Effects of 8-Methoxypsoralen and Ultraviolet Radiation on Human Lymphoid Cells in Vitro

Transcription

1 22-22X/8l/762-8$2./ THE JOURNAL OF lnvestigative DERMATOLOGY, 76:8-87, 1981 Copyright 1981 by The Williams & Wilkins Co. Vol. 76, No. 2 Printed in U.S.A. Effects of 8-Methoxypsoralen and Ultraviolet Radiation on Hman Lymphoid Cells in Vitro KENNETH H. KRAEMER, M.D., HAYWOOD L. WATERS, B.S., LAWRENCE F. COHEN, M.D., NICOLAE C. POPESCU, PH.D., SUZANNE c. AMSBAUGH, B.S., JOSEPH A. DIPAOLO, PH.D., DANIEL GLAUBIGER, M.D., PH.D., OWEN L. ELLINGSON, B.S., AND ROBERT E. TARONE, PH.D. Laboratory of Moleclar Carcinogenesis (KK & HW); Pediatric Oncology Branch (LC & DG); Laboratory of Biology (NP, SA, & JD); Biometry Branch (RT), National Cancer Institte, Bethesda, Maryland and Electrooptics Branch, Division of Electronic Prodcts (OE), Brea of Radiologic Health, Food and Drg Administration, Roclille, Maryland, U.S.A. Oral 8-methoxypsoralen (8-MOP) pls high-intensity long-avelength ltraviolet radiation (UV -A) is sed clinically to indce remissions of psoriasis and mycosis fngoides. Lekocytes in 8-MOP containing blood receive UV -A exposre hen circlating throgh the dermis dring therapy. The present stdy tilizes an in vitro assay system to permit qantitation and correlation of mltiple biological and physical alterations in hman lymphoid cells indced by 8-MOP pls UV-A treatment. Additive inhibition oflymphoid cell DNA synthesis by 8- MOP (.1 to 1tg/ml) pls UV -A (1, to 29, J /m 2 ) as accompanied by a synergistic potentiation of cell killing in the therapetic exposre range. Redction in tritiated thymidine ehtdr) incorporation to 65-7% of control vale as associated ith normal srvival; hile 3 HTdR incorporation of less than 5% of control indced by any 8-MOP pls UV-A combination tested as associated ith less than 1% srvival. 8-MOP-DNA-crosslin.ks ere detected by the alkaline eltion assay only hen 3 HTdR incorp9raion as redced to less than 5% of control. The relative nmber of crosslinks increased proportionately ith frther 8-MOP pls UV -A-indced redction in 3 HTdR incorporation. 8-MOP pls UV-A indced at most approximately a 2-fold increase in sister chromatid exchanges (SCE) per chromosome in lymphocytes or lymphoblastoid cells. Increasing 8-MOP pls UV-A exposre reslted in marked toxicity ith fe cells progressing to second division metaphases and no frther increase in SCE's per chromosome. Addition of 13-cis retinoic acid (1tg/ml) to the lymphoblastoid cells prior to 8-MOP pls UV -A treatment did not significantly alter the 3 HTdR incorporation or cell srvival. These stdies demonstrate that in vitro exposre of hman lymphoid cells to therapetic levels of 8-MOP and UV -A may decrease celllar DNA synthesis, prodce DNA 8- MOP interstrand cross-links, redce cell viability and indce small increases in sister chromosome exchanges. Oral 8:methoxypsoralen (8-MOP) pls high-intensity longavelength ltraviolet radiation (UV-A) is being sed clinically to indce remissions in psoriasis [1-3] and in mycosis fngoides [4]. 8-MOP hen activated by UV-A binds to DNA forming 8- MOP-DNA addcts [5] hich in trn inhibit celllar prolifer- Manscript received March 31, 198; accepted for pblication Jly 28, Reprint reqests to: Kenneth H. Kraemer, Laboratory of Moleclar Carcinogenesis, National Cancer Institte, Bilding 37, Room 3E24, Bethesda, MD Abbreviations: 3 HTdR: tritiated thymidine 8-MOP: 8-methoxypsoralen PHA: phytohemaggltinin SCE: sister chromatid exchanges TP A: tetradecanoylphorbol 8 ation. Becase the 8-MOP is present in the blood folloing oral administration [6,7] and UV-A penetrates throgh the epidermis [8], circlating blood cells might be sbjected to 8-MOP DNA photoreactions. Previosly, decreased DNA synthesis in circlating lekocytes from some psoriasis patients receiving 8- MOP pls UV-A (PUVA) therapy as reported [9,1]. An in vitro assay system as developed to facilitate examination of mltiple effects of 8-MOP and UV-A on fresh hman lymphocytes and on transformed lymphocyte cell lines [7]. The stdies presented here permit qantitation and correlation of mltiple 8-MOP pls UV -A indced biological and physical alterations in the same assay system. This investigation demonstrates that inhibition of DNA synthesis by UV-A and 8- MOP is additive, extends the observation of dose dependence of inhibition of DNA synthesis and of cell srvival to lo therapetic concentrations of 8-MOP, and describes the relationship of DNA synthesis inhibition to cell killing and to DNA- 8-MOP cross-link indction. Additional observations report sister chromatid exchange indction in vitro by 8-MOP pls UV -A and the lack of significant effect of 13-cis retinoic acid on 8-MOP pls UV-A indced DNA synthesis inhibition and cell killing. Cells MATERIALS AND METHODS Fresh lymphocytes ere obtained from normal donors by lekophoresis (performed by the Red Cross Blood Bank, Washington, D.C.) and frther prified by flotation in lymphocyte separation medim (LSM soltion, Litton Bionetics, Kensington, MD). An Epstein-Barr virstransformed lymphoblastoid cell line (E-1) from a normal donor as obtained from Dr. Alan Andres, Dermatology Branch, National Cancer Institte. The cells ere cltred in RPMI 164 medim (Grand Island Biological Co., Bffalo, NY or B & B Laboratories, Baltimore, MD) spplemented ith 17% fetal calf serm (Gibco, or HEM Research, Rockville, MD) ithot antibiotics, in closed 49 cm 2 plastic tisse cltre roller flasks (Corning No. 2513) at 37 C ith continos rotation. This procedre reslts in a mild blastogenic response in the fresh lymphocytes [11]. The stock lymphoblastoid cells ere fed at 1 to 3 day intervals to maintain exponential groth. 