The Chloroplast Genome Sequence of the Green Alga Pseudendoclonium akinetum

50 Pages·2005·0.48 MB·English

The Chloroplast Genome Sequence of the Green Alga Pseudendoclonium
akinetum (Ulvophyceae) Reveals Unusual Structural Features and
New Insights into the Branching Order of Chlorophyte Lineages
Jean-Franxcois Pombert, Christian Otis, Claude Lemieux, and Monique Turmel
De´partement debiochimieet demicrobiologie, Universite´ Laval, Que´becG1K7P4,Canada
Onemajorlineageofgreenplants,theChlorophyta,isrepresentedbythegreenalgalclassesPrasinophyceae,Ulvophyceae,
Trebouxiophyceae,andChlorophyceae.ThePrasinophyceaeoccupiesthemostbasalpositionintheChlorophyta,butthe
branchingorderoftheUlvophyceae,Trebouxiophyceae,andChlorophyceaeremainsunresolved.Thechloroplastgenome
sequences currently available for representatives of three chlorophyte classes have revealed that this genome is highly
plastic,withChlamydomonas(Chlorophyceae)andChlorella(Trebouxiophyceae)showingfewerancestralfeaturesthan
Nephroselmis(Prasinophyceae).Wereportthe195,867-bpchloroplast DNA(cpDNA) sequenceof Pseudendoclonium D
o
akinetum(Ulvophyceae),amemberoftheclassthathasnotbeenpreviouslyexaminedfordetailedcpDNAanalysis.This w
n
genomesharescommonevolutionarytrendswithitsChlorellaandChlamydomonashomologs.Thegenecontent,number lo
a
ofancestralgeneclusters,andabundanceofshortdispersedrepeatsinPseudendocloniumcpDNAareintermediatebe- d
e
tweenthoseobservedforChlorellaandChlamydomonascpDNAs.AlthoughPseudendocloniumcpDNAfeaturesalarge d
invertedrepeat,itsquadripartitestructureisunusualindisplayinganrRNAoperontranscribedtowardthelargesingle-copy fro
m
(LSC)regionandasmallsingle-copyregioncontaining14genesthatarenormallyfoundintheLSCregion.Twenty-seven h
group I introns lie in nine genes and fall within four subgroups (IA1, IA2, IA3, and IB); 19 encode putative homing ttp
eTnhdeohniugchlesaismeisl,aarintydo7bhsaevrveehdoammoolonggsthatei1d4enItAic1alainndse7rtiIoAn2siintetrsoinnsoatnhderthcheilroreonpchoydteedoernsdtroenputoclpehaysetesosurggagneesltlsetgheantommaensy. s://ac
a
intronsarosefromintragenomicproliferationofafewfoundingintronsinthelineageleadingtoPseudendoclonium.In- d
e
terestingly,oneintron(inatpA)andsomeofthedispersedrepeatsalsoresideinPseudendocloniummitochondria,pro- m
vidingstrongevidenceforinterorganellarlateraltransferofthesegeneticelements.Phylogeneticanalysesof58cpDNA- ic.o
encoded proteins and genes support the hypothesis that the Ulvophyceae is sister to the Trebouxiophyceae but cannot u
p
eliminate the hypothesis that the Ulvophyceae is sister to the Chlorophyceae. We favor the latter hypothesis because .c
o
it is strongly supported by phylogenetic analyses of gene order data and by independent structural evidence based on m
sharedgenelosses andrearrangement breakpointswithin ancestrally conserved geneclusters. /m
b
e
/a
Introduction rtic
le
The green algae are divided into the phyla Strepto- plast genome sequences are currently available in public -a
b
phyta and Chlorophyta. The Streptophyta (Bremer 1985) databases. The genomes of the prasinophyte Mesostigma stra
contains all land plants and the green algae belonging to (Lemieux, Otis, and Turmel 2000) and of the charophyte c
theclassCharophyceae(Graham,Cook,andBusse2000), Chaetosphaeridiumglobosum(Turmel,Otis,andLemieux t/22
whereas the Chlorophyta (Sluiman 1985) contains virtu- 2002)mostcloselyresembletheirlandplantcounterpartsin /9/1
ally all of the other green algae, i.e., the members of termsofoverallstructureandgeneorganization.Likemost 90
3
the classes Prasinophyceae, Ulvophyceae, Trebouxiophy- land plant cpDNAs, they feature a quadripartite structure /9
8
ceae,andChlorophyceae(LewisandMcCourt2004).The that is characterized by the presence of two copies of an 2
1
basal position of the Prasinophyceae in the Chlorophyta rRNA-encoding inverted repeat (IR) sequence separating 32
is generally well established, but the branching order of a small single-copy (SSC) and a large single-copy (LSC) b
y
the Ulvophyceae, Trebouxiophyceae, and Chlorophyceae region.IntheStreptophyta,eachofthesegenomicregions gu
(UTC) remains unresolved (Friedl and O’Kelly 2002; showsahighlyconservedgenecontent,andtherRNAop- es
Pombert et al. 2004). A third lineage at the base of the eron within the IR is transcribed toward the SSC region. t on
Streptophyta and Chlorophyta is possibly represented In contrast, the complete cpDNA sequences of the 18
byMesostigmaviride;however,somestudiessuggestthat prasinophyte Nephroselmis olivacea (Turmel, Otis, and N
o
this green alga traditionally classified within the Prasino- Lemieux 1999), the trebouxiophyte Chlorella vulgaris ve
m
phyceae represents a basal divergence within the Strepto- (Wakasugi et al. 1997), and the chlorophycean alga Chla- b
e
phyta (Bhattacharya et al. 1998; Marin and Melkonian mydomonasreinhardtii(Mauletal.2002)indicatethatthe r 2
1999; Karol et al. 2001). architecture of the chloroplast genome is very fluid in the 01
To improve our understanding of phylogenetic rela- Chlorophyta.NephroselmiscpDNAdisplaysancestralgene 8
tionships among green algae andalso to better understand clusters and the typical quadripartite structure observed in
howthechloroplastgenomehasevolvedinthisalgalgroup, the streptophytes, whereas Chlorella cpDNA lacks the IR
wehaveundertakenthecompletesequencingofchloroplast as well as several genes and has gained three group I in-
DNA(cpDNA)fromrepresentativesofvariouslineagesin trons. Chlamydomonas cpDNA is even more scrambled
theStreptophytaandChlorophyta.Fivegreenalgalchloro- initsstructureandgeneorganization.Althoughitfeatures
an IR, the two single-copy regions are about equal in size
Keywords:greenalgae,Ulvophyceae,Pseudendocloniumakinetum,
and each includes genes usually present in the SSC and
chloroplastgenomeevolution,groupIintrons,repeatedsequences.
