Deletion Mapping of Chromosome 3p and 13q and Preliminary Analysis of the FHIT Gene in Human Nonmelanoma Skin Cancer

Deletion Mapping of Chrotnosome 3p and 13q and Preliminary Analysis of the FHIT Gene in Human Nonmelanoma Skin Cancer Steph en K. Sikkink , [shtiag R ehm an , and J onathan L R ees De.p artlll ent of Derll!a to logy. U niversity of N ewcastle upo n Tyne, Newcastle upo n T yne, U .K.

Loss ofheterozygosity of chron1osotnes 3p and 13q occurs frequently in hun1an cutaneous squamous cell neoplastns, suggesting the presence of one or n1ore tun1or suppressor genes on these chrmnosome arms that tnay be involved in the pathogenesis of this tumor type . To date there is no clear evidence in cutaneous tumors where tl1ese putative genes are located. In this study we have analyzed 20 squamous cell neoplasn1s that show allelic loss at chromosmne 13q, and 22 squatnous cell neoplasn1s that show allelic loss at chromoson1e 3p, in an atten1.pt to define the smallest area of deletion. One commonly deleted region was identified on chron1oso1ne 13 tl1at centred around 13q13, and two con.lmmuy deleted regions were identified on chron1oson1.e 3 that mapped to 3p24-pter and 3p12-p14.1. Our fmdings suggest the presence of at least one tun1or suppressor gene on cluotnoson1e 13 and two tun1or suppressor genes on chron1o-

some 3p that may be involved in the progression of these neoplasn1s. Deletions within the Fragile Histidine Triad gene, located at 3p14.2, have been reported in several tumors, leading to the suggestion that this gene is involved in tumor development. To evaluate the role of the Fragile Histidine Triad gene in nonn1elanmna skin cancer, we have used reverse transcriptase polyn1erase chain reaction analysis to screen for deletions in 16 tumors (five basal cell carcinmnas, five squan1ous cell carcinomas, five actinic keratoses, and one case of Bowen's disease) and HaCaT and A431 cell lines. A normal transcript was found to be expressed in 14 of 16 tumors and both cell lines. This suggests that the Fragile Histidine Triad gene is not a cmnmon target for deletion in Bowen's disease and the cell lines HaCaT and A431. Key •vo.-ds: a/tematille splicing /loss of hctemzygosity I squauwus cell cauiuoma I tumor supp1'essor geue. j bwest Dermatol 109:801-805, 1997

T

go ne o n tO create a deletio n map of chro m oso mes 3 and 13 in a series of cutaneo us squam o.us cell tum o rs usin g a pan el of hi ghl y info rm ati ve nu crosa teLLire markers. In additi o n we have exa min ed a possible ca ndidate ge ne o n chro m oso me 3 . Th e Fragil e Histidine T 1iad (FHIT) gene that encompasses the FRA3B fi·agi le sire (G lover e1 a/, 1988) and renal cell carcin o ma-associated r(3;8) brea kp o int (Co he n e1 nl, 1979) was rece ntl y cloned fi·o m a co mlll o nly deleted regio n at 3p 14.2 (O hta cr a/, 1996) and is thought to be a pote ntial mmor supp ressor gene targe ted fo r deleti o n o n th is chro m oso m e. We have used reve rse tr:m scriptase po lym erase chajn reacti on (R T- PCR) analysis to screen for deleti o ns of the FH lT ge ne o n chrom oso me 3 in a se1;es of no nm elan o lll a sbn tum o rs and also the 1-l aCaT and A431 epithelial cell Lin es, bo th of-whi ch dispb y LO H at 3pl 4.2.

um orige nes is .is a multistage process in w hich multip.le ge netic abno rmaliti es lead to maLi gnant transformatio n. Evjd ence sugges ts th :J t genetic alterati o ns in o ncogenes and tum o r suppresso r genes combin ed w ith allel ic loss pla y a maj o r part in c;m ce r development (Knudson , 1985; Yokota and Sugimura, 1993; Kin zler and Vogelstein , 1996) . C haracteri zatio n of these ge neti c changes w ill be useful in determinin g the bio logic features necessa ry t:o r th e deve lo pment of neo plasia and ma y all ow the id entifi ca ti o n o f mo l<:c ular mark ers tl1at ca n prov id e a foundatio n Fo r therapeutic app roaches . Althou gh sq uamous cell neo plasms displa y di sti nct patterns o f chrom osom ;ll lo.>s w hen co mp;1red with o th er cutaneous tum o rs (Q uinn ct a!, l 994a), spec ifi c ge nes invo lved in the initiati o n and progress io n o f th ese tum o rs h:lVe ye t to be identified . Previo us studi es have used loss o f heterozygosity (LO H ) and deleti o n m appin g as a tool to locate sm aLl chro moso mal regio ns in w hi c h putati ve tum o r suppressor genes ex ist (Lundb erg et nl, 1987; Kastury eta/, 1996) . We ha ve alrea dy shown that c uta l H~o u s squ am o us ce Ll neo plasms show hi gh fi·equencies of LO l-l o n ch romoso m es 3 and chro moso m e 13 (Quinn el a/, 1994:J; l<..e hm an ct a/, 1994, 19')6). In the present study we have th erefo re

