신호전달체계에서의동종이량체의역할연구 Exploration of Homodimer Receptor: Homodimer Protein Interactions Abstract Background: Homodimerisation is a production of protein-protein complex composed of two identical molecules. Dimerisation is a phenomenon that often occurs in the regulation of biochemical systems like signal transduction pathways. We investigated if the existence of a homodimer-activated receptor and the activation of homodimer transducers is the general pattern in cell signaling. Methods: We developed a workflow to merge the data from the Gene Ontology and the BOND database to produce a list of interactions between homodimer receptors and homodimer proteins. To decide whether the results are significant from a statistical point of view, we performed two additional analysis. Furthermore, we verified the suspected affinity of homodimer receptors to homodimer transducers through χ² test and Yates correction. Results: With the analysis of Gene Ontology and BOND database, twenty-two interactions were detected between 12 potential homodimer receptors and 22 potential homodimer transducers. The ratio of homodimeric protein transducers interacting with the randomly selected receptors was 5.1% of the total number of interacting transducers, whereas the ratio in the homodimer-homodimer list was 23.9%. The p -value of homodimer receptors to interact with homodimer transducers was 2 10-5. Similarly, we calculated the ratio of homodimer receptors in the interactions with homodimer transducers and non-homodimer transducers. From the statistical test, the ratio of homodimer receptors was 60% and its p-value was 5.3 10-4. Conclusions: Our analysis indicates that homodimer receptor-homodimer transducer interactions are a frequent pattern in cell signaling. We detected a significant amount of these interactions and confirmed that there is a prevalence of such interactions when comparing their frequency with randomly selected non-homodimeric interactions between receptors and protein transducers. 질병관리본부국립보건연구원생물안전평가과권태수, 김대원, 강연호 1) 들어가는말 단백질이량체형성 (Protein dimerisation) 은동종 (Homodimerisation) 또는이종 (Heterodimerisation) 의단백질소단위가결합하여복합체를형성하는것으로정의할수있다. Klemm[1] 등은이합체형성이신호전달체계에있어서단백질 -단백질상호작용의방향을유도하거나, 단백질전송의 위치를결정하거나, 단백질이단량체에서이량체로전이하는것을조절하는등의생물학적기능을갖고있음을밝혔다. 뿐만아니라, 효소, 이온채널, 수용체, 전사조절인자와같은단백질의조절에있어서핵심적인역할을하고있다 [2]. 일례로, Brenda enzyme database(http://www.brenda. uni-koeln.de/) 를검색해보았을때 452개의인체유래효소중 125개의효소가동종이량체의구조를갖고있었다. Caspase 는 apoptosis 중에세포의사멸을조절하는효소인데한개의 1) 교신저자 (kyhfisher@nih.go.kr/043-719-8040) 138 www.cdc.go.kr
Vol. 8 No. 7 PUBLIC HEALTH WEEKLY REPORT, KCDC 사슬구조를갖고있다. 그러나실제단백질분해작용을할때는동종이량체를형성하여작용한다. 그외에도많은제한효소와전사조절인자와같은 DNA 결합단백질들이동종이량체구조를갖고있다. 동종이량체를형성하는또다른한가지예가세포표면의수용체인데, 작용물질 (Agonist) 에반응하고신호를전달하는과정에서다량체 (Oligomer) 를형성한다. 이때, 동종이량수용체는신호전달에관여하는 kinase와전사조절인자의이량체형성과활성에관여하는것으로밝혀졌다 [3][4]. 그대표적인예가 JAK2/STAT5 경로이다 [5]. 이때, 활성화된동종이량수용체인 EpoR은전사인자인 STAT5의활성화와 동종이량체형성을유도한다 (Figure 1). 활성화된동종이량수용체와동종이량전달체의활성이 존재한다는보고는이러한일련의과정이매우일반적인과정일 Figure 1. Scheme of the mechanism of activation and dimerisation suggested by the dynamics of interaction in JAK-STAT pathways. 