252022 Kiev-22, Ukraine

A new method of biological dosimetry of UVB radiation based on an in vitro model of previtamin D photosynthesis is presented. The important role of vitamin D synthesis upon UV solar radiation is briefly reviewed. Distinctive features and extended capabilities of the method are discussed in detail with reference to the results of laboratory and field tests.

Keywords: UV radiation, biological dosimetry, vitamin D synthesis, spectrophotometric analysis, mathematical modeling.

Although excessive exposure to sunlight is liable to cause skin cancer and premature aging of the skin, obviously these acute and chronic effects only occur upon excessive exposure to UV radiation. This suggests that the intensity of biologically active UVB (280-315 nm) solar radiation at the earth surface needs permanent control.

Two complementary methods of measuring UV radiation are spectroradiometry and biological dosimetry [1]. Spectroradiometers provide the most detailed spectrally resolved information on solar UV in physical units, but they are expensive bulky instruments requiring regular calibration. To give a biological dose, the spectrum measured by a spectroradiometer is weighted (integrated over the wavelengths) according to the action spectrum of a specific biological effect.

An alternative approach involves biodosimeters that use the simplest biologic systems (bacteria, spores, biomolecules) and measure the integrated biological dose directly weighting UV radiation by the absorption spectrum of a molecule whose excitation is a basis for a photobiologic effect. Biodosimeters are cheap small devices requiring no external power and thus are widely accepted for both natural and artificial UV dose control. However, for the most part biodosimeters have a number of disadvantages given below:

1. The vast majority of biodosimeters indicate a damaging effect of UV radiation because UV absorption of a DNA molecule is in the basis of biologic response;
2. They are not able to provide an in situ control of UV dose because after UV exposure laboratory analysis of the dosimeter material is required to determine the degree of UV-initiated change;
3. In view of the lack of a mathematical model every biodosimeter acts as "black box", the UV dose is measured in specific biological units that presents a considerable challenge for quantitative evaluation of the dose in physical units (J/m²) and comparison of the data measured by different biodosimeters;
4. They are not spectrally selective; that is they have only one parameter responding to UV exposure (the rate of a biomaterial damage) and are not able to reveal ozone depletion under conditions when total UV intensity is reduced by air pollution, aerosols or dust.

the disadvantages mentioned above:

1. It is based on the photosynthesis of previtamin D and hence represents a beneficial (antirachitic) effect of UV radiation;

2. Spectrophotometric analysis of the provitamin D photoisomer mixture makes it possible to provide an in situ control of UV dose;
3. Availability of an adequate mathematical model (system of differential equations) provides the possibility for quantitative evaluation of the UV dose both in biological (the photoproduct concentrations) and physical units;
4. The method possesses spectral selectivity because along with the parameter responding to the integral UV intensity (the photoreaction rate), it has an additional parameter (maximum achievable concentration of previtamin D) which is exclusively sensitive to the spectral position of the short wavelength edge of solar spectrum. This makes it possible to reveal ozone depletion under conditions of polluted atmosphere.

Health promoting properties of sunlight have been recognized from the beginning of civilization although beneficial influence of solar UV radiation has been discovered by Danish doctor Niels Ryberg Finsen only by the turn of the 19th century. Vitamin D synthesis is the most well known and well documented beneficial effect of solar UVB radiation. There is a good reason to believe that synthesis of vitamin D initiated in mammal skin by the UV photons was essential for the formation and evolution of the vertebrate skeleton and just was primarily responsible for the vertebrates beginning on the Earth.

History of vitamin D discovery is intimately linked to the history of industrialization in Northern Europe when rickets (commonly known as "English disease") has been widely occurred as a consequence of air polluted sunless environment in industrial centers. Later on it was found that UVB irradiation of rachitic children results in remission of the disease rickets.

Now it is acknowledged that vitamin D, traditionally perceived as an antirachitic vitamin essential for regulation of calcium homeostasis, is responsible for a wide array of biological processes via mechanisms analogous to classical steroid hormones [3]. In general, there is an opinion [4] that effects of UVB sunlight and equivalent artificial light on physiological and behavioral processes are probably mediated, in large part, through the skin - vitamin D - endocrine system.

All higher vertebrates have an endogenous mechanism involving solar UVB radiation that initiates vitamin D synthesis. Studies in plants have also demonstrated that at least some plants, when exposed to UVB radiation, possessed antirachitic activity [5]. Holick et al. [6] evaluated the vitamin D synthetic capacity of zooplankton, amphibians, reptiles, avian species and aquatic mammals and found that they all had the capacity to produce vitamin D upon exposure to sunlight.

