Photobiological Information Obtained from XPA Gene-deficient Mice


  • This invited paper is part of the Symposium-in-Print: Photobiology in Asia.

*email: (Takeshi Horio)


The XPA gene-deficient mouse, an animal model of xeroderma pigmentosum (XP), develops enhanced photobiologic reactions including acute inflammation, immunosuppression and skin carcinogenesis, because of the defect in the excision repair of ultraviolet-induced DNA lesions. The results strongly suggest that nuclear DNA is an important chromophore to initiate acute and chronic skin damages. The model mouse is a useful experimental animal not only to investigate the mechanisms of photosensitivity in XP, but also to study physiological photobiology in humans, because photobiologic reactions are greatly intensified in this mouse.


All living things on the surface of the earth have diverse, sophisticated defense mechanisms against injurious effects of solar ultraviolet (UV) radiation. The skin is the first barrier, which functions as such physically, biochemically and immunologically. However, when the skin is exposed to an excessive amount of UV radiation beyond the protective capacity, acute and chronic damages may develop. Nuclear DNA is the most important chromophore to initiate such photobiological changes in the skin. The DNA absorbs light energy in the range of UVB and UVC, and forms cyclobutane pyrimidine dimers (CPD) and pyrimidine–pyrimidone (6–4) photoproducts. These DNA damages are repaired by various processes in normal tissues, so that serious, unwanted changes do not easily develop after UV irradiation.

Xeroderma pigmentosum

Cultured cells from patients with xeroderma pigmentosum (XP) have a defect in an early step of the nucleotide excision repair of UV-induced DNA lesions and therefore exhibit greatly increased sensitivity to the lethal effects of UV radiation (1). Clinically, XP, an autosomal recessive disorder, is characterized by extremely high sensitivity to sunlight and a more than 1000-fold increased risk of developing cancers on the sun-exposed skin (2). While sunburn reaction reaches maximal intensity by 24 h after irradiation and subsides thereafter in healthy persons, erythema or blisters develop at a lower dose of UV radiation and take several days or weeks to resolve in XP patients. Kraemer et al. surveyed comprehensively 830 published cases of XP and reported that the median age of first nonmelanoma skin cancers among patients was 8 years, approximately 50 years earlier than that of the normal U.S. population and half of XP patients in the 10–14 age group had skin cancers (3). However, there is considerable variability in photosensitivity severity and tumor development among patients. Genetic complementation analysis by cell fusion has identified at least eight complementation groups of XP, designated group A through G, and a variant form (4). The clinical severity correlates roughly with the grade of the DNA repair defect. Group A XP is the most common form of this disease in Japan and shows the most severe clinical symptoms at an early age.

XPA gene-deficient mouse

Recently, mouse genes corresponding to each complementation group of XP have been identified and the encoded proteins were characterized biochemically and functionally. The XPA gene encodes protein of 273 amino acid residues with a C4 type zinc finger domain (5). The XPA protein plays a role in congregating nucleotide excision repair proteins at the UV-damaged site of DNA (6). The XPA gene and its mouse homolog have been cloned by Tanaka et al. (5). Although the cellular and molecular abnormalities in XP have been intently studied, how the defect in nucleotide excision repair causes the clinical features of XP has not been completely elucidated. In vivo studies using animal models are often helpful to understand the link between clinical and molecular abnormalities. By means of gene targeting, murine models of XP have been generated. In the model of group A-XP, the XPA allele was disrupted by insertion of neomycin cassette sequenced into exon 4 of the XPA gene by embryonic stem cell techniques (7). The XPA gene-deficient mice with CBA, C57BL/6 and CD-1 chimeric genetic background were backcrossed with hairless albino mice of the inbred strains Hos/HR-1, and the resultant hairless XPA (−/−), (+/−) and (+/+) mice were used in our experiments. The XPA (−/−) fibroblasts derived from XPA-knockout mice were hypersensitive to killing by UV irradiation, and UV-induced unscheduled DNA synthesis was absent. Removal of CPD and (6–4) photoproduct was not detected after UV irradiation in the XPA (−/−) fibroblasts. The model mice have defects in nucleotide excision repair without any obvious physical abnormalities or pathologic alterations. Similar to that of XP patients, these mice showed a high incidence of skin cancers initiated by low dose of UVB irradiation (Fig. 1).