M.ATE RIALS AND M ET H ODS LOH analysis T hi rty-fl ve parafrln -e rnbedded archi va l tum o r sa mples w ith a cl ini cal and histopatho logic diag nosis of acti ni c kera w sis (A K), .Bowen's di sease. or squamo us cell carci no ma (S ) tha t had bee n previously screen ed for LOH O il chro mosom es 3 and 13 were studi ed. In o rde r to in crease the num ber studi ed , :1 further 20 sarnples of Bowen's disease ' "'ere sc reened fo r LOl-l using l));)rkers D13S263 and D3S1293. A tata l of t·t ll owen's, four SCC, and f1ve AK were used fo r chro moso me 13 analysis; and 13 Bowen's, two SCC, and s~:ve n AK for chro nt o.sonl e 3 :tn::~.lys i s. Sa m plc.::s \·v e re nt_ic ro diS!\t.!cted ro sep~u-:-~ tc tumo r fi·01n no rmal stroma, ancl geno mi c D NA w:IS iso lated as previo usly described (Q uinn et of, '1994b) . I NA fi·o m matched adja cent uonna l skin and/ o r veno us blood sa mples were used as co ntro ls. No rmal and tumo r DN A was aJJ;Ji yzed for LO l-l by pol ymerase chai n reaction (PCR.) amplifica tion of th e mi crosarelli te po lym o rphisms (Figs 1, 2) o btain ed fi·o m R .esearch Geneti cs (Huntsvill e, AL). Ge no mi c DNA fi·o m Lhe A43 1 hutmn sq uamous c:n:ci n 111a

M anusc rip t ,·ece ived M arch I 0, 1997; revised Jul y 29, 1997; accepted for publi cation Au gust 29, 1997. Reprint 1·eq uests to: Professo r J: L. R ees, Depa rtlll cllt of Dcrmatolog)', Uni versity of N ewcastle upo n Tyne, M edical Sc hool, Framlin gto n Place, NE2 4 HH , U .l<. Abbrev iatio ns: At< , acti nic keratoses; FI-ILT, Fragi le Histidin e Triad ; LO H , loss ofheterozygosi.ry; RT- PC R , r·eve1·se transcript:Jsc pol)'ll1 crasc chain reacti o n.

0022-202X/\17 / $ 10.50 · Co pyright © 1997 by Th e Society fo r Investigative Derma to logy, Inc.

801

802

THE JOU FlN AL OF INVESTIG AT IVE DERMATOLOGY

SIKK fNI< ET A /..

6

13

7

X

9•

10

II

12

13 14

15

16

17

18

19

20

1 3~

p

Figure 1. Allelic loss on chromosome 13q in squamous cell neoplasms. Diagrammatic represe ntation displayin g the pattern of LO l-l in tumor sampl es 120 using n'li crosatellite markers that span chromosome 13. Vertical black bars represent th e approximate mappin g loca tion of markers acco rdin g to th e Genome Database. Tumors are denoted as shaded vertical bms with each block representin g th e mi crosatelute markers used for PCR a mplifi c:~ti o n. D, retenti on of heterozygosity; • . LOH ; lllllll.

I*

"'

I

-

q

un_info rnlatlve ;

~~

no produ ct; tftj, alleli c

shift. All tumors display LOH unless denoted by *.

2 1 22 23 H

zs

26 21

zs

29"30

Jl

n :n

34 35

JG

37 38 39 40 4 1 42

I =

I

@

1ii.

