The Mouse Gene Encoding the Testis-Specific Isoform of Poly(A) BindingProtein (Pabp2) Is an Expressed Retroposon: Intimations That GeneExpression in Spermatogenic Cells Facilitates the Creation of New Genes

Kenneth C. Kleene, Evan Mulligan, Daniel Steiger, Kevin Donohue, Mary-Ann Mastrangelo

Department of Biology, University of Massachusetts Boston, 100 Morrissey Boulevard, Boston, MA 02125-3393, USA

Received: 16 September 1997 / Accepted: 3 March 1998

Abstract. The gene encoding the testis-specific iso-form of mouse poly(A) binding protein (Pabp2) has beenisolated and sequenced. Unexpectedly, comparison ofthe sequence of genomic and cDNAs demonstrated thatthe Pabp2 gene lacks introns, whereas all other func-tional Pabpgenes in plants, amphibians, and mammalscontain introns. Thus, the mousePabp2gene is a retro-poson, created by synthesizing a reverse transcriptasecopy of a processed mRNA and inserting the copy intothe genome. ThePabp2retroposon is unusual because itis functional: previous work demonstrates that its pro-moter drives the accumulation ofPabp2mRNA in mei-otic and early haploid spermatogenic cells, and thePabp2mRNA encodes a protein whose size and RNA-binding specificities are characteristic of PABP in plants,yeast, and mammals (Kleene et al. 1994). Two novelfactors can be implicated in the retention of function ofthe Pabp2retroposon. First, the promoter of thePabp2gene is not derived from its intron-containing progenitor,Pabp1. Second, mRNAs encoding somatic PABP iso-form, PABP1, are present at high levels in meiotic andhaploid spermatogenic cells. Both features contrast withthe phosphoglycerate kinase 2 retroposon, which is be-lieved to compensate for the depletion of the somaticisoform due to X-chromosome inactivation in meioticspermatogenic cells. We also document that more func-tional retroposons are expressed in meiotic and haploidspermatogenic cells than in any other tissue and specu-

late that transcriptional derepression in spermatogeniccells favors the creation of expressed retroposons.

Key words: Intronless gene — Retroposon — Tran-scriptional promiscuity — X-chromosome inactivation— Poly(A) binding protein — Spermatogenesis

Introduction

The poly(A) binding protein (PABP) binds to thepoly(A) tail of eukaryotic mRNAs and is thought haveimportant functions in regulating mRNA stability andtranslation (reviewed by Sachs and Wahle 1993). PABPsare encoded by a single-copy genes in yeast andDro-sophila (Sachs et al. 1986; Lefrere et al. 1990) and bymultigene families including several functional genes inmammals andArabidopsis (Belostotsky and Meagher1993; Hilson et al. 1993; Grange et al. 1987; Kleene et al.1994; Wang et al. 1992; Yang et al. 1995). We havepreviously demonstrated that mRNAs encoding twoPABP isoforms are found in spermatogenic cells in mice:Pabp1mRNA, which is expressed in somatic cells and athigh levels in meiotic and early haploid spermatogeniccells; andPabptmRNA, which is expressed only in mei-otic and early haploid spermatogenic cells (Kleene et al.1994; Gu et al. 1995). Subsequently, cDNAs encoding athird isoform, iPabp, which shows a pattern of expres-sion similar to that of thePabp1mRNA, has been iden-tified in human (Yang et al. 1995). PABPt has beenrenamed PABP2 by the Mouse Genome Database, andCorrespondence to:K.C. Kleene;e-mail: kleene@umbsky.cc.umb.edu

J Mol Evol (1998) 47:275–281

© Springer-Verlag New York Inc. 1998

this term will be used below. The frequencies of ex-pressed sequence tags corresponding to the variousPABP isoforms in the databases indicate that PABP1 isthe predominant isoform in somatic and testicular cells inmice and human (Mastrangelo and Kleene, unpublished).The amino acid sequences of the mammalian PABP iso-

forms differ as follows: mouse PABP2 and mouse andhuman PABP1, ca. 18.5%; mouse PABP2 and humaniPABP, 31.5%; and humaniPABP and human andmouse PABP1, ca. 21%.

