RESEARCH ARTICLEDebashish Bhattacharya á Jamie J. CannoneRobin R. GutellGroup I intron lateral transfer between redand bro...
occurred at position 516 in small subunit rRNA. Thiscollection of ``tagged 516 introns and the sequences forthe SSU rRNA i...
tetraloop hairpin sequence at position 156:9±12(Fig. 1D) showed high identity (5¢-GYAA-3¢) to one ofthe canonical hairpin ...
516 introns, we used the BLAST program (BLASTNver. 2.1.2; Altschul et al. 1997) to search for other intronsequences with s...
constructed the phylogeny of the 516 introns, other SSUrRNA introns, and the host cells that contain them,using more tradi...
are postulated to be the sister group of the Florid-eophycidae (Oliveira and Bhattacharya 2000; MuÈ lleret al. 2001b), a h...
was again no support found for the inclusion of theNaegleria spp. 516 introns within this clade.Given this phylogenetic fr...
transfer. It is now generally believed that the plastid ofphotosynthetic stramenopiles has arisen through a singlesecondar...
Friedl T, Besendahl A, Pfei€er P, Bhattacharya D (2000) Thedistribution of group I introns in lichen algae suggests thatli...
Upcoming SlideShare
Loading in …5
×

Gutell 076.curr.genetics.2001.40.0082

139 views

Published on

Published in: Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
139
On SlideShare
0
From Embeds
0
Number of Embeds
2
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Gutell 076.curr.genetics.2001.40.0082

