REGU LAK AKTIC LES

Direct Correlation Between DNA Repair Capacity and Metastatic Potential of K-1735 Murine Melanoma Cells

Qingyi Wei, * Lie C heng, * Keping Xie,"j" Corazon D . BlIcana;j" and ZhongYlIll Dong"/" Departlllell ts of ' Ep i dellliolo~)' alld ,'Cell Uio l o~y. Th e Ulli\,\;rsity of Texas M. D. Anderson Cancer Center. Houston. Texas. U.S.A.

The purpose of this study was to determine whether the ability ofK-1735 murine melanoma cells to repair DNA damage correlates with their metastatic poten­tial. Three nonluetastatic clones, four metastatic clones, and three sonlatic-cell hybrids between met­astatic and nonnletastatic clones were exposed to incident ultraviolet (UV) light (254 nm). Cell survival was determined by the microculture tetrazolium as­say, which measures cell metabolic activity. DNA repair capacity was detertuined with a host-cell reac­tivation assay, which nleasures the activities of chlor­amphenicol acetyltransferase encoded by the re­porter gene in both UV -daluaged and undamaged

It is estimated that th e re will be 7,30n dea ths f.·01ll ll1elanonw and 38,300 new case~ in the United States in 1996 (Am er­ican C ancer Society. 1996) . The maj or cause o f death fi'om m e lanoma is m e tastases t hat are res istant to conven tiona l therapies (Auk erman and Fidler, 199 1) . By the time of

diagnosis , melan omas contain multiple ce ll populations with di ­verse g ro wth ra tes, karyotypes , cell-surface properties , antigenic ity. iJllm unogenic ity. abili ty to invade and metastasize , and sensitivi ty to various cytotoxi c drugs (Fidle r , 1978; 1985; Nico lso n. 1987). The heterogeneous respon se of primary tumors and m etastases to c h e motherapeutic agents is well recogni zed (Trope , 1975) and is due to modulation of drug response by host f.'lCtors (Fidl er, 1990 ; S impson-Herren cl nl, 1987) and to the intrinsi c resistan ce of tu m o r ce ll s (Weichselbaum cl ai , 1985). whi ch may be associa ted with e nh an ced DNA repa ir capacity (Nie 1:1 ai, 1995).

T here is now a large body of data sh owing th at neopla stic h e terogene ity ma y be associated w ith amplifi cation. rearrange m enr, a nd point mutation of va rious genes (Bishop. 1991) includin g /VAF11p2 1 (Vidal cf ai, 1995) and /iS3 (Albino cl ai, 1995; Sparrow

c l ai, 1995), which were iden t i ~icd in m e lan o m as. Kecen t data indicate th at humall m e lanoma ce ll lin es iso lated from m etastatic lesions di spla y in c reased resistan ce to killin g hy ultraviolet (UV) radiatio n because of an enhan ced ra te of post-repli ca tion recovery a nd 3n overa ll inc rease in photoproduct repa ir (Hatton cl ai, 1995) b y th e nucleo tide-ex cision repair system (Fri edberg. '1985). Be­c ause nu cleotide-exc ision repair removes damaged DNA, such as that CI'oss-linked by many anti cancer dl'llgs (San car, 1994) , it has

Malluscript received Mal' 6, 199(,; rev ised August I~ . 199(,; accepted rnr p ublica tioll Septelllber 5. 1996 .

Reprillt requests to: Dr. Qi ll ~y i Wei , Departmcllt or EpidclIlio logy, 130:< 189. Div isio n of Cance r Preve lltion, T h e Unive rsity o f Texas M . D .

Anderso n Cancer CClHer. 1515 Holcombe Bouleva rd, Houston. TX 770.,0.

plasmid (control) pCMVca.t 40 h after transfection. No discernible differences in transfection efficiencies were found between K-1735 clones with low and high metastatic potential or between cells transfected with UV -damaged and control plasmids. DNA repair ca­pacity directly correlated with cell survival (p < 0.05) and with metastatic potential in the K-1735 clones and somatic cell hybrids (p < 0.05). These data sug­gest that the intrinsic resistance of melanolua nletas­tases to systemic chemotherapy may be due, in part, to the cells' enhanced DNA repair capacity. Kq I11(JI'ds: DNA Jlepa idllletastasis!lIlela.llollla. ] ['Illest Del"lllatol

108:3-6, 1997

been implica ted in in creased resistall ce of tu m o r cells to che mo­th e rapy (Gclez iunas r l nl, '199 1; Dabho lknr rl ai, 1992) . Overall. cellula r DNA repai r ca pa city, whi ch consists large ly of nuclcotide­excision repair. can be measured by a host- cell reactivation (Protic­Sab ljic and Kraem er, 1986; Ath as el ai , 199 1) system that m easures the relative express ioll of a repaired Ilollrep li cati ng plasmid harbo r­in g a chlo ramphenicol acetyltransferase (CAT) reporter gene that is dama ged by UV radiatio n (Wei 1'1 ai, I ()93).

In murin e system s, the cells in m e tastases are m o re res istant to systemic che motherapy than are cell s in the parema l n eoplasm (Tsuruo and Fidler. 1981: Fidler, 1990: Fid le r (' I nl. 1994). W h ereas the o rgan envirollment call m odu late the response of tumor cells to chem otherapy (Fid le r. 1985: Dong cl nl. 1994a). w e h ypothesized that metasta tic ce ll s ma y also res ist chcmothe rapy by having an enhan ced overa ll cellular DNA re pair capacity. W e tes ted this h ypothesis by usin g th e K-1 73S murine m e lanoma sys te m of clones and so m atic ce ll hvbrids . which have stable and well-defined m c tastati c ch aracte l~i s ti cs. and report here a direc t corre latio n between cellul ar DNA repair capacity and metastatic po te mi a!.