8-MOP Treatment For irradiation 1,, cells/ml (lymphocytes) or 2, cells/ml (E-1) in 75 cm 2 plastic flasks (Costar Plastics) ere sspended in a salt soltion consisting of the salts and glcose present in RPMI 164 medim (NaCl 6. g/1, KCI.4 g/1, Sodim Bicarbonate 2. g/1, NazHP,.7HzO 1.5 g/1 Calcim Nitrate.4Hz.1 g/1, Magnesim Slfate. 7Hz1 g/1, and Glcose 2.9 g/1). 8-MOP (a gift from Pal B. Elder Co., Bryan, OH) as stored frozen in liqid nitrogen in the dark as a 1 mg/ml stock soltion in absolte ethanol. The 8-MOP as dilted in salt soltion immediately before addition to the cell sspension and incbated ith the cells at 37 C for at least 15 min before UV A exposre. All maniplations ere performed nder gold florescent lamp (Westinghose F4 Go, or General Electric F2 Tl2 Go) illmination. Irradiation T he flasks containing the cell sspensions ere placed horizontally in a holder above a plate glass filter (6-mm thick) hich as sspended

2 Feb above 4 parallel UV -A florescent lamps (FR4T12-PUVA-a gift from Phototherapy Lighting Prodcts Grop, GTE Sylvania, Danvers, MA). Identical type lamps are sed in clinical photo-chemotherapy [1-3] ithot the plate glass ftlter. Figre 1 shos the spectral irradiance of the nftltered "PUV A" lamp and of the lamp ith the plate glass ftlter. Inclded is a calclation of the predicted spectral irradiance sing pigmented or npigmented epidermis (data from reference 8) as a ftlter to mimic the ftltering effect of the epidermis on radiation impinging pon lymphocytes in the dermis. Ths, the plate glass filter eliminates radiation belo 32 nm hich is normally ftltered by the epidermis in vivo. The lamp sorce as described previosly [7]. The calclated irradiance ithin the flask as 22.5 J/m 2 /s ith an estimated ncertainty of ± 25%. Radiant intensity varied approximately 9% across the exposre field. The cells ere kept in sspension dring irradiation and exposed to different portions of the field by shaking the plate glass at approximately 4 revoltions per min. Control nirradiated sspensions ere placed in an opaqe box on the same shaker. DNA Synthesis Five replicate 1 ml samples of the cell sspension from each flask ere placed in tightly capped 12 x 75 mm plastic tbes (Falcon No. 258) ith 1 J.!Ci/ml methyl-tritiated thymidine ehtdr) (21-36 Ci/ mm ICN Pharmaceticals, Irvine, CA) and incbated in the dark for 2 hr in a 37 C ater bath. 3 HTdR incorporation into DNA as measred ' (; -,\ //., if. f!_ \/l \, "' I- - I I ' o--1 d4- -Unlilteredlamp,/',.! / --PiateGiassfilter Jt ti Pigmented Epidermis Filter Unpigmented Epidermis Filt Wavelength (Nanometers) FIG 1. Filtering of "PUV A" florescent lamp radiation by plate glass or by epidermis. The spectral irradiance of Sylvania FR4T12 PUV A lamp operated at 115 v as measred at a distance of 2 meters ith (- ) and ithot (-) 6-mm plate glass ftlter. Predicted spectral irradiance sing npigmented epidermis (- - -) or pigmented epidermis (. ) as a ftlter (data from reference 8) is also shon. Frther details in Materials and Methods. 15 M" 125 M I X E 1. z f= 75 <( "- 5 I CULTURE MEDIUM TEMPERATURE I C) FIG 2. Effects of pretreatment at different temperatres on DNA synthesis in hman lymphoblastoid cells. Cells ere sspended in cltre medim at the indicated temperatre for 3 min, 3 HTdR as added, and each cltre medim as incbated for an additional 2 hr at 37 before harvesting. Frther details in Materials and Methods. 8-MOP PLUS UV-A IN LYMPHOID CELLS 81 by a modification [7] of the Millipore assay techniqe of Robbins et al [12]. Samples ere dripped gently onto the Millipore ftlters, ashed sccessively ith saline, 5% trichloroacetic acid, and ethanol; vacm dried, and conted in RPI Scintillator PPO-POPOP (Research Prodcts International Corp., E lk Grove Village, IL) in Beckman model LS 25 Liqid Scintillation System. In preliminary experiments (Fig 2), a temperatre dependence of the rate of 3 HTdR incorporation into the lymphoblastoid cells as noted. Greatest activity as measred at 37, the highest temperatre tested. Ths, sbseqent experiments ere performed ith the cell sspensions maintained at 37 ± 3 by se of hot air heaters as monitored ith a thermoprobe (Tele-Thermometer model 44TD ith probe model 41, Yello Springs Instrment Co., Yello Spring, OH). Statistical Analysis For each experiment the mean 3 HTdR incorporation determinations ere normalized to the ntreated control vale. Using the method of least sqares, the crve y=a exp (bxc) as fit to the data for each concentration of 8-MOP here y denotes the normalized 3 HTdR incorporation vale and x denotes the dose of UV -A. At every exposre level of UV -A a linear regression analysis as performed sing y as the dependent variable and x as the independent variable. Groth Crve Analysis Groth assays of lymphoblastoid cells: After irradiation a soltion containing vitamins, amino acids, phenol red, and fetal bovine serm as added to each flask so that the fmal medim concentration as 1 x RPMI 164 ith 17% fetal calf serm and 1, lyrnphoblastoid cells/ mi. The flasks ere placed in a 37 C incbator ith 5% C 2 and 95% relative hmidity (Hotpack model 35192), in the dark for 4-14 days ithot medim change. The concentration of viable cells as estimated daily by dplicate hemocytometer conts of trypan ble (Trypan Ble Stain.4% in normal saline, Grand Island Biological