-
ANRV342-PP59-20 ARI 26 March 2008 20:26
Plastid EvolutionSven B. Gould, Ross F. Waller,and Geoffrey I.
McFaddenSchool of Botany, University of Melbourne, Parkville
VIC-3010, Australia;email: [email protected],
[email protected], [email protected]
Annu. Rev. Plant Biol. 2008. 59:491–517
The Annual Review of Plant Biology is online
atplant.annualreviews.org
This article’s doi:10.1146/annurev.arplant.59.032607.092915
Copyright c© 2008 by Annual Reviews.All rights reserved
1543-5008/08/0602-0491$20.00
Key Wordssecondary/tertiary endosymbiosis, complex plastids,
proteintargeting, genome evolution, intracellular gene transfer,
plastidbiochemistry
AbstractThe ancestors of modern cyanobacteria invented
O2-generatingphotosynthesis some 3.6 billion years ago. The
conversion of wa-ter and CO2 into energy-rich sugars and O2 slowly
transformedthe planet, eventually creating the biosphere as we know
it today.Eukaryotes didn’t invent photosynthesis; they co-opted it
fromprokaryotes by engulfing and stably integrating a
photoautotrophicprokaryote in a process known as primary
endosymbiosis. After ap-proximately a billion of years of
coevolution, the eukaryotic hostand its endosymbiont have achieved
an extraordinary level of inte-gration and have spawned a
bewildering array of primary producersthat now underpin life on
land and in the water. No partnership hasbeen more important to
life on earth. Secondary endosymbioses havecreated additional
autotrophic eukaryotic lineages that include keyorganisms in the
marine environment. Some of these organisms havesubsequently
reverted to heterotrophic lifestyles, becoming signifi-cant
pathogens, microscopic predators, and consumers. We reviewthe
origins, integration, and functions of the different plastid
typeswith special emphasis on their biochemical abilities, transfer
of genesto the host, and the back supply of proteins to the
endosymbiont.
491
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
ContentsINTRODUCTION. . . . . . . . . . . . . . . . . 492FROM
FREEDOM TO SLAVERY:
OUTLININGENDOSYMBIOTIC STEPS . . . . . 492Primary Endosymbiosis
. . . . . . . . . . . 493Eukaryotic Endosymbiosis . . . . . . . .
495Nature’s Playground:
The Evolution Continues . . . . . . 498PREPROTEIN TARGETING . .
. . . . 501
Targeting to Primary Plastids . . . . . 501Targeting Into and
Within
Secondary Plastids . . . . . . . . . . . . . 503BIOCHEMICAL
PATHWAYS . . . . . . 505
Starch Synthesis . . . . . . . . . . . . . . . . . .
505Isopentenyl Diphosphate
(Isoprenoid Precursor)Synthesis . . . . . . . . . . . . . . . .
. . . . . . 506
Heme Synthesis . . . . . . . . . . . . . . . . . . 507Aromatic
Amino Acid Synthesis . . . 508Fe-S Clusters . . . . . . . . . . . .
. . . . . . . . 508
INTRODUCTIONIn nature the counterpart of chaos is not cos-mos,
but evolution. The spark of life was ini-tially a chemical one,
leading to the synthesisof the first molecules. Some of these
persistedand evolved in a precellular period, perhapssimilar to
that described in the model of theRNA world, leading to the first
prokaryoticlife approximately 3.5 to 4 billion years ago(48, 74).
The invention of oxygenic photosyn-thesis by prokaryotic
cyanobacteria approxi-mately 500 million years later was the
nextmajor achievement of biological evolution. Ithad a major impact
on the earth by enrichingthe atmosphere with O2 to a level that
trans-formed the geochemistry of the planet.
The first molecular carbon skeletons typ-ical of cyanobacteria
can be identified instrata from approximately 2.75 billion yearsago
(15). At the same time a novel mineralknown as hematite (Fe2O3),
which can formonly in the presence of a minimum
criticalconcentration of oxygen, began to appear.
These geological indices testify to an ever-increasing
concentration of atmospheric oxy-gen due to photosynthetic
activity. Photosyn-thesis was also the evolutionary trigger for
thesweeping diversification of O2-dependent life.Indeed, oxygen has
become critical for mostliving things, acting as an acceptor for
theelectrons released from carbon-carbon bondsthat were ultimately
created using energy cap-tured by photosynthesis. Thus, a byproduct
ofphotosynthesis (oxygen) became an essentialcomponent for the
burning of the sugars pro-duced by photosynthesis. The balance of
thebiosphere was born.
Nineteenth century microscopists (Sachs,Altmann, and Schimper)
recognized the semi-autonomous nature and bacterial-like
stainingproperties of chloroplasts (then known aschlorophyll
bodies) and mitochondria (thenknown as cell granules) (4, 106), but
it tookanother 15 years before Mereschkowsky syn-thesized these
observations into the theorythat chloroplasts are derived from
cyanobac-teria (81, 109). Margulis later formalized theTheory of
Endosymbiosis, which posits thatplastids and mitochondria of
eukaryotic cellsderive from bacterial endosymbionts (71).
FROM FREEDOM TO SLAVERY:OUTLINING ENDOSYMBIOTICSTEPSAs far as we
know, all eukaryotes have mi-tochondria (or modified, anaerobic
forms ofmitochondria known as hydrogenosomes ormitosomes), and the
establishment of thispartnership is generally regarded as inte-gral
to the origin of eukaryotes (123). Theacquisition of plastids by
eukaryotes oc-curred later, after the establishment of a di-versity
of heterotrophic eukaryotic lineages,one of which adopted a
cyanobacterium-like endosymbiont to acquire photosynthe-sis and
become autotrophic. We refer to aninitial plastid-creating
endosymbiosis as theprimary endosymbiosis. Secondary (or
eu-karyotic) endosymbiosis refers to subsequentendosymbiotic events
in which the progeny
492 Gould ·Waller · McFadden
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
of the primary endosymbiotic partnershipbecome endosymbionts
within other het-erotrophic eukaryotes, thus transferring
thecaptured cyanobacterial symbiont laterallyamong eukaryotes.
Subsequently, the progenyof these secondary endosymbiotic
partner-ships have become endosymbionts in othereukaryotes,
creating tertiary endosymbioses,to weave an extraordinarily complex
set ofendosymbiotic relationships of cells withincells within cells
within cells (Figure 1). Inthis review we examine the cell biology
ofthese endosymbiotic events and examine howthe various
compartments and genomes ofthese extraordinary chimeras cooperate
as asingle cell, albeit one made up of parts frommultiple
individual cells.
Primary EndosymbiosisThe endosymbiotic integration of a
free-living, cyanobacterial-like prokaryote into aeukaryotic host
produced three major au-totrophic lineages: the glaucophytes,
thegreen algae (and their descendants, theplants), and the red
algae (2, 46) (Figure 1).Plastids in these primary endosymbionts
arecharacterized by having two bounding mem-branes, which are
derived from the two mem-branes (plasma membrane and outer
mem-brane) of the Gram-negative cyanobacterium(17, 20). If a
phagocytotic membrane sur-rounded the symbiont when it was first
inter-nalized by the host, it has disappeared with-out a trace
(20). The main lines of evidencesupporting homology between the
outer en-velope membrane and the outer membraneof a cyanobacterium
are (a) the presence ofgalactolipids (52), (b) the presence of
β-barrelproteins in both membranes (110), and (c) theoccurrence of
peptidoglycan (or rudiments ofpeptidoglycan synthesis machinery)
beneaththese membranes (117).
Phylogenetic analyses suggest that theglaucophytes were the
first primary endosym-biotic lineage to diverge, some 550 mya,and
that the red and green algae divergedlater (75, 82, 103). Plants,
which probably
diverged from their green algal ancestorsapproximately 400 to
475 mya (36), subse-quently conquered the terrestrial environ-ment,
paving the way for animals to followthem onto land. In accordance
with this se-quence, plastids in the glaucophytes (whichare
sometimes referred to as cyanelles but aredefinitely plastids) most
resemble theircyanobacterial ancestors in that they re-tain a
peptidoglycan, wall-like layer betweenthe inner and outer envelope
membranes(57). Additionally, the thylakoids inside theglaucophyte
plastid stroma are studded withphycobilisomes that are identical to
thoseof cyanobacteria, and the composition ofthe oxygen-evolving
enhancer complex isalso very similar to that of
free-livingcyanobacteria (117). Rhodophyte plastids alsouse
phycobilins in protein-based light har-vesting antenna
(phycobilisomes), but theirplastids have apparently lost the
peptido-glycan wall (31). The green algal/plantlineage plastids are
the most derived in the pri-mary endosymbiosis lineage.
Phycobilisomeswere replaced by chlorophyll b embeddedin thylakoid
membranes, and a rich panoplyof accessory pigments developed to
capturelight and protect the photosynthetic appara-tus from the
unfiltered terrestrial light (80).
Generally, primary plastids have under-gone major modification
during their tenurein the eukaryotic host; reduction of
theirgenome’s coding capacity is one of the moreconspicuous
attenuations. The genome ofthe cyanobacterium Anabaena sp. PCC
7120has 5366 protein-encoding genes, and othercyanobacteria possess
similar numbers ofgenes (53). In contrast, the most
gene-richplastid reported to date, that of the red algaPorphyra
purpurea, encodes a paltry 251 genes(99), and the plastids of the
parasitic plantEpifagus virginiana harbor a mere 42 genes(132).