MATERJ ALS AND METHODS

Cell Lines and Cell Culture The o ri ~iJlal K-1735 melanoma, which was inducco in a C3 H/ HeN mouse by exposure 1'0 UV li !,:hr f") \l owed by pa itHin~ with croran oil. was a ~itt fi'om Dr. Mar!,:aret L. Kripke (Depart­ment of llI1mLIIlology. The University of Texas M.D. Anderson Cancer Center. Houston. TX) (I<ripke. 1979) . T he parelltal rumor was cloned ;11 ";/1'0 by a double-di lu tion lIIethod (F idler. 1995). Of a large !Illlllber of cion", rhus iso la red, clones C-J. C- I n. C- 19. and C-23 w"re class ified as 1I0llmetastaric or slightly mcrasratic (F idler ,'r III . 19t; I). Clone C-4 is hi~hl r Illctastatic and produces 1I1l~ lal1ocy ti c t llltl o r foci ill the lungs of syngencic Inicc. Hig hl y 111 ctas tati c c10lles 1\11- 2 and X-2 1 wcn .. ~ d CI;vcd tro lll indepen­

dCllt solirary, ' I olltaneous lung mClilsrases produced by subcutaneous injecti on of the K- 1735 parenra llin c into nude mice (Talmadge and Fidler. 1(82), and hoth M-2 ;llld X-2 1 were shown to be clonal III orig in

0022-202X / 97 IS I n. So • Copyright il,l 1997 by The Society fo r Ill ves tigati ve Dermatology. In <.

3

4 WEI ET AL THE JOURNAL OF INVESTIGATIVE DEI~MATOlOG\

,oo~~--~~~~-----------------,

Figure 1. UV survival and DNA repair ca­pacity in selected K-1735 murine melanoma clones with different metastatic potentials. M-2, M-2/C-23 . and C - 23 represent high. inter­mediate. ,lIld low lJ1etilstatiC potential , respec­tively. (A) UV surv ival cu rves. C -23 cells showed significantl y lower UV wrv iva l than bo th M-2 and M-2/C-23 cells. (B) Pbsmid survival (ONA repair capaciey) c urve. The meas ured DNA repair capac­ity was positi vel y correlatcd w ith metastatic po­tential s in 1<-1735 ce lls. (Noee thac ehe log scales on the Y axes o f (rt) and (8) arc different).

iO .~ ~ =>

U)

C .. ~ Cl>

Q.

10

_ M·2

-0- M·2IC·23 --0- C·23

200 400

B

600 BOO 100 0

UV Dose to the Cells (J i m ' ) UV Dose to the Plasm id (J i m ' )

(Talmadge CI III. ! 982). The 1<-173 5 somatic ce il hybrids C-23H /C-23N. C-23 H / M-2N. and M-2H / M-2N were produced by fusion as described in detail previo ll sly (Starosclsky el III. 199 1). Tlte m c tastatic pote ntials of the seven K-1735 clones and three somatic-ce ll hy brids Were further charac­terized experime ntally by Dong el III (1994b).

The ce ll lines were used at low passage numbers (passages 6-10) . and cultures \lvcrc not: 111 a iJJ t:Jincd lo nger than 3 wk after recovery FrOl11 frozen stocks. Norm al skin fibroblasts from m ouse embryos (F, offspring of BALB/c and C3 H./ H e N mice) and UV20. a CHO lin e (C1U1862, American Type C ulture Colkctio ll. Roc!;ville . MD) de fi cient in repair of UV damage (Hoy ,'r al. 1985 ), Were uscd as a base lille (repair-proficient) control and a positi ve (rcpair-ddicient) control . respective ly. The mcla-11 0 111;.1 ce ll s and 11111rin e. fibroblasts were Inainta incd ill culture in Eagle' s minimal cssenti.,1 medium supplemellted with to'Yu feta l bov i" e se rum . sodium pyruvate. nonessential amino acids. L-glutamine . and 2 X vitamin so illti on (Flow Labo ratol'y, Roc!;vill e . MO) . VV20 cells were cultured in a -minimal essclltia l medium w itho ut ribonucleotides ,md deoxyribonucle­o tidcs . The ce ll cultures ,.vcrc Inai n ra inc d all pJastic and incubated in ;:1 5'Yo CO

2 atmosphere at 37°C. T he cul tllres were free of M)'(tJpla.wlfI and the

following tl'lurinc.: viruses : Senda i virus, 1l1illllCC virus. 1110USC adC IlO V!r ll S,

J)1 0 USC 11 epatiris v irll s, lymph ocyti c ch oriolTIcn ingltis virus, c ctro n1 c li a virus. and lactate deh yd rogenase virus (as assayed by M.A. Bioprodtlcts. Walk­

e rsville . MO).