Co.) exclding cells. The standard deviation of the dplicate conts as generally less than 1% of the mean. Before treatment greater than 9% of the cells exclded trypan ble. Srvival as estimated by a modification [13] of the method of Alexander and Miklski [14]. The groth crve of the concentration of viable cells as extrapolated back to zero time along the exponentially increasing portion. This method measres the proportion of the treated cell poplation hich srvives ith the capacity to proliferate at a normal rate. Microliter Well Groth Assay A modified limiting diltion assay as sed (described in detail in reference 13). Briefly, serial diltions of treated cells ere innoclated into 24 ells of a microtiter plate (Linbro) (.2 ml cltre medim per ell) and placed in a C 2 incbator for 2-3 eeks. The ells ere examined nder an inverted microscope (Leitz, Diavert) for evidence of cell groth. The fraction of negative ells at each diltion served as a basis for determining the srviving fraction by the Poisson formla. Retinoic Acid Treatment 13-cis-retinoic acid (Ro4-378) as obtained from Roche Pharmaceticals throgh Dr. G. Peck, Dermatology Branch, National Cancer Institte. Stock soltion of 1 mg/ml in absolte ethanol as stored frozen in liqid nitrogen. The 13-cis retinoic acid as dilted in salt soltion immediately before addition to the cell sspension and incbated ith the cells at 37 for 15 min prior to UV -A exposre. Sister Chromatid Exchange Determination Hman lekocyte cltres ere initiated sing 12 ml RPMI 164 medim spplemented ith 17% fetal bovine serm containing 5, cells mi. Phytohemaggltinin (Gibco M form) as added at a concentration of.1 ml per cltre coincidentally ith bromodeoxyridine (BrdU) 3 1-1g/ ml. For the lyrnphoblastoid cell line BrdU 1 }.lg/ ml as added to complete cltre medim containing 1, cells/ mi. All cltres ere maintained in complete darkness and incbated at 37 c, 72 hr for lekocytes and 48 hr for lymphoblastoid cells. Colcemid as added at concentrations of.55!lg/ml and g/ ml at 2 and 4 hr before harvest of lekocytes or lyrnphoblasts, respectively. Cells ere fixed in 3: 1 methanol glacial acetic acid and slides ere prepared by a standard air dry method. Slides ere processed for sister chromatid exchange analysis sing a modification of Kornberg and Freelender's Giemsa staining techniqe [15]. 5 metaphases of each sample ere scored.

3 82 KRAEMER ET AL Vol. 76, No.2 Alkaline Eltion Alkaline eltion as done as previosly described [16-18]. Cells ere prelabeled ith C-thymidine ((.2 J.LCi/ml, 6 mci/mm) and ere placed on ice immediately folloing UV -A exposre. Cells to be evalated for DNA crosslinking ere also exposed to x-irradiation (3R at -4 ). Cells ere then layered on a poly vinyl chloride filter, 2 Jl porosity (Millipore Corp., Bedford, Mass.) by gravity filteration. Cells ere lysed ith 5 ml of a soltion of sodim laroylsarcosinate (.2% /v) (Sarkosyl, ICN Pharmaceticals, Plainvie, NJ) 2 M NaCl,.2 M ethylenedinitrilo tetracetic acid (EDTA), ph 1. The residal DNA on the filter as elted ith a soltion consisting of.4 M H,-EDTA pls tetraporpyl ammonim hydroxide (RSA Corp., Ardsley, NY) 2% in ater ph The eltion rate as.4 ±.3 ml/min. Ninety minte fractions ere collected, and mixed ith 3.3 vol of scintillation flid (Aqasol, Ne England Nclear, Boston, MA) pls 1 9 ::> 8 "' 7 > <( ::> 6 t: 5 <( z ;:: 4 <( 2 ;; 2 I?.<' UV- A EXPOSURE (J I m 2 x 1- "l FIG 3. DNA synthesis in hman lymphocytes after 8-MOP pls UV- A. Relative effect of UV-A () and UV-A pls.111g/ml 8-MOP (e) on 3 HTdR incorporation in,lymphocytes from 4 donors. Each point represents the mean CPM of the 5 replicate treated cltres divided by the mean CPM of the 5 replicate ntreated cltres in the same experiment. The crves ere derived by least sqares analysis of the points. The crves for.1 Jlg/ml8-MOP pls UV -A and 111g/ml8-MOP pls UV-A are reprodced from reference 7. The crve for UV-A alone is significantly different from the O.Ol!Lg/ml 8-MOP pls UV-A crve at every UV-A exposre (p <.5). Frther details as described in the text. 1 9 ;::4 ii'jo 8 ;<;2 1 8"!! NoB-MOP!f o L L L W -A EXPOSURE IJ/m' x 1 - '1 FIG 4. DNA synthesis in hman lymphoblastoid cells after 8-MOP pls UV-A. Relative effect of UV-A ( ) and UV-A pls.111g/ml 8- MOP (e) on 3 HTdR incorporation in lymphoblastoid cell line E-1. Each point represents the mean CPM of 5 replicate treated cltres divided by the mean CPM of the 5 replicate ntreated cltres in the same experiment. The crves for.1 J.Lg/ml 8-MOP pls UV-A and 1 11g/ml 8-MOP pls UV-A are reprodced from reference 7. The crve for UV-A alone is significantly different from the.1 Jlg/ml 8-MOP pls UV-A crve at every UV-A exposre (p <.5). Frther details as described in the text..75% (V /V) acetic acid to redce chemilminiesence. Residal radioactivity on the fllter as solbilized by 1 N hydrochloric acid and heat (6, 1 hr) folloed by sodim hydroxide (.4 N, 1 hr) and the same scintillation flid as then added for conting. Calclation of Crosslink Coefficient The fraction of DNA retained on the fllter is calclated from the amont of the total DNA elted in the timed fractions. For evalation of crosslinking, both control and t'est cells are exposed to 3 R to indce single-strand breaks. When the DNA is sbseqently denatred nder alkaline conditions, control DNA eltes rapidly. Increased retention of DNA on the fllter from cells treated ith PUV A pls 3 R indicates DNA crosslinking. The crosslink coefficient (}{,) is calclated according to the forml [ 1- Ro]112 