I iiJ

-

-

il

-

w w

-

cell lin e (ECACC , Salisbuiy , U .K., G iard ct a/, 1\173) and the Ha CaT spo ntaneo usly imm ortali zed an euploid hum an keratin ocyte cell lin e (a gift fi:om N. Fuse ni g; Boubmp cl a/, 1988) was isolated by digesti o n overn ight at 37°C with proteinase K an d phenol-chl o roform extracti on. A431 and J-laCaT DNA was amplified along with a norm al bl ood co ntrol usu1 g chro moso me 3 specifi c m.i crosatelu te markers (Fig 2) and three co ntrol markers (D2S J 49, D I05 186, D 15S11 8) to detect LO l-l o n chrom oso me 3p. PC R. ampJification was performed in 10 ~I reac ti o ns using 100 ng of D NA, 200 rnM deoxynucl eotidc triphosphates, ·r pm ol of each prim er (o ne endlabeled with [y 2PJATP), l.5 mM MgCI2, and 1 unit ofTaq DNA polymerase (Advan ced Bi otechnologies, E psom, U .K.). PC R. consisted of 30 cycles of ·t n'lin at 95°C, I min at 55°C, and I min at 72°C with a fmal extension tim e of 10 n'lin at 72°C. PC R produ cts were heat denatured and electrop horesed throu gh 6% denatu ring polyacrylamid e gels. Ge ls were dri ed and ex posed to Fuji XR film (GRI, Essex, U.K.) fo r 24 h and all eli c loss sco red. Bands were sca nn ed by a ph osph orimager (M olecular Dynami cs, C hesham, U.K .) and analyzed w ith th e lmageg uant database V5.0 (Fig 3). FHIT analysis

RNA extmctio11 jm111 cell lines a11d 1111110rs Cell mo nolayers were in cubated at 37°C with 5% C0 2 un til co nAu ent in the appropriate culture medium ; D ulbecco's modi fied e:~ gl e medium (G ibco Hili, Paisley, U .K.) with 10% fetal calf serum for 1-l aC:~T :~ nd Fibroblasts; M CD I3153 (Sigma, Poole, U.K.) for kerat.in ocytes; Ea rl e's m od ifi ed eagle medi um (Earl e's balanced sal t so lu tion) , 2 mM glutamin e, 1% no nessential amin o ac ids (Gibco 1311-.L), 10% fe tal calf

Figure 2. Allelic loss on chromosome 3 in squamous cell neoplasn1s . Diagrammati c representation displaying th e pattern of LOH in tumor sa mples 2'142 usin g m.i crosateUi te markers that span chrom oso me 3. 1/ertica/ black bars represent the approximate mappin g locatio n of mark ers according to the Genome Database. Tum ors are denoted as shaded IJCrtical bars with each block represe nti ng the mi crosa telli te markers used for PCR. of amplification. D. retentio n heterozygosity; • . LOl-l; ~. uninformative; ~ . no produ ct; ~. alleLi c shift; g, LO l-l with aLl eli c shift. All tumors dispby LOl-l unless denoted by *.

se rum for A43 1. Co nAu ent cultures were spl.it 1:3- 1:6, i. e., seeding at 24 X 104 ceLls per cnl usin g 0.25% trypsin or t1y psin /ethylenediamine tetraace ti c ac id. RNA was extracted from dissected fi·esh tumors 0 11 the day of exc isio n or tumors were snap frozen in liquid nitrogen to be used at a later date. Samples were also taken fo r histopathologic diagnosis. Total RNA was extracted fi:o m cutan eous ce.lllines (keratin ocytes and fibroblasts to be used as a normal co ntrol for th e FH!T ge ne) , fresh or snap frozen tumors (fiv e basal cell ca rcinoma s, five SCC, five AK, and on e Bowen 's), and normal skin using the RN easy kit (Q IAGEN , C rawley, U.K.).

R everse lrr/1/scriptioll a11d RT-PCR First strand eDNA sy nth esis was ca rri ed out at 37°C in a 10 Ill final volume containing 2 !lg total R.NA , 1 11M oLigo(dT), 5 rnM deoxynucl eo tide triphosphate, and 200 units of M oloney-Mutip e Leuken'lia Virus reverse transcriptase (Prom ega, So uth ampton, U.K.). Th e reactio n was stopped by in activating th e enzyme at 75 °C for lO mi11 . PC R amplifi cations we re ca n·ied out in 25 ~LI reaction mi xtures containing l 111 eDNA, 1.5 mM MgCI2 , 10 mM Tris- H.C l, 50 mM K C I, 300 ~M of each deoxynucl eo tide triphosphate, 'I unit of Tag po lymerase, and 175 ng of primers MUR5 and R.P2 as desc ribed in Thiaga lin gham cl a/ (1996)). PCR consisted of ini tial denaturation at 95°C for 90s followed by 35 cycles of 95°C for 30 s, 62°C for 60 s, and 70°C for 60 s. PCR produ cts were resolved on 1.5% ethidium bromide-stain ed agarose gels. Nested RT-PC R l ~I of eDNA was used for a first round amplification in a 25 111 volum e co ntainin g 0.8 ~M of prim ers 5U2 and 3D2 (O hta cl a/, 1996), 50 ~LM of each dcoxynucl eo tide triphosphate, 1.5 111M MgCI2, '10 mM Tris-

VO L 109, NO. 6

C HRO MOSOME J P AND 13Q DELET ION MA P

DECEM BER 1997

T

N

(a)

803

N

(b)

T

sand1 ,_

Band 2 ,_

2-

1 ,_

N

N

T

6S

T

4S

2SK l4K

S2

1

2

2

2

36

20K

2

IBK J9

1

21

1 26

ll

15K

t2K

18

!OK 6.4K

9,0

S..OK.

