Carcinogenesis rol.l6 no.9 pp.2069 2074,1995

Nitric oxide and ethylnitrosourea: relative mutagenicity in the p53tumor suppressor and hypoxanthine-phosphoribosyltransferasegenes

Emanuela Felley-Bosco, Jovan Mirkovitchl,Stefan Ambs2, Katherine Mac63, Andrea pfeifer3,Larry K.Keefera and Curtis C.Harris2

Pharmacology and Toxicology Institute, Bugnon 27, 1007 Lausanne, lSwiss

Institute ol Experimental Research on Cancer, 1066 Epalinges, Switzerland,zlaborarory of Human Carcinogenesis. National Cancer lns-titute. Bethesda,MD 20892, USA, 3Nestec, Vers-chez-les-Blancs, Switzerland, alaboratoryof Comparative Carcinogenesis, National Cancer Institute, Frederick CancerResearch and Development Cenrer, Frederick, MD 21702, USA

Nitric oxide (NO) is a cellular messenger which is mutagenicin bacteria and human TK6 cells and induces deaminationof S-methylcytosine (SmeC) residues in vitro. The aims ofthis study were: (i) to investigate whether NO inducesSmeC deamination in codon 248 of the p53 gene in culturedhuman bronchial epithelial cells (BEAS-2B); and (ii) tocompare NO mutagenicity to that of ethylnitrosourea(ENU), a strong mutagen. Ttvo approaches were used: (i)a novel genotypic assay, using RFLP/PCR technology onpurified exon VII sequence of the p53 gene; and (ii) aphenotypic (HPRT) mutation assay using 6-thioguanineselection. BEAS-2B cells were either exposed to 4 mMDEA/NO (Et2N[N2O2]Na, an agent that spontaneouslyreleases NO into the medium) or transfected with theinducible nitric oxide synthase (rNOS) gene. The genotypicmutation assay, which has a sensitivity of 1 x 10-6, showedthat 4 mM ENU induces detectable numbers of G -+ Atransitions in codon 248 of p53 while S-methylcytosinedeamination was not detected in either i^^y'OS-transfectedcells or cells exposed to 4 mM DEANO. Moreover, ENUwas dose-responsively mutagenic in the phenotypic HpRTassay, reaching mutation frequencies of 24 and 96 timesthat of untreated control cells at ENU concentrations of 4and 8 mM respectively; by contrast, 4 mM DEA/NOinduced no detectable mutations in this assay, nor wereany observed in cells transfected with murine INOS. Weconclude that if NO is at all promutagenic in these cells, itis significantly less so than the ethylating mutagen, ENU.

Introduction

Nitric oxide (NO*), which is produced in vivo during theconversion of l-arginine to l-citrulline by'the nitric oxidesynthase (NOS) family of enzymes, is both an imporranrbioregulatory agent that mediates signal transduction, and acytostatic and cytotoxic agent (1). Recently, a role for NO asa potential endogenous mutagen has been suggested (2,3).Indeed, NO is mutagenic to Salmonella typhimurium TAl535

*Abbreviations: NO, nitric oxide; (i)NOS, (inducible) nitric oxide synthase;HPRI hypoxanthine-phosphoribosylrransferase; DEA/NO, Et2NlN(O)NOlNa;IBMX, isobutyhnethylxanthine; RFLP, resrricrion fragmenr lengrhpolymorphism; PCR, polymerase chain reaction; 6-TG, 6-thioguanine; 5meC,5 -methy lcytos ine.

O Oxford University Press

(2,4,5) and is able to deaminate deoxyribonucleosides, nucleo-tides, purines and calf thymus DNA in an aerobic environmenlin vitro (2,3). Exposure of a plasmid to NO in aerobic bufferedsolutions followed by replication of the transfected plasmid ineither human Ad293 cells or Escherichia coli resulted in base-pair mutations consistent with a mutational mechanism initiatedby deamination of DNA bases (6). NO is also mutagenic inhuman TK6 lymphoblastoid cells (3).

To test further the hypothesis that NO might serve as anendogenous mutagen we developed a human cell system whereNO is either endogenously produced or generated by a drugwith well-characterized NO-release properties. The resultsobtained by exposure to NO were compared with those obtainedwith ethylnitrosourea (ENU) since the mutagenic specificityof the latter is well known (7). In an attempt to be as close aspossible to an in uiyo situation, two genomic targets wereanalyzed: the p53 tumor suppressor gene, since mutations of itssequence, including C to T transitions at the CpG dinucleotidewithin codon 248,have been described in many human tumors(8,9) ; and the hypoxanthine-phosphoribosyltransferase (IlpRTlgene, often used to characterize induced mutational spectra(10). Using two different approaches-a genotypic assayand a phenotypic mutation aSSB!-1r,rs show that NO is notdetectably a mutagen in cultured human bronchial cells.

Materials and methods

Cell culture and mutagen treatmentBEAS-2B cells, an SV-40 large T antigen-immortalized bronchial epithelialcell line derived from a hurnan male donor (11), were cultured in dishescoated with a mixture containing l0 pgiml bovine fibronectin,20 pg/mlbovine serum albumin and 20 pg/ml type I collagen (Vitrogen 100, CollagenCorp., Palo Alto, CA) in serum-free medium (LHC-8, purchased fromBiofluids, Inc., Rockville, MD). LHC-8 contains insulin (0.g7 pM),hydrocortisone (0.2 pM), epidermal growth factor (0.825 nM), transferrin(0.125 pM) and bovine pituitary exrracr (35 p-glml).

