THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Copyright© 1970 by The Williams & Wilkins Co. Vol. 54, No. 2

Printed in U.S.A .

TISSUE CULTURE TECHNIQUES IN THE STUDY OF CELL PHOTOBIOLOGY AND PHOTOTOXICITY*

ROBERT G. FREEMAN, M.D., WENDY MURTISHAW AND JOHN M. KNOX, M.D.

ABSTRACT

Tissue culture monolayers exposed to 13 photosensitizing compounds demonstrated a phototoxic effect resulting in death of the cells when exposed to either broad spectrum or monochromatic ultraviolet or visible light of appropriate wavelength. Drug alone, light alone, and 2 non-photosensitizing chemicals with the light induced no such effect. The monolayer was damaged in six hours and destroyed within 24 hours. With chloroquin and quinacrine the reaction was delayed until 48 hours. The technique is sensitive to -methoxypsoralen in concentrations as dilute as 1:100,000. It provide~ a versatile and sensitive sy tern for photobiologic experiments.

ingl celled organisms were the first biologi­cal sy tern utilized in the study of photosensi­tization when Haab (1) observed that liO'ht killed protozoa placed in acridine dye solution. Obser­vations using such simple bioloO'ical systems un­d r a variety of conditions have aided in our understanding of photosensitivity of cells due to both ndogenous and exogenous photosensi­tizers. The cell systems used have included protozoa, bacteria, yeasts, viru es, eggs and p 'rm of various specie , erythrocytes, and a

variet of strains of mammalian cells growing in tis u culture. Chick embryo cell · in tissue culturr ''" rc u ·ed by L wi to demonstrate photo toxic rffect~ of cbemical carcinogens (2, 3), by tilwell to ob erve th effect of neutral red ( 4), and 1 y :l\lenke in tudyinO' the photody­namic action of phloxine and other dyes ( 5) . l\1or r cently Allison et al. u ed several cell lines in a study of phototoxic effect of tars, uroporph rin, and several dye (6). They con­clud d that lysosomes and cell membrane are important ites of damnge. The work of lVIath ws with bacteria i of particular interest b cau he elucidated a membrane effect for om ompound and a l thal interaction of -

mrtho ·ypRoralen with DNA (7-10). Th work

Thi .inv .:tigalion was upporte l b:v a Public H nllh 'crvicc Re earch Grant ~NB 036 9-0 fr m th at.ional Adv.i ory Neurological Di ea P and trok ouncil and Public H allh Re curch

rant ~ -0 40 -04 from the ntional AdYi ory Cane r ounril, Public H alth etYic .

R eciv d July 25, 1969; ace pt d for publi ation 1 ternb r 23 1969. * Fr m th D partm nt of Dermatoloay and

Patholog,,·, Baylor Coll ge of Medicine, Hou ton T X:l..

164

is too voluminous to summanze here and the reader is referred to excellent reviews of the subject (11, 12).

Photosensitization occurring in man as a re­sult of endogenous or exogenous drugs or chem­icals i a major concern of dermatologists. Bet­ter understanclinO' of the problem has been hampered by difficulties encountered in induc­ing photosensitization in experimental systems. The re-administration of a drug as an experi­mental procedure to a patient sensitive to that drug may be hazardous for the patient. At­tempt" to induce photosensitization in labora­tory animals has been uccessful in some instnnces. Several species have been photosensi­tized with p ,oralen compounds (13-17). ams ct al . u ·cd guinea pigs to demonstrate photo­sensitization to chlorpromazine, declomycin, and chlorthiazide (1 , 19). Ison and Blank were able to photosensitize hairless mice with several common photosensitizing drugs (20). The experimental conditions, particularly the conditions of light exposure, make animal ex­periment costly in time and effort for the amow)t of information gained.

We have adapted tissue culture techniques to utilize several mammalian cell lines as the bioloO'ical system for a study of drug photo­s n itivity. This technique has proven useful in the study of drug photosen itivity and provides a versatile and sen itive y tem for photobio­logic experiment .

MATERIAL ArD METHODS

1 issue culture. Four cell line were used for cul­ture including HEp-2 cell originating from a hu-

TISSUE CULTURE PHOTOTOXICITY 165

man laryngeal carcinoma, monkey kidney cells, rabbit kidney cells and rabbit skin fibroblasts. The HEp-2 cells were obtained in culture from Flow Biological Laboratories. Rhesus monkey kid­ney cells were provided by our Department of Virology. Rabbit kidney cells were harvested by standard techniques in which the kidney tissue is minced, trypsinized, and centrifuged. Then the pellet of cells is resuspended in growth media and transferred to culture containers. Rabbit skin cells were obtained using the method of Valenti and Friedman (21). All cells were grown and main­tained in Eagle's basal media with added serum, glutamine, penicillin, and streptomycin.