UV Survival Assay The microculture tc tra zo liul11 assay (Mosmann , 1983) was used to m e;\SUtc ce ll density as an indicato r o f UV surv ival. BrieR y. the vi:,bl e cells were seeded into a 96-well plate at a density of 200 celi s per well pc r 100/1.1 of l11 cdiulJl . Twellry-four ho urs lawr. the ce lls were irradiated with a 15-w Wltlltcred germicidal lamp (G1STS G I-15; Sank yo Oenki Co .• Ltd .. Tokyo . Japan) ,n doses of O. 20, 4(). 6(1. or 80 .II III ° in the same uncove red plates. The radiation SOllrCe produced all incident (254 n111 ) UV dose of 2 J per m 2 per s as detcnnined with a UVX radiometer (UVP, Inc .• San Gabriel. CA). After iHadiatio ll, 100 fl.l t) f fresh minimal essential 111cdjunl were ;,lddcd to each we ll . The c ultures \,yere thcll I'lwilltain cd at 51% CO , :';1' at JJOC for ')6 h. D uring the final 2 h . 40 fl.1 o ftetrazo liul1l (5 mg per -1111 in phosph,1te-bu!fe rcd sa line) (M 2 128; Sigma .Chemical Co ., Sc. Louis. MO) wcre "dded co each well. Af(er re m ov," o f unreacted d ye and (he medium. dimethyl sulfoxide ( 100 f..L1) was added to e'lch well. In this assay. 3_(4 .5_dimethylthiazol-2-yl)-2.5-dihel1yltetrazolillnl bromide is COn­verted to forl11 aza n b y metabolicall y viable ce ll s, and the absorbance at 570 I1Ill is m easured with a mjcro plate rcader (MR500; Dynatcch. Inc .. C han­tilly. VA). Cell s"rvival "vas detennincd from the formula: Survival C'!f.,) = AlB X 100, whe re A and [I arC the OD,1" of the tfeated alld control cells.

respectively.

Host-Cell Reactivation Assay Cellular DNA repair capacity was lllea­sured by the h ost-cell reactiVatio n <lssay as previollsly described (Athas ('I "I, 199 1; Wei "I ,,/. 1993). "Two milliliters of pCMVcnr (50 f..Lg pcr ml Tris(h yd roxymethyl)aminOnletlull1e-etilyle"cdi(lll1ine-tetraace tic acid) were placed en each of 35-111111 ti ~s lle clll ture di shes . The di shes were 1I11co;ered and separately irradiated WIth 0, 200. 400, or 800 J UV lIght per m- at a wave length of 254 nlll. These tre<ltlTlents were performed on one batch for all of the CAT assays. After trc;ltme llt. the conformation o f the plasmids Was assessed by 0.8 ,){, aga rose ge l electrophoresis. Treated and untn:atcd plas­mids (controls) were trans f<'.c ted into the ce ll s by the diethylarninocth yl­dextran procedure (Athas 1'I "I. 1991). T he rransfeetcd cell s were incllbated for 40 h to allow repair and gene expression. and the CA T <l cti v icy in ce ll extracts was then estimated by m easuring the radioactivity from the [J l-ljmo noace tylc hlorampllel)ico l and r' Hldiaccrylchlo ramphenicol lo rmed

by re~c tiol1 with the AT en zyme. after repeated extractio ns for remova l 0, free [. H lace t)'!. Thc ,,,,crage CA.,. actl vlty for the tnphcates was calc lllatc <,\ 'IS th e percenta ge of residual rcpair acti vity (percent CAT <l c ri"ity) t~ cstinl :'u c th e DNA repair c:l paciry at n given dose. The cS pCri1)lCJHS \\'er~ repeated three tinl es. and the averages \Ve rc used for statisti ca l co rnpariso l1 s ,

Transfection Efficiency A modified Auorescen ce ill Silll hybridizatio'\ procedure (\'inkel <I af. ·1986) "''' used to meaSure the transfcc(io" efri ciency by using pCMV CtfI labeled w irh digoxigenin b y ni ck tr 'll1 slatio n a& the probe and visllalizing by hyhridization rhodamine staining. BrieRy, the, bbelcd prohe (1 f.J.g) W<l S concenrrated with a Specdvac (Sava nt lnstru, me nts. Inc. . Farmingdale, NY) fo r I h and then dissolved before being completely dried in 5 f..L1 of dOllble-disrill ed water and stored at - 20°C fo l ,

later usc. The trans!cc ted ce ll s were grown on 100 X 2(J-mll1 cu lrure di shes. Before hybridization. the cell s were trypsinized ;ll1d then attached to (i'osted slides by low-speed centrifugation (600 rpm) with a Cytospin 3 (Shandon. Inc ., Pittsburgh , PAl. The slides w erC d ehydrated with a sel;es of e thanols (75'){,. 80%. 95%. and 1 OO'!/" each for 2 min). denatured with 70% (vol/vol) formal11ide/2 X standard sodium citrate at 75°C for 2 min , and dehydrated again with a series of ethanols (75'%. 80%. 95')1" . and 10M{, e"ch For 2 min) at - 20°C instead of the uSllal 4"C, whic h significandy inc reased h ybridiz a_ tion efficien cy. For h ybridization . 10 f.J.1 of probe cocktail containing 225 ng of DNA were denatllred at 75 °C fo r 2 min and then added to the slides on th e arca containing cells. The cell area was covered with a 2S X 25-nlln glnss coverslip , scaled with rubber cement , and incubated in " prcw,ml1ed humidified chamber at 37"C for 20 h. After hybridiza tion , the slides We r e was hed three times at 4 I °C for 5 min each time with 50% formamid e/2 X standard sodium citrate. and cwice af 4'1 °C fo r 5 min each rim" with 2 X standard sodium citrate with occasio nal shakin g and che ll washed with vigorou s shakjng twice at 41 °C tor 2 min each time with 0 .2 X standard sodium citrate. Oigox-igcnin-Iabeled probes were then d etected with a digoxigenin-rhodamine detecti on kit (SI 332-0R.) by fo llowing the proce­dures provid ed by rh e manufacturer (Oncor. Gaithersburg. MD). and the slides weTe \v:1shcd three tinlCS for 2 n1in each rinlc \vith sodill1U bicarbo nate and Tween 20 buffer. pH S.O. T h" coun tcrs(;ljn was perfOTln"d with 4.6-dial1lidilllo- 2-phenylindo le (Oncor) . The slides were then ex"m ined using a Z.e iss Photomi croscope equipped with a merc ury UV lalllp (Carl Zeiss, Inc .• Thernwood. NY). The imagc ~ wen.: digitized with an Optin)"s llllages Analysis Sys te m (Optimas Corp .. Bo thell . W A). The rransfectioll cfriciellcy was calculated as the percentage of the number o f positi vely stained cells in 300 cell s (l OOlrandolll field) examined microscopically.

