Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Chapter 3 MOLECULAR AND CELLULAR STUDIES SUMMARY A decrease in stratospheric ozone has biological consequences that reflect two processes: (1) the increase in the intensity of W -B reaching the earth as a function of decreasing wavelength, and (2) the increase in bio- logical sensitivity with decreasing wavelength. The latter function is called an action spectrum. Knowledge of action spectra is important in evaluating the hazards of ozone depletion but is not sufficient for making quantitative predictions. In the past few years, extensive advances have been made in understanding the effects of single-wavelength (monochromatic) light on simple biological systems such as bacteria and cultured mammalian cells. It is now known that the biological effects of W -B are quite similar to those of W -C (with which almost all experiments in the literature have been done), and that DNA is the target for many deleterious effects of UV. The action spectrum for damage to DNA is well characterized. However, despite the similarity between the DNA action spectrum and the action spectra for killing, mutating, or transforming mammalian cells in culture, irradiation of such cellular systems with broad bands of radiation does not give results that would be predicted from the sum of the effects at discrete wavelengths. There are good indications that the longer wavelengths in W-B, or W-A, may modify the effects of UV-B. For example, in a process called photoreactivation, cellular enzymes use longer wavelength light energy to reverse the effects of shorter wavelength light. Longer wavelengths may delay the growth of cells, leaving more time for repair processes to act on the short-wavelength damage. Thus 47
48 since sunlight comprises a broad band of wavelengths, simple action spectra do not permit reliable predictions of the responses of biological systems to sunlight. One of the more important recent findings concerns the fate of photoproducts in skin cells of animals and humans. Photoproducts result from the absorption of photons (quanta of light energy) by molecules. Sensitive ways have been designed to measure the most important UV-C photoproducts, pyrimidine dimers in DNA, in intact skin irradiated by sunlamps. The presence of these dimers has severe biological consequences, including mutation and cell death. Studies have shown that there are two processes that remove dimers from cellular DNA in viva; one of these repair mechanisms acts in the dark (excision repair), and the other acts in the presence of longer wavelength UV and visible radiation (photoreactivation). Photoreactivation is highly specific for pyrimidine dimers. Preliminary estimates indicate that photo- reactivation is very rapid in humans and takes place to an appreciable extent even while human skin is being irradiated by the W-B in sunlight. Consequently, the level of pyrimidine dimers in cellular DNA in viva depends upon the relative intensities of the dimer-forming wave- lengths in UV-B and the dimer-splitting wavelengths in UV-A and visible radiation. This argument suggests that the effects of sunlight exposure on people might depend upon their exposure habits. For example, exposure in the afternoon might be much less deleterious than exposure in the morning. This is because the intensity of W-B in relation to that of UV-A and visible light increases in the morning and is greatest at noon. If exposure ended at noon, dimer formation would be at a peak relative to dimer splitting. If exposure occurred in the afternoon, more photoreactivating (dimer splitting) activity would occur as the relative intensity of W-A and visible light to UV-B increased. These findings underscore the importance of obtaining quantitative measures of life style and effective exposure. Both excision repair and photoreactivation have now been found to occur more slowly in mice than in humans; this fact must be taken into account in extrapolating data from rodents to humans. An added complication in attempting to estimate the effects of UV on, say, skin cancer induction is the effect of UV-B on the immune systems of animals and humans. preliminary action spectrum has been determined for W-induced immunosuppression in mice. If this spectrum A