Figure 1.

 XPA gene-deficient mouse easily and rapidly develops squamous cell carcinomas by UVB irradiation.

DNA damage and inflammation

Sunburn erythema and blister are acute inflammatory reactions which first appear visibly after exposure to sunlight. Individuals with high UVB sensitivity show higher pyrimidine dimer yields than less sensitive persons (8), suggesting that DNA is a major chromophore for UV inflammation (9,10). Patients with group A-XP develop exaggerated and long-lasting sunburn reactions. Then, acute inflammatory reactions to UVB radiation were examined in XPA (−/−) mice (11).

The UVB source was a bank of fluorescent sunlamps (FL20SE.30; Toshiba Medical Supply, Tokyo, Japan) with an emission spectrum of 275–375 nm peaking at 305 nm. The irradiance of UVB was measured by a radiometer (UVR-305/365D(II); Toshiba Medical Supply). Acute inflammation was evaluated by ear swelling response to UVB radiation. Ear thickness was measured with a dial thickness gauge (Peacock, Tokyo, Japan) before and after irradiation at various time points. A single exposure to 100 or 250 mJ cm−2 of UVB radiation resulted in a significant ear swelling in the XPA (−/−) mice 24 h after irradiation, and the edema was still increasing at day 5 (Fig. 2). In contrast, the wild type and heterozygous (+/−) mice did not develop significant ear swelling. A higher dose of UVB radiation induced significant ear swelling in not only (−/−) mice but also (+/+) or (+/−) mice. However, ear swelling in (−/−) mice was significantly greater than that in (+/+) mice or (+/−) mice throughout the study.

Figure 2.

 XPA gene-deficient mice develop severe and long-lasting edema after UVB irradiation. Ears of mice were exposed to 250 mJ cm−2 of UVB and ear thickness was measured immediately before and 1, 2, 3 and 5 days after irradiation.

Histologic changes were more prominent in UVB-irradiated skin of XPA (−/−) mice (11) (Fig. 3). At 24 h after 500 mJ cm−2 UVB irradiation, skin samples from XPA (−/−) mice showed intracellular edema and necrosis in the epidermis and subepidermal bullae while only little changes were observed in the epidermis of (+/+) or (+/−) mice. Moreover, (−/−) mice revealed marked inflammatory infiltrates of lymphocytes, pronounced edema, vasodilatation and prominent extravasation of erythrocytes in the dermis. The XPA (−/−) mice developed enhanced sunburn cell formation after UVB radiation. At 24 h after 50 mJ cm−2 of UVB irradiation, significantly enhanced sunburn cell formation was induced in the epidermis of (−/−) mice (59.6 ± 32.2 cm−1; mean ± SD) compared with that in (+/+) and (+/−) mice (20.6 ± 23.8 and 17 ± 22.8 cm−1, respectively). Similarly, sunburn cells induced by UVB irradiation at a higher dose (100 mJ cm−2) in (−/−) mice were almost three times as numerous as those in (+/+) or (+/−) mice.

Figure 3.

 At 24 h after 500 mJ cm−2 UVB irradiation, abdominal skin samples from XPA (+/+) mice revealed only a little change in the epidermis (top). In contrast, the samples from (−/−) mice showed intracellular edema and necrosis in the epidermis and subepidermal bullae (bottom).

The present study using XPA (−/−) mice indicated that the defect in excision repair of pyrimidine dimers is one of the molecular mechanisms involved in the acute photosensitivity in XP. Formation of sunburn cells is regarded as an apoptotic process of keratinocytes that removes the DNA-damaged or mutated cells (12). Enhanced sunburn cell formation in the XPA (−/−) mice after UVB irradiation may reflect the increase in unrepaired pyrimidine dimers due to the deficiency in DNA repair. This is supported by the observation that sunburn cell formation could be prevented by topical treatment of skin of UVB-exposed mice with liposomal T4 endonuclease V, an enzyme preferentially involved in the excision repair of pyrimidine dimers (13).