••• Figure 3. Representative autoradiograph showin g allele loss, and graphs showing phosphorimager scans of normal (N) and tmnor (T) DNA at (a) D3S1289 and (b) D3S1217 for tumor 26. Acrybm.ide gels were exposed ove night in a phosphorimager cassette and individual lan es scanned using th e lmagequant database to produ ce a graph. Band imensity is give n in arbito ty units on th e y-ax:is. As th e phosphorimager peaks correspond to bands on the au to radiograph , each microsatellite band is numbered on the auroradiogr:tph and also on th e ph osphorimager scm1. Note the reduced intensiry of b:utd 1 in th e tumor sample for both markers. HCl, 50 mM KCI, and 1.25 units of Taq polymerase. PCR. co nsisted of denaturati on at 95°C fo r 3 m.in then 25 cycles of 94°C for 15 s, 62°C for 30 s, 72°C for 45 s, and a fin :il extension of 72°C for 5 min. I ~I of the amplifi ed produ ct was subj ected to a second round of PCR ampl ific ation using nested primers SUl and JD 'I (O hm e1 nl, 1996) und er the above conditi ons. PC R. produ cts were resolved on 1.5% ethidium bromid e-stain ed agarose gels. C lolfilfg of PCR pmd11c1s Three mi croliters of PCR produ ct was furth er amplifi ed und er identi cal co nditions fo r 10 cycles before cloni ng into a pC R.2. J vector and transformati on (TA cloning ki t; Invitroge n, Leek, The Netherlands)

acco rding to the 1n anufu cturer 's instructions. Coloni es w jrh inserts \Vere

identifi ed and DNA was extracted by th e alka line lysis method (B iruboim and Doly, ·1979) after ove rnight cultme. DNA fi·o m th e coloni es w:1s sequenced usin g a standard cycle sequencing kit (SequiTh enn ; Epi center Technologies, Madison, WI ). R ES ULTS Chromosome 13q is deleted in the m ajority of squamous ceU neoplasn1.s Fiftee n out of 20 tumo rs displayed LOl-l o f fi ve o r more markers spanning the length of chrom osome 13, sugges ting m o nosomy in th ese tmnors (Fig 1). One AK (numb er 16) di splayed an interstitial deletion ;1t l 3ql3 with LOH of the mark er at D1 3S263 and retention of all other markers o n chro moso me 13. Three tum o rs (1 0, 19, and 20) were found to retain th e t1Jark er Dl3Sl 65 mapped to 13q13q14. 3, and on e of th ese tum o rs (10) also showed retenti o n o f the marke-r D13S385 at l3q34. Tumo r 9 did not display LOB of an y of the markers o n chromoso m e 13; howeve r, it did show alleli c shifts o r nove l all eles with fi ve m.i crosatellite markers. T his patient gave no histo ry suggestive of a mismatch repair defect and alJ eli c shi fts we re no t detected with markers (ro m seve n oth er ch rom osomes (data not sh own). Two regions of deletion on chrmnosome artn 3p Six ·of 22 tumors sho wed LOH of fi ve o r mo re mi crosa teliite markers spa nning