Statistical Analysis The significance of ti lL' data were determined by Student's t test (two-tailed) and the Pearso ll correlatio n a,,,d ysis.

RESULTS

Inverse Correlation Between UV Survival and DNA Repair Capacity In the first set of experiments, we measured the UV surviva l and DNA repair capacity of one highly metastatic clone (M-2). one nonmetastatic clone (C-23), and a somatic cell hybrid be tween these cwo clones (M-2/C-23). As shown in Fig lA, irradiation W1th 20 - 80 J per 111 2 reduced survival of all three sllb lin cs of cell s in a dose-dependent manner. Com pared with M-2/C-23 (slightly m etastatic) and M-2 (highly mctastiltic) , the nonmetastatic clone C-23 had a signifi cantly lower UV surviv al at ali doses. The DNA repair curve (Fig IB) revealed a co nsistent

VOL. 108 . NO. I JANUARY 1997

tre nd: th e higher the metas ta ti c po tential , the better DNA repair o f UV damage.

Similar Transfection Efficiencies in Cells with Different Metastatic Potentials To verity that the observed diffe ren ces in DNA repair capacity were not due to diffe rences in transfection effic iencies, we perfo rm ed a m odified Au oresce nce ill sitll hybrid­ization usin g pCM Y cat as a pro be ro measure the transfection effic iency of M-2, M-2/C-23 , and C -23 ce ll s. The results, shown in Fig 2 , demonstrate that the trallSfection effi ciencies were similar fo r all three clones M-2, M-2 /C -23, and C -23 (75'1." 77%, and 68'X., respectively) for ullirradiated and (70%, 80%. and 71 %, respec­tive ly) for irradiated plasmids.

Inverse Correlation Between Metastatic Potentials and DNA Repair Capacity In the second se t of experiments, we eva luated the DNA repa ir capacity of 7 K-1735 clones with different me tasta tic potential s. In these studies, we a lso used primary m o use embryo fibro blasts as a baseline (negative) control, and th e CHO mutant UY20 , which is known to be deficient in excisio n repair, as a positive contro l. The data are su mmarized in Table I. The repair-defi cient UV20 cells had a signifi ca ntly lo wer DNA repair ca pacity th an did no rm al embl'Yo fibrob las ts (p <

A E

B F

c G

o H

Figure 2. FISH of the transfected plasmid pCMV ctl l in K-1735 murine melanoma cells. (A) and (E) lIn trall sfecred ce ll s showing on ly blue nuclei without Auoresccnce; (8) and (F) M-2 ce lls transfected with un irrad ia ted and irrad iated (800 J UV per Ill ' ) pl asm id. respectively. (C) and (G) M-2/C-23 hrb rid ce ll s transfected with unirradiatcd and irradiated (800 J UV perm ') plasmid. respectively. (0) and (H ) C-23 ce lls transfected with unirradiated and irradiated (800 J UV per m2

) plas mid, respec ti ve ly. T here \vere no substanti al differences in the transrection efficiencies with un irra­diatcd or irrad iated pl"sl11ids or of cells with different metast:ltic potenti"ls. Sca le bar (50 J.L 1ll ) is the same fo r each picture.

DNA REPA IR AND MELANOMA METASTASIS 5

Table I. Increased DNA Repair Capacity Correlates with High Metastatic Potential in K-1735 Murine Melanoma

Cells"

Cell Line

R.epair proficient Normal skin

fi broblasts R.epa ir deficient

UV20 NOll111ctastati c clones

C-3 C-I0 C-1 9 C- 23 (group l11e::1I1

I';

df = 13) Metastatic clones

M-2 C-4 X-2 1 (group l11ean":

df = 9) Somatic-ce ll hybrids

M-2 / M-2 M-2/C-23 C-23/C-23

DNA Repair Capacity ('X,)

Assay Assay 2 Assay 3 Mean :!: SD p value"

18.1

3.3

12.7 17.3 32.7 32.6

98 .1 69.5 92 .7

105.6 42.7 28.2

14.7

3.0

15.9 20.5 41.8 27.5

86.7 68.2 96.0

88.1 29.5 31..1

17.8

6.8

15.6 29.5 35.5 22.1

97.0 60 .0 94.7

90.3 46.7 16.7

16.9 :!: 1.9

4.4 :!: 2. 1

14 .7:!: 1.8 22.4 :!: 6.3 36.7:!:4.7 27.4 :!: 5.3

0.002

0.226 0.218 0.002 0.03 1

(25.3 :!: 9.2) (0. 150)

93.9 :!: 6.3 65 .9 :!: 5.2 94. 5 :!: 1.7

0.000 0.000 o.oon

(84 .8 :!: 16.3) (0.000)

94.7 :!: 9.5 39.6 :!: 9.0 25.4 :!: 7.7

Q.OOO 0.013 0. 136

"The resul ts ~rc exprcs!'cd as the pcrcc nwgc o f DNA repair capacity of UV-damaged (800 J lln1

) pCMVr(/f relati ve CO un dam aged plasmid ( 100%1). J. Two-sided ( tests. ( Reference for a ll the statistical comparisons unless otherwise stated. ,/ The I [est [or these twO group I1lc:ms (m etastatic and 11011IIl ct;')staric) WilS performed

ll sing e~ch experi ment for e~l ch ciolle (three expe riments per done) as one observation. w hi ch gave a p value of less than 0.00 I w ith 19 degrees of freedo m (df).

