ARTICLE
It is now commonly recognised that UVA as well as UVB irradiation is involved
in the development of cutaneous cancers. This relation to skin exposure
is particularly clear for squamous cell carcinomas (SCCs) and basal cell
carcinomas (BCCs) in Caucasian people [1]. Furthermore, the cutaneous SCC
and BCC carcinogenesis action spectrum in the albino hairless mouse established
by de Gruijl [2] shows two maximum efficiency wavelengths: the main in the
UVB and a second in the UVA region. The relative efficiency of the second
maximum is approximately 10,000 times lower than that of the first but it
is increased by the higher doses of UVA received. UVA represents 98% of
the total UV received at the Earth surface so that it can be considered
that the difference in effectiveness between UVB and A is only of 100 times.
For these reasons, the cosmetics industry has had to improve UVA sunscreen
protection (indeed, the first generation of sunscreens did not provide any
protection against UVA).
Immediate and persistent pigment darkening (IPD and PPD), aspects of
the in vivo macroscopic evidence of UVA exposure, were chosen to
determine UVA protection factors [4, 5]. Indeed the other in vivo
methods available have important disadvantages. For example, the method
based on UVA erythema induction requires psoralen [3] to be taken by the
volunteer and thus is not ethically acceptable because these compounds
are known to interact with DNA. Furthermore, protection factors obtained
from artificially sensitised skin are probably not relevant to those determined
with normal skin. The pigmentation method is a reproducible and useful
tool for routine assessment of UVA protection [4, 5]. It should be noted
that there is no consensus on the use of this (or any other) method to
determine protection factors in UVA.
Very little information is available about the nature and mechanism
of immediate pigment darkening, also called Meirowsky phenomena. Immediate
pigment darkening is a transitory darkening of the skin observed after
UVA exposure. Most authors agree on four points [6]. It involves the melanocytic
system: melanin precursors and/or melanin. It is often thought, with no
clear evidence, that redox reactions occur with these compounds. IPD is
UVA induced, partly reversible after light turn off and oxygen-dependent.
After light turn off, the coloration of the skin fades rapidly over 2
hrs, then more slowly during 24 hrs on average. In this second phase,
the coloration of the skin is called Persistent Pigment Darkening (PPD).
It appeared important to us to reproduce in vitro this phenomenon.
From a practical point of view, we could expect to use this in vitro
model to quantify the protection afforded by commercial sunscreens. This
in vitro "protection parameter" could be used either as an alternative
to human experiment (which often causes ethical problems) or as a predictive
value of in vivo protection prior to human experiments. Indeed,
few in vitro methods are now available for determining the UVA
protection factor. The best known and most widely used is the Diffey-Robson
calculation method [7]. But, it is quite unsuitable for it uses the action
spectrum for erythema induction in human whereas in vivo experiments
use the Meirowsky phenomena action spectrum.
In order to set up an in vitro model of immediate pigment darkening,
we investigated the evolution of different mixtures of melanocytic compounds
under irradiation. We mainly worked with melanin precursors rather than
melanin: melanin is almost insoluble and the reversibility of IPD suggests
that it takes place from precursors. Furthermore, during the past 10 years,
results reported concerning the biological role of the melanocytic system
suggest that it is related to the formation of soluble, labile and reactive
melanin precursors. For example, it has been shown that some of them (in
particular 5,6 dihydroxyindole-2-carboxylic acid (DICA) and 5-S-cysteinylDOPA)
have an important antioxidant activity [8-12]. DICA has also been found
to play a messenger role in NO synthesis [13]. Tyrosinase, the enzyme
responsible for the formation of melanin, is able to use superoxide anion
for its synthesis task thus eliminating this dangerous species [14-17].
In fact, the melanin synthesis machinery could be as protective as its
final product.
L-DOPA, a eumelanin and pheomelanin precursor, has been chosen for its
solubility in water and its stability. This compound does not absorb in
the UVA region so we had to introduce a UVA absorbing compound. We chose
pheomelanins for their ubiquitous presence in the melanocytic system and
their solubility. Furthermore, their UVA irradiation induces the formation
of semiquinonic radicals which are expected to play a primer role in our
model [18].
We examined spectral variations of solutions of DOPA and pheomelanin
occurring during UVA irradiation.
Then, we tried to develop a method to determine a UVA protecting parameter
(Puva) related to the model created. The criteria for determining
the Puva was the ability of sunscreen to inhibit the model
IPD reaction in vitro.
Materials and method
Chemicals: DOPA, mushroom tyrosinase, mannitol, cystein, Na2HPO4,
Na2EDTA, K3FeCN6 were purchased from
Sigma.