것임을시사해준다. 본글에서우리는동종이량수용체-동종이량전달체상호작용이신호전달체계에서일반적인현상인지알아보았다. 이를위해공개데이터베이스를검색하기위한생물정보학적분석을기획하고수행하였다. 몸말 Figure 2. Scheme of the search method used in our analysis to find homodimer-homodimer interactions. 우리는인간세포에서의신호전달체계에초점을맞추었기때문에분석대상은데이터베이스의정보중 Homo sapiens 에대한정보로제한하였다. 생물정보학적검색을위해단백질의특성과상호작용을대상으로하였다 (Figure 2). Gene Ontology (GO) 상에서수용체로분류된단백질만을수용체로정의하였다. 이수용체들의상호작용정보가 BOND(Biomolecular Object Network Databank)[6] 데이터베이스에있는데수용체자신과상호작용 (Self-interaction) 하는경우동종이량수용체로보았다 (Figure 2의 R-R 상호작용 ). 그다음으로신호전달에있어서동종이량수용체와상호작용하는모든단백질을찾았다 (Figure 2; A, B, P). 이단백질중에서 또다시자기상호작용 (Self-interaction) 하는단백질들을선택하였다. 그결과, 우리는동종이량수용체와동종이량전단체간의동종이량-동종이량상호작용만을볼수있게되었다. 이를위해, GO와 BOND 데이터베이스로부터실험적으로검증되고논문에인용된동종이량수용체와동종이량전달체의상호작용목록을만들었다. GO 데이터베이스는유전자와유전자의생산물에대해공통적이고구조적이며명확한용어를제공하는데신호전달체계에있어서특별한역할이있는단백질과역할이없는단백질을구분하는것은중요하였다. 수용체- 전달체상호작용을찾기위해 BOND를사용하였는데 www.cdc.go.kr 139
실험적으로검증된단백질과다른생물분자들과의상호작용정보를갖고있는가장완벽한데이터베이스이다. GO 데이터베이스를추출하기위해 GO 검색도구인 AmiGO(http:// amigo.geneontology.org/cgi-bin/amigo/go.cgi) 를사용하였다. 한편, BOND 데이터베이스로부터대량의상호작용정보를가져오기위해 BOND의 SOAP 인터페이스를이용하였다. 본연구에서사용한데이터검색과분석파이프라인은 Figure 3에순서대로나타내었다. 데이터분석은두단계로이루어졌다. 먼저 GO 용어중에서 Homo sapiens 의단백질중 receptor activity 를갖는것들을찾았다 ( 총 848개 ). 다음으로, 목록에 포함된수용체단백질들의상호작용대상단백질들을 BOND 데이터베이스로부터검색하였다. 이리스트중에서자기자신과상호작용하는수용체만을선택하였는데이들을동종이량체를만드는수용체로정의하였다. 이를통해 24개의동종이량수용체후보를찾았다. 다음으로, GO에서세포신호전달과연관되어있고 BOND 데이터베이스에서앞서찾은 24개동종이량수용체후보와상호작용하는단백질들을검색하였다. 그결과 146개의상호작용대상단백질들을동정하였다. 마지막으로, 146개단백질중자기 -상호작용 (Self-interaction) 을하는 22개의단백질들을동종이량전달체후보로선택하였다. 이들 동종이량전달체와상호작용하는동종이량수용체는 12개이었다. 최종동종이량수용체- 동종이량전달체목록은 Figure 4에나타내었다. 12개의동종이량수용체후보와 22개의동종이량전달체후보사이에 22개의상호작용이있었다. 동종이량수용체의절반이고유한동종이량전달체를상호작용대상으로했으나, 20% 의동종이량수용체는세개이상의동종이량단백질과상호작용하였다 (Figure 4). 그특별한예가 TGFBR1 동종이량수용체인데, 10개의동종이량전달체와상호작용하였다. TGFBR1 수용체가다른수용체에비해동종이량전달체와더많은상호작용을하는이유는명확하지않다. 매우재미있는것은, TGFβ 수용체시스템이다른수용체시스템보다도동종이량체형성을통한신호전달을더선호한다는것이다. 이상의상호작용분석에대해통계적유의성을검증하였다. 먼저, 동종이량수용체와상호작용하는동종이량전달체의빈도를무작위로선택한동종이량체를형성하지않는수용체와상호작용하는동종이량전달체의빈도와비교하였다. 이때 GO 데이터베이스로부터수용체로구성된 10개의집합을무작위로생성한뒤앞서서술한방법을통해이들수용체와상호작용하는단백질들을 BOND로부터검색하였다. 또한, 이들단백질들이동종이량체를생성하는지여부를 BOND 를이용하여확인하였고 Figure 3. Flow chart describing the database search performed 동종이량수용체 - 동종이량전달체목록과비교하였다. 중복을제거한뒤각무작위집합에서의평균적인수용체 140 www.cdc.go.kr
Vol. 8 No. 7 PUBLIC HEALTH WEEKLY REPORT, KCDC Figure 4. Homodimer receptor - homodimer transducer interactions for H. sapiens detected in our investigation. 개수는 42개였다. 무작위로선택된수용체와상호작용하는동종이량전달체의비율은상호작용하는전달체중 5.1% 인반면에, 동종이량수용체는 23.9% 의동종이량전달체와상호작용하였다. 이는무작위로선택된수용체보다 5배나높은수치였다. 덧붙여, 무작위로선택된수용체는평균적으로 0.21 개의동종이량전달체와상호작용하였는데이는동종이량수용체 (1.83개) 보다 9배나적은수치였다. 이러한동종이량수용체의동종이량전달체에대한친화도를검증하기위해 χ 2 테스트를수행하였다. 또한작은표본사이즈에대한보정을위해 Yates 보정 [7] 을하였다. 그결과 2 10-5 의 p-value를얻었고동종이량수용체와상호작용하는전달체는동종이량이아닌수용체와상호작용하는전달체와많은차이가있음을알수 있었다. 다음으로, 동종이량전달체와상호작용하는동종이량수용체개수와무작위로선택된동종이량이아닌전달체와상호작용하는동종이량수용체의갯수를비교하였다. 이를위해평균 28개의전달체로구성된 10개의전달체집합을무작위로선택하였고이들과상호작용하는수용체의목록을얻기위해 BOND 데이터베이스를검색하였다. 동종이량전달체와동종이량수용체의상호작용은 60%(20개중 12개 ) 이었으나, 무작위집합에서전달체와동종이량수용체상호작용의비율은 5.58% 이었다. 이는 11배나높은수치이었다. 더욱이, 무작위전달체와상호작용하는동종이량수용체의비율은 0.04% 에불과했으나동종이량전달체와상호작용하는 Figure 5. Results of the statistical analysis performed when considering the interacting partners of receptors. www.cdc.go.kr 141
Figure 6. Results of the statistical analysis performed when considering the interacting partners transducers. 동종이량수용체는 0.55% 로 14 배나높았다. 이결과에대한 χ 2 실험적인방법으로확인할수도있을것이다. 테스트와 Yates 보정을하였다. 그결과 5.3 10-4 의 p-value를얻었고동종이량전달체와동종이량이아닌수용체와의상호작용이동종이량전달체와동종이량수용체와의상호작용과많은차이가있음을알수있었다. 결국, 동종이량전달체는다른전달체보다동종이량수용체와상호작용을더많이한다고추정할수있었다. 마지막으로, 우리는동종이량수용체-동종이량전달체상호작용의실제적인근거를찾기위해 BOND 데이터베이스에기록되어있는상호작용의근거와단백질의 4차구조를참조하였다 (Table 1). 그결과, 대부분의동종이량수용체- 동종이량전달체상호작용이실험적, 구조적근거가있었지만아직실험적근거가부족한것도있었다. 또한 EGFR-GRB2 의경우처럼상호작용의실험적근거는있지만, 아직 GRB2의 맺는말 이상의분석을통해동종이량수용체-동종이량전달체상호작용이신호전달체계에있어서매우흔한현상임을알수있었다. 우리는이러한상호작용을찾아내었고수용체와전달체간의무작위적인상호작용과비교하여빈도가높음을증명하였다. 이러한동종이량체 -동종이량체상호작용이인간의다른신호전달체계에서도나타날것으로기대한다. 최근 Bornberg-Bauer 등의연구에서도신호전달중심에있는신호전달단백질들이동종이량체를생성하지않는경우보다동종이량체를생성하는경우가더많음을밝혔다 [8][9]. 단백질 4 차구조가밝혀지지않아동종이량체형성을확인할수 없는경우도있었다. 이러한한계에도불구하고, 향후더많은단백질상호작용데이터와단백질 4차구조가축적된다면본연구에의해예측된동종이량수용체-동종이량전달체상호작용이실험적인근거를갖게될것으로기대한다. 또한향후본연구에서예측한동종이량수용체- 동종이량전달체상호작용을 Yeast two hybrid, Co-immunoprecipitation, Affinity chromatography, 단백질 3차구조확인등의 참고문헌 1. Klemm, J.D. et al. 1998. Dimerisation as a regulatory mechanism in signal transduction. Annu. Rev. Immunol. 16:569-592. 2. Neelan J. et al. 2004. The power of two: protein dimerization in biology, Trends Biochem Sci. 24(11):618-625 142 www.cdc.go.kr
Vol. 8 No. 7 PUBLIC HEALTH WEEKLY REPORT, KCDC Table 1. Homodimer receptor-homodimer transducer interaction table. The interaction evidence, PubMedID and MMDB id were retreived from BOND database. Exceptionally, the homodimer structure of TGFBR1 was obtained from the published papers. Receptors(11) Transducer(24) Receptor-transducer interaction Receptor name EntrezID Quaternary Transducer Quaternary Quaternary PubMedID Evidence EntrezID PubMedID Evidence PubMedID Evidence structure BIND MMDB name structure BIND MMDB BINDID CD80 941 homodimer 93294 15806 11279502 Three Dimensional Structure CTLA4 1493 homodimer 93292 15805 11279501 Three Dimensional Structure 93295 11279502 EGFR 1956 homodimer 159814 27406 15093539 Three Dimensional Structure GRB2 Stat3 2885 6774 homodimer homodimer 337268 183815 35229 16165154 15653507 Three Dimensional Structure 128216 58094 8305738 12873986 Affinity Chromatography ERBB3 2065 homodimer 106689 20305 12154198 Three Dimensional Structure GRB2 2885 homodimer 337268 35229 16165154 Three Dimensional Structure 314233 Affinity Chromatography FGFR1 2260 homodimer 74595 6253 8752212 Three Dimensional Structure BNIP-2 663 130844 10954711 FKBP1A 2280 IGF1R 3480 100673 103205, 103206, 103207, 114489 948, 8015, 18337, 10937, 14199 19235, 22842 PDGFRB 5159 295854 15889147 Affinity Chromatography, Two Hybrid Test, Co- Immunoprecipitation 11735566 Three Dimensional Structure mtor 2475 178665 15467718 74597 8662507 11591350 Three Dimensional Structure, Cross Linking, Resonance