0.0 1). As a g l'O UP, the nonm etasta tic cells did n o t have a signi fi­cantly greater DNA repair capac ity than no rmal embryo fibroblasts. T he DNA repair capa city of the m etas tatic K-1735 clon es (M-2, C-4, and X - 21), however, was signifi cantly g reater than that of th e nonmetas tatic K- 1735 clones (C-3, C-I O, C-19, and C - 23) (p < 0.001) and th e no rmal fibroblasts (p < 0.001).

We also examined three somati c cell hybl·ids. The highly meta­static hybrid M-2/M- 2 had a signifi ca ntl y greater DNA repa ir capacity than no rmal embryo fibroblasts (p < 0.00 1) . The hybrid M-2/C-23 h ad a low-to-intermediatc metastatic potential and intermediate DNA repair capacity. T he DNA repa ir ca pa city of the nonm etastatic hybrid C -23/C-23 was similar to that of no rm al embryo fibro blasts and no nmetasta ti c c lo nes (Table I) .

DISCUSSION

T h e data reported here dem onstrate that highly mctastatic me la­no m a cells had a higher level of DNA repai r ca pacity than me lanoma ce lls with low metastatic potential. Because rum ors are highly he terogeneo us (Fidler, 1978, 1990), clonal analys is m ay be essential for dem o nstrating differences in the DNA repair capacity of metastatic and nonmetastatic cells. which may accoun t fo r the d iscrepa ncy bet',.vecn o ur resul ts and those of other studies that used no nclonal cell lines (Chalmers et ai, 1976; Good l' t a/, 1978; Musk and Parons, 1987).

R esults Ii-olll other in vestigators a lso suggest that enh anced DNA repair of UY-induced photoprodllcts was one of the underlying mechanism s fo r the UY -resistan t phenotype of the human meta­static melanoma ce ll line R YH421, which has twice th e no rm al resistan ce to UY. perhaps because of improved post-replication recovery (Hatton el ai, 1995). En hancement of UV resistance in hum an m etastatic melanoma cells has also been reported (Good ct ai, 1978; Konishi , 1981).

T he K-1735 melanom a, £i'om which the K-1735 cell lines were derived, arose in a C3 H / H eN mOllse exposcd to chronic UY radia tion and then painted with croron o il (Fidler, 1995). Such

6 WEI ET AI.

chro ni c expos ure can lead to mu tations (Anantba swa m y and Pierceall, 1990), and it is li kely that red uced DNA repair ca pa city for repair ofUV damage to DNA would increase the probabi li ty of carcinogenesis (Wei el nl, 1993, 1996). T he evolu tio n and progres­sion of m elan oma , however , is al so probab ly associated with an incre:lsed rate of mutation . In humans, it ha s been demonstrated that the percentage of 1153 mutations was higher in m etastatic les io ns (70'1.,) than in primary tum ors (35%) (S parrow cl (/ 1, 1995). One explan ation for thi s m ay be that metastatic m e lano ma cells develop a hi gher DNA repai r capacity, which ofFers surviva l advantage and a hi gher to lerance to mutations, such as those in p53, w hich leads to a growth advantage. T hi s is co nsisten t w ith the clonal expa nsio n theo ry, w hi ch predicts that a clone with hi gh DN A repa ir ca pacity and hi gh to lera nce of mutations in tum o r suppressor genes co uld ha ve a growth advantage (N ico lson,1 987) . T um or progression, however, ca n al so occur as a conseq uence of increa sed genomic in sta bili ty (Cifone and Fidler, 1981; Tsu ru o and Fidl er, 198J; Nowell, 1986; Usmani ('/ (/1, 1993), which may be re lated to defi cien t mism atch repair (Mod ri ch, 1994) . O ur find ing tha t metastati c ce ll s ha ve greater DNA repai r ca pacity than non­m etasta tic cell s ma y appear to con tradi ct thi s co ncept. T he host- ce ll reactivation assay, however, large ly measures nu cleotide-excisio n repail·, because the p lasmid docs not rep li ca te (A th as cl nl, 1991), and. mismatch repair usually occurs after rep li catio n (Modrich , 1994). Moreover, genomic instabili ty ma y no t be the o nl y conse­quence of defi cient mi smatch repa ir bu t m ay be re lated to the fide li ty of I·epair rather than to muta tio nal freque ncy in metastati c cells (Usmani, 1993; Usmani ct (/1,1993). Enhan ced overall DNA repa ir activity, coupled with poor mi smatch repair and poor fid e li ty of re pair , co ul d therefore cont ribu te to the geno mic in stabi li ty (Peris el (/1, 1995) that m ay be associated w ith mutations that lead to tumor progressio n.

Altho ugh the exact underlying molecular mechanisms of the en­hanced DNA repa ir capacity observed in metastatic melanoma cells are not known, it may be important to evaluate the DNA repair capacity of melanoma cells (especia lly metastatic ceUs) to develop successfu l treatment, because enhanced DNA repair may also be associa ted with in creased resistance to chemoth erapy (N ie et nl, 1995) .

