Eur. J. Biochem. 205,217-222 (1992) (CiFEBS 1992

Structure of the human ferrochelatase gene Exon/intron gene organization and location of the gene to chromosome 18 Shigeru TAKETANI', Johji INAZAWA', Yoshitsugu NAKAHASHI', Tatsuo ABE' and , Kikio TOKUNAGA' Department of Hygiene, Kansai Medical University, Osaka, Japan ' Departmcnt of Hygiene, Kyoto Prefectural Univcrsity of Medicine, Kyoto, Japan (Received October 29/Decembcr 16, 1991) - EJB 91 1450

We have determined the structure of the human ferrochelatase gene after isolation and characterization of lambda phage clones mapping discrete regions of the cDNA. This gene was assigned to human chromosornc 18 at region q21.3, by fluorescent in situ hybridization. The gene contains a total of 11 exons and has a minimum size of about 45 kb. The exonlintron boundary sequences conform to consensus acceptor (GTn) and donor (nAG) sequences, and the exons in the gene appear to encode functional protcin domains. A major site of the transcription initiation, determined by S1 nuclease mapping, was assigned to an adcnine base 89 bases upstream from the adenine base of the translation initiation ATG. The promoter region contains a potential binding site for Spl, NF-E2 and erythroidspecific transcriptional factor GATA-1, but not a typical TATAA or CCAAT sequence. Analysis of primer extension showed that the transcription starts at the same position between hepatoma HepG2 and erythroleukemia K562 cell mRNA, thereby suggesting that there can be a single transcript in erythroid and non-erythroid cells.

Ferrochelatase (protoheme lerro-lyase) is the terminal enzyme of the heme biosynthetic pathway and catalyzes the insertion of fcrrous ion into protoporphyrin 1X to form protoheme. The enzyme is located in the innermembrane ofmilochondria of various tissues including liver [l, 21, heart [3], kidney [4] and erythroid cells [4, 51. A genetic ferrochelatase deficiency in infants results in crythropoietic protoporphyria with excessive accumulation and excretion of protoporphyrin [6]. Protoporphyria in humans is a dominantly inherited disease. The enzyme activity decreases to about 50% compared with normal levels in all tissues and isolated cell preparations, including bone marrow [7], liver [8] and cultured skin fibroblasts [9]. The molecular basis of the defect has apparently not been reported. Regulation of the heme synthetic pathway differs between erythroid and non-erythroid cells. In liver and non-erythroid tissues, the rate of heme synthesis is controlled by the levcl of the first pathway enzyme, 5-aminolevulinic-acid synthase [6]. In differentiating erythroid cells, a different type of 5aminolevulinic-acid synthase is present and cannot be ratelimiting for heme production [lo]. It has been shown that hemc synthesis increased when mouse erythroleukemia cells were induced to differentiate to cells producing hemoglobin,

following treatment with dimethylsulfoxide [ll]. Mutationally or chemically induced blockage of ferrochelatase results in thc arrest of the normal erythroid differentiation [12]. Accordingly, control of heme synthesis and the induction of fcrrochelatase related to erythroid diffcrentiation has been the subject of increasing interest. Recently we cloned and Characterized mouse [13] and human ferrochelatase cDNA [14]. In mouse erythroleukemia cells, ferrochelatase mRNA increased during differentiation and the size of the mRNA in mouse erythroleukemia cells was the same as that in mouse liver. In addition, the size of ferrochelatase mRNA in human erythroleukemia K562 cells was similar to that seen in hepatoma cells, thereby suggesting that ferrochelatase in erythroid and hepatic cells can only be of one type. To elucidate the regulation of its expression as well as to determine molecular defects in erythropoietic protoporphyria, we characterized the ferrochelatase gene and mapped this gene to chromosome 18q21.3. The structural organization of the gene and the nucleotide sequence of its 5'-flanking region are described in detail.

