Initial sequencing and analysis of the human genome - Vitagenes

More documents

Recommendations

Info

articles2,000Number of distinct domain architectures1,8001,6001,4001,2001,000800600400TransmembraneExtracellularIntracellular2000Human Fly Worm YeastFigure 40 Number of distinct domain architectures in the four eukaryotic genomes,predicted using SMART 339 . The number of architectures is split into three cellularenvironments: intracellular, extracellular and membrane-associated. The increase inarchitectures for the human, relative to the other lineages, is seen when these numbersare normalized with respect to the numbers of domains predicted in each phylum. Toavoid artefactual results from the relatively low detection rate for some repeat types,tandem occurrences of tetratricopeptide, armadillo, EF-hand, leucine-rich, WD40 orankyrin repeats or C2H2-type zinc ®ngers were treated as single occurrences.architectures could not be an artefact resulting from erroneous genepredictions. Three-quarters of the architectures can be found inknown genes, which already yields an increase of about 50% overworm and ¯y. We expect the ®nal number of human architectures togrow as the complete gene set is identi®ed.)A related measure of proteome complexity can be obtained byconsidering an individual domain and counting the number ofdifferent domain types with which it co-occurs. For example,the trypsin-like serine protease domain (number 12 in Fig. 41)co-occurs with 18 domain types in human (including proteinsinvolved in the mammalian complement system, blood coagulation,and ®brinolytic and related systems). By contrast, the trypsin-likeserine protease domain occurs with only eight other domains in ¯y,®ve in worm and one in yeast. Similar results for 27 common domainsare shown in Fig. 41. In general, there are more different co-occurringdomains in the human proteome than in the other proteomes.One mechanism by which architectures evolve is through thefusion of additional domains, often at one or both ends of theproteins. Such `domain accretion' 340 is seen in many human proteinswhen compared with proteins from other eukaryotes. The effect isillustrated by several chromatin-associated proteins (Fig. 42). Inthese examples, the domain architectures of human proteins differfrom those found in yeast, worm and ¯y proteins only by theaddition of domains at their termini.Among chromatin-associated proteins and transcription factors,a signi®cant proportion of domain architectures is shared betweenthe vertebrate and ¯y, but not with worm (Fig. 43a). The trend waseven more prominent in architectures of proteins involved inanother key cellular process, programmed cell death (Fig. 43b).These examples might seem to bear upon the unresolved issue of theevolutionary branching order of worms, ¯ies and humans, suggestingthat worms branched off ®rst. However, there were other cases inwhich worms and humans shared architectures not present in ¯y. Aglobal analysis of shared architectures could not conclusivelydistinguish between the two models, given the possibility of lineage-speci®closs of architectures. Comparison of protein architecturesmay help to resolve the evolutionary issue, but it will requiremore detailed analyses of many protein families.New physiology from old proteins. An important aspect ofNumber of co-occurring domains70605040302010MustardweedYeastWormFlyHuman01 2 3 4 5 6 7 8 9 10 1112 131415 16 17 18 19 20 21 22 23 24Domain family25 26 27Figure 41 Number of different Pfam domain types that co-occur in the same protein, foreach of the 10 most common domain families in each of the ®ve eukaryotic proteomes.Because some common domain families are shared, there are 27 families rather than 50.The data are ranked according to decreasing numbers of human co-occurring Pfamdomains. The domain families are: (1) eukaryotic protein kinase [IPR000719];(2) immunoglobulin domain [IPR003006]; (3) ankyrin