LSCregions.ChlamydomonascpDNAcarriesfewergenes
E-mail:[email protected].
but more introns than its Chlorella homolog; moreover, it
Mol.Biol.Evol.22(9):1903–1918.2005 harbors fragmented ancestral operons, and the coding re-
doi:10.1093/molbev/msi182
AdvanceAccesspublicationJune1,2005 gionsofsomegenesaregreatlyexpandedorbrokenrelative
(cid:1)TheAuthor2005.PublishedbyOxfordUniversityPressonbehalfof
theSocietyforMolecularBiologyandEvolution.Allrightsreserved.
Forpermissions,pleasee-mail:[email protected]
1904 Pombertetal.
to the corresponding genes in other completely sequenced sertedatidenticalpositionswithinthesamegenewereiden-
cpDNAs. Many repeated sequence elements (Maul et al. tified by manual screening of the GOBASE database
2002) are dispersed throughout both the Chlamydomonas (O’Brien et al. 2003).
and Chlorella genomes; however, such repeats have not
beenidentifiedinNephroselmis,Mesostigma,andChaetos- Phylogenetic Analyses
phaeridium cpDNAs.
The 18S rDNA sequence alignment of Friedl and
Here we report the chloroplast genome sequence
O’Kelly (2002) was retrieved from TreeBASE (http://
of Pseudendoclonium akinetum, a unicellular green alga
www.treebase.org/treebase/), modified to include the
thought to belong to a deep-branching lineage within the
P. akinetumsequence,andanalyzedas described bythese
Ulvophyceae (Floyd and O’Kelly 1990). This genome
authors. To obtain the chloroplast amino acid data set, se-
hasauniquearchitecture,althoughitsharessomefeatures
quencesofindividualchloroplastproteinswerealignedus-
with Chlorella cpDNA and others with Chlamydomonas
ing T-COFFEE 1.37 (Notredame, Higgins, and Heringa D
cpDNA. Based on our phylogenetic inferences from o
2000),theambiguouslyalignedregionsofthesealignments w
chloroplast sequences and gene order data as well as on n
were removed with GBLOCKS 0.91b (Castresana 2000), lo
independent structural evidence, we strongly favor the a
andthefilteredsequenceswereconcatenated.Thenucleo- d
hsiyspteorthgersoiusptsh.at the Ulvophyceae and Chlorophyceae are tiinddeivdidatuaalspetrowteaisn-gceondeinragtegdenbeysoanligthneinbgastihseofsecqourreenscpeosnod-f ed from
ing protein alignments, filtering the alignments with h
MDNaAterIisaollsataionndaMndetSheoqdusencing GanBdLsOuCbsKeSqu0en.9t1lyb,recmonocvaitnegnatthinirgdthceodfiolnterpeodsistieoqnusenwceitsh, ttps://a
c
Pseudendoclonium akinetum (Tupa 1974) was ob- PAUP* 4.0b10 (Swofford 2002). Maximum likelihood a
d
tainedfrom theUniversity ofTexas Algal Culture Collec- (ML)treesinferredfromaminoacidsequenceswerecom- em
tion(UTEX1912)andgrowninmodifiedVolvoxmedium putedwithPHYML2.4.4(GuindonandGascuel2003)un- ic
.o
(McCracken, Nadakavukaren, and Cain 1980) under 12 h derthecpREV451Cmodel(Adachietal.2000),whereas u
p
light-dark cycles. An A1T–rich fraction containing ML trees inferred from nucleotide sequences were com- .c
o
m
cpDNA and mitochondrial DNA (mtDNA) was isolated puted with PAUP* 4.0b10 (Swofford 2002) under the
/m
and sequenced as described previously (Pombert et al. general time reversible 1 C 1 I model. Modeltest 3.6 b
e
2004).Thenuclear-encoded18SrRNAgenewasamplified (Posada and Crandall 1998) identified the latter model as /a
by polymerase chain reaction (PCR) from total cellular the one best fitting our nucleotide data. Bootstrap support rtic
le
DNA using primers NS1 (White et al. 1990) and 18L foreachnodewascalculatedusing100replicates.Thecon- -a
(Hamby et al. 1988) as described in Helms et al. (2001). fidencelimitsofalternativetreetopologieswereevaluated bs
SequenceswereeditedandassembledwithSEQUENCHER using the Shimodaira-Hasegawa test as implemented in tra
c
4.2.1 (Gene Codes, Ann Arbor, Mich.). CODEML3.14(Yang1997)andPAUP*4.0b10(Swofford t/2
2
2002). Support for the T1, T2, and T3 topologies by indi- /9
Sequence Analyses vidual proteins in the amino acid data set was estimated /19
0
with CODEML 3.14 using the MGENE 5 1 option. 3
Genes were identified by Blast homology searches /9
Intron sequences were aligned manually on the basis 8
(Altschul et al. 1990) against the nonredundant database 2
of secondary structure predictions, and regions judged to 13
of the National Center for Biotechnology Information 2
be ambiguously aligned were removed. Neighbor-Joining b
(NCBI) server (http://www.ncbi.nlm.nih.gov/BLAST/). y
analyses of the resulting data sets were carried out with g
Positions of open reading frames (ORFs) and protein- u
PAUP* 4.0b10 (Swofford 2002) using Hasegawa-Kishino- e
coding genes were determined using ORFFINDER at Yano 85 distances with 1,000 bootstrap replicates. st o
NCBIandvariousprograms oftheGCGWisconsin pack- n
Phylogenetic reconstructions from gene order data 1
age (version 10.2) (Accelrys, Burlington, Mass.), whereas 8
wereperformedwithGRAPPA2.0(Moretetal.2001)us- N
genescodingfortRNAswerelocalizedwithtRNAscan-SE o
ing the default parameters and the (cid:1)m (tighter circular v
1.23 (Lowe and Eddy 1997). Patterns of codon usage e
lower bound) option. Because this program requires that m
for protein-coding genes and ORFs were compared b
all genes analyzed be shared between the taxa examined, e
using the CORRESPOND and CODONPREFERENCE the genes present within copy B of the IR were excluded r 20
programs of the GCG package and the CAI program 1
from the data set to accommodate the lack of the IR in 8
of the EMBOSS package (version 2.6.0) (http://emboss.