Thus, most of the original genetic ma-terial of the endosymbiont
was clearly ei-ther lost or transferred to the host genomeduring
their coevolution. Selection likely fa-vored the initial loss of
genetic materialby the endosymbiont because it turned the
www.annualreviews.org • Plastid Evolution 493
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
Serial2o
Karlodinium DinophysisKryptoperidinium
Lepidodinium
Rhopalodiagibba
Cyanothece
Cryptophyta(4, PB, Ca/c)
3o
3o3o
Dinophyta(3, Ca/c)
Perkinsidae(4)
Apicomplexa(4) Heterokontophyta
(4, Fcx, Ca/c)
Haptophyta(4, Fcx, Ca/c)
Euglenophyta(3, Ca/b)
Chlorarachniophyta(4, Ca/b)
Cyanobacterial ancestor(PB,Ca/b)
Glaucophyta(2, PB, Ca)
Oxyhirris Ciliata
Rhodophyta(2, PB, Ca/d)
Embryophyta(2, Ca/b)
2o2o
2o
1o
Paulinella chromatophora(2, PB, Ca)
Chlorophyta(2, Ca/b)
1o
1o
Peridinium Karenia
Figure 1A schematic representation of plastid evolution.
Engulfment of a cyanobacterial ancestor and subsequentreduction to
a primary plastid (1◦) by a eukaryotic host initially led to the
formation of three lineageswith primary plastids: the chlorophytes,
and land plants, rhodophytes and glaucophytes. The subsequentuptake
of a green or a red alga by independent hosts to form secondary
endosymbioses (2◦) resulted ineuglenophytes, chlorarachniophytes,
and the monophyletic chromalveolates. Chromalveolates,
whichrepresent the association of chromists (Heterokontophyta,
Haptophyta, and Cryptophyta) and theAlveolata (Apicomplexa,
Perkinsidae, Dinophyta, Ciliata), unite an extremely diverse array
of protists andnot all authors accept the grouping. Different
Dinophyta have replaced their original secondary plastidwith a
green alga either by serial secondary endosymbiosis (Lepidodinium)
or even tertiary endosymbioses(3◦); e.g., Karlodinium harbors a
tertiary plastid of haptophyte origin. The heterokontophyte
Rhopalodiagibba engulfed a cyanobacterial Cyanothece species and
reduced it to so-called spheroid bodies, which arenot used for
photosynthesis, but rather act in N2-fixation. The plastid
organelles were apparently lost inthe case of the ciliates and the
dinoflagellate Oxyhirris. A possible nascent primary endosymbiosis
(1◦) isrepresented by Paulinella chromatophora, although whether
this endosymbiont is a true plastid organelleremains uncertain. The
number of membranes surrounding the plastid and the photosynthetic
pigmentsis shown in parentheses. PB, phycobilin proteins; Fcx,
fucoxanthin; Ca/b/c/d, chlorophyll a/b/c/d.
494 Gould ·Waller · McFadden
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
prokaryote-eukaryote consortium into an ob-ligate symbiotic
relationship. However, wenow know that a concerted and ongoing
trans-fer of genes from endosymbiont to host hasradically depleted
the endosymbiont’s genecatalog. Much of this intracellular gene
trans-fer was likely achieved prior to the divergenceof the three
primary endosymbiotic lineagesbecause they share a similar residue
of com-mon genes (75).
This transfer of genetic material mandatedthe development of a
mechanism to returnthe gene product to the organelle. We dis-cuss
this problem in detail below, but somegeneral concepts can be
outlined now. Host-encoded proteins destined for the plastid
aretypically translated as precursor proteins bear-ing an
N-terminal topogenic signal that is rec-ognized by a proteinaceous
receptor, whichis either soluble in the cytoplasm or boundto the
outer plastid membrane. After recog-nition, the precursor is
subsequently translo-cated across the plastid envelope by a suiteof
translocation machineries spanning thetwo bounding membranes. The
preprotein ispulled into the plastid and the topogenic sig-nal is
proteolytically removed to yield the ma-ture protein.
The protein import mechanism probablyevolved early on in the
conversion of the en-dosymbiont into an organelle and no
doubtfacilitated the relocation of genes from theendosymbiont to
the host. Transferred geneswould need to acquire expression and
to-pogenic signals for the gene product to bereturned to the
organelle. Rhodophtye andgreen algal/plant
plastid-protein-targetingmachineries appear to be fairly similar
(79);although virtually nothing is known about thetranslocation
machinery in glaucophytes, wepredict that it is also similar
because theseplastids have also relinquished so many oftheir genes
to the host genome (75).
Eukaryotic EndosymbiosisThe current consensus of molecular
phy-logeny recognizes six eukaryotic super-
Nucleomorph: theformer nucleus ofthe eukaryoticendosymbiont;
lostin most secondaryalgae, but stillpresent in a highlyreduced
form incryptophytes andchlorarachniophytes
clusters: Opisthokonta, Amoebozoa, Plantae(Archaeplastida),
Chromalveolata, Rhizaria,and Excavata (2, 54). The Plantae
su-percluster embraces the three lines (glau-cophytes, rhodophytes,
and chlorophytes)with primary endosymbiotic (two membrane)plastids
and their monophyly is consis-tent with a common origin for their
plas-tids (103). However, plastids also occur inthe Chromalveolata,
Rhizaria, and Exca-vata, and all these multi-membrane plastidsare
derived from secondary endosymbioses(Figure 1). These events
created a varietyof eukaryotic-eukaryotic chimeras referred toas
meta-algae (22). Secondary or complexplastids are derived from
eukaryotic, primaryplastid-containing endosymbionts and
haveundergone reduction during their tenure inthe secondary host.
The degree of reductionvaries; sometimes it is relatively minor,
suchas in the partially integrated secondary en-dosymbionts of
Hatena (85), and sometimesit is extensive, such as in the case of
eu-glenoids in the Excavata where the only traceof the eukaryotic
endosymbiont is an extra(third) membrane around the plastid
(130).Two important intermediate stages in thesecondary
endosymbiont reduction processare represented by cryptophytes and
chlo-rarachniophytes, in which a very reduced en-dosymbiont
nucleus, cytoplasm, and plasmamembrane can still be identified. The
rem-nant nucleus, known as the nucleomorph,is located inside the
periplastidial compart-ment (the former endosymbiont’s cytosol),and
the overall topology allows us to rational-ize the presence of four
membranes aroundrelated plastids in chromalveolates, in whichthe
endosymbiont nucleus has completelydisappeared. Reduction forces
have obviouslybeen at work in these endosymbionts becausethe great
majority of the endosymbiont nu-clear genes have been transferred
to the hostnucleus and most of the cytoplasmatic fea-tures, other
than a small collection of ribo-somes, have been lost (27, 33).
Secondary endosymbioses introducedplastids into heterotrophic
lineages, and
www.annualreviews.org • Plastid Evolution 495
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
Chromalveolatehypothesis:monophyly of
thechromists(Cryptophyta,Haptophyta andHeterokontophyta)and
Alveolata(Dinophyta, Ciliataand Apicomplexa);also, the
commonancestor contained acomplex plastidderived from a redalga
that is retainedin several of theselineages
much energy has been focused on establish-ing how many separate
times a eukaryoticsymbiont has been integrated into a pre-viously
nonphotosynthetic lineage. Theenvironmental and commercial
importanceof the lineages created, not to mention theimportance of
these events as drivers ofeukaryotic diversity, make this a
particularlyfascinating question. The antiquity of theseevents and
the reduction processes thathave occurred in the ensuing millennia
alsomake the question a difficult one to re-solve. The most
parsimonious hypothesis,put forward by Cavalier-Smith, invokesonly
two secondary endosymbioses: oneinvolving a green alga leading to
theCabozoa (which unites euglenophytes andchlorarachniophytes) and
one involving ared alga that created the Chromalveolata(which
unites cryptophytes, haptophytes,heterokontophytes,
dinoflagellates, perkin-sids, apicomplexa, and the
plastid-lackingciliates) (22). Interestingly, no examplesof a
glaucophyte secondary endosymbiontare known. Various lines of
evidence now
refute the Cabozoa hypothesis (7, 10, 33, 66,104) and it is now
clear that two separateacquisitions of green algal
endosymbiontscreated the euglenophytes and chlorarach-niophytes
independently. The veracity ofthe chromalveolate hypothesis
remainsuncertain, and whether or not the chrom-alveolates are
derived from a single or multi-ple secondary endosymbioses of
separate redalgal endosymbionts is still much debated.The
chromalveolate hypothesis finds somesupport from molecular
phylogenies (44, 86),and some unusual recruitments of enzymes tothe
endosymbiont shared by chromalveolatesalso lend credence to a
single secondaryendosymbiotic event (9, 30, 43, 44, 88). Itwas
argued early on that the mechanism ofhow proteins are targeted from
the host tothe complex plastid would give insight intothe
endosymbiont’s ancestry (22), and recentinsights into this process
(see Figure 2 andbelow) are congruent with the chromalveo-late
scenario. Drawn together, these differentanalyses support the idea
of a monophyleticorigin for chromalveolates from a single red
−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−→
Figure 2Models of the machineries that import nuclear-encoded
plastid proteins for select primary and secondaryplastids.
Nuclear-encoded factors are brown, plastid-encoded factors are
green, and nucleomorph-encoded factors are gray. Organisms with
primary plastids (green algae, plants, and rhodophytes) sharecore
components of the import apparatus, although land plants apparently
have a more elaborate set ofreceptors (Toc159 and Toc64) in the
outer envelope membrane (OEM) and other participating factors(Tic55
and Tic40) in the inner envelope membrane (IEM). These factors,
together with Tic62, might beinvolved in redox-regulated import.