K c = -R' here Ro is the fraction of DNA retained in the 3 R control sample and Ro' is the fraction of DNA retained in the PUV A pls 3 R treated sample both taken at a standard time after eltion begins. The derivation of this formla is fond in references 17 and 18. RESULTS Thymidine Incorporation in Fresh Lymphocytes The relative effect of UV -A (-29, J /m 2 ) alone and folloing.1 J.tg/ml 8-MOP on 3 HTdR incorporation into lymphocytes in 11 experiments ith for donors is shon in Fig 3. The 3 HTdR incorporation in the nirradiated cltres of these cells hich had been slightly stimlated by 1 to 4 days incbation in medim containing fetal bovine serm ranged from 796 to 4, cpm per 1 6 cells. The.1 J.tg/ml 8-MOP pls UV -A treatment cased a significant (p <.5) redction in 3 HTdR incorporation in comparison to that cased by UV-A alone at every exposre tested. Figre 3 also contains comparable data from reference 7 for lymphocytes treated ith.1 j.tg/ml or 1. j.tg/ml 8-MOP before UV -A exposre. The.1 J.Lg/ml 8-MOP reslt as significantly (p <.5) different from the.1 J.Lg/ml8-MOP data only for UV-A exposres of 15, and 29, J/m 2. Thymidine Incorporation in Lymphoblastoid Cells Figre 4 shos the relative effect of UV-A (-29, J/m 2 ) alone and folloing.1 J.tg/ml8-MOP on 3 HTdR incorporation in 15 experiments ith a lymphoblastoid cell line from a normal donor. The 3 HTdR incorporation in nirradiated cltres ranged from 41,416 to 263,734 cpm/2, cells. The degree of inhibition of 3 HTdR incorporation folloing exposre to UV -A pls.1 J.tg/ml 8-MOP as significantly different (p <.1) from that after UV-A alone at every exposre tested. Also shon are the reslts for similar data at 8-MOP concentrations of.1 J.tg/ml and 1 J.tg/ml from reference 7. There as no significant difference beteen the.1 J.tg/ml8-MOP crve and the previos.1 J.tg/ml 8-MOP crve for UV -A exposres of 7,5 J/m 2 or less. As ith the lymphocytes (Fig 3), the.1 J.tg/ml8-MOP crve as significantly different (p <.5) from the previos.1 J.tg/ml8-MOP crve only for UV-A exposres o.15, and 29, J/m 2 Redctions in DNA synthesis ere observed by treatment of lymphoblastoid cells ith either UV -A or 8-MOP alone (Fig 4). For most combinations of 8-MOP and UV-A the effect of treatment ith 8-MOP pls UV -A on inhibition of 3 HTdR incorporation as approximately eqal to the sm of the extent of inhibitions observed ith 8-MOP and UV-A separately (Fig 5). Evidence sggestive of a eak synergism as observed only ith the highest combinations tested. The percent of inhibition of 3 HTdR incorporation observed folloing combined treatment ith.1 J.tg/ml 8-MOP pls 15, J/m 2 UV-A as approximately eqal to the sm of the percent inhibition observed ith the separate treatments irrespective of the seqence of the treatments (Table I). 3 HTdR incorporation increased to control levels by 48 hr

4 Feb MOP PLUS UV-A IN LYMPHOID CELLS 83 7 experiments, by groth in microtiter ells in 2 experiments and by both methods in 1 experiments. There as a UV -A dose-dependent decrease in srvival at every 8-MOP concentration. UV-A exposre of 7,5 J i m 2 folloing pretreatment ith 1. gi ml,.1 giml or.1 J.Lgi ml 8-MOP reslted in srvivals of.3%, 6%, and 8% respectively. Neit her 8-MOP treatment alone nor UV -A exposre alone significantly redced cell srvival. The relationship beteen the inhibition of 3 HTdR incorpo UV-A EXPOSURE IJ/ m 2 x > ' FIG 5. Additivity of DNA synthesis inhibition by 8-MOP pls UV A in hman lymphoblastoid cells. The crves from Fig 4 ere evalated at each UV -A exposre for the percent inhibition of "HTdR incorporation and expressed as bars. The bar representing the inhibition of ''HTdR incorporation observed by each exposre of UV -A alone as added on top of the bar representing inhibition observed by the appropriate concentration of 8-MOP a lone and placed next to the bar representing inhibition observed by the combination of 8-MOP pls UV-A. For ajj except the highest combinations tested, the sm of the heights of the separate bars is not less than the heigh t of th e bar representing the combined treatment..j 16 ::E en.j.j 6.J CD q: > 8- MOP UV-A pg/ ml J / m2 6 I BEFORE AFTER 15. UV A UV A TABLE L Effect of seqence of 8-MOP and UV-A treatments on lym.phoblastoid cell DNA synthesis Treatment "HTdR lncorpo- Inhibition of ration CPM ± "HTdR lncor- 8-MOP g/ ml J / m' UV-A 2SE poration Crossl in k Coefficient 123,76 ± 9,334 Control.1 95,763 ± 7,719 22%. 15, 84,666 ± 3,84 31 % O.Ql7.1 after 15, 53,775 ± 4,85 56%. UV-A.1 before 15, 5,838 ± 7,639 59%.86 UV-A after treatment in cltres treated ith 15, J im UV-A alone,.1 gi ml 8-MOP alone or UV-A folloed by 8-MOP. In contrast, "HTdR incorporation progressively decreased in clt res treated ith.1 gi ml 8-MOP folloed by 15, Jim UV -A so that by 48 hr after treatment ahtdr incorporation as less than 5% of control cltres (data not shon). The cell viability as not appreciably altered ith respect to the ntreated control cells by the.1 giml 8-MOP alone, the 15, Ji m 2 UV-A alone, or the treatment ith 8-MOP after UV -A. In contrast, there as a marked inhibition of cell viability indced by the treatment ith 8-MOP before UV-A (Fig 6). Measrements of DNA crosslinking by the alkaline eltion techniqe revealed significant crosslinking only in the cells treated ith 8-MOP before UV-A (Table I). DNA-8-MOP Crosslinhing Figre 7 shos the relationship beteen the 'JHTdR incorporation folloing 8-MOP (1.,.1, or.1 giml) pls sbseqent UV-A treatment and the relative nmber of 8-MOP DNA