. The Xpa^(−/−) cells, with p53^(+/+) or ^(+/−) genotypes, had levels of MN and MN + C higher than the corresponding Xpa^(+/+) cells. The data show that CA is a major event induced by exposure to AZT in these cells, and that there is a complicated relationship between AZT and CA formation with respect to gene dosage of Xpa and p53. The loss of both genes resulted in high levels of damage, and p53 haploinsufficicency strongly protected Xpa^(+/+) cells from AZT-induced CA damage.

INTRODUCTION

Nucleotide excision repair (NER) removes DNA damage caused by many chemicals and ultraviolet (UV) radiation [Hanawalt, 1995]. Individuals with defects in NER, who suffer from xeroderma pigmentosum (Xpa), are not able to repair sunlight-induced DNA damage, and develop skin cancer at a young age when exposed to sunlight without protection [Hanawalt, 1994]. The XPA protein, an essential early component of the NER pathway, is involved in damage recognition, and initiation and stabilization of the NER complex.

Considered to be a tumor-suppressor protein, p53 is known to arrest DNA replication, in the aftermath of genotoxic damage, to allow DNA repair to occur, and thus avoid replication of DNA on a damaged template. This protein is involved in many other cellular pathways, such as apoptosis, development, differentiation, transcription, DNA recombination, chromosomal segregation and senescence [Oren and Rotter, 1999], and has a proofreading activity in mitochondria [Bakhanashvili et al., 2009].

Zidovudine (AZT) was the first nucleoside reverse transcriptase inhibitor (NRTI) approved for HIV-1 therapy. It has been shown to act as a transplacental carcinogen in mice [Olivero et al., 1997], and becomes incorporated into DNA [Poirier et al., 2004]. In vitro and in vivo, AZT-DNA incorporation induces mutations, MN, chromosomal instability, telomere shortening, and centrosomal deregulation, which leads to changes in cell cycle regulation and gene expression [Olivero, 2007]. A predicted consequence of these events is genomic instability. It has been hypothesized that NRTI-induced DNA damage is repaired by the base excision repair pathway [Sampath et al., 2003]. However, recent reports also suggest the involvement of Xpc, a component of NER in the removal of high concentrations of AZT [Wu et al., 2013]. In addition, mice deficient in p53 are more susceptible to mutagenesis [Dobrovolsky et al., 2007] and tumorigenesis induced by exposure to AZT, compared with wild type (WT) mice [Olivero et al., 1997; NTP, 2012].

In these experiments we focused on the processing of AZT-induced DNA damage in mesenchymal fibroblasts grown from mice of the following six genotypes: Xpa^(+/+) p53^(+/+); Xpa^(+/+) p53^(+/−); Xpa^(+/+)p53^(−/−); Xpa^(−/−)p53^(+/+); Xpa^(−/−)p53^(+/−); and Xpa^(−/−)p53^(−/−). The genotoxic end points examined included CA, MN, and MN + C. Overall the data has shown a complicated picture of interactions between cells with differing Xpa and p53 capabilities and the processing of AZT-induced DNA damage.

MATERIALS AND METHODS

Animal Model

Transgenic Xpa^(−/−)p53^(+/−) C57BL/6J mice have previously been shown to be more susceptible to DNA damage [John et al., 2012] and cancer induction by benzo[a]pyrene, than their WT counterparts [Hoogervorst et al., 2003]. These mice (provided by H. van Steeg of the Laboratory of Toxicology, Bilthoven, The Netherlands) were originally generated on a C57BL/6 background [de Vries et al., 1995; van Steeg et al., 2001; Divi et al., 2002]. In our laboratory they were bred to obtain the genotypes of interest: Xpa^(−/−) p53^(+/+); Xpa^(−/−) p53^(+/−); and Xpa^(−/−) p53^(−/−). C57BL/6J Xpa^(+/+) p53^(+/−) were obtained from the Immune Regulation Section of the Experimental Immunology Branch, NCI. C57BL/6 mice, Xpa^(+/+) p53^(−/−), were obtained from the Jackson Laboratory (Bar Harbor, ME). Although all mice were generated using the same background the authors cannot confirm pure isogenicity.