Oep16 imports protochlorophyllide oxidoreductase A from thecytosol
in Arabidopsis independently of the canonical Toc system. A more
complicated import pathway isnecessary for secondary plastids, as
in the case of the cryptophytes, which are surrounded by
additionalmembranes, namely the periplastidial membrane (PPM) and
rough endoplasmatic reticulum (rER). Incryptophytes preproteins are
cotranslationally inserted into the ER via the Sec61 complex and
the signalpeptide (SP) is cleaved by the lumenal signal peptide
peptidase (SiPP). The remaining transit peptide(TP) mediates
translocation across the remaining three membranes, before being
cleaved by the stromalprocessing peptidase (StPP) inside the
stroma, similar to primary plastids. Whether the secondary
plastidof P. falciparum and chlorarachniophytes is actually located
within part of the ER, as in cryptophytes andother chromists, is
uncertain. Morphology obviously has a significant impact on the
actual importpathway and machinery necessary. Proposed models for
complex plastids are mostly inferred fromgenome data mining and
lack experimental proof. Tic20 and Der1-2 have not yet been
identified incryptophytes (question marks) but genes that encode
proteins believed to be targeted into the plastid arepresent in
other chromalveolates for which full genome sequence is available.
PPC, periplastidialcompartment; EPM, epiplastid membrane. Topogenic
signals for stromal targeting are displayed beneaththe organisms’
names. The F-motif, which is apparently critical for stromal
targeting, occurs in plastidswith red algal origin.
496 Gould ·Waller · McFadden
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
CYTOSOL
14-3
-3
To
c15
9
To
c34
Hsp 70
SR
P80
S
Sec
61S
iPP
Bip
Bip
Oep
16
Grp
ES
tPP
Dn
aJ
Tic
110
FN
R
Cpn
60
Hsp 90
Tic
22
To
c64 Tic 40
Tic 62
Tic 55
Tic
20 Clp
C
To
c75
Gro
ELC
lpC
Grp
ES
tPP
Dn
aJ
Tic
22
14-3
-3
Tic
20
To
c75
SR
PR
e
SR
P
Ufd
1
Der
1-1
Tic
22
14-3
-3
Hrd
1U
bc
Ub
a1
Tic
22
80S
Sec
61
Gro
EL
Clp
C
Grp
ES
tPP
Dn
aJ
Tic
20 ?
Der
1-2
?
SiP
P
Cd
c48
Cpn
60C
lpC
Cpn
60
Tic
20
To
c75
Cpn
60
SR
P
80S
Sec
61S
iPP
StP
PD
naJ
Cpn
60
Tic
22
Ufd
1
Der
1-1 U
ba1
Clp
C
Ub
c
Der
1-2
Tic
20
Bip
Clp
C
Hsp 70
Hsp 70
STROMA
ER
ER
co
ntin
uous
with
pla
stid
?
Ves
icle
med
iate
d?
Em
bry
op
hyt
a(A
rabi
dops
is th
alia
na)
Ap
ico
mp
lexa
(Pla
smod
ium
falc
ipar
um)
Rh
od
op
hyt
a(C
yani
dios
chyz
on m
erol
ae)
Cry
pto
ph
yta
(Gui
llard
ia th
eta)
Ch
lora
rach
nio
ph
yta
(Big
elow
iella
nat
ans)
ER
ER
PPC
PPC
rER
rER
rER
PP
M
OE
M
IEM
OE
M
IEM
OE
M
IEM
OE
M
IEM
rER
rER
EP
M
PP
M
ER
co
ntin
uous
with
ap
icop
last
?
PP
M
EP
M
‘F’
SP
SP
‘F’
TP
‘F’
SP
Ves
icle
med
iate
d?
TP
TP
TP
TP
Hsp 70
Hsp 70 To
c34
Tic 62
Tic
110
Tic
110
Tic 62
Hsp 70
Hsp 70
Hsp 70
SR
PR
eS
RP
Re
Cd
c48
www.annualreviews.org • Plastid Evolution 497
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
Eukaryoticendosymbiosis: allevents in which theengulfed
organismthat was reduced toan organelle was aeukaryote
Endosymbiontmetabolicreplacement:replacement of anexisting
host-cellmetabolic pathwaywith one acquiredwith theendosymbiont
alga endosymbiont. However, some analyseswith genes encoding
cytosolic host proteinsdo not support the chromalveolate
hypothesis(86, 120), suggesting that the spread of asingle red
algal endosymbiont among thechromalveolate branch may have occurred
bysubsequent tertiary endosymbioses. Further-more, the clustering
of genes representingRhizaria together with the Chromalveolatain a
recent report by Hackett and colleagues(40) reminds us that
definite proof for themonophyletic origin of chromalveolates hasnot
been found.
One further aspect of eukaryotic endosym-biosis is tertiary,
maybe even quaternary, andserial secondary endosymbiosis. Tertiary
en-dosymbiosis is the uptake of a secondaryendosymbiosis-derived
alga by a eukaryote,and serial secondary endosymbiosis is the
re-placement of an original complex plastid witha new, primary
endosymbiosis–derived alga.Select dinoflagellate algal lineages
representthe best-studied cases of these higher orderendosymbiotic
events, and independent casesare represented by the genera
Lepidodinium,Kryptoperidinium, Karlodinium, and Dinophysis(for
detailed description of these unusual di-noflagellates lineages see
References 39, 46,and 54) (Figure 1). In each case, the host
di-noflagellate previously contained a secondaryplastid, so these
new endosymbionts representorganelle replacements. The mechanisms
fororganelle reduction and integration are likelythe same for
secondary endosymbionts; how-ever, in cases of organelle
replacement eventransferred genes from the first plastid
cancontribute to the integration of these new re-cruits (49,
91).
Nature’s Playground:The Evolution ContinuesPlastid loss and
reversion to obligateheterotrophy. A fascinating but often
over-looked element of endosymbiotic theory con-cerns organelle
reduction and loss. In a sense,all endosymbiotic organelles are
products ofmassive reduction of the metabolic complex-
ity and capabilities of the ancestral free-livingsymbiont. But
there is a tendency to regardfunctional organelles as having
reached a sta-ble suite of core metabolic functions—in thecase of
plastids photosynthesis is consideredthe cornerstone of organellar
function (seebelow for summary of plastid biochemicalfunctions).
Despite this mindset, an extensivenumber of lineages have
independently ad-vanced their plastids a further rung on theladder
of reduction by losing their photo-synthetic capability (21).
Parasitic plants andapicomplexan parasites such as the
malariaparasites are notable examples; many otherprotists have also
lost the ability to per-form photosynthesis but retained their
fur-ther reduced plastids (e.g., the euglenid As-tasia and the
dinoflagellate Crypthecodinium).These nonphotosynthetic plastids
apparentlystill provide essential services to the hostcells—for
instance, fatty acid synthesis, iso-prenoid synthesis, and heme
synthesis in thecase of the malaria parasites (96). Most ofthese
additional plastid pathways have likelyreplaced equivalent host
cell pathways thatoccurred in the ancestral host cell prior
toplastid acquisition (endosymbiont metabolicreplacement). Why a
plastid pathway shouldreplace an existing host cell pathway is
un-clear, although it is quite conceivable thatchance has played a
role in the eliminationof any one of the duplicated pathways
afterendosymbiotic merger. In any case, fixationof the plastid copy
of any essential metabolicpathway would commit a cell to plastid
reten-tion even if photosynthesis was subsequentlyabandoned. What,
then, is the likelihood ofsuch a cell achieving complete plastid
loss?
To date, members of at least six ma-jor eukaryotic lineages may
have achievedoutright plastid loss: ciliates, the apicom-plexan
(e.g., Cryptosporidium), dinoflagellates(several lineages),
heterokontophytes (e.g.,oomycetes), trypanosomatids, and
crypto-phytes (Goniomonas). However, for severalof these lineages
such claims have inspiredlively debate. The case for plastid loss
inalveolates (i.e., ciliates, apicomplexans, and
498 Gould ·Waller · McFadden
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
dinoflagellates) largely hinges on acceptanceof the
chromalveolata hypothesis, whichunites alveolates with chromists
(heterokon-tophytes, cryptophytes, and haptophytes) andproposes
plastid origin in a common ancestor(22). If this hypothesis is
correct, ciliates andbasal lineages of apicomplexans
(Cryptosporid-ium and gregarines) and dinoflagellates
(e.g.,Oxyrrhis, Amoebophyra, Noctiluca) that all lackplastids must
have independently lost theseorganelles (1, 107, 121, 134).
Challenging thisscenario is the lack of strong phylogeneticevidence
for the monophyly of chromalveo-lates host cells (40, 86, 120).
Thus, an alter-native explanation for plastid occurrence
inalveolates is that apicomplexans and dinoflag-ellates
independently gained their plastids,and that ciliates and the basal
members of eachancestrally lacked a plastid. The dinoflagel-late
lineage, however, may represent an in-dependent case for plastid
loss, because sev-eral nonphotosynthetic groups are
apparentlyscattered throughout photosynthetic di-noflagellates
[according to small subunitrRNA phylogenies and plate tabulation
data(107)], implying several independent losses.However, loss of
photosynthesis may notalways imply plastid loss, and the
recentdemonstration that at least one such taxon(Crypthecodinium)
retains a nonphotosyntheticplastid indicates that plastid loss
should bemore closely examined in this group (108).
