Graphic Art

11 pages
5 views

Expression of theSH2D1A gene is regulated by a combination of transcriptional and post-transcriptional mechanisms

of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
The SH2D1A gene, which is altered or deleted in patients with X-linked lymphoproliferative disease, encodes the small protein SAP (for SLAM-associated protein) that is expressed in T and NK cells. A 22-bp fragment in close proximity to an
Transcript
  Expression of the  SH2D1A  gene isregulated by a combination of transcriptional andpost-transcriptional mechanisms Susumu Okamoto 1 , Hongbin Ji 1 , Duncan Howie 1 , Kareem Clarke 1 , Charles Gullo 1 ,Stephen Manning 2 , Anthony J. Coyle 2 and Cox Terhorst 1 1 DivisionofImmunology,BethIsraelDeaconessMedicalCenter,HarvardMedicalSchool,Boston,USA  2 Millennium Pharmaceuticals, Cambridge, USA  The  SH2D1A  gene, which is altered or deleted in patients with X-linked lymphoproliferativedisease, encodes the small protein SAP (for SLAM-associated protein) that is expressed in Tand NK cells. A 22-bp fragment in close proximity to an initiator-like site was defined as thebasal promoterof mouse  SH2D1A , and a highly homologous 33-bp segment was defined asthehumanbasalpromoter.WhenanEtsconsensussitewasmutated,noreporteractivitywasdetectable. Gel mobility supershift assays revealed that the two transcription factors Ets-1and Ets-2 bind to the human and mouse sequences. The involvement of Ets-1 and Ets-2 inexpression of   SH2D1A  was functionally confirmed by overexpression studies of theirdominant-negative forms. We also found that SH2D1A mRNA decays very rapidly in mouseT cells, and its 3 0 untranslated region (UTR) has RNA-destabilizing activity in transfectionstudies with reporter/3 0 UTR constructs. As judged by RNA-gel mobility shift assays, thisrapiddegradationofSH2D1AmRNAwasduetoabalanceinbindingofthefactorsAUF1andHuR to its 3 0 UTR. Although the SH2D1A mRNA level decreased upon triggering of the T cellreceptor (TCR), the RNA degradation rate itself was not altered by TCR engagement. Key words:  XLP / Tcell / Ets family / ARE 1 Introduction  X-linked lymphoproliferative disease (XLP) is a geneticdisorder characterized by immune dysregulation andlymphoproliferation on exposure to Epstein-Barr virus(EBV) [1]. About two thirds of the patients develop fatalinfectious mononucleosis (IM), and others manifest adiverse phenotype that includes malignant lymphoma,dysgammaglobulinemia, and aplastic anemia. The cau-sative gene of XLP,  SH2D1A  [1–3], encodes a smallprotein that comprises a single SH2 domain and binds tothe cell surface protein SLAM (CD150) [4], 2B4 (CD244)[5, 6], Ly-9 (CD229), CD84 [7] and NTB-A/SF2000 [8, 9].Part of the diagnosis of XLP must include detection of amutation in  SH2D1A  [10–12]. However, germ-line muta-tions in the  SH2D1A  gene have been identified inapproximately two thirdsofthepatients. Itisconceivablethat a subgroup of XLP patients without detectable SH2D1A  mutations in the coding region might lackexpression of the SH2D1A protein because they harbormutations in the intronic sequences or the regulatoryregion of the  SH2D1A  gene [11, 12].SH2D1A-deficient mice have normal lymphocyte devel-opment, but on challenge with lymphocytic choriome-ningitis virus, they show greatly increased numbers of activated CD8 + and CD4 + cells and increased mortality[13, 14]. This is consistent with the observation that XLPpatients often have uncontrolled CD8 + cell proliferationleading to lethal parenchymal damage after EBV infec-tion. In mice and humans, SH2D1A mRNA expression islimited to thymocytes, peripheral T cells, and mature NKcells [4, 15, 16].The  SH2D1A  gene consists of four exons and threeintrons spanning approximately 25 kb. The nucleotidesequencesofhumanandmouseSH2D1AcDNAare80%identical,andthehumanandmouseSH2D1Aaminoacidsequencesshow 87% identity and 89% overall similarity,with the highest similarity in the SH2 domain. SH2D1A mRNA contains AU-rich elements (ARE) in its3 0 untranslated region (UTR), the core motif of which is[DOI 10.1002/eji.200324755] Received 21/11/03Revised 4/8/04 Accepted 23/8/04  Abbreviations: ARE:  AU-rich element  EMSA:  Electrophore-tic mobility shift assay  Inr:  Initiator  RT:  Reverse transcription UTR:  Untranslated region  XLP:  X-linked lymphoproliferativedisease3176 S. Okamoto et al. Eur. J. Immunol. 