49 were the correct one to use in estimating the carcinogenic effects of W on humans, the predicted effects of ozone depletion would be significantly smaller than those obtained by using an action spectrum determined for DNA alteration or for production of erythema in human skin. INTRODUCTI ON The studies described in this chapter aim at supplying the basic knowledge required to estimate the shapes of dose-response curves, to extrapolate from high to low dose rates, to determine the appropriate action spectra, and to assess synergistic and antagonistic phenomena in analyzing biological effects of ozone depletion. Without such data it is not possible to extrapolate from cellular studies to predict the effects on ecosystems and humans. Damage to DNA Many of the deleterious effects of W arise from the damage it does to DNA. Thus a great deal of effort has gone into understanding the biological effects of specific damages (photoproducts) in DNA. Early work in this field established definitively the shape of the action spectrum for damage to viral or bacterial DNA. It was supposed that this action spectrum was the appropriate one to use to calculate the biologically effective solar W dose for humans and the change in this dose as a result of ozone depletion (Setlow 1974). This Presumption was a big extrapolation. There were no reliable action spectra for W effects on mammalian cells, and there was a good possibility that the correct action spectrum might be different from that for bacterial DNA, because in higher organisms DNA is not a naked polymer but is closely associated with proteins. This association might change the action spectrum and might give rise to other deleterious photoproducts, for example, cross-links between protein and DNA. The effects of solar W on DNA are emphasized here because in much of the current work on carcinogenesis is assumed with reasonable assurance that DNA is an important target for initiating carcinogenic events. We enumerate in Appendix G a number of reasons for this presumption. It is important to realize, however, that other external or internal cellular factors, such as it
50 tumor promoters, hormones, or immunosuppressors, may be important in the development of cancer. In addition, UV could be both an initiator and a promoter, which would complicate the interpretation of dose-response curves. DNA Repair Although dosimetry for UV damage to cells is accurate and one can enumerate easily the DNA photoproducts existing immediately after irradiation (Setlow and Setlow 1972), the biological effects depend not only on the presence of such photoproducts but also on their lifetimes in cells. Most cells have repair mechanisms that remove the photo- products or permit cells to ignore them (Friedberg et al. 1979, Hanawalt et al. 1979, Setlow 1978). A great deal is known about two of these repair mechanisms, photo- reactivation and excision repair. Photoreactivation is a process by which visible light or W-A may reverse the effects of W-B or UV-C radiation. In this process, an enzyme binds to a DNA molecule containing pyrimidine dimers. The complex of enzyme and damaged DNA can absorb W-A or visible light, which causes the dimer to split, thereby repairing the damage. When the photoproducts are removed, the enzyme dissociates from the DNA. Photoreactivation is highly specific for pyrimidine dimers (Setlow and Setlow 1972), and when it is observed for a particular biological effect (such as survival, chromosome breaks, or mutation), it is taken as evidence that dimers caused that effect. Many cells, including normal human cells, contain repair systems such as excision repair that operate in the dark. In excision repair, products of W irradiation are removed from one strand of a DNA double helix by specific enzymes. The opposite, unaltered strand is then used as a template on which a new, unaltered strand is reconstituted. Excision repair is a very active process in normal human cells. Cells from individuals with a genetically inherited, sunlight-sensitive, cancer-prone disease called xeroderma pigmentosum are in almost all cases defective in excision repair. The high prevalence of skin cancer in such individuals is ascribed to the defect (Kraemer 1980). Mouse cells in culture also are defective in excision repair, and this defect must be taken into account in attempting to extrapolate from mice to humans.