In order to assess whether NO was able to incluce a mutation in codon 24gof p53, two strategies of NO exposure were employed: (i) exposure of BEAS-28 cells with an NO-releasing agent; and (ii) introduction of rhe murineinducible nitric oxide synthase gene (ilVOS) leading to endogenous productionof NO.

For the first approach BEAS 28 cells were incubated for 20 min with.{ mM DEA.t{O, an NO donor drug having the formula EI2N[N(O)NO]Na(CAS no. 92382-14-6) (12). The medium was rhen replaced and cells wereharvested after 72 h. The concentration of DEA,/I.{O (4 mM ) had been selectedfrom dose-response studies (unpublished data) showing that a survival of607c relative to the sham-treated control could be achieved at that concentration.

For endogenous production of NO, BEAS-2B cells were stably transfectedwith the murine /NOS gene (2b-nos3 and 2b-nos4) or control vector (2b-cmv2-1 and 2b-cmv2-2) as previously described (13). DNA was extractedfrom cells at passage zl.

As a positive control, BEAS-2B cells were exposed to 2, 4 and 8 mM ENU(Sigma) lor 20 min; the medium was then replaced and the cells wereharvested after 12 h. Cell survival after 2. 4 and 8 mM ENU treatment was1 O\Vo, 6'7 Va and, 30Vo respectively.

cGMP determination

To confirm that NO released into the medium was entering into the cells,NO-stimulated cGMP production was measured. Cells were plated in amultiwell culture dish (Costar; l0s cells/weli of 3.12 cm2). antj were then

2069

F..Feller-tsosco el a/.

in.ulrrtcd lbr l0 min in lhe presence or absence of isobutylmethylxanthinerlB\1X) and increasing concentrations of DEA/NO. All treatments were'ttrpped through the addition of cold ( 20.C) cthanol (final concentration66'l ). The santples were collected, frozen in liquid nitrogen, then evaporatedto dr\ness under vacuunt. They were redissolved in 500 ul of cGMp assavbufter (Amersham) and cGMP was determined using radioimmun.ur..yr.\mersham) following the manulacturer's instructions.

Soutlrcm anal,,-sis of meth,v-ltttion status oJ'cytusine in cod.on 24E of p53To assess thc methylation status ofcytosine in codon 248 ofp5J, l0 pg ofD\A from BEAS-2B cells were digested for 3 h at 37.C wirh either Hpallc,r .\Ispl. electrophoresed on 27c agarose gel and transf'ered onto Genescreenmembranes (Du Pont, Boston, MA). The filters were then hybridized (16 h at65'C1 with a 110 bp exon VII probe labeled by random priming. Hybridizednlters were washed twice at low stringency (2xSSC [SSC lx: 150 mMsodrum chloride, l5 mM sodium citratel, 25.C), then at increasinqlv strinsentconditions (65'C, 2xSSC/l TcSDS; 1 xSSC/l o/cSDS, 0.5 xSSC/I%SDS "and

0.1 xSSC/lToSDS).

Purifccttiort of p53 exon VII sequences from genomit DNAGenomic DNA was purilied from the cells by addition (20 ml/I08 cells) of50 mlVI Tris (pH 7.4), 5 mM EDTA, 200 mM NaCl ancl 0.zl7o SDS. incubarion$rth protetnase K.0.1 mg/rnl (16 h at 50"C), extraction by phenol/chloroforrn/isoamyl alcohol (25:24:1) and ethanol precipitation. Genomic DNA u.as thenre)trlcted (2 h at 37'C) with M.rpl (l U/pg DNA). The reaction was sroppedbr EDTA addition (2 mM final concentrarion) followed by phenolichlorolorm/isoamyl alcohol extraction and ethanol precipitation. DNA was resuspencledin 10 mM HEPES (pH 7.4) containing 5 mM EDTA. A DNA atiquot of100 pg was used fbr the isolation of mutated p53 exon VII sequencei usinga previously described method (14,15). Briefly, DNA was denarured in 0.2 MNaOH (5 min at 65'C), then the solurion was neutralized b,v, addition of0.2 M HCI and 100 mM Hepes (pH 7.4) added simultaneously. Sp6-biotinylated exon VII antisense RNA was then added to a final concentra_tion of 0.08 pg/ml. SP6 exon VII anrisense RNA was synrhesized frompSP72-X7 plasmid in which the enrire exon VII ofp53 (ll0 bp) has beeninserted into pSP72 (Promega, Madison, Wl). Hybridizations were canied outfor l6 h (65'C), then 5 pl of streptavidine agarose (Sigrra, St Louis, MO)uas added. The mixturc was incubated for,15 min at room Lemperarure.Agarose beads were spun in a microfuge and washed with l0 mM HEpES(pH 7.4) conraining I mM EDTA. purified genomic DNA was eluted in0.2 M NaOH to hydrolyze the RNA. Aftcr neutralization with 100 mM Tris(pH 7.4) plus 0.2 M HCI and addition of yeasr IRNA (0.5 [g/rnl), rhe samplesri'ere phenol-extracted and precipitated in silanized tubes with 3 vol ethanol.

In order to check the sensitivity of the method, 102, 103 or l0a moleculesol a standard exon VII mutant, pspstandardXT, were added to 200 pg ofDNA (containing 3X 105 copies/pg =+ 6X 107 copies) fiom unrreared cells andres-cued as described. pSPstandardXT plasmid contrins a I bp change in codon2,+8 (5'CGG ) CTG) and rwo additional base-pair changes in the exon VIIsequence. It was constructed using a method previously described for othermutants (16) and covers part of inron VI. eron VII anj part of intron VII ofthe p53 gene (2a6 bp).

Revriction fragment length polvnorphism/pol,-merase chain reactir.tn(RFLP/PCR)

]le n3f{l_e.{^ONA was resuspencted in l0 mM Tris (pH 9), 3 mM MgCl2 and0.3 mM dNTP containing CTTGCCACAGGTCTCCCCAÄ ano ACGTGGAT_CCAGGGGTCAGCGGCAAGCAGA (0.4 pM) as 5, and 3, primers respecr_ively, and amplified tbr five cycles (9,1.C fbr 30 s, 60"C lor I min and 7g.Cfor 30 s) using 0.5 U of pfu (Stratagene, La Jolla, CA) polymerase, Thesamples were then digested with M.tpI (30 min at 37.C). This pCR/restrictionstep was repeated once again and the samples were ampiified fbr 20 cyclesrn the same conditions as described above. Followrng a third digestion withMspI, the samples were amplified for 25 cyclei using AöGTGAAT_TCGTTGGCTCTGACTGTACCAC antl TGTGCAGGGTGGCAAGTGGC as5' and 3' primers respectively. The 146 bp fragmenr obtained was purifiedafter electrophoresis on agarose gel and sequencäd using pfu exo- (promega)fbllowing the manuf'acturer's instructions.

HPRT- selection

Cells .were,

grown lor 4 days in LHC-8 without atlenine before starting theselection.-ln preliminary studies it was determined that the expression rimerequired for 6-thioguanine (6-TG) resistance was 7 days. This lig phase wasadopted for DEA/NO- and ENU rreated ceils. Murine iNos-transficted cellswere tested at passage 5. To achieve HpRT- selection, 105 cells (l0l cells/cm') were cultured in the presence of r7.5 ng/mr 6-TG. Coronres were stained(0.25clo crystal violet) and counted after 2 weeks of selection. The mutantfraction was calculated as the ratio between plating efficiency ln the selective(6-TG) versus non-sclective medium.

2010

ß* HbP ar t'obF $t

G

EIltA.

*n- äCStuF

*x#1 Yll

BItttt*

g.

*,

*+ tr*:'*.

- - G - -:====

€ * *" .s* <_ i?obp

c{lon tirer-@*m;ö-Fig. 1. (A) Southern b)ot analysis of BEAS-2B cells DNA digested withHpaII (H) or MspI (M). Differential digestion indicates thar the inrernalc.ytosine of the MspUHpall site is methylated. The genomic map describes

-th_e_genomic restriction digestion bv MspI in the region of humin p53 exon

VII. Since the probe used in rhe Southern blot analysis was the pj3 exonVII, the 1 70 bp fragmenr corresponds to an hybridized probe of only 40 bpand.the signal appears u,eaker because of the stringent washing .onäitionr-used. (B) Southern blot anall'sis of DNA fiorn human tissues. The DNAwas digested with HpaII or MspL Differential digestion indicated rhat rheinternal cyrosine of the Mspl/HptLII site is rrethvLted.

Results

Assessment of methylation status of cr-tosine in codon 24goJ p53

A comparative Southern blot analysis (Figure I A) followingMspI and HpaIl reslriction indicates that the cytosine in thiCpG dinucleotide of codon 248 ot p53 is fully merhylared inBEAS-2B cells. Indeed, a 299 bp fragment is presenr afterdigestion with MspI, which is insensitive ro the merhylationof internal cytosine, but only a high molecular weight frigmentis hybridized after restriction wtth Hpal\ which is methylationsensitive. Thus, BEAS-2B cells are suitable to test the hvoo_thesis that the C -+ T transition is due to S-methvlcvtoslne(5meC) deamination. A similar result was obtained usineprimary human tissues from colon, liver, esophagus and lun!(Figure 1B).

'Fish-RFLP/PCR'

The 'fish-R.FLP/PCR' technique cleveloped ro derecr 5mecdeamination in codon 248 of the p53 gene is based on a two_

Co ntro I ENU

@NO and ENU mutagenicity in cultured human cells

$ {?8$1*

A"

aa-F

CH:

fJOOOCX

CTGGmec->T

CCAG+CCGAG>A

+Mcnlo

s"

*

Fe F-e/-f>ä\ /ly.tt'\ r},,J-.A\_^_\_r \_J__1.-/ eaj\_l.,I

f r , atkati

ooI

Vf 5xPCR,rf +\ Mspl

| ( intronic *PCR { evonicprlmers +

| | MsnlV

sequencing

Fig. 2. Flou'charr of lhe ti\h-RFLP/PCR' protocol. Follorving the exposureof the cclls ro a mull-genii accnt. DNA is cxn.acted and digested witÄ Mrylto climinate the *ild r\pc \equence. The DNA is thcn enriched for thc 25.1gene exon VII regrtrn L,r 'ti.hing' using biotinyJated antiscnse RNAfollowed by intrrobilizrtiolt on strcptavidin-agarose. The rue mutants canthcn be anrplilied usin-g thc polvmcrase chain reaction and characterizecl bysequenclng.

step enrichmenr of a mutant target (Figure 2). In the lirst stepthe target gene is exrracted from total genomic DNA, and inthe second the mutated sequence is arnplified by coupledRFLP/PCR. This allows the derecrion of mutants fiöm a lärgeamount of DNA that would not fit in a standarcl pCR assay.

The sensitivitl, of this method was assessed by mixing lb2,103 or l0a molecules of standard mutated eion Vll with 2'or200 pg of M.spI-digesred DNA (Figure 3). Ethidium bromiclestaining showed that the amount of pCR produci was propor-tional to the number of molecules acided (Figure 3A). Theanalysis .of the sequence of the pCR products showed thatwhen 10* artificially mutated molecules were added to 200 [gof DNA (-1:l0a dilution) only the sequence of the standarämutant could be derected (Figure 3B). When fewer (102, 1031artificially mutated molecules were added (-l:106 and -l:l0sdilution), both the wild-type and the mutated sequence couldbe seen. Since after addition of 102 molecules we could stilldetect the mutated sequence, we estimate that the sensitivitvof the ussay is I in 106.

C o

s"

*-=$+*Fr

**dsä4S

Ä{]#T A*üT A#TT

DEA/NO (pM) _IBMX +IBMX

0

0.1

I

l050

r00

13 + 6.15 + l583+26

170 + 31

168 + 13

155 + l0

29+53+

176 +398 +483 +362 +

t4l725

6029083

Values (l'mol cGMP/l0s cells) are expressed as mean + SD. n - 3_4."P < 0.05, DEA/NO versus control. Student,s l-test.

Stimulation of cGMP production in BEAS-2U cells treatedwith DENNOTo ascertain that NO released from DEA,/NO in the culturemedium was efficiently reaching the BEAS-2B cells, wemeasured the stimulation of the NO-sensitive guanylate cyclase(Table I) which is presenr in BEAS-2B (13). As shown fornormal human epithelial bronchial cells (13), 100 nM DEA/NO significantly increased cGMp production in BEAS-2B andno further stimulation could be observed at doses >10 pM.cGMP production was higher in presence of IBMX probablyas a result of inhibition of cGMp breakdown.

The cGMP production in INOS-transfected cells has beendescribed in previous studies (13).

)o(})cüo

---+ biotinylated exon Vll p53 RNA

=J-oo

+ streptavid ine_agarose

I

V

Fig. 3. Assessment of the sensitivity of the .fish-RFLp/pCR' protocol bvmixing an artiflcial standard mutant with DNA fiom unt.eat"d cells. tAiAnalysis of the PCR products by electrophoresis: lane l, molecular weightmarker; fane 2,200 pg of DNA digested by MspI; lane 3,200 pg of DNAdigested tty M^spI processed in rhe absence of biotinylated antisense RNA;lanes 4-6, 102, 103 or l0a molecules of artificial mutant standard added to2 trtg of DNA digestcd with M.r.pl; lanes 7-9, 102, 101 or 104 molecules ofartificial murant srandard added to 200 prg of DNA digested with MspI; lane

l9; "ll DN* rB) Sequencing analysis of rhe samples obtained using (a)

l0-. (b) l0' or 1c; l(tl molecules of artificial mutant standard added to200 pg of DNA digested with MspI, corresponding to lanes 7 9 of (A).

Table I. Stimulation of guanylate cyclase in BEAS-28 cells exposed toincreasing concentrations ol DEA,ß{O in presence or absence oi ßI4X(0.-5 mM)

2071

E.Fellev-Bosco el al.

ACC? ÄC*T ÄflGT

.il 5'

ACGT AfCT ACCT

Fig. 4. Sequencing analysis of samples obtained from (1) control BEAS,2Bceils, (2) cells transfected with murine rNOS gene, (3) cells exposed to 4mM DEA/NO, (4) cells exposed to 4 mM ENU, experiment 1, and (5) cellsexposed to 4 mM ENU, experiment 2. The 5' sequence analysis confirmsthe mutations found in samples 4 and 5.

Mutations in codon 248 of the p53 gene

To test whether NO was able to induce 5meC deamination incodon 248 of the p53 gene we applied the 'fish-RFLP/PCR'protocol to DNA samples obtained from BEAS-2B cellsexposed to NO either by endogenous production (2b-nos3 or2b-nos4) or by treatment with DEA/NO. In these samples, noC -+ T transitions were detected (Figure 4, samples 2 and 3).However, induced G -+ A mutations could be observed incodon 248 of p53 in cells treated with 4 mM ENU inexperiments 1 and 2 (Figure 4, samples 4 and 5). Thesemutations alter the MspI restriction site, providing a positivecontrol for the assay. From comparison with the calibrationexperiment using the standard mutant, we estimated that thefreq_uency of FNU-induced mutations was in the range of10-r to 2x l0-b.

HPRT selection

To buttress the results obtained using the 'fish-RFLP/PCR'method, we performed the HPRT- phenotypic mutation assay.The X-linked hprt gene is present once per diploid genome incells from male origin such as BEAS-2B. Only cells containinga mutated (HPRT ) gene will survive in the presence of theselective agent, 6-TG.

The median spontaneous frequency for sham-exposed (0 mMENU, 0 mM DEA/NO) BEAS-2B was 1.7X10-5 per colony-forming cell. Since the size of the hprt target has been estimatedas 1000 bp (17), it is possible to calculate the mutationfrequency at the level of the base pair. Assumifrg that crypticmutations will be 10 times more prevalent than 6-TG resistancemutations (18), base-pair mutation frequency is estimated tobe 1.7x10 5xl0 /1000 bp : 1.1x10-7.

The introduction of INOS or control expression vector didnot significantly modify HPRF frequency; neither did thetreatment of exponentially growing BEAS-2B cells with 4 mMDEA/NO (Figure 5). However, exposure of the cells to 2, 4and 8 mM ENU resulted in mutant frequencies of 2X104,4X l0-4 and 16X 10+ respectively. This represents a mutationfrequency l2-, 24- and 96-fold higher than the untreatedcontrol mutant frequency and varies in direct proportion withcell killing. However, at each ENU concentration the relative

2072

Af,ST

ffiACS?

ÄCGT

I

=

*

ne#?

ö"!'- r oo

z!

ti-

1--F:aasä E33*]itE=o E::

= 5; i E;EEE;;-N'Nf€NN

Fig. 5. Frequency of mutatton at the HPRT locus in BEAS-2B cells exposedto NO or ENU. Mean + SD (n - .1).

increase in mutant fraction was >1/survival. For example, at4 mM ENU the inverse of the survival is 1.5 (l/0.67). Thatmeans that plating cells of the 4 mM ENU treated populationmight cause a 1.5 enrichment of spontaneous mutants if weassume that they do survive ENU toxicity better. Since at4 mM ENU we observed an increase in the mutant fractionof 24-fold, preferential survival of spontaneous mutant cellscould not have caused the resoonse.

Discussion

Cytosines in CpG dinucleotides of the p53 gene are oftenmethylated (19). In this study we show that cytosine in codon248 (CGG) is fully methylated in different human rissues. Thisis consistent with the hypothesis that C to T transitionsobserved in CpG dinucleotides in human cancers might bedue to 5meC deamination (8,9), especially in the absence ofan efficient GIT mismatch repair (20-23).'spontaneous' 5meCdeamination occurs at a very low rate in the normal cell (24).However, deamination can be increased in vitro by aerobicexposure to NO (2,3), which is an endogenous product, witha basal production rate estimated as 1 mmol/person/day (25).NO is also a toxic air pollutant and cigarette smoke constituent,leading to significant inhalation exposure as well (26). Finally,many dn-rgs of the nitrovasodilator series release NO onactivation in vivo (27), and gaseous NO deliberately added asa pulmonary vaso- and bronchodilator to the breathing air ofpatients with certain types of respiratory distress is seeingincreasing clinical use (28). To test the hypothesis that exposureto NO might result in deamination of DNA in cells of thehuman bronchial epithelium, we have used a method whichallows the detection of a low proportion of mutant in a largepopulation of wild-type gene. This approach uses the RFLP/PCR technique (16,29-35) combined wirh affinity purificationof the target sequence (14,15). In the experimental conditionsdescribed in this study the 'fish-RFLP/PCR' method alloweddetection of G -+ A transitions in codon 248 of the p53 geneinduced by 4 mM ENU at a 2X10-6 to l0-s frequency. Asimilar frequency of single bp mutation has been recentlydescribed for 3229 fibroblasts exposed to ENU (29,33). Sincewe estimated a spontaneous base-pair mutation frequency of

t./'t - -\_)pair n--.-. - :mofe -:;--:'' --... .:Consirl-:.. - - -' -- -

f"-^, , -.rrtYu! ,.

fibrob '.- . :,: .,- -

...inr-treated cells. ENU-induced base-

.: l1S appears ro be 12- to 60-fold..r:re oLrS mutation, These results are

I t"lJ increa.e ir hr.c-pair mutation:'.rfrcd in codon l-18 of p53 in3229

r \1 E\U i33). In addition. the relative

NO and ENU mutagenicity in cultured human cells

In conclusion, the results obtained using both a new geno-typic approach and the classical HPRT mutation test, on cellseither endogenously producing NO or exposed to an NO-releasing agent. showed that NO is weakly or non-mutageniccompared to ENU. It cannot tre excluded that oxidized derivat-ives of NO (43) not fbrmed in our experimental conditionsmight produce mutations, and further studies are needed toaddress this question.

Acknowledgements

This work was supported in part by the Swiss National Science Foundation(fellowship to E.F.B. and granr 3l-39845.93 ro E.F.B.) and by the DeurscheForchungsgemeinschaft (fellowship to S.A.).

References

LLowenstein,C.J. and Snyder,S.H. (1992) Nitric oxide. a novel biologicmessenger. C e I l, 70, 7 05-'1 0'7 .

2, Wink.D.A., Kasprzak,K.S., Maragos,C.M. et al. (1991) DNA deaminatingability and genotoxicity of nitric oxide and its progenitors. Science,254,1001 1003.

3.Nguyen,T., Brunson.D., Crespi,C.L., Penman,B.W., Wishnok,J.S. andTanncnbaum,S.R. (.1992) DNA damage and mutation in human cellsexposed 1o nitric oxide in vitro. Proc. Natl At:ad. Scl. USA, 89, 3030-3034.

4. Isomura,K., Chikahira.M., Teranishi.K. and Hamada,K. (1 984) lnductionof mutations and chromosomc aberrations in lung cells tbllowing ln vluoexposr-rre of rats to nitrogen oxrdes. Mutat. Res., 136, ll9-12-5.

-5. Anoyo,PL., Hatch-Pigott,V., Mower,H.F. and Cooney,R.V (1992)Mutagenicity of nitnc oride and its inhibition by antioxidants. Maral.Res..28l. 193-202.

6. Routledge,M.N., Wink,D.A., Kecfcr,L.K. and Dipplc,A. ( 1 993) Murationsinduced by saturated aqueous nitric oxide in the pSPl89 slpF gene inhuman Ad293 and E.co1r MBM7070 cells. Carcino gene s i s, 14, I 25 1 -l 254.

7.Singer,B. and Essismann.J.M. (1991) Sitc-specific mutagenesis:retrospectrve and prospective. Carcinogenesis, 12, 949-955.

8. Hollstein,M., Sidransky,D., Vogelstein,B. and Harris,C.C. ( l99l ) p53mutations in huntan canccrs. Sclence, 253, .19--53.

9. Hanis.C.C. and Hollstein,M. ( 1993) Clinical implications of thc p53 tumorsuppressor gene. Nerr Engl. J. Med.,329, 1318-1321.

10.Cariello,N.F. and Skopek,T.R. (1993) Analysis of murations occurring atthe hurnan /rprt locus. J. Mol. Biol. . 231, 41 5'7 .

l l.Reddel,R.R., Ke,Y, Gerwin.B.L, McMenamrn,M.G., Lechner,J.F., Su,R.T.,Brash,D.E.. Park.J.-B., Rhirn,J.S. and Hanis,C.C. ( I 988) Transfbrmationof human bronchial epithelial cells by rnfection with SV40 or adenovirus-I 2 SV40 hybrid virus, or transfection via strontium phosphatecoprecipitation with a plasmicl containing SV.10 early region gener. CancerRes..48. 1904-1909.

I 2. Maragos.C.M.. N{orlcy,D., Wink,D.A. et uL. (1991\ Complcxcs of NO withnucleophiles as zrgents for rhe conrrollrd biological rclease of nitric oxide.Vasorelaxanr ef lccrs. "i. lr4 ed. C hern., 34. 32.42-3211 .

l.3.Felley Bosco,E., Ambs,S., Lowensiein,C.J., Keet'er,L.K. and Harris.C.C.( 199,1) Constitutive expression of inducible nitric oxide synthase in humanbronchial epithelial cells induces c-y'r_r and stimulatcs the cGMp pathway.Arn. J. Resprr. Cell Mol. Biol.,ll. 159 164.

14. Mirkovitch.J. and Darnell,J.E.,Jr ( l99l ) Rapid ln r,ilo footprinringtcchnique identifics proteins bound to the TTR gcnc in the mouse liver.Genes Dev.,5, 83 93.

l5.Mirkovitch.J. (1995) Mapping DNA-protein inreracrions in cclls andnuclei: genonic sequencing by ternplate purification. Methotls Enlmol.,1n press.

l6.Felley-Bosco,E., Pourzand.C., Zijlstra,J., Amstad,P and CeruttiP (1991)A genotypic u.tutation systcm nreasuring mutations in rcstriction recocnirionsequences. Nucleic Acid.s Re.s., 19, 2913 2919.

l7.Thilly.WG. (1990) Mutational spectrometry in animal toxrcrry testing.Antttt. Rey. Phannacctl. Toicol..30. 369 3U5.

18. Thilly,WG.. Liu,V.F., Brown,B.J., Cariello,N.F.. Kat.A.G. andKeohavong.P ( I989) Direct tneasurement of mutational spectra in hum:rns.Getutme, 3l, 590-593.

l9.Rideout.W.M.,lll, Coerzee,G.A.. Olumi.A,F änd Jones.PA. (1990) 5-Methylcytosinc as an cndogenous mutagen in the human LDL receptorand p.5-l gencs. Scrence, 249, 1288-1290.

20.Brown.T.C. and Jiricny,J. (1987) A specific mismatch repf,ir e\cnr prorecrsnammalian cells fl'om loss of 5 methvlcytosine. Cel1, 50. 9.15-9-50.