the length of chromosom.e 3, suggestin g monosomy in th ese tLUTtors (Fig 2). Twelve of 22 tum.ors displayed a w hole or partial deletion of 3p w ith retention of markers on 3q . Tumor 37 show ed LOH of D3S1307 and 0 3S1293 but retention of ali oth er informative markers o n chromoso m e 3, pointing to a terminal deletion of 3p in this tumo r. Tu m o r 26 sho wed retention of three m arkers at th e termi nal end o f 3p (D3Sl 307, D 3S1554, and D3S1293) but lost alJ other m arkers o n chro m osome 3 (Fig 3). Another tumor (27) displayed LOB fi o m D3S1307 (3p26.5) to D3S1 611 (3p2l) , retelltion of the n ext three markers fi·om 3p24.2 to p14.2, then LOH of markers at D 3S 12 17 and D 3S 1251. As 3q was also retained, this suggests two interstiti a.! deletio ns o n chrom osom e 3 in this tum o r. H :JCaT and A4 31 cells w ere hemizygous for al.l m arkers di stal to D 3S125l (3p12) but retain ed D 3S1 2l5 (3p11) and D 3S151 2 (3q2325.1) , sugges tin g deletio n of 3pter-3p13. Full-length FHIT transcripts were detected in most ttunors U sing primers MUK5 , FP1 , and RP2 (Fig 4), two sampl es of no rma.! keratinocytes, two samples of fibrobla sts, H nCaT and A 43 1 ce lls, fiv e AK , three SCC , four basa.l ceLl carcin o mas , and o ne B o wen 's disease all showed bands of the predicted size (Fig 5). C ultured fibro blasts as w ell as displaying a di stin ct normal eDNA also showed a 150-bp band of va riable intensity that on clo ning into a TA vecto r and sequ ence anal ysis turn ed o ut to b e an artefa ct composed o f concatenated primers. One basa.l cell carcinoma o ut o f fi ve sho wed products 650 bp and 550 bp as well as a no rm al p.r odu ct o f 750 bp . Two of fi ve SCC showed abnormal transcripts of =690 bp , 550 bp , and 450 bp with o ut a no m1 al product. W hen nested primers we re used, ho w ve r, to am.plify the FHI'T gene a large number of produ cts ran ging in size fi·o m full lengt h to 150 bp were seen fo r both n ormal skin and tumor tissue. H aCa T cells also sho wed multiple bands as well as a stro ng 300-bp transcript using primers SUl and 3Dl. Wh en the

804

T H E JOU ilNAL 0 1' INVEST IGAT IVE DE ilMATOLOGY

SII
I TCCCCGCT CT GCTCrGTCCG GTCACAGGAC ITTTTGCCCT C rGTTCCCGG GTCCCTCAGG

61 CGGCCA CCA GTGGGCACAC T CCCAGGCGG CGCTCCGGCC CCGCGCTCCC TCCCTCrGCC

SU2 12 1 TTTCA TTCCC AGCrGTCAAC A TCCTGGA;IG C71TGAAGCT OI GGMAGAA GAGAAATCCA

SU I 18 1 CTGAGAACAG TC'/'G'/'AAAGI7 TCCCi1;1muq /ITCf1ICATC CAGACGGTGG AAGGGAGAGA

MURS 241 AAGAGAAAGA AGGTATCC.TA GGAATACCTG CCTGCTTAGA CCCTCTATAA AAGCTCrGTG 301 CATCCTGCCA CTGAGGACTC CGAAGAGGT A GCAGTCnC'r GAAAGACITC AACTGTGAGG

Fl't 36 1 ACATG7CG1T O IGAT/TGGC OIAOITC/'C'A TCAAGCCCTC TGTAGTGHT CTCAAAACAG

421 AACrGTCGn CGCTCTTGTG AATAGGAAAC CTGTGGTACC AGGACATGTC CTrGTGTGCC 48 I CGCTGCGGCC AGTGGAGCGC TTCCATGACC TGCGTCCTGA TGAAGTGGCC GA TTTGTnC

S41 AGACGAeCCA GAGAGTCGGG ACAGTGGTGG AAAAACA nT CCATGGGACC TCTCTCACCT 60 1 TTTCCATGCA GGATGGCCCC GAAGCCGGAC AGACTGTGAA GCACGTTCAC GTCCATG'n'C

66 1 1TCCCAGGAA GGCTGGAGAC 1TTC ACAGGA ATGACAGCAT CTATGAGGAG CTCCAGAAAC

72 1 ATGACAAGGA GGACITICCT GCCTGnGGA GATCAGAGGA GGAAATGGCA GCAGAAGCCG 781 CAGCTCTGCG GGTCTACT1T CAGTGACACA GATGITTTTC AGATCCTGAA 1TCCAGCAAA

30 1 84 1 AGAGCT A1TG CCAACCAGTT TGAAGACCGC CCCCCCGCCT CTCCCCA AGA i'iGi!ACTGAII T

RP2 90 I ( /1 GCA TGA AA ATGb l GTI7'C 7TO I TC7'CAC CA 1'CCTG1/I T TC17'0 1A CCA G'f'G11TCCCCC

302 96 1 ACCr CGGTCA C"/'CCAACTCC GnAAAATAC CTAGACCI'AA ACG GC.TCAGA CA GGCAGATr I 02 1 TGAGGTTTCC CCC.TGTCTCC 1TA TTCGGCA GCCTr ATG AT TAAAC.TICCT TCTCr GCTGC AAAAAAAAAA AAAAA

108 1

Figu re 4. T h e eDNA seq uen ce of the FHlT geue showing the position of primers used in PCR reaction s.