EXPERIMENTAL PROCEDURES

-~ .-

Materials

C'orrespondence t o S. Taketani, Department of Hygiene, Kansai Medical University, Mariguchi, Osaka, Japan 570 Enzymes. Ferrochelatase (EC 4.99.1 .l); 5-aminolevulinic acid synthase (EC 2.3.1.37); porphobilinogen deaminase (EC 4.3.1.8). Now. The nucleotidc scquence data published hcrc have been deposited with the EMBL scquence data bank with an accession number X63760.

[y-32P]ATP(6 kCi/mmol), [z-"PIdCTP (3 kCi/mmol) and nylon membrancs were obtained from Amersham. Restriction cndonucleases and other nucleic-acid-modifying enzymes were from Toyobo Co., and Takara Shuzo Co. The human genomic libraries in EMBL 3 were kind gifts from Dr. M. Yamamoto, Tohoku University. Oligonucleotides were

21 8 IF 7

IF8

IF 1

I F 5

,I F 4

1

2

I

I

3 4 I

I

I

5 6 7

I

I I

I

I

8 91011 1

1 1

1

I I

I I

BE

B0 E

E

B E

B

E

E

E

BE

EE E

B

u

I1 I

I

I I

I

I

I

I

II

I1

I

I

c-3 5 kb

Fig. 1. Exon/intron organizationof the human ferrochelatasegene. The direction of transcription is from left to right. Five overlapping genomic clancs wcrc used to characterize the gene. Exons are represcntcd by boxes: closed and open boxes represent protein-coding and untranslated regions. respectively. The restriction map obtained with two of the enzyme used (B, BumHI; E, EcoRI) is shown.

synthesized on an Applied Biosystem Model 381 A DNA synthesizer. Human hepatoma HepG2 cells were grown in Dulbecco's modified Eagle's medium containing 10% fetal calf serum, and erythroleukemia K562 cells were grown in RPMI 1640 medium containing 10% fetal calf serum [14]. All other chemicals used were of analytical grade. Screening of genomic libraries and sequence analysis Human genomic libraries in EMBL 3, a lambda replacement vector were screened by plaque hybridization [I 51 for cDNA segments encoding ferrochelatase. The hybridization probes used were an equimolar mixture of the EcoRI - EcoRI fragment (nucleotide -25 - 1508) and the EcoRI-EcoRI fragment (nucleotide 71 3 -2488) excised from the human ferrochelatase cDNA, I.HF2-1 and AHFI-2, respectively [14], and labeled with [ M - ~ ~ P I ~by C the T P random-priming method [13]. Hybridization-positive clones were isolated by repeated plaque hybridization and the isolated DNA inserts were subcloned in the pUC 18 plasmid vector. The subcloned DNA fragments were used for further analysis and nucleotide sequences were determined. Fluorescent in situ hybridization To identify the chromosomal localization of the human ferrochelatase gene, fluorescent in situ hybridization was performed on human metaphase chromosomes, essentially as described by Lawrence et al. [16], with minor modifications. Chromosome preparations were made from normal female lymphocytes by the thymidine-synchronization 5-bromodeoxyuridine release technique, for replicating the chromosome bands [17] and subjected to fluorescent in situ hybridization after staining with Hoechst 33258, followed by exposure to ultraviolet light. The 15-kb genomic ferrochelatase DNA fragment designated as AF4 (see Fig. 1) was labeled with biotin-1 6-dUTP by nick translation. After overnight hybridization, the biotinylated probes were detected with avidin/fluororescein isothiocyanate (Boehringer) and metaphase chromosomes were counterstained with propidium iodide. The hybridization signals were observed using a Nikon fluorescent microscope (filter combination: Nikon B-2A or B-2E) and replication G-bands were visualized on the same metaphase chromosomes through an ultraviolet-2A filter (Nikon), thereby allowing for an unambiguous assignment of the hybridization signals.

SI nuclease-mapping analysis For the S1 nuclease-mapping analysis, total cellular RNA was isolated from human erythroleukemia K562 cells [14].