After initial cultures in flasks, cells were grown in monolayer on coverslips in either of two con­tainers . For use with broad spectrum lamps and wavelengths 300 nm or longer, constricted tubes (Leighton*) were utilized since these provided fo r efficient handling of cell monolayers and al­lowed irradiation of the monolayer through a plane surface on the fl at side of the tube (Fig. 1) . The transmission characteristics of the glass tubes and coverslips preclude their use for irradia­tion with wavelengths 300 nm or shorter (Fig. 2). Therefore. in such cases, ykes-Moore metal tis­sue culture chambers* were used (Fig. 3) provid­ing a 2.5 mm working area between two round 25 mm diameter coverslips separat d by a silicone rubber ga ket. One cover lip was replaced with a quartz coverslip allowing approximately 90 % transmission uniformly th roughou t the ultraviol et spectrum (Fig. 2).

Light expostae. Light sources used in the ex­periments described herein consisted of a bank of 2 Westinghouse FS20T12 fluorescent sun lamps and 2 ylvania cool white lamps or a high intensity diffraction grating monochromator (22) adjusted for emission of a 5 nm half-power bandwidth and selected wa\ elengths as described in each experi­ment.

Irradiation of the tissue culture monolayers was p0rformed after the nutrient media had been re­placed with sterile phosphate buffered saline (pH 7.4) . The culture chamber was placed in the out­put beam of the light source so that tl1e ra~·s

pn o..osed through the fla t glas. wall and coverslip or through the quartz chamber wall to which the monolayer wa attached. After irradiation the buff­ered saline was replaced with nutrient medium and the cells were obserYed for 48 hours or w re processed a specified for each experim nt.

C:onlrol experiments. Before testing the photo­toxic effect of specific drugs in this system, several preliminary and control experiments were per­formed: 1) The effect of light alone on the mono­l a~·er wa d termined b~' exposincr monolayers without drug to the light sources under condi­tions id ntical to those to be used later in drug te ting. 1 o gro~s or microscopic damage to the monolayers was ob erved with long ultraviolet or

*Obtained from Bellco Biological Glassware & Equipment Co ., Vineland, N.J.

Frc. I. Constricted (Leighton) tub s provided a plane surface for light expo ur s of c 11 monolayers growing on a removable overslip. Black tape was u cd to protect a portion of each monolay r from light xposurc thu providing an int mal control.

, . i ~ ibl c " ·a,·clf'ngt ll s undrr such condi1ions. With broad spect rum or mid-U' wavelength , a th resh­old encr~y dose for killing of cells of th mono­layer was determined by serially testing cliff rent exposure tim . Th i.hre hold dose with the bank of lamps at n. distance of .5 em was 10 minutes for all 4 cell lines. For the monochromator, the threshold varied with wavelength in the mid ultra­violet rnn~e. To injury was obseryed with the long ultraYiolet '"a,·elength actually reported in this study (T able II). 2) The effect of the drug or chemical on cell monolayer without light exposure was determined for each compound by incubating monol a~·e rs for 18 hours with a series of double dilution of each drug. After 18 hours the drug solution was replaced with nutrient medium and the cell ob. en·ed for 48 hour for any gross or micro copic ffect. From this, a non-toxic concen­t ration of the drug was chosen for final t sting (Table I). Drug were obtained in pure form from the manufacturer where possibl e or from the pharmac~' in a form ui table for pr paring

1 6 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

.,T

100

80

60

40

20

---------------------

Tl SSUE CULTURE TUBE QUARTZ COVER SLIP

0 ======~L-------~------~------~ 200 300 400 500 600

>.. n m

1 ra. 2. The pyrex constricted tubes transmitted wav lengths as short a 390 nm. Quartz coversliJ? in th small chambers (F1g. 3) were u ed for m1d UV wavel ngths.