In the present study, we have determined the structureof the mousePabp2gene. Unexpectedly, thePabp2gene

Fig. 1. Nucleotide sequence of the mouse PABPt gene. The mousePABPt gene was cloned and sequenced as described under Materialsand Methods. The bases are numbered using the A of the translationinitiation codon as +1. Bases upstream are designated by negative

numbers. The sequences of the PABPt gene and cDNA that lack intronsinclude bases −237 to 2052. The polyadenylation signals areunder-lined and the poly(A) addition sites are indicated inbold face type.

276

lacks introns: therefore, it appears to be a rare example ofan expressed retroposon. Retroposons are a type of genethat is generated by making a reverse transcriptase copyof an mRNA and inserting the DNA copy into genomicDNA in a germline cell (reviewed by Vanin 1985;Weiner et al. 1986; Brosius and Tiedge 1996). Retro-posons are distinguished by the absence of introns, a 38A-rich terminus representing the remnant of a poly(A)tail, short flanking direct repeats resulting from insertionof the retroposon into a staggered break in genomicDNA, and a chromosomal locus differing from the in-tron-containing progenitor gene. Although retropositionhas occurred frequently in mammalian evolution (Dein-inger et al. 1993), the vast majority of retroposons de-rived from mRNAs are nonfunctional processed pseudo-genes (Vanin 1985; Weiner et al. 1986). We documenthere that functional retroposons tend to be expressed intestis and suggest that this tendency may be related to theunusual patterns of gene expression in spermatogeniccells.

Materials and Methods

The Pabp2gene was identified by screening about 5 × 105 plaques ina mouse genomicl library (Clontech ML1009D) with the 58 proximalEcoRI–PstI fragments of mousePabp2andPabp1cDNAs (Kleene etal. 1994). Fourty-four positive plaques were identified in the firstscreen of the genomic library. Ten plaques were rescreened by plaquehybridization, one of which hybridized to the 58 proximal EcoRI–PstIfragment of mousePabp2cDNA but not the corresponding fragment ofPabp1cDNA. DNA sequence analysis showed that this recombinantphage contained the 58 end of thePabp2gene, but it lacked the 38 end.To identify the 38 UTR of Pabp2cDNAs, a CD-1 mouse testis cDNAlibrary (Stratagene 937308) was screened with aPabp2-specific oligo-nucleotide labeled with T4 polynucleotide kinase andg-ATP, CTGAC-CAGAGACCTCAAAAAAGAAACTCACACTTTGCA. The se-

quence of the cDNA clones with the longest 38 ends were used todesign specific PCR primers for the 38 end of thePabp2 mRNA,G C A A A T G C T T G G G T G A A A G G C T A ( u p s t r e a m ) a n dTCAAAGTTTCCTTTCCTTTT (downstream), which were then usedto identify l bacteriophage containing the 38 end of thePabp2gene.The PCR reactions (50ml) contained 50 mM KCl, 10 mM Tris–HCl(pH 9.0), 0.1% Triton X-100, 2.0 mM MgCl2, 1–10 ng of DNA tem-plate, 20 to 30 pmol of each primer, 0.25 mM dNTPs, and 1.0 U ofTaqDNA polymerase (Promega-Biotec). PCR reactions were overlaid withmineral oil and amplified using an MJ-Research Thermocycler: 5 minof denaturation at 95°C, followed by 25 cycles consisting of 1 min ofannealing at 62°C, 1 min of extension at 72°C, and 1 min of denatur-ation at 92°C.