Energy Transfer CSK 1445 258571 12588871-197285 7559507 262780 10551883 Two Hybrid Test Three Dimensional Structure, Two Hybrid Test IRS1 3667 homodimer 82263 10940 10411883 Three Dimensional Structure 45374 7541045 Two Hybrid Test PKD1 5587 315546 16100512 Immunoprecipitation 194829 11884618 SRC 6714 homodimer PRKCH 5583 262388 12223477 - PKD1 5587 77009, 77012, 77013, 77016 315546, 262390 7564, 7565, 7566, 7570 9174343 Three Dimensional Structure 64774 2173144 16100512, 12223477-262385 12223477 PRKCZ 5590 209434 15665819 - Stat3 6774 183815 15653507 300769 10446219 TGFBR1 7046 homopentamer heterodimer * TNFRSF1A 7132 homodimer 97792, 97793, 97794 99072, 70705, 73022 17203 11583628 Three Dimensional Structure CDK6 1021 92448 15491 Three Dimensional Structure 218432 15761153 17787, 3995, 5348 8387891, 7768931, 11592999, 8196061 Three Dimensional Structure FKBP1A 2280 100673, 100674, 100675, 66062, 77888, 82257, 88865 18337, 948, 8015, 10937 11735566, 7678431, 9871618, 9448138, 9448137, 10425089 Three Dimensional Structure 81294 10025408 Three Dimensional Structure IKK-beta 3551 178827 9346485 218380 15761153 SMAD2 4087 SMAD3 4088 MADH4 4089 TGFBR2 7048 UBA52 7311 302368, 218605 302355, 218607 89721, 83448, 83449, 83450 262674, 197391 111384, 68114 29559 29554 14492, 11496 21862, 2261 15350224, 15761153 15350224, 15761153 11224571, 10647180 15840165, 8774881 12507430, 8107144 Three Dimensional Structure, Three Dimensional Structure, 217980 15761153 196744 8774881 Three Dimensional Structure 197393 8774881-194832 14612425 Three Dimensional Structure 218258 15761153 SAT2 112483 homodimer 338413 35739 16596569 Three Dimensional Structure 218574 15761153 CLIK1 140901 193841 11973348-218360 15761153 Sentrin 7341 homodimer 55020 11112409 Two Hybrid Test 55754 8906799 MAP-1 64112 homodimer 153633 11060313, Affinity Chromatography Two Hybrid Test, Affinity Chromatography 297458 15949439 heterodimer * : J Cell Biol. 1998. 140(4): 767-777; J Biol Chem, 1998. 273(48): 31770-31777 www.cdc.go.kr 143
3. Reich Z. et al. 1997. Ligand-specific oligomerisation of T-cell receptor molecules. Nature. 387:617-620. 4. Heldin C.H. 1995. Dimerization of cell surface receptors in signal transduction. Cell. 80:213-223. 5. Ferrajoli A. et al. 2006. The JAK-STAT pathway: a therapeutic target in hematological malignancies. Curr. Cancer Drug Targets. 6:671-679. 6. Bader G.D. et al. 2000. BOND-a data specification for storing and describing biomolecular interactions, molecular complexes and pathways. Bioinformatics. 16:465-477. 7. Yates F. 1934. Contingency table involving small numbers and the χ 2 test. Supplement to the Journal of the Royal Statistical Society 1(2):217 235. 8. Amoutzias G.D. et al. 2004. The evolution of protein interaction networks in regulatory proteins. Comp. Funct. Genomics. 5(1):79-84. 9. Amoutzias G.D. et al. 2007. A protein interaction atlas for the nuclear receptors: properties and quality of a hub-based dimerisation network. BMC Syst. Biol. 1:34. 144 www.cdc.go.kr