51 ADVANCES IN KNOWLEDGE Transformation of Cells in Culture Transformation, an inheritable alteration of cells, can lead to cancer formation. W -C and W-B are able to transform mouse, hamster, and human cells in vitro, so that their growth on surfaces is no longer inhibited by contact with neighboring cells. The cells grow into piled-up clumps of cells instead of monolayers and can grow without being attached to a surface . (Chan and Little 1976, DiPaolo and Donovan 1976, Sutherland et al. 1980a). The fraction of transformed cells per surviving cell increases with dose. of W-transformed rodent cells are usually tumorigen~c when injected into certain mouse strains, but no tumori genicity has been shown for the W-transformed human cells described in the experimental results shown in Figure 3.1. Transformation of mouse and hamster cells is accomplished by single acute doses of W , but transforma tion of human cells thus far has only been effected by several small W doses (Sutherland et al. 1980a) or after a single dose under rather special conditions (Milo et al. 1981). - In numerous exper~ments, co~on~es Photoproducts in DNA Evidence that pyrimidine dimers are one of the major W-C photoproducts in cells of higher organisms comes from studies on photoreactivation. About 65 percent of the lethal damage to frog cells, which have high levels of photoreactivating activity, is photoreversible (Rosenstein and Setlow 1980). Since under experimental conditions only 80 percent of the pyrimidine dimers in the cells are returned to monomers, the results indicate that approximately 0.65/0.8 (~80 percent) of the lethal damage can be ascribed to pyrimidine dimers. In other systems with high levels of photoreactivating activity (frog cells and chicken embryonic fibroblasts), there is extensive photoreactivation of W -C-induced chromosome aberrations and sister-chromatic exchanges (Griggs and Bender 1973, Natarajan et al. 1980). Between 75 percent and 95 percent of the dimers are reduced to monomers, and the effective reduction in sister-chromatic exchanges was calculated to be between 65 percent and 80 percent, indicating again that a major fraction of this particular
52 0° _ ~ 2 _ O -1 ~ 10 LLI c, 5 He Is 2 L1J CL 1o2 I1J ~ 5 > 2 103 LLI ~5 ~\~: 2 _ Human Cells 0 Killing (Kantor et al. 1980) · Transformation (Sutherland et al. 1981 ) Chinese Hamster V79 Cells Killing (Rothman and Setlow 1979) ~ Dimers (Rothman and Setlow 1979) Syrian Hamster Cells O Kil I ing ( Doniger et al. 1981 ) · Dimers (Doniger et al. 1981 ) O Transformation ( Doniger et al. 1981 ) Mouse Cells X Mutation (Jacobson et al. 1981 ) Frog Cells V Killing (Rosenstein and Setlow 1980) Dimers (Rosenstein and Setlow 1980) DNA Absorption Spectrum (Sutherland and Griffin 1981) \~\ ~Action Spectrum for Lethal Effects in 1 \ Bacteria and Viruses (Setlow 1974) I\\ \ - 4 1 , 1 1 1 1 ' 10 250 270 290 310 WAVE LE NGTH ( n m ) FIGURE 3.1 Points on action spectra of a variety of higher organism cells (normal- ized to 1.00 at 265 nary). Also shown are the absorption spectrum of purified mam- malian DNA (solid line) and the action spectrum for lethal effects In bacteria and viruses (dashed line). type of chromosomal damage arises from pyrimidine dimers in DNA (Natarajan et al. 1980). The initiation by UV-C irradiation of the transformation of human cells has now also been found to be photoreversible (Sutherland et al. 1980a). That these conclusions may be extended to the UV-B region is strongly indicated for certain lesions. The killing of frog cells is photoreversed with equal effectiveness at each damage-producing wavelength tested between 252 nm and 313 nm (Rosenstein and Setlow 1980), and transformation of human cells by W -B is photo- reactivable (Sutherland et al. 1980a). W-C makes other photoproducts in DNA in addition to pyrimidine dimers, but these products have generally not been analyzed for their biological consequences. Examples
53 include other photoproducts of thymine, DNA-protein cross-links, and single-strand breaks (but of a different character than those made by ionizing radiation) (Erickson et al. 1980, Setlow and Setlow 1972). The ratio of other photoproducts to pyrimidine dimers appears to be a function of the wavelength of irradiation (Hariharan and Cerutti 1977). If other products were important, they might distort an action spectrum for affecting DNA from the shape of the action spectrum for dimer formation. The action spectrum for dimer formation itself is a complicated function of wavelength since all types of pyrimidine dimers may be formed, such as thymine-thymine, thymine-cytosine, and cytosine-cytosine. The ratio of cytosine-thymine to thymine-thymine dimers appears to increase with wavelength from 290 nm to 313 nm (Ellison and Childs 1981). It should be recognized that cell killing by W -A is produced mostly by mechanisms that are quite different from those produced by W-C, although the lethal lesion is still primarily damage to DNA. This is evidenced by the fact that bacterial mutants lacking DNA repair systems are very sensitive to W-A. Roughly 90 percent of the W-A killing requires oxygen (Webb 1977), whereas W-C killing does not. Further evidence that W-A damage is different is that the action spectrum for aerobic killing has a specific structure that suggests absorption of W by dyelike molecules (Webb 1977). The fraction of W-A killing that does not require oxygen (10 percent) may be due to direct production of damage in DNA, as evidenced by the fact that the action spectrum below 350 nm is without structure (Webb 1977) and similar to the absorption spectrum of DNA (Cabrera-Juarez et al. 1976, Peak and Tuveson 1979, Sutherland and Griffin 1981). Thus a small fraction of the lethal damage produced by W-A may be similar in mechanism to that produced by W -C. There are many indications that the mecnan~sms causing effects on DNA from W-B irradiation represent a mixture of the W-C and the W-A mechanisms, although the W-C mechanisms clearly predominate, as shown, for example, by the data of Figure 3.1. Thus, even though damage to DNA from W-B (the waveband critical in ozone depletion effects) is quite similar to that produced by W-C, it is not identical to it.
