Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Look up keyword
Like this
1Activity
0 of .
Results for:
No results containing your search query
P. 1
Elysia chlorotica

Elysia chlorotica

Ratings: (0)|Views: 77|Likes:
Published by Sergio L.

More info:

Published by: Sergio L. on Apr 14, 2012
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

04/15/2012

pdf

text

original

 
Transcriptomic Evidence for the Expression of HorizontallyTransferred Algal Nuclear Genes in the Photosynthetic SeaSlug,
Elysia chlorotica
Sidney K. Pierce,
 ,
*
 ,1,2
Xiaodong Fang,
 ,3
 Julie A. Schwartz,
1
Xuanting Jiang,
3
Wei Zhao,
3
Nicholas E. Curtis,
1
Kevin M. Kocot,
4
Bicheng Yang,
3
and Jian Wang*
 ,3,4
1
Department of Integrative Biology, University of South Florida
2
Department of Biology, University of Maryland
3
BGI-Shenzhen, Shenzhen, China
4
Department of Biological Sciences, Auburn University
These authors contributed equally to this work.*
Corresponding author:
E-mail: Wangjian@genomics.org.cn; pierce@usf.edu.
Associate editor:
Charles Delwiche
Abstract
Analysis of the transcriptome of the kleptoplastic sea slug,
Elysia chlorotica
 , has revealed the presence of at least 101chloroplast-encoded gene sequences and 111 transcripts matching 52 nuclear-encoded genes from the chloroplast donor,
Vaucheria litorea
. These data clearly show that the symbiotic chloroplasts are translationally active and, of even moreinterest, that a variety of functional algal genes have been transferred into the slug genome, as has been suggested byearlier indirect experiments. Both the chloroplast- and nuclear-encoded sequences were rare within the
E. chlorotica
transcriptome, suggesting that their copy numbers and synthesis rates are low, and required both a large amount of sequence data and native algal sequences to find. These results show that the symbiotic chloroplasts residing inside thehost molluscan cell are maintained by an interaction of both organellar and host biochemistry directed by the presence of transferred genes.
Key words:
transcriptome, HGT, horizontal gene transfer, chloroplast symbiosis, kleptoplasty,
Elysia chlorotica
 ,
Vaucheria litorea
.
Introduction
Certain cells lining the digestive tubules of several species of sacoglossan, opisthobranch, sea slugs are able to sequesterchloroplasts from their algal food. The plastids remainintact inside the digestive cells for some length of time, de-pending on the species involved (Pierce and Curtis 2012).Also, in several of these slug species, the captured plastids,often called kleptoplasts, are capable of photosynthesis,and in a few of the species, photosynthesis continuesalmost unabated for as long as a year in the completeabsence of any contact with the algal source of thechloroplasts (Pierce and Curtis 2012).Most detailed information about various aspects of themechanism of long-term maintenance of the chloroplastsymbiosis has come from work on
Elysia chlorotica
(Gould),where the sequestered chloroplasts come from the chro-mophytic alga,
Vaucheria litorea
(C. Agardh). Once thisslug sequesters the chloroplasts, it can continue to pho-tosynthesize for 10–12 months in the absence of any algalfood. Several chloroplast proteins and chlorophyll a aresynthesized during that starvation period, and polymerasechain reaction (PCR) experiments have demonstrated thepresence of at least 11 algal nuclear genes, all involvedin photosynthesis, in
E. chlorotica
adult and veliger larvalgenomic DNA as well as in adult slug RNA (Pierce et al.2007 ,2009 ;Rumpho et al. 2008 ,2009 ;Schwartz et al. 2010). All of these results show that translationally com-petent algal nuclear genes are present in the slug andthat plastid protein and pigment turnover, necessaryfor sustained photosynthesis, is taking place in the slugcell, supported by horizontal gene transfer (HGT) be-tween the two multicellular species. In addition, a varietyof chloroplast-encoded chloroplast proteins are also syn-thesized while the plastid resides inside the
E. chlorotica
digestive cell (Mujer et al. 1996 ;Green et al. 2000). How- ever, recent partial analyses of the transcriptomes of twoother slug species,
E. timida
and
Plakobranchus ocellatus
 ,failed to find any transcriptome sequence reads corre-sponding to algal nuclear genes, which lead to the provoc-ative conclusion that, in spite of the entire foregoing, HGThas not occurred between slug and algae and that ‘‘saco-glossan are not, in genetic terms, what they eat’’ (Wa¨geleet al. 2011). Perhaps not, at least in the case of the twospecies investigated byWa¨gele et al. (2011) , but severalaspects of this study seem to make such a broad conclusionpremature.Forinstance,thedatasetrepresentedonlyasmallfraction of the transcriptome. Also, the RNA came from justa few specimens and was extracted from whole animals, al-though only a small fraction of cells contain chloroplasts. Inaddition,Wa¨gele et al. (2011)assumed that expressionlevels
©
The Author 2012. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, pleasee-mail: journals.permissions@oup.com
Mol. Biol. Evol.
doi:10.1093/molbev/msr316 Advance Access publication December 23, 2011 1
 R  e  s  e  a  r  c  h   a  r  t   i     c  l     e 
 