In contrast to the conspicuous photo-synthetic members of the
heterokontophytessuch as kelp and diatoms, many mem-bers (e.g.,
thraustochytrids or oomycetes) arenonphotosynthetic (23). Oomycetes
are wellknown plant pathogens, responsible for sig-nificant
historical events such as the Irishpotato famine. Most
nonphotosynthetic het-erokontophytes fall into basal clades (23),
andtherefore again the question of plastid losshinges on whether
heterokontophytes sharea common plastid with other major
lineages(i.e., other chromists, or indeed all chromalve-olates) or
whether a plastid was independentlyacquired within the
heterokontophyte radi-ation. Trypanosomatids are a
heterotrophic
group of parasites whose sister relationship tothe euglenids
(many of which are photosyn-thetic) has inspired suggestion that
these par-asites also once contained a plastid but havesince lost
it (41). However strong evidence forsecondary plastids as a recent
gain in the eu-glenoid lineages (66), along with the rebuttalof the
Cabozoa hypothesis (see above), under-mines the case for plastid
loss in trypanoso-matids.
Perhaps the strongest case for plastid lossoccurs in the
Cryptophyta. Most cryptophytesare photosynthetic, although some
have ap-parently lost photosynthesis but retain a relictplastid
(113). Conversely, Goniomonas is abasal heterotrophic cryptophyte
that appar-ently lacks a plastid (77). Recent phylogeniesbased on
molecular data have strongly identi-fied haptophytes as the sister
lineage to cryp-tophytes (40, 86), and a gene replacement
ofplastid-encoded rpl36 is uniquely shared bythese taxa, implying
they share a commonplastid (55, 101). Thus, reasonable
evidenceexists that the common ancestor of crypto-phytes and
haptophytes contained a plastid,and therefore Goniomonas has lost
its plastid.
Although further cases of plastid losswill likely be
substantiated as global phy-logenies develop better resolution, a
casefor plastid loss in cryptophytes at leastlooks well supported.
How then, is a eu-karyote able to reverse the
endosymbioticprocess—particularly, how can endosymbiontmetabolic
replacement be reversed? Two sce-narios are possible. One is that
the plastid islost relatively early in endosymbiont integra-tion,
before endosymbiont metabolic replace-ment occurs. Cavalier-Smith
(19) has sug-gested that this accounts for why plastidlesstaxa are
often basal to photosynthetic lineages;they represent the period
before a cell startsto rationalize its own biochemistry and relyon
elements of the symbiont’s biochemistry. Ifthe chromalveolate
hypothesis is correct, thenonly after the major lineages diverged
did theplastid become essential beyond photosynthe-sis, because
most of the major lineages haveplastidless basal members. A second
scenario
www.annualreviews.org • Plastid Evolution 499
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
is that through heterotrophy a cell can satisfyits requirement
for the macromolecules it hadcome to rely upon from the plastid.
Many sec-ondarily nonphotosynthetic groups are eitherpredators or
highly adapted parasites, with ac-cess to a rich supply of
macromolecules. Ifthis diet can satisfy the need for fatty
acids,isoprenoids, and heme, for instance, a cellmight be well on
the way to making its plastidobsolete.
In either case, if a plastid is lost from a eu-karyotic lineage,
would evidence of the plas-tid’s tenancy remain? The many hundreds
tothousands of nucleus-encoded plastid geneswould initially be
present, but without a func-tion these would likely degrade quite
rapidly.However, any plastid genes that had cometo fulfill an
alternative, nonplastid-localizedfunction would remain useful even
after plas-tid loss. Such genes, referred to as EGT(endosymbiont
gene transfers) have been esti-mated to represent between 10% of
nucleus-encoded genes derived from the plastid inglaucophytes and
50% in plants (73, 100).This amounts to ∼150 and ∼2250 genes in
to-tal in the two groups, respectively. Thus, lossof a plastid
could conceivably leave a conspic-uous footprint if these genes
remained usefulin the absence of the plastid. In the oomycetesone
such plastid-derived gene ( gnd ) has beenhailed as evidence for
the former plastid in thislineage (6). However, to form a
compellingcase for a former plastid we should antici-pate a large
collection of such genes whenthe annotation of oomycete genomes is
com-plete. Conversely, ciliates have revealed nosuch plastid
footprint; the complete sequenceof Tetrahymena thermophila was
recently inves-tigated for just such relict plastid genes (29).
Itwill be very interesting to undertake the sameanalysis for
Goniomonas in pursuit of a betterunderstanding of the process of
plastid loss.
Symbioses in progress. In addition to full-fledged plastid
organelles, numerous organ-isms demonstrate that endosymbiosis is a
con-tinual driving force in evolution. Here wepresent four
interesting cases of organisms at
various points of negotiation of these cellularmarriage
contracts.
Rhopalodia gibba is a diatom that hostsboth a secondary red
algal–derived plas-tid and a novel cyanobacterial symbiont(Figure
1) known as the spheroid body (32).Unlike cyanobacterial-derived
plastids, whichare typically photosynthetic, the spheroidbody has
apparently lost this ability. However,the spheroid body has
retained another corecyanobacterial function, the ability to fix
ni-trogen, which it performs for its diatom hostduring the day
(93). Spheroid bodies are in-herited vertically from one generation
to thenext, and their numbers are regulated in thehost cell, which
implies a high level of hostcontrol over the endosymbiont. As yet
no ev-idence for spheroid body genes in the diatomhost has been
found, so whether the final stageof organelle integration has been
achieved inthis case remains undetermined. However, thespheroid
bodies have clearly suffered geneloss, and they are most likely
incapable ofagain living outside the host (93). Nitrogenfixation is
another of the innovations specificto prokaryotes, so it is
noteworthy that en-dosymbiotic capture again has played a rolein
extending such fundamental capabilities toeukaryotes.
A further independent example of pri-mary endosymbiosis is seen
in the freshwa-ter amoeba Paulinella chromatophora, whichalso hosts
a cyanobacterium-like symbiont(56). In this case photosynthesis is
retained,and P. chromatophora accordingly has con-verted from
heterotrophy to autotrophy. ThePaulinella endosymbiont, referred to
as acyanelle, occurs in the cytoplasm without anyadditional
bounding membranes, and sym-biont numbers are strictly regulated,
againsuggesting a higher level of host-symbiontinteraction (133).
Attempts to culture thesymbiont separate from the host have thusfar
been unsuccessful, but analysis of thePaulinella cyanelle genome
reveals neither ob-vious gene loss nor transfer of genes to
thehost, so it appears that there has been rela-tively little
genetic response to this union so
500 Gould ·Waller · McFadden
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
far (133). The Paulinella symbionts are thusenigmatic; however,
these symbionts poten-tially represent a second case of primary
en-dosymbiosis enabling photosynthetic capture.Fortunately,
Paulinella chromatophora hassymbiont-lacking sister species (P.
indentataand P. ovalis), which offers the possibility
ofunderstanding how this organism can acquirea permanent
prokaryotic endosymbiont in aprocess reminiscent of the origin of
plastids.
Nascent secondary endosymbioses arealso in evidence. The
enigmatic flagellateHatena arenicola harbors a
quasi-permanentprasinophyte-like endosymbiont (related tothe genus
Nephroselmis), which exhibits sub-stantial structural modification
when withinthe host. Permanent integration of this pho-tosynthetic
symbiont is apparently pendingbecause division of the symbiont is
not yetcoordinated with that of the host (85). Never-theless, a
degree of integration has apparentlyoccurred, because host cell
division resultsin one daughter cell inheriting the endosym-biont
while the other daughter cell is left with-out a symbiont and
presumably sources a newsymbiont from the environment.
Perhaps one of the more startling cases ofendosymbiosis in
progress is the plastid theftperformed by the sea slug Elysia
chlorotica.This sea slug feeds upon algae and can sal-vage the
plastids from their diet of Vaucherialitorea and maintain them,
generating pho-tosynthate that can nourish the animal formany
months. The plastids, which are arrayedin specially generated
diverticulae of the sluggut to be exposed to incoming light,
remaintranscriptionally and translationally active forup to nine
months (83). Circumstantial evi-dence suggests that the plastids
even receiveproteins synthesized by the sea slug (42).
Ifsubstantiated, this would implicate horizontaltransfer of a gene
from the alga to the sea slug,which by one definition would make
this cap-tured plastid an animal organelle. However,these plastids
are unable to divide, and arenot passed on from one slug generation
to an-other, nor do they occur in the animal’s germline. In fact,
stolen plastids such as these could
probably never achieve permanent endosym-biont status because
many essential plastidgenes would have been left behind when
thealgal nucleus was digested. Sea slug plastidsare thus examples
of kleptoplastids: photosyn-thetic organelles stolen from another
organ-ism but not permanently acquired.
PREPROTEIN TARGETINGIntracellular gene relocation is dependent
onthe existence of a system to reimport thegene product back to the
compartment oforigin. Given the massive scale of plastid-to-nucleus
gene relocation, this system mustrecognize and sort a large number
of plastid-destined proteins from all other proteins syn-thesized
in the cytoplasm. Elements of thissystem, including proteins of the
import ap-paratus embedded in the plastid membranesand some
features of plastid precursor pro-teins, are shared throughout
phototrophiceukaryotes and reflect the common originof primary
plastids. Core elements of thesystem shared by red algae and green
al-gae/plants clearly arose early, prior to the di-versification of
the primary endosymbiont-containing lineage (79). In plants
additionalelements such as extra receptors and redox-sensing
components of plastid protein importclearly have arisen to optimize
this systemand also facilitate the biogenesis of differentplastid
types (e.g., amyloplasts, chromoplasts,and chloroplasts) (14).
Great insight has nowbeen achieved into the complex and
sophis-ticated plastid protein import machinery ofplants, the
details of which are reviewed else-where (14, 38, 115). Here, we
confine our-selves to a synopsis of the primary plastidimport
system and give more focus to theless well-understood machinery for
proteinimport into secondary plastids with multiplebounding
membranes.