The PvuII-AcyI fragment (nucleotide -118 - 135) was labeled at the 5' ends with [ Y - ~~ P I A Tusing P , T4 polynucleotide kinase. The resulting labeled fragment was hybridized with K562 RNA (90 pg) at 46°C for 3 h in 20 p180'X formamide containing 0 .4 M NaCI, 1 mM EDTA and 40 mM Pipes, pH 6.4 [15], and digested with S1 nuclease at 37'C for 30 min. The products were fractionated by electrophoresis on a sequencing gel. Primer-extension analysis Total RNA was isolated from human hepatoma HepG2 and erythroleukemia K562 cells and poly(A)-rich RNA was isolated by oligo(dT) -cellulose column chromatography [13]. A synthetic oligonucleotide 5' d(CTCCCAGACAGTCGGTACCTCCACCTTCAGT) complementary to nucleotide position 84- 114 of the cDNA [14] corresponding to a part of exon 2, was used as an extension primer. After completion of the 5'-extension reactions, the samples were digested with 100 ng ribonuclease A at 37'C for 30 min [15]. The remaining DNA was ethanol-precipitated and denatured in 90% formamide. All the samples were separated by electrophoresis on a 5 % polyacrylamide/urea gel.

RESULTS AND DISCUSSION Screening and isolation of the ferrochelatase gene Two types of human genomic libraries in EMBL 3 were screened with 32P-labeled cDNA segments for human ferrochelatase [14]. Among 5 x lo6 plaques, four positively hybridizing clones were obtained, and these were purified and characterized. Restriction mapping and Southern blot analyses of the four clones (1F1, IF4, 3.F5 and E.F7) suggest that the four clones represent other distinct and overlapped loci, but do not contain the 5' upstream region from the internal Sac11 site of the human ferrochelatase cDNA. Screening of the libraries with the 5'-end fragment (0.5 kb) of clone i F 4 yielded clone 1F8 (Fig. 1). The insert of E.F8 was positively hybridized with the 5'-end fragment of the cDNA. The DNA fragments carried by overlapping genomic clones, AFl, IF4, iF5, I F 7 and i F 8 encoded the entire region of the human ferrochelatase gene. Blot hybridization analysis of human lymphocytes DNA revealed that the cloned ferrochelatase gene retains the same sequence organization as present in the genomic DNA and that there is a single gene for ferrochelatase in the human genome (data not shown).

219

Fig. 2. Fluorescent in situ hybridization of the human ferrochelatase gene. (A) Partial metaphase showing the specific fluorescein-isothiocyanate signals (arrows) of hybridization to both chromosomes 18. (B) The same metaphase was subsequently identified by replication G-banding with the filter combination of ultraviolet-2A.

5 ’ End

Exon

I - TCCGCgtaagtgggtctgtcc

ctttaatttgtcacagTGGCA

-

Exon

2

-

AAGAGgtatgagtgtctaaca

attttattttatatagGAAGC

-

Exon

3

-

CAGAAgtgagatatatataac

Organization of the human ferrochelatase gene

To characterize the ferrochelatase gene, fragments spanning most of the gene were subcloned into pUC 18 for DNA sequence analysis. The data derived from the systematic analya t g t t a a a t g a t t t a g T A A G C - Exon 4 - CACAGgtatggtgtttcttca sis of all five plaques demonstrated that a single gene of at t g t g t c t t c c t a a t a g C C C C T - Exon 5 - CACAGgtaagggcatctctct least 45 kb encodes ferrochelatase (Fig. 1). The gene consists of a total of I 1 exons and 10 introns (Figs 1 and 3). The 5’6 - TCCAGgtaagccagtgcttgc t c t c t g t a t a t t t c a g G C A G C - Exon flanking region, the first exon and about 4.5-kb of intron 1 t c t a t t a c t c a c t c a g T G C T T - Exon 7 - TGTCTgtaagtaagaacattt are included in AF8. Exon 1 is composed of a 5‘ untranslated g t g t c c t c t g c t g c a g G T G G T - Exon 8 - CCAAGgtaagtggcttccaca sequence and the first 22 amino terminal residues, which corresponded to the first half of the putative leader sequence t c t a t c t t t c g c a t a g G T T G G - Exon 9 - AGGAGgtaaatgcacccatct [14]. Exon 11 encodes the cdrboxyl-terminal domain of ferrot t c t c a t t t g a t t c a g T G T G G - E x o n 10 - C T A A G g t a t c t a c a g t g t t a c chelatase of 0.27 kb and contains the entire 3‘ untranslated sequence of 1.15 kb ferrochelatase messages. The sequences c t t g t c t a t a t t c c a g G C C C T - Exon 1 1 - T C A G h t c c a g a g t c t g a c t t c of the exon/intron boundaries for the human ferrochelatase Fig. 3. Nucleotide sequence of intron/exon junctions in the human gene are shown in Fig. 3. The consensus dinucleotides GT at ferrochelatase gene. Thc exon sequence is shown in uppercase lettering the 5’ and AG at the 3’ intron boundaries are conserved for and the intron sequence is shown in lowercase lettering. The start of each intron [ 181. In addition, pyrimidine-rich sequences which the poly(A) tail of the human ferrochelatase cDNA is underlined. immediately precede the 3’ AG boundaries of the intron are present for all of the introns in the ferrochelatase gene. We made use of S1 nuclease mapping to examine the 5’ end of ferrochelatase messages. The PvuII --AcyI fragment Chromosome localization of the ferrochelatase gene (nucleotide residues - 118 - 135) labeled at the AcyI site (see To determine the chromosomal localization of the human Fig. 6) was used in conjunction with total cellular RNA from ferrochelatase gene, fluorescent in situ hybridization was K562 cells. One major protected DNA fragment was detected performed on human metaphase chromosomes using the (Fig. 4). By comparison of the position of the major protected genomic fragment of ferrochelatase, 1F4. Of 16 metaphase fragments with the sequence ladder of the S1 probe, we cells observed, 6 (37.5%) had a green spot located symmetri- assigned the A residue of the nucleotide at position I , as a cally on each chromatid in both homologs of chromosome 18 major transcription/initiation site. Primer extension was performed with an antisense oligoat region q21.3 (Fig. 2). The others showed either incomplete (50?4) or undetectable signals (12.5%). Thus, the human nucleotide primer complementary to part of the exon 2 seferrochelatase gene was mapped to human chromosome 18 at quence, which corresponded to nucleotide position 84- 1 14 region q21.3. of the human ferrochelatase cDNA [14]. When mRNA from

220

A

Fig. 4. Transcriptionlinitiation site of the human ferrochelatase gene. The initiation site was determined by S1 nuclcase mapping. The S1 probc was the PiwlI - AcyI fragment (nucleotide residues - 118 135) end-labclcd at the AcyI site. The protcctcd fragments were analyzed on a 6% polyacrylamide/7 M urea gel. Lane 1, control products with yeast tRNA. Lanc 2, protected fragments with K562 cell RNA. Complementary nucleotides around the initiation site are shown along the sequence laddcrs. An asterisk indicates the major transcriptionlinitiation site. ~

HepG2 and K562 cells were used as templates, a major and common extended product consisting of 202 nucleotides was obtained in both RNA (Fig. 5). Thus, the 5’-untransbated region of the ferrochelatase gene consists of 88 nucleotides, a finding consistent with the results by S1 nuclease mapping. In addition, 5’dternative splicing of the transcript of the ferroehelatase gene does not occur in erythroid-type and hepatic-type cells. Thus, regulation of expression of the ferrochelatase gene in erythroid cells can differ from the mechanism by which tissue-specific expression of porphobilinogen deaminase, the third enzyme in heme biosynthetic pathway, is controlled by the utilization of an erythroid-specific promoter situated 3’ to the house-keeping promoter [19, 201. 5’-Flanking region of the ferrochelatase gene The sequencc of the 5’-flanking region of the human ferrochelatase gene is shown in Fig. 6. Sequences upstream as

Fig. 5. Primer-extension analysis of the human ferrochelatase messages. (A) Schematic representation of the experiments.The oligonucleotide primer (thick line) complementary to the sequence within exon 2 corresponding to nucleotides at positions 84- 114 of the human

ferrochelatasecDNA, was synthesized as dcscribcd under Expcrimental Procedures. The primer was hybridized to HepG2 or K562 cell mRNA (1Opg) and extended. The nucleotide at position 1 of the human ferrochelatasecDNA is shown in Fig. 6. (B) Autoradiography of extended fragments. The reaction products were separated by electrophoresis on a 5% polyacrylamide/7 M urea gel. Lanc 1, size marker (HinfI-digested fragments of pBR322). Lane 2. control product of yeast tRNA. Lane 3, the extension product of HcpG2 ccll mRNA. Lane 4, the extension product of K562 cell mRNA. The major transcription/initiation site is shown by an arrow.

well as downstream of the transcription start site were examined to search for distinguishing elements of the gene promoters. No obvious TATAA and CCAAT boxes were evident at their consensus positions. The region was found to be abundant in G C , a structural feature which has been shown to demarcate the 5’4anking regions of various genes [21]. A prominent feature is the preponderance of G C boxes, G G G C G G (CCGCCC), the recognition site of the Spl transcription factor [22]. We found these G C elements at positions - 600 - - 595, - 591 - - 586, - 115 - - 110 and -73 - -68 (Fig. 6), thereby indicating that the gene is comparable to many house-keeping promoters [23]. The sequence CTGGGA (nucleotide residues - 580 - - 575) is consistent with the putative sequence of the interleukin-6 responsive element [24]. Sassa and co-workers [25, 261 suggested that

+

221 -680

-670

-660

-650 -640 -630 GAGGTGTAGG AATGAAATGC GCGCCTGTTC

CTGAACATAT ATAGGCAGGG TGGGCTGACG

-620 -610 CTCCCAGTGA TCTCAGTGTG

-600

-590

-580

-570

C G A C C T T G G F ~ C A C B G G C G C T A ATGGGAAGGTG AC

-560 -550 -540 -530 -520 -510 CGCGGATATG ATACAGACAA ACCTCATCAT ACCGTCCCTC CAAGAAATGC ACTTGCCAGG

-500

-490

CTGGCCTTCA CTTTTCCTAA

-480 -470 -460 -450 TGCTCTTCAG TGAGTTTCAC CTGTCAAATT CTGCTACACA

-440 -430 -420 GAGAGGGTGC AGAGAAACTG AATAATCCCT

-410 -400 -390 T T T T C T A T T C TGAAAATAAA GCAAAACAAA

-380 -370 -360 -350 -340 -330 ACAGGATCTG CACACTTGTG TTGTGTGACT AGGCAGGCAA ACTGGAGAAC TGAGGCAAAA -320 -310 GCATCCTTAT ATCTAAATAA

-300 -290 -280 TAGTAATGGA AACGTCGAAA AAAGTTATTT

-180 -170 -160 -150 TGCAGTGACC TGCGATGGTG GTGGGGAGCG GGCTTCTAGC TCAGCCCCCG

-190

-

I40 -130 -120 -110 -100 -90 GGCCCATTTT CACGCAGGGA G C G G C G C C C A G - G T m GCTCTCCGCG AGGGGCGTGT

Pna -80 -70 -60 CTCTGCCTGG CC I C G I $ m @ G G C C C G G

-50 -40 -30 CAGCGAATGA GCGGGCGCCG GGAGGGCGCG

-1r -10 11 21 -20 31 AGGTCAGGGG GCTGGGGACG CGCGTGGGGA TCGCTACCCG GCTCGGCCAC TGCTGGGCGG 41

51

ACACCTGGGC GCGCCGCCGC

101

111

61 71 81 91 GGGAGGAGCC CGGACTCGGG CCGAGGCTGC C C A G G C A A ~

*

121

131

141

CGTTCACTCG GCGCAAACAT GGCTGCGGCC CTGCGCGCCG CGG-GGGJGCT

We thank President I. Tsukahara of Kansai Medical University for his encouragement, Dr. M. Yamamoto for the kind gift of human genomic libraries and valuable advice, K. Yasaka for excellent technical assistance, M. Ohara for comments and F. Shigenobu for sccretarial services. This study was supported in part by grants for Scientific Research from the Ministry of Education, Science and Culture of Japan (No. 03265201,03202138).

-270 GCTGAGTCAT

-240 -260 -250 -230 -220 -210 GGCTGAGGAT CCTGACTTCG CTAGTTTGGC AGATGCAAAC AGGGCACGCA ACTAGGAGlC -200 CAGCAGGTTT

regulation of the ferrochelatase gene expression are underway.

151 GCTCCGCGAT

Acyl

Fig.6. Nucleotide sequence of the 5'-flanking region of the human ferrochelatase gcne. Nucleotide residues arc numbered in the 5'-to-3' direction, bcginning with the transcription start site. The nucleotides on the 5' sidc of residue 1 arc shown by negative numbers. An arrow points 10 thc transcription/initiation site. Thc ATG codon for initiation melhionine is overlined. The nucleotidc at position l of the human l'crrochelatase cDNA is indicated by an asterisk. Spl-binding scquences are boxcd. Underlined scquences include. NF-E2 binding site (TGAGTCA); interleukin-6-responsive element (CTGGGA); GATA-I binding site (ATATCT).

stimulation in heme biosynthesis might occur during the acutephase response a n d th at treatment of H e p G 2 cells with interleukin 6 resulted in stimulation of 5-aminolevulinic-acid synthase [27]. It is possible t h at this p r o mo t er element in the ferrochclatase gene functions in expression of the enzyme. Finally, there is a possible GATA-1 binding site a t position - 321 - - 31 6 (ATATCT) (Fig. 6), in the opposite orientation. which conforms to five o u t of six bases to t h c consensus sequence (WGATAR: W = A/T and R = A / G ) [29]. This DNA motif serves in p a r t as a signature for erythroid-specific gene expression and is present in t h e promoters o f globin [29, 301, erythroid-specific 5-aminolevulinic-acid synthase [31, 321 a n d porphobilinogen deaminase genes [33].It should be noted that t h c GATA-1 binding site of t h e mouse a n d h u m a n erythroid-specific 5-aminolevulinic-acid synthase genc also differs f r o m the consensus sequence by t h e last single base [31, 321. T h e scquence T G A G T C A (nucleotide residues -277 - -271) is identical to the consensus sequence o f the N F - E 2 binding site where NF-E2 recognizes a n AP-I-like scquence [34]. T h e NF-E2 motif was first demonstrated in the erythroid porphobilinogen deaminase p r o mo t er [33, 341 a n d more recently, multiple sites h av e been f o u n d in the upstream cnhancer or the /3-globin gene 135, 361. The NF-E2 sequence in th e ferrochelatase gcne is identical t o those observed in human erythroid 5-aminolevulinic-acid synthase [32] and porphobilinogen deaminase promoters [34]. Attempts t o clarify a functional role f o r a n y of these sequences in the

REFERENCES 1. Taketani, S. & Tokunaga, R. (1981) J . Biol. Chem. 256, 1274812753. 2. Tdkctani, S. & Tokunaga, R. (1982) Eur. J. Biochem. 127,443447. 3. Taketani, S., Tanaka-Yoshioka, A., Masaki, R., Tashiro, Y. & Tokunaga, R. (1986) Biochim. Biophys. Acta 883, 271283. 4. Nakahashi, Y., Taketani, S., Sameshima, Y . & Tokunaga, R. (1990) Biochim. Biophys. Acta 1037, 321 -327. 5 . Hanson, J. W. & Dailey, H. A. (1984) Biochem. J . 222, 695700. 6. Kappas, A,, Sassa, S., Galbraith, R. A. & Nordmann, Y . (1989) in The metabolic basis of inheriteddisease (Scriver,C.R., Beaudt, A. L., Sly, W. S. & Valle, D., eds) pp. 1305- 1365, McGrawHill Book Co., New York. 7. Bottomley, S. S., Tanaka, M. & Everett, M. A. (1975) J . Lab. Clin. Med. 86, 126-131. 8. Bonkowsky, H. L., Bloomer, J. R., Ebert, P. S. & Mahoney, M. J . (1975)J. Clin. Invest. 56, 1139-1148. 9. Bloomer,. J . R. (1980) J . Clin. Invest. 65, 321 -323. 10. Riddlc, R. D., Yamamoto, M. & Engel, J. D. (1989) Proc. Nut1 Acad. Sci. USA 86,792 -796. 11. Marks, P. & Rifkind, R. A. (1978) Annu. Rev. Biochem. 47,419 448. 12. Sassa, S. (1983) in Regulation of hemoglobin biosynthesis (Goldwasser, E., ed.) pp. 359-383, Elsevier, New York. 13. Taketani, S., Nakahashi, Y., Osumi, T. & Tokunaga, R. (1990) J . Bid. Chem. 265, 19377 - 19380. 14. Nakahashi, Y . ,Taketani, S., Okuda, M., Inoue, K. & Tokunaga, R. (1990) Biochem. Biophys. Res. Commun. 173, 748755. 15. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular cloning: a laboratory manual, 2nd edn, Cold Spring Harbor University Press, Cold Spring Harbor. 16. Lawrence, J. B., Villnave, C. A. & Singer, R. H. (1988) Cell 52, 51-61. 17. Drouin, R., Lemieux,N. & Richer, C.-L.(1990) Chromosoiiia 99. 273 - 280. 18. Breathnach, R. & Chambon, P. (1981) Annu. Rev. Biochem. 50, 349 - 383. 19. Chretien, S., Dubart, A., Bcaupain, D., Raich, N., Grandchamp, B., Rosa, J., Gosscns, M. & Romeo, P.-H. (1988) Proc. Natl Acad. Sci. U S A 85, 6 - 10. 20. Beaumont, C., Porcher, C., Picat, C., Nordmann. Y. & Grandchamp, B. (1989) J . Biol. Chem. 264,14829- 14834. 21. Bird, A. P. (1986) Nature 321,209-213. 22. Kadonaga, J. T., Jones, K . A. & Tjian, R. (1986) Trends Biochc~m. Sci. I I , 20-23. 23. Wilcox, D. E., Affara, N. A., Yates, J. R. W. & Ferguson-Smith. M. A. (1985) Hum. Genet. 70,365-375. 24. Hattori, M., Abraham, L. J., Northemann, W. & Fey, G. H. (1990) Proc. Natl Acad. Sci. U S A 87,2364-2368. 25. Iwasa, F., Galbraith, R. A. &L Sassa, S. (1988) Biochem. J . 253, 970 - 973. 26. Galbraith, R. A., Sassa, S. & Fujita, H. (1988) Biochem. Biopph-vs. Res. Commun. 153,869 - 874.

222 21. Iwasa, F., Sassa, S. & Kappas, A. (1989) Biochem. J . 25Y, 605607. 28. Whitelaw, E., Tsai, S.-F., Hogben, P. & Orkin, S. H. (1990) Mol. Cell. Bid. 10, 6596 -6606. 29. Wall, L., de Boer, E. & Grosveld, F. (1988) Genes Dev. 2, 10891100.

30. Orkin, S. H. (1990) Cell63, 665-672. 31. Schoenhaut, D. S. & Curtis, P. J. (1990) Nucleic Acids Rex 17, 7013 -7028. 32. COX,T. C., Bawden. M. J., Martin, A. & May, B. K. (1991) EMBO J . 10, 1891-1902.

33. Mignotte, V., Wall, L., de Boer, E., Grosveld, F. & Romeo, P.-H. (1989) Nucleic Acids Res. 17, 3 1 54. 34. Mignotte, V., Eleouet, J. F., Raich, N. & Romeo, P.-H. (1989) Proc. Natl Acad. Sci. U S A 86, 6548 - 6552. 35. Philipson, S., Talbot, D., Fraser, P. & Grosveld, I;.(1990) EMBO J . 9,2159-2167. 36. Ney, P. A,, Sorrentino, 13. P., Lowrey, C. H. & Nienhuis, A . W. (1990) Nucleic Acids Res. 18, 601 1-6017. ~