interstrand crosslinks indced. When ahtdr incorporation as less than 5% of control there as a dose-dependent increase in 8-MOP-DNA crosslinking detected. Cell Srvival Stdies Figre 8 shos s vival of lymphoblastoid cells after treatment ith UV-A or ith 8-MOP (1.,.1, or.1 gi ml) pls UV-A. The reslts of 29 experiments are plotted. Srvival as measred by extrapolation from groth crves (as in Fig 6) in 1 4 L--L--J---L--LL-J- J DAYS AFrER UV- A EXPOSURE F1G 6. Viability of hman lymphoblastoid cells after 8-MOP pls UV-A exposre. Log phase cells ere treated ith.1 JLg! ml 8-MOP before () or after (.a.) exposre to 15, J / m 2 UV-A. Control cells ere ntreated ( ), or exposed only to.1 J!g/ ml 8-MOP ( ) or to 15, J / m 2 UV-A (). Hemocytometer conts of viable (trypan ble excl ding) cells ere performed daily folloing treatment. Frther details in Materials and Methods. ;e!: 1 9 a 8?S > q: 7 6 t;: <! z f= q: a i"! J= DNA CROSSLINK COEFFICIENT FI G 7. Helationship of DNA synthesis inhibition after 8-MOP pls UV-A to DNA 8-MOP crosslink indction. Lymphoblastoid cells ere incbated ith.1 j!g/ ml (e),.1 Jlg/ ml (.a.), l j!g/ ml (), or no ( ) 8- MOP and exposed to UV-A. "HTdH incorporation as determined from the crves in Fig 4 for each combination of 8-MOP and UV-A. T he DNA crosslink coefficient as determined sing the alkaline el tion assay. Frther details are described in Materials and Methods.

5 84 KRAEMER ET AL c 1 1 NOBMOP i '\: A e t " A A A Q) A Q).9- ::J (f) " " A.1 f<q / ml e i.1 A.1 f'g/ ml Vol. 76, No.2 ith.1 t-tg/ml 8-MOP pls 7,5 J/m" UV-A reslted in approximately a 65% increase in the nmber of SCE's per metaphase and per chromosome. The lymphocyte cltres ere also analyzed 'for the freqency of first, second, and third division metaphases on the basis of the Giemsa pattern. The freqency of first division metaphases increased and second division meta phases decreased after 8-MOP pls UV -A. The combined regime also increased SCE in lymphoblastoid cells (Table III). Hoever, treatment ith.1 t-tg/ rnl 8-MOP pls 3, J/m 2 UV-A reslted in an approximately 2-fold increase in SCE's per chromosome or per metaphase. 8-MOP alone (.1 t-tg/ml) or UV-A alone (15, J/m 2 ) ere ineffective in altering the nmber of SCE's. Treatment ith.1 t-tg/rnl 8-MOP pls 15, J /m 2 UV -A sbstantially redced the nmber of second division metaphases. DISCUSSION The in vitro assay system described approximates some of the conditions of 8-MOP pls UV -A on hman lymphoid cells in vivo dring photochemotherapy. It is designed to measre direct effects of 8-MOP pls UV-A. The UV-A radiation delivered to the cells in in vitro as ftltered to remove the shortave ltraviolet radiation hich is absorbed by the epidermis before reaching the lymphoid cells circlating throgh the dermis in vivo. The plate glass sed in these experiments reslted in a spectral distribtion similar to that passing throgh pigmented epidermis at avelengths belo 33 nm (Fig 1). This short-avelength ltraviolet radiation (in the UV B range) is more toxic to lekocytes than the UV-A [19] and UV-A Exposre (J /m' x 1-1 FIG 8. Srvival of hman lymphoblastoid cells after 8-MOP pls UV-A. Log phase cells ere treated ith O.Olj.!g/rnl (, e), O.lj.!g/ ml (6, A), 1 j.!g/ ml (, ), or no ('il, '9') 8-MOP and exposed to UV-A. Srvival as measred by extrapolation from groth crves (closed symbols ) and by groth in micro titer ells (open symbols ). The crves ere calclated by the method of least sqares. Frther details in Materials and Methods. ration indced by 8-MOP pls UV-A (from Fig 4) and ceil srvival (from Fig 8) indicates that srvival as naffected at levels of 'lhtdr incorporation of 6 to 7% of control (Fig 9). ahtdr incorporation less than 6% of control as associated ith decreased cell srvival. The correlation coefficient beteen the log of the srviving fraction and the fraction of 3 HTdR incorporation remaining as.91 over the linear portion of the crve. Similar srvival as observed at a similar extent of 3 HTdR incorporation inhibition obtained by different combinations of 8-MOP pls UV-A. Ths, 5% of control 3 HTdR incorporation as observed ith 1 t-tg/ rnl 8-MOP pls 3, J/ m 1 UV-A, ith.11-lg/ rnl 8-MOP pls 15,5 J / m 2 UV-A, and ith.1 t-tg/ml 8-MOP pls 58, J/m 2 UV-A and as associated ith approximately.5% srvival ith each combination. Effects of Retinoic Acid Preincation of lymphoblastoid cells ith 1 t-tg/rnl 13 cisretinoic acid as associated ith a decrease in ahtdr incorporation nder each of the 4 conditions shon in Fig loa, hoever, this decrease as not significant at the.5 significance level (p =.8). Srvival as naffected by preincbation ith 13 cis retinoic acid prior to 8-MOP, UV-A, or 8-MOP pls UV-A treatment (Fig lob). Indction of Sister Chromatid Exchanges The indction of sister chromatid exchanges in lymphocytes after treatment ith 8-MOP pls UV-A as significantly increased (Table II). Second division metaphases ere analyzed for the freqency of sister chromatid exchanges. Treatment c Q) Q).9- > "' ::J (f) 1 \ '\7. \ i' 1.1 \ E1 l I ' HTdR Incorporation Remaining {percent) FJG 9. Relationship of hman lymphoblastoid cell srvival to DNA synthesis inhibition after 8-MOP pls UV-A. The mean srvival aftej treatment ith O.Olj.!g/ ml (e),.1 1-'g/ ml (A), 1 1-'g/ ml (). or no ('il). MOP pls UV-A as calclated from Fig 8 and compared to the mean determinations of "HTdR incorporation r&maining (in Fig 4). The coefficient of correlation beteen the log of the srviving fraction and the percent "HTdR incorporation as.91.

