Mutations induced by UVB (313-nm) radiation, a wavelength in the region of peak effectiveness for sunlight-induced skin cancer in humans, have been analyzed at the sequence level in simian ceUs by using a plasmid shuttle vector (pZ189). We find that significant differences exist between the types of mutations induced by this solar wavelength and those induced by nonsolar UVC (254-nm) radiation. Compared with 254-nm radiation, 313-nm radiation induces more deletions and insertions in the region sequenced. In addition, although the types of base substitutions induced by the two wavelengths are broadly similar (in both cases, the majority of changes occur at G-C base pairs and the G-C to A-T transition is predominant), an analysis of the distribution of these base changes within the supF gene following irradiation at 313 nm reveals additional hot spots for mutation not seen after irradiation at 254 nm. These hot spots are shown to arise predominantly at sites of mutations involving multiple base changes, a class of mutations which arises more frequently at the longer solar wavelength. Lastly, we observe that most of the sites at which mutational hot spots arise after both UVC and UVB irradiation of the shuttle vector are also sites at which mutations arise spontaneously. Thus, a common mechanism may be involved in determining the site specificity of mutations, in which the DNA structure may be a more important determinant than the positions of DNA photoproducts.

Cultured fibroblasts derived from normal human skin have been irradiated at a series of monochromatic wavelengths throughout the ultraviolet region and exposed to the specific α polymerase inhibitor, aphidicolin (1 μg/ml, 2 days) prior to assay for colony forming ability. Repair of 75‐80% of the lethal damage induced by UVC (254 nm) or UVB (302 nm, 313 nm) radiation is inhibited by aphidicolin suggesting that such damage is repaired by a common α polymerase dependent pathway. Exposure to aphidicolin after irradiation at longer UVA (334 nm, 365 nm) or a visible (405 nm) wavelength leads to slight protection from inactivation implying that the processing of damage induced in this wavelength region is quite distinct from that occurring at the shorter wavelengths and does not involve α polymerase.

Mutations induced by UVB (313-nm) radiation, a wavelength in the region of peak effectiveness for sunlight-induced skin cancer in humans, have been analyzed at the sequence level in simian cells by using a plasmid shuttle vector (pZ189). We find that significant differences exist between the types of mutations induced by this solar wavelength and those induced by nonsolar UVC (254-nm) radiation. Compared with 254-nm radiation, 313-nm radiation induces more deletions and insertions in the region sequenced. In addition, although the types of base substitutions induced by the two wavelengths are broadly similar (in both cases, the majority of changes occur at G-C base pairs and the G-C to A-T transition is predominant), an analysis of the distribution of these base changes within the supF gene following irradiation at 313 nm reveals additional hot spots for mutation not seen after irradiation at 254 nm. These hot spots are shown to arise predominantly at sites of mutations involving multiple base changes, a class of mutations which arises more frequently at the longer solar wavelength. Lastly, we observed that most of the sites at which mutational hot spots arise after both UVC and UVB irradiation of the shuttle vector are also sites at which mutations arise spontaneously. Thus, a common mechanism may be involved in determining the site specificity of mutations, in which the DNA structure may be a more important determinant than the positions of DNA photoproducts.

We have used the eukaryotic DNA polymerase alpha inhibitor, aphidicolin, and the polymerase beta inhibitor, dideoxythymidine, to examine the role of these enzymes in excision repair of ultraviolet (u.v., 254 nm) damage induced in non-dividing (arrested) human skin fibroblasts. The effects of these drugs on u.v.-treated cells have been monitored using a simple and reproducible repair synthesis assay in parallel with viability measurements to determine the degree of inhibition of repair of potentially lethal damage. In agreement with previous studies using density gradients, repair synthesis induced by low fluences of u.v. (less than 3 J m-2) is relatively insensitive to inhibition by aphidicolin compared to high fluences where approximately 85 per cent inhibition is observed at the highest (20 micrograms/ml) aphidicolin concentration employed. However, repair of potentially lethal damage is inhibited by at least 90 per cent over the entire fluence range. Although dideoxythymidine led to considerable inhibition of repair synthesis, the result is probably an artifact under these in vivo conditions. The polymerase beta inhibitor was not toxic to u.v.-treated cells nor did it add to the toxicity of aphidicolin when the drugs were used in combination. We conclude that if the beta polymerase is involved in excision repair then its temporary (4 h) inhibition by dideoxythymidine is entirely reversible. In contrast, polymerase alpha appears to be an enzyme essential to the majority of biologically effective excision repair over the entire u.v. fluence range tested.