MBE Advance Access published February 21, 2012
  a  t  M c  Gi  l  l   Uni   v e  s i   t   yi   b  a i   e  s  onA pi  l   , 0  t   t   p :  /   /  m b  e  . oxf   o d  j   o un a l   s  . o g /  D o wnl   o a  d  e  d f   om
 
thanin itsfood alga(Soule 2009).Thus, generalizing negative results obtained from problem-atical data to the rest of the species of kleptoplastic slugs,especiallyinthepresenceofthelargeamountofbiochemicaldata,includingPCR,andshowingthepresenceoftransferredgenes, seem inappropriate.However, another recent report failed to find any evi-dence for transferred algal nuclear genes in the transcrip-tome of 
E. chlorotica
 , although 19 chloroplast-encodedgene sequences emerged from the analysis (Pelletreauet al. 2011). Unfortunately, while this study investigated
E. chlorotica
 , the transcriptome sequencing yielded
,
20% (14,000) of the contigs of the
P. ocellatus
transcrip-tome and was not compared with
V. litorea
native sequen-ces. Thus, the lack of both sequencing depth and databasevoids suggests that rare transcripts could have been easilymissed. Indeed, a conclusion of this analysis was that ‘‘moreexhaustive sequencing may be required’to adequatelytest for the presence of transferred genes in
E. chlorotica
(Pelletreau et al. 2011).In spite of the issues with these negative studies(Pelletreau et al. 2011 ;Wa¨gele et al. 2011), both nonetheless point to the likelihood that if transferred algal nucleargenes are present in sacoglossan slugs, they will be of verylow copy number, their expression will be low, and knowl-edge of the native algal sequence will facilitate, perhapseven be required for, the annotation. Therefore, insteadof assuming that a large number of transcripts for nuclear-encoded and plastid-encoded proteins would be present inthe slug cell, we have hypothesized that such would be ex-ceedingly rare in the transcriptome and have done our ownanalysis of 
E. chlorotica
creating large amounts of Illumina-generated sequence data. In addition, to facilitate an accu-rate annotation, wehavesequenced the genome of 
V. litorea
as well as the algal transcriptome, to provide a databaseof native algal transcript sequences. Our
E. chlorotica
tran-scriptome data contain a variety, albeit rare, of chloroplast-encoded transcripts and, in addition, rarer still, at least 111reads that match 52 algal nuclear-encoded sequences, in-cluding one that matches exactly an algal nuclear sequencefoundbyPCRin
E.chlorotica
genomicDNAandcDNAintheearlier studies.
Materials and Methods
Animals and Algae
Specimens of 
E. chlorotica
were collected from a salt marshnear Menemsha on Martha’s Vineyard, MA, and shipped toTampa, FL. The sea slugs were placed into aerated aquariacontaining artificial seawater made from Instant Oceansalts dissolved in sterilized deionized water at 1,000 mOsm.The aquaria were kept at 10
°
C in a cold box equipped withfluorescent lights set on a 14:10 light–dark cycle. The slugswere starved for at least 2 months under the foregoingconditions before use in experiments.Filaments of 
V. litorea
