Like intracellular GPX (5) and plasma GPX (61,. PHGPX proved to be a selenoprotein (2,4) but, according to the limited sequence information available, ...
THEJ o m u . OF BIOWCIC~. CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc
Vol. 269,No. 10,Issue of March 11, pp. 7342-7348,
1994 Printed in U.S.A.
Phospholipid-hydroperoxide Glutathione Peroxidase GENOMIC DNA, cDNA, AND DEDUCED AMINO ACID SEQUENCE* (Received forpublication, February 16,
1993, and in revised form, September 24, 1993)
Regina Brigelius-FloheSI, Klaus-Dieter Aumann$ Helmut Bliickefl, Gerhard Gross$ Michael Kiessn, Klaus-Dieter Kloppeln, Matilde Maiorinoll, Antonella Roverill, Rainer SchuckeltS, Fulvio Ursini**, Edgar Wingender$ and Leopold FlohenSS From the llGesellschaft fur Biotechnologische Forschung (GBF), 0-38124Braunschweig, Germany, the [Department of Biological Chemistry, University of Padoua, I-35121Padova, Italy, the SGrunenthal GmbH, Center of Research, 0-52078 Aachen, Germany, and the **Department of Chemistry, University of Udine, I-33100Udine, Italy
The complete amino acid sequence of the selenoprotein phospholipid-hydroperoxide glutathione peroxidase (PHGPX) frompig heart hasbeen deduced from the corresponding genomic DNA, the cDNA covering the coding region,and by sequencing the N terminus of the protein. The maximum length of the peptide chain derived from the cDNA amounts to 170 amino acid residues. By protein sequencing the N-terminal residues methionine and cysteine of the deduced sequence were found to be cleaved. The molecular mass of 19,671 Da obtained by laser desorption mass spectroscopy, however, significantly exceedsthe mean molecular massof 19,257.09 calculated for the sequence 3-170 of PHGPX, thus indicating posttranscriptional modification. In contrast to glutathione peroxidase (GPX) the coding area of the PHGPX gene is composed of seven exons. Only the amino acid sequences encoded by the thirdand fifth exon are highly homologousto GPX sequences. The aminoacid residues selenocysteine, tryptophan, and glutamine forming the catalytic site inbovine GPX are conserved in homologous positionsof PHGPX, whereas the arginine residues presumed to bind GSH in GPX are not. Gapsin thePHGPX sequence correspond to subunit interaction sites of the tetrameric GPX The data suggest an identical catalytic mechanism of the selenoperoxidases, a less stringent substrate specificity of PHGPX, and explain the monomeric nature of PHGPX As in otherselenoproteins,the selenocysteine residue of PHGPX is encoded by UGA. The 3”untranslated region ( U T R ) of the PHGPX shows a limited consensus with that of GPX and 5’-deiodinase, whereit was shown to be responsiblefor the decoding of UGA as selenocysteine. The 3’-UTR of PHGPX can form a s t e d o o p as in other mammalian selenoprotein genes. The 5’-UTRand the first intronof the PHGPX gene contain a variety of putative regulatory elements indicating hormonal control.
Phospholipid-hydroperoxide glutathione peroxidases (PHGPX)I were functionallycharacterized asa peculiar typeof glutathione peroxidases (GPX) reducing hydroperoxides of phospholipids (1, 2) and cholesterol (3). Recently, partial sequencing of pig heart PHGPX and of the encoding cDNA (4) revealed that PHGPX, although functionally and structurally related topreviously known GPX species, represents a distinct gene product. Like intracellular GPX (5)and plasma GPX (61, PHGPX proved to be a selenoprotein (2,4) but,according to the limited sequence information available, appeared to be a remotely related member of the glutathione peroxidase family (4). PHGPX first attracted interest by its ability to directly interact with peroxidized lipids even if these complex lipids are integrated in biomembranes, whereas GPX can only interfere with lipid peroxidation in a concerted action with phospholipases (1-3, 7, 8). PHGPX has therefore been considered the main lineof enzymatic defense against oxidative biomembrane destruction, whereas the role of GPX is believed to consist in the protection of membrane lipids from attack by soluble hydroperoxides or, supported by phospholipases, in biomembrane of PHGPX in repair (7). The presumed outstanding importance the defense against oxidative stress was corroborated by the observation that it is hardly affected by dietary selenium deprivation and most avidly utilizes the supplemented trace element in refeeding experiments (8). Other observations, however, do not favor the assumption that PHGPX could play a relevant role in the defense mechanism againstoxidative stress. Firstof all, compared with GPX, its activity is lower by more than 1 order of magnitude in most tissues (8, 9). Also, the tissue distribution of PHGPX is exit is tremely unusual. is It particularly high in testes, but there only expressed after puberty (101, and, most surprisingly, the testicular PHGPX level of rats could be kept close to zero by hypophysectomy and was shown to respond to gonadotropin treatment (10). Finally, a putative phosphorylation site (4) and a change in its subcellular distribution during spermatogenesis have been reported (10).Evidently, such findings suggest a specific role in cellular differentiation rather than a general antioxidant function. The intriguing characteristics of this enzyme prompted US to isolate and sequence the corresponding full-length cDNA and the genomic DNA and to determine theamino acid sequenceof the mature protein by direct proteinsequencing. The DNA
* The costsof publication of this article were defrayed in part by the payment of page charges. This article must thereforebe hereby marked “advertisement” in accordancewith 18 U.S.C.Section 1734 solely to indicate this fact. The nucleotide sequence(s) reportedin this paper has been submitted to the GenBankTMIEMBLData Bank with accession numbed4 X76008 and X76009. 5 Presentaddress: Deutsches Institut fur Ernahrungsforschung, Arthur Scheunert-Allee 114-116, D-14558 Bergholz-Rehbriicke, Ger1 The abbreviations used are: PHGPX, phospholipid-hydroperoxide many. glutathione peroxidase; GPX, glutathione peroxidase; UTR, untrans$$ To whom correspondence should be addressed: Gesellschaft fur Weg 1, D-38124 latedregion;HPLC, high pressure liquid chromatography; bp, base BiotechnologischeForschung(GBF),Mascheroder paids); PCR, polymerase chain reaction. Braunschweig, Germany.