  1. 1. RESEARCH ARTICLEDebashish Bhattacharya á Jamie J. CannoneRobin R. GutellGroup I intron lateral transfer between redand brown algal ribosomal RNAReceived: 12 February 2001 / Accepted: 26 April 2001 / Published online: 17 July 2001Ó Springer-Verlag 2001Abstract How group I introns originate in nuclear rib-osomal (r)RNA genes is an important question in evo-lutionary biology. Central to this issue is the multitudeof group I introns present in evolutionarily distantlyrelated plant, fungal, and protist lineages, together withan understanding of their origin and lateral transferfrom one exon to another, between cell organelles, andbetween cells. These introns vary considerably in pri-mary and secondary structure; and their provenancefrom a few or perhaps many mobile elements that havespread in rRNAs is unknown. Here we show that anovel lineage of group IC1 introns inserted at posi-tion 516 (Escherichia coli gene numbering) in the smallsubunit rRNA in bangiophyte red algae and a brownalga (Aureoumbra lagunensis) are speci®cally related,although their host cells are not. These bangiophyte andAureoumbra introns are the only known cases that havea helical insertion in the P5b helix. The highly conservedprimary and secondary structure of the extra P5b helixsuggests that it is important, although its speci®c func-tion is unknown. Our study attempts to understand theorigin and movement of these IC1 introns.Keywords Group I intron á Stramenopiles áLateral transfer á PhylogenyIntroductionGroup I introns are a distinct class of RNA enzymesthat are characterized by a conserved RNA primary andsecondary structure essential for splicing and are oftencapable of ``self-splicing (Kruger et al. 1982; Cech1985). Clarifying the origin and evolutionary history ofgroup I introns helps us understand how these sequenceshave persisted for billions of years and, more generally,addresses the possible role(s) of catalytic RNAs inmodern cells. The sporadic and wide distribution ofgroup I introns in the nuclear-encoded small (SSU) andlarge (LSU) subunit ribosomal (r)RNA and organellargenes of green algae, red algae, fungi, ciliates, amoebae,and other organisms (Cannone et al., unpublished data)suggests that these sequences are highly successfulat invading and maintaining themselves in eukaryoticgenomes (Bhattacharya et al. 1994, 1996a; Turmel et al.1995; Bhattacharya 1998; Nishida et al. 1998; Watanabeet al. 1998; Suh et al. 1999; Friedl et al. 2000).In this study, we use secondary structure and phylo-genetic analyses to address the lateral transfer of group Iintrons in nuclear rRNA. To approach this issue, uniquestructural features of introns are tracked through evo-lutionary time. We also use traditional phylogeneticmethods to determine the evolutionary relationships ofthese introns and their host cells. Currently, there areapproximately 1,214 group I introns available in Gen-Bank. Our analyses of these sequences [summarizedat the Comparative RNA web site (CRW), http://www.rna.icmb.utexas.edu/] resulted in their classi®ca-tion into ®ve major structural subgroups [Michel andWesthof 1990 (subgroups A±D); Suh et al. 1999 (sub-group E)]. Within the order Bangiales in the Rhodo-phyta (red algae), IC1 introns at position 516 revealed aunique insertion in the P5b helix, which is absent fromall other IC1 introns (Gutell and Cannone, unpublisheddata). Subsequently, we found a group I intron in abrown algal stramenopile, Aureoumbra lagunenesis, withthe same unusual insertion that, coincidentally, alsoCurr Genet (2001) 40: 82±90DOI 10.1007/s002940100227Communicated by R.W. LeeD. Bhattacharya (&)Department of Biological Sciences, University of Iowa,239 Biology Building, Iowa City, IA 52242-1324, USAE-mail: dbhattac@blue.weeg.uiowa.eduJ.J. Cannone á R.R. GutellInstitute for Cellular and Molecular Biology,University of Texas, 2500 Speedway, Austin,TX 78712-1095, USA
  2. 2. occurred at position 516 in small subunit rRNA. Thiscollection of ``tagged 516 introns and the sequences forthe SSU rRNA in the host cells have presented us withthe opportunity to explore the lateral transfer of a groupI intron. There are two major scenarios that couldexplain this situation:1. Introns early. The 516 intron was present in the SSUrRNA of the common ancestor of red and brownalgae and has been maintained only in the Bangialesand Aureoumbra. This scenario would be supportedby a close evolutionary relationship between red andbrown algae, whereas a distant relationship wouldsupport independent acquisition of the intron.2. Introns late. The presence of the 516 intron in theBangiales and Aureoumbra is best explained by in-dependent acquisition of the intron after the separa-tion of red and brown algae. We used phylogeneticanalyses of introns and SSU rRNA coding regions toaddress the origin of these 516 introns.Materials and methodsIntron and rRNA sequencesThe publicly available sequences analyzed here were retrieved fromGenBank (http://www.ncbi.nlm.nih.gov/Genbank/index.html), atthe National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/). The rRNA and rRNA introns were usu-ally identi®ed by searching for the words ``rRNA and ``introns,and for ``rRNA in the feature key ®eld in the Entrez web-searching utility (http://www.ncbi.nlm.nih.gov/Entrez/). However,we have utilized the BLAST utility (http://www.ncbi.nlm.nih.gov/BLAST/) to identify some intron sequences that are similar to thered algal rRNA intron sequences, but are not always annotatedproperly in their GenBank entry. Although the boundaries for therRNA exons and introns are usually annotated properly, we haveindependently determined the intron/exon borders.Alignment and secondary structure models and diagramsSecondary structure diagrams for the red and brown algal group Iintrons were modeled from the current comparative structuremodels for the IC1 introns (Michel and Westhof 1990; Dambergerand Gutell 1994; Gutell and Cannone, unpublished data). Thesestructural models are