lkb

A B

500bp

Figure 5. PCR amplifi catio n ofFHIT eDNA fro m SCC (la11 cs 2, 3) with primers MURS and RP2, A431 (la ue 4) w ith FP1 and RP2, Normal fibroblasts (lau e 5) w ith MURS and FP2, H.aCaT (/a11 c 6) with MURS and FP2 and Normal kerati11ocytes (/a11 c 7) with F Pl and RP2 . ~ indi cate normal transcripts usin g primer pa irs MURS and RP2 (A) and FPl and RP2 (B) .

reJction was carri ed out w ith o ut the us<:: of nc::sted primers, usin g MUR5 and l'l... P2 , ;111 d FP1 and RP2, respecti ve ly, full- le ngth tran scripts were seen and th e mul tipl e bands were not oppare nt (Fig 5). D ISCUSS ION .Prev io us studies have ch arac teri zed th e pattern of all e li c loss in hum an cutan eo us squamou s ce ll neo plasm s (Q uinn e/ a/, 1994b ; l'l...ehm aJl cf a/, J994) . W e have used LOH anal ysis to m op areas o f deletio n o n chro moso m es 3 and 13 in cutan eo us sq uamo us ce ll tum ors. T hi s study shows that loss of a w hole co py of chro m oso m e '13, resultin g in monoso m y, is a fre qu ent findin g in thi s typ e of skin tu m or; howeve r, in fo ur cases partial de letions we re also fo un d. T um o r '16 di spla yed an

interstitial de letio n between J 3q 13 and 13 g 14. Deletion m app ing in seve ral tum o r types in cludin g breast (Sch ott ef nl, 1994), hepatocellular (M aes tro Cf a/, ca rcinoma (Kuroki eta /, .1995), hea d and nec k 1996) , and pros tate can ce r (Coo ney ef a/, 1996) have all detec ted LO H in thi s area. Th is is a parti cular " ho tspot" fo r de.letio n o n chro m oso m e 13 du e to seve ral targe t gen es co ntain ed in the regio n. Poss ib le ca ndidates in clu de th e lef illob /nstol!la- ·1 tumor suppresso r ge n e that is loca ted at 13q ·J4. Jnac ti vation of th e rcfill oblasfoll w- ·1 gene has been assoc ia ted with aLl el ic loss in seve ral tum o rs incl udin g li ve r (Z hang ef a/, 19':!4), bbdde r (Xu eta/, 199 1), and lun g ca nce r (Re issmann cf a/, 1993); however, fin e delet io n m appin g :tro und the relill ob /asfoii W- ·1 locus has imp li cated o rh e r ge m:s to be in volved (B row n el a/, 1993; i'e i ef a/, '1995; Kuro ki el a/, '19':!5; D ev ildt:r ct a/, 1995; Coo ney ct a/, 1996; M aestro et nl, 1996). Fo r <::xa mpl e , Bmslt - l situ ated pro xim al to th e rctillob/asfO!IW-1 ge ne at 13q12-q 13 di spla ys red uced m RNA expressio n in p1·ima1y breast tum o rs a11d breast ca ncer cell lin es (Sc ho tt cf a/, J 994) and th e breast ca nce r susceptibility ge ne (BRCA2) , located at 13q l2-q'1.3 (Wooste r cf a/, 1994), co ul d also be a targe t ge ne. T hree tumo rs (10, 19 , and 20) di splayed rete ntion of isobted mi crosatelli te m arkers o n chro m osom e 13 at .IY I3S165 and D ·I3S285. T h e reaso n fo r thi s is no t clear, but it is poss ib le that thi s regio 11 co uld be ho m ozygo usly deleted in these tumo rs and retentio n o f th e two m arkt: rs co ul d be a res u lt o f co n ta min atin g n o rm al st ro rn:1. ln co ntrast to l3q only six tu m ors :1ppca red to lose a w hole copy o f chro m oso m e 3. M o re frequ e ntl y w ho le o r partial d ele ti o ns of 3 p occ ured. T h e regio ns ofloss we fo und are co nsiste n t w ith chro m oso m e 3p d eletio n m app in g studi es in ot he r cance rs (E hl en and Debea u, J':!90; 1-libi ef a/, 1992; M aestro ct a/, 1993; hen cf a/, 1994; W u cf a/, 199 4; Dru ck ct a/, 1995). Seve ral gro ups have sh ow n th e presen ce of three d isti net de le ted areas o n chro m osome 3p in severa l tum o rs. Two o f the deleted regio ns w ere simi lar to the o nes fo u nd in o ur stud y (3p243pter and 3p 12-p '14.2), but a third regio n of del etio n was also d etected at 3p21. We did n ot d etect a minim al de leted r<::gio n in thi s area but thi s co uld be du e to th e tum o r populati o n we selected ; howeve r, 3p2 1 was lost in 1. 2 o f 22 tum o rs as part o f a large r deletion. T umo r 37 showed .LO l-l of two markers at the termin al end of 3p but rete ntion of 3p and 3q. Co n ve rsely, tum o r 26 reta in ed the terminal <:: nd of 3p b ut lost th e rest of the chro m osome. T he m ain ca ndidat<:: ge ne in this regio n is th e l/o11 Hippei-Lir rdnu ge ne (.Latif ef a/, 1993) that is mu tated in f:11nilia l ren:1l ce ll carcin o m a (reviewed in G narra et a/, 199G) and he m angio bl asto m as (Kann o cr a/, 1994) and d ispla ys tu m o r supp resso r properties in certain cell lin es (ILi opo ul os ct a/, 1995). Tumor 27 di sp layed two regio ns of .LO H from the termi nal e nd of 3p to 3 p2 '1 and also 011 in terstitial de letio n at 3p l4- p1 2. T hi s reg io n of loss co ntai ns the rece ntl y clo ned PI-li T ge ne th at in turn en co n1 passes th e Fl'l...A 3B fi·agil e site an d also the renal ce ll carcin o m a-asso ciated t(3;8) brea kp o in t. D istributed ove r 500 kb and co mposed of I 0 exo llS, the FJ-J IT prote in is a m embe r of a hi ghly co nserved hum an hi stidine triad ge ne f:1 mil y and shows simi larity to the Schi zosacca ro m yces pomb e Ap"1 A asymm e tri ca l hydro lase, w hi ch is p ro posed to have a role in DNA rep li cation (Baxi cl a/, 1994) and in the ce ll 's respo nse to n1 etabo lic stress (Ba ker ct a/, 1986; l0 arr ef a/, 19H
sec