used in the experiments camefrom a culture that has been maintained in our lab for morethan 10 years. The initial filament used to establish the cul-ture came from the same marsh that provided the slugs.The
V. litorea
filaments are grown in a modified F/2enriched seawater at 250 mOsm in an incubator at20
°
C, illuminated by fluorescent lights on a 14:10 light–dark cycle (Pierce et al. 1996). Since the original culturewas established from a single filament and has only grownvegetatively, the filaments used here are clonal.
Extraction of Genomic DNA from
V. litorea
Genomic DNA was extracted from 10 g of algal filamentsusing a protocol modified fromAl-Samarrai and Schmid(2000).The filaments were rinsed with fresh culture media,blotted to remove excess liquid, frozen in liquid nitrogen,and ground to a powder with a precooled mortar andpestle. Lysis buffer (40 mM Tris–acetate, 20 mM sodiumacetate, 1 mM ethylenediaminetetraacetic acid, and 1%sodium dodecyl sulfate, pH 7.0) was added to the frozenalgal powder and extracted for 10 min at room tempera-ture (RT) while rotating. To facilitate the precipitation of proteins and polysaccharides, 5 M NaCl was added to eachtube (2:5 v/v), the tubes were vortexed and then centri-fuged at 12,000
Â
g at 4
°
C. The supernatant was extractedwith phenol/chloroform (1:1 v/v) and spun at 10,000
Â
g4
°
C. The aqueous phase was extracted with chloroform(1:1 v/v) and centrifuged again at 10,000
Â
g 4
°
C. DNAwas precipitated from the aqueous phase by adding isopro-panol (1:1 v/v) and spun at 10,000
Â
g 4
°
C to pellet thenucleic acids. The precipitated DNA was resuspended inlysis buffer containing 100
l
g/ml RNase A (Qiagen, Valencia,CA) and rotated for 30 min at RT. Following thatincubation, the DNA was run through the purificationprocess a second time. The final precipitated DNA waswashed twicewith 75% ethanol, air dried, resuspended in nu-clease-free water, quantified spectrophotometrically at 260nm, and then express shipped to BGI in Hong Kong, China.
Algal Transcriptome
Two
V. litorea
transcriptome data sets were used in ouranalysis. One set was produced earlier by 454 sequencing(Schwartz et al. 2010) and the second set using the Illuminaplatform as described below. The 454 data are referred to asEST and the Illumina data as RNA-seq, hereafter.
Extraction of RNA from
V. litorea
Total RNA was isolated from about 1 gm of algal filamentsafter 5 h of exposure to light. The filaments were groundinto a frozen powder as described above, RNA was ex-tracted using the RNeasy Plant mini kit (Qiagen, Valencia,CA) following the manufacturer’s instructions, quantifiedspectrophotometrically at 260 nm, placed on dry ice,and express shipped immediately to BGI in Hong Kong, CN.
Extraction of RNA from
E. chlorotica
Total RNA was isolated from
.
2-month starved slugs af-ter 8 h of light exposure by homogenization in Trizol Re-agent (Invitrogen, Carlsbad, CA). The homogenate was
Pierce et al.
·
doi:10.1093/molbev/msr316
MBE
2
  a  t  M c  Gi  l  l   Uni   v e  s i   t   yi   b  a i   e  s  onA pi  l   , 0  t   t   p :  /   /  m b  e  . oxf   o d  j   o un a l   s  . o g /  D o wnl   o a  d  e  d f   om
 
3
C
6
H
5
O
7
In total, 8 Gb, or 86-fold coverage, of high-quality reads wereused in the assembly. The algal genome was assembled usingSOAPdenovo (Li, Zhu, et al. 2010) software (http://soap. genomics.org.cn), using the same procedures described forassembly of the giant Panda genome (Li, Fan, et al. 2010).A total of 7.9 Gb (or 85.9X) data were retained for as-sembly. All high-quality paired-end reads were aligned intocontigs for scaffold building. This paired-end informationwas subsequently used to link contigs into scaffolds,step-by-step, from short insert sizes to long insert sizes.About 3.9 Gb (or 42.4X) data were used to build contigsfor the algal genome, and all the high-quality data wereused to build scaffolds. Some intrascaffold gaps were filledby local assembly using the reads in a read pair, where oneend uniquely aligned to a contig while the other end waslocated within a gap. The final total contig size and N50were 83.6 Mb and 59.6 Kb, respectively. The total scaffoldsize and N50 were 93.2 Mb and 333.3 Kb, respectively.More than 97% of long ESTs (
.
500 bp) mapped to thealgal genome, which indicated the high quality of the ge-nome in the transcribed regions.
Gene Annotation Pipeline and Evaluation of GeneQuality
Since there is only scant
V. litorea
sequence informationin the public databases, we used both homology basedand de novo methods to localize gene sequences in thealgal genomic data incorporating reads from both theRNA-seq and EST data. To identify homologous genes,protein sequences from algal species that were availablein NCBI (http://www.ncbi.nlm.nih.gov/), including
Ectocarpussiliculosus
 ,
Phaeodactylum tricornutum
 ,
Micromonas sp
.,
M. pusilla, Ostreococcus lucimarinus
 ,
O. tauri
 , and
Volvoxcarteri
 , were mapped to the
V. litorea
genome using TBlastN(Kent 2002). Then, homologous genome sequences werealigned against the matching proteins, using GeneWise(Birney et al. 2004) to define gene models. For de novodiscovery of coding genes, AUGUSTUS (Stanke andWaack 2003), GlimmerHMM, and SNAP were used withappropriate parameters. ESTs were mapped to the ge-nome with BLAST and assembled to genes with PASA.RNA-seq were mapped to the genome using TopHat(http://tophat.cbcb.umd.edu/), and transcriptome-basedgene structures were obtained with Cufflinks (http://cufflinks.cbcb.umd.edu/). Finally, the homology-basedde novo derived EST prediction and transcript gene setswere merged to form a comprehensive and nonredundantreference gene set using GLEAN (http://sourceforge.net/projects/glean-gene/), removing all the genes which hadonly de novo method support. This procedure produceda reference set of 17,988
V. litorea
genomic coding genes.
Transcriptome Sequencing and Assembly
To prepare the slug and algal transcriptomes for sequenc-ing, poly (A)
þ
RNA was enriched from the total RNA of each species, sheered into fragments, and cDNA wassynthesized by reverse transcription. The cDNA from eachspecies was then sequenced using standard high through-put techniques (Illumina HiSeq2000). All high-quality readswere assembled into contigs longer than 100 bp usingSOAPdenovo software (Li et al. 2010). Contigs were linkedinto scaffolds by mapping reads back to contigs andcombining paired-end information.
Sequence Analysis
Alignment and annotation of the
E. chlorotica
transcrip-tome sequences were done using the
V. litorea
coding se-quence database as a reference source (fig. 1) utilizing MPIBLAST 1.5.0 software (Darling et al. 2003 ;Lin et al. 2005), by means of the University of South Florida’s 120 node com-puter cluster platform consisting of dual Intel Xeon X5460Quad Core processors each with 16 GB of memory. Briefly,the slug contigs, scaffolds, and raw reads from the tran-scriptome data were formatted as databases and then com-pared with all 17,988
V. litorea
coding sequences using theBlastN algorithm (Altschul et al. 1990) with a cutoff of 1
Â
10
À
10
. Slug transcripts with significant hits were aligned tothe corresponding
V. litorea
reference sequence using theClustalW2 (http://www.ebi.ac.uk/Tools/msa/clustalw2/) sequence alignment program todetermine the location of the matching slug transcripts inthe coding sequence. The corresponding
V. litorea
match-ing sequences were then analyzed using the BlastX algo-rithm (Altschul et al. 1990) to determine if they werenuclear- or chloroplast-encoded in origin by comparisonwith the
V. litorea
chloroplast genome database in NCBI.Last, slug transcripts matching to algal nuclear-encodedsequences were compared with the assembled slug andalgal transcriptomes using BlastN to determine if the se-quences were located in the contig or raw-read data sets.
Results
Chloroplast-Encoded Transcripts in
E. chlorotica
OursequencingrunsofthecDNAmadefromthe
E.chlorotica
transcriptome produced 98,238,204 reads, which assembledinto 459,299 contigs and 378,851 scaffolds (table 1).
Algal Nuclear Genes in Sea Slug Transcriptome
·
doi:10.1093/molbev/msr316
MBE
3
  a  t  M c  Gi  l  l   Uni   v e  s i   t   yi   b  a i   e  s  onA pi  l   , 0  t   t   p :  /   /  m b  e  . oxf   o d  j   o un a l   s  . o g /  D o wnl   o a  d  e  d f   om

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->