PhutochrmOfrv und Plrotohiulogy Vol. 44. N o . 5 , pp. 561-564. 1986 0031-8655186 $03.00+0.

00
Printcd in Grcat Britain. All rights reservcd Copyright 0 1986 Fcrgamon Journals Ltd

ENDOGENOUS GLUTATHIONE PROTECTS HUMAN SKIN FIBROBLASTS

AGAINST THE CYTOTOXIC ACTION OF UVB,
UI’A AND NEAR-VISIBLE RADIATIONS

Rex M. Tyrrell and Mireille Pidoux

Swiss Instritute for Experimental Cancer Research
1066 Epalinges/Lausanne, Switzerland

(Received 26 August 1986; accepted 2 September 1986)

Abstract--Both the UVI3 (290-320 nm) and UVA (320-380 nm) regions of sun-
light damage human skin cells but, particularly at the longer wavelengths,
information is scant concerning the mechanism(s) of damage induction and
the roles of cellular defense mechanisms. Following extensive glutathione
depletion of cultured human skin fibroblasts, the cells become strongly
sensitized to the cytotoxic action of near-visible (405 nm), UVA (334 nm,
365 nm) and UVB (313 nm) but not UVC (254 nm) radiations. I n the critical
UVB region, the magnitude of the protection afforded by endogenous gluta-
thione approaches that of the protection provided by excision repair. The
results suggest that a significant fraction of even UVB damage can be
mediated by free radical attack and that a major role of glutathione in
human skin cells is to protect them from the cytotoxic action of sunlight.

INTRODUCTION

The important wavelengths in sunlight for causing skin cancer in man are believed to lie in
the UVB (290-320 nm) region of the solar spectrum (Setlow, 1974). Radiation in the less
energetic UVA (320-380 nm) range may be involved in tumor promotion (Cerutti, 1985) and cer-
tainly makes a major contribution to the overall cytotoxic action of sunlight on human skin
cells (Keyse &., 1983; Smith and Paterson, 1982; unpublished observations, this labora-
tory). Radiation absorption by critical cellular macromolecules drops sharply with increase
in wavelength through the solar range and while photochemical alteration of DNA has been
implicated in UVB-induced lethal damage, our understanding of the molecular mechanisms
leading to cell death is minimal at lower photon energies. The oxygen dependence of cellular
inactivation by UVA and UVB in bacteria (Webb, 1977; Miguel and Tyrrell, 1982) and by UVA in
mammalian cells (Danpure and Tyrrell, 1976) implicates reactive oxygen intermediates in the
mechanism of cell death by these agents and raises the question as to whether cellular defense
mechanisms other than DNA repair are a significant factor in protecting cells against the
toxic effects of solar radiation. This report provides strong evidence in support of this
concept by showing that a large sector of the lethal damage to human skin fibroblasts that
results from irradiation in vitro by UVB as well as UVA and near-visible radiations is pre-
vented by normal cellular levels of glutathione, a ubiquitous intracellular thiol with
unique protective properties (Meister and Anderson, 1983).

MATERIALS AND METHODS

Diploid human fibroblasts (lBR biopsy 3 and XP 2161, Xeroderma pigmentosum complementation
group A) were kindly donated by the MRC Cell Mutation Unit, Brighton, and cultured by standard
techniques in Earle’s MEM with 15 percent foetal calf serum and cloned in 15 percent new born
calf serum with A-irradiated feeder cells. For experiments, 2-3 day old cultures were tryp-
sinised and resuspended in PBS for irradiation at 4°C. The irradiation source was a 2.5 k J
mercury-xenon lamp in combination with a Schoefel GM 250 monochromator. Details o f half
-band widths, fluence-rates, optical filtration and dosimetry appear elsewhere (Tyrrell
_
et _
al., 1985). Cells were depleted of glutathione by incubating them for 18 h in 500 1JM BSO
prior to trypsinisation. This treatment reduced cellular glutathione levels to approximately
5 percent as measured by the method of Tietze (1969) while leaving a surviving fraction of
0.67 t 0.052 ( 8 determinations). The mean of duplicate experiments is shown in Fig. 1.

For the experiments with cysteamine (Figs. 2 and 3 ) all aliquots were maintained in sus-
pension for a similar period and then centrifuged to remove the potentially toxic thiol com-
pound. Cells were then counted electronically before plating to determine clone-forming
efficiency. Under these conditions, cysteamine can be used at concentrations of up to 15 mM
without significant toxicity.