VOL. IIJ9 , N O. 6

C HR OMOS

DE EMUET>... \997

th at the aberrant eDNA detec ted by Ohta et n/ usin g nes ted PC R co uld be ca used by al ternati vely spli ced H ·:I IT n:ansctipts being overrepresented in th e nested PC R. We have also detected results co mpatibl e with thi s usin g both PC R approaches and this has mad e aberrant eDNA transcrip ts difficult to distin guish from possible a l tern~tivcl y spli ced tran sc ripts (Fig 5, laHes 2, 3). In th e abse nce of dir ct seq uencing of th e multip le bands it is also possible that so m e of th e <1 dditi o n<1l bands we have detected could be a res ult of the nonspecific amp lification of unrelated sequ ences. T he res ults presented h ere, h oweve~:, suggest that as a ful.l-lengt h tran sc ript was detec table in 14 of 16 sa.m p les, deleti o n of FH IT is un common in norunelanoma ski n ca ncer. These res ults, however, do not exclude the p ossib iliLy of alterations in RNA processin g o r of differenti al expression o f va ri o us transcripts in no nrnelan oma skin can ce r. T he technica l pro blems raised by alternative sp li cin g and th e use of P C R, exacerbated by th e use of nested P C R approaches, m ean that further analysis will requjre dire ct sequ encing and s tudi es of RNA expressio n usin g north ern blottin g, so m ethin g that will be tec hni cal ly demanding o n small clinical samples, although it is clearly feasibl e for ce ll lLnes.

S KS is a rcseatr.fl associat.eji11tded by tlt e No rth o{ E 11gla11d Ca11ccr R.escatdt Ca t11pa(g11 (NECJ~C). Tlds work 111as SIIJ!J!Mtcd by gm111s jiv111 NECRC to JLR .