repeat [IPR002110]; (4) RING ®nger[IPR001841]; (5) C2H2-type zinc ®nger [IPR000822]; (6) ATP/GTP-binding P-loop[IPR001687]; (7) reverse transcriptase (RNA-dependent DNA polymerase) [IPR000477];(8) leucine-rich repeat [IPR001611]; (9) G-proteinb WD-40 repeats [IPR001680];(10) RNA-binding region RNP-1 (RNA recognition motif) [IPR000504]; (11) C-type lectindomain [IPR001304]; (12) serine proteases, trypsin family [IPR001254]; (13) helicaseC-terminal domain [IPR001650]; (14) collagen triple helix repeat [IPR000087];(15) rhodopsin-like GPCR superfamily [IPR000276]; (16) esterase/lipase/thioesterase[IPR000379]; (17) Myb DNA-binding domain [IPR001005]; (18) F-box domain[IPR001810]; (19) ATP-binding transport protein, 2nd P-loop motif [IPR001051];(20) homeobox domain [IPR001356]; (21) C4-type steroid receptor zinc ®nger[IPR001628]; (22) sugar transporter [IPR001066]; (23) PPR repeats [IPR002885];(24) seven-helix G-protein-coupled receptor, worm (probably olfactory) family [IPR000168];(25) cytochrome P450 enzyme [IPR001128]; (26) fungal transcriptional regulatory protein,N terminus [IPR001138]; (27) domain of unknown function DUF38 [IPR002900].NATURE | VOL 409 | 15 FEBRUARY 2001 | www.nature.com © 2001 Macmillan Magazines Ltd905
articlesaYW, F, VBr Br Ba Br Br Br Br Br Br Ba hmgARSC1/2* CG11375* *ZnfZnfbYEp1 Ep2 Ep1 Ep2 BrEp1PHDPHD*YPR031wWPHDPHDZnfF, H WF, HPHDPHDEp2 Br BMB Ch Ch SWI2 Sa hmg Ch Ch SWI2 SaPHDPHDPHDPHD*Lin-49**peregrin*CHD-3/T14G8.1 Mi-2cY, F, VEp1 Ep2E(Pc)-likeCommonancestorPHDPHDPHD2...5PHDPHDBrPHDhmgFsja* TrxW, F, HALRsjaCCCCCCSETSETCCA A A*Figure 42 Examples of domain accretion in chromatin proteins. Domain accretion invarious lineages before the animal divergence, in the apparent coelomate lineage and thevertebrate lineage are shown using schematic representations of domain architectures(not to scale). Asterisks, mobile domains that have participated in the accretion. Speciesin which a domain architecture has been identi®ed are indicated above the diagram(Y, yeast; W, worm; F, ¯y; V, vertebrate). Protein names are below the diagrams. TheMe*PHDPHDPHDHBrHrxPHDsja CCC SET Cdomains are SET, a chromatin protein methyltransferase domain; SWI2, a superfamily IIhelicase/ATPase domain; Sa, sant domain; Br, bromo domain; Ch, chromodomain; C, acysteine triad motif associated with the Msl-2 and SET domains; A, AT hook motif; EP1/EP2, enhancer of polycomb domains 1 and 2; Znf, zinc ®nger; sja, SET-JOR-associateddomain (L. Aravind, unpublished); Me, DNA methylase/Hrx-associated DNA binding zinc®nger; Ba, bromo-associated homology motif. a±c, Different examples of accretion.vertebrate innovation lies in the expansion of protein families. Table25 shows the most prevalent protein domains and protein familiesin humans, together with their relative ranks in the other species.About 60% of families are more numerous in the human than in anyof the other four organisms. This shows that gene duplication hasbeen a major evolutionary force during vertebrate evolution. Aab60102Conserved domain architectures in chromatin proteinsConserved domain architectures in apoptotic proteins3331816Human and flyHuman and wormWorm and flyAll threeHuman and flyHuman and wormWorm and flyAll threeFigure 43 Conservation of architectures between animal species. The pie charts illustratethe shared domain architectures of apparent orthologues that are conserved in at leasttwo of the three sequenced animal genomes. If an architecture was detected in fungi orplants, as well as two of the animal lineages, it was omitted as ancient and its absence inthe third animal lineage attributed to gene loss. a, Chromatin-associated proteins.b, Components of the programmed cell death system.comparison of relative expansions in human versus ¯y is shown inFig. 