The 38 end of thePabp2gene was isolated by inverse PCR ampli-fication (Triglia et al. 1986). The recombinant bacteriophage DNAcontaining the 38 nontranslated region of the PABPt mRNA was puri-fied, digested withSau3A, ∼100 ng self-ligated in a 50 reaction, andPCR-amplified with the primers, AGCTGTTGGCAGTACCTCTG andACACCGAGACACCTGGAGATTC, using conditions describedabove. PCR andPstI fragments of genomic DNA were subcloned,respectively, into pGEM-T and pGEM-3 (Promega-Biotec) and se-quenced using Sequenase II (US Biochemicals) and35S-dATP and anApplied Biosystems automatic sequencer. NewPabp2sequences weredetermined on both strands. The sequence of thePabp2gene has beendeposited in GenBank under Accession Number AF001290.

Results and Discussion

ThePabp2Gene Is a Retroposon

The mousePabp2 gene was isolated from a genomiclibrary by a strategy combining conventional screeningand inverse PCR. Figure 1 shows the sequence of 3424,bases which includes at least 800 bases of 58 flankingsequence, based on estimates that thePabp2mRNA is2400–2700 bases long including a poly(A) tail of 50–150bases (Kleene et al. 1994). Two poly(A) addition siteswere identified from the sequence ofPabp2cDNAs with

Fig. 1. Continued.

277

the poly(A) tracts too long (>30 dA’s) to have beengenerated in cloning. ThePabp2polyadenylation signalcorresponds to an upstream polyadenylation signal in thehumanPabpl 38 UTR (Grange et al. 1987).

Comparison of the sequences ofPabp2genomic DNAand cDNA demonstrates that thePabp2gene lacks in-trons. The possibility thatPabpgenes generally lack in-trons is eliminated by the observation that the codingregions of the humanPabplgene and thePabpgenes inXenopus laevisand Arabidopsis thalianacontain 6–14introns (Belostotsky and Meagher 1993; Hilson et al.1993; Nietfeld et al. 1990; Hornstein and Meyuhas, un-published; Accession Numbers U68093–U68105). Theabsence of introns indicates that thePabp2 gene is aretroposon, a class of genes created by reverse transcrip-tion of a mature mRNA and insertion of the DNA copyinto the genome (Vanin 1985; Weiner et al. 1986).

It is unclear whether thePabp2 gene contains thedA-rich 38 terminus and short flanking direct repeatsfound in most, but not all, retroposons (Boer et al. 1987;Brosius and Tiedge 1995). The high proportion of dA’s(60.9%) in the 123 bases proximal to the 38 terminus ofthePabp2gene is probably not the relic of a poly(A) tail,because this sequence is similar to an A-rich region in themiddle of the 38 UTR of humanPabp1mRNA (Grangeet al. 1987). A plausible explanation for the origin of the38 end of thePabp2gene is that the reverse transcriptaseinitiated within the A-rich region of thePabp138UTR.We cannot identify the flanking direct repeats due touncertainties about the exact 38 terminus of the insertedDNA.

ThePabp2Retroposon Is Expressed and Its Functionhas been Selected for

ThePabp2retroposon is unusual because it is functional.Most retroposons evolve into nonfunctional processedpseudogenes, because mRNAs lack the promoter in the58 flanking sequences of genomic DNA (Vanin 1985;Weiner et al. 1986; McCarrey 1994). Consequently, theretroposon is not expressed, its function cannot be main-tained by natural selection, and the coding region is rap-idly inactivated by deleterious mutations. The promoterof thePabp2gene is active, becausePabp2mRNAs aredetectable by Northern blots in meiotic and early haploidspermatogenic cells (Kleene et al. 1994). The PABP2protein also appears to be functional, because its size andribohomopolymer binding specificities resemble those ofother PABPs in mice, humans, yeast, andArabidopsis(Belostotsky et al. 1993; Burd et al. 1991; Kleene et al.1994).