FIG. 3. yk s-Moore chambers were utilized with ·hort wav length monochromatic light.

th buff r d olulion . 3) Th ability of c lls to ·urviY a 2 hour p riod of incubation in pho phal buff r d alin \Vithou t nutrients '~as confirmed. 4) lutions of the druO's li ted in Table I were irradiat d for 7 minut u ing th Wr tinghouse sunlamps nt a di tanc of .5 em. These solu­tion w rc then appli d to cell monolayers for 2 h ur without an~' ob ITabl damage to the cell . indicating that thi pr -irradiation of the dru did not r .nd r it toxic. 5) Pre-irradiation of cell did not render them mor or le s s n itive to 8-m tho:xyp oral n hematoporphyrin. demethyl­chlort tra y line, or thi ridazine. 6) The drug in elution wa further valuated b. xaminin()' the ff t of licrh t xpo d dru elution on a portion

of h · monolay r hi ldcd from th ligh t ourc

by wrapping part of the chamber with black elec­trical tape. Shielded cells thrived while unshielded cells were killed by light exposure even though both parts of the same monolayer were in the same chamber exposed to the same drug solution.

Technique of drug evaluation. Observation of the pbototoxic effect of various drugs (Table I) was evaluated in the following way : 1) Nutrient media in the chambers was replaced with a buffered solution of the drug in non-toxic concentration. 2) Th e chambers were returned to the dark for 2 hours incubation. 3) Duplicate chambers were irradiated for 7 minutes using Westinghouse sun­lamps at a distance of 8.5 em. 4) Nutrient medium was replaced and the chambers returned to the dark. 5) Gross and microscopic observations of

TABLE I Results of testing for phototoxicity

Response observed Non- after light

Drug toxic cone %

Phototoxic Non-toxic

Sv.lfa gro11p Snlf anilamide 0.1 X Tolbutamide 0.1 X

Tetracyclines Demethyl- 0.1 X

chl ortetr a-cyclin e

Te tra.r~' cl i ne 0.1 X Penicillins

Penicillin G 0.1 X Thiazide H~·d rochlor- 0. 1 X

thiazide Ph enolhiazines

Chlorproma- 0.1 X zine IJCl

Thioridazi11e 0.01 X Porph yn·ns

Hematopor- 0.01 X phyrin HCI

Furoco wnarin s -Methoxy- 0.1 X psoralen

A ntimalaria.ls Chloroquine 0.1 + (4 hr) X (24 hr) Qninacrine 0.1 + (4 hr) X (24 hr)

Dyes Toluidine Blue 0.001 X Eo in Y 0.01 X Methylene 0.001 X

Blue Phloxine 0.01 X

Othe1· Cyclamate 0.1 X

TISSUE CULTURE PHOTOTOXICITY 167

TABLE II

Pholotoxicity using monochTomatic light

HEp-2 cell monolayer

Threshold Drug X, nm energy

mjj cm2

-methoxypsoralen 320 20

0.1C7£ 330 20 360 45 405 (-)

Hematoporphyrin 320 (-)

0.01 0 360 (-)

405 1300

Demethylchlortetracycline 330 800

0.1%

Thioridazine 320 120

0.01% 330 110

monolayers were made at 24 and 48 hours. 6) Control preparations for each experiment included (a ) cells kept in the dark without drug or light exposure, (b ) cells receiving sub-threshold light exposure alone, and (e) a portion of each chamber covered with black electrical tape to block direct light exposure.

pecific experiments were designed to evaluate 1) the feasibility of demonstrating in our labora­tor~· photoxic responses to a number of chemicals 2) the l e~gth of drug incubation required, 3) the sensitivity of the procedure for detecting photo­toxic responses induced b~· dilute concentrations of 8-methoxypsoralen. and 4) the reliability and applicability of the technique as a method for de­tecting phototoxic responses induced b~' different drugs.

RE SULTS

Initial experiments on drug or chemical in­

duced phototoxicity of tissue cultures were car­

ried out using a{Tents known to have a trong

I hototoxic action, pecifically -methoxyp or­

~len, hematoporphyrin, and demethylchlortetra­

cycline. These drugs in a final concentration

of 0.1 o/o, or 0.01 o/o for hematoporphyrin, caused

no vi ~ible cytotoxic effect either grossly or mi­

cro copically. After broad spectrum light ex­

posure of sensitized cells to sublethal amount

of energy, a phototoxic effect was observed if

the cells were incubated with the drug for one

hour or longer prior to irradiation. All 4 cell

line reacted similarly. The cytotoxic effects ob­

:'lerved wi th in 6 bourn were inc rea ed den ity

and granularity of the cytoplasm, shrinkage of

the cell, nuclear pykno is progressing to cell

death and complete destruction of the monolayer

within 24 hours (Fig. 4).

The procedure was shown to be useful as an

instrument for demonstrating a phototoxic ef­

fect of a variety of chemical compounds includ­

ing many commonly used photosensitizing

drugs . The ten classes of compounds, the spe­

cific agent tested in each clas , and the ob­

served re ponses are listed in Table I. Two

drugs selected for study (madribon and grise­

ofulvin) were immiscible in the solvents tested.