44.Many of the families that are expanded in human relative to ¯yand worm are involved in distinctive aspects of vertebrate physiology.An example is the family of immunoglobulin (IG) domains,®rst identi®ed in antibodies thirty years ago. Classic (as opposed todivergent) IG domains are completely absent from the yeast andmustard weed proteomes and, although prokaryotic homologuesexist, they have probably been transferred horizontally frommetazoans 341 . Most IG superfamily proteins in invertebrates arecell-surface proteins. In vertebrates, the IG repertoire includesimmune functions such as those of antibodies, MHC proteins,antibody receptors and many lymphocyte cell-surface proteins. Thelarge expansion of IG domains in vertebrates shows the versatility ofa single family in evoking rapid and effective response to infection.Two prominent families are involved in the control of development.The human genome contains 30 ®broblast growth factors(FGFs), as opposed to two FGFs each in the ¯y and worm. Itcontains 42 transforming growth factor-bs(TGFbs) compared withnine and six in the ¯y and worm, respectively. These growth factorsare involved in organogenesis, such as that of the liver and the lung.A ¯y FGF protein, branchless, is involved in developing respiratoryorgans (tracheae) in embryos 342 . Thus, developmental triggers ofmorphogenesis in vertebrates have evolved from related but simplersystems in invertebrates 343 .Another example is the family of intermediate ®lament proteins,with 127 family members. This expansion is almost entirely due to111 keratins, which are chordate-speci®c intermediate ®lamentproteins that form ®laments in epithelia. The large number ofhuman keratins suggests multiple cellular structural support rolesfor the many specialized epithelia of vertebrates.Finally, the olfactory receptor genes comprise a huge gene familyof about 1,000 genes and pseudogenes 344,345 . The number of olfactoryreceptors testi®es to the importance of the sense of smell invertebrates. A total of 906 olfactory receptor genes and pseudogenescould be identi®ed in the draft genome sequence, two-thirds ofwhich were not previously annotated. About 80% are found inabout two dozen clusters ranging from 6 to 138 genes and encompassingabout 30 Mb (,1%) of the human genome. Despite theimportance of smell among our vertebrate ancestors, hominids906 © 2001 Macmillan Magazines Ltd NATURE | VOL 409 | 15 FEBRUARY 2001 | www.nature.com
Page 2 and 3: articlesGenome Sequencing Centres (
Page 4 and 5: articleswell as targeted regions of
Page 6 and 7: articlesultimate goal of a complete
Page 8 and 9: articlesthat were used to `seed' th
Page 11 and 12: articlesFingerprint clone contigPic
Page 13 and 14: articlesof the end of the contig, s
Page 15 and 16: articles®ngerprint map. However, m
Page 17 and 18: articlesnucleotides to collections
Page 19 and 20: articlesTable 10 Number of CpG isla
Page 21 and 22: articlesClasses of interspersed rep
Page 23 and 24: articlesposable elements in categor
Page 25 and 26: articlesProportion of genome compri
Page 27 and 28: articlesgenomes 184±187 . By study
Page 29 and 30: like termini show the typical diver
Page 31 and 32: articlesvelocardiofacial/DiGeorge a
Page 33 and 34: articlesaSum of aligned bases (kb)c
Page 35 and 36: articlesthe IAC anticodon for a Val
Page 37 and 38: articlesTable 21 Characteristics of
Page 39 and 40: articles30% to 50%. The correlation
Page 41 and 42: articlesTable 22 Properties of the
Page 43 and 44: articlesTable 23 Properties of geno
Page 45: articlesmultifunctional proteins an
Page 49 and 50: articlesSegmental history of the hu
Page 51 and 52: articlesThe most extreme mechanism
Page 53 and 54: articlesinclude the DiGeorge/veloca
Page 55 and 56: articlesbecause of the availability
Page 57 and 58: articlesNature 380, 152±154 (1996)
Page 59 and 60: articles265. Sorensen, P. D. & Fred
Page 61 and 62: articles443. Roth, F. P., Hughes, J

Initial sequencing and analysis of the human genome - Vitagenes

Create successful ePaper yourself

Delete template?

Save as template?