/tI1(' ",ish 10 1IIIIIIk Dr. Isainh Fidlcrji.,. his ,·IIC<W""gCllletll ill Ih" illilialilJ lI 4 1his

.,,"fldy (/1ll1 crith'a l n~IJ ;r'" I!f lite dnta m,d lIu",,,srr;pl , I) ,.. Lm l/n'lIre' C r{lSS l/lall ji'I" Iholl.eh!li.! dis(IIss inll n",1 «/111111 (' Ills , Dn. R,."hcII LOlnll alld MfII~~fII"" Slli lz./i'rlhl'ir

(r;l;enl ITllj(' IfI , Dr, /Vlm /H'l'" Goode (D l'/wr'lIIclIl '!J" Sd(,lIfUic PlIlJlirilf ;olls) .fi,,· sriclIl Ulr editing, AI/s . .. '(i"/I11/ Q. Yn ll,<! .Ii,r h{'r exce fl t'" , fer/l11irn1 nssistfll1fC, aud 1\4s.

LO/tl Lopez (J lld I\I/s. Sherr), Pmflcll je". their {':rpiT' ((ss is/nl/ f(' ill /1rt '/u lrill,t! ,he

"'flllI'S(l'il'l. T his ",ork ",ns slIl'/",rl('(/ ill I'" rl h), Crn llis R29CA 7033 -1 fw d

IU5CA 42 107 ji-O lll lit" Nnliollnl Caliri.,. IIISliIIlIC, Nnliollal IlISlillll,'S of H enllh .

REFERENCES

Albillo AI'. Vidal M.I. McNutt NS. Shea CR. Prie lo VG. Na llus DM , Pa lmer J M, Hayward NI{: Mutalion ~lI l(i c.x p ress ion or p5 . ~ ~cn c in hU1I1an m al ig na nt 1I1ch1l1oma. iVIl'lmlOlI/tJ Res 4:35- -15. 11)1)5

A n :lI llhasw i'lI11 y H. Pi erce;1 1I WE: Molecu la r mcc han ism s of ul t ra v iolet rad iation

c;ucill ogc ll cs is. PIWfclr/lfllf PhIl/obit" 52: III Y- 11 36, 19 1)0 Amcrica n C;lI1 ce r Socicty: C(1I1t1'I' Sfalisf ;rs. Amc rican C;:mcer Society. A llallta. GA

1'1% Adw s A t:: . H cdayali M. Malil l1 0ski G M . Fan l1 er E I ~. Gross m all L: J)cve lop llI CIH and

ti e ld-test va lid at'iorl or ;111 :Iss:!y for DNA re pnir ill circul ari ng human IYlll p llO­

cyrcs. Cflll(('r R,.s 5 1:57R6-5793. I f)9 1

Auk~rT1lall SL. Fid ler IJ : T ile heterogen cou s 11al'llrc Ofl11ct:lstaLic ncop iasill s: rclev:lll cl' to hi oril er<lpy. In : O ld h a lll Ill< (cd.). }Jri ll r il,lcs (!f" C alm:,. UiOllwffI/')'. Marce l

Dekker. New York . 1~91. 1'1' 23-53 Bisho p JM : M o lccula r th c m es ill o n coge n e!l is. Cell &·k235-24~. 199 1

C halme rs A I-! . Lav in M. Atio nr o rnkul S. M:Ulsbridge .J , K in soll C: Rxsis r;lI1 cc o flllllll :J n 111e1:1110 Ill il ('c ll s 10 r<ldi .u-ion . Cftll(('r Res :\G: I 1):\0- 1934. I1J7(1

Cil~)f1c M. Fidler IJ. III .. Tc:ls in g IIletaSla li<: po t c l1ti~ 1 is assoc i:1tcd with il1creasing gC l1 ctic insl,lbili lY 0(' c10llcs isola le d fro 111 Illllrin e neopla sm s. Prot N II,I i l(ad Sri US/J 7H:694'1-(,952. I 'IH I

Dahho lkar M. BOSlick-l3 ru ron F. Wchcr C. 1I0hr V, E!:wlI af! u C. R eed E: ERCCI and ERCC2 cxprc\sio ll ill m .dig l1 ;11lI tisslle li·mll (lv'lri a l1 c an Cer p ;1I ic n ts . .J N ntl C(/ l/nT

t IIsl H4: I 5 /2-1 5 17. I ~n

T H E JOUItNAL OF INVESTIGATIV E DEItMATOLOGY

DOllg Z. Iladinsk y It... Fa n I). T"all R. BlIC:1 lla CD . \Xl ill11:1lln s C. Fidlc r IJ : Orga ll -spcc· il-ic m o dulatio n of src:1dy-statc IIIdr- 1 ge ne cxprcs"illll a11d d r ll ~

rcs isra ll Cc ill IIll1ri ll C culon c;m ce r cells. J N IIII ClllltlT IIISI Hfi:lJ 13-920. 11)t)4 a Dung Z. S tarosclsky A H . Q i X. Xic K. Fid ler IJ : In versc correlation hctweel1

cx pn:so;ion nfi nrl ucihlt: nitri c ox id e sYlithasc aniv i!"y an d produ c tio n o flllcta sra sis ill 1<- 1735 lI'lurin e m e la no m a ce ll :;. e m/f l '" I<('s 54:7H9-793. 1( 1) -lh

Fid le r IJ. Cruys MA. C i'<H1e MA. B ;l fll CS Z . Bu ca na C: DC llI o1l.;rr;lIio n u t'l1Iulti p le

pilcl1l.' t )'pic divcrsity ill ;1 Illurin c l11 e la no l11a or rece n r o ri g in . I ,\(lfl e dlin, l 'I ., f

67:9·17-95(,. I 'IH I Fidlcr IJ . W illll il llll S C. S taroscisky A . n ... ad in .;ky R. Doug Z . F~111 I): Modul at io n u f

tUlll n .. c e ll respOII SL' Ch l! lll nr herapy hy I' h l' Il rg: 111 c ll v irllllllll!lIt . em/!'t.,. }\/I·'lIsr,'Si ...

