Phylogenetic Classification of Prokaryotic and Eukaryotic Sir2-like Proteins
Roy A.Frye 1
Pittsburgh V.A.Medical Center (132L),Department of Pathology,University of Pittsburgh,Pittsburgh,Pennsylvania 15240
Received May 31,2000
Sirtuins (Sir2-like proteins)are prent in pro-karyotes and eukaryotes.Here,two new human sir-tuins (SIRT6and SIRT7)are found to be similar to a particular subt of inct,nematode,plant,and pro-tozoan sirtuins.Molecular phylogenetic analysis of 60sirtuin conrved core domain quences from a di-ver array of organisms (including archaeans,bacte-ria,yeasts,plants,protozoans,and metazoans)shows that eukaryotic Sir2-like proteins group into four main branches designated here as class I–IV.Pro-karyotic sirtuins include members of class II and III.A fifth class of sirtuin is prent in gram positive bac-teria and Thermotoga maritima.Saccharomyces cer-evisiae has five class I sirtuins.Caenorhabditis elegans and Drosophila melanogaster have sirtuin genes from class I,II,and IV.The ven human sirtuin genes include all four class:SIRT1,SIRT2,and SIRT3are class I,SIRT4is class II,SIRT5is class III,and SIRT6and SIRT7are class IV.©2000Academi
c Press
马达加斯加3歌曲Key Words:molecular phylogeny;evolution;Sir2;chromatin;aging;DNA stability;epigenetic;NAD me-tabolism;deacetyla.
In addition to silencing genes within the silent mating-type loci and telomeric regions the yeast Sir2protein is targeted to the nucleolus (1,2)where it caus veral effects such as alteration of chromatin structure (3),gene silencing (4–6),modulation of the meiotic checkpoint (7),decread recombination of rDNA (8),and a decread rate of aging (9).All Sir2-like proteins have a sirtuin core domain which contains a ries of quence motifs conrved in organisms ranging from bacteria to humans (10,11).Bacterial,yeast,and mammalian sirtuins are able to metabolize NAD and possibly act as mono-ADP-ribosyltrans-feras (11,12).The enzymatic function of sirtuins is not yet completely understood but recent reports of
histone-activated Sir2-mediated NAD metabolism (12,13)and NAD-activated Sir2-mediated histone deacety-lation (13,14)suggest a possible coupled reciprocal activation mechanism involving interactions of Sir2with NAD and the N ⑀-acetyl-lysine groups of acety-lated histones.It is possible that an enzymatic mech-anism of this sort that couples binding and metabolism of both NAD and N-li
nked acetyl groups is a key fea-ture of all sirtuins;however not all sirtuins function physiologically by interacting with acetylated histones becau prokaryotes lack histones yet many pro-karyotes posss sirtuins,also some eukaryotic sir-tuins appear to be cytoplasmic and thus not accessible to histones (15,16).Here the conrved sirtuin core domain quences from a variety of prokaryotic and eukaryotic organisms are compared and it is found that the eukaryotic sirtuin quences can be grouped into four main class.
MATERIALS AND METHODS
Characterization of human SIRT6and SIRT7cDNAs.Using the human SIRT4quence as a probe,a BLAST arch of human quences in GenBank yielded veral expresd quence tags (ESTs)for SIRT6and SIRT7and the genomic quence for SIRT6.The Clontech human spleen Marathon cDNA library rved as a source for cDNA clones that were isolated using PCR primers de-rived from the GenBank quences and the Clontech AP1primer.Pfu-turbo high-fidelity thermostable polymera from Stratagene was ud for amplification and the PCR products were cloned into the InVitrogen pcr4TOPO cloning vector and quenced with an ABI PRISM 377.
BLASTP similarity score analysis.The full length amino acid quences for each sirtuin was compare
d with the quence of the other sirtuin using the “Blast 2quences”utility at the NCBI Blast website (bi.v/gorf/bl2.html).For the D.mel3-quence the currently available GenBank Drosophila genomic DNA quence has a gap of 17codons just upstream of the region encoding the conrved FGE motif.