Solutions: All tested solutions were made in phosphate buffer (4 mM,
pH = 7).
Preparation of pheomelanins: Synthesis of pheomelanin was performed
according to the method described by Ito [19] using DOPA, cystein and
mushroom tyrosinase.
Irradiation source and dosimetry: The irradiation source is a Muller
Xe 150 W lamp equipped with a WG 335 filter (thickness 3 mm) and a water
filter. The UVA output was about 12 mW/cm2 measured with a
radiometer Oriel 70380.
Model solutions: The solution of DOPA (4.5 mM) and pheomelanin (1.6
x 10- 2 mg/ml) in phosphate buffer was UVA-irradiated
using the apparatus described above. Absorbance variations of the solutions
were monitored using a Hewlett Packard (HP 8452) UV-Vis spectrophotometer.
Synthesis of DOPAchrome: This compound was obtained by the addition
of two equivalents of sodium periodate on one equivalent of DOPA according
to the procedure described by Kagedal [20].
Detection of DOPAchrome: Detection of DOPAchrome in irradiated solutions
of DOPA and pheomelanin was performed by HPLC using a Bischoff Ultrasep
column (C18, 6 mum, 150 x 4 mm) and UV-Vis detector (lambda = 300 nm).
The mobile phase was composed of 83% phosphoric buffer (pH = 4.30 mM)
and 17% of methanol.
Preparation of sunscreen samples: Sunscreen was sprayed on a quartz
cuvet at a concentration of 2 mg/cm2. Then the cuvet was filled
with 3 ml of the model solution and irradiated with UVA (the sunscreen
facing the lamp output).
Results
Spectroscopic studies of mixed solutions of DOPA
and pheomelanins under irradiation
DOPA (4.5 mM) and pheomelanin (1.6 x 10- 2 mg/ml, 1.1
muM considering the molecular weight to be to 600 g/mol) solutions in
phosphate buffer (pH = 7, 4 mM) were irradiated by UVA using a 150 W xenon
lamp equipped with WG335 (3 mm thickness) and a water filter. The absorbance
of the solutions in the UV-Vis region was followed during the irradiation.
Figure 1 shows the absorption
spectrum of such a solution before and after its exposure to UVA (about
45 J/cm2). For all wavelengths, an increase in solution absorbance
occurred with a maximum amplitude at 480 nm. When the irradiation was
stopped, the absorbance decreased (Fig.
2). The decrease reached 8% at 340 nm and 20% at 480 nm (average)
one hour after the light was turned off. Thus, the phenomenon was partially
reversible. For longer observation times, black melanin-like polymers
were formed in the solutions. Indeed, after several hours in solution
(even in the absence of irradiation), DOPA is unstable and leads to the
formation of auto-oxidation melanin-like polymer [23]. As a control, a
solution of DOPA alone was treated as previously described (without irradiation
UVA) to ensure that the auto-oxidation process does not participate in
the reaction observed in the presence of pheomelanin.
Figure 3 shows the absorbance
variations at 480 nm as a function of the irradiation time for aerated
(atmospheric conditions) and deaerated (argon bubbling) solutions of DOPA
(4.5 mM) and pheomelanin (1.6 x 10- 2 mg/ml). The absorbance
of the deaerated solution varied much less than that of the aerated one.
When oxygen was bubbled through the solution (Fig.
3), no increase in absorbance variations was noted. Addition of a
free radical scavenger, mannitol (0.37 mM), induced a 50% decrease in
the absorbance variations at 480 nm (Fig.
4). The same phenomenon was observed when EDTA (70 muM), a metal chelator,
was added (Fig. 4).
The DOPA and pheomelanin solutions irradiated as described were analysed
by HPLC. A compound which absorbs at 480 nm was separated from DOPA (Fig.
5). Pure DOPAchrome, obtained as described by Kagedal et al.
[20], allowed us to identify the compound absorbing at 480 nm as DOPAchrome.
We showed that the UVA irradiation of DOPA and pheomelanin solutions
leads to an increase in their absorbance with a maximum amplitude at 480
nm. This phenomenon is partially reversible and oxygen-dependent. The
mechanism of the reaction is probably a radical one because it is inhibited
by mannitol and EDTA. We can propose a two-step mechanism. In the first
step, the irradiation of pheomelanin leads to the formation of semiquinonic
radicals, already observed in the skin [21]. In the second step, these
radicals induce oxidation of DOPA probably with the involvement of oxygen
reactive species as intermediates suggested by the oxygen-dependence of
the reaction in vitro and its inhibition by EDTA (chelating iron
inhibit Fenton type reaction which lead to reactive oxygen species). The
oxidation of DOPA has been studied by pulsed radiolysis [22, 23] and found
to proceed in several steps: DOPA is first oxidised to DOPAsemiquinone
and then to DOPAquinone. This compound undergoes rearrangement and leads
to the formation of DOPAchrome which was detected in our solutions by
HLPC. Its molar extinction coefficient is 3,700 cm- 1M-
1 at 475 nm [20]. Thus, the average concentration of DOPAchrome
in model solutions after irradiation (90 min) was estimated at 30 muM.