54 Action Spectra for Affecting Simple Cellular Systems Action spectra for killing a wide variety of higher organism cells in culture have been obtained for wave- lengths up to 313 nm. Action spectra have been obtained for killing human cells, mutating mouse cells, and bans forming hamster cells up to 313 nm and for transforming human cells up to 297 nm. The data points for a number of action spectra, normalized to a value of 1.00 at 265 nm, are shown in Figure 3.1. All the spectra that could be drawn for the various sets of points are very similar and coincide with the spectrum for forming pyrimidine dimers in the DNA of these cells. Figure 3.1 illustrates that (1) the action spectra for effects on mammalian and frog cells are very similar to those for effects on bacteria and viruses, (2) from 297 nm to 313 nm, the shapes of the spectra that could be drawn for the effects on cells of higher organisms are all approximately the same, and (3) there is a substantial deviation at long wavelengths between the values for all the action spectra and the absorption spectrum of purified mammalian DNA. At long wavelengths, the biological effects indicated in Figure 3.1 are smaller than would be predicted by the DNA absorption spectrum, probably because much of the W is absorbed in the purine (guanine) residues of DNA whereas the lethal photoproducts are primarily in the pyrimidine (cytosine and thymine) residues that absorb little at long wavelengths. Possible effects at wavelengths longer than 313 nm have not yet been determined in higher biological systems because the energies needed are higher than those achiev- able with the monochromatic UV sources used in past studies. The lethal responses of a large number of normal human cell strains have been examined at 254 nm and 313 nm (Smith and Paterson 1981), and the ratios of their sensitivities are close to those shown in Figure 3.1. A similar ratio is obtained for xeroderma pigmentosum cells (Smith and Paterson 1981), indicating that these repair-deficient human cells show equally enhanced sensitivity to 254 nm and to 313 nm. Thus for monochromatic radiation sources and the effects shown in Figure 3.1, almost all cells follow the DNA action--not absorption--spectrum. The primary conclusion drawn from the current understanding of action spectra is that all have similar shapes and hence the DNA action spectrum for mammalian cells may be taken to represent an average spectrum (not drawn in the figure).