Vitamin D synthesis consists of two basic stages. At the first stage previtamin D is formed from initial provitamin D photochemically, and at the second stage it converts into vitamin D by thermoisomerization. Despite the terminology vitamin D is employed in the general sense, there are two chemical species of vitamin D. Vitamin D₂, or ergocalciferol (C₂₈H₄₄O), is synthesized upon UVB irradiation from ergosterol (provitamin D₂) mostly in plants. Vitamin D₃, or cholecalciferol (C₂₇H₄₄O) is photochemically produced in animals from 7-dehydrocholesterol (7-DHC, provitamin D₃). It is significant, that basic monomolecular isomerizations of the two steroid species occur in perfect analogy.

A steep decrease in provitamin D absorption spectrum within the interval 280-310 nm causes high sensitivity of vitamin D synthesis to the spectral composition of UVB radiation. Despite sunlight has long been recognized as the principal source of vitamin D₃ for humans, dramatic influence of seasonal and latitudinal changes of solar UVB radiation on vitamin D₃ synthesis (in vivo and in vitro) has been revealed [7]. From this fact high sensitivity of the process to ozone layer depletion can be deduced.

Another important factor affecting vitamin D synthesis is air pollution. Growing incidence of rickets in children has been observed in Mexico City in period 1990-1992 as a result of severe air pollution [8]. Positive correlation of breast cancer mortality with low sunlight levels and high air pollution levels was found in major U.S. cities, in Italy and in the former Soviet Union [9].

These data testify that the problem of the in situ control of solar UV radiation in respect to its ability to initiate vitamin D synthesis deserves much concentrated attention, and adoption of an in vitro model of previtamin D photosynthesis has great potential to attain these ends.

The photochemical stage of previtamin D photosynthesis is considerably complicated by the previtamin D side reversible and irreversible photoisomerizations (Fig.1). Since the absorption spectra of the photoisomers overlap in the same spectral region (Fig.2), UV irradiation of starting provitamin D produces the multicomponent mixture and is accompanied by the transformation of provitamin D absorption spectrum. This opens the way of UV dosimetry by the spectral monitoring in a manner like

Considerable gain in the accuracy and enhancement of capabilities of the method can be achieved in a "concentration mode" when the photoisomer concentrations are determined for the every exposure time. In principle, the dose dependence of each photoisomer can serve as a measure of the UV effect, but provitamin D decay and previtamin D accumulation are the most suitable. The decay of provitamin D is exponential in wide range of doses and has analogy to biological response of many biodosimeters in the form of some damage caused by UV radiation. The kinetics of previtamin D accumulation is directly associated with the biological endpoint of interest, e.g. vitamin D synthesis. This kinetics is more complicated (Fig.3b), it is linear up to C» 10%, then concentration of previtamin D reaches the maximum, and further slowly decreases due to the side photoconversions.

The biological dose is determined by measuring provitamin D and previtamin D concentrations depending on the exposure time. Method of high performance liquid chromatography (HPLC) is commonly used to perform concentration analysis, but it is quite expensive, time consuming and thus is not convenient for UV dosimetry. Considerable progress has been achieved with the development of spectrophotometric analysis whereby the absorption spectra are recorded by a spectrophotometer at different exposure times and then the photoisomer concentrations are derived from the spectra by computer processing [10]. In such a manner the in situ control of biological UV dose can be provided. (For this in mind elaboration of portable "VitaD" biodosimeter with a built-in computerized spectrophotometer is planned).

Although the second,- thermal,- stage of vitamin D formation is useless for UV dosimetry, the temperature dependence of the Previtamin D<=>Vitamin D conversion sets the limits of the exposure time when the thermoreaction can be neglected. It was found [11] that at 20C, in vitro, it took 20 hours to form 10% of vitamin D from previtamin D, and Pre<=>D equilibrium was established in 30 days. At 40C these times reduce to 2,3 hours and 3,5 days correspondingly. Consequently, in most cases using "Vitamin D" biodosimeter we can exclude the thermoreaction from consideration.

To date the photoreaction scheme of previtamin D photosynthesis is known in sufficient detail mainly due to comprehensive studies of E. Havinga and co-workers [12,13]. Representing a biological process, the reaction of previtamin D photosynthesis at the same time can be treated as a physical system with known parameters (the photoisomers absorption spectra and the quantum yields of partial photoreactions) that allowed development of adequate mathematical model [14,15].