6 Feb e 13 E 12 al 11 ;;; 1 E ::J 9 1f1 z >= <{ 8 7 cr: 6 c cr: 5 4 cr: "C f- if \. A. 13-cis Retinoic Acid 1 g l ml S Methoxypsoralen. 1g / ml.\ UV A 1.5 J / cm' I 8 13-CIS R.A MOP + + UV A f- H- B.A, I N e: cis Retinoic Acid 1 g / ml 8-Methoxypsoralen.1 g / ml UV A 1.5 Jicm' CIS R.A MOP t' UV-A FIG 1. Effect of 13-cis retinoic acid on DNA synthesis and cell s rvival in hman lymphoblastoid cells treated ith 8-MOP pls UV A. Cells ere treated ith.11-'g/ ml 8-MOP (6,.&), 15, J / m 2 UV-A (,.), 8-MOP pls UV-A (, +), or ntreated (, e) ith (closed symbols) or ithot (open symbols), exposm e to 11-'g/ ml 13-cis retinoic acid. A, "HTdR incorporation determined for 2 hr post-treatment (mean± SE). The difference in "HTdR incorporation beteen cl tres receiving 13-cis retinoic acid and those not receiving 13-cis retinoic acid as not statistically significant. B, Srvival measred by microtiter ell assay. There as no significant difference in s1 viva l beteen cltres receiving 13-cis retinoic acid and those not receiving 13-cis retinoic acid. 8-MOP PLUS UV-A IN LYMPHOID CELLS 85 ths its removal is important in seeking to mimic the in vivo sitation. The in vitro assay as performed at 37 for increased DNA synthesis (Fig 2) and to mimic the temperatre of circlating lymphoid cells. 8-MOP concentrations sed ere in the range measmed in the serm of patients ndergoing photochemotherapy (cited in reference 7). Hoever, no attempt as made to stdy the effect of circlating metabolites of 8-MOP hich are also present in patients [6], or other possible indirect effects of 8-MOP pls UV-A in vivo. UV-A exposre to circlating lymphocytes has been estimated to be 1-5% of the skin s face dose (19]. T his estimate as based on assming a ctaneos blood volme of 1-5 ml by analogy to experiments in small mammals. With typical skin UV-A exposre of 1, J / m (1 J / cm 2 ) the circlating lymphocytes old be estimated to be receiving UV-A exposres on the order of 1, to 5, J im. A rapid, dose-dependent inhibition of DNA synthesis occrred ith 8-MOP pls UV-A treatment of hman lymphocytes (Fig 3) and a lymphoblastoid cell line (Fig 4). Scherer, Kern, and Bran-Falco [2] and Krger, Christophers, and Schlark (21] measred DNA synthesis 3 days after 8-MOP pls UV -A treatment of phytohemaggltinin (PHA) stimlated lekocytes. They fond a qalitativiely sin1ilar dose-dependent inhibition of DNA synthesis for the combinnd treatment bt no effect ith 8-MOP or UVA alone. Hoever, they did not measre immediate inhibition of DNA synthesis. In or stdies decreased DNA synthesis as observed hen measred in1- mediately after treatment ith 8-MOP or UV-A alone, bt by 48 hr, DNA synthesis had increased to the control level. The separate effects of 8-MOP and UV-A treatments on immediate inhibition of DNA synthesis ere approximately additive (Fig 5) (for most combinations of 8-MOP and UV-A) and independent of the seqence of treatments (Table I). 8- MOP in the dark is knon to bind noncovalently ith DNA [5]. UV -A alone has been shon to alter transfer RNA metabolism in bacteria [22]. Hoever, the inhibition of DNA synthesis indced by the lesion(s) reslting from treatment ith 8-MOP alone or UV-A alone as nonlethal (Fig 8). Only in those cltres treated ith UV-A after 8-MOP as cell srvival significantly altered (Fig 6). Frther, only cells from those cltmes contained detectable DNA interstrand crosslinks. Smvival of lymphoblastoid cells after 8-MOP pls UV-A treatmet as markedly dose-dependent (Fig 8). Depending on the 8-MOP preincbation concentration, srvival after a UV-A exposre of 3,5 J / m 2, presmably in the therapetic range, varied from.3% to 1%. Wlf and Wettermark [23] stdying PHA stimlated lekocytes reported similar alterations of cell trnover depending on the 8-MOP concentrations. We fond that long-term cell smvival as predictable from the alteration of DNA synthesis measmed in the first 2 hr after treatment of the cells ith 8-MOP folloed by UV-A, irrespective of the 8-MOP concentration (Fig 9). Ths DNA synthesis of 65-7% of the control vale or greater as associated ith virtally normal s-vival, hile DNA synthesis of less than 5% of the control vale as associated ith less than 1% smvival. There as a similar threshold vale for DNA crosslink detection occrring ith DNA synthesis less than 5% of the control vale. This sggests that DNA crosslinks are associated ith lethality. Or previos stdy [9] detected a greater than 5% redction in DNA synthesis in circlating lekocytes of pso- TABLE II. Sister chromatid exchanges (SCE) indced by 8-MOP pls UV-A in hman blood lymphocytes Treatment SCE's per metaphase 8-MOP (J.!g/ ml) UV-A (J i m' ) Range Mean± SE SCE's per chromo- Metaphase freq ency some First division Second division Third division ± % 6% 2% 7, ± % 56% 14% ± % 47% 17%.1 7, ± % 34% 14%