ReI' 13:209 -222. 1994 Fid ler I.J: C ri tiGIi [;Icwrs ill th c hiology ur hUlllan GlIIccr Ill elastasi!l . CallffT J{{',.;

50:(, 130-(.1 38. 1'1911

Fid lc r IJ : M ,\(:rnph ag:c:-. a n d III t.:taSl:lseS - :I bio logic;11 'lppn)ach 10 cam:e r th l!rap y. Calln". Res 45:-17 14 - 472(,. I 'IXS

Fidler IJ: M odula tiOIl o r th c org:1I1 lI1i c ruell vi ron lllc nt fur th e trC;It IllClll 0 1' l·a l1 l't .. T

1Ilclas tas is (CO l1l1l1 c lllar),) . .J i\'1I 11 C llllft'r /11 51 t)7: 15H~ - 1 5c)2. 1 C)')5

r=idlc r IJ : TUllIor he tc roge neit y alld til c b io logy oj' can ce r in vas io n :Jn d m e Lastasis. COlin". Re.' 38:2(,, 1-2(>(,0. 1'I7H

Friedbcrg EC: DN rl Uf/",ir, Frcccl lll an :In d Compan y. New York. 19H5

Gel ezi ulla ~ It .. McQl1i ll ;1I1 A. Ma lapc rs:1 A. Hu tt.: h insu ll M . Ku pri va D . Wai nherg MA. Hi!lcon J, Bra l1lson J. Pan as<.: i L: In crc;t:,ed I )NA ';Y lllh~" i !l ,l1lel repair-e l1 zYIll l!

ex pre ssion in IYlllph ocytes Ji·(l 111 J1i1rielll ~ wit h c hro nic Iymph ocyric le l1 kL'llli :1 rcs ist:1llt to lI itr()~c l1 111t lstarci s. J Narl CtlHlTr IIISf B3:5S7-5(1-1. 19 1) 1

Good M. L:1vi ll M. C h c lI P. K idsoll C: Depcndencc 0 11 d Ollin g llI e rillld or :-' lIrvi va l o f hum a n lIl C I ~III Ullla cc lls ancr ultra v io lct alld ioniz.in g rad ia tio n . C (II/(er R es .'X:4671-467'i. I97H

H a ll'ol1 D H . M itche ll DL. StI'ick lalld PT. J o hllson RT: Enhan ced p ilulOp ro till t.' l' rc pair:

its ro le in th e DNA d amaJ.!e-rcsi~ta ll cc ph c1lotypc of' hul1l an l1Ialig nant Ill c \;\n n llla cc ll s. C(l l/ca J{f'S 5S: I C; I- I tW • . 11)1)5

H oy CA . T IH)III PSOll LH. Moo nc)' CL. ':1i:12ar EP: Defectivc DNA uo'\s- lin k re lllo\·a l

in C hin ese h :lIl1sre r ccll IIHl ta 1l1 h ypcrse lls ili vc to hiflillcdo nal :dkylatiOI1 agcll ts. COH,.,.,. R,.s 45: 1737- 1743. I~ ll'i

I{l· ipkl: MJ..: Spec ulatioll s 0 11 dI e ro l ~ of ul rr:1vio lcc radiation ill the develop m c ll t or

1I1 a lig nant m e lano ll1a . J NaI l C m/('r IIrst 63:54 1-5·18. 1979 Modrich P: M isll l;'1tcil l'ellair. gC llCriC stab ili ry. ;111(\ ClIIcer. Sdt'llff 260: 1959- 1 C) (j( I. 199-1

M o .. nwlll1 T: Rapid colurimctric assar fl)r ce llu);lr growt h ~Ild surviv:1l: appli..;::tl'iol1 t o pro l i{l::r~tion :I nd cytotoxic it), :lS:-'~)'S. J 1,,11,"1111 ,1 J\lcl llc,ds (15:55-(,3, I YS:\

Musk P. Parolls PC: R esistan cc o f' pi~IIlC I HCd human ccll s 10 k illin g by sunli g la ~l1ci oxygen r~l d i ca l. IJ/' ct/od' /' /11 P/JclI(lhi,,/ 11 6:4~)c)--I1)4. 1 1)~7

Nicolson GL: TU lIlo r ce ll in stah ilit y. di vcrsi l·icatio ll . and pro~ressi\lll t'O thc 111e rastacic p llCllntype: IrOll 1 OI ICOgCII L'. to OI lco lcla l eXllrcssiu ll . e ll//re.,. T<t~~ 4 7: 14 73- 1-I ~7. 1987

Nie Z R , Paterso n J . Al l'en L. Tsao MS .. Vi all et.l. Alaoui-Jama li MA: Elcvated D NA rcp~lir c~pacity is ~ssocii1tcd wilh illtrin sic re sista llcc o r lu ng c an cc l' to ch c ll H)_ rh e rapy. e m/ft,,. J{"J SS:47GD-47(14. I I)();

Nowcll PC: Mec h alliSlil S of tU11Ior progrcssion . e (lli fer R f'.,· -1(1:2203-2207. I c)~(, Pcris K, Ke ller G, C hilll c ll t i S. A IlWIlt'ca A. Ked I-! . 1-1 6 J1 cr H : M1crosat c lli rc ill srabilirv

;1110 loss of he Lc rozygosity ill melallo m;l . J 11I1'l"st Demlt/ftll I05:G25-()28. 1/)1)5 .

