J. Biochem. 110, 407-411 (1991)

Genetic Linkage of Lung Cancer-Associated MspI Polymorphisms with

Amino Acid Replacement in the Heme Binding Region of the Human

Cytochrome P450IA1 Gene1

Shin-ichi Hayashi,* Junko Watanabe,* Kei Nakachi,** and Kaname Kawajiri*"Department of Biochemistry and "Department of Epidemiology, Saitama Cancer Center Research Institute, Ina-machi, Kitaadachi-gun, Saitama 362

Received for publication, April 25, 1991

Individuals with high genetic risk of lung cancer had previously been identified by Mspl

polymorphisms of the cytochrome P450IA1 gene. In the present study we analyzed the

structures of individual P450IA1 genes by PCR direct sequencing of genomic DNA of each

genotype raised by the MspI polymorphisms, which were ascribed to a single point

mutation in the 3•L-flanking region. We then found a novel point mutation in the coding

region of the gene which results in the substitution of Ile for Val at residue 462 in the heme

binding region. We further analyzed the genetic association between this amino acid

replacement and MspI polymorphisms in the general population, using a new method to

detect polymorphisms not recognized by restriction enzymes. The results showed that there

are at least two forms of human P450IA1 protein with different primary structures and that

one of the forms is closely linked with the lung cancer-susceptible genotype of MspI

polymorphisms. Thus MspI polymorphisms, which are associated with increased risk of

lung cancer, are linked to at least one amino acid substitution, which gives an important

clue, at the molecular level, toward elucidation of increased susceptibility to lung cancer.

A large proportion of human cancers are known to be caused

by synthetic or natural chemical compounds in the environ

ment (1, 2). Many chemical carcinogens are metabolically

activated to forms that have deleterious effects on organ-

isms (3), and this metabolic activation is an obligatory

initiation step in human chemical carcinogenesis. The

microsomal electron transport system, including cyto

chrome P450s, plays the most important role in oxidation

of chemical carcinogens (4, 5), and involves a variety of

isozymes of P450. This oxidative activation shows genetic

variation (4, 6), which may be responsible for individual

differences in susceptibility to chemical carcinogenesis.

Cytochrome P450IA1 is important in the initiation of lung

cancer because it is responsible for the activation to

mutagens of benzo [a] pyrene and other aromatic hydrocar

bons in cigarette smoke (7, 8).

We recently found a close correlation between develop

ment of lung cancer and MspI polymorphisms in the

3•L-flanking region of the P450IA1 gene, where three

genotypes were determined the predominant homozygote

(genotype A), the heterozygote (genotype B), and a homo

zygous rare allele (genotype C) (9). Namely, the incidence

of genotype C of the P450IA1 gene was 21.2% in lung cancer

patients, but only 10.6% in healthy controls. This genetic

change was specifically correlated with increased risk of the

Kreyberg I type of lung cancer, which is closely associated

with cigarette smoking, but not with the Kreyberg II type.

Furthermore, our case-control study showed that the

relative risk of genotype C was much higher (7.31-fold)

than those of genotypes A and B at low levels of cigarette

consumption (10). Therefore, it is indispensable to investi

gate whether this MspI polymorphism is genetically as

sociated with differences in protein structure or gene

expression of P4501A1. On the basis of this investigation,

the interindividual difference in the basal or induction

response of P450IA1 can be elucidated.

In the present study we analyzed individuals with

genotypes A and C by PCR direct sequencing. We found that a novel point mutation in genotype C resulted in an

amino acid replacement in the heme binding region of

P4501A1 and also showed that at least two forms of

P4501A1 protein with different primary structures exist

among humans. We further determined the genetic associa

tion of this novel mutation with MspI polymorphisms in the

general population, developing a new detection method to

identify the polymorphisms not recognized by restriction

enzymes. It was found that these two loci, causing the

observed DNA polymorphisms, were very closely linked.

EXPERIMENTAL PROCEDURES

Materials-Restriction endonucleases were purchased

from Takara Shuzo (Kyoto). [ƒ¿-35S]dATP (1,000Ci/

mmol) and [ƒÁ-32P]ATP (5,000Ci/mmol) were obtained

from Amersham (Arlington Heights, IL). All other re-

agents were of the highest quality commercially available.

Isolation and Sequence Analysis of the 3•L-Region of the

P4501A1 Gene-A human gene library of Charon 4A

1 This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan, and a research grant from the Ministry of Health and Welfare

of Japan.Abbreviations: P450IA1, previously called P450c; PCR, polymerase chain reaction; RFLPs, restriction fragment length polymorphisms; XRE, xenobiotic responsive element; BTE, basic transcription element.