56 1
562 Rapid Communications

F L K N C E UVC ( 254 nm ) J m 2 FLUENCE UVB ( 3 1 3 nm ) Jm-2

I\ \

dJ
l + L z T T - z i d ' 2; Id' ;I 10; ;XI$ .Ix 105

FLUENCE UVA ( 334 nrn J M ~ FLUENCE UVA ( 365 nm ) Jm.2

Figure 1. Inactivation of normal human skin

fibroblasts (IBR/3) by radiation at 5 mono-
chromatic wavelengths. Cells were either
untreated prior to irradiation (X) or
totally depleted of endogenous glutathione by
exposure to BSO ( 0 ) . A, 254 nm; B, 313 nm;
C, 334 nm; D, 365 nm and E, 405 nm. Also
shown Fig. 1A and B are the sensitivities of
the excision deficient Xeroderma pigmentosum
10' cell strain XP1261 to radiation at 254 nm and
FLUENCE N E A R . VISIBLE ( 405 nm ) Jm.2 313 nm h).

RESULTS AND D I S C U S S I O N

Most of the agents commonly used to deplete cellular non-protein thiols are n o t specific
for glutathione and act neither by forming a covalent bond or by oxidising thiols. A
metabolic inhibitor, DL-buthionine-S, R-sulfoximine ( B S O ) , is now available which specifi-
cally prevents de novo synthesis of glutathione by preventing 1-glutamyl cysteine synthetase
activity (Griffith &., 1979). In contrast with the results obtained at 254 nm (Fig. lA),
depletion of endogenous glutathione strongly sensitises cell populations to radiation at all
solar wavelengths tested (Figs. 1B to IE). The ratio of final slopes ( 2 glutathione) of the
inactivation curves is close to 3 at 313 nm and somewhat higher at longer wavelengths (5.7 at
334 nm, 4.1 at 365 nm and 4.2 at 434 nm). These values indicate that between 67 (313 nm) and
8 2 (334 nm) percent of the lethal damage potentially induced by radiation at these wavelengths
in normal human cells is prevented or reversed by endogenous glutathione. Excision repair
removes DNA damage induced by radiation at 313 nm as efficiently as that induced by radiation
at 254 nm (Keyse &., 1983; Smith and Paterson, 1982; Niggli and Cerutti, 1983; compare
the relative sensitivities of the repair competent and the excision deficient strain in
Fig. 1A and 1B). However, the result in Fig. 1B indicates that after absorption of radiation
at 313 nm, a wavelength in the region of peak effectiveness for sunlight induced skin cancer
in humans, cells are protected from lethal damage by a major glutathione-dependent radical
scavenging mechanism to an almost as large an extent as by the highly efficient human excision
repair system.

The large sensitizing effect that is observed for non-ionizing radiation in the range 313
-405 nm (Fig. 1B-1E) may be compared with the relatively small response of cells treated
 Rapid Communications 563

' " 7 7
'OD r-----
FLUENCE UV (36inm) 5 6 '

Figure 2. Cysteamine protection of human Figure 3. Fluence dependent of cystearnine

fibroblasts irradiated at 365 nm as a
function of drug concentration. Cells were
either untreated prior to irradiation with a
fluence of 6x105 Jm-2 ( 0 ) or depleted of
endogenous glutathione (BSO, 500 , ! u 18 h)
prior to irradiation with a flucince of
Zx105 Jm-2 ( 0 ) .
protection against the lethal action of
radiation at 365 nm on human fibroblasts.
Fibroblasts were either untreated ( ~ , o )or
treated with 500 UM BSO for 18 h (+,n)
to
deplete glutathione prior to irradiation at
365 nm. The populations were irradiated
either with ( o , + ) or without ( ( l , A ) 10 mM
cysteamine. Experiments were carried out
between 2 and 4 times. Horizontal bars
depict standard errors or the limits of
duplicate experiments.

under aerobic conditions with low LET ionizing radiation in a similar state of BSO-mediated
et
glutathione depletion (Biaglow ._ -
al., 1983). Slope ratios for inactivation (t glutathione)
of both A549 (human lung carcinoma) and V79 (Chinese hamster) by ionizing radiation are of
the order of 1.3. Because of the relatively large protective effect of glutathione against
cytotoxicity induced by solar radiation under aerobic conditions, it is unlikely that
glutathione protects against these radiations by chemical reduction of radical damage in
critical macromolecules (i.e. "radical repair", Biaglow g &., 1983). More realistic
alternatives are either that glutathione quenches intermediate oxygen radicals directly (e.g.
it has a very high reaction rate with -OH,) or that organic hydroperoxides are formed which
can be detoxified by glutathione peroxidase. These two alternatives should be distinguished
by adding cysteamine exogenously to the glutathione depleted cells. This thiol is a
powerful radical scavenger but cannot be used as a hydrogen donor by the peroxidase.