7342
Phospholipid-hydroperoxide Peroxidase Glutathione
7343
was obtained and sequencedas follows. Total RNA (6 pg) andpoly(A)+) RNA (1pg) were isolated fromporcine parathyroid glands and reverse transcribed withMoloney murine leukemiavirus reverse transcriptase (Life Technologies A. G., Basel, Switzerland) primedby oligonucleotide WW774 corresponding to positions 2207-2188 of the PHGPX DNA EXPERIMENTAL PROCEDURES shown in Fig. 2. Reaction products were separated by a short run in low gelling agarose (1.5%), and the nucleic acids of 250-1300 bp were exPHGPX Measurement tracted.Theisolated cDNA-RNA hybridsweredenatured,andthe PHGPX activity was measured by the coupled test procedure devel- cDNA products were dA-tailedat the 3'-end with terminal transferase. oped for GPXdetermination (11) butwithphosphatidylcholine hyThe specific cDNA was then isolated essentially according to the RACE droperoxidase as substrate (12). One unit of activity is defined as the protocol designed for the rapid amplification of cDNA ends of rare amount of enzyme catalyzing the reduction of 1p o l of hydroperoxide/ transcripts by repeated polymerase chain reaction(13).The first round min at 37 "C and pH 7.4. of 30 PCR cycles was performed with the nested primer WW774 and %R,-T17 (14) as the reverse primer. ThePCR products of 200-600 bp Preparation of PHGPX were isolatedas above and subjected toa second round of 30 PCR cycles Procedure I-Porcine heart PHGPX was prepared as described by using the outer adapter primerFb (13) and the nested primerWW773 Ursini et al. (1) but replacing KSCN during chromatography by a n (positions 2161-2137 in Fig. 2). A 450-bp band was isolated from the equimolar concentrationof KCl. The specific activity of the preparation PCR products and amplified in a third round of 30 PCR cycles performed as before. The material thus obtained was sequenced directly by was 131 unitdmg protein. means of a commercially available Sequence Kit with T7DNA polymerProcedure 2-700 g of ventricular muscle from pig heart were minced ase (United States Biochemical GmbH, Bad Homburg, Germany) and and homogenized in 3 volumes of ice-cold 10 m phosphate buffer, pH 7, containing 5 m mercaptoethanol. The homogenate was centrifuged for5'-end-labeled sequencing primers, WW772, UK 1079, and UK 1078, 30 min at 10,000 x g , and the supernatant was centrifuged for 60 min corresponding to positions1840-1815, 1679-1669, and 616-635 of the at 100,000 x g . The supernatant was filtered through gauze and, aRer PHGPX DNA (see Fig. 2), respectively (seebelow). adding 5% glycerol, was applied to a bromosulfophthalein glutathione Genomic DNA Analysis affinity column. The column was equilibrated with 10 xm phosphate, pH 7, containing 5 xm mercaptoethanol and 10% glycerol and washed The sequence of the genomic DNA encoding PHGPX was derived in two steps with (i)1 liter of 10 m phosphate, pH 7, 5 xm mercapto- from overlapping fragments isolated from a twice amplified pig liver ethanol, and 10% glycerol and (ii) 1 liter of 300 xm phosphate, pH 7.6, genomic DNAlibrary obtainedfrom Clontech LaboratoriesInc. via ITC 5 m mercaptoethanol, and 10% glycerol. Elution was carried out with Biotechnology GmbH, Heidelberg, Germany. The gene bank contained 150 mlof 300 m phosphate, containing 25xm Tris base, pH 8.5, 5 m genomic DNA fragments cloned into a BamHI restriction site of the mercaptoethanol, and 10% glycerol. The active fractions were pooled phage EMBL3. Suitable hosts were Escherichia coli K12 NM538 (ATCC and concentrated to approximately 5 ml usingan Amicon ultrafiltration 35638) or E. coli K12 K802 (ATCC 33526). A first492-bp fragment was apparatus equipped with an Y " 1 0 membrane. The concentrated ma- obtained by PCR using the primers P2 (5"ACATCGACGGGCACATGterial was then applied to a Sephadex G-75 column (5 x 100 cm) equili- GTGAACCTGGAC, bases 1698-1726 i n Fig. 2) and PCB4i,, (5'-CCAbrated with 300 m phosphate, pH 7.6,5 m mercaptoethanol, and10% CAGAGGGTGGGCATCGTCCCCATTCAC, complementary t o bases glycerol. The active fractions, eluting 565 ml aRer thevoid volume of the 218s2159 in Fig. 2) derived from the known cDNA sequence (4). The column, were pooled and concentrated with the same ultrafiltration resulting 492-bp fragment corresponded to bases 1698-2188 in Fig. 2. apparatus as before. 0.4 mg of PHGPX correspondingto a n overall yield Part of this fragment was amplified i n a n asymmetric PCR (50-fold of 30.8% were obtained. The specific activity of 80 unitdmg wasslightly excess of P2 over the inverse primerP3 = 5'-CCAGAATCCGTAAACCAbelow that usually achieved by the conventional procedure (2). CACTCAGC, positions 1927-1902 in Fig. 2)inthepresence of [CY-~*PI~CTP. The single-stranded radiolabeled 230-nucleotide fragment Amino Acid Sequencing thusobtainedwasusedforplaquehybridizationexperimentsthat 160 pmol of PHGPX prepared by procedure 2 were centrifuged onto yielded two identicalpositive clones from 5 x lo7 plaques. Thepositive a polyvinylidene difluoride membranein a ProSpin Cartridge (Applied clones contained an insert of about 10,000 bp. Overlapping restriction as as of Biosystems, Foster City, CA), washed twice with 1ml of 20% methanol, fragments of this insert comprising 279,606, and 1700 bp well and then subjected to N-terminal sequencing by means of a gas-phase the originally obtained 492-bp fragment were cloned into pUC19, amplified in E. coli K12 JM103 (ATCC 39403), and completely sequenced