Vol. 110, No. 3, 1991 407

408 S. Hayashi et al.

bacteriophage was re-screened using an EcoRI-EcoRI

1-kbp fragment of the Tend of the clone AhP450mc-1 as a

probe (11). The clone ăhP450mc-2 which contained the

3•L-flanking region of the gene was isolated, and its EcoRI-

EcoRI 4-kbp fragment was subcloned in pUC18 (named as

phP450mc-2). Unidirectionally deleted subclones gener

ated by the Erase-a-Base system (Promega, Madison, WI)

were used to obtain sequence information on the 3•L-flanking

region. Sequencing was carried out for both strands by the

chain-termination method (12).

Detection of MspI Polymorphisms-Individual DNAs

were isolated from peripheral lymphocytes in blood sam

ples obtained from a cohort of 2,500 Japanese persons all

over 40 years old. The genotypes of the P450IA1 gene

ascribed to the Mspl site were identified as RFLPs by the

polymerase chain reaction (PCR) (13) and were incomplete

agreement with our previous Southern blotting (9). The

PCR-amplified DNA fragments including the polymorphic

site were digested with Mspl and subjected to electropho

resis in 1.8% agarose gel.

PCR Direct Sequencing of Individual Genomic DNAs-

DNA fragments of 0.5 to 1.5kbp were prepared by

PCR-amplification and used as templates for direct se

quencing. Synthesized 21-nucleotide oligomers were end-labeled with [ƒÁ-32P]ATP by kination, and 4pmol of the

labeled primers were used for sequencing reaction together

with 200ng of purified PCR template using a Sequenase

sequencing kit (USB, Cleveland, OH). Oligo-nucleotides

for sequencing primers and PCR primers (named as C1 to

C51) were synthesized on an Applied Biosystems Model

381 DNA sythesizer.

Detection of Polymorphisms Not Recognized by RFLP-

Two oligonucleotides of 20 mer (primer 2a, primer 2b),

both of which contained the polymorphic site at the 3•L end,

were synthesized and each of them was used as a primer for

PCR-amplification together with another strand of 21 mer

primer (primer 1) which is located about 200 by upstream of a polymorphic site detected by sequencing. PCR was

performed by 25 cycles under the following conditions: 1

min at 95•Ž for denaturation and 1 min at 70•Ž for primer

annealing and primer extension. The other conditions were

as described by Saiki et al. (13). The PCR products were

then subjected to electrophoresis in 1.8% agarose gel.

RESULTS AND DISCUSSION

Structural Analysis of the 3•L-Flanking Region of

P450IA1•\Previously we cloned the human P450IA1 gene

and determined the nucleotide sequence (11), although the

cloned gene did not include the 3•L-flanking region where the

polymorphic Mspl site is located. To identify the MspI site

and analyze the 3•L-flanking region of the gene in this study,

we re-screened a human genomic library and obtained clone

ăthP450mc-2 for analysis. A 4-kbp fragment including the

3•L-flanking region was subcloned in a plasmid vector

(phP450mc-2) and then sequenced. Based on the sequence

information, primers were synthesized, and DNAs of types

A and C were subjected to PCR direct sequencing. Figure 1

Fig. 1. Structural analysis of the 3•L-region

of the P450IA1 gene. (A) Nucleotide sequence of

the 3•L-flanking region of the gene. Clone

A hP450-mc2 was isolated, and subclone

phP450mc-2 was sequenced as described under "EXPERIMENTAL PROCEDURES

." PCR direct

sequencing was carried out using the primers

indicated by broken lines, and the polymorphic

site was identified. The nucleotides with numbers

up to 5985 in the margins were transcribed in

mRNA, and untranscribed nucleotides are num

bered from 1. (B), RFLPs of PCR-amplified

fragments by MspI or SmaI. The primers indicat

ed by solid lines in (A) were used for PCR

amplification, and the PCR products from type A,

B, and C DNAs were digested with MspI or SmaI,

and subjected to agarose gel electrophoresis.

Lanes 1, 2, 3, and M are type A (ml/ml), B (ml/

m2), C (m2/m2), and size markers (SV40-

HindIII), respectively. Allele ml and m2 are

defined by the absence and presence, respective-

ly, of the one MspI site detected by RFLPs in our

previous paper (9).