Mouse offspring were paw tattooed and tail clipped 8–10 days following birth. DNA from the tail clips was used to genotype the mice, using published primer sequences for Xpa and p53 [de Vries et al., 1997; Divi et al., 2002]. Animals were bred at the NCI Animal Facility under conditions specified by the Association for the Assessment and Accreditation of Laboratory Animal Care in accordance with humane principles for laboratory animal care, and under a protocol approved by the NCI Animal Care and Use Committee.

Culture and Treatment of Mesenchymal Bone Marrow Cells

Bone marrow cells from two to four C57BL/6J mice/genotype were rinsed from the femur and tibia using a disposable LO-DOSE ¹/₂ cc U-100 Insulin Syringe with a permanently attached 28G ¹/₂ needle (Becton Dickinson, Franklin Lakes, NJ). The material was resuspended in RPMI media (ATCC, Manassas, VA) containing 10% fetal bovine serum (ATCC) and grown in T-75 Falcon flasks (BD Falcon, Bedford, MA) until they reached 95% confluency. To remove cells from flasks the monolayer was washed with Dulbecco’s phosphate buffered saline (DPBS, Gibco Invitrogen, Carlsbad, CA) without Ca and Mg, twice and incubated with 0.05% Trypsin (Gibco Invitrogen) for 4 minutes at 37°C. After neutralizing with an equal volume of media and centrifuging at 2,000 rpm for 4 minutes, cells were resuspended and distributed into 4-well culture slides (BD Falcon). Following a 4-hr attachment period, the media was replaced with media containing 0, 10, or 100 μM AZT (Sigma-Aldrich Co, St. Louis, MO) and cells were incubated for 24 hr. To prepare a stock solution, AZT was dissolved in DPBS and the final concentration was calculated using an extinction coefficient of 11,500 m²/mol at an absorbance wavelength of 266 nm.

Experimental Design for Centrosomal Amplification (CA) Analysis

To analyze CA, two to four mice were used per genotype, and from each mouse 2 to 10 slides were examined (Table I). The total numbers of cells scored is also shown in Table I. Since the cells were cultured on four chamber slides, control and treated cells could be grown and exposed on the same slide.

TABLE I.

Experimental Protocol Showing Number of Mice, Number of Slides, and Number of Cells Scored for the Genotoxicity Biomarkers, CA, MN, and MN + C

				Total number of scored cells
Endpoint	Genotype	Number of mice	Number of slidesa	Control	AZT 10 μM	AZT 100 μM
CA	XPA−/− P53+/−	3	7	1,464	1,447	1,304
	XPA−/− P53−/−	2	3–5	1,059	806	761
	XPA+/+ P53−/−	2	3	843	960	866
	XPA−/− P53+/+	3	5	1,226	1,148	1,249
	XPA+/+ P53+/+	4	9–10	2,092	2,118	2,091
	XPA+/+ P53+/−	2	2	1,030	1,056	1,042
MN and MN+C	XPA−/− P53+/−	1	2	1,060	1,139	1,075
	XPA−/− P53−/−	1	2	1,094	1,058	1,050
	XPA+/+ P53−/−	1	2	1,138	1,023	1,065
	XPA−/− P53+/+	1	2	1,125	1,158	1,124
	XPA+/+ P53+/+	1	2	1,031	1,167	1,093
	XPA+/+ P53+/−	1	2	1,049	1,113	1,057

^aVaries by concentration.

Experimental Design for Micronucleus (MN) and MN + Chromosomes (MN + C) Analysis

One mouse/genotype was used to evaluate the incidence of MN and MN + C. Two cell lines/genotype were grown on two separate slides, and controls and treated cells were scored on the same slide (Table I). There were 1,023–1,158 cells scored/genotype, and because MN + C is a subset of MN, both were scored using the same four chamber slide (Table I).

Cell Proliferation and Viability

Mesenchymal derived cells from mice of the following genotypes: Xpa^(+/+)p53^(+/+), Xpa^(+/+)p53^(+/−), Xpa^(+/+)p53^(−/−), Xpa^(−/−)p53^(+/+), and Xpa^(−/−)p53^(−/−) were used to evaluate proliferation and viability. Cells, obtained as described, were seeded in 12-well plates containing fresh media. One 12-well plate was used per genotype, and triplicate wells were used to examine each treatment and each genotype. At 24 hr after seeding, media was replaced with fresh media containing 0, 10, or 100 μM AZT. Cells were grown for an additional 24 hr, at which time the plates were trypzinized and cells counted twice in a Coulter Counter. The three control wells were averaged and considered 100%, and the three wells treated with 10 or 100 μM AZT were averaged and designated as percentage of the control. The termination of the mouse colony precluded evaluation of Xpa^(−/−)p53^(+/−) cells.

Immunofluorescence Staining of Centrosomes and Centromeres (Kinetochores)

Mesenchymal derived bone marrow fibroblasts, cultured in slide chambers and exposed as described above, were fixed with ice-cold methanol for 20 minutes at −20°C, washed two times for 5 minutes with 0.1% Tween-PBS and permeabilized for 4 minutes with 0.1% Triton X-100-PBS. Cells were washed three times with Tween-PBS and immersed in 1.25% bovine serum albumin (BSA) in DPBS blocking solution for 2 hr at 37°C in a humidifying chamber.

To identify the centrosomes and CA, a dilution of 1:500 rabbit anti-pericentrin antibody (Covance, Emeryville, CA) in blocking solution was used for 2 hr at 37°C in a humidifying chamber. After washing with Tween-PBS, the secondary antibody, anti-rabbit Alexa 488 (Invitrogen, Life Technologies, Grand Island, NY) was applied at a 1:750 dilution in blocking buffer for 45 minutes at 37°C. Using this approach the centrosomes were stained green.

To identify the kinetochores, chromosomal proteins at the site of microtubule attachment to the centromere, slides were incubated in a 1:1,000 dilution of human anti-CREST (calcinosis, Raynaud’s phenomenon, esophageal dysfunction, sclerodactyly, and telangiectasia) antibody-FITC conjugated (Antibodies Incorporated, Davis, CA) for 1.5 hr at 37°C in a humidifying chamber. The kinetochores appeared as light blue dots localized within the nucleus. Cells were stained with DAPI to identify nuclei and MN (blue).

Immunofluorescence Imaging Via Confocal Microscopy

Stained cells were visualized and scored using a Nikon Eclipse E-400 (Nikon, Inc, Melville, NY) microscope fitted with a Plan Apo 100× objective with a 1.40 numerical aperture. Cells were photographed on a Zeiss Axiovert 100M microscope equipped with a Zeiss Plan Apochromat 100×/1.4 oil Dichromic objective. Confocal images were generated using a Zeiss LSM 510 scanning laser microscope. The LSM 510 zoom software was used to produce a final magnification of 2,000×. Images shown are three-dimensional maximal projections generated from a series of images through the Z-plane. Approximately 1,000 cells per genotype were evaluated for the presence of CA and MN using fluorescent microscopy. Cells containing greater than two centrosomes were considered positive for CA. MN were blue (DAPI) circular fragments separated from the whole nucleus. MN stained positive for CREST antiserum contained whole chromosomes (MN + C), while the parent nucleus was considered to be aneuploid.

Statistical Methods

A three-factor factorial mixed model analysis of variance (ANOVA) was performed on the CA data using a binomial distribution (CA cells/total cells). The three factors considered were Xpa, p53, and AZT dose. Random effects were included to account for mouse and slide variation. For all ANOVAs, residuals were examined for homogeneity and were partitioned if found to be heterogeneous. All reported P-values are two-tailed, and Holm’s or Dunnett’s method was used to adjust pair-wise comparison P-values. In view of the large number of tests performed, we consider P < 0.005 to be statistically significant, while those for which 0.005 < P < 0.05 would be considered trends.

RESULTS

Representative Examples of CA, MN, and MN + C

Confocal micrographs showing the cellular aberrations CA, MN, and MN + C are presented in Figure 1. Panel (a) shows normal centrosomes (arrows) in untreated interphase nuclei from Xpa^(+/+)p53^(+/−) mesenchymal fibroblasts. Panel (b) shows a likely recent mitosis (two nuclei) in Xpa^(+/+)p53^(+/−) cells exposed to 100 μM AZT, where multiple centrosomes are shown in green (arrows). Panels (c) and (d) show two MN (arrows) in Xpa^(+/+)p53^(+/+) cells exposed to 100 μM AZT, as well as punctate light blue nuclear signals representing CREST staining for centromeres (kinetochores). In Figure 1d, the MN inside the circle, and in the insert, stained positive for CREST and therefore contained centromeric material considered to be a whole functional chromosome (MN + C), while the parent nucleus is considered to have lost a functional chromosome.

Fig. 1.

Confocal photomicrographs showing examples of CA, MN, and MN + C. DAPI (blue) stains nuclei. (a, b) Xpa^(+/+)p53^(+/−) mouse mesenchymal-derived fibroblasts where pericentrin antibody stains centrosomes green (arrows) in unexposed cells (a), and (b) cells exposed to 100 μM AZT for 24 hr (arrows). (c, d) show Xpa^(+/+)p53^(+/+) cells with kinetochores (centromeres) stained light blue with anti-CREST antiserum. (c) MN (arrow) in an unexposed cell. (d) A cell exposed to 100 μM AZT for 24 hr shows MN (arrow) and MN + C (circle and insert).

Cell Proliferation and Viability

Analysis of cell proliferation and viability for five of the genotypes used in this study (Table II) revealed that the unexposed cell types proliferated at about the same rate. In addition we observed moderate decreases in growth and viability in cultures exposed for 24 hr to either 10 or 100 μM AZT for the following five genotypes: Xpa^(+/+)p53^(+/+), Xpa^(−/−)p53^(+/+), Xpa^(+/+)p53^(−/−), Xpa^(+/+)p53^(+/−), and Xpa^(−/−)p53^(−/−).

TABLE II.

Index of Cell Growth During the 24 hr of AZT Exposure in Unexposed Cells, and Mean Percentage of Viable Cells Treated with 10 or 100 μM AZT During the Same 24 hr, Based on One Mouse and Triplicate Samples

				% survival with AZT dose (μM)
Genotype	Time 0a	Time 1a	Cell growth indexb	0	10	100
Xpa(+/+)p53(+/+)	39,102	86,691	2.20	100	77.6	65.5
Xpa(+/+)p53(+/−)	9,513	16,733	1.76	100	94.1	145.0
Xpa(+/+)p53(−/−)	63,276	136,787	2.16	100	84.9	75.1
Xpa(−/−)p53(+/+)	22,522	30,369	1.35	100	92.3	88.6
Xpa(−/−)p53(−/−)	20,102	35,286	1.75	100	93.0	92.1

^aAt (T0) AZT was added, and at (T1), 24 hr later the experiment was terminated.

^bcell growth index T1/T0.

Centrosomal Amplification (CA)

Values for CA in mouse fibroblasts exposed to 0, 10, and 100 μM AZT for 24 hr, are shown in Figure 2. Analysis of the factorial treatment structure revealed a three-way interaction effect, for percentage of cells with CA, among the factors tested (Xpa, p53, and AZT concentration) with P = 0.013. There were two particular aspects of note with respect to this data: first, the varying dose response effect; and second, the extent of CA formation in Xpa^(−/−) cells, which were hypothesized to sustain more damage than their Xpa^(+/+) counterparts. Dose-related increases in CA were found with AZT in cells of most genotypes (Fig. 2a). The two exceptions were the Xpa^(+/+)p53^(+/−) cells and the Xpa^(−/−)p53^(−/−) cells, neither of which showed an AZT dose-response. Figure 2b shows the least squares (LS)-means (±95% confidence intervals [CI]) for Xpa, p53, and the different AZT doses for Xpa^(+/+) mice (solid line) and Xpa^(−/−) mice (broken line). For Xpa^(+/+) cells, the figure shows dose responses among the AZT concentration means for p53^(+/+) and p53^(−/−) (F-test P = 0.0057 and P < 0.0001, respectively), but not for the p53^(+/−) cells (F-test P = 0.099). Figure 2b also shows dose responses, among the AZT concentration means, for Xpa^(−/−) cells from p53^(+/+) and p53^(+/−) mice (both F-test P = 0.0002), but not for cells from the p53^(−/−) mice (F-test P = 0.50).

Fig. 2.

CA analysis of Xpa, p53, and AZT dose by Factorial ANOVA. Analysis revealed a three-way interaction effect with P = 0.013. (a) Percent of cells with CA (and loess lines), in mesenchymal cells taken from two to four mice of each of the six different genotypes and exposed to 0, 10, or 100 μM AZT. (b) Mean fraction of cells (±95% CI) of different genotypes [Xpa^(+/+),—0—, and Xpa^(−/−), --+--] plotted on the ordinate, with AZT dose (0, 10, and 100 μM) plotted on the abscissa. The p53 genotypes (+/+, +/−, −/−) are shown in the left, middle, and right panels, respectively.

As mentioned above, cells of two genotypes, the Xpa^(+/+)p53^(+/−) and the Xpa^(−/−)p53^(−/−), showed no dose response for CA formation (Figs. 2a and 2b). There was, however, a marked difference between the two genotypes. The Xpa^(−/−)p53^(−/−) cells had high CA levels and the Xpa^(+/+)p53^(+/−) cells had low CA levels. The high CA levels were presumably due to the loss of both genes, which may have had more than an additive effect in protecting the cellular genome. In contrast, the Xpa^(+/+)p53^(+/−) cells, appeared to be protected from CA damage by their p53 haploinsufficient status.

Table III shows the pair-wise P-values for these comparisons, confirming the graphic observations represented in Figure 2b. Only in the p53^(+/−) mice was there a significant increase in CA in the Xpa^(−/−) cells, compared with the Xpa^(+/+) cells (all P ≤ 0.0001). This may be because of the extra protection provided by the p53^(+/−) genotype [see also Fig. 2b, p53^(+/−) panel].

TABLE III.

Pair-Wise P-Values for %CA in Xpa^(+/+) Versus Xpa^(−/−) Mice by p53 Status and Dose of AZT

	μM AZT
Genotype	0	10	100
Xpa(+/+) p53(+/+) vs. Xpa(−/− p53(+/+)	0.31a	0.074	0.12
Xpa(+/+) p53(+/−) vs. Xpa(−/− p53(+/−)	<0.0001	<0.0001	<0.0001
Xpa(+/+) p53(−/− vs. Xpa(−/− p53(−/−)	0.13	0.63	0.94

^aAll p values are two-tailed and adjusted using Holm’s method.

MN and MN + C

Both MN (Fig. 3a) and MN+C (Fig. 3b) reflect loss of genetic material occurring as a result of genotoxic damage. Cells with MN have sustained the loss of part of a chromosome, which does not include the kinetochore. Cells with MN + C are a subset of cells with MN where the MN contains a centromere and presumably a whole functioning chromosome. The resulting chromosomal loss, or aneuploidy, in the parent nucleus is a biologically challenging event. In this study our profiles for MN and MN + C were very similar. Figure 3a, shows the percentage of cells with MN in each of the six genotypes, and Figure 3b shows the incidence of MN + C in each of the six genotypes. (Supplementary Information Table I shows the frequencies of both MN and MN + C.) The prediction here was that Xpa^(−/−) cells would have higher damage levels than Xpa^(+/+) cells. For both MN and MN + C, values appeared higher in the Xpa^(−/−)p53^(+/+) and Xpa^(−/−)p53^(+/−) cells compared with the Xpa^(+/+)p53^(+/+) and Xpa^(+/+)p53^(+/−) cells, respectively. Only for the Xpa^(+/+)p53^(−/−) cells did the damage appear higher than for the Xpa^(−/−)p53^(−/−) cells for both MN and MN + C.

Fig. 3.

Percent of cells with MN (a ordinate) and MN + C (b) plotted as a function of AZT dose (abscissa), by mouse genotype, as dot plots with loess lines. Similar MN + C incidence is observed in cells of the Xpa^(+/+) genotype (- - -) for the three p53 genotypes, but this is not the case for Xpa^(−/−) cells (o). There was one mouse per genotype, and two slides (cell lines) per mouse.

Overall the data suggest that, whereas AZT induces marginal dose-related MN + C effects in these cells, there are significant differences in endogenous damage levels, related to cellular protective capacity provided by Xpa and p53 status. Additionally, chromosomal breaks and losses are influenced by intact NER capacity. However unlike CA which was observed at high levels in the Xpa^(−/−)p53^(−/−) cells, these same cells had very low levels of MN and MN + C. Although the percentage MN + C and MN are highly correlated, the percentage MN + C does not vary much among the three p53 genotypes when Xpa is ^(+/+). Conversely, the percentage MN + C varies considerably among the three p53 levels within the Xpa null genotype.

DISCUSSION

This study was designed to address the interaction between Xpa and p53 in the disposition of AZT in cultured mesenchymal cells taken from mice of six different genotypes. The genotoxic end points evaluated included: CA, which occurs frequently in almost all types of cancer and is a major contributing factor to chromosome instability in cancer cells; MN, comprising loss of chromosomal fragments that do not contain a centromere; and MN + C, defined as MN containing chromosomal fragments, which include a centromere, resulting in aneuploidy in the parent cell. The overall conclusion here is that both Xpa and p53 contribute to the protection of cells from AZT-induced genotoxic damage, but the type and extent of protection also depend on the end point being examined. For CA it is clear that there is an interaction between AZT dose, Xpa status and p53 status, and that in cells haploinsufficient for p53 the Xpa^(+/+) protection is more effective than even in the WT cells. In addition, in Xpa^(−/−)p53^(+/−) cells there were high levels of CA damage in the absence of an AZT dose-response. For the MN and MN + C endpoints, there was no dose response for AZT, so the effects observed are presumably due to endogenous events. Loss of both Xpa alleles resulted in increased levels of MN when p53 was either ^(+/+) or ^(+/−). However in p53^(−/−) cells, Xpa null status resulted in lower levels of both MN and MN + C, compared with the corresponding Xpa^(+/+) mice. The data suggest that events required for the formation of MN and MN + C are somehow impaired in the Xpa^(−/−)p53^(−/−) cells.

For the CA endpoint, the data showed increases with dose of AZT for four of the six genotypes. There was no dose response with the Xpa^(−/−)p53^(−/−) cells, which also sustained the highest levels of endogenous CA damage. In addition, the Xpa^(+/+)p53^(+/−) cells showed no dose-response and had very low levels of damage, suggesting that, compared with the WT cells, p53 haploinsufficiency is protective against AZT-induced damage when paired with the Xpa^(+/+) genotype. Interestingly the Xpa^(−/−)p53^(+/−) cells had significantly higher levels of CA damage than their Xpa^(+/+) counterparts, and this was the only combination in which CA incidence significantly increased in the absence of the Xpa gene. Although the protective role of p53 haploinsufficiency remains obscure, a similar phenomenon has been reported [van Gijssel et al., 2007]. These investigators found low levels of cisplatin-DNA adducts in mouse p21 (WAF/Cip1) heterozygous keratinocytes, compared with their WT or null counterparts. Apparently DNA replication arrest and apoptosis do not completely explain the low levels of cisplatin-DNA adducts in the p21^(+/−) cells, suggesting that this genotype results in either low cisplatin-DNA adduct formation or enhanced cisplatin-DNA adduct removal. In our study p53 haploinsufficiency was protective against AZT-induced DNA damage suggesting that these interactions may be worthy of further investigation. Also, because p53 plays a role in centrosomal cycle regulation, lack of both p53 alleles may result in unregulated centrosome replication, leading to centrosome amplification and chromosome instability [Fukasawa, 2008].