Chlorella cpDNA.
sourceforge.net). Repeated sequences were identified with
PipMaker(Schwartz etal.2000) andREPuter 2.74(Kurtz
Results
et al. 2001). Repeats were sorted with REPEATFINDER
Phylogenetic Affiliation of P. akinetum
(Volfovsky, Haas, and Salzberg 2001), and the retrieved
classification was refined manually. Putative stem-loop Because it has not been previously demonstrated
structures and degenerated repeats were identified with thatP.akinetum(UTEX1912)belongstotheUlvophyceae,
PALINDROME and ETANDEM, respectively; these two wehaveascertainedthisaffiliationbyconductingphyloge-
programs are part of the EMBOSS package. Homologous netic analyses of nuclear-encoded 18S rDNA sequences
introns were identified by BlastN searches (Altschul et al. with ML, maximum parsimony, and distance methods
1990)againstthenonredundantdatabaseofNCBIusingan (Supplementary Fig. S1, Supplementary Material online).
E value threshold of 1 3 10(cid:1)6. Homologous introns in- Pseudendoclonium akinetum was found to clearly affiliate
ChloroplastGenomeofPseudendoclonium 1905
Table1
ComparedFeaturesofPseudendoclonium andOther Green Algal cpDNAs
Feature Mesostigma Nephroselmis Chlorella Pseudendoclonium Chlamydomonas
Size(bp)
Total 118,360 200,799 150,613 195,867 203,827
IR 6,057 46,137 —a 6,039 22,211
LSC 83,627 92,126 —a 140,914 81,307
SSC 22,619 16,399 —a 42,875 78,088
A1T(%) 69.9 57.9 68.4 68.5 65.5
Codingsequences(%)b 74.5 68.7 60.9 62.3 50.1
Genes(no.)c 137 128 112 105 94
Introns(no.)
GroupI 0 0 3 27 5
GroupII 0 0 0 0 2 D
o
w
a ChlorellacpDNAhasnoIR. nlo
b Conservedgenes,uniqueORFs,introns,andintronORFswereconsideredascodingsequences. a
d
c GenespresentintheIRwerecountedonlyonce.UniqueORFsandintronORFswerenottakenintoaccount. e
d
fro
with other members of the Ulvophyceae, being part of putative homing endonuclease (see table 4). In addition to m
a clade including Pseudendoclonium basiliense and five these intron ORFs, 12 free-standing ORFs larger than 100 http
other members of the Ulotrichales. The P. akinetum and codons are present in Pseudendoclonium cpDNA; only 2 s
P. basiliense 18S rDNA sequences differ at only seven (orf286 and orf521) show a codon usage not significantly ://a
c
a
positions in the data set analyzed. different from that observed for the protein-coding genes. d
e
Genes are less tightly packed in Pseudendoclonium m
ic
Genome Structure and Gene Partitioning cpDNA than in Mesostigma and Nephroselmis cpDNAs .o
u
(table 1).Theintergenicspacersinthisulvophytegenome p
Pseudendoclonium cpDNA (195,867 bp) contains .c
areupto3,215bpinsize,withanaverageof600bp.The o
two copies of an IR sequence that are separated from m
higher proportion of coding sequences found in Pseuden- /m
one another by LSC and SSC regions (table 1 and fig. 1). doclonium cpDNA relative to Chlorella cpDNA is ex- b
e
The Pseudendoclonium IR encodes only the rRNA op- plained by the greater number of introns in the former /a
egrroenen, pinlancotncptrDasNtAtostthhaetiInRcsluodfeaaldldpitrieovniaolugselynesse.qSuuernpcriesd- genome (table 1). rticle-a
tionwglayr,dththeePLsSeuCderengdioocnlornatihuemrtrhRaNnAtowopaerdrotnheisStSraCnsrcergiiboend. Expansion of Coding Regions bstra
c
Another unusual feature of Pseudendoclonium cpDNA Eight protein-coding genes in Pseudendoclonium t/2
2
concerns the pattern of gene partitioning. The SSC region cpDNAarenearlytwoorthreetimeslargerthantheirMes- /9
ofthisgenomefeatures14genesthatmaptotheLSCregion ostigmahomologs(table3).Forallofthesegenes,withthe /19
0
inMesostigma,Nephroselmis,andlandplantcpDNAs(fig. exceptionofcemA,expansionofcodingregionsappearsto 3
/9
1). However, the Pseudendoclonium LSC region contains bethemaincauseoftheirincreasedsizes.Becausethesizes 8
2
no genes that are usually found in the SSC region. of the corresponding Mesostigma genes match those of 1
3
2
theircounterpartsinstreptophytesandthecyanobacterium b
Gene Content and Gene Density Synechocystissp.PCC6803,weeliminatedthealternative y g
u
hypothesisthatshrinkageoftheMesostigmacodingregions e
Table 2 compares thegene contentof Pseudendoclo- is responsible for the observed size variations. The ftsH, st o
nium cpDNA with those of other green algal cpDNAs. It n
rpoB, rpoC1, and rpoC2 genes representthe most notable 1
can be seen that a large number of genes found in both 8
casesofgeneexpansion;thelevelsofexpansionnotedfor N
Mesostigma and Nephroselmis (including all ndh genes) these genes are comparable or greater than those for the ov
havebeenlost beforetheemergenceoftheUTC lineages. e
correspondinggenesinChlamydomonas(table3).Theex- m
With its 105 genes, the chloroplast gene repertoire of b
trasequencesaccountingfortheexpansionofthePseuden- e
Pseudendoclonium is intermediate in size between those doclonium ftsH and rpo genes map mainly to internal r 2
0
of Chlorella and Chlamydomonas (see table 1). Eight of 1
coding regions. In Chlamydomonas, rpoB, rpoC1, as well 8
the genes identified in Chlorella (chlB, chlL, chlN, cysA,
asrps2eachconsistoftwoseparateORFs,andeventhough
cysT, trnL(gag), trnS(gga), and trnT(ggu)) are absent
theseORFsareadjacent,nointronhasbeendetectedinthe
from Pseudendoclonium, whereas all of the genes found
noncoding sequence separating them. In Pseudendoclo-
in Pseudendoclonium except trnR(ccu) are present in
nium, each of these three genes occurs as a single ORF.