After turning off the light, the absorbance of model solutions decreased
suggesting that the formation of DOPAchrome becomes less than its oxidation
to melanin-like polymer. Indeed, the molar extinction coefficient of the
polymer is lower than that of DOPAchrome at 480 nm [20].
Thus, the irradiation of DOPA and pheomelanin solutions in these conditions
satisfactorily mimics the effects of a UVA irradiation on the skin as
far as IPD is concerned. They are good candidates for an in vitro
model of immediate pigment darkening. Furthermore the compounds used are
soluble and present in melanocytes and keratinocytes of Caucasian subjects.
Interaction between DOPA and pheomelanins could occur in vivo as
described to occur in vitro.
New method to quantify UVA photoprotection
Definition
We used the inhibition of the model reaction described above as an indicator
of the filtering efficiency of the sun-screen tested. In order to increase
the absorbance variations observed on irradiated model solutions, K3FeCN6
(45 muM) was added to the solution. This compound is expected to play
a double role: promoting radical formation by supplying iron (via Fenton
processes) and favouring the formation of DOPAchrome by complexing DOPA.
The sunscreen to be tested was placed on a quartz surface (2 mg/cm2)
between the irradiation source and the model solution. The absorbance
of the solution at 480 nm was measured before and after irradiation. Its
variation was used to define the level of protection afforded by the sunscreen.
We calibrated the response of the reaction to filtering inhibition using
some neutral* attenuators filtering 10 to 70% of the incident light (*they
do not modify the emission spectrum of the lamp). They were placed between
the model solution and the lamp. The absorbance of the solutions was measured
before and after the irradiation. Its variation varied linearly with the
percentage of light cut off by the attenuator (Fig.
6) giving a calibration curve. Measuring the absorbance variation
for sunscreen samples allowed us to determine the percentage absorbance
of a beam attenuator which would have the same inhibitory effect as the
tested sunscreen as shown in figure
7. The parameter obtained was called Puva (for parameter
in UVA) and is in an indicator of the filtering efficiency of the sunscreen.
Puva was determined for 10 sunscreens (Table
I). The values of Puva varied from 37 to 72. The method
developed is reproducible with a low standard deviation of 6.2%.
Another possibility was to define the protection parameter in the same
way as the in vivo parameter is defined i.e. using pigmentation
method [4, 5]. The evolution of the model solution absorbance at 480 nm
was followed during irradiation to determine the curve delta(OD480nm)
versus time (time is proportional to the dose received). This must be
done for two samples: one protected by sunscreen and the other not. Thus,
we obtained two curves (Fig.
9). The in vivo protection factor is defined as the ratio of
the irradiation dose necessary to obtain the same variation of skin pigmentation
on a protected skin area (with sunscreen) to that on an unprotected skin
area (without sunscreen). So, we defined P* calculated as below:
P* = (Dp/Dnp)
Where Dp is the dose necessary to obtain an absorbance variation
(DELTAOD480nm) of 0.2 for the protected model solution and
Dnp is the dose necessary to obtain an absorbance variation
of 0.2 for the unprotected model solution. The value of 0.2 was chosen
because it is the largest absorbance variation we could obtain without
any artefact due to pigment polymerisation.
The measurement of P* appeared to be simpler and faster than that of
Puva because only two samples are needed to determine the protection
parameter whereas 5 were necessary for Puva determination (4
for calibration and one for the tested sample).
But, numerical values of P* (2 to 10) are much lower than Puva
values (1 to 70) or to in vivo values (2 to 35). This scale difference
could increase the difficulty in making a parallel with traditionally
determined protection factors (in vivo or in vitro). Thus,
the following experiments were made with the first method described here.
Correlation with classical methods of determination
of protecting factors
Correlation Puva/IPuva
The method described by Diffey-Robson [7] is an in vitro spectroscopic
method. The absorption spectrum of the sunscreen is measured on a glass
slide and combined with the action spectrum for erythema induction in
humans and the emission spectrum of the sun (40° latitude North).