55 Effects of UV on Photoproducts in Animal and Human Skin Sensitive enzymatic techniques have recently been developed to investigate the amounts of pyrimidine dimers in irradiated animal and human skin. These experiments indicate that excision repair (toy et al. 1977) and photoreactivation (toy et al. 1978) are negligible in mice, except for neonatal mice, in which there is a low level of photoreactivation (Ananthaswamy and Fisher 1980). For human skin irradiated with one minimal erythemal dose by a sunlamp, it is possible to measure the number of pyrimidine dimers immediately after irradiation. If incubation is continued in viva in the dark, there is an appreciable loss of dimers within 20 minutes, presumably by excision repair (D'Ambrosio et al. 1981a, Sutherland et al. 1980b). If instead the skin is illuminated with light from an incandescent lamp (predominantly visible wavelengths) for 20 minutes immediately after the sunlamp, there is more loss of dimers than from incubation in the dark. With even higher photoreactivating illumination, 80 percent of the dimers are lost in 4 minutes (D'Ambrosio et al. 1981b). These data indicate that normal human skin has both an active excision repair process and an active photoreacti- vation process, as was inferred from experiments on cells in culture (see the section below, "Mitigation and Enhancement of W -B Effects by Light at Other Wave- lengths"). The experiments imply that the illumination of human skin by sunlight results in a rather complex set of reactions for exposures that cover an appreciable period of time. W -B exposure makes pyrimidine dimers, but during the exposure dimers are being excised, and the W -A and visible components of sunlight are reversing the dimers by enzymatic photoreactivation. Thus in some . ~ · · ~ . ~ _ _ situations low, chronic UV exposures might have little effect, especially if the exposure continues into the later parts of the day when the W-B component of sunlight is relatively low and the photoreactivating (W -A) component is relatively high. Action Spectra for Immune Responses Irradiation of certain strains of mice with sunlamps ( W-A and UV-B) suppresses two immune responses, rejection of W-induced tumors (DeFabo and Kripke 1979,
56 Fisher and Kripke 1977) and contact hypersensitivity to the chemical trinitrochlorobenzene (TNCB) (see Chapter 5). Such systemic effects in mice irradiated at low doses (less than 1 minimal erythemal dose) could well be a contributing factor to the efficacy of W as a carcinogenic agent. The abolition of contact hypersensitivity to TNCB requires much lower doses than those needed for the lowered rejection of UV-induced tumors. This makes it experimentally possible to obtain an action spectrum for inhibition of contact hypersensitivity. Preliminary descriptions of this spectrum (De Fabo and Noonan 1980) indicate that it is in the UV-B region and falls off rapidly as wavelength increases above 290 nm. The action spectrum matches closely the absorption spectrum of several compounds known to be in mammalian skin, such as urocanic acid, and its values are appreciably greater than the absorption spectrum of DNA (Figure 3.1) at wavelengths greater than 290 nm. However, the target(s) (as yet unknown) for this effect may lie below the surface of the skin, and light absorption of skin is greater at shorter wavelengths. Hence, if this action spectrum were expressed in terms of quanta incident on the target as they are in Figure 3.1, rather than on the surface of animals, the values for the shorter wavelengths would be increased in relation to the values for the longer ones. As a result, the shape of the action spectrum for the depression of contact hypersensitivity would be closer to, but probably not identical to, the DNA absorption spectrum. If the suppression of these immune responses is important in UV carcinogenesis, and if their biological sensitivities at longer wavelengths are greater than that for damage to DNA, the effects of ozone depletion would be less than those computed for a DNA action spectrum. This is because it is the steepness of the action spectrum in the UV-B that makes ozone depletion important (see Figure 2.2). Experiments using broad-band sunlamps have shown suppression of contact hypersensitivity in mice. The shapes of the dose-response curves are similar to those of the narrow-band UV used to determine the action spectrum (Noonan et al. 1981a). It is important to determine how the effects of heterochromatic radiation on these immune responses compare quantitatively with the sum of the effects of monochromatic radiation.