J. Biochem.

Genetic Polymorphisms of Cytochrome P450IA1 409

Fig. 2. Structural analysis of individual P450IA1 genes by PCR-direct sequencing. The DNA fragments amplified by PCR are shown

by horizontal bars. Closed circles and squares show the primers used for the two opposite directions in PCR. These PCR products were sequenced directly by the chain-termination method using synthesized sequencing primers. Coding sequences are indicated by boxes, and leader and trailer sequences of mRNA are indicated by boxes of lower height. The solid arrow shows the site of MspI polymorphisms, and the open arrow shows that of the novel mutation in the coding region. Solid triangles, the open circle, and the open triangle indicate the locations of the XRE, BTE, and TATA box, respectively.

Fig. 3. Novel polymorphic mutation in the coding region. DNAs from individuals were examined by PCR-direct sequencing. A, B, and C show the results for an Ile/Ile homo-zygote, Ile/Val heterozygote,

and Val/Val homozygote, respectively. The point mutation is indicated by an arrow.

Fig. 4. Comparison of the amino acid sequences of the heme binding region in the P450IA1 subfamily. The heme-binding cysteine residue is designated as 0 (22). The amino acid residues at the polymorphic position are boxed. The sequences of all members of the IA1 subfamily identified so far are shown (23).

summarizes the results. Replacement of thymine in genotype A by cytosine in genotype C was observed at the 264th base downstream from the poly A additional signal, form-ing an MspI or Smal site in genotype C. For confirmation of this observation, PCR-amplification was carried out with DNAs of types A, B, and C as templates and the oligonucleotide primers indicated in Fig. IA. The products were then digested with MspI or Smal. The fragment amplified from type A DNA gave only a single undigested band at the

position of 340 bp, the fragment from C type gave bands of digested DNA at 200 and 140 bp, while the fragment from

B type gave three bands at 340, 200, and 140 by (Fig. 1B).

This procedure was used for determining the genotypes of

P450IA1 in a large number of individuals, and allowed

identification of persons at genetically high risk of lung

cancer.

Amino Acid Replacement in the Heme Binding Region of

P450IA1-For determination of whether there was any

other mutation site in the P450IA1 gene linked with MspI

polymorphism, the regulatory regions in the 5•L-flanking and

coding regions of the P450IA1 gene were sequenced direct-

ly using PCR-amplified fragments from the DNAs of types

A and C (Fig. 2). In the 5•L-flanking region including the XRE

(xenobiotics responsible element) (14), BTE (basic tran

scription element) (15), and TATA box, the DNA sequences

of type A and C were identical.

However, we found a difference of one base at position

4889 in the 7th exon. As shown in Fig. 3, adenine in type A

was replaced by guanine in many individuals of type C. This

novel point mutation resulted in replacement of Ile by Val

at residue 462 in the HR2 region (16), which was well

conserved among P450 families. The cysteine in this region

has been shown to be the heme-binding thiolate ligand (17,

22), and thus this region is essential for the catalytic

activity of this enzyme. As this polymorphic position is

conserved as Ile among members of the IM subfamily (Fig.

4), the substitution of Val may affect the catalytic activity

of this enzyme.

Vol. 110, No. 3, 1991

410 S. Hayashi et al.

Fig. 5. Detection of the point mutation of Ile-Val polymorphism. (A) Scheme of PCR for the detection. Two primers (primer 1 and primer 2a, or primer 1 and primer 2b) were used for each DNA sample, and PCR was performed according to the method described in "EXPERIMENTAL PROCEDURES." The point mutation site is indicated by asterisks. (B) Representative results of PCR to identify Ile-

Val polymorphisms. A, Ile/Ile homozygote; B, Ile/Val heterozygote; C, Val/Val

homozygote. Primers 1 and 2a were used for lane 1 and primers 1 and 2b for lane 2; lane

M, size markers (ƒÉ-EcoRI•EHindIII).

TABLE I. Estimated frequencies of combined genotypes. Numbers of subjects examined are shown in parentheses.

'These frequencies were observed in our previous study (10) .

The RFLP method cannot be used to detect this polymor

phism because there is no suitable restriction site. There-fore, for use in screening, we developed the new detection method described in "EXPERIMENTAL PROCEDURES." The primers used in this method are shown in Fig. 5A. Figure 5B shows the clear profiles that identified the two homo-zygotes (Ile/Ile and Val/Val) and the heterozygote (Ile/ Val). The combination of primers 1 and 2a gave a PCR product of 210 by (lane Al) when (Ile/Ile) was used as a template. The combination of primers 1 and 2b gave no PCR product (lane A2): the product appeared only with the combination of primers 1 and 2b with (Val/Val) as tem-plate (lanes C1, C2), and with both combinations with (Ile/ Val) as template (lanes B1, B2). The results by this new method were completely consistent with those obtained by direct sequencing.

