Gene, 122 (1992) 377-380 0 1992 Elsevier Science Publishers

GENE

377

0378-1119/92/$05.00

06830

A pseudogene (Gene;

B.V. All rights reserved.

evolution;

for human glutathione

UGA:

Alan M. Diamond

sequencing;

cloning;

peroxidase

retrotransposition)

a, Rebecca Cruz a, Craig Bencsics a and Dolph Hatfield b

aDepartment of Radiation and Cellular Oncology. University of Chicago, Chicago, IL 60637, USA: and h Laboratory ofExperimentalCarcinogenesis, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892, USA. Tel. (301) 496-9263 Received by Y. Sakaki:

1 May 1992: Revised/Accepted:

20 July/ 21 July 1992; Received

at publishers:

3 September

1992

SUMMARY

Glutathione peroxidases (GPx) serve a bioprotective function in the reduction of peroxides to less toxic substances. Both cellular and secreted forms of the protein have been reported, as well a number of distinct cDNA sequences. Previous efforts have described three distinct loci on human chromosomes 3, 21 and X which hybridize to a GPX cDNA and these authors have speculated that only the chromosome 3 locus encodes a functional GPX gene. This conclusion was based on mapping studies showing a precise deletion of intron sequences in the GPX loci on chromosomes 21 and X despite strong conservation among these sequences in both the coding and 3’-untranslated regions. To pursue this issue, we have isolated the chromosome 21 GPX locus by molecular cloning and determined its nucleotide sequence. Consistent with the expectations of McBride et al. [Biofactors 4 (1988) 285-2921, the sequence does reveal a highly conserved processed pseudogene. It is suggested that a retrotransposed copy of the GPX gene integrated into chromosome 21 and may have maintained activity prior to the accumulation of inactivating mutations.

INTRODUCTION

Glutathione peroxidase (GPx; EC 1.11.1.9) is a selenoprotein which functions in protecting cells against oxidative damage. Selenium exists as a selenocysteine moiety within protein (Cone et al., 1976) and it occurs at the active site of GPx (Forstrom et al., 1978). Selenocysteine is donated to the growing polypeptide chain in response to certain UGA codons by a specific selenocysteine-inserting tRNA (Hatfield et al., 1990; Bock et al., 1991). UGA codes for selenocysteine in mammalian GPx (Chambers et al., 1986; Mullenbach et al., 19X8), selenoprotein P (Hill et al., 1991)

Correspondence to: Dr. A.M. Diamond, University of Chicago, MC0085, Chicago, IL 60637, USA. Tel. (312) 702-9193; Fax (312) 702-1968. Abbreviations:

bp, base pair(s): GPx, glutathione

encoding GPx; kb, kilobase reading frame.

peroxidase;

or 1000 bp; nt, nucleotide(s);

GPX, gene ORF. open

and type-1 iodothyronine 5’ deiodinase (Berry et al., 1991), as well as a number of prokaryotic proteins (Stadtman, 1991). At least two distinct but related forms of GPx have been described, a cellular and secreted plasma form (Takahashi et al., 1987). The peptide sequence for these proteins are also different from that predicted for two GPxrelated cDNAs isolated from human liver recombinant libraries (Dunn et al., 1989; Akasaka et al., 1990). Furthermore, McBride et al. (1988) detected three distinct loci in the human genome which hybridize to a GPX cDNA probe. These loci map to chromosomes 3,21 and X and only the locus on chromosome 3 hybridizes to a GPX intron probe. It was therefore suggested that the locus on chromosome 3 encodes a functional protein while the other loci encode processed pseudogenes. Moscow et al. (1992) have recently sequenced the gene on chromosome 3 and found it to be functional. Here, we have characterized the GPX hybridizing locus on chromosome 21 and show that it is a GPX pseudogene.

378 EXPERIMENTAL

TABLE

AND DISCUSSION

(a) Molecular cloning chromosome 21

of the GPX-hybridizing

locus on

To isolate genomic clones representing previously identified loci which hybridize to a GPX cDNA probe, a recombinant human ICharon28 phage library was screened by hybridization to a 32P-labelled human GPX cDNA probe (McBride et al., 1988). One million phage were screened and two clones with identical restriction enzyme digest maps were isolated. DNA of one of these clones was prepared for analysis. To map this clone in the region of the GPX homology, recombinant phage DNA was digested with EcaRI, BglII, PstI and XbaI alone and as double digests, electrophoresed in a 1% agarose gel, transferred to GeneScreen Plus hybridization paper by capillary action and probed with radioactively labelled human GPX cDNA. The resulting autoradiograph is shown in Fig. 1. Examination of these data reveals a digestion pattern which is consistent only with the restriction enzyme cleavage for the GPX-hybridizing locus on chromosome 21 (McBride et al.,

123456

78

9

IO

Fig. 1. Southern analysis of GP21. Phage DNA was digested as follows. Lanes: 1, X&I; 2, XbaI+BgZII; 3, BglII; 4, XbaI+Ps& 5, BgZII+PstI; 6, PsrI; 7, EcoRIiXbaI; 8, EcoRI+&$II; 9, EcoRI+PstI: 10, EcoRI. The digest in lane 7 is partial. Markings on the left margin indicate the positions of the n/l, standards obtained from HindIII digest of phage 1 DNA (9.4, 6.6, 4.4, 2.3, 2.0 and 0.56 kb). DNA was electrophoresed in a 1% agarose gel and transferred to GeneScreen Plus hybridization membrane (Du Pont) by capillary action. Human GPX cDNA ~ullenbach et al., 1987) was labelled by the random primer method (Feinberg and Vogelstein, 1983) and then used to probe the GeneScreen Plus membrane employing procedures suggested by the vendor. The resulting autoradiograph was used to determine Table I).

the A&s of the hybridizing

fragments

(see

I

Identification chromosome

of GPX-hybridizing 2 1a

sequences

as the Iocus from human

21

x

3

7.0; 0.9

5.0, 1.2

4.0, 1.0

3.9, 0.5

5.0, 0.6

5.0, 0.9

2.5, 1.0, 0.1

0.6, 0.5, 0.1

EcoRI + BglII

2.3, 0.6

2.1, 0.9

5.0, 2.5

2.9, 0.5

EcaRI + XbaI PSI Psi1 + 3grrr

2.5, 0.8

2.4, 1.3

5.6

6.0

1.5, 0.6 2.5, 2.3

3.3 5.4

3.1 3.4

3.3 3.3

3.1 3.1

Digest(s)

Size(s)

EcoRI EcoRI + PstI

PstI + XbaI Bg/II BglII t XbaI

a Digests and resulting

2.9, 2.3 2.9, 2.3 1.5

5.2, 0.6 3.7

4.0

2.9

sizes (in kb) of hybridizing

mined from the data shown in Fig. 1. Expected GPX-hybridizing the restricion

loci on chromosomes

1.7, 0.5 1.1, 0.6 0.9, 0.6

fragments

were deter-

sizes expressed

as kb of

21, 3 and X were determined

maps for these loci presented

in McBride

from

et al. (1988).

1988). As can be seen in Table I, only the loci from chromosomes 3 and 21 would be expected to produce a BglII fragment of approx. 3.0 kb, but only the locus on chromosome 21 lacks an internal PstI site. Similarly, the lack of an XbaI site between the Eg!II sites which iIank the hybridizing region of the clone is consistent with a chromosome 21 origin and elimates both 3 and X. Finally, the presence of only a single EcoRI site between these same BglII sites is also consistent with this clone being derived from chromosome 21. Based on this analysis, we conclude that this clone was derived from chromosome 21 and refer to it as GP21. (b) Nueleotide sequence of GP21 From the mapping data presented above, it was apparent that all the GPX hybridizing sequences in GP21 were present on a 3.3-kb BglII fragment which was subsequently subcloned into the BumHI site .of plJC18 in both orientations. It was also apparent that the hybridizing sequences contained an intern&&o RI -site which was eerate deletion clones. Sequencing was therefore performed with these deletion clones using universal plasmid sequencing primers and sequencing across the EcoRI site. As the sequence was generated, oligo primers were prepared to extend the analysis such that data were obtained from alternate strands. The nt sequence is shown in Fig. 2. (c) GP21 is a processed pseudogene The nt sequence of GP21 is compared to that of a human GPX cDNA (Fig. 2). The two sequences are very similar: 94% in the coding sequences and 88 “/;, in the portion of the cDNA 3’ to the translational stop codon. High similarity also occurs for 47 nt 5’ of the start codon of the cDNA sequence at which point there no longer are any significantly similar sequences in the 5’ direction. The sequence

379

GE21

TCTAGATACA

ACAAGATTTA

CAAAGTATTT

TTTTTGACAG

GACAGTTGCG

CCCTGTGTGC

GACAATTGCG *

CCATGTGTGC **

TGCTA

ACAG

wx

*

50

TTACTATTAT

TTTTGTTATT

TGTTTGTTCG

GGGCGCTCCC

CTAGCTTCTC

TGCTTGTTCG *

GGGCGCTCCG

CTGGCTTCTTG *

TGCTGGGCTA

GCGGCGGCGG

CTGCCCAGTC

TGCTCGGCTA

GCGGCGGCGG

CGGCCCAGTC

loo

Iso

GGTGTAAGCC

TTTTCCGCGC

GCCCGCTGGC

CGGCGGGGAG

CCTGTGAGCC

GGTGTATGCC

TTCTCGGCGC

GCCCGCTGGC *

CGGCGGGGAG

CCTGTGAGCC

TGGGCTCCCT

GCGGGGCAAG

GGACTACTTA

TCGAGAATGT

GGCGTCCCTC

TGGGCTCCCT *

GCGGGGCAAG

GTACTACTTA

TCGAGAATGT

GGCGTCCCTC *

GGAGGCACCA

CGGTCCGGGA

CTACACCCAG

ATGAACGAGC

CGCAGCGGCG

TGAGGCACCA *

CGGTCCGGGA

CTACACCCAG

ATGAACGAGC

TGCAGCGGCG *

CCTCGG

l

CCCCGGGGCCTGGT

GGTGCTTGGC

TTCCCGTGCA

ACCAGTCTGG

CCTCGGACCCCGGGGCCTGGT

GGTGCTCGGC

TTCCCGTGCA *

ACCAGTTTGG

GCATCAGGAG

AACGCCAAGA

ACGAAGAGAT

TCTGAATTCC

CTCAAGTACG

GCATCAGGAG **

AACGCCAAGA

ACGAAGAGAT *

TCAGAATTCC

CTCAAGTACG *

2oo

250

3oo

35o

4oo

TCCAACCTGG

TGGTGGGTTC

GAGCCCAGCT

TCATGCTCTT

GGAGAAGTGC

TCCGGCCTGG

TGGTGGGTTC

GAGCCCAACT

TCATGCTCTT * *

CGAGAAGTGC *

GAGGTGAACG GAGGTGAACG *t *

GTGCGGGGGC GTGCGGGGGC

GCACCCTCTC GCACCCTCTC **

TCCGCCTTTC TTCGCCTTCC *

TGCGGGACGC TGCGGGAGGC

5oo

CG GCCAGCCCCCAGCGACGA

CGCCACTGAG

CTCATGACCG

ACCCCAAGCT

55o

CCTGCCAGCTCCCAGCGACGA

CGCCACCGCG

CTTATGACCG

ACCCCAAGCT

CATCACCTGG

TCTCCGGTGT

GTCGCAACGA

TGTTGCCTGG

AACTTCTTTG

CATCACCTGG

TCTCCGGTGT

GTCGCAACGA

TGTTGCCTGG

AAC

** *

***

*

*

AGAAGTTCCT

GGTGGGCCCT

GACGGTGTGC

CTGTATGCAG

GTATAGCTGC

GGTGGGCCCT

GACGGTGTW

CCCTACGCAG

GTACAGCCGC

CGCTTCCAGA

CCATTGACAT

CGAGCCTGAC

ATCGAAGCCC

TGCTGTCTCA

CGCTTCCAGA

CCATTGACAT

CGAGCCTGAC

ATCGAAGCCC

TGCTGTCTCA

AGGGCCCAGA

TGTGCCTAGG

GCGCCCCTCC

TACCCCGACT

GCTTGGCAGT

AGGGCCCAGC

TGTGCCTAGG

GCGCCCCTCC

TACCCCGGCT

GCTTGGCAGT

TGCAGCGCTG

CTCTCT

TGCAGTGCTG

CTGTCTCGGGGGGGTTTTCATCTATGAGGGTG

TTTCCTCTAA

ACCTGCAAGG

AGGAACACCTGATCTTGCAGA

AAATACCCCC

TCGAGATGGG

AAATACCACC

TCGAGATGGG

ACCTACGAGGGAGGAACACCTGATCTTACAGA TGTCG

TTCATCCGA

TCTCTGCCA

6oo

TTTG

AGAAGTTCCT

GGGGGGTTTTCATCTATGAGGGTG

450

TTTCCTCTAA

ACCAGGGCGAGTTTCCCCACTAA

650

loo

750

800

851

a91

TGCTGGTCCTGTTGATCCCAGTCTCTGCCAGACCRAGGCG TAAAGTGC

GGGTAG

TAU,GTGCCGGGT

Fig. 2. Comparison cDNA

for human

AGC934

GTCAGCAAAAAAAAAAAAA

of the nt sequence determined GPX. The nt sequence

for GP21 to that of the

of GPZl (GenBank

accession

No. M93083) was aligned to the reported sequence of human GPX cDNA as reported (Sukenaga et al., 1987). The start and stop codons as well as the selenocysteine-encoding

TGA

are shown

in boldtype.

A putative

‘TATA’ box and polyadenylation signal are underlined. Differences in the coding portion of these sequences are indicated by an asterisk above the mismatch.

Cloning

and sequencing

busing both the Sequenaseversion Cleveland,

were as described 2.0 sequencing

OH) or a Model 373A automated

terns) such that both strands

in sections

mology between

the coding regions

of GP21 and GPX is

higher than that observed for other pseudogenes. For example, the pseudogenes for human a-globin and rabbit fl-globin have retained 76% and 82% sequence identity in their coding regions, respectively (Proudfoot and Maniatis, 1980; Lacy and Maniatis, 1980). Furthermore, there is a 3-nt in-frame addition in GP21 as compared to the GPX cDNA which would be expected to maintain an ORF. This retention of the coding potential of the cDNA might be indicative of evolutionary pressure to preserve the coding capacity of GP21. It is interesting to note that the UGA codon which specifies selenocysteine in the ORF of the

GPX cDNA has been changed to a glycine codon, GGA. It is possible that there was evolutionary pressure to limit the amount of mRNA with an in-frame UGA codon. If the GPX cDNA was transcribed following its reintroduction into the genome, then it would be predicted that integration needed to occur in the vicinity of nt sequences which could serve as a promoter. Such a sequence, displaying similarity to a ‘TATA box’ (Maniatis et al., 1987), is present at the -93 position of GP21. This is in a region described above which shows no similarity to the comparable region of the fuctional GPX gene on chromosome 3 (Moscow et al., 1992) and therefore was present on chromosome 21 prior to retrotransposition. (d) Conclusions We have isolated by molecular cloning the GPX hybridizing locus on chromosome 2 1. Analysis of the nt sequence reveals a processed pseudogene that has lost its start codon and ORF but maintains strong similarity to the coding GPX sequence present on chromosome 3. Thus, the presence of a poly(A) tract in the exact location of the poly(A) sequence of GP21 mRNA and the abrupt loss of similarity 5’ to the start codon argue strongly that GP21 represents a retrotransposed copy of a GPX mRNA.

a and

kit (U.S. Biochemicals,

sequencer

(Applied Biosys-

ACKNOWLEDGEMENTS

were determined.

of GP21 displays a number of properties which strongly suggests it is a processed pseudogene (Vanin, 1985). It lacks an intervening sequence present in the GPX locus on chromosome 3. In addition, it has lost the start codon, has an in-frame TAA stop codon at position 155 and has two frame-shift deletions at positions 307 and 503. A run of 20 A’s are also seen 3’ to the stop codon at a position corresponding to the polyadenylation site of a precursor GPx mRNA. Collectively, these data are consistent with the reverse transcription of a GPX mRNA and its subsequent introduction into chromosome 21. The 94% sequence ho-

The authors wish to thank Dr. Howard Lieberman for his assistance with DNA sequencing and Dr. G. Mullenbath for the human GPX cDNA probe. This work was supported by NIH grant ROl CA54364 to A.M.D.

REFERENCES Akasaka, M., Mizoguchi, J. and Takahashi, K.: A human cDNA for a novel glutathione peroxidase-related protein. Nucleic Acids Res. 18 (1990) 4619. Berry, M.J., Banu, L. and Larsen, P.R.: Type I iodothyronine deiodinase is a selenocysteine-containing enzyme. Nature 349 (1991) 438-440. Chambers, L., Frampton, J., Goldfarb, P., Affara, N., McBain, W. and

380 Harrison, P.R.: The structure of the mouse glutathione peroxidase gene: the selenocysteine in the active site is encoded by a termination

Maniatis, T., Boodbourn, S. and Fischer, J.A.: Regulation of inducible and tissue specific gene expression. Science 236 (1987) 1237-1245.

codon TGA. EMBO J. 5 (1986) 1221-1227.

McBride,

Cone, J.E., Del Rio, R.M., Davis, J.N. and Stadtman, characterization reductase:

of the selenoprotein

identification

of selenocysteine

ety. Proc. Natl. Acad. Dunn,

D.K.,

human

Howells,

cDNA

selenopeptide, Feinberg,

D.D.,

sequence GPRP.

Richardson,

moi-

B.: A technique

P.S.: A

DNA

to high specific activity. Anal. Bioof the cat-

as selenocysteine.

Bio-

17 (1978) 2639-2644. and function

Crit. Rev. Biochem..

Lee. B.J., Worland

of suppressor

P.J. and Oroszlan,

tRNAs

S.:

in higher eukaryotes

P contains

10 TGA codons

frame. J. Biol. Chem. 266 (1991) 10050-10053. Lacy, E. and Maniatis. T.: The nucleotide sequence pseudogene.

Cell 21 (1980) 545-553.

Moscow,

chromosomes J.A.,

in the open reading of a rabbit p-globin

Morrow,

Structure

tathione

G.T.,

Proudfoot,

to

and Cowan glu-

B.D.,

Bell, G.I. and Hahewell,

peroxidases.

T.: The structure

of a human

to alpha-globin

and function

and evolution reProt. Eng. 2 (1988) alpha-globin

duplication.

of selenocysteine

Cell 21

containing

J. Biol. Chem. 266 (1991) 16257-16260. K., Avissar,

characterization coprotein

G.T.

selenium-dependent

of incorporation

of three glutathione

and its relationship

maps

4 (1988) 285-292.

J. Biol. Chem. 267 (1992) 5949-

A., Irvine,

(1980) 537-544. Stadtman, T.C.: Biosynthesis enzymes.

cytosolic

mechanism

N.J. and Maniatis,

pseudogene

peroxidase

He, R., MulIenbach,

gene (&xl).

Tabrizi,

R.A.: Selenocysteine’s vealed in cDNAs ‘239-246.

C.S.,

Cl. and Hatfield, D.:

glutathione

3,21, and X. Biofactors

of the human

peroxidase

Takahashi:

Mol. Biol. 25 (1990) 71-96.

Hill, K.E.. Lloyd, R.S., Yang, J.-G., Reed, R. and Burk, R.F.: The cDNA for rat selenoprotein

human

5958. Mullenbach,

peroxidase-related

for radiolabelling

peroxidase

W.O., Mitchell, A., Lee, B.J., Mullenbach,

The gene for selenium-dependent

K.H.:

J.P. and Goldfarb,

J.J. and Tappel, A.L.: Identification

Hatfield, D.L., Smith, D.W.E., Structure

as the organoselenium

for a novel glutathione

alytic site of rat liver glutathione chemistry

glycine

Nucleic Acids Res. 17 (1989) 6390.

A.P. and Vogelstein,

J., Zakowski,

of clostridial

Sci. USA 73 (1976) 2659-2663.

restriction endonuclease fragments them. 132 (1983) 6-13. Forstrom,

component

T.C.: Chemical

N., Whitin, J. and Cohen, H.J.: Purification

of human plasma glutathione

peroxidase:

distinct from known cellular enzyme. Arch. Biochem.

phys. 256 (1987) 677-686. Vanin, E.F.: Processed pseudogenes: Rev. Genet.

19 (1985) 253-272.

characteristics

and

a selenoglyBio-

and evolution. Annu.