In Europe, the protective factor obtained for the UVA region is called
IPuva (indice de protection UVA). IPuva and Puva
were determined for 10 sunscreens (Table
I). When the IPuva increased, the Puva increased.
The linear correlation coefficient was 0.86 (Fig.
8). This value is quite low but it should be taken in consideration
that standard deviation for IPuva determination is about 20%.
Furthermore different action spectrums are involved in the two methods.
The Diffey-Robson [7] method involves: the action spectrum for erythema
induction in man, the absorption spectrum of the tested compound, the
emission spectrum of the sun. However, the method described here involves:
the action spectrum for model reaction induction, the absorption spectrum
of the tested compound, the emission spectrum of the lamp.
Correlation SPF/Puva
Sun Protecting Factor (SPF) characterises the ability of a sunscreen
preparation to inhibit the appearance of sunburn under irradiation UVA
and B. The SPF was determined for the tested preparations according to
the classical in vitro Diffey-Robson method [7] (this value is
well correlated to the in vivo experiments). The figure
10 shows Puva as a function of SPF for the tested compounds.
Surprisingly, a linear variation could be extrapolated. Puva
and SPF increased simultaneously. The linear correlation coefficient was
low: 0.77. This value is lower than that obtained for the previous correlation.
This is in agreement with the obviously different nature of the phenomenon
involved. Indeed, there is no reason for the existence of a simple correlation:
SPF is a UVA + B erythemal parameter whereas Puva is a UVA
biochemical parameter. Thus the correlation IPa-Puva
seems to be much more pertinent, Puva been defined in UVA only.
About the correlations: Both the correlations gave a linear tendency
with a low correlation coefficient, Puva increased when SPF
and/or IPuva increased. This is due to the fact that different
processes and action spectrum are involved. However this result is quite
satisfactory because it show that our parameter can be effectively linked
to existent parameters and that its use as a routine tool in the quantification
of the efficiency of new products is completely justified.
Discussion
We succeeded in building an in vitro model which reproduces the
main characteristics of immediate pigment darkening in vivo: UVA
induction, reversibility, oxygen-dependence, increase in absorbance. This
was done using melanocytic compounds DOPA and pheomelanin.
The involvement of melanin precursor in immediate pigment darkening,
(suggested by the model build here and by other authors [9]) could explain
the difficulties encountered in predicting the solar sensitivity of Caucasian
subjects (erythemal sensitivity as well as pigmenting capability). Actually,
some classifications of Caucasians are based on skin colour though melanin
precursors are colourless.
On the basis of this model (pheomelanin catalysed), a UVA protecting
factor has been developed (Puva and P*). in vivo protection
factors are defined as the ability to inhibit erythema in the case of
UVA + B irradiation and immediate pigment darkening (in most cases) in
the case of UVA irradiation. Erythema was chosen as an end point because
it is the most evident and visual deleterious effect of a solar irradiation
on skin. But some authors argue that it could be not representative of
UV cutaneous damage for long term modification such as aging and cancer.
They propose different end points such as sunburn cells, global DNA damage
(comet assay), P53... As far as UVA is concerned, IPD is the endpoint
frequently used but its biological role is still misunderstood: is it
the scar of a deleterious effect or a defence reaction of the skin? Whatever
the answer, it is a visually evident trace of UVA irradiation on the skin.
Our model tries to mimic this response. The protection parameter defined
here is obviously not an absolute nor global evaluation of the protection
against real biological damage induced on skin. But, it allows the definition
of a novel scale with some advantages and drawbacks. The major advantages
are:
- It is an in vitro method thus useful for rapid estimation
of the protection ability of new products in development.
- It is an in vitro method linked to biological events in
the skin. So it can be used as an alternative method to human experiments
for very high protection factors which require very long exposure time
in vivo.
- It enables quantification of UVA photoprotection to be proposed
for very light skin which is not measurable in vivo for technical
and ethical reasons. Indeed, fair skinned people do not exhibit any immediate
pigment darkening.
The essential drawback is the in vitro nature of the method,
one can obviously claim it is only a model restricted to one reaction.
It does not represent the overall events occurring in the skin under a
UVA irradiation.
Article accepted on 4/6/02
REFERENCES
1. International agency for research on cancer. IARC Monographs
on the evaluation of carcinogenic risks to humans: solar and UV radiation.
Lyon: IARC, 1992; 55.
2. De Gruijl FR, Sterenborg HJ, Forbes PD, et al. Wavelength
dependence of skin cancer induction in albino hairless mice. Cancer
Res. 1993; 53: 53-60.
3. Lowe NJ, Kingston TP, Weingarten DP. Evaluation of UVA sunscreen
on subjets photosensitized with topical methoxsalen. Br J Dermatol
1985; 113: 772.