Linkage Analysis in the General Population-We found the DNA polymorphisms at two loci of the P450IA1 gene to be due to point mutations. Therefore, we next investigated the genetic association between these two loci in the general population. In our previous study on MspI polymorphisms, we observed frequencies of 0.443, 0.451, and 0.106 for genotypes A (ml/ml), B (ml/m2), and C (m2/m2), respectively, in 375 healthy controls (10). Our next problem was to determine the frequencies of the genotypes (Ile/Ile), (Ile/Val), and (Val/Val) in the populations with genotypes A, B, and C. To do this, we randomly selected 43

and 47 individuals from among those with genotypes A and B, respectively, and all the 45 individuals with genotype C observed so far. The observed frequencies of (Ile/Ile), (Ile/ Val), and (Val/Val) in the groups of genotypes A, B, and C were then weighted by the frequencies of A, B, and C. The estimated frequencies are shown in Table I as a 3 x 3 genotype table.

From the table we could calculate the respective fre

quencies f(ml; Ile), f(m2; Ile), f(ml; Val), and f(m2; Val) of the genes ml with Ile at residue 462, m2 with Ile, ml with Val, and m2 with Val, assuming the Hardy-Weinberg equilibrium for f (ml; Ile) and f (ml; Val) and also for f (m2; Ile) and f (m2; Val) weighted by the observed gene frequencies f(ml)=0.668 and f(m2)=0.332. The gene frequencies were thus estimated as f(ml; Ile)=0.661,

f(m2; Ile)=0.092, f(ml; Val) =0.007, f(m2; Val)=0.240, f (Ile) =0.753, and f (Val) =0.247. The frequencies expect-ed from Table I agreed well with observed values within a range of 0.03.

Linkage disequilibrium D (24) was then calculated to be 0.158, and the linkage coefficient R described by Hill and Robertson (25) to be 0.778, showing that these two loci are closely linked. To understand the genetic association more clearly, we calculated the occurrence of linkage, i.e., p(ml; Ile)=f(ml; Ile)/f (ml)=0.990, p(ml; Val)=0.010, p(m2; Ile)=0.272, p(m2; Val)=0.723, p(Ile; ml)=f(ml; Ile)/ f(Ile)=0.878, p(Ile; m2)=0.122, p(Val; ml)=0.028, and p(Val; m2)=0.972. These results clearly showed 99% linkage of ml with Ile and 97% linkage of Val with m2.

We analyzed the gene structures of the individuals with different susceptibility to lung cancer identified by Mspl polymorphisms of P450IA1. As a result, a novel polymorphic site was then found in the DNA sequences of these individuals which results in an amino acid mutation in the heme binding region. The existence among human of at least two forms of P450IA1 protein in association with MspI polymorphisms was thus revealed. To clarify this

genetic association, we further assessed the frequencies of the genotypes with a combination of these two polymorphic loci among the general population using the new detection method and showed that these two loci were very closely linked.

Recently the reasons for phenotypes of extensive and

poor metabolism of debrisoquine have been explained by DNA polymorphisms of P450IID at plural polymorphic sites (26, 27). However, there is no previous report of

J. Biochem.

Genetic Polymorphisms of Cytochrome P4501A1 411

analysis of the linkage between multiple polymorphic loci of human P450s in the general population. Such analyses are essential for estimating susceptibilities to drugs and chemical carcinogens, because the susceptibility or risk must be evaluated as the ratio of the genotype frequency in the susceptible subpopulation (or patients) to that in the general population (28). It is, of course, important to characterize the effect of this novel mutation and MspI

polymorphisms on phenotypic expression of P450IA1 at the molecular level. We are now investigating these successive questions, which involve biological assays of the activity of the enzyme with the mutation, influence of MspI polymorphisms on expression, association with arylhydrocarbon hydroxylase activity together with epidemiological analysis of cancer risk.

We are grateful to Dr. Yusaku Tagashira for his support during this study, and Nahomi Shinoda for technical assistance.