The function of thePabp2retroposon has been pre-served by natural selection. As noted in the Introduction,the amino acid sequence of PABP2 is more similar tothat of PABP1 than to that ofiPABP: thus, it is likely that

the Pabp2gene is derived from thePabp1mRNA. Theobservation that the nucleotide sequence of mousePabp2cDNA differs from those of mouse and humanPabp1cDNAs by similar extents, 19.3 and 19.0%, re-spectively, implies that thePabp2 retroposon was cre-ated close to the time of divergence of primates androdents∼80 million years ago (Li and Grauer 1992).Thus, thePabp2 gene is much older than the typicalprocessed pseudogene, most of which were created andinactivated during the past 10 million years (Vanin1985). In addition, the mousePabp2coding region con-tains five deletions that are absent from the mouse andhumanPabp1cDNAs (Grange et al. 1987; Wang et al.1992; Kleene et al. 1994). Since all five deletions involveeither three or nine nucleotides, they have not altered thereading frame. Assuming equal probabilities of creatingdeletions that shift the reading frame by zero, one, or twobases, the probability (P) of randomly acquiring five de-letions, none of which cause a reading frame shift, can becalculated:P 4 0.335 4 0.0039. This low probabilitysuggests that frame shift mutations have been eliminatedby natural selection.

Retroposons Are Preferentially Expressed inSpermatogenic Cells

The expression of thePabp2retroposon in meiotic andhaploid spermatogenic cells (Kleene et al. 1994) exem-plifies a common phenomenon. Table 1 lists 14 intron-less genes in mammals that are closely related to intron-containing genes whose expression can be detected byNorthern or Western blots in vivo. This list does notinclude several categories of intronless genes that maynot be functional retroposons: (1) members of gene fami-lies that are typically intronless (eg., histones and G-proteins); (2) retroposons whose mRNAs have only beendetected by reverse transcriptase-PCR, because the levelsof transcripts may not be biologically significant; and (3)retroposons whose promoters have only been demon-strated to be functional in vitro (Ludwig et al. 1992).Brosius and Tiedge (1995) independently compiled asimilar list of functional retroposons.

Six of the 14 retroposons in Table I are known to beexpressed in meiotic and/or haploid spermatogenic cells,including the phosphoglycerate kinase 2, pyruvate dehy-drogenase E1a subunit 2, C gamma subunit of cAMP-dependent protein kinase glucose-6-phosphate dehydro-genase,Zfa and Pabp2 retroposons (Boer et al. 1987;McCarrey and Thomas 1987; Fitzgerald et al. 1992,1994; Kleene et al. 1994; Hendrickson et al. 1997; Ash-worth et al. 1990; Reinton et al. 1998). The number ofretroposons that is expressed in meiotic and haploid sper-matogenic cells is greater, six, than in any other tissue,two, in liver and brain. This is also likely an underesti-mate, because the testicular cell types that express theglycerol kinase 2,S-adenosylmethionine decarboxylase,

278

SCIP, and glutamate dehydrogenase 2 retroposons havenot been identified.

Do the Patterns of Gene Expression in SpermatogenicCells Favor the Creation or Expressionof Retroposons?

Previous workers have pointed out that functional retro-posons require an active promoter to express the geneand a selective advantage to preserve the function of theretroposon. The possibility that the tendency of retro-posons to be expressed in meiotic and haploid spermato-genic cells may be related to unusual patterns of geneexpression is discussed below.

The vast majority of retroposons evolve into nonfunc-tional processed pseudogenes because the inserted DNAlacks the promoter in the 58 flanking sequences of geno-mic DNA. Soares et al. (1985) and McCarrey (1994)have pointed out that functional retroposons can acquirepromoters in three ways, and we propose a fourth: (1)reverse transcription of an aberrant mRNA that contains58 flanking promoter elements normally lacking inmRNAs, (2) insertion of the retroposon downstreamfrom a preexisting promoter, (3) mutations create pro-moter upstream of the site of insertion of the retroposon,and (4) the 58 UTR of the progenitor mRNA containscryptic promoter elements. The first mechanism is nor-mally considered to be the most likely and is supportedby the observation that 58 flanking sequences of someexpressed retroposons exhibit similarity to the 58 flank-

ing sequences of their progenitor genes (Soares et al.1985; McCarrey 1990).