2004.  34:  3176–3186 f 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji.de  the pentanucletide AUUUA [17]. ARE are found in manymRNA with short half-lives, such as the TNF- a  transcript[18], and are believed to promote degradation of thesemRNA. Thus, ARE provide a mechanism that facilitatesthe rapid reduction in the number of copies of proteinsthat have become dispensable to the cell. BecauseSH2D1A plays an important role in T cell signaling andbecomes down-regulated upon activation, we hypothe-sizethatits expression ispost-transcriptionally regulatedthrough ARE.In this study, we examined the mechanisms that regulatethe  SH2D1A  gene both at the transcriptional and thepost-transcriptional level. We found that Ets-1 and Ets-2are important in the  SH2D1A  gene basal promoteractivity. We also demonstrated that the SH2D1A mRNA 3 0 UTR binds two RNA-binding proteins and has RNA-destabilizing activity. 2 Results 2.1 An Ets binding site is central to the activity ofthe mouse and human  SH2D1A  promoter The 3-kb sequence upstream of the translation start siteof mouse  SH2D1A  (Fig. 1A, B) contains four regions withhigh homology to the human upstream sequence(  > 80%). Some of the predicted consensus binding siteswould be specific for transcription factors known to beimportant in lymphocyte development, including PU.1,Tcf-1 and Ets-1. The transcription start site was definedby5 0 rapidamplification ofcDNA ends (RACE)formouse SH2D1A [4,16].Althoughthe putative  SH2D1A promoterregion does not contain a TATA box, segments thatconform to the initiator (Inr) consensus sequence(YYANT/AYY) [19] are found in the proximity of thetranscription start site.To identify the critical element(s) required for promoteractivity, firefly luciferase reporter constructs containingDNA segments 5 0 upstream of the mouse and human SH2D1A  ATG (+1) were transfected into Jurkat cells. Theresults represent firefly luciferase activity normalized onthe basis of the Renilla luciferase activity derived fromcotransfected pRL-TK plasmid.The mouse  SH2D1A  promoter sequence from –3162 to–1efficientlydirectedluciferaseexpression,asitsactivitywas ten times higher than that of the empty vector(Fig. 1A). By contrast, no promoter activity wasdetectable in A20 (B cell) or L929 (fibroblast) cells (datanot shown), in accordance with the restricted expressionofSH2D1AinTandNKcells.Upondeletionofsequencesbetween –3612 and –2162 and between –2162 and–1362, luciferase activity was reduced by 15% and 20%,respectively. Likewise, deletion of the –562 to –312segment decreased the activity by 50%. This set of resultssupportsthenotionthatthedistalsequencesmaycontain positive regulatory elements.However, the most dramatic change was observedfollowing deletion of the sequence between –185 and–172, which reduced transcriptional activity to back-ground levels. Deletion of the sequence between –1 and–83 did not show any effect, while deletion between –83and –163 reduced luciferase activity by 50%. Deletion of sequences between –163 and –185 completely abol-ished the transcriptional activity. Taken together, 5 0 RACE[16] and deletion analyses shown in Fig. 1A stronglysuggest that the core promoter is located within thesequence between –185 and –163.Becauseofthe sequence similarities betweenthemouseand human 5 0 flanking segments, the human  SH2D1A promoter region was examined in the same fashion(Fig. 1B). Again, the most striking decrease in promoteractivity occurred when the sequence between –167 and–134 was deleted. Deletion of the sequence between –1and –80 reduced the transcriptional activity to near-background level. As this region is downstream of theputative transcription start site [16], it possibly containsother crucial elements or binding sites for RNA-bindingproteins involved in post-transcriptional regulation.Taken