There is cross talk between the cell cycle, DNA damage signaling and the centrosomal cycle. Recent studies have shown that several DNA repair proteins localize to the centrosome and are involved in the regulation of centrosome number, possibly through cell cycle checkpoints or direct modification of centrosome proteins [Shimada and Komatsu, 2009; Smith and Costanzo, 2009]. It has been hypothesized that one of the mechanisms for CA formation is the lack of synchronization between centrosomal cycle and cell cycle regulation. Hence, DNA damage signaling resulted in an arrest of the cell cycle and amplification of centrosomes [Rai et al., 2008; Saladino et al., 2009]. Additionally it has been shown that centrin-2, a structural protein located on microtubules binds to XPC, most probably regulating the dynamic of both, cell and centrosomal cycles [Nishi et al., 2005; Thompson et al., 2006; Yang et al., 2006; Miron et al., 2008; Sugasawa, 2008; Acu et al., 2010; Clement et al., 2010]. Interestingly, direct interaction between centrosomal proteins and DNA damage has been reported [Acu et al., 2010], and mounting evidence suggests that the cell and centrosomal cycles are regulated in order to prevent abnormal cells from progressing into cell division [Fukasawa, 2005]. Following DNA damage, MCF-7 cells experience a translocation of the centrosome protein centrin-2 from the cytoplasm to the nucleus reducing the cytoplasmic availability of the protein and preventing CA [Acu et al., 2010]. There is evidence that individuals with Seckel syndrome, a disorder of markedly reduced brain and body size, have a mutation in pericentrin, a protein associated to centrosomes, that abrogates the signaling checkpoint ATR (ataxia-telangiectasia mutated and RAD3-related [Cimprich and Cortez, 2008] damage [Griffith et al., 2008]. Furthermore, cells of individuals with Seckel syndrome due to mutations in pericentrin have defects in ATR-dependent checkpoint signaling, providing the first evidence linking a structural centrosomal protein with DNA damage signaling [Griffith et al., 2008].

Overall this study, performed in cells cultured from mice of different Xpa and p53 genotypes and treated with AZT, shows that CA induced by AZT can be elevated in cells that lack an intact NER pathway, and that p53 also plays a role in protection against some manifestations of AZT-induced DNA damage. Because NRTI-induced DNA damage is currently considered to be repaired by base excision repair (BER) [Sampath et al., 2003], with just one report indicating the potential role of NER in the removal of molar amounts of AZT in transformed hepatic cells [Wu et al., 2013], we found these findings unexpected and worthy of further investigation.

Abbreviations

ATR

ataxia-telangiectasia mutated and RAD3-related

AZT

3′-azidoG′-deoxythymidine, Zidovudine, Retrovir®

BSA

Bovine serum albumin

CA

centrosome amplification defined as greater than two centrosomes/cell

CI

95% confidence interval

CREST antiserum

calcinosis, Raynaud’s phenomenon, esophageal dysfunction, sclerodactyly, and telangiectasia antiserum

DAPI

4′,6-diamidino-2-phenylindole

DPBS

Dulbecco’s phosphate-buffered saline

LS

least squares

MN

micronucleus

MN + C

micronucleus containing ≥1 whole chromosome(s)

NER

nucleotide excision repair

NRTI

nucleoside reverse transcriptase inhibitor

PBS

phosphate buffered saline

RPMI

Roswell Park Memorial Institute

WT

Wild type

Xpa

xeroderma pigmentosum complementation group A