Significantly, the 58 flanking sequence of the humanPabp1gene [bases 1–378 of Accession Number U68093(Hornstein and Meyuhas, unpublished)] exhibits no simi-larity to any sequence near the 58 end of the mousePabp2gene. In contrast, the 58 UTR of humanPabp1mRNA [bases 379–883 (Grange et al. 1988)] is identicalat ∼59% of sites of the mousePabp2 gene. Thus, thepromoter of the mousePabp2gene is derived either froma prexisting promoter near the site of insertion, by mu-tation, or a cryptic promoter in the 58 UTR.

We wish to suggest that the probability of creatingactive promoters in spermatogenic cells by any of thesemechanisms may be far higher than previously thought.The levels of RNA polymerase II and basal transcriptionfactors are extraordinarily high in haploid spermatogeniccells (Schmidt and Schibler 1995), creating a conditiontermed ‘‘transcriptional promiscuity’’ (Schmidt, 1996),characterized by the recruitment of cryptic promoters,the utilization of multiple start sites, and the accumula-tion of extraordinarily high levels of many mRNAs(Davies and Willison 1993; Hecht 1995) of which thePabp1 mRNA is an excellent example (Kleene et al.1994). Transcriptional derepression may be a widespreadphenomenon in spermatogenic cells because P-elementinsertion indicates that about 50% of all genes are tran-scribed in Drosophila spermatocytes (Bownes 1990).Transcriptional promiscuity could dramatically increasethe likelihood that retroposons are expressed in sper-

Table 1. Expressed retroposonsa

Name of gene(species of mammal)

Pattern ofexpression Reference

Glucose-6-phosphate dehydrogenase (M) Primary spermatocytes, round spermatids Hendrickson et al. (1998)Calmodulin-like mRNANB-1 (Hu) Epithelial cell types Yaswen et al. (1992)C gamma subunit of cAMP-dependent

protein kinase (Hu)Primary spermatocytes, round spermatids Reinton et al. (1998)

Glutamate dehydrogenase 2 (Hu) Retina, testis Shashidharan et al. (1994)Glycerol kinase (Hu) Testis Sargent et al. (1994)N-myc2(WC) Liver tumors, brain Fourel et al. (1990)Phosphoglycerate kinase 2 (M, Hu) Primary spermatocytes,

round spermatidsMcCarrey and Thomas (1987);

Boer et al. (1987)Preproinsulin I (R) Pancreatic islet Soares et al. (1985)Pyruvate dehydrogenase E1a2 (M, Hu) Primary spermatocytes,

round spermatidsDahl et al. (1990); Takakubo and Dahl (1992);

Fitzgerald et al. (1994)Poly(A) binding protein 2 (M) Primary spermatocytes,

round spermatidsKleene et al. (1994)

r-Pem2homeobox (R) Epididymis Nhim et al. (1997)S-Adenosylmethionine decarboxylase (M) Liver, spleen, kidney,

testisPersson et al. (1995)

Tst-1/Oct6/SCIP transcription factor (M) Schwann cells, manyneural cell types, testis

He et al. (1989); Kuhn et al. (1991)

Zfa (M) Primary spermatocytes Ashworth et al. (1990)

a The list of expressed retroposons was compiled by searching MEDLINE with the key words retroposon, retrotransposon, intronless, andpseudogene. Intronless genes were excluded that did not meet criteria described in the text. Species of mammals: M, mouse, Hu, human, R, rat, WC,woodchuck.

279

matogenic cells, even if the DNA copies of mRNAs lackthe promoter of the progenitor gene.