Pill kcl D. Str;llllllC T. G ra y J W: CYLogenetic analysis usillg llu;lIltiwt;ve. hig h sClIsitiv itv. flu oresccllCc hyhridizOltiun. Prj/r Noti 1Ir(l(I Sri USA H3:2934 -21)3 1) . 198(, .

Protic-Sahljic M. Kraelller KH : R educed repair of 1l0 1l- diJl1I.:r pho to pro dllc t' :-. ill a gCllc transrccted illto xcrudel'll1 :1 piglllCllt OSlIlIl l'c ll s. PIW1c!(/1l'I1I P/wlllbiol-!J:5(1 4)-5 13, 1986

Sancar A: M ecllailis I11S ur DNA CX...-iS iOII rcpair . SriCllft' 2(,6: I 1)54 - I ') 5H . I C)C)4

S impson- H errc ll L. Nokcr PE. \V;:lgoncr S)): Vari ~lbi li ry or tUlll o r respon se ro

Chc llI olh er;IP Y. I. COlitTibuli o ll of' h ost heterogell eity. e (1/lf,T C lu'/lwlln'l' fJlwf/lw(j!/ 20:297-304. I ~H7

Spnrrow LE. Soong R. Dawki n s HJ S. lacopcn:1 OJ. H ec11a n PJ : p 53 Ge llc 1l1ll t:lrio ll alld cxprcssiOll ill Il aevi ;llld Illc lan OIII:IS. /\ I l'iIIllOI/ Itl u.. c.~ 5: C):l-IOI). 1995

Srarosclsk y A I-I. P.uhak S. C h c rn;ljovsk y Y. Tucker S I. Fid lc r Ij : Predo11linance ofLh c IIICr:lSr:llic p he nor-ype ill sOlllaric ce ll h yhrid.s of rhe K- 1735 11Iurinc mel ano ma . Ca",,,r Res 5 1 :(,292-G29H. I ~~ I

Talllladge J E. Fidle r Ij : Enhan ced nlcras l"atlc potc ntial o r tuli lo r ce lls harvcsted li·o 1l1 ' pOnt:llleO IlS IIlct :nit:ISCS or he lerogc lleolls 111lIriJ1L' rU11Iors . J '1111 Ct/llrcT 1!I~u

(,~:97S-9HO . I ')H2a Ta hll adgc .I E. Wol1ll a n SR. Fid lcr IJ: Ev idence ji11" lhe c1unal origin of sp o nl a n eolls

IllctaStascs . . StiCllrl' 2 17:3(, 1-363. 19H2h

Trope C: D iffe rcnt sCll sili v iry 1.0 c),[os (;Jtic d rugs o f prilll ary 1"11111 01' a lld Illc rastasis OrtiJ e Lewis can.: inoi ll a. NL'OI'/ flS lll fl 22: 1 7 1 - 1 ~(J. 1975

Tsunlo T. r:icllcr I): Di lfc rcn ccs in drug SCJlsiti vil ), :1 l1lo n l:; nIlIlo r t:c lls li·om }Jarcmal ttl lll0rs.

selectcd v;ui :1 l1 ts, al ld spo II Can colls Ille taSl;ISCS. Clllln.,. /{t's -I I :3058-3064. 19H 1

USIII :Jn i HA: Ge l1 o mi c.: inst'lb ili ry and llI ~t'as t ;ll i c prog ress io l1 . Pflflt,1hiolog), 6 1: I 09-\ I U. 1993"

USIl1:lI1i I3A. LUl1 cc J. Sh erbet GV: DNA repair tln d rcpair fidc lit ), in IIlct:lsraric \ '" ria ll r

of thc 13 16 Illurin c m e la no ma . J Cdl Bit1fltcIlI 5 1 : 336-3~ "' . 1993b Vidal MJ. Logan zo F J r. d e O li ver ia Alt. Ila yw:ud NI< , Alhin o f\ P : M ut at io ns and

dercc tivc cxpress io n or thc IVA FI 1121 tu ill o r-suppressor gC 11 C ill llI ;li ig ll ;t m IlIc l;lIlo ll1a . /l'll'in ll (l /l lfl I~ t '''' 5:2·13-:!50. 1(j!J5

\:Vei Q. M ;ltallo :-; ki GM. Far l1lc r ER. H c da)';lt i M. Gro!ls ll ta ll L: DNA repair :1nd :1g:i l1 g

in hasal ce ll c:1rt.." in o llla : a mo lccu lar ~pidc l1l i() l op;:i<: sHirl y. P,.,)( Not/ rlc-"d Sd USA 911:I(d 4-1(.JH. 1993

We i Q. SlrOll( SS. C heng L. Lee .I E. J)uvi .. M. Wd,e r ~S. Han kin s 1'1) : Red nced DNA rc pair c:l p;lc ilY in p;lt iellts \vir-h c ut illlCOUS lIlal ig n :IIH ItIcJaIlOlIl :L JJ,.c)( .... 1", ;JJJ(lf C fl/I(j ;r Ut'J 37:263 . 19!J(,

Weich scl h;1l11l1 RR. Dahlhc rg W. Lit d cJU: Il1h c rc ntl y radiorc!oi i:-; ra llt l'c ll s CX iM i11 'OIllC

huma ll ('tlJIlo rs. Pi'llI' N flrl i lead Sd USA t:>2:4732- -I 735 . I I)S~