Chlorella. Fourteen of the genes identified in Pseudendo-
clonium(accD,chlI,infA,minD,psaI,psaM,rpl12,rpl19,
Introns
rpl32, ycf20, ycf62, trnL(caa), trnR(ccg), and trnR(ccu))
areabsentfromChlamydomonas,whereasallofthegenes The27intronsinPseudendocloniumcpDNAaccount
found inChlamydomonasexceptchlB, chlL, andchlN are for14.8%ofthegenomesizeandarepreferentiallylocated
present in Pseudendoclonium. ingenescodingforphotosyntheticproteins(table4).They
Atotalof19ofthe27intronsdisplayedbyPseuden- all belong to the group I family and fall within subgroups
doclonium cpDNA carry an internal ORF encoding a IA1,IA2,IA3,andIBaccordingtotheclassificationsystem
1906 Pombertetal.
D
o
w
n
lo
a
d
e
d
fro
m
h
ttp
s
://a
c
a
d
e
m
ic
.o
u
p
.c
o
IR IR m
/m
b
e
/a
rtic
le
-a
b
s
tra
c
t/2
2
/9
/1
9
0
3
/9
8
2
1
3
2
b
y
FIG.1.—GenemapofPseudendocloniumcpDNA.Genesoutsidethemaparetranscribedclockwise.ThetranscriptiondirectionoftherRNAoperons gu
isindicatedbyarrows.Thegenesshowninyellow,blue,andredmaptotheIR,LSC,andSSCregionsinMesostigmacpDNA,respectively.Genesand e
s
ORFsabsentfromMesostigmacpDNAareshowningray.GeneclusterssharedspecificallywithChlorellaaredenotedbythickbrackets,whereasthe t o
clusterssharedwithChlorellaandoneormoreadditionalchlorophytesarerepresentedbythinbrackets.Onlythefree-standingORFslargerthan100 n 1
codonsareshown.tRNAgenesareindicatedbytheone-letteraminoacidcodefollowedbytheanticodoninparentheses(Me,elongatormethionine;Mf, 8
initiatormethionine). N
o
v
e
m
of Michel and Westhof (1990). The subgroup IA1 is well that potentially codes for an endonuclease of the LAGLI- b
e
represented by the Pseudendoclonium introns, with 14 in- DADGfamily(fig.2E).FiveIA2intronsexhibitanORFin r 2
0
trons assigned to this category. L6thatcodesforaputativeGIY-YIGendonuclease;these 1
8
Structural and phylogenetic analyses of the IA1 and IA2 intron-encoded proteins can bedivided into two cate-
IA2 introns revealed close relationships among members gories based on sequence similarity (fig. 2D and E).
of the same subgroup. These introns are well conserved The IA3 and IB introns show more sequence diver-
in both primary sequences and secondary structures (fig. gencethantheirIA1andIA2counterparts.AsshowninSup-
2A, B, C, and D); moreover, those sharing similar ORFs plementaryFigureS2(SupplementaryMaterialonline)and
tendtoclustertogetherinphylogenetictreesinferredfrom table4,thetwoIA3introns(Pa.psbA.3andPa.rrl.1)differ
core sequences (compare fig.2Ewith fig. 2C and D). The greatlywithrespecttotheircorestructuresandencodeden-
seven IA1 intron-encoded H-N-H endonucleases are all donucleases.TheP3.1andP3.2pairingsarepresentinPa.
specified by an ORF in L5 (table 4) and can be assigned psbA.3butnotinPa.rrl.1,andPa.psbA.3encodesaputative
to three categories based on sequence similarity within GIY-YIGendonuclease,whereasPa.rrl.1encodesaputative
theregioncontainingtheH-N-Hmotif(fig.2E).Pa.psbA.1 endonuclease with a single LAGLIDADG motif (table 4).