together, the data suggest that the highlyhomologous mouse –185 to –163 and human –167 to–134 sequences (Fig. 1C) comprise the respective corepromoter regions.The importance of the putative Ets binding site was nextexamined by mutational analysis. When the Ets con-sensus site was mutated (ATCC to ATAG) in either themouse construct (m–312/–1) or the human  SH2D1A promoter construct (h–514/–1), a complete loss of transcriptional activity was observed (Fig. 2). Takentogether, the data suggest that the Ets consensus site iscrucial to the basal promoter activity of both the mouseand the human  SH2D1A  gene. 2.2 Ets-1 and Ets-2 bind to the  SH2D1A  corepromoter Next, the Ets consensus site in the  SH2D1A  promoterregion was further examined by gel mobility shift assays.To this end, a labeled 20-bp probe representing themouse sequence from –187 to –168, which is identical tothe human sequence from –154 to –135, was used.Incubation of the 20-bp probe with nuclear extracts fromthe human Tcell line Jurkat resulted in the formation of  Eur. J. Immunol. 2004.  34:  3176–3186 Regulation of   SH2D1A  gene expression 3177 f 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji.de  three complexes (Fig. 3A). Whereas an excess of unlabeled probe competed with the labeled DNA in allthree complexes, a mutated Ets consensus motif probe(ATCC to ATAG) did not. This indicated that complexformation was specific for the Ets consensus site.The identity of the factor(s) bound to the  SH2D1A  Etsconsensus site was next investigated using electro-phoretic mobility shift assay (EMSA) and monoclonal Abthat recognize various Ets family members (Fig. 3B).When the nuclear extracts were pre-incubated with anti- Fig. 1.  The basal promoters of the mouse and human  SH2D1A  genes. (A) The mouse  SH2D1A  promoter region is outlined at thetop. Inr-like elements, putative PU.1 and Ets binding sites are indicated. The translation start site (ATG) is designated as +1. Thevarious segments upstream of the ATG codon were inserted into the luciferase reporter vector and transfected into Jurkat cells.FireflyluciferaseactivitywasnormalizedonthebasisofRenillaluciferaseactivity.Theresultsarerepresentativeofsixindependentexperiments. (B) The same experiments for the human  SH2D1A  promoter. The results are representative of six independentexperiments. (C) Nucleotides that differ between the two species are shown in boldface in the human sequence. The DNA segments that are indispensable for  SH2D1A  promoter activity as determined in (A, B) are indicated by arrows.3178 S. Okamoto et al. Eur. J. Immunol. 2004.  34:  3176–3186 f 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji.de  Fig. 3. Ets-1 and Ets-2 bind to theEts consensus site in the SH2D1A  promoter. (A) Jurkat nuclearextracts (10  l g) were incubatedwithamouse SH2D1A Etsprobe(m–187to–168)alone(lane 2),orinthepresenceofatenfoldexcess(lane 3)ora50-foldexcess(lane 4) of unlabeled probe. In lane 5, a 50-fold excess of an unlabeled nucleotide probe with a mutation in the Ets binding site(ATCC to ATAG) was added. Lane 1 shows the mobility of the  SH2D1A  Ets probe in the absence of nuclear extract. (B) The SH2D1A  Ets probe was incubated with 10  l g of a Jurkat nuclearextract in the presence of anti-Ets-1 (lane 3), anti-Ets-2 (lane 4),or anti-Elk-1 Ab (lane 5). (C) The  SH2D1A  Ets probe was incubated with 10  l g of a Jurkat nuclear extract in the presence of anexcess of unlabeled probes containing a generic Ets consensus site (TCGAACTTCCTGCTCGAGATC) (lane 5) or a mutated Etssite (TCGAACTTCTTGCTCGAGATC) (lane 6). Fig. 2.  Site-specific mutation of an Ets binding site in the human and mouse  SH2D1A  promoter abolishes its activity. The Etsbinding site most proximal to the putative transcription start site of   SH2D1A  was mutated (AACATCCTG to AACATAGTG) in themouse m–312/–1 construct and in the human h–514/–1 construct. Firefly luciferase activity was determined as described inFig. 1A. The results are representative of three independent experiments.Eur. J. Immunol. 2004.  