sequenator (model 475 A, Applied Biosystems). Amino acid composition et al. (15)by means of the primer walking strategy was analyzedby means of a 420 A aminoacid analyzer. The experiment according to Sanger the T7 Sequenase Kit (Pharmacia, Freiburg, Germany).All fragwas repeated with a PHGPX sample prepared according to procedure using 1. ments were sequenced in both directions. Condensations observed in Cyanogen bromide fragments of PHGPX were analyzed correspondthe 5'-untranslated region and the first intron (1700 bp) could be reingly. solved with the 7-deaza-dGTP technique with exception the of positions Molecular Mass Determination 917-919 (nnn in Fig. 2). Fig. 1schematically shows the genomic DNA fragments sequenced (A), the corresponding cDNA (boxes)and primers PHGPX samples were concentrated 10-fold by means of a Centricon used ( B ) , and the restriction enzymes and restriction sites used for 3 microconcentrator (Amicon, Witten, Germany) and freed from disfragmentation of the - 10,000-bp insert ( C ;position numbersof restricturbing solvent constituents by HPLC chromatography on aVydac C4 tion sites correspond to Fig.2). column (4 x 150 mm; The Separation Group, Hersperia,CA). The column wasdeveloped by a gradientof 0.06% trifluoroacetic acid in water Amino Acid Sequence Alignment (solvent A) and up to95% solvent B(0.052% trifluoroacetic acid in 80% Sequencealignment of PHGPXwithotherrelatedproteinswas acetonitrile) at 40 "C. Peak fractions monitored by absorption at 215 nm were collected and used for mass determination. Matrix-assisted laser achieved by means of the GENMON program package (16). desorption mass spectra were obtained on a REFLEX@-TOF mass specSearch for Putative Regulatory DNA Sequences trometer (Bruker-Franzen Analytic GmbH, Karlsruhe, Germany). Ions were produced usinga pulsed nitrogen lastera t 337 nm with sinapinic The sequence TGACC and its reverse sequence was considered a acid a s matrix. External mass calibration was achieved with cytochrome c. Spectra were recordedin the linear mode. The identityof the putative estrogen-responsive element, in particular if complemented by protein thusprocessed was reinvestigated by amino acid analysis, and a palindromic sequence as in GGTCANNNTGACC (17-19). Glucocorticoid /progesterone-responsive elementsare primarily charsequencing cyanogen bromide fragments were prepared as described (4) and isolatedby the HPLC procedure givenabove but with the gradient acterized by the sequence TGTYCT. Complementation as in GGTWCANNNTGTYCT is suggested as supportive (20). An additional supranging from 2to 93% solvent B. portive criterion is considered the presence of neighboring CACCC boxes (21). cDNA Analysis The search for putative cAMP-responsive elements was based on the Preparation, cloning, and sequencingof cDNA fragments coding for motif TGACG or TGACGTCA (22, 23). partial structures of PHGPX are described in detail in Ref. 4. The CCAATlenhancer-binding protein and NF-1-binding elements are sequence information residing in the unidentified 5'-end of the cDNA considered to be characterized by the motifs TKNNGYAAK (K = T/G)
sequences are screened for putative hormonal control regions and consensus sequences relevant to the incorporation of selenium.
7344 A
Phospholipid-hydroperoxide Glutathione Peroxidase
-
1700 bp
++ bp
276
606
tgtctgcgacctacaccacagctcacgggc ccagggatcaaacccgcatcctcatggata ctagttggattcgcttccgctgcgccacaaccgactccgaaaatcccgttttcaaggagc tggagcgacttttttgagagtccagccgccgcctcccctcacaggtttcctgagtatcta
bp
492 bp
B
5#7-
aatctgtcgggcaccgcgctgcgcagcagtcaaccacccagataactgattttaaccaga
-
P,
fl[1 p3 +
3'
t
PCB4,
caaccacagcctggccc cgtctacagtcccg tggtccgctcggcgagggg gaggggaggggcgc tttagccgcctgttccgcc ccccggagcccccg
gtcggcccctcaccccgacctcccattggc
tgagagggatggagggcgt
540
600
Met GCCCGGCCTGGCCGGCACCATGg cggagggccgttgmgcn nncggcctgcctctcggcgcgcggggcag ttaggcagggggcgcgcggacaa tcggccgccgcctccgcggccgtc ggccagcgcccgggccgcgatcacgatc ggtggggctgggggcaaggctgggcggt ccaggctgggtcggggacgggggg gaggacccccgggtggnnngcccc agagttgctgcctcggggcaacggggtg gcgccgtcggctactggtactacg ggagagcggggggcgccctctacggggt CtgCCCCtCCCCggCCCtgtgtg gccccgcgcccttggctaccggc ctgcatgtccgaagacgggtggg catgggccgcaccagccgccggttccccgggtcgccgtggccagcggcgccggttgccgg ccgggcggcgacgcagggcccctcggaggcggagggctcgtttgtgccgcagaagggcgc gcccccggagaaggcagccgg gagcggctgcagcgccgagtc ggggtagcctaggtctcagtc gcctctgcggaggaggagcct tggcaccctgggttgagagagttgcgggcggccttagccccgag~tctgcaggctgctca tgCtgaagCCagcgCTTCTCTGffiGGACC
C
SalI BglII EcoRI +/ : :
SalI
Erg111
Xhol
1682
1882
2151
2877 2151
1
1607
sequencing strategy of the genomic clone of PHGPX Based on the partial cDNA sequence a 492-bp genomic fragment was obtained by PCR. This fragment was used to identify a genomic cloneof about 10,000 bpthat was cleaved as indicated to yield overlapping fragments. A shows the genomic DNA fragments cloned and sequenced; B shows the corresponding coding cDNA sequences (boxes)and the primers used for PCR (see text); and C shows the restriction sites and enzymes used for fragmentation. The position numbers of restriction sites correspond to sequence positions in Fig. 2; the last number (2877)represents the last base sequenced. FIG. 1. Schematic representationof the
60 120 180 240 300 360 420 480
tgagccatggtgagctaggcgctgccgac
660 720 780 840 900 960 1020 10a0 1140 1200 1260 1320 1380 1440 1500 1560 1620
CysAlaSerArgAspAspTrpArgCysAlaAr&e%~H~
(24) and TGGMNNNNKCCA (M= NC; K = T/G) (25, 261, respectively.
RNA Secondary Structure Prediction Secondary structure prediction of the 3'-untranslated region of the
PHGPX mRNA was carried out by the RNAFOLD program of Zuker and Stiegler (27). The graphic presentation module is part of the GENMON program package (16). RESULTS
Complete Structural Gene and Deduced Amino Acid Sequence-In our previous report (4) theN-terminal part of the PHGPX sequence could not be identified, because it resisted sequencing; no full-length cDNA clone could beisolated despite repeated efforts. Also the present approach, the analysis of rare transcripts by the RACE protocol, primarily yielded cDNA fragments appreciably shorter than calculated from the position of the nested primer and the presumed 5'-end of the structural gene. Up to the second PCR round the bulk of the material obtained by the poly(A+)RNA amplification had a molecular weight ranging from 250 to 350 bp, and sequencing of these PCR products indicated that a major RNA processing site should reside in the 5"terminal part of the PHGPX mRNA, obviously close to the start signal for protein synthesis. Only after amplification of a hardly visible band of about 450 bpin a third round of PCR, a defined cDNA fragment of appropriate size was obtained. Sequencing of this material finally yielded the so far missing information on the 5'-terminalcoding region. The sequenceconfirmed the positions 1673-1734 in Fig. 2 knownfrom our previouswork (4); it was extended at the 5'-end for another 42 nucleotides that coincided with the genomic sequencethat meanwhile had become available (see below). After bridging an intron of about 1000 bp, it again coincided with the genomic sequence for 30 nucleotides comprising part of the 5"untranslated region and the start codon ATG at its 3'-end. The transition from the 5'-UTR to the start codon and the structural gene was then confirmed by resequencing with a primer ( U K 1078) starting from the 5"untranslated region. These results taken together with those of Ref. 4 provide the maximum length of the PHGPX sequence, i.e. 170 amino acid residues (Figs. 2 and 3). Determination of the N-terminal Sequence of PHGPX-When PHGPX isolated in active form according to procedure 2 was reduced and carboxymethylated and then subjected to gasphase sequencing, an unequivocal sequence of 24 residues was obtained. It started with position 3 of the deduced amino acid sequence (Ala) and extended into the sequence previously es-
gtgtgcccttgccttgcagTCGCGTCCCGffiACGACTGGCGATGTGCTCGCTC~TG~1680
sGluPheserAlaLysAspIleAspGlyHisMetValAsnLe~sPLysTyIA1 CGAATTCTCAGCCAAGGACATCGACGGGCACATGGTGAACCTGGACAAGTACCGgtgggt 1740 acctgcccgcttggtctgggaaggcaggggtaa accgtctca la00 4 cgGlyTyrValCysIleValThrAsnValAlaSerGln Thr accgcctccagGGGffACGTGTGCATCGTCACCAATGTGGCCTCTCAA Affi 1860 " "
~~~~~~~~~~~~
~
~
~
~
~~
~~~
"""_
~
" " " "
GluValAsnTyrThrGlnLeuValAspLeuHisAlaArgTyrAlaGlu
Arg
GAGGTAAACTACACTCAGCGTCGACCTGCACGCCCGATATGCTGAG
CGG
1920
IleLeuAlaPheProCysAsnGlnPheGlyArgGln ATTCTGGCCTTCCCTTGCAACCAGTTTGGGAGGcAGgtacgtcttggttgccctggatgt agggcctgtgtgaggaatctgggtccaaagagggtaca
1980 2040
- - -- - - - - - - -
---
GluProGlySerAspAlaGl gggccctcccccctcacctccagcacccctttctgcccagGAGCCAGGGAGTGA~CTGA 2100
uIleLysGluPheAlaAlaGlyTyrRsnValLysPheAsPMetPheSeILysIleCYsVa GATCAAAGAATTTGCTGCTGGCTAC ATTTGATA TmGTGT
2160
~s~G~A~~~l~His~oteuTrpLysTrpMetLysValGlnProLysGlyArgG1 GAATGGGGACGATGCCCACCCTCTGTGGAAGTGGATGAAGTGGATGAAAGTCCAGCCCAAGGGGAGGGG
2220
e " yMetLeuGlyAs CATGCTGGGAAAgtgagttggggggctggggtgagagtggagggcagtggggatctgcag 2280 - --
-- - - - - -
nAlaIleLysTrpAsnPheTh1 ctgccacgggattactgatgacacatttctttttgcagTGCTATC~TGGAACTTTACC
2340
"
ggtgccacttctag gcccagtccaccac
2400 2460
" "
PheLeuIle~LysAsnGlyCys ggtggccttctgtccc~~~c~~ctgtcctgcagTTC~CA~GATAAGAACG 2520 GCTGT
tgggggggtatgggaga@accacm@&tgtctcacctac
ValIieGluLysAsp A G A ~ G G A C 2640
LeuProCysTyrLeu' C'TGCCGTGCTACCTCTAGCTCAACAAGTGTGTGCTGAGC~GCTGCCTGTGCCCTCAGAG2700 CCTTCCACCTGG~TGACAGTCTGCCTAAAAACCAGCCCTGGTGGGGCAGA~~CGA 2760 2820 CACCCCAGGATGCTGCCTTGTGGGAATMCCATACMTTGGCT2877
GAACCTGGCGTGCAACCCCGCCAGAGGAAGGTCCCTTGGCCCAGCTTGCAGmCGCACCC
FIG. 2. DNA and deducedaminoacidsequence of porcine PHGPX. The cDNA sequence, as far as determined, is given in capital letters. Amino acid sequences confirmed by peptidesequencing are marked with a solid (this work) or a dashed (see Ref. 4) line. Sec indicates the position of selenocysteine. Nucleotide sequences suggestive of regulatory relevance according to the criteria listed under "Experimental Procedures" are shaded: estrogen-responsive elements at positions 12-24,1467-1480,2404-2415,2598-2610, and 2625-2635; progesterondglucocorticoid-responsive element at 1774-1790; CACCC boxes in the 1400-1600- and 2300-2400-bp region; cyclic nucleotideresponsive elements at 925-929,1103-1107,2128-2132,2411-2414, and 2478-2480; CCAATIenhancer-binding protein consensus a t 18491856 and 2142-2152; AP-2 consensus at 1992-2001; AP-3consensus at 1909-1916; SPl sites in the first intron and upstream from the start codon; and NF-1 site at 50a520.
Phospholipid-hydroperoxide
Glutathione
Peroxidase
7345
poly(A) tail, was sequencedand is depicted in Fig. 2. It confirms the cDNA sequence of the structuralgene and thus the deduced amino acid sequence (Fig. 2) as well as the sequence of the m O D B 3'-UTR sequenced as cDNA before. The introdexon structure BOVINE OPX 3l:EPFHLS of the PHGPX gene differs substantially from that of the GPX LAP wrun_PoPX :lY YAB PIO~PHGPX :x. gene. Whereas only one intron hasbeen found in theGPX gene (28), the PHGPX coding area is interruptedby six introns. All TTTT " TTTTHwlMA IUHPLFA FEP BOVINE OPX 81: introdexon transitions comply with the GT/AG rule. In the KEQKFYT HrmAW-PoPX : Ivp DAHPLWX PIG-FHGPX X first exon the coding area is represented just by the ATG start BlMRBBTTTT lTTT TTTTDBDBTTTT BOO codon. It is followed by an intron of about 1000 bp in the BOVINE OPX 1 3 l : ~ P T P S D D A T - P ~ I ~ S ~ genomic gene. HUXAN B P X :~SCPPTSE~TSDRF....~PXltVHD PIO-PRGPX XWMWQ............ .....PKCRGULCNA With regard to the recently described hormonal control of testicular PHGPX levels in rats, the genomic DNA sequence BOVINE OPX 1811 was screened for putative regulatory elements. The sequence HuEulS_pDPX x PIG-PHGPX : GGTCA, which is indicative of estrogen responsivityif partially FIG.3. Alignment of porcine PHGPX (lower line) with bovine complemented by a palindromic sequence (TGACC) at a discystolic GPX (upperline) and humanplasma GPX (middle line). tance of three nucleotides(17, 181, is indeed found several Position numberingof the bovine GPX sequence is used throughout. # indicatesselenocysteine. . . ., HHH . . ., and ppp . . . above the times in thePHGPX gene (see Fig. 2). The gene also contains a sequencesindicateaminoacidsinvolved in turns, a-helices, and putative progesterone-responsive element (TGTCCT partially p-sheets, respectively, in bovine GPX (29). Residues conserved in the complemented by GGTCGA) that, however, cannot be distinthree types of GPX are sha&d. guished from a glucocorticoid-responsive one (19-21). Suggestive of regulatory importanceare also multipleCACCC boxes in tablished by sequencing of peptides (4) of the cDNA (see below) the vicinity of the glucocorticoid- and estrogen-responsive eieand of the genomic DNA (see below; Fig. 2). A second sample of ments. Further, typical cyclic nucleotide-responsive elements PHGPX prepared in theconventional way but omittingKSCN (TGACG or CGTCA, respectively), putative CCAATlenhancerduring chromatography proved more heterogeneous during se- binding protein consensus sequences (TKNNGYAAK with K = T/G),AP-3 and AP-2 consensus sequences,SP1 sites (GGCGGG quencing despite slightly higher specific activity but again be tentatively identified. yielded a n N-terminal sequence of 13 residues starting with and reverse), and an NF-1 site can Alignment of PHGPX with Other n p e s of GPX-The alignposition 3 of the deduced sequence, thus confirming the data at least one form of ment of the PHGPX amino acid sequence deduced from the obtained before. We therefore can state that mature PHGPX has an unprotected N terminus starting with cDNA with cytosolic bovine GPX and human plasma GPX is position 3 of the deduced sequence. In both experiments, how- shown in Fig. 3. The overall homology of PHGPX and cytosolic ever, the sequencing yields appeared poor in comparison with and plasma GPX is low. There are, however, some sequence the amount of material applied suggesting thatonly a copuri- clusters with pronounced similarity. The conserved sequences fied minor portion of the protein sample had anunprotected N correspond largely to those parts of the protein that form the active site in bovine GPX (29), i.e. a plated sheet with the terminus (see below). Molecular Mass of PHGPX-The molecular mass of purified adjacent selenocysteine (positions 47-53of the bovine GPX PHGPX was determinedby laser desorption mass spectroscopy sequence), the area around glutamine 87 in the turnbetween to be 19,672 Da (M + H+). This value reasonably complies with the p sheet at positions 74-80 and thehelix in positions 91-96, the 20,000 Da previously estimated by SDS-polyacrylamide gel and finally tryptophan 165 at the beginning of a turn followed electrophoresis (2) but is significantly higher than the mean by a p sheet (positions 168-172). Interestingly, the most conmass calculated for the PHGPX sequence of 168 amino acid served parts of the sequence are provided by only two of the residues (19,257.09 Da). Since the methodological error of the seven exons of the PHGPX gene, i.e. exons 3 and 5. Between the mass determination applied here hardly exceeds 0.1%, the dif- conserved clusters and in particularat the C- and N-terminal ference of at least 400 Da between calculated and measured extensions the relatednessof the proteins is hardly detectable. molecular mass has to be attributed toa still undefined deriva- Not surprisingly, the N-terminalextension of plasma GPX reptization of the peptide chain. A presumed phosphorylation at resenting a signal peptide is missing in the intracellulartypes tyrosine 96 (4) alone cannot account for the difference. of glutathione peroxidases. It is further worthmentioning that In order to further clarify the molecular species of PHGPX a large part of the bovine GPX sequence is deleted in PHGPX exhibiting themolecular mass of 19,671 Da,we again subjected (positions 136-152) and thatalso part of the a-helix of positions the material processed for mass spectrometry to N-terminal 94-102 is missing. sequencing and now found the protein resistant to EdmandegThe J'-Untranslated Region-As was shown for two mamradation. Amino acid analysis of the material, however, was malian selenoproteins, type I 5"deiodinase and GPX (301, the compatible with the amino acid composition of PHGPX. Fur- translation of the stopcodon TGA as selenocysteine is encoded ther, after cyanogen bromide cleavage, a fragment wasisolated, by a sequence of the 3'-UTR, which is assumed toform a charcould be sequenced completely, and proved to be the C terminus acteristic stem loop. Consensus sequences in the 3'-UTR of of PHGPX (positions 157-170 of the deduced sequence). A si- 5'-deiodinase and GPX have been identified. We therefore inmultaneously readable sequence of a contaminating cyanogen vestigated the 3'-UTR of PHGPX for consensus sequences with bromide fragment (about 10% of the major component) repre- the relevant parts of the deiodinase and GPX genes. Fig. 4 sented the positions 130-144 of PHGPX. The data unequivo- shows the consensus in the3'-UTR of two 5'-deiodinase species cally show that the 19,671-Da molecule is PHGPX but a de- and in the next line the pertinentconsensus of the GPX 3'-UTR rivative different from the PHGPX form that yielded the free of four differentspecies. The consensus sequencesderived from N-terminal sequence. Unfortunately the latterspecies could so these two enzyme familiesshown in the third line correspond to far not be recovered after HPLC chromatography. the one worked out by Berry et al. (30). Alignment of the 3'Genomic DNA Sequence-A genomic DNA fragment of 2877 UTR of PHGPX (fourth line) then yields a n overall consensus of bp comprising the complete coding area of PHGPX, almost 600 the 3'-UTR of the different selenoproteins. Evidently, the sec5'-untranslated bp and the 3"untranslated region up to the ond and third consensus sequences of the proposal of Berry et
...
7346
Phospholipid-hydroperoxide 5 ’ DI T..ATATTTG
TTTATGATGG ..CACAG..T
GPX .GG...TT.. .C.ATGA.GC
Peroxidase
AAAGT.CACA C.GCTGTG.A CTTGATT...
TGTT.CCTCT AAA..T.C.. GG....A... CCTGAT.T..
SECYS c ~ n s e n s u so € GPX and 5 ’ DI TT.. ATGA.GG T
......
Glutathione
...
.........
ARR.......
PHGPX CCACCTGGCA CCCATGACAG T-CTG-CC-T AAAAACCAGC
.......... C.TGAT.... --“------
. . . . . . m.......
...AAAAT.C
...AAAAT..
CCTGGTGG-- GGCAGACTCG
overall SECYS consensus
.......... ...-..E
...AAAAT..
C.TG.2..
”
...m..
FIG.4. 3’-UTR consensus of different selenoproteins. 5’-DZ,5”deiodinase (two species),SECYS, selenocysteine.
. c
n
C
c
c c
A ‘CGC CQ CG AU .E
L
c
A Q C Q C A U
.”
”
G C
ALu m
cc G
A Q
UQ
c L
FIG.6. Alignment of the partial sequences of the GPX supelc family corresponding to the deduced amino acid sequence of exon 3 of PHGPX Amino acid residuesconserved in all sequencesare shaded; # stands for selenocysteine. For references and abbreviations, see Fig. 7.
U
u QCC
In contrast toprevious experience(4), the N terminus of the two PHGPX samples here investigated and prepared differA ently could be sequenced by conventional Edman degradation. A 3 C The protein chain started with position 3 of the deduced se5’ cu quence. Since the C terminus of PHGPX was establishedbefore FIG.5 . Predicted secondary structure of the complete 3’-un- by amino acid and cDNA sequencing, the pertinent molecular translated region of the PHGPX RNA (207 nucleotides). The structure prediction has been carried out according to the RNAFOLD species could comprise positions 3-170 of the deduced sequence program (27). The graphic presentation model is part of the GENMON with a corresponding theoretical mean molecular mass of the program package (16).The region marked with heavy bars couldbe protein moiety of 19,257.09. The sequencing yields, however, considered consensus sequences involved in the decoding of TGA as had alreadysuggested that themolecular species exhibiting an selenocysteine (see also Fig. 6). unprotected N terminus might haveonly been a minor component of the samples. After the material hadbeen rechromatoal. (30) are reasonably confirmed, whereas arbitrary stretch- graphed on a C4 HPLC column, it was indeed no longer susings of the PHGPX sequence have to be made to construct a ceptible to Edman degradation, and its molecular mass proved marginal conservation of the fourth and fifth consensus cluster. to be substantially larger than calculated for the maximum Interestingly, however, when the complete 3’-UTR of PHGPX is theoreticalsequence as 3-170 butwas clearlyidentified subject to a calculation of possible secondary structure forma- PHGPX by partial sequencing of its cyanogen bromide fragtion, a stem-look structure of considerable stability (-41.9 kcal ments. It is therefore evident that PHGPX is prone to postat 37 “C) is obtained, and theoverall consensus sequences pre- translational modification inter alia affecting the N terminus. sumed to be relevant to thecorrect TGA translation arefound The nature of such modifications awaits clarification and deat or near the terminal loops of the stem-loop (Fig. 5). Like in serves interest, since derivitization could well be responsible the genes of the other mammalian selenoproteins (30) a poly(A) for the translocation of PHGPX during cell maturation (10). sequence is located in the loop, and also the UGXU sequence The sequence data presented demonstrate that PHGPX is a remainsunpairedinthe predicted stem-loop structure of gene product distinct from “classical” cytosolic GPX and extraPHGPX. cellular GPX. They alsounequivocally show that all three types of mammalian glutathioneperoxidases are phylogenetically reDISCUSSION lated. However, the overall similarity between pig PHGPX and With the data presented the complete amino acid sequence of other mammalian GPX species is only close to 30%, whereas PHGPX is established.73% of the sequence including boththe between the GPX sequences of different mammals it is nearly N terminus and the C terminus was determined by peptide 90%. Based on thelow percentage of conserved amino acids,the sequencing. The arrangement of the peptide fragments se- phylogenetic divergence of PHGPX and GPX has been disquenced is evident from the cDNA now covering the complete cussed to have happened approximately one billion years ago coding area. The cDNA in turn was confirmed by sequencing (4). Since highly conserved sequences are, however, restricted to only two out of seven exons of the PHGPX gene, it could be the corresponding genomic DNA. Q
G U C G ~ L C C C C A G U
u
A
Phospholipid-hydroperoxide
Glutathione
7347
Peroxidase
ex-MOUSE-GPX ex-RAT-GPX
-
ex-BOVINE-GPX ex-HUMAN-GPX ex-RABBIT-GPX ex-HUMAN-GIGPX ex-SMANS-GPX ex-NTAB-GPXH ex-PIG-PHGPX ex-HUMAN-PGPX ex-RAT-PGPX ex-MOUSE-EGPXH
1-
I
ex-RAT-EGPXH ex-MFASC-EGPXH ex-RAT-GPXH ex-ECOLI-GPXH
r l
I
ex-NEMATO-GPXH
FIG.7. Dendrograms of the GPX superfamily calculatedfrom conserved amino acid residue of aligned protein sequences according to Ref. 16.Only the sequence parts correspondingto exon 3 of PHGPX are used forcalculation. The pertinent sequencesare mouse GPX (taken from Ref. 28), rat GPX (311, rabbit GPX (321, bovine GPX(33,34), human GPX (351,human gastroinestinal GPX (36), human plasma GPX (6),rat plasma GPX (37),mouse epididymal cysteine homolog of GPX (381,rat and monkey epididymal cysteine homolog (A.C. Perry, R. Jones, L. S. Niang, R. M. Jackson, and L. Hall, unpublished data), ratGPX homolog (40), nematode GPX homolog (41), GPX of S. mansoni (42), GPX homolog of N. tabaccum (43), and cobalamin binding protein of E. coli (44).The sequence of PHGPX corresponds to that in Figs. 2 and 3. Abbreviations usedare: GZGPX, gastrointestinal GPX, PGPX, plasma GPX, GPXH, cysteine-containing homologs of GPX EGPXH, epididymal GPX homolog; MFASC, Macaca fascicularis: SMANS. S . mansoni: NTAB.N. tabaccum: ECOLI,Escherichia coli; NEMATO,nematode; ex, the sequence encodedby exon 3 of PHGPX and homologous'sequences, respectively.
envisaged that thelow overall homology originated from shuffling some unrelated exons into the PHGPX gene at a later stage of evolution. Acandidate of this typecould be exon 4 of the PHGPX gene. The location of the single intron in the GPX gene (44)corresponding exactly to that of intron 3 of the PHGPX this view. It therefore appearsmore reliable gene also supports to derive the relatednessof the members of the GPX superfamily from a single exon that is clearly homologous (Fig. 6). A dendrogram based on the comparison of the amino acids sequences corresponding to exon 3 of PHGPX is shown in Fig. 7. PHGPX still branches off very early from the (classical) cytosolic and plasmaGPX species, which form a separate clade with their cysteine-containing homologs. In fact, PHGPX appears more closely related to GPX of Schistosoma mansoni and the cysteine-containing GPX homolog of Nicotinianatabaccum than to all vertebrate GPX types. Irrespective of the approach, therefore, PHGPX has to be rated as a phylogenetically old achievement of the GPX superfamily. Evidently, the dendrogram of related proteins (Fig. 7) does not only comprise glutathione peroxidases. A peroxidasenature may certainly be questioned for poorly related members like the cobalamin-binding protein of E. coli but also for the epididymal homologs of plasma GPX that containcysteine instead of selenocysteine in a homologous positions, since the substitution of sulfur for selenium will definitely lower the reaction rates of the proteins withperoxides by orders of magnitude. According to the x-ray structure of bovine GPX (29),its peroxidase nature dependson the characteristic triad composed of selenocysteine, tryptophan 165, andglutamineresidue 87. Theseresiduesare conserved in homologous positions in
PHGPX and plasma GPX. Also, according to preliminary molecular models of PHGPX and plasmaGPX (45),the triad there is likely to form a n almost identical catalytic center. All GPX types thusfklfill the basic structural requirementsfor the catalytic peroxide reduction. Nevertheless, it may be questioned whether PHGPX is not misnamed as a glutathione peroxidase, since all residues invoked to bind GSH in GPX (7) are mutated or deleted in PHGPX. Certainly, the PHGPX molecule in its reduced state reacts with hydroperoxides at rates comparable with those of GPX ( 2 ) . Certainly, the oxidized form of the enzyme can be stepwise reduced again withGSH to complete the catalytic cycle as in GPX (7). The limitingreaction rate for the reductive part of the cycle, however, is about 1order of magnitude slower for PHGPX (2) than for GPX (7). Also, comparative studies on the selective inhibitionof the reduced enzymes with iodoacetate (2), which carboxymethylates the reduced selenocysteine only (7), revealed that PHGPX is reduced less specifically by various thiols, whereas thereduction of GPX proved to be fairly restricted to GSH (46).This indicates that a specific enzyme-donor substrate complex, in contrast to GPX, is not formed with PHGPX because of the exchanged or deleted arginine residues near the catalytic site. In fact, all attempts to model the substrate GSH onto the active site of PHGPX so far failed to yield a convincing proposal for an enzyme-substrate complex. Nevertheless, preliminary substrate specificity studies using the polarographic thiol determination (11, 46) did so far not offer any alternative substrate;a variety of physiologically occurring thiols like cysteine, cysteamine, ergothioneine, and dihydrolipoamide were not utilized faster than GSH by
7348
Phospholipid-hydroperoxide Peroxidase Glutathione
2. Ursini, F., Maiorino, M., and Gregolin, C. (1985)Biochim. Biophys. Acta 839, PHGPX.' Similarly, Maiorino (47) found considerable activity 62-70 of PHGPX with dihydrolipoamide and dihydrolipoate, but GSH 3. Thomas, J. P., Mariorino, M., Ursini, F., and Girotti, A. W. (1990)J. Biol. Chem. again was the best of the substrates investigated. Thus, left 266,454461 R e d e n s , J., 4. Schuckelt, R.,Brigelius-FlohB,R.,Maiorino,M.,Roveri,A,, with no better alternative, we have to consider PHGPX as Strapburger, W., Ursini, F., Wolf, B., and FlohB, L.(1991)Free Radical Res. glutathione peroxidase, and the intriguingquestion remains to Commun. 14,343-361 be answered how the different types of mammalian selenoper5. FlohB, L., Giinzler, W. A., and Schock, H. H. (1973)FEES Lett. 32, 13s134 6. Takahashi, K., Akasaka, M., Yamamoto,Y., Kobayashi, C., Mizoguchi, J., and oxidases have functionally diversified. Koyama, J.(1990)J. Biochem. (Tbkyo) 108, 145-158 Among the mammalianmembers of the GPX family PHGPX 7. FlohB, L. (1989)in Glutathione: Chemical, Biochemical and Medical Aspects (Dolphin, D., Poulson, R., and Avramovic, O.,eds) . pp. is unique inbeing monomeric and membrane-associated (2,lO). . 643-731,John Wiley and Sons, Inc., New York While the latter property is not easily deduced from presented 8. Weitzel, F., Ursini, F., and Wendel, A. (1990)Biochim. Biophys. Acta 1036, structural data, its monomeric nature can be explained by its 8a94 Zhang, L., Maiorino, M., Roveri, A., and Ursini, F. (1989)Biochim. Biophys. 9. sequence. The large gaps in PHGPX the sequence when aligned Acta 1006,140-143 to the sequences of the other GPX types very precisely corre- 10. Roveri, A., Casasco, A,, Maiorino, M., Dalan, P., Calligam, A,, and Ursini, F. (1992)J . Biol. Chem. 267,61424146 of bovine GPX identified spond to the subunit interaction sites L.(1989)in Handbmk ofFree Radicals and Antioxidants in Biomedicine by x-ray crystallography (29),and subunit interaction energies11. FlohB, (Miguel, J., Quintanilha, A. T., and Weber, H., eds) Vol. 111, pp. 281-286, calculated for bovine GPX are considerably higher than those CRC Press, Inc., Boca Raton, FL for a molecular model of PHGPX (45). It appears conceivable 12. Roveri, A, Maiorino, M., and Ursini, F. (1994)Methods Enzymol.233,20!&212 13. Frohman, M. A,, Dush, M. K.,and Martin, G . R.(1988)P m . Natl. Acad. Sci. that themonomeric nature of PHGPX is the reason for both its u. s. A. ES, 899a9002 ability to interact withbulky hydroperoxide substrates and its 14. Fmhman, M. A. (1990)Amplifications 6, 11-15 15. Sanger, F., Nicklen, S., and Coulson, A. R. (1977)Proc. Natl. Acad. Sci. U.S. affinity to biomembranes. A . 74, 54636467 The sequence data may also provide a tool to solve another 16. Lincoln, D., Bliicker, H., Leuner, U., Lehnherg, W., and Graf, G . (1989)GENMON Version 4.1 Manual, GesellschaR fiir Biotechnologische Forschung puzzling problem. As has been amply demonstrated by Behne mbH, Braunschweig, Germany and co-workers (48), alimentary selenium is not utilized at 17. Beato, M. (1987)Biochim. Biophys. Acta 91495-102 random for selenoprotein biosynthesis but ina tissue- andpro- 18. Glass, C. K., Holloway, J. M., Devary, 0. V., and Rosenfeld, M. G. (1988)Cell 54,313-323 tein-specific manner. According to Wendel and co-workers ( 8 ) 19. Klein-Hitpass, L.,RyfFel, G. U.,Heitlinger, E., and Cato, A. C. B. (1988)Nucleic PHGPX benefits from a low selenium supply much more effecAcids Res. 16,647463 Beato, M. (1989) Cell 56,335-344 20. tively than GPX. In this respect it resembles the 19,700-Da 21. Schiile, R., Muller, M.. Otsuka-Murakami, H., and Renkawitz, R. (1988)Naprotein of rat testis (48) that most likely represents PHGPX. ture 332,87-90 One of the possible explanations for a selective incorporation of 22. Montminy, M. R., and Bilezikian, L. M. (1987)Nature 328,175178 M. R., Sevarino, K. A., Wagner,J. A,, Mandel, G., and Goodman, R. the trace element selenium intoa specific protein would be a 23. Montminy, H. (1986)Proc. Natl. Acad. Sci. U.S. A . 83,6682-6686 preferential affinity of a hypothetical translation factor (anal- 24. Ryden, T. A,, and Beemon, K. (1989)Mol. Cell. Biol. 9, 1155-1164 ogous to SelB of bacteria (39))for the pertinent mRNA. Since 25. Borgmeyer, U., Nowock, J., and Sippel, A.E. (1984)Nucleic Acids Res. 12, 4295-4311 the recognition signals for this translationfactor are presumed 26. Leegwater, P. A. J.,van der Vliet, P. C., Rupp, R. A. W., Nowock, J., and Sippel, A. E. (1986)EMBO J. 5,381386 to reside in the3'-UTR of the mammalianselenoprotein genes, 27. Zuker, M., and Stiegler, P. (1981)Nucleic Acids Res. 9, 133-148 the hypothesis could now be checked by exchanging the 3'-UTR 28. Chambers, I., Frampton, J.,Goldfarb,P.,AtTara, N., McBain,W., and Harrison, between such genes. P. R. (1986)EMBO J. 5, 1221-1227 29. Epp, O.,Ladenstein, R., and Wendel, A. (1983)Eur. J. Biochem. 133,5149 PHGPX is indeed the most puzzling member of the GPX 30. Berry, M. J.,Banu, L., Chen, Y., Mandel, S . J.,Kieffer, J. D., Harney, J. W., and family with regard to its structural pecularities, its membrane Larsen, P. R. (1991)Nature 383, 273-276 association, its occurrence in trace amounts only, its unusual 31. Ho, Y., Howard, A. J., and Crapo, J. D. (1988)Nucleic Acids Res. 16, 5207 M., Mizoguchi, J., Yishimura, S., and Watanabe, K. (1989)Nucleic tissue distribution, and its hormone dependence (4,8, 10).Age- 32. Akasaka, Acids Res. 17, 2136 and sex-dependent variations of levels, some increase upon 33. Mullenbach, G. T.,Tabrizi, A,, Irvine, B.D., Bell, G. I., Tainer, J. A., and Hallewell, R.A. (1988)Protein Eng. 2, 239-246 castration, and decrease upon testosterone application have 34. Giinzler, W. A,, Steffens, G. J., Grossmann, A,,Kim, S.". A,, Otting, F.. alsobeen described for cytosolic GPX in rodents. Also, the Wendel,A,, and Flohe, L. (1984)Hoppe-Seylerb Z. Physiol. Chem. 385, 195-212 genomic gene structure of GPX suggested susceptibility to horK., Morino, T., Takagi, K., and Sukenaga, Y. (1987)Nucleic Acids Res. monal control at the transcriptionallevel (for review see Ref. 7). 35. Ishida, 15, 10051 By no means, however, was anyobserved influence of hormones 36. Chu, F.-F., Domshow, J. H., and Esworthy, R. S . (1993)J . Biol. Chem. 268, 2571-2576 on GPX as pronounced as reported for PHGPX by Roveri et al. 37. Yoshimura, S., Watanabe, K., Suemizu, H., Onozawa, T., Mizoguchi,J., Tsuda, (10); hypophysectomy virtually abolished the PHGPX in rat K., Hatta, H., and Moriuchi, T.(1991).Biochem. J. 109,918-923 (1990)Nucleic Aciag Res. 18,7144 testes, and gonadotropin treatment restored the levels. That a 38. Ghyselinck, N. B., and Dufaure, 39. Heider, J., Baron, C., and Bock, A. (1992)EMBO J. 11, 3759-3766 human (female) gonadotropinenables the synthesis of PHGPX 40. Dear, T. N., Campbell, K , and Rabbits, T. H. (1991)Biochemistry 30,10376in testesof (male) hypophysectomized rats (10)may be rated as 10382 an odd finding but intrigues in suggesting PHGPX that expres- 41. Cookson, E., Blaxter, M. L., and Selkirk, M. E. (1992)Proc. Natl. Acad. Sei. U. S. A . 89, 58375841 sion strictly depends onsome gonadal steroid hormone.In this 42. Williams, D. L., Pierce, R. J., Cookson, E., and Capron, A. (1991)Mol. Biochem. Parasitol. 52, 127-130 context, the abundance of consensus sequences suggestive of M. C., Jamet, E., Parmentier, Y., Marbach, J., Durr, A., and Fleck, J. estrogen and progesterone responsivity in the PHGPX gene 43. Criaui, (i992)P h t MOL Bioi. 18,62%627 (Fig. 2) offers an opportunity to further analyze the hormonal 44. Friedrich, M. J., DeVeaw, L. C.,and Kadner, R. J. (1986)J. Bacteriol. 167, 928-934 control of this enzyme. Evidently, such studies would also be 45. FlohB, L.,Aumann, K.-D., Brigelius-FlohB, R.,Schomburg,D., Strapburger, W., helpful to get a better perception of the physiological role of and Ursini, F. (1993)in Active Oxygens, Lipid Peroxides, and Antioxidants (Yagi, K.,ed) pp. 29-11, CRC Press, Inc., Boca Raton, FL PHGPX.
REFERENCES 1. Ursini, F., Maiorino, M., Valente, M., Fem, L., and Gregolin. C. (1982).Biochim. Biophys. Acta 710, 197-211
R. Brigelius-FlohB,M. Maiorino, and L. FlohB, unpublished data.
46. FlohB,L.,Giinzler, W. A,, Jung, G., Schaich, E., and Schneider, F. (1971) Hoppe-Seyler's Z. Physiol. Chem. 352,159-169 47. Maiorino, M. (1994)in Biological Oxidants and Antioxidants: New Deuelopments in Research and Health Effects (Packer, L., and Cadenas, E., e d ~ ) Hypocrates Verlag, Stuttgart, in press 48. Behne, D., Hilmert, H., Scheid, S . , Gessner, H., and Elger, W. (1988)Biochim. Biophys. Acta 966, 12-21