The experiment depicted in Fig. 2 demonstrates that the presence of the thiol compound,
cysteamine, during irradiation protects only partially against the radiation-sensitizing
effects of glutathione depletion. A small cysteamine protection is also evident for cells
which have not been depleted of glutathione. Additional information is obtained from the
set of experiments depicted in Fig. 3 which shows the fluence dependence of protection at a
constant cysteamine concentration (10 mM). From these experiments it is evident that cystea-
mine does not marked1 protect glutathione-depleted cells against radiation at 365 nm at
fluences below 105.Jm-3 but that a significant protection occurs at higher fluences which
probably involves direct scavenging activity by cysteamine. The lack of cysteamine protec-
tion against the cytotoxic action of low fluences of UVA radiation in glutathione depleted
cells suggests that the glutathione plays a more specific role under these conditions, such
as acting as the hydrogen donor for glutathione peroxidase scavenging activity. A small but
significant protection of cells containing normal levels of glutathione is also observed at
high fluences so that, at least under these conditions, the reducing equivalents provided by
glutathione are inadequate to provide total protection against radical-mediated UVA damage.

Using a similar approach to that used in the experiments described in this report, it
should now be possible to determine whether human cells are protected against potentially
mutagenic damage by a glutathione-dependent pathway. Such a study is encouraged by the
indication from these results that even the erythemogenic and carcinogenic UVB region of
sunlight has the potential to induce a large sector of radical-mediated damage under condi-
tions where cellular Scavenging mechanisms are compromised. For the present, this study
provides the first evidence that a major role for endogenous glutathione in human skin cells
is to protect them from potentially lethal solar radiation damage.

Acknowledgements--The investigation was supported by grants from the Swiss National Science
Foundation (3.108.085) and the Swiss League against Cancer.
564 Rapid Communications

REFERENCES

Biaglow, U.E., M.E. Varnes, E.P. Clark and E.R. Epp (1983) The role of thiols in cellular
response to radiation and drugs. Radiation Res. 95, 437-455.

Cerutti, P.A. (1985) Prooxidant states and tumor promotion. Science 227, 375-381.

Danpure, H.J. and R.M. Tyrrell (1976) Oxygen-dependence of near-UV (365 nm) lethality and
the interaction of near-UV and X-rays in two mammalian cell lines. Photochem.
Photobiol. g,171-177.

Griffith, D.W., M.E. Anderson and A. Meister (1979) Inhibition of glutathione biosynthesis
by prothionine sulfoximine (S-n-propyl homocysteine sulfoximine), a selective inhibitor
of Y-glutamyl cysteine synthetase. J. Biol. Chem. 254, 1205-1210.

S.H. Moss and D.J.G. Davies (1983) Action spectra for inactivation of normal
K e y s e , S.M.,
and xeroderma pigmentosum human skin fibroblasts by ultraviolet radiations. Photochem.
Photobiol. 31, 307-312.

Meister, A . and M.E. Anderson (1983) Glutathione. Annu. Rev. Biochem. 2 , 711-750.
Miguel, A.G. and R.M. Tyrrell (1983) Induction of oxygen-dependent lethal damage by mono-
chromatic UVB (313 nm) radiation: strand breakage, repair and cell death. Carcino-
genesis 5 , 375-380.

sion in human skin fibroblasts after irradiation with 313 nm ultraviolet light.
chemistry 22, 1390-1395.
=-
Niggli, H . J . and P. Cerutti (1983) Cyclobutane-type pyrimidine photodimer formation and exci-

Setlow, R.B. (1974) The wavelengths in sunlight effective in producing skin cancer: a theo-
retical analysis. Proc. Natl. Acad. Sci. US. 71,3363-3366.

Smith, P.J. and M.C. Paterson (1982) Lethality and the induction and repair of DNA damage in
far-, mid- or near- UV-irradiated human fibroblasts: comparison of effects in normal,
Xeroderma pigmentosum and Bloom's syndrome cells. Photochem. Photobiol. 36, 333-344.

Tietze, F. (1969) Enzymic method for quantitative determination of nanogram amounts of

total and oxidised glutathione: Applications to mammalian blood and other tissues.
Anal. Biochem. 27, 502-522.

Tyrrell, R.M., P. Werfelli and E.C. Moraes (1984) Lethal action o f ultraviolet and visible
(blue-violet) radiations at defined wavelengths in human lymphoblastoid cells: action
spectra and interaction sites, m c h e m . Photobiol. 2, 183-190.

Webb, R.B. (1977) Lethal and mutagenic effects of near ultraviolet radiation in Photochem.
Photobiol. Reviews Vol. 2 (Edited by K.C. Smith) pp. 169-261, Plenum, New York.