Targeting to Primary PlastidsPrimary plastids contain three
distinct setsof membrane: the outer envelope membrane
www.annualreviews.org • Plastid Evolution 501
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
(OEM), the inner envelope membrane (IEM),and the thylakoid
membranes, which thuscreate three separate compartments
(inter-membrane space, stroma, and thylakoid lu-men). Proteins can
therefore be targeted tosix regions within plastids: three
membranesand three soluble compartments. Dedicatedtranslocation
machineries and peptide target-ing information within the
nuclear-encodedplastid proteins are used in concert to achievethese
targeting feats (Figure 2).
The majority of proteins is targeted tothe plastid
posttranslationally, facilitated by anN-terminal transit peptide
extension. Thispeptide leader can vary in length from
approx-imately 20 to 150 amino acids and no primarysequence
consensus or common secondarystructure has been identified in the
large col-lections of transit peptides known from dif-ferent
plants. General characteristics includehydrophobicity at the
extreme N terminus,enrichment of hydroxylated amino acids, anda
depletion in acidic residues that leads toa positive charge,
particularly toward the Nterminus (16, 89, 95). Some transit
peptidesare phosphorylated by an ATP-dependent cy-tosolic kinase,
which leads to an interactionwith a guidance complex; these
peptides arethen preferred for import (76). After translo-cation,
transit peptides are cleaved off by thestromal processing
peptidase, which belongsto the M16 family of metallopeptidases,
re-leasing the mature protein into the stroma(102).
Transport of the majority of preproteinsacross the two envelope
membranes is thejob of the Toc (translocator of the
outerchloroplast membrane) and Tic (transloca-tor of the inner
chloroplast membrane) ma-chineries. These two apparatuses
comprisemultiple soluble and membrane-bound pro-teins named for
their molecular masses (seeFigure 2). Toc75 is the main
translocationpore in the outer membrane (28) and togetherwith Toc
33/34 and 159 makes up the Toccore (111, 112). Other Toc components
ap-parently have subsidiary roles; for example,Toc64 is implicated
in plastid protein recog-
nition and delivery to the Toc pore. In plantssuch as
Arabidopsis multigene families encodedifferent (partially
redundant) isoforms of Toccomponents and differential isoform
expres-sion probably generates import complexes tai-lored to
particular plastid states in differenttissues (12, 51). Conversely,
the haploid mossPhyscomitrella patens appears to lack these
iso-forms, and thus is emerging as a superiormodel for gene
knockout studies of Toc/Ticfunction (47). Interestingly, in the
genomeof the red alga Cyanidioschyzon merolae, onlyToc34 and Toc75
have thus far been identi-fied (Figure 2), which might mean that
majorreceptor components of the outer membrane(e.g., Toc 159 and
64) of land plant plastidsare specific to this green lineage, and
that red-specific Toc receptors await discovery (79).
The core of the Tic complex includesTic20, Tic22, and Tic110
(Figure 2). Tic22is a soluble protein in the intermembranespace and
is thought to be the first Tic com-ponent to interact with incoming
precursors(59). Either or both of the two membrane pro-teins Tic20
and Tic110 could be involved inpore formation but details are
unclear (50, 60,122). Tic110 also interacts with the chaper-one
Hsp93 (ClpC) inside the stroma (3). Asimilar function is proposed
for Tic40, be-cause it possesses a conserved domain knownfrom Hsp70
cochaperones (24). Several re-ports suggest the other Tic
components ofland plants (Tic32, Tic55, and Tic62) mightbe involved
in the redox-regulated import ofpreproteins (13, 61). The presence
of ho-mologs for Tic20, Tic22, Tic62, and Tic110in the genome of
the red alga C. merolae, com-bined with the apparent absence of
Tic32 and55, might suggest that the former are essen-tial and the
latter dispensable for functionalplastid import (79), but
experimental confir-mation is needed. Finally, on the stromal
sideof the membrane chaperones such as GroELand CplC interact with
the Tic complex toreceive the imported proteins and fold them,after
cleavage of the transit peptide, to theirmature conformation (125).
However, someproteins require further targeting and the
502 Gould ·Waller · McFadden
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
thylakoid membranes contain no less thanthree independent sets
of protein transloca-tion machineries for this purpose: the sig-nal
recognition particle–dependent (Albino3)pathway, the Tat (twin
arginine translocon)pathway, and the Sec-dependent pathway.
Inaddition, spontaneous insertion of proteinsinto thylakoid
membranes is also known tooccur, thus offering at least four
alterna-tive routes into the membrane or lumen ofthylakoids
(38).
Alternative, Toc/Tic-independent routesto plastids are also
becoming apparent (60,97). For instance, the outer envelope
pro-tein 16 (Oep16)—a homolog of bacterialpreprotein and amino acid
transporters—serves as the translocase for one plastid pro-tein,
NADPH:protochlorophyllide oxdidore-ductase A (98). Another
noncanonical importpathway through the outer envelope mem-brane was
recently revealed with the identi-fication of nuclear-encoded
plastid proteinspossessing N-terminal signal peptides ratherthan
the standard plastid transit peptides (94,126). These plastid
proteins traverse the en-doplasmic reticulum (ER), and most
likelyalso traverse the Golgi apparatus where theyare glycosylated,
and are subsequently tar-geted to the outer envelope membrane of
theplastid (94, 126). The details of this alternateroute remain
mysterious.
Targeting Into and WithinSecondary PlastidsTranslocation of
precursor proteins to sec-ondary plastids must surmount
additionalobstacles in the form of extra bounding mem-branes, which
also creates additional compart-ments that have their own specific
proteomes.Three membranes surround dinoflagellateand euglenophyte
complex plastids, whereascryptophyte, heterokontophyte,
haptophytes,apicomplexan, and chlorarachniophyte plas-tids are
surrounded by four membranes (22).Independent origins of secondary
plastidshave resulted in distinct targeting solutionsto these
advanced trafficking needs; however,
Epiplastidmembrane (EPM):the outermostmembranesurrounding
complexplastids
remarkably, some unifying principals haveemerged. Virtually all
known complex plastidpreproteins encoded in the nucleus possess
anN-terminal topogenic signal composed of atleast two parts: a
signal peptide and a transitpeptide. The ubiquity of what appears
to bea canonical signal peptide in complex plastid-targeted
proteins is consistent with the out-ermost membrane being a
component of thehost cell’s endomembrane system, apparentlyderived
from the formative phagocytic event(22). The signal peptide
mediates cotransla-tional import into the ER lumen, where
signalpeptidase removes the signal peptide to exposethe transit
peptide, which is responsible fortargeting across the remaining
membranes.An unusual elaboration of this bipartite leaderoccurs in
the two cases of complex plastids sur-rounded by three membranes:
dinoflagellatesand euglenoids. Here an additional signal,
ahydrophobic membrane anchor, is embeddedin the transit peptide
region of most plastidproteins (84, 87, 118). Thus, insertion of
pre-proteins into the ER lumen in these taxa isapparently delayed;
plastid preproteins are an-chored to endomembranes until they are
de-livered to the plastid and the complete plastidimport is
enabled.
In heterokonts, haptophytes, and crypto-phytes the outer plastid
membrane is con-tinuous with the host rough ER and thusis studded
with ribosomes (18). Plastid pro-teins encoded in the nucleus have
thereforealready passed through the first of four mem-branes after
cotranslational insertion into theER lumen by the N-terminal signal
peptide(Figure 2). However, in other complex plas-tid systems no
such continuity of plastids andER is apparent, so plastid
preproteins mustbe delivered from the ER lumen to the outerplastid
membranes [termed epiplastid mem-branes (EPM) in Figure 2],
presumably byvesicular traffic (84, 118). After signal pep-tide
cleavage, plastid proteins must be dis-tinguished from secretory
proteins; mutage-nesis experiments in several complex
plastidsystems indicate that the transit peptide is re-sponsible
for this discrimination (34, 62, 129).
www.annualreviews.org • Plastid Evolution 503
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
Periplastidialcompartment(PPC): the reducedcytosol of
theengulfed alga;harbors thenucleomorph and80S ribosomes
incryptophytes andchlorarachniophytes
Periplastidialmembrane (PPM):represents theformer
cytoplasmicmembrane of theendosymbiont and isthe
secondoutermostmembranein four-membrane-bounded plastids
The signal peptide and transit peptide thusact sequentially to
mediate targeting into thestroma of complex plastids.
At least for some complex plastids, proteinsare also targeted
into the periplastidial com-partment (PPC); these proteins also
utilize abipartite leader. The most N-terminal residueof the
transit peptide is critical in dictatingwhether a protein travels
all the way throughthe three innermost plastid membranes to
thestroma or whether it stops in the PPC aftertraversing only the
periplastidial membrane(35, 116). In the cryptophyte Guillardia
thetaand the heterokontophyte Phaeodactylum tri-cornutum, this +1
transit peptide residue istypically a phenylalanine for stromal
proteins(in a few exceptions other aromatic aminoacids fulfill this
role). In the absence of thisphenylalanine the preprotein
accumulates inthe PPC (34, 37). In other chromalveolates anaromatic
amino acid–based motif (F-motif) isalso a conspicuous feature of
the N terminusof transit peptides, and likely plays a role
incorrect stromal targeting (35, 58, 90). Inter-estingly, this
F-motif also occurs in the transitpeptides of plastid-targeted
proteins of glau-cophytes and rhodophytes (117), which sug-gests
that the F-motif could be an ancient tar-geting element for plastid
import. Thus, incomplex plastids the role of this F-motif
hasapparently been extended to discriminate be-tween proteins that
are required to be targetedfully into the plastid stroma and those
thatmust be halted in the PPC. The corollary isthat the remainder
of the transit peptide is suf-ficient for targeting across the
periplastidialmembrane (PPM) (Figure 2). Curiously, theF-motif does
not occur in transit peptides ofthe green algal/plant lineage, and
thus thisancient targeting signal has apparently beenabandoned
here. This might explain the ap-parent need for extra receptors
like Toc159and 64 in these plastids (Figure 2).