Involvement of DNA damage in prostaglandin production after UV irradiation

Possible chemical mediators which are involved in UV-induced inflammation include prostaglandins (PGs), histamine, serotonin and kinin. Among them, at least in human skin, PGE2 is known to be a major mediator in sunburn reaction (14,15), which can be partially inhibited by indomethacin application (16). Then, we examined the effect of indomethacin, a potent inhibitor of PG biosynthesis, on the increased ear swelling response after UVB irradiation in the XPA (−/−) mice. Topical application of 1% indomethacin immediately after UVB irradiation significantly suppressed ear swelling in XPA (−/−) as well as (+/+) mice. A similar level of percent suppression was observed in both mice (17).

Inhibitory effects of indomethacin in the enhancement of acute UV reactions in the XPA (−/−) mice suggested that UVB radiation produced a high amount of PGs in these mice. The amounts of PGD2, PGE2 and PGF2 in mouse ears at 0, 24, 48 and 72 h after irradiation with 250 or 500 mJ cm−2 of UVB were determined by enzyme immunoassay. The amounts of PGs in the XPA (−/−) mouse skin significantly increased at 48 and 72 h after irradiation. Among three PGs, PGE2 most markedly increased to levels 4- to 10-fold higher than PGD2 and PGF2. Furthermore, the amount of PGE2 in (−/−) mice was approximately 8- and 15-fold higher than that in (+/+) mice at 48 and 72 h after irradiation, respectively (Fig. 4). The amount of PGE2 in the skin of UV-irradiated (−/−) mice was not detected by treatment of 1% indomethacin at 24 and 48 h after irradiation. Increased production of PGs in the skin of irradiated mice suggested that UVB exposure induced synthesis of cyclooxygenase (COX). To analyze the expression of COX-1 and COX-2 genes, mRNA was isolated from the mouse ears after an exposure to 250 mJ cm−2 UVB irradiation, and subjected to RT-PCR analysis. The expression of COX-2 mRNA in (−/−) mice increased from 24 h after UV irradiation in a time-dependent manner. COX-1 mRNA was expressed nearly constantly and was not affected by the UV irradiation in either (+/+) or (−/−) mice.

Figure 4.

 Both ears of each mouse were removed at various intervals after 250 mJ cm−2 of UVB irradiation. The amount of prostaglandin E2 (PGE2) in ears was determined by enzyme immune assay.

The molecular mechanisms of the severe inflammation found in XPA-deficient mice after UV exposure have not been clarified. This experiment indicated that the mechanisms would be due to increased expression of mRNA for COX-2 and subsequent overproduction of PGE2 after UV exposure.

Enhanced UVB damage of Langerhans cell in XPA gene-deficient mice

As epidermal keratinocytes of XPA (−/−) mice were severely damaged by UVB radiation, the alteration of Langerhans cell (LC), another epidermal population, was also examined. In nontreated skin, approximately the same numbers of LC were found in the XPA (−/−), (+/+) and (+/−) mice. Single exposure to 25 mJ cm−2 of UVB reduced the number of LC by 59% of the preirradiated level in XPA (−/−) mice 24 h after irradiation. On the other hand, the percentage reductions in the XPA (+/+) and (+/−) mice were 33% and 38%, respectively. After UVB irradiation at 100 mJ cm−2, the number of LC decreased by almost 100% in the XPA (−/−), but by only 62% in (+/+) mice. Furthermore, the recovery of LC density was slower in the XPA (−/−) mice than in the (+/+) or (+/−) mice (11).

Enhanced UV immunosuppression in XPA gene-deficient mice

It has become apparent that UVB radiation has immunosuppressive effects, which may contribute to the etiology of sunlight-induced skin cancer as well as to mutagenic properties of UVB (18–20). Kripke and colleagues proposed that DNA may serve as the photoreceptor for UV-induced immunosuppression. They observed that irradiating the South American opossum with photoreactivating light immediately after UVB exposure prevented UV-induced immunosuppression. In this animal, irradiation with visible light activates the photoreactivating enzyme that repairs pyrimidine dimers (21). Furthermore, UVB-induced systemic immunosuppression was prevented nearly completely by topical treatment of UVB-irradiated mice skin with functionally active T4 endonuclease V in liposomes (22).

We have demonstrated that the expression of COX-2 and production of PGE2, which is one of the possible factors involved in UV immunosuppression (23), are greatly increased in XPA (−/−) mice after UVB irradiation. Moreover, epidermal LC, an antigen presenting cell, was easily damaged by UVB radiation in the mice. Based on these observations, the effect of UVB radiation on contact hypersensitivity (CHS) was examined in XPA (−/−) mice (11).

The mice were sensitized by epicutaneous application of 25 μL of 1% dinitrofluorobeneze (DNFB) solution on abdominal skin. CHS was elicited by application of 20 μL of 0.2% DNFB solution on the surface of each left ear 6 days after sensitization. Ear thickness was measured before and 24 h after application of the challenge dose. The XPA (−/−) mice developed CHS to DNFB similar to the XPA (+/+) and (+/−) mice. Sensitization with DNFB in the skin that had been exposed to 100 mJ cm−2 UVB resulted in a significantly decreased CHS response in both XPA (+/+) and (−/−) mice (59.1% and 54.1% suppression, respectively). In the XPA (−/−) mice, almost the same degree of suppression was induced by lower doses of UVB radiation such as 40 and 10 mJ cm−2 (69.4% and 56.3% suppression, respectively), whereas less suppression was induced in the (+/+) mice (18.6% and 21.7% suppression, respectively) and in the (+/−) mice (30.9% and 18.4% suppression, respectively) (Fig. 5). The results indicate that UVB-induced local immunosuppression was enhanced in the XPA (−/−) mice.

Figure 5.

 Local immunosuppression. Groups of mice were exposed to UVB at a dose of 10, 40 or 100 mJ cm−2. Sensitization was attempted 1 day after irradiation by applying 1% dinitrofluorobeneze (DNFB) to the irradiated skin. Five days later, the mice were challenged with 0.2% DNFB. Data are expressed as the mean of percentage suppression (±SD).

Next, the experiments of systemic immunosuppression were performed. Exposure to UVB at a dose of 500 mJ cm−2 5 days before sensitization with DNFB on nonirradiated skin induced stronger suppression of CHS in the XPA (−/−) mice than in the (+/+) or (+/−) mice (100%vs 76.5% or 76.7%, respectively). Although UVB radiation at 125 mJ cm−2 produced only a little suppression on (+/+) or (+/−) mice (43.1% or 41.9% suppression, respectively), it produced a pronounced suppression in (−/−) mice (92.0% suppression) (Fig. 6).

Figure 6.

 Systemic immunosuppression. Dorsal skin of mice was exposed to UVB at a dose of 0.125 or 0.5 J cm−2. Sensitization was attempted 5 days after irradiation by 1% dinitrofluorobeneze (DNFB) to the unirradiated abdominal skin. Five days later, the mice were challenged epicutaneously with 0.2% DNFB. Data are expressed as mean of percentage suppression (±SD).

There are reports that XP patients have defect in cell-mediated immunity such as impaired cutaneous responses to recall antigens (24,25) and contact sensitizers (24,26). Our investigations suggest that these immune deficiencies are not innate but acquired after exposures to sunlight.

Enhanced UVB damage of natural killer cell in XPA gene-deficient mice

Natural killer (NK) cells play an important part in tumor surveillance. Norris et al. reported reduced NK cell activity in five patients with XP, but subsequently this group described that an XP patient with multiple skin cancers had normal NK cell activity (27–29). Gaspari et al. demonstrated that NK cell function was 40% of the normal control in five of eight XP patients (30). To study whether UV radiation induced suppression of NK cell function involved in the high incidence of skin tumors in patients with XP, we analyzed the number and activity of NK cells in UVB-irradiated XPA model mice (31). The number of NK cells in peripheral blood significantly decreased after UVB irradiation only in XPA (−/−) mice, but those in the spleen were not affected. Compared with the wild-type mice, the XPA (−/−) mice displayed a higher level of spontaneous splenic NK cell activity (10–15%vs 3%) and inducible NK activity (30–50%vs 20–25%) after injection of polyinosinic:polycytidylic acid. At 24 h after the last irradiation of three and five daily consecutive exposures to 500 mJ cm−2 UVB, however, the NK activity in XPA (−/−) mice decreased to 60% and 30% of the preirradiated level, respectively, but it did not in the wild-type mice. The depression of NK activity in XPA (−/−) mice recovered to a normal level at 10 and 15 days after last irradiation, respectively.

The investigation suggests that the enhanced UV-induced impairment of NK function could be partially involved in cancer development of XP patients. The mechanism for suppression of NK activity in vivo by UV radiation is not known. It is possible that UV radiation exerts a direct effect on NK cells circulating through the dermal capillaries. The defect in repair of DNA damage in XP cells may be involved in the enhancement of UV-induced suppression of NK activity in XPA mice. Alternately, soluble mediators may be released from the epidermal or dermal cells on UV irradiation and indirectly affect NK cell function. The proliferation and activation of NK cells are induced by various cytokines such as IL-2, IL-12 and IFN-γ (32,33). In contrast, cytokines such as TGF-β and IL-10 and PGE2 have been shown to inhibit NK cell activity (34,35). Excess production of both PGE2 and IL-10, induced by UV exposure, may cause the enhancement of UV-induced impairment of NK cell activity in XPA mice.

UV-induced impairment of tumor rejection in XPA gene-deficient mice

The underlying mechanisms of UV-induced immunosuppression have been extensively studied in T cell-mediated immune reactions including CHS and delayed type of hypersensitivity. It is well established that the tumor immunity plays an extremely important role in regulating the growth and metastasis of tumors. Especially, cell-mediated immunological reactions play a crucial role in tumor immunity. Thus, to examine directly whether UVB-induced DNA damage effects is involved in the impairment of immunological reactions against tumors, UVB radiation on tumor rejection were compared between XPA (−/−) and wild-type mice (36).

Tumor cells established from UVB-induced squamous carcinoma in XPA mice were injected subcutaneously. No difference in the development of tumors was observed between the nonirradiated XPA and wild-type mice. Tumors developed, grew in size and reached the maximum at 7–10 days after the inoculation. Thereafter, all tumors decreased in size and were completely rejected by 4 weeks in both strains of mice. When tumor cells were inoculated into the skin that had been irradiated with 50–150 mJ cm−2 of UVB, tumor grew in 60% (12 of 20) of the XPA mice, but only in 4% (one of 23) wild-type mice (Fig. 7). Phenotyping of tumor-infiltrating cells revealed that the migration of natural killer cells and CD8 (+) T cells was inhibited in UVB-irradiated XPA mice. These data suggest that enhanced UVB-induced impairment of tumor rejection could be partially involved in the cancer development of XP patients.

Figure 7.

 UVB radiation suppressed rejection of transplanted tumor cells from XPA mice. Mice were exposed to UVB at a dose of 100 mJ cm−2 for two consecutive days. One day after the second exposure, a suspension of 3 × 107 squamous cell carcinoma was injected subcutaneously into UVB-irradiated XPA (bottom) and wild-type mice (top). Seventeen days after tumor cell inoculation, wild-type mice completely rejected tumors, whereas XPA mice had large tumors at 6 weeks.

Chemical carcinogen-induced inflammation and immunosuppression in XPA gene-deficient mice

Similar to UVB radiation, a chemical carcinogen, dimethylbenz(a)anthracene (DMBA) forms adduct with DNA. XPA protein is required to repair DMBA–DNA adducts. Actually the XPA (−/−) mice easily developed cancers by the topical application of DMBA as well as by UVB radiation (7). An application of DMBA to the skin depletes LC and suppresses CHS, when a sensitizer is applied to DMBA-treated skin (37–39). Based on these backgrounds, we investigated if DMBA-induced inflammation and immunosuppression are enhanced in the XPA (−/−) mice (40). A single application of 0.1% or 0.5% DMBA resulted in a significant ear swelling in the XPA (−/−) mice 2 days after application. In contrast, the wild-type mice did not develop any significant ear swelling even with a 2.5% DMBA application (Fig. 8). There were no differences in ear swelling responses to other primary irritating chemicals such as croton oil or phenol between the XPA (−/−) mice and wild-type mice.

Figure 8.