7 86 KRAEMER ET AL Vol. 76, No. 2 8-MOP (l"g/ ml) TABLE III. Sister chromatid exchanges (SCE) indced by 8- MOP pls UV-A in hman lymphoblastoid cells Treatment Average No. of SCE's per metaphase SCE's UV-A chromosomes per (J/ m") per metaphase Range Mean± SE chromosome ± , ± L ± , "± , No Metaphases Detected riasis patients ndergoing photochemotherapy hen "HTdR incorporation as measred in blood samples taken immediately after UV-A. A similar redction in DNA synthesis in lekocytes of photochemotherapy treated patients as reported by Friedmann and Rogers [1). If the in vitro data can be extrapolated to the in vivo sitation, this old imply that a portion of the circlating lekocytes old be killed folloing 8-MOP pls UV -A. In this regard Ortonne et al [24], and Haftek et al [25] reported decreased circlating E rosette forming cells and Dahl, Nyfors, and Brodthagen [26] observed a decrease in the percentage of peripheral blood netrophils in psoriasis patients treated ith 8-MOP pls UV-A. Retinoic acid derivatives have been reported to shorten the clinical corse of photochemotherapy in psoriasis patients [27). Epstein [28] fond that in hairless mice, retinoic acid inhibited ltraviolet-indced epidermal hyperplasia as measred by atoradiographic assessment of 3 HTdR incorporation. Kensler and Meller [29] reported that retinoic acid, bt not 13 cis retinoic acid, as a potent inhibitor of the increased DNA synthesis indced by tetradecanoylphorbol acetate (TP A) pls PHA in bovine lymphocytes. Ths the mechanism of action of th e retinoids in psoriasis patients receiving photochemotherapy cold involve a potentiation of the effects of 8-MOP pls UV -A on DNA synthesis or cell-srvival. In or stdies, hoever, ith a short exposre to a single concentration of 13 cis-retinoic acid, in the presmed therapetic r ange [3], no significant alteration in DNA synthesis (Fig loa) or cell srvival (Fig lob) as observed in comparison to the vales observed after treatment ith 8-MOP, UV-A or both in the absence of the retinoic acid derivative. 8-MOP pls UV -A treatment of hman lymphoid cells in vitro has been shon to indce an increased rate of sister chromatid exchanges [31-36]. In the present stdy at most a 2- fold increase in SCE's after 8-MOP pls UV-A (Tables II and III), as fond. In lymphocytes and lymphoblastoid cells, 8- MOP pretreatment folloed by UVA exposre as associated ith cell toxicity or mitotic delay as indicated by the smaller fraction of cells attaining second mitoses dring the incbation period. With.1 JJ.g/ ml8-mop pre-incbation, 15, J/m 2 UV A folloed by BrdU had an adverse effect on cell progression as indicated by the lo nmber of second division meta phases seen in the lymphpblastoid cells. Ths, based on 5 metaphases, there as a very limited range of 8-MOP pls UV -A concentrations in hich increases in SCE cold be detected; and the increase _in SCE freqency as relatively small. In contrast, treatment of lymphocytes ith other DNA crosslinking agents may indce 'an increase in SCE to 5 times control vale (mitomycin-c) (36], or to 1 times t he control vale (bslfan) [37]. Ths in hman lymphoid cells, SCE indction by 8-MOP pls UV -A is a relatively insensitive indicator of cell damage. Lambert et al [33] demonstrated a 31% increase in SCE/cell hen cells ere removed from patients 2 hr after oral administration of 8-MOP sbseqently exposed to 18, J/m 2 UV-A. Wolf Schreiner et a1 [35] fond a 34% increase in SCE/c;ell ith a similar protocol sing 1,5 J /m 2 UV -A. Since circlating lymphocytes may have received less than 1, J/m 2 UV-A, as discssed by Brger and Simons [38] the nmber of SCE indced per cell m ay be too lo to be detected. Other stdies, hoever, have docmented alterations in circlating lymphoid cells of patients treated ith 8-MOP pls UV -A. In this regard, Kraemer and Weinstein [9] and Friedmann and Rogers [1] detected decreased DNA synthesis of circlating lekocytes of some psoriasis patients ndergoing photochemotherapy and Strass et al [39] detected an increased freqency of thioganine resistant circlating lymphocytes dring 8-MOP photochemotherapy for vitiligo. REFERENCES l. Parrish JA, Fitzpatrick TB, Tanenbam L, Pathak MA: Photochemotherapy of psoriasis ith oral methoxsalen and longave ltraviolet light. Ne Engl J Med 291: , Wolff K, Fitzpatrick TB, Parrish JA, Gschnait F, Honigsmann H, Pathak MA, Tanenbam L: Photochemotherapy for psoriasis ith orally administered methoxysalen. Arch Derma to! 112:943-95, Melski JW, Tanenbam L, Parrish JA, Fitzpatrick TB, Bleich HL: Oral methoxsalen photochemotherapy for the treatment of psoriasis: Cooperative clinical trial. J Invest Dermatol 68: Roenigk HH: Photochemotherapy for mycosis fngoides. Arch Dermmatol113: , Tjerneld F, Norden B, Ljngeren B: Interaction beteen DNA and 8- methoxy psoralen stdied by linear dichroism. Photochem Photobiol 29: , Bsch U, Schmid J, Koss FW, Zipp H, Zimmer A: Pharmacokinetics and metabolite pattern of 8-methoxypsoralen in man folloing oral administration as compared to the pharacokinetics in rat and dog. Arch Dermatol Res 