4. Moyal D, Chardon A, Kollias N. Determination of UVA protection
factors using the persistant pigment darkening (PPD) as the end point.
Part 1: calibration of the method. Photodermatol Photoimmunol Photomed
2000; 16: 245-9.
5. Moyal D, Chardon A, Kollias N. UVA protection efficacy of
sunscreens can be determined by the persistant pigment darkening (PPD)
method. Part 2. Photodermatol Photoimmunol Photomed 2000; 16: 250-5.
6. Routaboul C, Denis A, Vinche A. Immediate pigment darkening:
description, kinetic and biological function. Eur J Dermatol 1999;
9: 95-9.
7. Diffey BL, Robson J. A new substrate to measure sunscreen
protection factors throughout the ultraviolet spectrum. J Soc Cosmet
Chem 1989; 40: 127-33.
8. Borovansky J. Free radical activity of melanin and related
substances: biochemical and pathobiochemical pathway. J Sbornik Lekarky
1996; 97: 49-70.
9. Prota G, Misucarra G. Melanin-related metabolites: a new look
at their functional role in melanogenesis and malignant melanoma: biochemistry,
cell biology, molecular biology, pathophysiology, diagnosis and treatment.
In: Hori Y, Hearing VJ, Nakayama J, eds, 1996: 49-61.
10. Pavel S, Smit N. Detoxification processes in pigment producing
cell. In : Melanogenesis and malignant melanoma: biochemistry, cell biology,
molecular biology, pathophysiology, diagnosis and treatment. In: Hori
Y, Hearing VJ, Nakayama J, eds, 1996: 161-8.
11. Napolitano A, Menoli S, Nappi A, d'Ischia M, Prota G. 5 s
cysteinyl DOPA, a diffusible product of melanocyte activity, is an efficient
inhibitor of hydroxylation/oxidation reactions induced by the Fenton system.
Biochim Biophys Acta 1996; 1291: 75-82.
12. Memoli S, Napolitano A, d'Ischia M, Misucara G, Palumbo A,
Prota G. Diffusible melanin related metabolites are potent inhibitors
of lipid peroxidation. Biochim Biophys Acta 1997; 1346: 61-8.
13. Roméro-Graillet C, Aberdam E, Clément M, Ortonne
JP, Balloti R. Nitric oxide produced by UV irradiated keratinocytes stimulates
melanogenesis. J Clin Invest 1997; 99: 635-42.
14. Valverde P, Manning P, McNeil CJ, Thody AJ. Tyrosinase may
protect human melanocytes from cytotoxic effects of the superoxide anion.
Exp Dermatol 1996; 5: 247-53.
15. Valverde P, Manning P, McNeil C, Thody J. Activation of tyrosinase
reduces the cytotoxic effect of superoxide anion in B16 mouse melanoma
cells. Pig Cell Res 1996; 9: 77-84.
16. Wood JM, Schallreuter KU. Studies on the reactions between
human tyrosinase, superoxide anion, hydrogen peroxide and thiols. Biochim
Biophys Acta 1991; 1074: 378-85.
17. Karg E, Odh G, Wittbjer A, Rosengren E, Rorsman H. Hydrogen
peroxide as an inducer of elevated tyrosinase level in melanomas cells.
J Invest Dermatol 1993; 100: 209s-13.
18. Sealy RC, Hyde JS, Felix CC, Menon IA, Prota G. Eumelanins
and pheomelanins characterization by ESR spectroscopy. Science
1982; 217: 545-7.
19. Ito S. Optimisation of conditions for preparing synthetic
pheomelanin. Pigment Cell Res 1989; 2: 53-6.
20. Kagedal B, Konradsson P, Shibata T, Mishima Y. Hight performance
liquid chromatographic analysis of DOPAchrome and dihydroxyphenylalanine.
Anal Biochem 1995; 225: 264-9.
21. Sawamura D, Sato S, Kiuchi H, Hashimoto I, Katabira Y. UVA
induced darkening of lower epidermal cells as an in vitro system
of immediate pigment darkening and mechanism of IPD. J Dermatol
1986; 13: 101-7.
22. Land EJ. Pulsed irradiation studies of some reactive intermediates
of melanogenesis. Rev Chem Int 1988; 10: 219-40.
23. Bensasson RV, Land EJ, Truscott TG. Melanogenesis and melanoma.
In: Excited states and free radicals in biology and medecine: contributions
from flash photolysis and pulsed radiolysis. Oxford Science Publication,
1993: 228-59.
|
Version imprimable