Two known photosensitizin{T drugs, chlorproma­

zine and sulfanilamide yielded negative results

as did tetracycline and penicillin.

In the survey tests of various drug classes, a

difference in the response of cells to the two an­

timalarials, chloroquine and quinacrine, was ob­

served. The monolayers expo ed to the e drugs

appeared viable at 24 hours but were dest royed

at 4 hours whereas all other phototoxic reac­

tions were apparent by 24 hours.

To evaluate the sensitivity of the te t pro­

cedure as a method of detecting small concen­

t rations of the drug, serial dilutions of 8-me­

thoxypsoralen were prepared and the cells were

incubated with these dilute solutions before

light exposure. This drug in a fiEal concentra­

tion of 1 part in 100,000 caused definite but

incomplete killing of the monolayer and was the

thre hold concentration of this drug inducin{T a

cytotoxic effect. More dilute solutions produced

no detectable effect under these conditions.

MONOLAYER PHOTOTOXICITY (HEP--2 'CELLS)

NEGATIVE POSITIVE

L'IGHT

DARK

~ ·-

FIG. 4. Monolayers on coverslips stained darkly except when destroyed by phototoxic action (Neu­t ral red tain) .

168 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Using monochromatic light, threshold energy values were determined at selected wavelengths within the absorption bands for the compounds shown in Table II.

DISCUSSION

The use of simple biological systems such as single cell d organisms and tissue cultures has been useful in exploring various parameters of cell function altered by photosensitizing agents. Outstanding examples of interest to dermatolo­gists are illustrated by the experiments of Mathew (7-10) and the recent study of Alli­son et al. (6). We have found that the tissue culture syst m is an adaptable and versatile one for u with different light sources and varying conditions of photosensitization. The param t rs one can explore are varied and in­clude direct observation of the cell monolayer, application of an almost endless variety of his­toloCTical and hi toch micnl staininCT techniques, chemical examination of cells and their frac­tions, examination of the supernatant for changes in electrolyte content, nucleic acid con­tent or other constituents, biologic manipula­tion of th c ll system and ultrastructural stud­Ies.

The technique we have adapted as a test system i simpl , reproducible, and sensitive as a d t ctor or screening test for phototoxic ef­f ct of ch micals. Vve have observed excellent eorr lation between re ults of this simple survey trst and knowledCTe gained from clinical experi-nc about the photosensitizing ability of 17

diff r nt drugs selected from 10 classes of chem­ical compounds. Two notable exceptions are hlorpromazine and sulfanilamide. Compounds

w r cho en to test the applicability of the proc dur for u e with a variety of different t:; p s of ch m.icals and was not intended as an xhaustiv urv y.

Di advant[JCTes of the technique include the fact that a knowledCTe of tissue culture techniques i r quir d. pecialized equipment i necessary, but thi is available commercially at modest co t. From th biolo ic tandpoint, sinCTle cell

t m do no imulate the mammalian organism in many wa e.CT. druCT metabolism is altered, and th immune mechanism is lackinCT. While this rna ? useful at times, it will, obviously, impo e a limit on the application of results ob in d from tis ue cultur studies. ince ti -

sue culture requires a sterile system, bacterial and chemical contamination may be trouble­some. Photoallergic reactions obviously will not be detected by the technic we have describedr however, it may well be feasible to adapt the procedure to the phenomenon of lymphocyte transformation and reveal photoallergy. Two compounds, madribon and griseofulvin, were im­miscible in the solvent used.

Our results, though preliminary, indicate that ba ic differences in the mechanisms of pho­totoxicity can be demonstrated. The delayed cytotoxic effect of the antimalarials on tissue culture cells is an interesting observation that deserves further investigation. An interesting speculation would be that the immediate pho­totoxic effect induced by the antimalarial in­volves a cell component vital to cell division but not lethal for the cell until it attempts to divide, thus delaying a grossly visible change in the monolayer. It is too early to confirm an in­terpretation of this sort from our data.

The one hour incubation period required be­fore a phototoxic response could be induced sug­ge ts that mere contact with the drug is not sufficient but that some interaction between the crll and the drug takes place before the cell can be injured by light. Such an interaction might be expected since -metho:x-ypsoralen is known to react with DNA ( , 23, 24) and other photo-ensitizers may affect the cell membrane or ly­osome or multiple sites (6, , 10, 25).

Thi technique is adaptable to many other experiments and measurement of many param­eters rela ted to photosensitivity and photobiol­ogy, and it should enable researchers to learn much about mechanisms of photosensitivity.