Molecular phylogenetic analysis.Most of the 60sirtuin quences listed in Table 1were obtained from GenBank,some of the quences were obtained from preliminary quence data from The Institute for Genomic Rearch (TIGR)accesd via the “Microbial Genomes:Finished and Unfinished”Blast website (bi.v/Microb_blast/unfinishedgenome.html).In the TIGR-relead prelim-inary DNA quence data for the B.per sirtuin gene there are three
1
To whom correspondence should be addresd at V.A.Medical Center (132L),University Drive C,Pittsburgh,PA 15240.Fax:412-688-6872.E-mail:frye01@pitt.edu.
Biochemical and Biophysical Rearch Communications 273,793–798(2000)doi:10.1006/bbrc.2000.3000,available online at
on
X’s designating indefinite bas;in the analysis for this paper the three X’s were provisionally substituted with G’s becau the molec-ular phylogeny analysis software(e below)would not process -quences with X’s.For some quences the putative protein-encoding quences were derived by splicing together gments of genomic quence;putative splice sites were determined using the Gene Finder software(dot.du:9331/gene-finder/gf.html) and by asssing homology to similar known sirtuin cDNAs.The conrved core domain quences were aligned by ClustalW and further adjustments to the alignment were made manually using the SeqPUP program.The multiple quence alignment data was con-verted to Phylip format and the60aligned quences were analyzed by the MOLPHY:protML program(17)and the resulting“outtree”datat was converted to a treefile using the PHYLIP:connsus program.The treefile was converted to an u
nrooted dendrogram using the PHYLIP:drawtree program.The MOLPHY and PHYLIP programs were utilized via the Pasteur Institute Bioweb site( bioweb.pasteur.fr/).
sherlock holmesRESULTS
SIRT6and SIRT7human sirtuin genes.The hu-man sirtuin gene family is comprid of ven known members.Each sirtuin contains a conrved core do-main,in some instances additional N-terminal or C-terminal quence is prent(Fig.1A).For some of the human sirtuin genes(SIRTs1,2,3,4,and6)the genomic quence is available(Table1).A SIRT5-like pudogene is prent on chromosome1p31.2-32.1 (GenBank Accession No.AL157407bas46500–48000).The SIRT6gene is on chromosome19p13.3 between CDC34and D19S325.The SIRT7gene has been mapped(as Unigene EST cluster Hs.184447)to chromosome17q between D17S784and qTEL.The SIRT6and SIRT7cDNAs encode proteins with pre-dicted M
r
of39.1and44.9KDa,respectively.The SIRT6and SIRT7quences are much more similar to each another than they are to Sir2(BLASTP similarity scores:SIRT6:SIRT7ϭ178,SIRT6:Sir2ϭ61,SIRT7: Sir2ϭ53).The SIRT6and SIRT7quences are highly similar to some sirtuins from Drosophila melano
gaster, Caenorhabditis elegans,Oryza sativa(rice),Arabidop-sis thaliana,and the protozoan malaria parasite Plas-modium falciparum(Fig.1B).
Drosophila homologues of human sirtuins.The genomic quence of Drosophila melanogaster has re-cently been determined(18).A arch of this Drosoph-ila databa revealedfive sirtuin genes(Table1).
Thefive Drosophila sirtuins are homologues offive of the human sirtuins.Very high BLASTP similarity scores are en when each Drosophila sirtuin is com-pared with the corresponding human homologue (D.mel1:SIRT1ϭl2:SIRT2ϭl3: SIRT4ϭ263, D.mel4:SIRT6ϭ294,and D.mel5: SIRT7ϭ364).The data suggest that sirtuins from organisms reprenting diver phyla can be grouped into distinct quence-related class.
Molecular phylogeny of prokaryotic and eukaryotic sirtuin core domain quences.Molecular phyloge-netic analysis software,the MOLPHY:protML pro-gram(17),was ud to analyze the aligned conrved core domains of60sirtuins from a variety of pro-karyotes and eukaryotes.This yielded an unrooted tree diagram with the eukaryotic sirtuins grouped into four main branches(Fig.2).
Class I:The main branch leading toward Sir2is designated as class I.The yeast Sir2and proteins suc狠狠地
h as Hst1,SIRT1,C.ele1,l1form a subgroup called class Ia.The yeast Hst2and proteins that in-clude SIRT2,SIRT3,l2form the subgroup class Ib.The yeast Hst3and Hst4genes are subgroup Ic;the currently available data indicates that
sirtuins FIG.1.SIRT6and SIRT7are two new members of the human sirtuin family with high homology to certain eukaryotic sirtuins.(A) The placement of the conrved core domain(darkly shaded)in Sir2 visiae and the ven human sirtuins(sizes range from SIRT5at33.9kDa to SIRT1at81.7kDa).(B)Aligned amino acid quences comparing human SIRT6and SIRT7with similar sirtuins from fruitfly,nematode,two plants,and the malaria parasite.Black shading indicates residues that are conrved in at least50%of the quences.The portrayed quence gments all begin with the initiation methionine except for l5which begin at the55th and29th amino acids respectively.See Table1for accession information on quences.
of this Ic subgroup are found only in yeasts.Class I sirtuins are not prent in prokaryotes.
Class II:The human SIRT4and the fruitfll3are class II sirtuins (Fig.2).The two class II sirtuin genes of Caenorhabditis elegans are both located on the same cosmid clone,perhaps the result of a gene dupli-cation event.Interestingly,some bacteria have class II sirtuin genes (e.g.,Streptomyces coelicolor A3(2),My-cobacterium avium,and Bordetella pertussis ).
Class III:The human SIRT5is a class III sirtuin (Fig.2).Candida albicans and Plasmodium falcipa-rum also have class III sirtuins.Most bacterial sirtuin quences are of class III type;in addition to tho n
oted in Table 1there are other bacterial class III quences prent in the TIGR databa (e.g.,-quences from Pudomonas aeruginosa and Pasteu-rella multocida ).There are some completely-quenced bacterial genomes which lack sirtuin genes,the in-clude:Rickettsia prowazekii,Borrelia burgdorferi,Chlamydia muridarum,Chlamydia pneumoniae,Chlamydia trachomati,Mycoplasma genitalium,Chlamydophila pneumoniae,Synechocystis,Neisria meningitidis,Treponema pallidum,and Ureaplasma urealyticum.Several archaeans (e.g.,Archaeoglobus fulgidus,Aeropyrum pernix,Pyrococcus furiosus,Pyro-coccus horikoshii,and Pyrococcus abyssi )have class III quences but the completely-quenced archaeans Methanobacterium thermoautotrophicum and Meth-anococcus jannaschii lack sirtuin genes.
Class IV:The SIRT6and SIRT7sirtuins are both class IV sirtuins.The class IV sirtuins are further subdivided into class IVa (SIRT6, D.mel4, C.ele4,P.Fal2,O.sat,and A.tha)and class IVb (l5).Class IV sirtuins are not prent in pro-karyotes.
Class U:Several Firmicute (gram positive)bacteria and Thermotoga maritima have sirtuin genes with -quence motifs that em to be intermediate between class II and III and the class I and IV;this undif-ferentiated form of sirtuin gene is called “class U.”The gram positive bacterium Streptomyces coelicolor A3(2)has a class II sirtuin (S.coe1)and a cond sirtuin (S.coe2)that is class
U.
Characteristic quence motifs of the various class of sirtuins.Figure 3illustrates the specific amino acid quence motifs that characterize the different class of sirtuins.The core domain quences of ven members from each of the five class were first shaded to indicate intraclass-conrved resi-dues and then the five grouped quence ts were aligned to facilitate interclass comparison.There are veral short motifs of conrved amino acids prent within the sirtuin core domain;the include GAGISXXXGIPXXR,PXXXH,TQNID,HG,two ts of CXXC that may be a zinc finger domain (10),FGE,GTS,and (I/V)N (Fig.3).In class III sirtuins the
TABLE 1
Sirtuin Sequence Abbreviations,Sirtuin Class Designa-tion,Source Organism,and GenBank Accession Numbers (Genomic Regular Font/cDNA Italic Font)
Abbreviation (Class)Organism (Genomic #/cDNA #)
A.act (III)Actinobacillus actinomycetemcomitans (AF006830)A.aeo (III)Aquifex aeolicus (AE000776)
A.ful1(III)Archaeoglobus fulgidus (AE000987)A.ful2(III)Archaeoglobus fulgidus (AE001098)A.per (III)
Aeropyrum pernix (AP000062)A.tha (IVa)Arabidopsis thaliana (AB009050)
B.per (II)Bordetella pertussis (TIGR)B.sub (U)Bacillus subtilis (Z99109)
erar的音标C.ace (U)Clostridium acetabutylicum (TIGR)C.alb1(Ia)Candida albicans (O59923)C.alb2(Ib)Candida albicans (CAA22018)C.alb3(Ic)Candida albicans (TIGR)C.alb4(III)Candida albicans (TIGR)C.dif (U)Clostridium difficile (TIGR)
C.ele1(Ia)Caenorhabditis elegans (Z70310)C.ele2(II)Caenorhabditis elegans (Z50177)C.ele3(II)Caenorhabditis elegans (Z50177)C.ele4(IVa)Caenorhabditis elegans (U97193)C.jej (III)Campylobacter jejuni (AL139077)
Hst1(Ia)Saccharomyces cerevisiae (U39041)Hst2(Ib)Saccharomyces cerevisiae (U39063)Hst3(Ic)Sa
ccharomyces cerevisiae (U39062)Hst4(Ic)Saccharomyces cerevisiae (Z48784)H.pyl (III)Helicobacter pylori (AE001545)
L.maj1(Ib)Leishmania major (AL160493/L40331)L.maj2(II)Leishmania major (AL117324)M.avi1(II)Mycobacterium avium (TIGR)M.avi2(III)Mycobacterium avium (TIGR)
M.tub (III)Mycobacterium tuberculosis (Z95584)O.sat (IVa)Oryza sativa (AF159133)
P.aby (III)Pyrococcus abyssi (AJ248286)
P.fal1(III)Plasmodium falciparum (AL049185)P.fal2(IVa)Plasmodium falciparum (TIGR)P.hor (III)Pyrococcus horikoshii (AP000004)S.aur (U)Staphylococcus aureus (M32103)
<1(II)Streptomyces coelicolor A3(2)(AL121596)
<2(U)Streptomyces coelicolor A3(2)(AL035205,AL035206)S.pom1(Ia)Schizosaccharomyces pombe (AL035637)S.pom2(Ib)Schizosaccharomyces pombe (AL121807)
S.pom3(Ic)Schizosaccharomyces pombe (AL136499/p (III)Salmonella typhimurium (U89687)Sir2(Ia)Saccharomyces cerevisiae (X01419)SIRT1(Ia)Homo sapiens (AL133551/NM_0122
38)SIRT2(Ib)Homo sapiens (AC011455/NM_012237)SIRT3(Ib)Homo sapiens (AF015416/NM_012239)SIRT4(II)Homo sapiens (AC003982/NM_012240)SIRT5(III)Homo sapiens (NM-012241)瘾科技
SIRT6(IVa)Homo sapiens (AC006930/AF233396)SIRT7(IVb)Homo sapiens (AF233395)
T.bru (Ib)Trypanosoma brucei (AF102869)T.mar (U)Thermotoga maritima (AE001726)Y.pes (III)
Yersinia pestis (TIGR)
GAGISXXXGIPXXR motif is usually GAGISAESGI-PTFR whereas in class II it is GAGISTESGIPDYR.Sirtuins of class U,I,and IV also have GIPD within this motif,thus the prence of GIPT within this motif is indicative of a class III quence.In class IV quences there is a GVWTL motif prent four residues C-terminal to the GAGISTXXGIPDFR motif.In class II quences there is an RXRY-WARXXXGW motif prent 27residues C-terminal to the GAGISTESGIPDYR motif.In quences of class II,III,and U,the PXXXH motif is PNXXH while in the eukaryotic class I and IV quences a T or S usually follows the P.The HG motif is of interest becau the point mutation that changes HG to YG caus loss of sirtuin-mediated ADP-ribosylation and deacetylation (11,12,14)and it converts yeast SIR2to a dominant negative gene (12).The HG mo
tif is strictly conrved in all known sirtuins.In class Ia the HG motif reads CHG,in class Ib it is AHG,in class Ic (and in most non-class I sirtuins)it is LHG.The three residues located five residues C-terminal to the GTS motif are rather uful in differenti-ating between the types of sirtuins.The three residues usually en at this position for each class and subclass are:Ia(PVA or PVS),Ib(PFA),Ic(GVK),U(PAA),II(SGY),III(PAA),IVa(PXX),and IVb(KKY).