The second step of protein trafficking intocomplex plastids (the
translocation from theER lumen into the PPC, see Figure 2)
pre-sumably requires a translocon in the PPM.Termed the Top
translocon (22), this hy-
pothetical membrane transporter and its in-triguing evolutionary
pedigree may recentlyhave been identified. The nucleomorph ofthe
cryptophyte G. theta encodes ERAD (ER-associated degradation)
components, includ-ing a Der1p (degradation in the ER) mem-brane
translocon able to complement ERAD-deficient yeast (116). Because
no ER is presentinside the periplastidial space, the location
ofthis nucleomorph-encoded ERAD machin-ery was intriguing.
Preliminary immunolo-calization studies suggest that the
Der1ptranslocon is located in the PPM of the cryp-tophyte’s complex
plastid (116), leading tospeculation that it could be the long
sought af-ter Top translocon. Further credence for thishypothesis
comes from the identification ofan extra set of ERAD machinery
(distinct fromthe canonical host ER ERAD machinery) thatis
apparently targeted to the complex plas-tids of other
chromalveolates such as diatomsand Plasmodium (116). Because these
plastidshave lost all traces of the endosymbiont cy-toplasm it is
highly plausible that this ERADmachinery could localize to the
periplastidialmembrane and have a role in translocatingtransit
peptide–bearing preproteins intothe complex plastids. The
ERAD-derivedtranslocon is proposed to recognize the transitpeptide
(which might resemble an unfoldedprotein similar to the normal ERAD
sub-strate) and to pull the precursor proteins outof the ER and
into the periplastidial compart-ment (Figure 2).
Although the role of this ERAD appara-tus in targeting proteins
to complex plastidsremains to be substantiated, it provides
tan-talizing support for the chromalveolate hy-pothesis. As alluded
to above, solving the“protein-import problem” was a major hurdlein
the establishment of secondary endosym-bionts (18). The apparent
co-option of a nor-mally ER-based protein translocation systeminto
plastid transport by cryptophytes, het-erokontophytes, and
Apicomplexa is sugges-tive of a common origin for the
plastids.Similarly, use of the F-motif to discriminatebetween
periplastidial and stromal-directed
504 Gould ·Waller · McFadden
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
proteins by cryptophytes, diatoms, and per-haps even apicomplexa
is also congruent witha common origin for their plastids.
BIOCHEMICAL PATHWAYSThe union of a heterotroph and an au-totroph
in an endosymbiotic partnershipamalgamates two suites of metabolic
path-ways into a single organism (131). The driverfor the union is
typically believed to be theacquisition of photosynthesis by the
host.Thus, both primary and secondary endosym-bioses likely
converted heterotrophs into pho-totrophs. Some serial secondary
endosym-bioses and tertiary endosymbioses may havesimply exchanged
one photoautotrophic en-dosymbiont for another, but in general
wecan frame the question in terms of het-erotroph + autotroph = new
autotroph.From a metabolic perspective this fusion cre-ates
interesting possibilities. Autotrophs aretypically self-sufficient
metabolically; somerequire vitamins, but in general they
synthe-size everything they need from scratch. Con-versely,
heterotrophs have access to a rangeof preformed macromolecules in
their dietand are able to salvage various building blocksfrom these
macromolecules and utilize themin their metabolism. Thus, as a
general princi-ple, the endosymbiosis likely introduced
extrametabolic capability beyond just photosyn-thesis to the host’s
repertoire. A key challengeis to unravel which pathways were
introducedinto the amalgam from the endosymbiont. Asdiscussed
above, the host can become depen-dent on endosymbiont pathways
other thanphotosynthesis, and this dependency can im-pact plastid
persistence should the organismsubsequently revert to a totally
heterotrophiclifestyle.
What do we know about the metabolicrepertoires of the original
hosts and endosym-bionts? For the hosts we can say very little.
Wehave a relatively poor understanding of thenature of the host for
the primary endosym-biosis, and, similarly, we are largely
ignorantof the host’s affinities for the three known sec-
ondary endosymbioses (euglenophytes, chlo-rarachniophytes, and
chromalveolates). It isthus rather difficult to speculate on what
kindof metabolisms these hosts could have had atthe outset of the
endosymbiotic relationship.However, we are in a better position to
hy-pothesize about the metabolic repertoire ofthe endosymbionts.
For primary endosymbio-sis we can postulate that the
endosymbionthad a suite of metabolic capabilities similar tothose
in modern-day cyanobacteria. For sec-ondary endosymbioses we can
assume that theendosymbionts had a metabolic potential sim-ilar to
that in the modern representatives ofred or green algae as
appropriate.
Weeden (131) was the first to ponder froma metabolic perspective
the consequences offusing an endosymbiont and host. He recog-nized
that the endosymbiont introduced novelpathways into the host and he
outlined howamino acid, heme, and starch pathways wereinducted into
hosts via endosymbiosis. Wehave subsequently learned that cells
have alsoexercised considerable creativity during thesemetabolic
mergers, and complex amalgams ofhost and symbiont pathways have
also beenthe fruits of these partnerships.
Starch SynthesisExcess photosynthate is generally stored
asglucan polymers. Plants and green algae storestarch (α-1,4
glucan) in the plastid, whereasred algae store starch in the
cytosol (127). Onthe basis of these localities of starch
synthe-sis, red algae were assumed to utilize a host-derived glucan
synthesis mechanism whereasthe green algae and plants were assumed
toemploy a system derived from the endosym-biotic ancestor of the
plastid (Figure 3). Thestarting points for each of these
pathways—UDP-glucose precursors for red algae andADP-glucose
precursors for green algae andplants—also reflect the dichotomy
betweeneukaryotic and prokaryotic glucan pathways.However, in
reality both host and symbiontproteins have been recruited in
starch syn-thesis in both red and green algae, and only
www.annualreviews.org • Plastid Evolution 505
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
UDP-G
Florideanstarch
ADP-G
GlycogenGlycogen
UDP-GADP-G
Starch
Rhodophyta Chlorophyta
Nuc
NucNuc
Figure 3Schematic representation of starch synthesis before and
after primaryendosymbiosis. Nuc, nucleus; UDP-G,
uridine-diphosphate glucose;ADP-G, adenosine-diphosphate
glucose.
the localities, either cytosolic or organellar,have been derived
from either host or sym-biont (87). Why red algae retained the site
ofhost glucan storage whereas green algae (andtheir descendants)
adopted the endosymbiontstorage site remains unknown (Figure
4).
In secondary endosymbiosis glucan stor-age distribution differs:
Sometimes it is in the
Nuc
MevPyruvate(DOXP)
IPPIPP
Nuc
Pyruvate(DOXP)
Mev
IPP IPP
Figure 4Schematic representation of the isopentenyl diphosphate
(IPP) synthesispathway before and after primary endosymbiosis. Mev,
mevalonate;DOXP, 1-deoxy-D-xylulose 5-phosphate.
host, and sometimes it is in the endosym-biont. For instance,
euglenoids store paramy-lon (β-1,3 glucan) in the cytosol,
althoughtheir endosymbiont is thought to have beena green alga that
presumably stored starchin the plastid (26, 128).
Chlorarachniophytesand heterokontophytes also store β-1,3 glu-cans
in the secondary host cytosol and haveabandoned the glucan storage
systems of thegreen and red algal endosymbionts, respec-tively
(78). Conversely, cryptophytes storestarch in the PPC (remnant
endosymbiont cy-toplasm), thus conserving the endosymbiontglycan
storage system of the red algal en-dosymbiont (34). Dinoflagellates
also storestarch; however, rather than in a PPC, storageoccurs in
the host cytoplasm, implying reloca-tion of this pathway from the
red algal cytosolto that of the host (25). Thus, the storage
ofsurplus photosynthate in either the host or en-dosymbiont
compartments has taken a rangeof alternatives in both primary and
secondaryendosymbiotic partnerships.
Isopentenyl Diphosphate (IsoprenoidPrecursor)
SynthesisIsopentenyl diphosphate (IPP) is a buildingblock for
terpenes, sterols, carotenoids, andisoprenoids that are important
components ofa diverse range of cellular molecules such
aschlorophylls and quinones. IPP synthesis wasonly recently
discovered to occur in plastids(68). In plants the cytosol harbors
the canoni-cal mevalonate pathway for IPP synthesis (8),and for
many years it was presumed that thiswas the sole source of
isoprenoid precursorsin plants. Given the extensive use of
isoprenesubunits in plants for secondary metabo-lites such as
terpenes, chlorophylls, ubiquinol,prenylated proteins, and
isopentyl tRNAs, itis sobering to reflect that a plastid-based,
non-mevalonate, deoxyxylulose (DOXP) pathwayfor IPP synthesis was
overlooked, or at leastunrecognized, by plant physiologists.
How-ever, once it emerged that bacteria synthe-size IPP from
pyruvate and glyceraldehyde3-phosphate and not from mevalonate
like
506 Gould ·Waller · McFadden
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
eukaryotes, it was a simple step to identifya DOXP pathway in
plastids (105). Indeed,the discovery of the plastid DOXP
pathwayreconciled some previously incongruous pre-cursor
incorporation and inhibitor data (67).Synthesis of IPP in plastids
also simplifiesthe delivery of these entities to
isopentenylatetRNAs for plastid translation and to isoprenechains
for chlorophyll production (63).
In plants, IPP synthesis occurs in boththe host compartment
(cytosol) and the en-dosymbiont compartment (plastid). The
twodifferent pathways coexist and are even inte-grated to an extent
(65), but their differencesare congruent with one
(acetate/mevalonate)being derived from the host and the other(DOXP)
being introduced with the cyanobac-terial symbiont (Figure 4).
Exactly why bothpathways persist is not known but the require-ment
for products in both the host and en-dosymbiont compartments
perhaps necessi-tated the retention of two pathways.
Heme SynthesisThe synthesis of the tetrapyrroles that actas
temporary electron carriers in various re-dox reactions is
reminiscent of IPP synthe-sis in that there are two very different
path-ways that begin with different substrates andutilize some, but
not all, different enzymes.Many plastid-lacking eukaryotes utilize
theso-called C4 or Shemin pathway, which com-mences by fusing
succinyl-CoA and glycineto create δ-aminolevulinic acid (ALA),
cour-tesy of aminolevulinic acid synthase (ALAS)(Figure 5). In
animals and yeast, ALAS (114)is located in the mitochondrion and
ALA isthen exported to the cytosol where a series ofsteps convert
it to coproporphinobilogen III(CPIII). CPIII is then routed back
into themitochondrion by a recently identified trans-porter for the
last four steps to eventually pro-duce heme (Figure 5).
Cyanobacteria have a different initial sub-strate, commencing
C-5–type heme synthe-sis from glutamyl-tRNA rather than
succinyl-CoA and glycine (5) (Figure 5). Glu-tRNA
Nuc
Glycine+
Suc-CoA
Nuc
Glu-tRNA
Heme
Chl
Heme
Heme
Heme
Glu-tRNA
Heme
Chl
Figure 5Schematic representation of heme synthesis before and
after primaryendosymbiosis. Nuc, nucleus; Glu tRNA, glutamyl tRNA;
Suc-CoA,succinyl-Coenzyme A; Chl, chlorophyll.
reductase followed by Glu-SA aminomutaseconvert the
aminoacylated tRNA to ALA.Steps from ALA to heme are then identical
incyanobacteria and the Shemin pathway, andthe enzymes involved are
homologous. How-ever, in cyanobacteria the pathway forks
atprotoporphinobilogen IX. One branch leadsto heme as per the
Shemin pathway, but theother branch involves the addition of
Mg2+
to protoporphinobilogen IX to generatechlorophyll (124).
Because the original host for primaryendosymbiosis likely had
aerobically respir-ing mitochondria, we can assume it had aShemin
pathway to generate heme for itscytochromes. The acquisition of a
cyanobac-terial endosymbiont almost certainly in-troduced the
glutamyl-tRNA–based path-way into the first eukaryotic autotrophs.
Atthe outset this organism would have hadtwo heme synthesis
pathways: a Sheminpathway in the mitochondrion/cytosol anda
cyanobacterial-like C5 pathway in theendosymbiont (Figure 5). To
carry onphotosynthesis, the endosymbiont likelycontinued to
synthesize chlorophyll fromglutamyl-tRNA; indeed, plant plastids
stillsynthesize chlorophyll entirely within the
www.annualreviews.org • Plastid Evolution 507
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
plastid using a pathway homologous to that ofcyanobacteria
(124). Interestingly, plants havedisposed of the early portions of
the Sheminpathway and do not use glycine or ALASto commence heme
synthesis (92). Rather,they export protoporphinobilogen IX fromthe
plastid to the mitochondrion, which thenperforms the last two steps
of heme synthe-sis using enzymes homologous to those ofthe
animal/yeast Shemin pathway (Figure 5).It is noteworthy that
plastids also completethe conversion of protoporphinobilogen IXto
heme independently for the benefit of fur-nishing the prosthetic
group for their own cy-tochromes.
For heme we thus see a picture of two path-ways rationalized
into one pathway that forksthree ways: chlorophyll and heme
synthesisfrom glutamyl-tRNA in the plastid, and hemesynthesis in
the mitochondrion commencingnot from its original ALAS but from
plastid-synthesised ALA (Figure 5).
Aromatic Amino Acid SynthesisThe essential amino acids are a
necessary partof the animal diet because we lack a pathway
Mt
Nuc
Mt
Nuc
E-3-P+PEP
Phe
Trp
Tyr
Trp
E-3-P+PEP
Phe
Tyr
E-3-P+PEP
Phe
Trp
Tyr
Figure 6Schematic representation of aromatic amino acid
synthesis before andafter primary endosymbiosis. Nuc, nucleus; Mt,
mitochondrion; E-3-P,erythrose-3-phosphate; PEP,
phosphoenolpyruvate; Tyr, tyrosine; Phe,phenylalanine; Trp,
tryptophan.
to synthesize tryptophan, phenylalanine, andtyrosine. Autotrophs
lack a diet and must syn-thesize these and all 17 other amino
acids.In plants the shikimate pathway located inthe plastid
synthesizes the precursors for thearomatic amino acids. There are
two ver-sions of the shikimate pathway: a prokaryotic-style
version, which is what occurs in theplant plastid, and a
cytosolic-based versionwith different enzymes, as occurs in
fungi(45). In plants the plastid has clearly retainedits ancestral
ability to synthesize tryptophan,phenylalanine, and tyrosine and
supplies theseamino acids to the cytosol (host) (Figure 6).Whether
or not the original host possessed ashikimate pathway prior to
primary endosym-biosis remains unclear. If it did, all traces
arenow lost and the plastid bears sole responsi-bility for this
task in members of the red algae,green algae, and plants.
Fe-S ClustersFe-S clusters are important prosthetic groupsof
various metalloproteins that participate inredox reactions, sensing
of iron and oxygen,and catalysis (69). The Fe atom in Fe-S
clus-ters is able to take up an electron reversibly,thus providing
the required electron carriercapacity. Fe-S–containing proteins are
perva-sive in life and ancient; well-known examplesinclude the
ferredoxins, NADH dehydroge-nase, and Coenzyme Q. The clusters
containdifferent numbers of iron and sulfur depend-ing on cluster
type and are coordinated intothe protein through cysteinyl ligands.
Syn-thesis and insertion or removal of Fe-S clus-ters into and out
of proteins is managed bya number of enzymes, not all of which
havebeen identified. At least three different sys-tems for Fe-S
cluster formation and inser-tion [iron sulfur cluster (ISC),
nitrogen fix-ation (NIF), and mobilization of sulfur (Suf )]are
known thus far. Plastids harbor a Suf-type Fe-S cluster formation
system homolo-gous to that of cyanobacteria (11, 119).
Sufsystem–generated Fe-S clusters are probablyincorporated into a
range of plastid proteins
508 Gould ·Waller · McFadden
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
including ferredoxin and the Rieske iron sul-fur protein.
The original host of the primary plastid en-dosymbiont almost
certainly contained an Fe-S cluster formation system, but it was
prob-ably not cytosolic. Eukaryotes all appear toform Fe-S
clusters, but the initial steps are mi-tochondrial and typically
utilize the ISC sys-tem (64). An export machinery translocates
the Fe-S cluster into the cytosol where ma-chinery exists to
insert the cluster into apopro-teins. This mitochondrial-based
synthesis waslikely obtained with the α-proteobacterialendosymbiont
that gave rise to the mito-chondrion (70). The host cell for
primaryendosymbiosis likely had this system, and in-deed, plants
utilize an ISC system in their mi-tochondria and a Suf system in
the plastid (11).
SUMMARY POINTS
1. All plastids ultimately arose from a single prokaryotic
endosymbiosis, where acyanobacterium was engulfed and retained by a
eukaryotic cell.
2. Eukaryotic endosymbiosis has occurred multiple times, and
this process has lead tothe spread of plastids throughout a great
diversity of eukaryotes.
3. Further endosymbiotic events continue to occur in nature, and
a wide continuum existsbetween temporary symbiotic relationships,
stable interdependent partnerships, andthose that are intimately
integrated at a molecular-genetic level.
4. Pivotal to the integration of plastids was the establishment
of protein delivery systemsthat enabled a shift of genetic control
from the organelle to the host nucleus. Prokary-otic endosymbiosis
required the generation of a novel protein import system,
whereaseukaryotic endosymbionts have co-opted and adapted existing
protein translocationsystems to achieve this task for complex
plastids.
5. Eukaryotes have gained numerous metabolic capabilities
through endosymbiosis, in-cluding but not restricted to
photosynthesis. Some of these capabilities were uniqueprokaryotic
inventions, and have thus extended the capabilities of eukaryotes.
Othersrepresented duplications of existing host pathways, and in
many cases rationalizationof redundancy has generated novel
chimeric pathways in eukaryotes.
FUTURE ISSUESIn recent years the broadening of genome sequencing
programs has encompassed agreater diversity of plastid-bearing
eukaryotes, and we are now seeing great advances inour
understanding of plastids. This diversity includes the molecular
integration of plastidgenomes with those of the host, the
mechanisms and trafficking routes of transferred geneproducts on
their return journey to the plastid, and the metabolic integration
and tradebetween plastids and their diverse hosts. Several key
research directions now presentthemselves and are conceivably
within greater reach than ever before. (1) Although thephylogenetic
affinities between the major eukaryotic lineages are begining to
slowlyresolve into focus, considerable controversy continues to
surround the question of howmany endosymbiotic events have
generated the plastid diversity observed in eukaryotes—notably with
respect to the plastids of the Chromalveolates. Resolution of these
issues
www.annualreviews.org • Plastid Evolution 509
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
is critical to our interpretation of evolution of the diversity
we see among plastids inthis group, and our understanding of the
frequency and mechanisms of plastid loss.(2) Although our
understanding of protein targeting to primary plastids has reached
arelatively advanced state, equivalent insight into the mechanisms
for targeting proteinsto complex plastids lags behind. These
details may be vital in tackling some of the moreinsidious complex
plastid-bearing eukaryotes such as the apicomplexan parasites,
whereplastid-targeted pathways offer tantalizing possibilities as
drug targets for diseases suchas malaria. (3) Genomic analyses have
presented some insights into the broader suite ofmetabolic
functions of plastids beyond photosynthesis; however, little is
known aboutdelivery of the products of these pathways to the host
cell, or vice versa. Although thispresents one of the larger
challenges to plastid researchers, the realization of this goal
isnecessary to provide a full appreciation of the significance of
plastid gain in eukaryotesthrough endosymbiosis.
DISCLOSURE STATEMENTThe authors are not aware of any biases that
might be perceived as affecting the objectivity ofthis review.
ACKNOWLEDGMENTSG.M. is an ARC Federation Fellow and Howard
Hughes International Scholar and supportedby a Program Grant from
the NHMRC. S.B.G. is supported by the ARC grant DP0664097.
LITERATURE CITED1. Abrahamsen MS, Templeton TJ, Enomoto S,
Abrahante JE, Zhu G, et al. 2004. Complete
genome sequence of the apicomplexan, Cryptosporidium parvum.
Science 304:441–452. Adl SM, Simpson AG, Farmer MA, Andersen RA,
Anderson OR, et al. 2005. The
new higher level classification of eukaryotes with emphasis on
the taxonomy of protists.J. Eukaryot. Microbiol. 52:399–451
3. Akita M, Nielsen E, Keegstra K. 1997. Identification of
protein transport complexes inthe chloroplastic envelope membranes
via chemical cross-linking. J. Cell Biol. 136:983–94
4. Altmann R. 1890. Die Elementarorganismen und ihre Beziehungen
zu den Zellen. Leipzig:Veit
5. Andersen T, Briseid T, Nesbakken T, Ormerod J, Sirevåg R,
Thorud M. 1983. Mecha-nisms of synthesis of 5-aminolevulinate in
purple, green and blue-green bacteria. FEMSMicrobiol. Lett.
19:303–6
6. Andersson JO, Roger AJ. 2002. A cyanobacterial gene in
nonphotosynthetic protists—anearly chloroplast acquisition in
eukaryotes? Curr. Biol. 12:115–19
7. Archibald JM, Longet D, Pawlowski J, Keeling PJ. 2003. A
novel polyubiquitin structurein Cercozoa and Foraminifera: evidence
for a new eukaryotic supergroup. Mol. Biol. Evol.20:62–66
8. Bach TJ, Boronat A, Campos N, Ferrer A, Vollack KU. 1999.
Mevaloate biosynthesis inplants? Crit. Rev. Biochem. Mol. Biol.
34:107–22
510 Gould ·Waller · McFadden
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
9. Bachvaroff TR, Sanchez Puerta MV, Delwiche CF. 2005.
Chlorophyll c-containing plas-tid relationships based on analyses
of a multigene data set with all four chromalveolatelineages. Mol.
Biol. Evol. 22:1772–82
10. Baldauf SL, Roger AJ, Wenk-Siefert I, Doolittle WF. 2000. A
kingdom-level phylogenyof eukaryotes based on combined protein
data. Science 290:972–77
11. Balk J, Lobreaux S. 2005. Biogenesis of iron-sulfur proteins
in plants. Trends Plant Sci.10:324–31
12. Bauer J, Chen K, Hiltbunner A, Wehrli E, Eugster M, et al.
2000. The major proteinimport receptor of plastids is essential for
chloroplast biogenesis. Nature 403:203–7
13. Becker T, Hritz J, Vogel M, Caliebe A, Bukau B, et al. 2004.
Toc12, a novel subunit of theintermembrane space preprotein
translocon of chloroplasts. Mol. Biol. Cell 15:5130–44
14. Bedard J, Jarvis P. 2005. Recognition and envelope
translocation of chloroplast prepro-teins. J. Exp. Bot.
56:2287–320
15. Brocks JJ, Logan GA, Buick R, Summons RE. 1999. Archean
molecular fossils and theearly rise of eukaryotes. Science
285:1033–36
16. Bruce BD. 2000. Chloroplast transit peptides: structure,
function and evolution. TrendsCell Biol. 10:440–47
17. Cavalier-Smith T. 1982. The origins of plastids. Biol. J.
Linn. Soc. 17:289–30618. Cavalier-Smith T. 1986. The kingdom
Chromista: origin and systematics. In Progress in
Phycological Research, ed. FE Round, DJ Chapman, pp. 309–47.
Bristol: Biopress19. Cavalier-Smith T. 1993. The origin, losses and
gains of chloroplasts. In Origins of Plastids:
Symbiogenesis, Prochlorophytes and the Origins of Chloroplasts,
ed. RA Lewin, pp. 291–348.New York: Chapman & Hall
20. Cavalier-Smith T. 2000. Membrane heredity and early
chloroplast evolution. Trends PlantSci. 5:174–82
21. Cavalier-Smith T. 2002. Chloroplast evolution: secondary
symbiogenesis and multiplelosses. Curr. Biol. 12:R62–64
22. Cavalier-Smith T. 2003. Genomic reduction and evolution of
novel genetic membranesand protein-targeting machinery in
eukaryote-eukaryote chimaeras (meta-algae). Philos.Trans. R. Soc.
London Ser. B 358:109–33
23. Cavalier-Smith T, Chao EE. 2006. Phylogeny and
megasystematics of phagotrophicheterokonts (kingdom Chromista). J.
Mol. Evol. 62:388–420
24. Chou ML, Fitzpatrick LM, Tu SL, Budziszewski G, Potter-Lewis
S, et al. 2003. Tic40,a membrane-anchored co-chaperone homolog in
the chloroplast protein translocon.EMBO J. 22:2970–80
25. Deschamps P, Haferkamp I, Dauvillee D, Haebel S, Steup M, et
al. 2006. Nature of theperiplastidial pathway of starch synthesis
in the cryptophyte Guillardia theta. Eukaryot.Cell 5:954–63
26. Dodge JD. 1969. The ultrastructure of Chroomonas
mesostigmatica Butcher (Crypto-phyceae). Arch. Mikrobiol.
69:266–80
27. Douglas S, Zauner S, Fraunholz M, Beaton M, Penny S, et al.
2001. The highly reducedgenome of an enslaved algal nucleus. Nature
410:1091–96
28. Eckart K, Eichacker L, Sohrt K, Schleiff E, Heins L, Soll J.
2002. A Toc75-like proteinimport channel is abundant in
chloroplasts. EMBO Rep. 3:557–62
29. Eisen JA, Coyne RS, Wu M, Wu D, Thiagarajan M, et al. 2006.
Macronuclear genomesequence of the ciliate Tetrahymena thermophila,
a model eukaryote. PLoS Biol. 4:e286
30. Fast NM, Kissinger JC, Roos DS, Keeling PJ. 2001.
Nuclear-encoded, plastid-targetedgenes suggest a single common
origin for apicomplexan and dinoflagellate plastids. Mol.Biol.
Evol. 18:418–26
www.annualreviews.org • Plastid Evolution 511
Ann
u. R
ev. P
lant
Bio
l. 20
08.5
9:49
1-51
7. D
ownl
oade
d fro
m a
rjour
nals.
annu
alre
view
s.org
by W
IB62
42 -
Uni
vers
itaet
s- u
nd L
ande
sbib
lioth
ek D
uess
eldo
rf on
01/
28/1
0. F
or p
erso
nal u
se o
nly.
-
ANRV342-PP59-20 ARI 26 March 2008 20:26
31. Gabrielson PW, Garbary DJ, Sommerfeld MR, Townsend RA, Tyler
PL. 1990.Rhodophyta. See Ref. 72, pp. 102–18
32. Geitler L. 1977. Zur Entwicklungsgeschichte der
Epithemiaceen Epithemia, Rhopalodiaund Denticula (Diatomophyceae)
und ihre vermutlich symbiontischen Sphaeroidkoerper.Plant Syst.
Evol. 128:295–75
33. Gilson PR, Su V, Slamovits CH, Reith ME, Keeling PJ,
McFadden GI. 2006. Completenucleotide sequence of the
chlorarachniophyte nucleomorph: nature’s smallest nucleus.Proc.
Natl. Acad. Sci. USA 103:9566–71
34. Gould SB, Sommer MS, Hadfi K, Zauner S, Kroth PG, Maier UG.
2006. Protein tar-geting into the complex plastid of cryptophytes.
J. Mol. Evol. 62:674–81
35. Gould SB, Sommer MS, Kroth PG, Gile GH, Keeling PJ, Maier
UG. 2006. Nucleus-to-nucleus gene transfer and protein retargeting
into a remnant cytoplasm of cryptophytesand diatoms. Mol. Biol.
Evol. 23:2413–22
36. Grambast L. 1974. Phylogeny of the Charophyta. Taxon
23:463–8137. Gruber A, Vugrinec S, Hempel F, Gould SB, Maier UG,
Kroth PG. 2007. Protein
targeting into complex diatom plastids: functional
characterisation of a specific targetingmotif. Plant Mol. Biol.
64:519–30
38. Gutensohn M, Fan E, Frielingsdorf S, Hanner P, Hou B, et al.
2006. Toc, Tic, Tat et al.:structure and function of protein
transport machineries in c