REFERENCE' 1. Raab. 0 .: Z. Biol., 39: 524, 1900, cited by Hill.

R. B. Jr.. Bensch, K. G. and King, D. W.: Photosensitization of nucleic acids and pro­t ins: The photodynamic action of acridine orange on living cells in culture. Exp. Cell R es ., 21: 106, 1960.

2. Lewis, M. R.: The photosensitivity of chick embryo cells growing in media containing certain carcinogenic substances. Amer. J . Cancer, 25: 305, 1935.

3. L wi , M. R.: The injurious effect of light upon dividing cells in ti sue culture con­taining fluorescent sub tances. Anat. Rec., 91: 199, 1945.

4. tilwell, E. F.: The photodynamic action of neutral red on embryonic chick c lls grown in vitro. Anat. Rec., 84: 193, 1942.

5. M nke, J. F.: Photodynamic action of normal

TISSUE CULTURE PHOTOTOXICITY 169

and malignant cells in vitro, Contributions to Dermatology 1/:149, Vol. 25 . Carnegie Inst. of Washington, 1935.

6. Allison, A. C., Magnus, I. A. and Young, M . R.: Role of lysosomes and of cell membranes in photosen itizat ion. Nature, 209 : 874, 1966.

7. Mathews, Micheline M. and Si trom, W. R. : Function of carotenoid pigments in non­photosynthetic bacteria. Nature, 184: 1892, 1959.

8. Mathew , Micheline M.: Comparative tudy of lethal photosensitization of sarcie lutea by 8-methoxypsoralen and by toluidine blue. J. Bacterial., 85: 322, 1963.

9. Mathews, Micheline M.: Protective effect of B-carotene against lethal photosensitization by haematoporphyrin. Nature, 203: 1092, 1964.

10. Roth, M . Mathews: Carotenoid pigments and photokilling by acridine orange. J . Bacterial., 93: 506, 1967.

11. Giese, Arthur C .: Photophysiology, Vols. I-IV, Academic Press, New York, 1964.

12. Blum, H arold F .: Carcinogenesis by Ultra­violet L ight, Princeton University Press, Princeton, New J ersey, 1959.

13. H arber, Leonard C. and Baer, Rudolf: Effect of humidity on the photosensitive response to 8-methoxypsoralen. J. Invest. Derm., 44: 61 , 1965.

14. Perone, Vernon B., Scheel, Lester D. and Meitus, Robert J.: A bioassay for the quan­titation of cutaneous reactions associated with pink-rot celery. J. Invest. Derm., 42: 267, 1964.

15. Cloud, T . M., H akin R. and Griffin, A. C.: Photosensitization of the eve with methox­salen. Arch. Ophthal., 66 : 689, 1961.

16. Daniels, Farrington: A simple microbiological method for demonstrating phototoxic com­pounds. J. Invest. D rm. 44: 259 1965.

17. Owens, Donald W., Glick man, J. M., Free­man R. G. and Carnes, R.: Biologic action spectra of 8-methoxyp oralen determined b monochromatic light. J. Invest. D rm., 51: 435, 1968.

18. Sams, W. Mi tchell : The experimental produc­tion of drug phototoxicit) in guin a pig . Arch. Derm ., 94 : 773, 1966.

19. Sams, W. Mitchell and Epstein, John H.: The experimental production of drug photo­toxicity in guinea pigs. J. Inve t. D rm., 4 : 89, 1967.

20. Ison, Arnold and Blank H arv y: TestinO' dru photo toxicity in mice. J. Inve t. Derm., 49: 508, 1967.

21. Valenti, Carlo and Friedman, Eli: Long term cultivation of diploid rabbit kin cells. Tex. R pt. Biol. Med., 26 : Fall, 1968.

22. Knox, J. M., Warshawsky, J. , Lichodziejewski, W. and Freeman, R. G.: Design of a high intensity monochromator. Arch. Derm., 95: 319, 1967.

23. Musajo, L. and Rodighiero. G.: The me ha­nism of action of the skin photo ensitizing furocoumarins. Acta dermatoven r ., 47: 298, 1967.

24. Musajo, L., Rodighiero , G.. Calombo, G., Torlone, V. and Dall'Acqua, F.: Photosensi­tizing furocoumarium: Interactions with DNA and photo-inactivation of DNA-con­taining viruses. E:xperientia, 21: 22, 1965.

25. Bastos, A. L., Terrinha, A. M., Vigario, J. D., Moura Nunes, J . F. and Nune P ti ca, J. L.: UV-induced changes in lyso ornes tagged with quinacrine. Exp. Cell Res ., 42 : 84, 1966.