34:  3176–3186 Regulation of   SH2D1A  gene expression 3179 f 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji.de  Ets-1 Ab, a supershift occurred in the lowest band (C3).With the addition of anti-Ets-2 Ab, the lowest band againbecame undetectable, indicating that the Ab preventedtheformationofthecomplex.Ontheotherhand,additionof either anti-elk-1 Ab or anti-elf-1 Ab showed no effecton complex formation (Fig. 3B and data not shown). TheinvolvementofEts-1wasalsoconfirmedusing adifferentapproach (Fig. 3C). All three complexes disappearedwhen an excess of unlabeled Ets consensus probe,whichisa21-bpprobecontainingtheconsensusbindingsite for Ets-1 (AC TTCC TG), was used in a competitionassay. On the other hand, the mutated Ets consensusprobe (AC TTCT TG) did not show any effect. 2.3 Expression of dominant-negative Ets-1or Ets-2 interferes with  SH2D1A  promoteractivity  The role of Ets-1 and Ets-2 was further investigated byusing  b -actin promoter-driven vectors expressing domi-nant-negative mutants of either Ets-1 or Ets-2. WhilepCXN Ets-1 and pCXN Ets-2 encode the normal Etsproteins, both pCXN DN Ets-1 (N70) and pCXN DN Ets-2(  D 1–328) only encode the DNA binding domain. Becausethe expressed proteins do not have the transactivationdomain, they work as dominant-negative proteins intransfected cells. When Jurkat cells were cotransfectedwith  SH2D1A  promoter constructs and wild-type Etsexpression vectors, no significant effect was observed(Fig. 4). This is probably because Jurkat cells constitu-tively express sufficient levels of endogenous Ets-1 andEts-2 proteins. However, when they were transfectedwith dominant-negative Ets-1/2 vectors, the luciferaseactivity decreased by 50% in both cases. Together, theEMSA and transfection data confirm a positive roleof theEts-1 and Ets-2 transcription factors in the functionalactivity of the  SH2D1A  basal promoter. 2.4 Destabilizing activity in the 3 0 UTR of SH2D1A mRNA  While there are two transcripts identified for humanSH2D1Adiffering only in the length oftheir 3 0 UTR, only asingle 0.9-kb transcript was detected thus far for mouseSH2D1A (Fig. 5A). However, there remains a possibilitythat larger transcripts of mouse  SH2D1A  exist, becausesequence analysis beyond exon 4 revealed highlyhomologous regions, including polyA addition signals.The 3 0 UTR of mouse SH2D1A mRNA is highly homo-logous to that of human SH2D1A transcripts andcontains three AUUUA motifs. To investigate whetherthe SH2D1A 3 0 UTR regulates the stability of the mRNA,the mouse SH2D1A 3 0 UTR sequence was cloned inbetween the luciferase cDNA and the polyA signal in thepGL3 control vector carrying an SV40 promoter. Tran-scripts from the resulting plasmids would thereforecontain the luciferase-coding region and the SH2D1A 3 0 UTR.Asexpected,theluciferaseactivitydecreasedby40–50%, both in EL4 and A20 cells (Fig. 5B).To prove that the SH2D1A 3 0 UTR actually has adestabilizing effect on SH2D1A mRNA, the half-life of SH2D1A transcripts with or without 3 0 UTR was com-pared upon transfection into A20 cells. As shown inFig. 5C, SH2D1A mRNA without 3 0 UTR had a half-lifesimilar to that of GAPDH transcripts. By contrast,SH2D1A mRNA with 3 0 UTR was degraded significantlymore rapidly than GAPDH.To identify the factors that can bind to SH2D1A 3 0 UTR,RNA EMSA was performed. Cytoplasmic extracts fromEL4 cells were incubated with the  32 P-labeled,  in vitro transcribed 300-nucleotide SH2D1A 3 0 UTR RNA probe. After RNase T1 digestion, they were run on nativeagarose gels. As shown in Fig. 5D, four distinct Fig. 4.  Dominant-negative forms of Ets-1 and Ets-2 interferewith  SH2D1A  promoter activity. The expression vectorsencoding Ets dominant-negative forms were cotransfectedwith the mouse  SH2D1A  reporter construct m–312/–1 (A) orthe human  SH2D1A  reporter construct h–514/–1 (B). Fireflyluciferase activity was determined as in Fig. 1A. The resultsare representative of four independent transfection experi-ments.3180 S. Okamoto et al. Eur. J. Immunol. 2004.  34:  3176–3186 f 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.eji.de
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks