MATERIALS AND METHODS

Patient recruitment and UV source

Ethical approval for our study was obtained from Beaumont Hospital ethics committee. Seventeen control volunteers with benign skin disorders unrelated to sun exposure and 19 patients with BCC attending the dermatology clinics were recruited after informed consent. Volunteers were irradiated with three times the predetermined minimal erythema dose using solar simulated radiation (SSR), equivalent to exposure to natural sunlight at midday at 40° latitude, as described by Murphy et al.⁴⁰ No significant difference between the UV doses received by the patients with BCC and the doses received by the normal controls (table 1) was found using the Student’s t test. Four millimetre punch biopsies were taken at 4.5, 9, 24, 33, 48, and 72 hours after SSR. A control biopsy was taken from each volunteer from non-irradiated buttock skin. Some individuals gave more then one irradiated biopsy (in some cases, more than one irradiated biopsy was taken from the same individual; the dose of radiation was the same but the biopsy was excised at a different time point). Therefore, in total, 24 post UV irradiation punch biopsies were taken from controls (three at 4.5 hours, four at 9 hours, four at 24 hours, seven at 33 hours, four at 48 hours, and two at 72 hours) and 26 post UV irradiation punch biopsies were taken from patients with BCC (two at 4.5 hours, four at 9 hours, four at 24 hours, eight at 33 hours, five at 48 hours, and three at 72 hours). All tissues were fixed in 10% unbuffered formalin, paraffin wax embedded, and 4 μm serial sections were cut. Results from all biopsies were used in the statistical analysis.

p53, p21, and Bax immunohistochemistry

Serial sections were stained with each antibody, using the avidin–biotin peroxidase complex method, as described previously.^40,43 Positive control slides were included in all tests and consisted of paraffin wax sections from tissues known to be positive for p53, p21, and Bax. In addition, a negative control was included, in which primary antibody was replaced with buffer.

Detection of apoptosis

The induction of apoptosis was determined using two different approaches.

Haematoxylin and eosin (H&E) staining

As described previously, serial sections were stained with H&E. Features characteristic of apoptosis^18,40 were identified using microscopy by two different observers. It is important to note that sunburn cells were regarded as apoptotic cells in a different phase of apoptosis (prefragmentation) than cells where DNA fragmentation was visible on H&E. That is, sunburn cells were identified by their eosinophilic cytoplasm, whereas apoptotic cells were identified by their shrunken fragmented nuclei.

The in situ cell death detection kit

Nick end labelling of fragmented DNA, the TUNEL (terminal deoxynucleotidyl transferase mediated dUTP nick end labelling) method, was performed using the in situ cell death detection kit (Boehringer Mannheim, Diagnostics and Biochemicals Ltd, East Sussex, UK) according to the manufacturer’s instructions.⁴⁵ Both apoptotic cells and sunburn cells were stained brown by TUNEL.

Counting and analysis

Apoptotic cells on H&E were defined as cells with visible nuclear fragmentation or pyknotic nuclei with eosinophilic cytoplasm, and apoptotic cells staining with the in situ cell death kit in addition to immunopositive (p53, p21, and Bax) cells were counted at high magnification (×400). Sunburn cells with condensed chromatin, but no fragmented nuclei, were counted separately to determine whether these cells behaved similarly to cells with fragmented DNA. The morphology and percentage of positive cells and the intensity of staining were assessed by two independent observers (MM, MJEMFM). The average number of positive cells in each field was calculated. We used the Mann Whitney test to see whether there were any differences between normal controls and patients with BCC in the response to UV.

RESULTS

p53 expression

In normal controls, the expression of p53 was induced immediately after UV exposure at 4.5 hours (mean, 17.67 positive cells/field; SD, 7.33) (fig 1A). Peak expression was seen between 9 (mean, 50.9 positive cells/field; SD, 14.50) and 33 hours after UV irradiation (mean, 65.42 positive cells/field; SD, 13.84). Expression began to level off between 48 hours (mean, 40.99 positive cells/field; SD, 10.16) and 72 hours after UV irradiation (mean, 48.42 positive cells/field; SD, 13.37). The time course in patients with BCC was found to be similar to normal controls. The staining of positive cells in both groups was of similar intensity and positivity was located throughout the epidermis.

