Long Terminal Repeats Are Used as Alternative Promoters for the Endothelin B Receptor and Apolipoprotein C-I Genes in Humans

Abstract

To examine the potential regulatory involvement of retroelements in the human genome, we screened the transcribed sequences of GenBankTM and expressed sequence tag data bases with long terminal repeat (LTR) elements derived from different human endogenous retroviruses. These screenings detected human transcripts containing LTRs belonging to the human endogenous retrovirus-E family fused to the apolipoprotein CI (apoC-I) and the endothelin B receptor (EBR) genes. However, both genes are known to have non-LTR (native) promoters. Initial reverse transcription-polymerase chain reaction experiments confirmed and authenticated the presence of transcripts from both the native and LTR promoters. Using a 5′-rapid amplification of cDNA ends protocol, we showed that the alternative transcripts of apoC-I and EBR are initiated and promoted by the LTRs. The LTR-apoC-I fusion and native apoC-I transcripts are present in many of the tissues tested. As expected, we found apoC-I preferentially expressed in liver, where about 15% of the transcripts are derived from the LTR promoter. Transient transfections suggest that the expression is not dependent on the LTR itself, but the presence of the LTR increases activity of the apoC-I promoter from both humans and baboons. The native EBR-driven transcripts were also detected in many tissues, whereas the LTR-driven transcripts appear limited to placenta. In contrast to the LTR of apoC-I, the EBR LTR promotes a significant proportion of the total EBR transcripts, and transient transfection results indicate that the LTR acts as a strong promoter and enhancer in a placental cell line. This investigation reports two examples where LTR sequences contribute to increased transcription of human genes and illustrates the impact of mobile elements on gene and genome evolution. To examine the potential regulatory involvement of retroelements in the human genome, we screened the transcribed sequences of GenBankTM and expressed sequence tag data bases with long terminal repeat (LTR) elements derived from different human endogenous retroviruses. These screenings detected human transcripts containing LTRs belonging to the human endogenous retrovirus-E family fused to the apolipoprotein CI (apoC-I) and the endothelin B receptor (EBR) genes. However, both genes are known to have non-LTR (native) promoters. Initial reverse transcription-polymerase chain reaction experiments confirmed and authenticated the presence of transcripts from both the native and LTR promoters. Using a 5′-rapid amplification of cDNA ends protocol, we showed that the alternative transcripts of apoC-I and EBR are initiated and promoted by the LTRs. The LTR-apoC-I fusion and native apoC-I transcripts are present in many of the tissues tested. As expected, we found apoC-I preferentially expressed in liver, where about 15% of the transcripts are derived from the LTR promoter. Transient transfections suggest that the expression is not dependent on the LTR itself, but the presence of the LTR increases activity of the apoC-I promoter from both humans and baboons. The native EBR-driven transcripts were also detected in many tissues, whereas the LTR-driven transcripts appear limited to placenta. In contrast to the LTR of apoC-I, the EBR LTR promotes a significant proportion of the total EBR transcripts, and transient transfection results indicate that the LTR acts as a strong promoter and enhancer in a placental cell line. This investigation reports two examples where LTR sequences contribute to increased transcription of human genes and illustrates the impact of mobile elements on gene and genome evolution. long terminal repeat endogenous retrovirus human ERV apolipoprotein C-I endothelin B receptor polymerase chain reaction reverse transcription-PCR rapid amplification of cDNA ends kilobase pair(s) base pair(s) hepatic control region splice donor endothelin A very high proportion of mammalian DNA consists of retroelements that have arisen via RNA reverse transcription and reintegration into the genome (1Smit A.F.A. Curr. Opin. Genet. Dev. 1999; 9: 657-663Crossref PubMed Scopus (728) Google Scholar). Retroelements are found in most, if not all, species, where they have amplified to high copy numbers during evolution (2Boeke J.D. Stoye J.P. Coffin J. Hughes S.H. Varmus H.E. Retroviruses. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1997: 343-435Google Scholar). The sheer number of such mobile elements suggests that they affect the host genome, and several observations indicate that retroelements impact on the species in a number of ways by acting as insertional mutagens or contributing regulatory functions to genes (3Kidwell M.G. Lisch D. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 7704-7711Crossref PubMed Scopus (438) Google Scholar). While transposable elements can be harmful to their host, the vast majority of transposable elements present in humans are derived from ancient transpositional events which are fixed in Old World primates. Potential long term effects of the majority of these elements must be either neutral or beneficial; otherwise they would be eliminated by selection. Human DNA contains essentially two classes of retrosequences, (i) the non-long terminal repeat (non-LTR)1 retroposons represented by LINE and Alu sequences, and (ii) the LTR retroelements in which the endogenous retroviruses (ERVs), solitary LTRs derived from ERVs, and other LTR-like sequences fall (4Smit A.F.A. Curr. Opin. Genet. Dev. 1996; 6: 743-748Crossref PubMed Scopus (488) Google Scholar). Human ERVs (HERVs) are classified into different families based on sequence similarity and monophyletic clustering (5Larsson E. Kato N. Cohen M. Curr. Top. Microbiol. Immunol. 1989; 148: 115-132Crossref PubMed Scopus (128) Google Scholar, 6Tristem M. J. Virol. 2000; 74: 3715-3730Crossref PubMed Scopus (266) Google Scholar). The thousands of ERVs and solitary LTRs that are present in human DNA are the result of infections and transposition events during primate evolution. Solitary LTRs are common features in the human genome, and they probably arose from a recombination event between the 5′ and 3′ LTR of a full-length provirus. Despite their evolutionary age, many ERVs are still transcriptionally active in human cells, where different ERV families show quite different sites and levels of transcription (7Wilkinson D. Mager D.L. Leong J.-A.C. Levy J. The Retroviridae. Plenum Press, New York1994: 465-535Crossref Google Scholar). The LTR and ERV elements are especially interesting in this regard, since they naturally possess enhancer, promoter, and polyadenylation functions within their LTRs, which probably accounts for differences in transcription of the various HERV families. Besides promoting transcription of retroviral genes, several studies have demonstrated that ERVs and LTRs can assume gene regulatory functions (8Britten R.J. Gene (Amst.). 1997; 205: 177-182Crossref PubMed Scopus (159) Google Scholar, 9Sverdlov E.D. FEBS Lett. 1998; 428: 1-6Crossref PubMed Scopus (109) Google Scholar, 10Brosius J. Gene (Amst.). 1999; 238: 115-134Crossref PubMed Scopus (304) Google Scholar). For example, the paratoid-specific expression of amylase in humans is dependent and under control of an HERV-E element (11Ting C.N. Rosenberg M.P. Snow C.M. Samuelson L.C. Meisler M.H. Genes Dev. 1992; 6: 1457-1465Crossref PubMed Scopus (232) Google Scholar). HERV-E also appears to be involved in the expression of human pleiotrophin in placenta (12Schulte A.M. Lai S. Kurtz A. Czubayko F. Riegel A.T. Wellstein A. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 14759-14764Crossref PubMed Scopus (177) Google Scholar). It has also been demonstrated that a HERV-K LTR encodes the last 67 amino acids of one form of the leptin receptor OBR (13Kapitonov V.V. Jurka J. J. 1999; PubMed Scopus Google Scholar). These indicate that an LTR to or within a gene have a of effects gene gene expression or the transcription levels of the which be are LTR sequences to the involvement of retroelements in gene we have the expressed sequence tag and transcribed of GenBankTM for retroviral gene we two human genes that are by ERV LTRs, the apolipoprotein C-I (apoC-I) gene and the endothelin B receptor (EBR) show that these two genes a HERV-E LTR as an alternative promoter, the presence of the transcripts in human tissues, and the of the LTRs the of apoC-I and transcription with the reaction as M. J. J. Virol. 1992; PubMed Scopus Google Scholar). of cDNA of the RNA were either or from different of as J.D. Mager D.L. J. Virol. PubMed Google Scholar). The were to the different in for LTR-apoC-I fusion native apoC-I fusion native EBR by of cDNA with the one of for by the apoC-I or the EBR of for for and for and one for In the of different EBR the were fusion native EBR total EBR In these cDNA in the as but with to effects during the The of the amplification from the for RNA to the of the in the DNA or in the to the of the in the DNA or in the in a of EBR transcripts in of the LTR and native by The LTR-driven form detected by derived from the native form with from and the total EBR amplified derived from on cDNA of different of of the DNA are to the amplification were the different The of the different is to the of the EBR levels of the different are with a and show the of the native and LTR-driven were to the total in and cDNA were from and a cDNA in as in the with the The amplification EBR and the by the and with the apoC-I and The by EBR and the by the and with the apoC-I and The for one for by for and and for The were the were for and The to of apoC-I, with and of the with of total RNA in a containing for The were to and of and of were to the reaction and to a of with a of and The reaction by for for and for The reaction were on a containing and by to The of the were the on essentially as Mager D.L. J. Virol. 