REFEREN C ES 13ak er J C, .Jacob$011 MK: Alteratio n ofac\eJJ )'I diJl\t cleo tide JH etabo lism by e nviron m e ntal stress. !'me Ntal Aotd Sci USA 83:2350-2352, 1 9R6 \3 nx i M D, M c Le nn an AG, Vis il wn natha JK : C hara cteri sa ti o n of th e H eLa cell DNA polym er:Jsc :1\ph:t-associ:Jted Ap4A bindi11g prote in b y photoaffinity bbeling. Bioc/ICinistry 33: 14601 - 14607. 1994 Birnbo im H C, Doly J: A rap id al kaliu ~ cx t:rn ction procedure fo r sc ree nin g rcco 1nb inant plasmid DNA . N ndcir A cids Res 7: 15 13, 1979 .Bo uk J nlp 1', Pctw sscvskn RT, Breitkreutz D , HornU JJg j , Markam A, Fuseni g N.E: N orn1al kerati ni zatio n in a spo n tan eo usly i11Hn ortali scd :llu'! u p lo id human kcratin ocytc cdl line.) Cell Biv/1 06:76 1- 77 1, 198H Drow n AG, !'loss I' M , Dunn e EM , Stcd C M , Wei r-Th ompso n EM: Evidence fo r a new tu mo ur suppresso r locus (Dt3M) in hum an B-cl·ll n eo pl::tsi:~ tdoml!ri c to th e rctin o b!:lsto m a gene . tVt1f11re Gcll crics 3:67-72, I 993 C hell LC, Mat:sumura K, De ng C, C/. lll: Dele ti o n or tV•1 0 sc p:u-ate n: gio ns Oil c hro m oso m e 3p in breast cancers. G rllccr}{emrn:lr 54 :302 1- 3024, \994 Co h en AJ , Li FP, B e rt~ S, Marchetta DJ , Tsai S, J acobs SC, Brow n R S: H credi~1ry renaleel! ca rcinOIIl:J as.wci ::~tcd with :1 c!Jrom osoJJial trnnsl oc:~ l"ion . N _E1(g J t\,fcd 30 I :5925~5,

Jn9

Coo n ey J Dcvildc r M C . l' rnn co is S, \)osic C, ct a/: Deleti o n ca rtograph y aro und th e D 13S25 lows in 13 cell c hro ni c lymph ocytic leukemia and accurate m appi n g o rrh e in vo lved tum o r suppresso r ge ne. Catrccr Rrs 55: J 355- 1357, '1995 Druck T , Ka stury K, 1-l:rdaczck P, "' trl: Loss o f heterozygosity at th e Iilllli lial It C t (3: l:$) locus in most clear cd l rena.\ carcino rn ;'IS. Caut:er Res 55:5348- 5353, 1995 E hl en T , Dubeau L: Loss of h c.: tc rozygosity o n c hro moso m~ l segm ents 3p, 6q and I I p in human OV
or

M E 31' AND 13 Q DELETION MAP

805

Hibi K, T :1 kah:1s hi T, Y:una k::l \V:-1 K. ct
1-\ippd-Lindou gene prod uct. Nn ll/11' Mcd 1:822-826, \995 K:1111J O H , Ko ndo K, Ito S, ct ,·1/: So m ati c.: mut:nio ns of th t.:: vo n H ip pd-Lind:lu tumor suppn;ssor ge n e in spo radic ccntr:1l n e rvo us ~ystcm h c mnn giob b sto m:1s. Cmu:cr R es

54:4845-4847, \994 Kastu ry K, U:lHh Jl , Dru ck T. t•r al: Potenti al b>:lSt:ro inti..'.Stin al t\.11 1\ 0 lll' su pprcs5o r lo cus :'It th e 3p 14. 2 I=R.A313 site identified by h o m ozygo us. deletions in wm o ur ce ll lines.

Cn llcer Res 56:978-983, 1996 r< in zkr I{W, Vogc!stcin 13 : Lesso ns fi·o 111 hcrcd ita1·y co lo n,;c tal C..'Hl C<:r'. Cell 87 (2):159170, 1996 lanngopou los \, 1':11\ dis N , T helin S, rt a/: Th<• FHIT n11cl J>TPRG ge nes are deleted in bl.: ni gn pro lifer!ltive breast d ise5:2088-2090. \994 \Vu l, Sloa n P, R ead AP, H arri s R , Th::tkk cr N: Dclerio n m:~ pp in g on th e sh ort arm chro m oso m e 3 in squ amous cell ca rcin o m a o f th e ora l cav ity. Ctl HCc r Rcscmrh 54:6484-6488. '\994 X u 1-1 - T , Hu SX, Cagle PT. Moo re GE, Oen ed ict \VF: Abse n ct: o[ rctinobbstont n expressio n in primary n o n -Stn :l !l-cdllu ng ca rcit\ 0 1\l:ls. Crwccr Res 5 1:2735-2739. 199 1 Yokot:t j , Sug inn1ra T: Multiple steps in C::trCill ogcnc.:!sis in volvi n g altcr:lti o ns of nwltipl e tum or suppressor ge nes. /R eview /. H'ISE/3 J 7:920-925. 1993 Z hang X . X u Hj, fV\u rakami Y, c.t ttl: Deletions of chro m oso m e 13q, m ut:\tions in R cti1 JO bbsto111:1 1, a11d rerinob l;"tstom;J protei ll .sta te in humaJl ll cp;twcel lu l:lr carc ino nu. Crrrrrcr Res 54: 41 77-4182, \994

or

or