differsfromtheotherIA1intronsincarryinganORFinL9 EachofthefourIBintronscontainsanORFinL8thatcodes
ChloroplastGenomeofPseudendoclonium 1907
Table2
Gene ContentinPseudendoclonium andOther GreenAlgal cpDNAs
Genea Mesostigma Nephroselmis Chlorella Pseudendoclonium Chlamydomonas
accD s d d d s
bioY d s s s s
chlB d d d s d
chlI d d d d s
chlL d d d s d
chlN d d d s d
cysA d d d s s
cysT d d d s s
ftsI d d s s s
ftsW d d s s s
infA d d d d s D
o
minD d d d d s w
ndhA d d s s s nlo
ndhB d d s s s ad
ndhC d d s s s ed
nnddhhDE dd dd ss ss ss from
ndhF d d s s s h
ndhG d d s s s ttp
nnddhhHI dd dd ss ss ss s://ac
ndhJ d s s s s ad
ndhK d d s s s em
odpB d s s s s ic
petN d d s s s .ou
ppssaaIM dd ds dd dd ss p.co
m
rne s d s s s /m
rnpB s d s s s b
rpl12 s d d d s e/a
rrppll1292 dd ds ds ds ss rticle
rpl32 d d d d s -a
rpl33 d s s s s bs
rps15 d s s s s tra
rspsrsA16 dd ss ss ss ss ct/22
yyccff2407 ds sd ds ds ss /9/19
0
ycf61 d s s s s 3
ycf62 d d d d s /98
ycf65 d s s s s 21
ycf66 d s s s s 32
ycf81 d d s s s b
y
trnA(ggc) d s s s s g
trnL(caa) d d d d s ue
ttrrnnLR((gccagg)) ds dd dd sd ss st on
trnR(ccu) s s s d s 18
trnS(cga) d d s s s N
trnS(gga) d d d s s ov
trnT(ggu) d d d s s em
trnV(gac) d s s s s b
e
a Afilled/opencircledenotesthepresence/absenceofagene.OnlythechloroplastgenesthataremissinginoneormorecpDNAs r 20
1
areindicated.Atotalof91genesaresharedbyallcomparedcpDNAs:atpA,B,E,F,H,I,ccsA,cemA,clpP,ftsH,petA,B,D,G,L,psaA, 8
B,C,J,psbA,B,C,D,E,F,H,I,J,K,L,M,N,T,Z,rbcL,rpl2,5,14,16,20,23,36,rpoA,B,C1,C2,rps2,3,4,7,8,9,11,12,14,18,19,
rrf,rrl,rrs,tufA,ycf1,3,4,12,trnA(ugc),C(gca),D(guc),E(uuc),F(gaa),G(gcc),G(ucc),H(gug),I(cau),I(gau),K(uuu),L(uaa),
L(uag),Me(cau),Mf(cau),N(guu),P(ugg),Q(uug),R(acg),R(ucu),S(gcu),S(uga),T(ugu),V(uac),W(cca),Y(gua).
foraputativeLAGLIDADGendonucleasewithoneortwo (Pa.atpA.1andPa.atp1.1)insertedatthesamegeneposition
copiesofthismotif(table4);however,theseORFsaredis- inthechloroplastandmitochondrialgenomesofPseuden-
tantlyrelatedinsequence. doclonium(fig.3).Inadditiontohighlysimilarprimaryse-
Our BlastN searches of the GenBank database for quences and secondary structures, these introns feature in
homologs of the Pseudendoclonium chloroplast introns L8 very similar ORFs encoding putative endonucleases
identified only introns from green algae and land plants with a double LAGLIDADG motif.
(Evaluethresholdof1310(cid:1)6).Themostsurprisingresult Table 5 reports the homologous group I introns of
of this analysis was the finding of homologous introns chloroplast origin that proved to be inserted at the same
1908 Pombertetal.
Table3 Table4
ComparedSizes of ExpandedGenesin Pseudendoclonium, GroupI Intronsin Pseudendoclonium cpDNA
Chlorella, andChlamydomonascpDNAs
ORF
Chlorella Pseudendoclonium Chlamydomonas
Size Size
Size Size Size Designation Subgroupa (bp) Locationb Typec (codons)
Gene (bp) Expansiona (bp) Expansiona (bp) Expansiona
Pa.atpA.1 IB 1,634 L8 LAGLIDADG(2) 302
cemA 801 1.6 909 1.8 1,503 3.0 Pa.atpA.2 IA1 344 — — —
ftsH 5,163 1.9 7,791 2.9 8,916 3.3 Pa.psaA.1 IB 1,520 L8 LAGLIDADG(2) 301
rpoA 837 0.9 1,734 1.8 2,213b 2.3 Pa.psaB.1 IB 1,368 L8 LAGLIDADG(2) 234
rpoB 3,906 1.2 6,537 2.0 4,967c 1.5 Pa.psaB.2 IB 1,488 L8 LAGLIDADG(1) 253
rpoC1 2,511 1.3 4,737 2.4 5,739c 2.9 Pa.psaB.3 IA1 1,273 — — —
rpoC2 4,689 1.3 10,389 2.8 9,423 2.6 Pa.psbA.1 IA1 1,068 L9 LAGLIDADG(1) 156
ycf1 2,460 2.0 2,505 2.0 5,988 4.9 Pa.psbA.2 IA1 1,120 L5 H-N-H 241 D
ycf62 1,515 1.6 1,803 1.9 —d — Pa.psbA.3 IA3 1,452 L3.2 GIY-YIG 294 ow
Pa.psbA.4 IA1 1,092 L5 H-N-H 217 n
a SizerelativetothecorrespondinggeneinMesostigma.Eachvaluewasob- Pa.psbA.5 IA2 1,216 L6 GIY-YIG 231 loa
tainedbydividingthesizeofthegreenalgalgenebythesizeoftheMesostigmagene. Pa.psbA.6 IA1 978 L5 H-N-H 222 de
thatabpTohrteioinndoicfatthedesripzoeAis-cdoedriinvgedsefrqoumenocueriusnmpuisbsliinsgheidnsaecqcueesnscioendaBtKa.0W00e5f5o4unads PPaa..ppssbbAA..78 IIAA11 13,30373 L—5 H—-N-H 2—54 d fro
m
asproensducilntTgohfetoaintshdeeiqcua5et#endacnisndigz3ee#irnrpocorlurtidinoetnrsostdhouefciitnnhtgeeraggeefnnraeicm. sepshacifetrmseuptaatriaotnin.gtheORFscorre- PPPaaa...pppsssbbbAAB...9110 IIIAAA112 139,1604645 L——6 ——GIY-YIG 2——23 https
d ycf62isabsentfromChlamydomonascpDNA. Pa.psbB.2 IA1 1,282 L5 H-N-H 105 ://a
c
Pa.psbB.3 IA1 1,230 L5 H-N-H 105 a
d
sites as their Pseudendoclonium counterparts. Five Pseu- Pa.psbC.1 IA2 887 — — — e
m
dendocloniumintronshaveknownhomologsinchlorophy- Pa.psbC.2 IA1 907 — — — ic
Pa.psbC.3 IA2 960 L6 GIY-YIG 212 .o
ceangreenalgae,whereashomologsofPa.rrl.1havebeen Pa.psbC.4 IA2 1,051 — — — up
observedinprasinophytes,trebouxiophytes,chlorophycean Pa.psbD.1 IA2 1,118 L6 GIY-YIG 214 .c
o
green algae, and the hornwort Anthoceros punctatus. The Pa.psbD.2 IA2 921 L6 GIY-YIG 143 m