262: , Kraemer KH, Waters HL, Ellingson OL, Tarone RE: Psoralen pls ltrav iolet radiation- indced inhibition of DNA synthesis and viability in hman lymphoid cells in vitro. Photochem Photobiol 3:263-27, Everett MA, Yeargers E, Sayre R, Olson HL: Penetration of epidermis by ltraviolet rays. Photochem Photobiol 5: , Kraemer KH, Weinstein GD: Decreased thymidine incorporation in circlating lekocytes after treatment of psoriasis ith psoralen and long-ave ltraviolet ligh t. J Invest Dermatol 69: , Friedmann PS, Rogers S: Photochemotherapy of psoriasis: DNA damage in blood lymphocytes. J Invest Derma to! 74:44-43, Zielski JV, Golb SH: Fetal calf serm-indced blastogenic and cytotoxic responses of hman lymphocytes. Cancer Res 37: , Robbins JH, Gart JJ, Levis WR, Brk PG: The millipore fllter assay techniqe for measring tritiated thymidine incorporation into DNA in lecocyte cltres. Clin Exp lmmnol 11: ' 13. Kraemer KH, Waters HL, Bchanan JK: Srvival of hman lymphoblastoid cells after DNA damage measred by groth in microtiter ells. Mtat Res, 72: , Alexander P, Miklski ZB: Differennes in the response of lekaemia cells in tisse cltre to nitrogen mstard and to dimethyl myleran. Biochem Pharmacal 5: , Kornberg JR, Freelender EF: Giemsa techniqe for the detection of sister chromatid exchanges. Chromosoma (Berl) 48: ' 16. Kohn KW, Erickson LC, Eig HAG, Friedman CA: Fractionation of DNA from mammalian cells by alkaline el tion. Biochemistry 15: , Eig HAG, Kohn KW: DNA protein cross-linking and DNA interstrand cross-linking. Cancer Res 38: , Kohn KW, Eig RAG, Erickson LC, Zelling LA: Measrements of strand breaks and cross-links in DNA by alkaline eltion, Handbooh of DNA Repair Techniqes. Edited by E Friedberg P Hanaalt. Marcel Dekker, in press ' 19. Morison WL, Parrish JA, Anderson RR, Bloch KJ: Sensitivity of mononclear cells to UV radiation. Photochem Photobiol 29: , Scherer H, Kern B, Bran-Falco : UV-A-indced inhibition of proliferation of PHA-stimlated lymphocytes from hman

8 F eb trea ted i th 8- methoxypsoralen. Br J Dermatol 97: , 1977 Krger J P, Christophers E, Schlark LA: Dose-effects of 8-methoxypsoralen and UV A in cl tred hman lymphocytes. Br J Dermatol 98: 141-l44, 1978 J agger J: E ffects of near-ltraviolet radiation on microrganisms. Photochem P hotobiol 23: , 1976 Wlf H C, Wettermark G: T oxic effects of 8-methoxypsora.len on Ortonne JP, Clady AI, Alario A, Thivolet J : Decreased circlating E rosette forming cells in psoralen UV A treated patients. Arch D ermatol Res 258:35-36, 1977 H aftek M, Glinski W, J ablonska S, Obalek S: T Lymphocyte E J Invest Dermatol 72: , 1979 D a hl KB, Nyfors A, Brodthagen H: Decrease in netrophils observed in vivo in psoriatics after PUV A therapy. Arch Derma to! R es 262: , 1978 F ritsch P O, H onigs mann H, J aschke E, Wolff K : Agmentation of oral methoxsalen photochemotherapy ith an oral retinoic acid derivative. J Invest Derma to! 7: , 1978 E pstein J, cited in Kripke ML, Glasman HN: Retinoic acid and photocarcinogenesis orkshop. J Am Acad Dermatol 2: , 198 Kensler TW, Meller GC: Retinoic acid inhibition of the comitogenic action of merezin and phorbol esters in bovine lymphocytes. Cancer Res 38: , 1978 Frolik CA, T avela T E, Peck GL, Sporn MB: High pressre liqid chromatographic determination of 13-cis-retinoic acid and alltrans-retinoic acid in hman plasma. Analyt Biochem 86:743- lymphocyte devision. Arch Dermatol Res 26:87-92, 1977 rosette fnction dring photochemotherapy (PUV A) of psoriasis. 8 -MOP PLUS UV-A IN LYMP HOID CELLS 87 75, Popesc NC, Amsbagh SC, DiPaolo JA: The relevance of caffeine post-treatment to SCE incidence indced in Chinnse ham ster cells. Mtat Res 6:313-32, Carter DM, Wolff K, Schnedl W: 8-methoxypsoralen and UV-A promote sister chromatid exchanges. J Invest Dermatol 67: , Lambert B, Morad M, Bredberg A, Sanbeck G, Thyresson-Hok M: Sister chromatid exchanges in lymphocytes from p oriasis patients treated i th 8-methoxypsoralen and longave ltraviolet ligh t. Acta Dermatovener (Stockh) 58: 13-16, 1978 chromosomes. Hm Genet 38: , Wa ksvik H, Brogger A, Stene J : Psoralen/ UV A treatment and 35. Wolf-Schreiner EC, Carter DM, Scharzacher HG, Wolf K: Sister chromatid exchanges in photochemotherapy. J Invest Dermatol 36. Faed MJW, Morelatos : Enhancement by caffeine of sister 69: , 1977 chromatid exchange freqency in lymphocytes from normal sb 37. Honeycomb JR: The effects of bslphan on the chromosomes of jects after treatment by mtagens. Mta Res 49:437-44, 1978 normal hman lymphocytes. Mtat Res 57:35-49, Brger PM, Simons J : Mtagenicity of 8-methoxypsoralen and long-ave ltraviolet irradiation in diploid hman skin fi broblasts: An improved risk estimate in photochemotherapy. Mtat Res 63: , Strass GH, Albertini RJ, Krsinski PA, Baghman RD: 6-thioganine resistant peripheral blood lymphocytes in hmans folloing psora len, long-ave ltraviolet light (PUV A) therapy. J Invest Dermatol 73: , 1979