2l . Palombo,F.. Hughes,M., Jilicny,J., Truong,O. and Hsuan.J.( 1994) Mismatch

:-..r':..\ induced br E\U in the codon- .' .: r-;D9e similar to that observed in

thf i l/ri.i Sr':-.: .: - -:In thi: ::-.-. :'" r-.,JJ G -+ .\ trans:itions \\'ere observed

on the n.,:'- - -_- - ',:'.-.;>\unting that mutation results frometh\'lliti '- -- --:-.:t; irrllrrr\ed by replication. A strandbir: r' :r --'-- --;r,r'-.,:lerENUexposure(10);however,th-r- '.-: :, -: --- ,:d tt is not yet clear whether thisis due : t:':::-. - :.:'.:,ri the transcribing DNA strancl(36r rr : -- .*. - .:-: :i ir has recently been shown thatc)tosjn. :- :.-.. " ^ .. ihe alkylation of nearby guanine(38). sr. .r: *' - :\L -induced mutations might give adiffere nt :.' .. - . -:. iir is included. However. it isnote\\onh'. . - : "': - -.:o:ine did not fully protect nearbyguanine ;r. B: -.--: .- . :rudr ) or 3229 cells (33).

Usin.'s rh- .---'.. -: ?CR protocol with 1 in 106 sensitivitywe \\er. :-. - -^: ,: r.ii _.meC deamination in bronchialepithelial .;. , : .-i- .:-: .:-.-red with murine INOS or exposedto the \O-:::-. _. .,tr-:i DEAiNO. 2b-nos cells produce-100 pnro, \'-- '- -':: t:lrrein (13), which is similar to NOproduction , r.: : - .. :r'tulated macrophages. DEA/NO wasused at it i,-r-:' r: . l nt\{) -400-fo1d higher than theconcentratii t' i:: -: : : - ..*.i ntr\in1al stimulation of intracellu-lar solubic j-- . -: -,.t;t:e. Such a high dose was usedbecause rt f.:: ::: . .i. :t rre \ iously that a high concentrationof NO ts n:',-.-,,-. .r.ir\e a mutagenic effect in humancells (l). I::r:- - IK6 .ells exposed to such high NOcoilc€ntrJ.ilr,:.. :-: :1PRT ntutant fraction increased 4-foldcompared itr r:-r -: ,. :. rr hile cell survival was 35olc. Whencompared :! :,.: .:::r r!.iults. the lack of detectable NOmutagent!-rr,, .:: :: iS-lB cells may be due to at least twopossibilitie: F.:-. :: :-.-hial epithelial cells mighr be moreresistant th;f. i ::,--.. .\pes to NO's mutagenic effect due toa better s.;,. -:-:.:--: . :3nair svstem. Indeed, we have alreadyshown thar \O ,:. r. .\irrcssed in these cells at high amountsdurin-l scr ei;. r:,-:Sis I l3). Second, the mutagenic NOderit'atire \,,: .:. .. ILrnled in aerob:ic conditions might bemore reildil\ r:,,r,r.:d uhen NO is bubbled into the mediumor delir ered '. ir J:iiu:ron thror_rgh a gas-permeable membrane(39) than rn .rur .'.rndirions \\'here NO is intracellulariy synthe_sized or delir:rl.j bi an \O donor drug.

This rssue i:.-ontplicated by the recent discovery thatNO-induced nturarions in plasmids may differ depencling onwhether \O i: adminisrered as a gas (6) or released intosolution br dissolred NO donors (40). In änother plasmidstudy no -5-methr lcr tosine deamination was observed by eitherexposure route {+l ).

The results obtained using the genotypic mutation assay areconsistent with the analysis of the HPRT mutation assay.Exposure of the cells to 4 mM ENU increased -24-fold theHPRT mutation frequency. A similar induction has beendescribed in TK6 cells exposed to ENU when data obtainedwith the same cell survival are compared (42). Furthermore,ENU exposure affects both hprt and p53 genes to the sameextent. Consistent with the results of the genotypic assay, noincrease of the mutation frequency could be detected in 2b-nos or DEAA{O-exposed ce1ls.

201 3

l

1

i1

IIt

n

tlra

1frCl

2

E.Felley-Bosco el al.

repair and cancer. Ndture,367 ' 411 '

:z.r_"'"fr.,,n.s., Nicolaides,N.C,, papadopoulos,N . et .al. (1993) Murations oJ_a

lrlilt S homolog in hereditary'nonpolyposis colorectal cancer' Ce11' 75'

1r1< lfl<

Z:.nlstret,R.,Lescoe'M.K'Rao'MRS'Copeland'NG'Jenkins'NA'-- Cu.bei,t., Kane,M. antl Kolodner'R (1993) The human mutator gene

homologMsH2andltsassociationwithhereditarynonpolyposiscoloncancer. Cell, 75' 1027-1038'

24. Shen,J.-c., Rideout,wM. ill, and Jones,eA (199a) Th1 rate of hydrolytic

deamination of 5 methylcyttlsine in double-stranded DNA' Nucleic Acids

Res., 22, 9'12-916.25.Tannenbaum,S.R, Wishnok,J S and Leaf'C D . (1991) Inhibition of

nitrosamine fbrmation by ascorbic acid An J' Clin Nutr' 53' 211-250'

26. United Nations h,nvrronment Progrlmme \1911) Orirles of Nitrogen world-"

Healtfr Organization (Environmental Health Criterir 4,)' Geneva'

ZZ.ign"no,t-.i, Ross,G. and Tillisch,J' (1991) Pharmacology of endothelium

derived nitric oxi.le and nitrovasodilators ' West' J' Med" 154' 5l-62'

28.iapol,W.M., Rimar'S., Gillis'N, Marletta'M and -Bosken'C'H (1994)

Nitric oxide and the tung An J Respir' Crit Cure Med '149' 1375-1380'

29.Chiocca,S.M ' Sandy,MS ana Ceruiti'p'A (1992) Genotypic analysis of--

lv.iftyf-lV-"itiosourea-incluced mutations t:y TaqI restriction fragment

length polymoryhism/polymerase-chain reaction in c-H-ras1 gene Proc'

Natl Acatt. Sci. USA' 89,5331-5335'30.s;;;y,M.i., crrio.cu,s.v. unJ ceruni,pA. (1992) Genotypic analysis of

mutations in Zl4l restriction recognition sites by- restriction fragment

i."girt p.fvrrl",pi-ti,*/polyln"'u" Äain reaction Pror: Natl Acad' Sti'

usA,89, 890-894.3l . Ä;;ii";,F., Hussain'S.P and Cerutti'P ( 1993) Aflatoxin B I induces

-the- trin.u".rion of G_> T in codon 249 of the p53 tumor suppressor gene ln

hu."n h.p"to.ytes Proc Natl Acatl sci' USA' 90' 3586-8590'

32.Pourzand,C. and Ceruttr'P (1993) Mutagenesis of H-ras codons ll and

12 in human fibroblasts by N cthyl-N nitrosourea Carcittogenesis' 14'

2193-2196.33'Hussain,S.P,Aguilar'F'anclCerutti,P(1992{)Mutagenesisofcodon248

of the human p5-3 tumor suppressor gene by N-ethyl-N-nltrosourea'

Oncogene,9' 13-18'

34.;g;ii;;,F.,' ilanis'C C , Sun,T ' Hollstein'M .and Cerutti'P (1994)'

C?og-pfti. variation of p5-l mutational profile in nonmalignant human

liver. Sclence' 264' 1311-1319

35. Hussain,S.P, Aguilar,F., Amstaa'p and Cerutii'P (1994) oxy-radical

induce<l mutagenests of hotspot codons 248 and 249 of the human p53

gene. O n c o g ene, 9, 22T 7 -228136.Evans,M.K., Taft'e,BG, Harris'CC' and Bohr'VA (1993) DNA strand'"

Ui., it the repair of itre pS3 gene in normal iuman and xeroderma

pie-"ntoru- group C ltUrot'tastJ Cancer Res ' 53' 5377-5381'

3?. ö;i;;;;,F-' rir'i.ia,,s., c"lcagniie'A' Nehls'P and Doeliotti'E ( 1 992)

N-Methyl-N-nrtroi;outtu-'ndut"i mutations in human cells: effects of the

transcriptional activity of the target gene' J' Mot .-Biol" 223' 587-594'

38.Mathison,B.H ' Said'ts. anO snan't,Rlc (1993) Effect of 5 methylcytosine

as a neighboring ttur" on ncthylation of DNA guanine by N-methyl-N-

nitrosouiea. Carcinogenesis' 14' 323-321 '

39.Tamir,S., Lewis'R.S.' a" nojot Walker'T' Deen'WM ' Wishnok'J S and

Tannenbaum,S.R (1993) Th; influence of delivery rate on the chen.ristry

and biological etlects ofi nitric oxide Chem' Res Toxicol" 6' 895-899'

,10.Rout1edge"M.N., Wink.D.A' Kcefcr'LK and Dipple'A (1994) DNA-

.;;"""; changes inclucecl by two nitric oxide-donor drugs in the srpF

asiay. Chent. Res. Toricol . T ' 628 632'

4l.Schmutte,C., Ritteout'WM IIl, Shen'i -C' "n9 Jones'PA (1994)

' t ,tut"g.ni.iry of nitric oxide is not caused by deamination of cytosine or 5

methylcytosine rn double-stranded DNA iarcirtoginesis' 15' 2899-2903'

zl2.Bronstein,S.M.' co.n,"n",: g ' Craft'T'R ' Sqenb-crg'J A and Skopek'TR'

(1991) Toxrcrty' muu-g"nitiiy, and mutational -spectra of N ethyl N-

nitrosourea in human J"tt tintt with dilfcrent DNA repair phenotypes'

Cancer Res.,51, 5188 5197'

43.Beckrnan,J.S., g".t*"n,iW, Chen'J ' Marshall'PA and Freeman'B A-'

ii;öÄo;rln, rtuaro"vr radical Production by peroxynitnte: implications

fbr endothelial injury irorn nitric oride and superoride' Prtr' Nutl Atttd'

Sci. USA' 87, 1620-1624'

Ret:eivetl on Febrtutrl' t3, ]995; rel.ised tnl Juttt, 5, ]oa5: Ll(Lep|(\l tül 'Iunr'

16.1995

2014