1998; PubMed Google Scholar). DNA from World World and human cell Mager D.L. PubMed Scopus Google in were to the presence or of the solitary LTR in the apoC-I of different primates. The LTR of the EBR amplified by sequence the LTR and sequence from the The apoC-I and EBR were from numbers and The of apoC-I and EBR were amplified from human DNA and of the gene in the To into were with terminal sequences for The is where the are The numbers in indicate the and of the in the DNA to the of the of the two genes. The sequences are the of the EBR LTR and the hepatic control region to EBR and apoC-I are the sequences to the EBR LTR and the are where sequence in the The EBR were For the region of the native EBR transcription and and amplification were For the LTR amplified with and and were by the LTR a of the native promoter region of The LTR amplified with amplification were into the of which is from the on either in or with to the native EBR promoter region were The of the based on the of GenBankTM number The LTR were derived from GenBankTM number The apoC-I were For the region of the apoC-I transcription and This contains both the native and LTR promoter For the LTR of the apoC-I amplified with and in of For the LTR from the apoC-I by the non-LTR of with and The two amplification were with and into This has the as the that the LTR is and were by the LTR a of the native apoC-I promoter region the LTR The LTR amplified with amplification were into the of either in or with to the apoC-I promoter region were The of the is with to the apoC-I of GenBankTM number The apoC-I were For the apoC-I amplified from DNA with and For amplified with and The two amplification were with and This into The LTR that amplified with into the of of the the LTR in the apoC-I the as in into of the are from GenBankTM number and with to the of the apoC-I The from human DNA and and The sequences were derived from and of GenBankTM number The with and the into the of apoC-I which is 3′ to the DNA by the to were in with were to transfections in a of Transient transfections of were by of DNA and of as in the with the were in with were in a of and with of DNA and of of as in the from the the were and the were the and to the were in and were an and an Using the in we GenBankTM and the human expressed sequence tag data bases the LTR and the region of HERV elements (7Wilkinson D. Mager D.L. Leong J.-A.C. Levy J. The Retroviridae. Plenum Press, New York1994: 465-535Crossref Google Scholar). with the of the LTR and other HERV sequence probably by the whereas transcripts with or probably by an data base we two transcripts where HERV-E J. Virol. PubMed Google sequences were fused to the EBR gene number and the apoC-I gene number The of these transcripts suggests that the EBR the splice donor in the region of the HERV which is of the 5′ LTR of an provirus. The is in of HERV-E transcripts, that this the of HERV-E M. J. Virol. PubMed Google Scholar). The apoC-I fusion LTR-driven which is derived from a solitary LTR and into the To the presence of fusion transcripts, we to the retroviral and the of the this in is to the presence and the of the fusion transcripts in human of the genes were as a different promoter from the retroviral LTR D. J. PubMed Google Scholar, M. S. S. M. Scholar, S. S. Scholar). to these two as the apoC-I and EBR promoter. To of the LTR and native transcripts, we also a to transcripts of the native of the two genes. of the on a of derived from different human tissues are in The EBR is to where levels appear with that of the expressed native In the of apoC-I, transcripts from the native promoter in were high as D. J. PubMed Google but are also by in many of the other from the solitary LTR were detected in two both of which were also detected in many The result of this the presence of fusion transcripts between LTRs of HERV-E and the genes for EBR and The LTRs the EBR and apoC-I in their with the EBR LTR in of the the of the fusion transcripts, where the two fusion of apoC-I are the result of in the 5′ To that the apoC-I and EBR fusion transcripts within the LTRs and not transcripts from a promoter of the LTRs, we the of both LTR fusion gene Using a protocol, we that both the apoC-I and EBR fusion within their LTRs and of several showed that the apoC-I and EBR is of a of HERV-E J. Virol. PubMed Google Scholar, M. J. Virol. PubMed Google Scholar). This is the also by other HERV-E a full-length transcribed HERV-E element number of the of the apoC-I In the of the EBR fusion the sequence the also within the but a 3′ to the apoC-I the apoC-I and EBR were the of The retroviral of the the apoC-I is a solitary which is of the native apoC-I promoter. The two sites are by where the sites of the native and LTR are and of the apoC-I The EBR LTR not present in the sequence of the EBR native promoter which is of the EBR M. S. S. M. Scholar, S. S. Scholar). A containing both the HERV-E element and the EBR in GenBankTM The sequence of this is in a of and contains of The of the LTR and region that were in EBR are to the element of It has been that several other alternative transcripts are by and of the and alternative of the M. Gene (Amst.). 1999; PubMed Scopus Google Scholar). The 5′ LTR of the HERV-E element is from the EBR gene and is by to the splice in the as are the The and the of the different transcripts these two are in to the retroviral the fusion transcripts have different with the native but the apoC-I and EBR Using the sites of the LTRs in of different primate we the of of various HERV-K elements during primate evolution Mager D.L. J. Virol. 