Another unique feature of gene expression in sper-matogenic cells, the absence of an active X chromosomein meiotic spermatogenic cells, is also believed to selectfor functional retroposons (McCarrey 1987, 1994). Sev-eral retroposons that are expressed soley in testis areautosomal genes in which an X-linked paralogous genecontaining introns is expressed in somatic and early sper-matogenic cells. Examples include the genes encodingphosphoglycerate kinase 1 and 2, pyruvate dehydroge-nase E1a subunits 1 and 2,Zfa and Zfx, glucose-6-phosphate dehydrogenases 1 and 2, and glycerol kinases1 and 2 (Ashworth et al. 1990; Dahl et al. 1990; Fitzger-ald et al. 1992, 1994; Sargent et al. 1994; Hendricksen etal. 1997). McCarrey (1987, 1994) has proposed that tran-scriptional inactivation of the X chromosome in meioticspermatogenic cells results in depletion of the somaticisoform and that the scarcity of the somatic isoform se-lects for a testis-specific isoform.

Significantly, the levels ofPabp1mRNA are at least10-fold higher in meiotic and haploid spermatogeniccells than in somatic cells in mice (Kleene et al. 1994).Thus, X-chromosome inactivation does not correlatewith disappearance of thePabp1 mRNA, and Pabpgenes are absent from the X chromosome in humans(Morris and Bodger 1993). We would like to suggest thatthe combination of high levels of poly(A) (Kleene et al.1983) and strong translational repression ofPabp1mRNAs in meiotic and early haploid spermatogenic cells(Kleene et al. 1994, Gu et al. 1995) creates a situation inwhich the levels of somatic PABP isoforms are too lowfor the amount of poly(A), thus selecting for a testis-specific PABP. Another possibility is that PABP2 has aspecial function in regulating mRNA stability and trans-lation in spermatogenic cells, cell types in which post-transcriptional gene regulation is unusually important(Davies and Willison 1993; Hecht 1995; Kleene 1996).

In conclusion, the mousePabp2gene exemplifies atendency for retroposons to be expressed in meiotic andhaploid spermatogenic cells. ThePabp2retroposon dif-fers from other retroposons that are expressed in sper-matogenic cells because thePabp1 mRNA is not de-pleted during spermatogenesis andPabp2 promoter isnot derived from thePabp1promoter. We speculate thattranscriptional promiscuity in meiotic and haploid sper-matogenic cells dramatically increases the likelihood thatnew genes are transcribed and that widespread transla-tional repression in the these cells (Kleene 1996) neu-tralizes the deleterious effects of translating high levelsof mRNAs and provides a selective advantage for newgenes. The hypothesis that the patterns of gene expres-sion in spermatogenic cells function in the creation ofnew genes could explain several puzzling observations:many mRNAs in spermatogenic cells encode truncatedproteins of questionable function (reviewed by Ivell

1992), other testis-specific mRNAs are not conserved indifferent species of mammals (Ivell 1992), and manymRNAs in spermatogenic cells are very strongly trans-lationally repressed with no evidence that a significantproportion of these mRNAs is ever translated (Kleene,1996).

Acknowledgment. This work was supported by NSF Grants DCB-90128486 and IBN-9418285.

References

Ashworth A, Skene B, Swift S, Lovell-Badge R (1990)Zfa is anexpressed retroposon derived from an alternative transcript of theZfx gene. EMBO J 9:1529–1534

Belostotsky DA, Meagher RB (1993) Differential organ-specific ex-pression of three poly(A)-binding-protein genes fromArabidopsisthaliana.Proc Natl Acad Sci USA 90:6686–6690

Boer PH, Adra CN, Lau Y-F, McBurney MW (1987) The testis-specific phosphoglycerate kinase genepgk-2is a recruited retropo-son. Mol Cell Biol 7:3107–3112

Bownes M (1990) Preferential insertion of P elements in two genesexpressed in the germ-line ofDrosophila melanogaster.Mol GenGenet 222:457–460

Brosius J, Tiedge H (1995) Reverse transcriptase: mediator of genomicpasticity. Virus Genes 11:163–179

Burd CG, Matuni EL, Dreyfuss G (1991) Multiple RNA-binding do-mains of the mRNA poly(A) binding protein have different RNA-binding specificities. Mol Cell Biol 11:3419–3424

Dahl H-H, Brown RM, Hutchison WM, Maragos C, Brown GK (1990)A testis-specific of the human pyruvate dehydrogenase E1a subunitis coded for by an intronless gene on chromosome 4. Genomics8:225–232