When the action spectra of vitamin D synthesis is compared with the CIE erythema action spectrum [16] (Fig.4), it is apparent that the curves behaviour in the UVA spectral region differs dramatically.

"Vitamin D" biodosimeter has been put to the laboratory and field tests within the framework of the EC Project "BIODOS". We investigated the validity of the model by simultaneously measuring artificial and solar UV radiation using the spectroradiometer and "Vitamin D" biodosimeter. Here we will show and briefly discuss some results.

Several experiments were performed using narrow band filters and calibration facility at the Belgium Institute of Space Aeronomy and the required doses to get 10% of previtamin D have been determined (Table 1).

Laboratory tests with artificial UV sources were conducted at the Institute of Biophysics, Budapest (1996). Ethanol solutions of 7-DHC were irradiated in quartz cuvettes by the three UV lamps: (1) FS20, (2) TL01, and (3) Solar Simulator. The spectral irradiance of the lamps at the cuvette positions has been measured by spectroradiometer, and corresponding data are presented in Fig.5a) together with the provitamin D absorption spectrum. Figure 5b) shows associating concentrational kinetics of provitamin D decay and previtamin D accumulation. As is easy to see, not only the photoreaction rate differs essentially for the three lamps but maximum achievable concentrations of previtamin D are distinguishable as well. This last-named parameter affords spectral selectivity of the method (exceptional sensitivity to the spectral composition of UV radiation) [2].

Additional test has been done at the UMIST (Manchester) with the standard lamp BBH4 (1000W FEL, Osram Sylvania Ltd) which is commonly used for spectroradiometer calibration. Cuvette with ethanol solution of 7-DHC was irradiated at the standard distance of 50 cm. The absorption spectra were recorded by a Perkin-Elmer Lambda 15 UV/VIS spectrophotometer before and after several exposures, and were further processed by computer to derive the photoisomer concentrations. The photoreaction kinetics was calculated using the spectral data of the BBH4 lamp as input for the system of differential equations [14,15]. Close correspondence of the experimental and calculated data has been found (Fig.6a).

The same experiment was performed for a 25cm distance between the lamp and the cuvette to increase the doses without requiring excessively long exposure times. It is hardly surprising that a doubling of the distance between the lamp and cuvette led to a four fold increase of the exposure time to get the same concentration of previtamin D (Fig.6b)

First measurement of solar radiation with "Vitamin D" biodosimeter has been performed during the field intercomparison in Thessaloniki in July, 1997. The cuvettes were exposed during a day and the absorption spectra were recorded before exposure and every hour (30 min around noon) during the day. The photoreaction kinetics was calculated using the time-averaged spectroradiometer data. The corresponding experimental and calculated kinetic are presented in Fig.7 where the non-exponential character of provitamin D decay is readily apparent. The method reproducibility is demonstrated in Fig.7 by good coincidence of the results for the two cuvettes exposed in the same conditions.

In June, 1998 the solar UV measurement was conducted in Manchester. Seven identical cuvettes with ethanol solution of 7-DHC were exposed in the horizontal position close to the spectroradiometer diffuser. The absorption spectra were recorded every hour. During the time when the cuvettes were exposed, the solar spectrum was recorded every ten minutes by the spectroradiometer, a Bentham DTM300 double monochromator. Then the data obtained with the spectroradiometer were averaged over every hour and used as input for the mathematical model of the "Vitamin D" dosimeter.

Additionally, the minimal erythema doses (MED) were calculated from the spectroradiometer data for every one hour exposure (1 MED = 200 J/m2), and good correlation of experimentally measured concentrations of previtamin D with calculated MED was found (Fig.8). ). However, calculated concentrations of previtamin D were permanently lower than the experimentally measured. Two factors were suggested to be responsible for the discrepancy. The first one is the geometrical factor. The diffuser of the spectroradiometer has an angular response that is a reasonable approximation to a true cosine response, while this is far from the case for the "Vitamin D" biodosimeter. This difference in angular acceptance of the two devices is especially important when the zenith angle is large (high latitudes, early morning and evening time). A second, related, cause lies in the mathematical model where direct irradiation of the absorbing layer is suggested. This condition is fulfilled in the laboratory tests under lamp irradiation, but obviously was not met for solar irradiation in Manchester (contrary to clear midsummer day in Greece where more radiation was in the direct beam, and entered the cuvette at an angle closer to normal).