1998; PubMed Google Scholar). Using the we were to the two HERV-E LTRs in the primate The apoC-I LTR detected in whereas Old and New World not have this LTR in the apoC-I that the the of and the Old World about J. PubMed Scopus Google Scholar). we the presence of the EBR LTR both in and but not in New World we that the LTR the EBR is the apoC-I the between the New and Old World J. PubMed Scopus Google Scholar). of the 5′ and 3′ LTRs of the EBR HERV-E that they are The of is by that the two LTRs an of since in the primate a of into M. J. Virol. 2000; 74: 3715-3730Crossref PubMed Scopus (266) Google Scholar, M. PubMed Scopus Google Scholar). not between the LTR-driven and native transcripts for either the apoC-I and EBR genes. For the apoC-I transcripts, are in from to and the in this of is For the sequence of the LTR-driven is from the retroviral and is not to this region as a of the of the LTR sequences in human a from several tissues and an derived from of Using this we detected transcripts of to the native and the LTR-driven detected transcripts to the apoC-I LTR fusion As expected, the of transcription in which is the of apoC-I transcription D. J. PubMed Google Scholar). we that the derived from the LTR promoter of the total in sites of transcription and were also detected this However, in these tissues the of transcription in also a the region of apoC-I in The in for of the other tissues, that the the of apoC-I in A of this be effects in the of the or of the To the of the fusion we a This by of the reverse transcribed RNA to effects during the LTR-driven EBR fusion detected in we RNA from different of the placenta in with for the native and the LTR-driven also for the which would of the total of EBR in the different on the of the cDNA of the total of EBR transcripts were by to be derived from the native promoter, and of the total derived from the LTR promoter. As has been we for EBR expression in A.M. 1997; PubMed Scopus Google Scholar). To the of the LTR in expression of the apoC-I we the native promoter which naturally contains the LTR and the native promoter, of a also the activity of the LTR by and the native where the LTR transient transfections to the levels of promoter activity of the different The LTR also a in with the apoC-I promoter where the LTR to for the that the LTR acts as an enhancer of the native promoter. The expression in is dependent on a D. S. J. J. J. PubMed Scopus Google and we promoter activity of the apoC-I the presence of this The results of the transfections of with a of apoC-I are in and suggest that the LTR by is not contributing to the expression levels of apoC-I in However, the LTR is from the apoC-I the promoting activity of the region about in that the presence of the LTR in the apoC-I to the activity of the native promoter However, we found that the LTR acts as an enhancer in a from the native promoter. A of the the LTR the of in the primate would be to the LTR into the apoC-I of a species that naturally the showed that have the LTR in the apoC-I but that is in the The sequence of apoC-I has been M. S. 1992; PubMed Scopus Google and sequence of the human and the of the LTR in the the LTR into the apoC-I the and the promoting activity between the with and the The LTR into the in increased to that in the human that the LTR a in the primate To the of the LTR in EBR the native EBR promoter region or the LTR were of the gene of also the LTR a in and with to the native EBR promoter to for potential of the LTR on the EBR native promoter The cell and the cell were with these and the results are in In both and the activity of the native EBR promoter is and is that the native EBR promoter is dependent on an enhancer element not present in the or on a that is in the cell However, the LTR in either a with to the native promoter a significant in activity that the LTR can as an enhancer of the native promoter region containing the LTR of the gene were into a very high activity with that in and in with the other in or the promoter control The high activity of the LTR in and activity in with the where the fusion transcripts were detected in placenta. As an control of activity of the with the LTR of the promoter were into of the of the LTR with to the promoter, a in activity to with the promoter not that the LTR also the promoter in placental In this we detected and alternative transcripts of the apoC-I and EBR genes with HERV-E sequences their 5′ fusion transcripts are expressed in a of human tissues and were by to of a within HERV-E LTRs, that the LTRs are alternative for these genes in For apoC-I, we found that a of transcripts is derived from the LTR promoter in The of apoC-I in other tissues is not and the transcription levels are in However, the LTR and native appear to be active in many of the other tissues tested. In the of be that the fusion from a placental by S. J. PubMed Google but to be a gene or to the LTR in the which from the region of EBR M. S. S. M. Scholar, S. S. Scholar). In of placental the form to derived from the native promoter, that this is the form in placenta. However, the cDNA derived from different of the placenta that of the total derived from the LTR promoter, on the placental cDNA In and the total of amplified EBR to be for the LTR and native derived amplification which be for by the these transcripts are also expressed in placenta M. Gene (Amst.). 1999; PubMed Scopus Google Scholar). the LTRs of the apoC-I and EBR are in the expression of the two fusion transcripts is where activity of the is to placenta and the apoC-I transcripts are detected in many It is that expression of the is to of the HERV in is a by mammalian to the expression of gene and retroelements Genet. 1997; PubMed Scopus Google Scholar, Genes Dev. 1999; PubMed Scopus Google Scholar). The apoC-I LTR be from and expressed in to to the native apoC-I promoter for the different transcription is that have or transcription sites in the two LTRs. The of the LTRs is probably a of their into the that the LTR into the EBR about and that the LTR into the apoC-I It is that HERV-E elements were in the primate during this the HERV-E element and the HERV-E of the pleiotrophin are of A. M.G. PubMed Scopus Google Scholar, A.M. Wellstein A. J. Virol. 1998; PubMed Google Scholar). As is the for many other HERV this endogenous family have been that the HERV-E elements are fixed in the primate Transient transfections were to the of the LTRs in the of apoC-I and these experiments the the results the apoC-I the in where the LTR not to contribute to the expression levels of apoC-I in However, the LTR from the apoC-I the promoting activity of the apoC-I about in This result suggests that the presence of the LTR in the apoC-I to the activity of the native promoter by which in with the native regulatory The genes the human and are in a on and with the to with M. S. J. PubMed Google Scholar). The different have in where apoC-I is to with apolipoprotein in the levels and in the of in the M.H. 1999; PubMed Scopus Google Scholar). that have the LTR in the apoC-I but is in the LTR in the apoC-I we an increased expression to that for the the of is that the LTR by either neutral or on It is that the presence of the LTR would have been if a strong impact on apoC-I in M. S. J. PubMed Google which has been as a for of a apoC-I gene in humans S. J. 2000; PubMed Scopus Google Scholar). both the in and transfection results suggest that the LTR has a on the expression levels of apoC-I, one is that the LTR an for the apoC-I is that the LTR a for in the of to tissues, functions during of limited In contrast to the LTR the apoC-I a significant of the EBR transcripts is derived from the LTR promoter in placenta. The LTR also increases the activity of the native EBR promoter region in transient transfection that this LTR has a in acting both as promoter and enhancer for the expression of EBR in placenta. In human are in the via and and as of placental S. J. Google Scholar, J. 1992; 74: PubMed Scopus Google Scholar). The for and in placental is by studies in where an in and receptor with a rapid in placental A. S. M. 1996; PubMed Scopus Google whereas are in of placental J.D. Sci. PubMed Scopus Google Scholar). the of the of and the in different of the placenta are and not studies show that the LTR to expression of While the of EBR in placenta be an evolutionary event is that an increased receptor would as a for the high levels of that are present in the which in have in placental and As is in this the of LTRs and other retroelements to in other genes is sequences are present within the potential is not the element not be of the to on gene expression are to be common to on the and of the element with to genes. It is that such elements have been as evolutionary in the of many in that they in the of and levels of gene in have an in In LTRs and other retroelements are in the evolution of gene for example, leptin is under control of a repeat element that acts as an enhancer for this gene in placenta S. M. J. 1997; PubMed Scopus Google Scholar). However, leptin is not expressed in placenta the element is in examples where gene control elements have during primate evolution of an enhancer element the of or of a regulatory region the of by HERV-E (11Ting C.N. Rosenberg M.P. Snow C.M. Samuelson L.C. Meisler M.H. Genes Dev. 1992; 6: 1457-1465Crossref PubMed Scopus (232) Google Scholar, A.M. Lai S. Kurtz A. Czubayko F. Riegel A.T. Wellstein A. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 14759-14764Crossref PubMed Scopus (177) Google Scholar). It is that HERV-E elements are found involved in gene regulatory functions these elements are not as as other HERV families in the human genome (7Wilkinson D. Mager D.L. Leong J.-A.C. Levy J. The Retroviridae. Plenum Press, New York1994: 465-535Crossref Google Scholar). a for the LTR is not is that the of HERV sequences or LTR functions gene In we have two HERV-E elements that increased transcription of the EBR and apoC-I genes in humans by of promoter and enhancer functions from their LTRs and to the where LTRs have been to gene regulatory and for