REFERENCES

1. Nagao, M. & Sugimura, T. (1978) Annu. Rev. Genet. 12,117-1592. Doll, R. & Peto, R. (1981) The Causes of Cancer, pp. 1256-1260,

Oxford University Press, Oxford, New York3. Heidelberger, C. (1975) Annu. Rev. Biochem. 44, 79-1214. Guengerich, F.P. (1988) Cancer Res. 48, 2946-29545. Conney, A.H. (1982) Cancer Res. 42, 4875-49176. Gonzalez, F.J. (1988) Pharmacol. Rev. 40, 243-2887. Aoyama, T., Gonzalez, F., & Gelboin, H.V. (1989) Mol. Carcinog.

1, 253-2598. McManus, M.E., Burgess, W.M., Veronese, M.E., Huggett, A.,

Quattrochi, L.C., & Tukey, R.H. (1990) Cancer Res. 50, 3367-3376

9. Kawajiri, K., Nakachi, K., Imai, K., Yoshii, A., Shinoda, N., & Watanabe, J. (1990) FEBS Lett. 263, 131-133

10. Kawajiri, K., Nakachi, K., Imai, K., Hayashi, S., &Watanabe, J. (1991) in Proceedings of the 21th International Symposium of the Princes Takamatsu Cancer Fund (Ernster, L., ed.) Japan Sci. Soc. Press, Tokyo/Taylor & Francis, London, in press

11. Kawajiri, K., Watanabe, J., Gotoh, 0., Tagashira, Y., Sogawa,

K., & Fujii-Kuriyama, Y. (1986) Eur, J. Biochem. 159, 219-22512. Sanger, F., Nicklen, S., & Coulson, A.R. (1977) Proc. Natl. Acad.

Sci. U.S.A. 74,5463-546713. Saiki, R.K., Gelfand, D.H., Stoffel, S., Scharf, S.J., Higuchi, R.,

Horn, G.T., Mullis, K.B., & Ehrlich, H.A. (1988) Science 239, 487-491

14. Fujisawa-Sehara, A., Sogawa, K., Yamane, M., & Fujii-Kuriyama, Y. (1987) Nucleic Acids Res. 15, 4179-4191

15. Yanagida, A., Sogawa, K., Yasumoto, K., & Fujii-Kuriyama, Y. (1990) Mol. Cell. Biol. 10, 1470-1475

16. Gotoh, 0., Tagashira, Y., Iizuka, T., &Fujii-Kuriyama, Y. (1983) J. Biochem. 93, 807-817

17. Poulos, T.L., Finzel, B.C., Gunsalus, I.C., Wagner, G.C., & Kraut, J. (1986) J. Biol. Chem. 260, 16122-16130

18. Heilmann, L.J., Sheen, Y.Y., Bigelow, S.W., & Nebert, D.W. (1988) DNA 7, 379-387

19. Okino, S.T., Quattrochi, L.C., Barnes, H.J., Osanto, S., Griffin, K.J., Johnson, E.F., & Tukey, R.H. (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 5310-5314

20. Kimura, S., Gonzalez, F., & Nebert, D.W. (1984) J. Biol. Chem. 259,10705-10713

21. Sogawa, K., Gotoh, 0., Kawajiri, K., & Fujii-Kuriyama, Y. (1984) Proc. Natl. Acad. Sci. U.S.A. 81, 5066-5070

22. Kawajiri, K., Gotoh, O., Sogawa, K., Tagashira, Y., Muramatsu, M., & Fujii.Kuriyama,Y. (1984) Proc. Natl. Acad. Sci. U.S.A. 81,1649-1653

23. Nebert, D.W., Nelson, D.R., Adesnik, M., Coon, M.J., Estabrook, R.W., Gonzalez, F.J., Guengerich, F.P., Gunsalus, I.C., Johnson, E.F., Kemper, B., Levin, W., Phillips, I.R., Sato, R., & Waterman, M.R. (1989) DNA 8, 1-13

24. Vogel, F. & Motulsky, A.G. (1986) in Human Genetics, pp. 162-164, Springer-Verlag, Berlin, Heidelberg, New York, London, Paris, Tokyo

25. Hill, W.G. & Robertson, A. (1968) Theor. Appl. Genet. 38,226-231

26. Kagimoto, M., Heim, M., Kagimoto, K., Zeugin, T., & Meyer, U.A. (1990) J. Biol. Chem. 265, 17209-17214

27. Gough, A.C., Miles, J.S., Spurr, N.K., Moss, J.E., Gaedigk, A., Eichelbaum, M., & Wolf, C.R. (1990) Nature 347, 773-776

28. Khoury, M.J., Adams, M.J., Jr., & Flanders, W.D. (1988) Am. J. Hum. Genet. 42, 89-95

Vol. 110, No. 3, 1991