 XPA mice, but not wild-type mice, developed significant ear swelling in response to dimethylbenz(a)anthracene (DMBA) application dependent on the concentration. Twenty microliters of various concentrations of DMBA was applied to the ears of mice, and ear thickness was measured immediately before and then 1, 2, 3, 4 and 5 days after application.

The number of epidermal LC significantly decreased after application of 0.5% DMBA in both the XPA (−/−) and XPA (+/+) mice. The DMBA application of a lower concentration (0.1%) also greatly reduced LC in the XPA (−/−) mice, but not in the wild-type mice. Sensitization with DNFB on skin that had received 0.1% or 0.5% DMBA application resulted in almost complete suppression of the CHS response in XPA mice (96% and 94% suppression, respectively). In contrast, less suppression was induced in the wild-type mice with pretreatment using 0.1% or 0.5% DMBA (24% and 36% suppression, respectively) (Fig. 9). Application of 0.1% DMBA on the back 4 days before sensitization with DNFB on the abdominal skin induced a stronger suppression of CHS in the XPA mice than in wild-type mice (82%vs 45% suppression).

Figure 9.

 Enhanced local immunosuppression was induced by dimethylbenz(a)anthracene (DMBA) application in XPA mice. Groups of mice received applications of 60 μL of DMBA at concentrations of 0.1% and 0.5%. Sensitization was attempted 4 days after DMBA treatment by applying 25 μL of 1% dinitrofluorobeneze (DNFB) to the site that had received DMBA treatment. Six days after sensitization the mice were challenged at the left ears epicutaneously with 20 μL of 0.2% DNFB.

Dimethylbenz(a)anthracene application induced pronounced production of PGE2, IL-10 and TNF-α in the skin of the XPA (−/−) mice. Treatment with indomethacin inhibited DMBA-induced inflammation and immunosuppression.

Development of melanoma model (XPA-deficient, SCF-transgenic) mouse

Epidemiological studies strongly suggest that sunlight plays a critical role in the induction of malignant melanoma. To date, however, there have been few appropriate animal models for studying the role of UV radiation in melanoma carcinogenesis, mainly because of the lack of the epidermal melanocyte, which is the major source of origin of human melanoma. As described in this article, XPA gene-deficient mice easily and rapidly developed squamous cell carcinoma of keratinocyte origin, but never developed melanoma, as they have no epidermal melanocytes. The growth and differentiation of melanocytes require stem cell factor (SCF) and the ligand for the c-Kit receptor tyrosine kinase. The lack of SCF expression in keratinocytes is thought to result in the melanocyte-deficient skin of the mouse except for the hair follicle. Then, we established XPA gene-deficient, SCF-transgenic mice, which are defective in the repair of damaged DNA and do have epidermal melanocytes (41). The mice were exposed to UVR three times a week for 10 weeks. At 4 to 6 months after the termination of the exposure, more than 30% of the irradiated mice developed tumors of melanocyte origin (Fig. 10) that metastasized to the lymph nodes. Histologically, proliferated cells exhibited lentigo maligna melanoma or nodular melanoma. Immunohistochemistry confirmed that the tumor cells were characteristic of melanoma. Nonirradiated mice did not develop skin tumors spontaneously. The newly generated mouse model might be useful for studying the photobiological aspects of human melanoma, because the mice developed melanoma from melanocytes only after UVR exposures.

Figure 10.

 Nodular malignant melanoma developed on the irradiated skin of XPA (−/−), stem cell factor-transgenic mouse at 6 months after termination of UVB exposure.


The xeroderma pigmentosum model mouse is a useful experimental animal not only to investigate the mechanisms of photosensitivity in the disorder, but also to study physiological photobiology in humans, because photobiologic reactions are greatly intensified in this mouse. It has been well known that high incidence of skin cancers in XP patients is due to the defect in repair of UV-induced DNA damages. Our investigations using XPA gene-deficient mice indicated that acute inflammation and immunosuppression after UVB irradiation are greatly enhanced due to hyperproduction of chemical mediators and cytokines, when the excess DNA photoproducts remain in the exposed skin. The results strongly suggest that nuclear DNA is an important chromophore to initiate acute as well as chronic photobiologic reactions.