57 Mitigation and Enhancement of W-B Effects by Light at Other Wavelengths Sunlight and UV-B in Bacterial Systems Much of what is known concerning photobiological effects in animal cells originates from studies of bacteria. Bacteria are very small, have large populations in a small volume, and have a very short generation time, permitting efficient study of rare events, such as mutation. Although bacteria are different in many ways from human cells, the fundamental biochemistry and genetics are analogous. It is generally true therefore that advances in bacterial photobiology have preceded and have pointed to advances in the photobiology of cells of higher organisms. It is now evident that UV-B acts on bacteria in much the same way as do shorter wavelengths (W-C), namely, through absorption of energy in DNA. Sunlight, however, contains much more of the longer wavelength W-A than UV-B. Although UV-A by itself can kill and mutate bacteria, it does this with only very low efficiency; the primary effect of the UV-A in full spectrum sunlight is a modification of the action of W-B. This modification may be either antagonistic or synergistic. Because of these modifications, it is clear that most actions of sunlight on biological systems cannot be understood from experimental work using monochromatic light alone. Among the known antagonistic processes are photoreacti- vation and photoprotection. Photoreactivation was defined earlier as a process in which UV-A or visible light eliminates the pyrimidine dimers produced by UV-B. The great excess of UV-A in sunlight suggests that much of the W-B damage will be repaired in the same exposure to sunlight that produced the damage (Webb 1977). Such effects have been observed in bacterial systems, which are normally killed (to a 10 percent survival level) by about 30 minutes of exposure to bright sunlight, but which have been shown to be more sensitive to killing by sunlight under conditions where photoreactivation is prevented (by low temperature or by using a system possessing defective photoreactivating enzyme). It is not easy to demonstrate such effects in animal cells irradiated with sunlight, but one may confidently expect that any animal tissue (such as human skin) that contains photoreactivating enzyme will in fact have some of its lethal damage repaired in this way (see the section
58 above, "Effects of W on Photoproducts in Animal and Human Skin"). Another way in which UV-A may decrease the damage caused by W -B is through the phenomenon of photoprotec- tion, that is, protection resulting from a preceding illumination with W-A (see the review by Webb (1977)). Photoprotection in bacteria may be induced by as little as 10 minutes of exposure to bright sunlight. It involves the induction by UV-A of a delay in growth, allowing more time after W -B irradiation is completed for error-free dark-repair systems to repair the damaged DNA (Tsai and Jagger 1981). In addition, the UV-A effective in Photoprotection may actively inhibit error-prone repair (Turner and Webb 1981). In wild-type Escherichia cold bacteria, for example, it has been shown that sunlight does not induce mutations. The UV-A wavelengths effective in reducing mutations are the same as those that delay growth, so this process may be a Photoprotection effect (Tyrrell 1980). Photoprotection and growth delay in E. cold are produced by the absorption of light by an unusual base in transfer RNA, 4-thiouracil (Tsai and Jagger 1981), but this base has not yet been shown to exist in the cells of higher animals. However, growth delay in Bacillus subtillis has been shown to be due to the absorption of light by quinones in the bacterial membrane (Taber et al. 1978). The process could also occur in the cells of higher organisms, although it is not yet tested experimentally. Synergistic effects have been observed in E. cold between W-A and W-B wavelengths, at high doses of W-A (Turner and Webb 1981, Tyrrell 1978, Webb et al. 1978). Some of these synergisms appear to be due to the destruc- tion of error-free DNA repair systems by the UV-A radia- tion. Error-prone recombination repair is responsible for mutation induced by all UV wavelengths in bacteria. At high W-A doses, the destruction of error-free repair systems results in an enhancement of mutation; at low doses, enhanced mutation is seen only in cells that are defective in error-free repair systems and thus are not capable of photoprotection. Such synergisms may operate in the sunlight induction of skin cancer in those humans whose skin is defective in error-free repair systems, such as those with excision-defective xeroderma pigmentosum (Maher and McCormick 1976). Finally, consideration must be given to UV-induced repair systems. In the W-C region, for example, W-C itself induces the error-prone repair system that is