DISCUSSION
Although some prokaryotes lack Sir2-like proteins it appears that sirtuins are prent in all eukaryotes.The five sirtuin genes of the completely-quenced Sac-charomyces cerevisiae are all of class I subtypes;thus this yeast has evolved to survive in the abnce of non-class I sirtuins.However,except for yeasts most eukaryotic organisms em to harbor an assortment of both non-class I and class I sirtuin genes.Many types of eukaryotes have class II and class IV sirtuins.In addition to the human,inct,nematode,and Leish-mania class II sirtuins described here (Table 1),partial-length quence data indicate that class II sirtuins are prent in Trypanosoma brucei (TIGR_5691|T.brucei_27H6.TJ),a mold fungus (As-pergillus nidulans AA788405),and plants (A.thaliana N38434,Soybean AW395917).The class IV sirtuins are also widely distributed and en in vertebrate,inct,nematode,protist,and plant lifeforms (Fig.1).In summary,the currently available d
skyviewata indicate that class I,class II,and class IV sirtuins could be prent in all metazoan organisms and that some protists also posss the class of sirtuins.
The genomic quence data ts lanogaster and C.elegans are complete and no class III sirtuin quences are found;thus the two organisms
熟能生巧的故事
中秋节快乐的英文are
FIG.2.The molecular phylogeny of sirtuins.An alignment of the conrved domains of 60sirtuin quences was analyzed by the MOLPHY:protML program and the PHYLIP:connsus and PHYLIP:drawtree programs were then ud to construct this unrooted tree diagram.See Table 1for accession information on quences.
examples of metazoans which lack class III sirtuins.The class III sirtuin is the most widely distributed form found in prokaryotes,thus it may be a very ancient version of the sirtuin gene.Becau class III genes are found in most non-gram-positive bacteria and most archaeans it is unclear in which prokaryotic domain this gene originated.The class III sirtuin gene may have been prent prior to the divergence of the two prokaryotic domains,otherwi the class III gene might have traverd the boundary between Archaea and Bacteria via lateral gene transfer.It ems incon-gruous that prokaryotes,Candida albicans,and mam-mals have a class III gene while some eukaryotes (e.g.,S.cerevisiae,C.elegans,lanogaster )lack a class III gene.Current theory on the origin of eu-karyotes has modified the endosymbiont hypothesis of Margulis (19)by postulating the fusion of a bacterium with an archaean (20–22).A preliminary model for the evolution and distribution of eukaryotic sirtuins can be propod (Fig.4)in which the first eukaryote may have acquired sirtuin gen
es from each of its prokaryotic parents,the genes were of the types which are known to occur in lass U,class II,and class III.Protozoans and metazoans from diver phyla have class I and class IV sirtuins,so the types of sirtuins must have developed at an early stage of eu-karyotic evolution.Thus this model postulates that an early eukaryote possd all four types of sirtuins
and
FIG.3.The conrvation of specific quence motifs within the different class of sirtuins.Seven sirtuins from each class (U,I,II,III,and IV)are aligned and residues that are intraclass-conrved (in at least five of the ven quences)are shaded black.Numbers in parenthes indicate amino acid residues prent in variable regions that have been omitted to conden the size of the figure.
that later some eukaryotes lost one (e.g.,Drosophila and C.elegans which lack class III)or more (e.g.,S.cerevisiae which lacks class II,III,and IV)of the non-class I sirtuins.This model predicts that all four class of sirtuins are likely to be prent in some organisms (e.g.,echinoderms and chordates)that are phylogenetically similar to lifeforms that were ances-tral to vertebrates.
Much current rearch effort is aimed at determin-ing the specific molecular interactions and physiologi-cal functions of the sirtuins in both prokaryotic and eukaryotic organisms.To date the most extensively-studied sirtuin is the yeast Sir2protein itlf which is a class Ia sirtuin that appears to function via an inter-action with NAD and acetylated histones to cau mod-ification of chromatin structure (12–14).The physiolog-ical functions and molecular interactions associated with other class of sirtuins are still largely unknown.It is likely that particular class and subclass of sirtuin
s will be found to interact physiologically with particular types and subtypes of biomolecules that rve as sirtuin modulators and sirtuin substrates.Perhaps this classification of the sirtuins into specific quence-defined groups will provide a framework to help integrate rearch from diver biological systems on the molecular signaling pathways and physiological functions associated with the various class of Sir2-like proteins.