Davies PO, Willison KR (1993) Molecular mechanisms of differentia-tion in mammalian spermatogenesis. Semin Dev Biol 3:179–188

Deininger PL, Batzer MA (1993) Evolution of retroposons. In: HechtM, Macintyre RJ, Clegg MT (eds) Evolutionary biology, Vol 27.Plenum Press, New York, pp 157–196

Fitzgerald J, Hutchison WM, Dahl H-HM (1992) Isolation and char-acterization of the mouse pyruvate dehydrogenase E1a genes. Bio-chim Biophys Acta 1131:83–90

Fitzgerald J, Dahl H-HM, Ianello RC (1994) Differential expression oftwo testis-specific transcripts of the mouse Pdha-2 gene duringspermatogenesis. DNA Cell Biol 13:531–537

Fourel G (1992) Expression of the woodchuck N-myc2 retroposon inbrain and liver tumors is driven by the cryptic N-myc promoter.Mol Cell Biol 12:5336–5344

Grange T, Martins de Sa C, Oddos J, Pictet R (1987) Human mRNApolyadenylate binding protein: evolutionary conservation of anucleic acid binding motif. Nucleic Acids Res 15:4771–4787

Gu W, Kwon Y, Oko R, Hermo L, Hecht NB (1995) Poly(A) bindingprotein is bound to both stored and polysomal mRNAs in the mam-malian testis. Mol Reprod Dev 40:273–285

Hecht, NB (1995) The making of a spermatozoon: a molecular per-spective. Dev Genet 16:95–103

Hendrickson PJM, Hoogerbrugge JW, Baarents WM, DeBoer P, Vree-burg JTM, Vos EA, van de Lende T, Grootegoed JA (1997) Testis-specific expression of a functional retroposon encoding glucose-6-phosphate dehydrogenase in the mouse. Genomics 41:350–359

Nhim RP, Lindsey JS, Wilkinson MF (1997) A processed homeoboxgene expressed in a stage-, tissue- and region-specific manner in theepididymis. Gene 185:271–276

Hilson P, Carroll KL, Masson PH (1993) Molecular characterization of

280

PAB2, a member of the multigene family coding for poly(A) bind-ing proteins inArabidopsis thaliana.Plant Physiol 103:525–533

Ivell R (1992) ‘‘All that glitters is not gold’’—common testis genetranscripts are not always what they seem. Invest J Androl 15:85–92

Kleene KC (1996) Patterns of translational regulation in the mamma-lian testis. Mol Reprod Dev 43:268–281

Kleene KC, Distel RJ Hecht NB (1983) cDNA clones encoding cyto-plasmic poly (A)+ RNAs which first appear at detectable levels inhaploid phases of spermatogenesis. Dev Biol 98:455–461

Kleene KC, Wang M-Y, Hall C, Cutler M, Shih D (1994) Develop-mental expression of a testis-specific variant of the poly(A) bindingprotein in mouse. Mol Reprod Dev 39:355–364

Kuhn R, Monuki ES, Lemke G (1991) The gene encoding the tran-scription factor SCIP has features of an expressed retroposon. MolCell Biol 11:4642–4650

Lefrere M, Vincent A, Amaric F (1990)Drosophila melanogasterpoly(A)-binding protein: cDNA cloning reveals an unusually long3-untranslated region of the mRNA, also present in other eukaryoticspecies. Gene 96:219–225

Li W-H, Grauer D (1992) Fundamentals of molecular evolution.Sinauer Associates, Sunderland, MA

Ludwig T, Ruther U, Metzger R, Copeland NG, Jenkins NA, Lobel P,Hoflack B (1992) Gene and pseudogene of the mouse cation-dependent mannose 6-phosphate receptor. J Biol Chem 267:12211–12219

McCarrey JR (1987) Nucleotide sequence of the promoter region of atissue-specific human retroposon: comparison with a housekeepingprogenitor. Gene 61:291–298