Using the Mann-Whitney test no significant differences were found between the two groups.

p21 expression

In normal controls, p21 expression was initiated at 9 hours (mean, 6.78 positive cells/field; SD, 7.97) (fig 1B) after UV irradiation. This expression peaked between 33 hours (mean, 9.79 positive cells/field; SD, 7.41) and 48 hours (mean, 9.67 positive cells/field; SD, 10.36) and then declined at 72 hours after UV irradiation (mean, 2.45 positive cells/field; SD, 2.08). In patients with BCC, a trend towards higher p21 expression was seen: 9 hours (mean, 12.27 positive cells/field; SD, 11.13), 24 hours (mean, 12.21 positive cells/field; SD, 8.70), and 33 hours after UV irradiation (mean, 18.69 positive cells/field; SD, 7.73). Although none of these values was significantly different from normal controls, at 33 hours it was close to significance (p = 0.54).

Bax expression

In normal controls, Bax expression was first seen at 4.5 hours (mean, 0.61 positive cells/field; SD, 0.63) (fig 1C) after UV irradiation, reached a peak at 24 hours (mean, 1.45 positive cells/field; SD, 2.06), and declined from 33 hours (mean, 1.09 positive cells/field; SD, 1.32) to 72 hours (mean, 0.39 positive cells/field; SD, 0.01). In patients with BCC, the time course appeared to be similar except that Bax expression peaked slightly earlier at 9 hours (mean, 1.14 positive cells/field; SD, 1.99) and was sustained until 33 hours (mean, 1.31 positive cells/field; SD, 1.14). It then declined from 48 hours (mean, 0.46 positive cells/field; SD, 0.49) to 72 hours (mean, 0.36 positive cells/field; SD 0.36). Using the Mann-Whitney test no significant difference was found between the two groups.

The induction of apoptosis

Apoptosis was detected and scored by means of the in situ cell death kit and H&E staining. The pattern of response was found to be the same for both methods. This indicated that data generated by counting apoptotic cells as assessed by H&E staining were comparable to data generated by the in situ cell death detection kit (fig 2). Sunburn cells and DNA fragmented cells stained with TUNEL, indicating that both represent apoptotic cells.

In normal controls, apoptosis was initiated between 9 hours (mean, 0.66 apoptotic cells/field; SD, 0.32) (fig 1D) and 24 hours after UV irradiation (mean, 2.08 apoptotic cells/field; SD, 1.52), peaked at 48 hours (mean, 4.22 apoptotic cells/field; SD, 2.73), and declined at 72 hours (mean, 1.61 apoptotic cells/field; SD, 0.39). This response was altered in patients with BCC in that apoptosis peaked earlier at 24 hours after UV irradiation (mean, 3.69 apoptotic cells/field; SD, 1.61) and remained high up to 48 hours (mean, 3.02 apoptotic cells/field; SD, 2.18). The difference between the two groups at the 33 hour time point was found to be significant (p = 0.04) using a Mann-Whitney test.

DISCUSSION

BCCs are the most common skin cancer in the Western world.^9–11 Causes of BCC involve UV induced or inheritable p53 or patched gene mutations.^41–43 p53 mutations result in the loss of cell arrest and hence damaged cells are not repaired or deleted.^28–31 Patched gene mutations or mutations of other genes involved in the hedgehog pathway result in the upregulation of pathways that cause increased proliferation, such as the ras–ERK pathway.⁴⁴ Hence, disruption of cell cycle arrest and or of hedgehog signalling predisposes to BCC.^28–31,45

UV is a major risk factor for BCC and it has been suggested that inefficient apoptosis has a role in the genesis of neoplasia.^6–8 Apoptosis is a truly error free method of deleting DNA damaged cells, whereas DNA repair mechanisms are error prone. It was shown that resistance to Fas mediated apoptosis develops in a cell model of human proliferative breast disease (a precursor to breast cancer),⁴⁶ and also that tumour angiogenesis, which promotes growth and metastasis of lung cancer, inhibits apoptosis.⁴⁷ These findings have led to the proposal that inefficient apoptosis could be a risk factor that leads to the development of mutations seen commonly in BCCs.

Therefore, in our study, buttock skin (previously non-exposed skin) from patients with BCC and normal controls was irradiated, biopsies were excised, and these were examined for p53, p21, and Bax protein expression and the induction of apoptosis.

“Patients with basal cell carcinoma seem to be inherently predisposed to UV induced cellular damage”

We found that apoptosis occurred via the p53 pathway. That is, in normal controls, p53 expression began at 9–33 hours, followed by p21 expression (33–48 hours), and then Bax expression (24 hours), and apoptosis was induced 48 hours after UV irradiation. In patients with BCC, the response to UV was altered in two ways. First, there was a trend towards higher p21 expression. Second, apoptosis was induced earlier in patients with BCC and was significantly higher than that seen in controls 33 hours after UV irradiation (p = 0.4).

Although the time course for apoptosis varied between patients with BCC and normal controls, the time course for Bax expression did not. This, coupled with the fact that Bax expression was present at very low levels (peak expression, 1.45 cells/field), suggests that other apoptosis inducing proteins are present—for example, PERP, Fas ligand, tumour necrosis factor α, and p53 inducible genes.^48–50 These genes all initiate cell death and have all been shown to be upregulated by p53. In addition, apoptosis could also occur through a p53 independent pathway.⁵¹ Therefore, in the future it would be useful to look at the expression of these proteins to determine whether different apoptotic pathways function in patients with BCC compared with controls.

However, although more work is needed to elucidate which apoptotic pathways are occurring in BCC in comparison with controls, the end result is the same—apoptosis is occurring earlier in BCCs and is significantly higher at 33 hours. This could suggest that more DNA damage is occurring in patients with BCC (that is, they are more sensitive to insult) or that in these patients cell cycle arrest is faulty.

Higher p21 expression in patients with BCC indicates that an increase in cell cycle arrest has occurred. This, coupled with the fact that apoptosis occurs earlier in patients with BCC, suggests that these patients have suffered more DNA damage than normal controls after equivalent UV exposure.

Therefore, patients with BCC seem to be inherently predisposed to UV induced cellular damage. This seems plausible because individuals with increased sensitivity to UV have been recorded⁵² and, in addition, it has been shown that increased vulnerability to sun exposure comes with old age, as does the development of BCC.⁵³

p21 expression is upregulated in patients with BCC after UV exposure. However, this does not always mean that proper cell arrest has occurred³⁹ because other factors, such as a mutant retinoblastoma gene⁵⁴ or a lack of essential cofactors, could prevent it.^55,56 Inefficient cell arrest or inefficient cell repair (which has been shown to be a risk factor for BCC)^27,57,58 could result in incomplete surveillance—that is, pushing cells into apoptosis earlier and allowing cells with low damage to go through the cell cycle unchecked, ultimately resulting in duplication of damage and possibly the generation of a tumour. An analogous situation has been seen in LEC cells (a cancer sensitive cell line), where in response to UV irradiation there was an increased sensitivity to apoptosis without cell cycle arrest.⁵⁹ Further support for this argument comes from the fact that BCC tumours have high levels of apoptosis,⁶⁰ indicating that the apoptotic machinery is intact in patients with BCC and that the initial fault must lie elsewhere.

In conclusion, it was found that the response to UV irradiation is altered in patients with BCC when compared with normal controls. This suggests that individuals with BCC have an altered defence mechanism, and further investigations at the molecular level should be carried out to identify this predetermining BCC factor.