ACKNOWLEDGMENT
This work was supported by the Competitive Pilot Project Fund of the Veterans Affairs Stars and Stripes Healthcare Network.
REFERENCES
1.Gotta,M.,Strahl-Bolsinger,S.,Renauld,H.,Laroche,T.,Kennedy,B.K.,Grunstein,M.,and Gasr,S.M.(1997)EMBO J.16,3243–3255.
2.Straight,A.F.,Shou,W.,Dowd,G.J.,Turck,C.W.,Deshaies,R.J.,Johnson,A.D.,and Moazed,D.(1999)Cell 97,245–256.
3.Fritze,C.E.,Verschueren,K.,Strich,R.,and Easton Esposito,R.(1997)EMBO J.16,6495–6509.
4.Bryk,M.,Banerjee,M.,Murphy,M.,Knudn,K.E.,Garfinkel,D.J.,and Curcio,M.J.(1997)Genes Dev.11,255–269.
5.Smith,J.S.,and Boeke,J.D.(1997)Genes Dev.11,241–254.
6.Smith,J.S.,Brachmann,C.B.,Pillus,L.,and Boeke,J.D.(1998)Genetics 149,1205–1219.
7.San-Segundo,P.A.,and Roeder,G.S.(1999)Cell 97,313–324.8.Gottlieb,S.,and Esposito,R.E.(1989)Cell 56,771–776.
9.Kaeberlein,M.,McVey,M.,and Guarente,L.(1999)Genes Dev.13,2570–2580.
10.Brachmann,C.B.,Sherman,J.M.,Devine,S.E.,Cameron,
E.E.,Pillus,L.,and Boeke,J.D.(1995)Genes Dev.9,2888–2902.
11.Frye,R.A.(1999)Biochem.Biophys.Res.Commun.260,273–
279.
12.Tanny,J.C.,Dowd,G.J.,Huang,J.,Hilz,H.,and Moazed,D.
(1999)Cell 99,735–745.
13.Landry,J.,Sutton,A.,Tafrov,S.T.,Heller,R.C.,Stebbins,J.,
Pillus,L.,and Sternglanz,R.(2000)Proc.Natl.Acad.Sci.USA 97,5807–5811.
14.Imai,S.,Armstrong,C.M.,Kaeberlein,M.,and Guarente,L.
(2000)Nature 403,795–800.
15.Zemzoumi,K.,Sereno,D.,Francois,C.,Guilvard,E.,Lemesre,
alike是什么意思J.L.,and Ouaissi,A.(1998)Biol.Cell 90,239–245.
16.Afshar,G.,and Murnane,J.P.(1999)Gene 234,161–168.
17.Adachi,A.,and Hagawa,M.(1996)MOLPHY Version 2.3:
Programs for Molecular Phylogenetics Bad on Maximum Like-lihood ,Computer Science Monographs,No.28(Inst.Statistical Mathematics,Tokyo).the reader
18.Adams,M.D.,Celniker,S.E.,Holt,R.A.,Evans,C.A.,Gocayne,
J.D.,Amanatides,P.G.,Scherer,S.E.,Li,P.W.,Hoskins,R.A.,Galle,R.F.,et al.(2000)Science 287,2185–2195.
19.Margulis,L.(1970)Origin of Eukaryotic Cells Yale Univ.Press,
New Haven,CT.
20.Gray,M.W.,and Spencer,D.F.(1996)Symp.Soc.Gen.Micro-biol.54,109–126.
21.Brown,J.R.,and Doolittle,W.F.(1995)Proc.Natl.Acad.Sci.
USA 92,2441–2445.
22.Martin,W.,and Muller,M.(1998)Nature 392,
37–41.
FIG.4.Hypothetical model for the evolution and distribution of the four eukaryotic sirtuin class.Some yeasts have only class I (e.g.,S.cerevisiae )and some have both class I and class III (e.g.,C.albicans ).