McCarrey JR (1990) Molecular evolution of the human Pgk-2 retro-poson. Nucleic Acids Res 18:949–955

McCarrey JR (1994) Evolution of tissue-specific gene expression inmammals. BioScience 44:20–27

McCarrey JR, Thomas K (1987) Human testis-specific PGK gene lacksintrons and possesses characteristics of a processed gene. Nature(London) 326:501–505

Morris CM, Bodger MP (1993) Localization of the human poly(A)-binding protein gene (PAB1) to chromosomal regions 3q22, 12q13-q14, and 13q12-q13 byin situhybridization. Genomics 15:209–211

Nietfeld W, Mentzel H, Pieler T (1990) TheXenopus laevispoly(A)binding protein is composed of multiple functionally independentRNA binding domains. EMBO J 9:2699–3705

Nhim RP, Lindsey JS, Wilkinson MF (1997) A processed homeoboxgene expressed in a stage-, tissue-, and region specific manner inepididymis. Gene 185:271–276

Persson K, Holm I, Heby O (1995) Cloning and sequencing of an

intronless mouse S-adenosylmethionine decarboxylase gene codingfor a functional enzyme strongly expressed in the liver. J Biol Chem270:5642–5648

Reinton N, Haugen TB, Orstavik S, Skalhegg BS, Hansson V, JahnsenT, Tasken K (1998) The gene encoding the C gamma catalyticsubunit of cAMP-dependent protein kinase is a transcribed retro-poson. Genomics 49:290–297

Sachs A, Wahle E (1993) Poly(A) tail metabolism and function ineucaryotes. J Biol Chem 268:22955–22958

Sachs AB, Bond MW, Kornberg RD (1986) A single gene from yeastfor both nuclear and cytoplasmic polyadenylate-binding proteins:domain structure and expression. Cell 45:627–635

Sargent CA, Young C, Marsh S, Ferguson-Smith MA, Affara NA(1994) The glycerol kinase gene family: structure of the Xp gene,and related intronless retroposons. Hum Mol Genet 3:1317–1324

Schmidt EE (1996) Transcriptional promiscuity in testis. Curr Biol6:768–769

Schmidt EE, Schibler U (1995) High accumulation of components ofthe RNA polymerase II transcription machinery in spermatids. De-velopment 121:2373–2383

Shashidharan P, Michaelidis TM, Robakis NK, Kresovali A, Pap-matheakis J, Plaitakis A (1994) Novel human glutamate dehydro-genase expressed in neural and testicular tissues and encoded by anX-linked intronless gene. J Biol Chem 69:16971–16976

Soares MB, Schon E, Henderson A, Karathanasis SK, Cate R, ZeitlinS, Chirgwin J, Efstradiadis A. (1985) RNA-mediated duplication:the rat preproinsulin gene is a functional retroposon. Mol Cell Biol5:2090–2103

Triglia T, Peterson MG, Kemp DJ (1988) A procedure forin vitroamplification of DNA segments that lie outside the boundaries ofknown sequences. Nucleic Acids Res 16:8186

Vanin EF (1985) Processed pseudogenes: characteristics and evolution.Annu Rev Genet 19:253–272

Wang M-Y, Cutler M, Karimpour I, Kleene KC (1992) Nucleotidesequence of a mouse poly(A) binding protein cDNA. Nucleic AcidsRes 20:3519

Weiner AM, Deininger PL, Efstratiadis A (1986) Nonviral retroposons:genes, pseudogenes and transposable elements generated by reverseflow of genetic information. Annu Rev Biochem 55:631–661

Yang H, Duckett CS, Lindsten T (1995)iPABP, an inducible poly(A)binding protein detected in activated human T cells. Mol Cell Biol15:6770–6776

Yaswen P, Smoll A, Peehl DM, Trask DK, Sager R, Stampfer MR(1990) Down regulation of a calmodulin-like gene during transfor-mation of human mammary epithelial cells. Proc Natl Acad SciUSA 87:7360–7364

281