threepsbAandrrlintronsinC.reinhardtiicpDNAexhibit Pa.rbcL.1 IA1 1,682 L5 H-N-H 246 /mb
Pa.rrl.1 IA3 816 L6 LAGLIDADG(1) 168 e
ahighdegreeofprimarysequenceandsecondarystructure /a
conservation with their Pseudendoclonium counterparts a IntronswereclassifiedaccordingtoMichelandWesthof(1990).Thesubca- rtic
(ESvuepnptlheemuennutasruyalFPi3g..1Sa2n,dSPu3p.p2lpemaireinntgasryfoMunadteirniaPlao.pnslbinAe.)3. tegorbiesLoffolIlBowinetdrobnysaconuulmdbneortrbeefeirdsetnotitfiheedlouonpamexbtiegnudoiunsglyt.hebase-pairedregion le-ab
are present in Cr.psbA.2. identicfieTdhebycotnhseernvuemdbmero.tiAfindatshhepdreendoitcetsedtheendabosneunccleeaosefiasngOivRenF,.withthenumber strac
ItisintriguingthatL9ofPa.psbA.1codesforaputative ofcopiesoftheLAGLIDADGmotifindicatedinparentheses. t/2
2
singleLAGLIDADGendonucleaseandthatnoendonucle- /9
ase motif has been assigned to the ORF of 102 codons /1
9
0
found within the same loop in the homologous Cr.psbA.1 The SDRs map to intergenic spacers and/or introns 3
/9
IA1 intron (Holloway, Deshpande, and Herrin 1999). Our and can be classified into four groups of repeat units (A, 8
2
analysis of the published Chlamydomonas cpDNA se- B,C,andD)onthebasisoftheirprimarysequences(table 1
3
quence mapping at this locus revealed the presence of 6). They occur frequently as palindromic sequences sepa- 2 b
an ORF of 46 codons encoding a LAGLIDADG motif ratedby7–8bp,andthestem-loopstructuresoftencontain y g
29 nt upstream of orf102; interestingly, the predicted morethanoneSDRunit.Asdeletionofpalindromicstruc- ue
s
proteinsequenceofthisORFexhibitshighsimilaritywith turesandotherrepeatedelementsisknowntooccurduring t o
the N-terminal domain of the Pa.psbA.1-encoded endonu- cloning in Escherichia coli, we confirmed by direct se- n 1
clease(fig.2E).Totestthepossibilitythatthereisaframe- quencing of PCR products the sizes and sequences of 8 N
shift in the sequence originally published by Holloway, thePseudendocloniumSDRsthatweidentifiedbyanalysis ov
e
Deshpande,andHerrin(1999),wesequencedaPCRprod- ofclonedfragments.Tooursurprise,wefoundthatnumer- m
b
uctcontainingtheCr.psbA.1intron;however,wefoundno ous SDRs in Pseudendoclonium cpDNA bear identical or e
nucleotide difference. We conclude that a nonsense muta- closesequencesimilaritytothosepresentinthemitochon- r 2
0
tionledtolossoftheLAGLIDADGmotifintheCr.psbA.1- drialgenomeofthisalga(Pombertetal.2004).Theseho- 18
encoded protein. mologousmitochondrialSDRsfallwithinthreeofthefour
above-mentionedgroupsofrepeatunits(A,B,andC)(table
Repeated Elements 6)andalsooccurasstem-loopstructurescomposedofone
or more repeat units. To our knowledge, the presence of
The Pseudendoclonium chloroplast genome contains
closely related SDRs in different organelles of the same
a large number of repeated elements (fig. 4). Two types
eukaryotic cell has not been previously reported.
of repeated sequences can be distinguished: short tandem
repeatsandshortdispersedrepeats(SDRs).Theshorttan-
Genome Organization
demrepeatsarefoundinthevicinityoftrnMe(cau)inare-
gion spanning about 2 kb. The units forming this repeat Our pairwise comparisons of Pseudendoclonium
regionare10–20bpinsize,richinA1T,anddegenerated cpDNAwithpreviouslysequencedgreenplantcpDNAsin-
in sequence. dicate that its gene organization is most similar to that of
ChloroplastGenomeofPseudendoclonium 1909
A B
D
o
w
n
lo
a
d
e
d
fro
m
h
ttp
s
://a
C D c
a
d
e
m
ic
.o
u
p
.c
o
m
/m
b
e
/a
rtic
le
-a
b
s
tra
c
t/2
2
/9
/1
9
E 03
/9
8
2
1
3
2
b
y
g
u
e
s
t o
n
1
8
N
o
v
e
m
b
e
r 2
0
1
8
FIG.2.—ComparativeanalysesofgroupIintronsandintronORFsinPseudendocloniumcpDNA.(AandB)Consensussecondarystructuremodels
ofintronsdisplayedaccordingtoBurkeetal.(1987):(A)IA1intronsand(B)IA2introns.Arrowsdenotethe5#splicesitesbetweenexonandintron
residues.Highlyconservedresidues(in11IA1and5IA2introns)andslightlylessconservedresidues(innineIA1andfourIA2introns)areshownin
uppercaseandlowercasecharacters,respectively;theotherresiduesarerepresentedbyfilledcircles.Conservedbasepairings(in11IA1and6IA2introns)
aredenotedbybars.Numbersinsidethevariableloopsindicatethesizevariationsoftheseloopsinthecomparedintrons.NotethattheP2pairingis
missinginoneIA1intron(Pa.psbA.7).(CandD)Neighbor-Joininganalysesofintroncoresequences:(C)IA1intronsand(D)IA2/IA3introns.The
Chlamydomonasreinhardtii(Cr)andChlamydomonaseugametos(Ce)intronshomologoustothePseudendoclonium(Pa)introns(seetable5)were
includedintheseanalyses.Nodesthatwereidentifiedinatleast80%ofthebootstrapreplicatesarelabeledwithasterisks.Theintronsdenotedin
green,red,blue,andorangeencodeputativehomingendonucleases;ineachpanel,thosedenotedbythesamecolorspecifycloselyrelatedproteins.
(E)AlignmentoftheendonucleaseregionscontainingtheH-N-H,LAGLIDADG,orGIY-YIGmotif.Theaminoacidresiduesmakingupeachmotifare
underlined.
1910 Pombertetal.
D
o
w
n
lo
a
d
e
d
fro
m
h
ttp
s
://a
c
a
d
e
m
ic
.o
u
p
.c
o
m
/m
b
e
/a
rtic
le
-a
b
s
tra
c
t/2
2
/9
/1
9
0
3
/9
8
2
1
3
2
b
y
g
u
e
s
t o
n
1
8
N
o
FIG.3.—ComparativeanalysesofthechloroplastPa.atpA.1andmitochondrialPa.atp1.1introns.(A)Consensussecondarystructuremodelsofthe v
twointronsdisplayedaccordingtoBurkeetal.(1987).Thearrowdenotesthe5#splicesitebetweentheexonandintronresidues.Identicalresiduesare em
showninuppercasecharacters,whereasdifferentresiduesarerepresentedbyfilledcircles.Conservedbasepairingsaredenotedbybars.Upperandlower be
numbersinsidethevariableloopsindicatesizevariationsinthePa.atpA.1andPa.atp1.1introns,respectively.(B)ClustalWalignmentoftheputative r 2
LAGLIDADGendonucleasesencodedbythetwointrons.TheLAGLIDADGmotifsareunderlined. 01
8
Chlorella.Thesetwogreenalgalgenomesshare16distinct cestrally conserved clusters, the Pseudendoclonium ge-
gene clusters, 3 of which (rps18-rpl20, psbK-ycf12-psaM, nome shares 9 of the 10 break points found in Chlorella
and trnE(uuc)-trnMf(cau)) have not been identified in any cpDNA and features 13 additional break points (fig. 5).
othersequencedchloroplastgenomes(fig.1).Asinglegene AlloftheseninebreakpointsarealsoobservedinChlamy-
cluster(trnC(gca)-trnT(ugu))issharedspecificallybetween domonascpDNA(fig.5).Ofthe33breakpointsexhibited
Pseudendoclonium and Chlamydomonas cpDNAs. bythischlamydomonadgenome,11arespecificallyshared
Twelveofthegeneclustersthatareconservedbetween with Pseudendoclonium, whereas a single one is specifi-
MesostigmaandNephroselmiscpDNAshavebeenbroken cally shared with Chlorella. The psbB-psbT-psbN-psbH
in Pseudendoclonium cpDNA (fig. 5). As revealed by our cluster displays only two break points that are unique
analysisoftherearrangementbreakpointswithinthesean- to Pseudendoclonium. Although the psbN gene of this
ChloroplastGenomeofPseudendoclonium 1911
Table 5 oftaxawasobservedonlyfortheUTCclade.Thealterna-
GroupI Introns atIdentical GeneLocations in tivetopologyshowingPseudendocloniumassistertoChla-
Pseudendoclonium cpDNA,Other Green Algal cpDNAs, mydomonas (T2) was recovered in 9% and 20% of the
andLand Plant cpDNAs bootstrap replicates in the ML analyses of the amino acid
HomologousIntron andnucleotidedatasets,respectively,whereasthetopology
placing Pseudendoclonium at a basal position in the UTC
Pseudendoclonium Accession
Intron GreenPlanta/IntronNumberb Number clade(T3)wasnotdetectedinanyofthe100bootstraprep-
licates.Theconfidencelevelsofthesealternativetopologies
Pa.psbA.1 Chlamydomonasreinhardtii(C)i1 BK000554
were assessed using the statistical test of Shimodaira-
Pa.psbA.3 C.reinhardtii(C)i2 BK000554
Pa.psbA.5 C.reinhardtii(C)i3 BK000554 Hasegawa.T3,butnotT2,provedtobesignificantlyworse
Pa.psbC.1 Chlamydomonaseugametos(C)i1 M90639 thanT1atP,0.05intheanalysesofboththeaminoacid
Pa.rbcL.1 Chlorogoniumcapillatum(C)i1 AB010236 and nucleotide data sets (fig. 7).
Chlorogoniumeuchlororum(C)i1 AB010224 D
SeparateMLanalysesoftheaminoaciddatasetwith o
Pa.rrl.1 Chlamydomonasagloeformis(C) L43351 w
Chlamydomonascallosa(C) L43501 CODEML revealed no strong disagreement among the nlo
Chlamydomonasiyengarii(C) L43354 phylogenetic signals provided by the individual proteins a
d
Chlamydomonasmexicana(C) L43538 (SupplementaryTableS1,SupplementaryMaterialonline). ed
CChhllaammyyddoommoonnaassnpievtaerlifisi((CC)) LL4423959308 TopologiesT1,T2,andT3werefoundtobesupportedby from
22,25,and11proteins,respectively,withRELLbootstrap
C.reinhardtii(C) BK000554 h
Carterialunzensis(C) L42986 valuesrangingfrom39%to93%.However,all58proteins, ttp
CHaaretmeraitaocoolicvciuersil(aCcu)stris(C) LL4439510501 wsuiftfihctihenetesxtcreenpgtitohntoofrerjpeocBt,ofnaeileodr btootphroovfitdheeaalstiegrnnaatlivoef s://ac
Pediastrumbiradiatum(C) L49156 a
Neochlorisaquatica(C) L49155 topologies at P , 0.05 in the Shimodaira-Hasegawa test. de
m
Scenedesmusobliquus(C) L43360 rpoB supported T1 and rejected T3 (but not T2) at P 5 ic
Trichosarcinamucosa(U) AY008341 0.02. Thus, it appears that general homoplasy throughout .o
u
CMholnooremllaastvixulsgpaerciises(TM)722(P) LN4C4_102041865 trheseodluattiaonsebteitswteheenmTo1stanlidkeTly2.explanation for the lack of p.com
MonomastixspeciesOKE-1(P) L49154
Phylogenetic relationships were also inferred from /m
Scherffeliadubia(P) L44126
b
Anthocerospunctatus(E) AF393576 geneorderdatausingtherelativepositionsofthe80chlo- e
/a
a Theletterinparenthesesindicatesthespecificchlorophyte/streptophytelin- roplast genes shared by Pseudendoclonium and the seven rtic
eTa,gTerceobmoupxriisoipnhgytcheeage;reCen,Calhglaolr/opplahnytcienadei;caatnedd.EP,,PErmasbirnyoopphhyycteaa.e;U,Ulvophyceae; gurseeednfoprlathnetssemaennatliyosneesd,raebcoovnest.rGucRtsApPhPyAlog2e.0n,ietshebypraosgsuramm- le-ab
s
b Theintronnumberisgivenonlywhenmorethanoneintronispresent. ingthatgenerearrangementsoccurbyinversions.Thebest tra
tree was found to display the T2 topology and to feature ct/2
atotallengthof240inversionsteps(fig.8).User-treeanal- 2
ulvophytehasretaineditslocationbetweenpsbTandpsbH, /9
ithasbeenrelocatedtotheDNAstrandencodingtheother ysesconstrainedtotheT1andT3topologiesyieldedtrees /1
9
with 6 and 10 extra steps, respectively. 0
psb genes (fig. 1). 3
Mapping of shared gene losses and rearrangement /9
8
break points located within ancestrally conserved gene 2
Phylogenetic Analyses 13
clusters on topologies T1, T2, and T3 also revealed that 2
b
The amino acid and nucleotide sequences derived these two sets of structural characters independently sup- y
g
fromthe58protein-codinggenes(seeSupplementaryTable port topology T2 as the most parsimonious scenario (fig. u
e
S1,SupplementaryMaterialonline)thataresharedbetween 7). A total of 36 rearrangement break points and 41 gene st o
the cpDNAs of Pseudendoclonium, Chaetosphaeridium, losses were mapped on T2 compared to 47 break points n
1
Chlamydomonas, Chlorella, Marchantia, Nephroselmis, and46genelossesonT1and46breakpointsand45gene 8
N
and Nicotiana were concatenated and analyzed with ML losses on T3. In contrast to T1 and T3, T2 shows several o
v
inference methods using the homologous sequences of rearrangement events and gene losses that are specifically e
m
Mesostigma as the outgroup. As expected, analyses of shared between sister taxa in the UTC lineage. b
e
both the amino acid (11,225 positions) and nucleotide r 2
0
(25,318positions)datasetsyieldedabesttree(T1)inwhich 1
Discussion 8
the chlorophytes and streptophytes form two distinct line-
Unusual Features of Pseudendoclonium cpDNA
ages (fig. 6). In the chlorophyte lineage, the prasinophyte
Nephroselmis occupies the most basal position, whereas Like the chloroplast genomes of the chlorophycean
thetrebouxiophyteChlorella,theulvophytePseudendoclo- green algae belonging to the genus Chlamydomonas,
nium, and chlorophycean alga Chlamydomonas form Pseudendoclonium cpDNA is unusual with respect to its
a clade in which Pseudendoclonium is sister to Chlorella. quadripartitestructure(fig.1).Fourteenofthegenesfound
TopologyT1accountedfor91%and80%ofthebootstrap in the SSC region of this ulvophyte genome map to a dif-
replicatesintheanalysesoftheaminoacidandnucleotide ferent genomic region (most often the LSC region) in all
data sets, respectively, and was also found to be the most streptophytesfeaturinganIRaswellasintheprasinophytes
highlysupportedtopologyindistance(MLandLogDetdis- NephroselmisandMesostigma.Moreover,thePseudendo-
tances) and maximum parsimony analyses (data not cloniumIRfeaturesanrRNAoperonthatistranscribedto-
shown). In all analyses, instability in the branching order wardtheLSCregion.Duringtheevolution ofulvophytes,
1912 Pombertetal.
D
o
w
n
lo
a
d
e
d
fro
m
h
ttp
s
://a
c
a
d
e
m
ic
.o
u
p
.c
o
m
/m
b
e
/a
rtic
le
-a
b
s
tra
c
t/2
2
/9
/1
9
0
3
/9
8
2
1
3
2
b
y
g
u
e
s
t o
n
1
8
N
o
v
e
m
b
e
r 2
0
1
8
FIG.4.—PipMakeranalysisofPseudendocloniumcpDNA.Thisgenomesequencewasalignedagainstitself.Atthetopofthealignment,genesand
theirpolaritiesaredenotedbyhorizontalarrows,andcodingsequencesarerepresentedbyfilledboxes.TheSDRrepeatunitsdescribedintable6are
denotedbyboxedletters.TheshorttandemrepeatsneartrnMe(cau)areshownasgraydots.Similaritiesbetweenalignedregionsareshownasaverage
percentidentitybetween50%and100%identity.