59 responsible for most of the mutation produced by UV-C in bacteria (Witkin 1976). It has recently been discovered that UV-A will induce a repair system in bacteria that is capable of repairing damage caused by UV-A (Peters and Jagger 1981). It is not yet known if the W -A system will repair damage caused by UV-B, or if it is error-prone and would therefore produce mutation. Sunlight or Broad-Band Radiation and Mammalian Cells In mammalian cells, comparisons have been made between 254-nm radiation, sunlamp radiation in the range from 290 nm to 365 nm, sunlamps filtered to remove wavelengths below 300 nm, and sunlamps filtered to remove radiation below 310 nm. The results depend on the response being studied. For example, in Chinese hamster ovary cells, sister-chromatic exchanges are proportional to the amounts of pyrimidine dimers made by 254-nm radiation or by sunlamp radiation at wavelengths greater than 290 nm (Reynolds et al. 1979). However, the ratio of killing efficiency to dimer production, or mutation efficiency to dimer production, increases as the shorter wavelengths are removed from the radiation bands with mutation per dimer increasing more rapidly (Zelle et al. 1980). Similar results are obtained with Chinese hamster V79 cells, where the longest wavelength band used (greater than 310 nm) produces, in the time of a typical irradiation, no cell killing but appreciable mutation, and with mouse cells, where it produces very little killing but considerable transformation (Elkind et al. 1978, Suzuki et al. 1981). Thus it seems as if heterochromatic light in the longer wavelength regions of UV-B does not act as the sum of a series of monochromatic wavelengths. On the other hand, it is not clear which wavelengths are interacting to give the apparent synergistic effects for mutation and transformation. The interaction may arise between wave- lengths in the 310-nm to 315-nm region and longer wave- lengths such as WV-A. At present, the quantitative response to an enhancement in UV intensity in the region of 305 nm to 310 nm as a result of ozone depletion, with the other, longer UV wavelengths remaining constant, is not known. There is some evidence for the existence of UV-induced repair systems in mammalian cells (Bockstahler and Lytle 1977, Rommelaere et al. 1981). Experiments investigating
60 these mechanisms have used UV-C irradiation, and, in view of the shape of the average mammalian action spectrum in Figure 3.1, one would expect similar findings for UV-B radiation. The evidence for the existence of W-induced repair is the ability of irradiated cells to reactivate W-irradiated viruses that are used to infect the cells. Irradiation of cells before virus infection results in an enhanced survival of the infecting W-irradiated viruses, and in some instances a higher frequency of mutations is observed in irradiated viruses infecting irradiated cells (Des Gupta and Summers 1978). The magnitudes of the observed effects are small, and the extrapolation of such data from effects on viruses to effects on the cells themselves has not been made. RESEARCH RECOMMENDATIONS The following list of unclarified issues is intended as a guide for future research. The list is not exhaustive. It has been limited to those issues that should receive attention first, but it is not organized according to priority. 1. It would be useful to know the shape of the action spectrum for affecting so-called DNA functions of mammalian cells at wavelengths greater than 313 nm. 2. An understanding is needed of why broad bands of UV (heterochromatic radiation) do not seem to act like a simple sum of monochromatic wavelengths in terms of their effects on DNA. Studies of synergistic effects between W-A and W-B (for example, in bacteria) are fundamental to understanding the mechanisms of cancer induction by sunlight. 3. The quantitative aspects of the immune response of mice to monochromatic wavelengths versus the response to broad bands of W -B should be explored. The molecular and cellular mechanisms for immune system effects and wavelength dependence should be investigated. 4. An understanding of the mechanism of neoplastic transformation by UV in vitro is needed. In some rodent systems, the level of transformation is so high--close to 100 percent--that this transformation process looks suspiciously like a triggering mechanism that controls the regulation of cell growth rather than like an effect on a specific gene or genes (Kennedy et al. 1980).
61 5. The characteristics of W-A-induced repair systems in bacteria need to be determined. The possibility of the existence of such systems in mammalian cells should be examined. 6. Data are needed on the rates of repair, in the dark and in the light, of UV-irradiated human skin cells as a function of UV dose. The differences, if any, between acute and chronic irradiations should be deter- mined. One might be able to study the responses of individuals who are exposed to high levels of W-B as part of therapy for psoriasis. The aim of such studies would be to determine whether the kinetics of dark repair of damage from dimers in human skin show two components, a slow and a fast one, as is true for human cells irradi- ated in vitro. (The fast component represents repair of DNA in the so-called linker regions of chromatic, and the slow reaction is the repair in the core regions of chromatin (Cleaver 1977, Smerdon et al. 1978). The latter is not as readily accessible to enzymes as is the former.) Equally important questions are, what other types of biologically important damages occur in skin, what are their lifetimes, and are any of them persistent?