The results obtained confirm the ability of the "Vitamin D" biodosimeter to provide a reliable measure of the antirachitic UV solar radiation, although further work is needed to address the angular response of the dosimeters, and the model assumption of normal incidence. The method is sensitive enough to enable the UV dose measurement with a time resolution of 0.5-1 hour.

The availability of a mathematical model for previtamin D photosynthesis (from measured or modelled spectra) will allow prediction of profiles of antirachitic solar UV radiation over the globe.

Further studies are required to establish the correspondence between the in vitro "Vitamin D" dosimeter readings and previtamin D accumulation in vivo.

The current research of the author is supported by grant from the EC (ENV4-CT95-0044) and the Fellowship from the British Royal Society. The author is grateful to Dr. A. Webb, Dr. D. Bolsee and Dr. O. Galkin for cooperation in this investigation, and to Dr. G. Horneck, Dr. D. Gillotay, Dr. A. Bais and Prof. G. Ronto for helpful discussions.

1. A.R. Webb, Measuring UV radiation: a discussion of dosimeter properties, uses and limitations, J.

Photochem. Photobiol. B: Biology, 31 (1995) 9-13.

2. I.P.Terenetskaya, Provitamin D photoisomerization as possible UVB monitor: kinetic study using

tunable dye laser", SPIE Proceedings, 2134B "Ultraviolet Radiation Hazards", (1994) 135-140.

3. A.C. Mayar and A.W. Norman, Vitamin D, in: Encyclopedia of Human Biology, Academic Press, Inc.

v.7 (1991) 859-871.

4. W.E.Stumpf and T.H.Privette, Light, vitamin D and psychiatry, Psychopharmacology, 97 (1989) 285-

294.

5. M.F.Holick, Vitamin D₃ and sunlight: an intimate beneficial relationship, in M.F. Holick et al. (eds.),

Biologic Effects of Light, , W. de Gruyter, Berlin, New York, 1992, pp. 11-33.

6. M.F.Holick, S.A.Holick and R.L.Guillard, 1982. In: Comparative Endocrinology and Calcium

Regulation, (C. Oguro and P.K.T. Pang, eds.) Sci. Soc. Press, Tokyo, 85-91.

7. A.R.Webb, L.W.Kline and M.F.Holick, Influence of season and latitude on the cutaneous synthesis of

vitamin D₃: exposure to winter sunlight in Boston and Edmonton will not promote vitamin D₃ in

human skin, J. Clin. Endocrinol. Metabol., 67 (1988) 373-378.

8. I. Galindo, S. Frenk and H. Bravo, Ultraviolet irradiance over Mexico city, J. Air & Water Manage

Assoc. 45 (1995) 886-892.

9. a) F.C. Garland, C.F. Garland, et al. The positive association of breast cancer mortality with low sunlight

levels and high air pollution levels in major U.S. cities, in M.F. Holick et al. (eds.), Biologic Effects of

Light, , W. de Gruyter, Berlin, New York, 1992, pp.34-38.

b) C.F. Garland, F.C. Garland, et al., Sunlight, vitamin D, and mortality from breast and colorectal

cancer in Italy, ibid. pp.39-43.

c) E.C. Gorham, F.C. Garland and C.F. Garland, Geographic variation in breast cancer incidence and

sunlight in the USSR: the possible protective effects of vitamin D₃ and 25-hydroxyvitamin D₃, ibid.

pp.68-72.

10. I.P. Terenetskaya, N.A. Bogoslovsky, et al. Routes to optimization of previtamin D photosynthesis

using irradiation by a sunlamp, Pharm. Chem. J., 28 (1994) 589-596.

11. K.H. Hanewald, M.P. Rappoldt and J.R. Roborgh, The antirachitic activity of previtamin D₃, Rec.

Trav.Chim. 80 (1961) 1003-1014.

12. E. Havinga, Vitamin D, example and challenge, Experentia, 29 (1973) 1181-93.

13. H.J.C. Jacobs and E. Havinga, Photochemistry of vitamin D and its isomers and of simple trienes, in:

J.N. Pitts et al.(eds.), Advances in photochemistry, Wiley, New York, 11 (1979) 305-373.

efficiencies in the photochemistry of provitamin D using a tunable laser, Sov. J. Quantum Electron.,

21 (1991) 339-343.

16. Mc. Kinlay and A.F. Diffey, A reference action spectrum for ultraviolet induced erythema in human