derived by comparative sequence analysis, amethod based on the simple premise that RNAs within the samefamily (e.g., tRNA, group I intron) that have similar functions willhave similar secondary and tertiary structures (reviewed in Gutell1996). The red and brown algal intron sequences were aligned withthe IC1 sequences, because these were the most similar in sequenceand structure. All sequences were manually aligned using the SUNMicrosystems UNIX-based alignment editor AE2 (T. Macke,Scripps Research Institute, La Jolla, Calif.) to maximize sequenceand structure identity, to maintain (when possible) previouslyproposed base-pairings, and to minimize the number of insertionand deletion events. Once the secondary structure model had beendetermined, the computer version of the diagram was drawn withthe interactive SUN Microsystems UNIX-based, interactivegraphics program XRNA (B. Weiser and H. Noller, University ofCalifornia at Santa Cruz, Calif.).Phylogenetic analysesFor the group I intron data, a set of 62 group IC1 and IE intronsequences from nuclear (16S and 16S-like) SSU rRNA in greenalgae, red algae, fungi, Acanthamoeba grini, and Naegleria spp.were manually aligned, using primary and secondary structuresimilarity to juxtapose homologous regions (e.g., Michel andWesthof 1990; Bhattacharya et al. 1994; Damberger and Gutell1994). An alignment with a total length of 4,523 characters, in-cluding insertions in the di€erent introns, was submitted to a pair-wise distance analysis, using the HKY-85 model (Hasegawa et al.1985) with equal rates across sites and the transition/transversionratio=2. Missing data and gaps were excluded from each pair-wisecomparison. This distance matrix was used as input for neighbor-joining tree-building. A total of 1,000 bootstrap replicates (Fel-senstein 1985) were analyzed with the distance method. We alsoused the Modeltest program (ver. 3.04; Posada and Crandall 1998)to survey 56 possible models of DNA substitution, to identify themodel that best ®t the intron data. This analysis showed the generaltime-reversible model (RodrõÂguez et al. 1990) with estimations ofnucleotide frequencies (A=0.244, C=0.251, G=0.271, T=0.234),the shape parameter of the gamma distribution (a=0.764) to ac-commodate rate variations across sites, and the proportion of in-variant sites (0.016) that are unable to accept substitutions(GTR+G+I) as best ®tting the intron data. The GTR+G+Imodel was used to calculate bootstrap values (1,000 replicates) formonophyletic groups identi®ed by the HKY-85 distance tree. Inthis way, we had estimates of support for di€erent groups in theintron tree, using evolutionary models that were either relativelysimple (HKY-85), or parameter-rich (GTR+G+I). The IE intronswere used to root the intron tree.The complete SSU rRNA sequences from a broadly diverse setof red algae and stramenopiles were also aligned, using conservedsecondary structures, and 1,581 nt that could be simultaneouslyaligned in all the sequences were submitted to a distance analysisusing the HKY-85 model (as described above). Modeltest alsoidenti®ed the GTR+G+I model as best ®tting the rRNA data;and so it was used in the distance bootstrap analysis (1,000 repli-cates). The parameter value estimates for this data set were:A=0.251, C=0.206, G=0.281, T=0.262, proportion of invariantsites=0.359, a=0.548). These data were also analyzed with anunweighted maximum parsimony method. Initial trees were builtstepwise with ten random sequence additions; and the trees wererearranged with tree bisection/reconnection, to ®nd the most par-simonious phylogeny. A total of 1,000 bootstrap replicates wereanalyzed with this method. The stramenopile sequences were usedto root the SSU rRNA tree. All phylogenetic analyses were donewith PAUP* (ver. 4.0b8; Swo€ord 2001). The ®gures presented inthis text, along with other secondary structure diagrams, areavailable at the web site http://www.rna.icmb.utexas.edu/PHYLO/ALGINT and are referred to in this text as the CRW Algal Introns.Results and DiscussionSecondary structure analysis of the 516 intronsAn analysis of the group I introns at the online CRWweb site revealed that 59 bangiophytes in the mono-phyletic red algal order Bangiales (Bangia and Porphyraspp; MuÈ ller et al. 1998, 2001a; Oliveira and Bhattach-arya 2000) contained a IC1 intron at SSU rRNA posi-tion 516 with the following characteristic: they each hadan insertion in the P5b helix (see Fig. 1A,C) which isabsent from the other IC1 introns, except for an intron inA. lagunensis (Fig. 1B, Table 1). The Aureombra group Iintron was also in subgroup C1, had the P5b insertion,and mapped to the 516 site in SSU rRNA. Detailedanalysis of the P5b helix provided strong evidence for itshomology in the red and brown algal 516 introns. Theinsertion was conserved in size (19 nucleotides) and the83
  3. 3. tetraloop hairpin sequence at position 156:9±12(Fig. 1D) showed high identity (5¢-GYAA-3¢) to one ofthe canonical hairpin tetraloop sequences initially iden-ti®ed in 16S rRNA (Woese et al. 1990). All of the basepair positions have some underlying variation, and theyall contain G:C, A:U, or G:U base pairs in all of thesequences with this insertion. The three base pairs withthe purest covariance are 3:18, 7:14, and 8:13 (see darkboxes in Fig. 1D, Table 2). Taken together, these ob-servations were consistent with a single origin of the P5binsertion and, by extension, of the 516 introns contain-ing the extra helix. The pattern of sequence conservationof the P5b helix also suggested that it is involved in aRNA±RNA tertiary structure interaction within theP5abc subdomain (Cate et al. 1996; Doherty et al. 1999;Doherty and Doudna 2000). Tetraloops often dock intostem-loop structures in other regions of RNA. An ex-ample of such an interaction is between domains 5 and 1of group II introns (Abramovitz and Pyle 1997). Theconserved nature of such higher-order interactions alsogenerally requires that the helices capped by tetraloopsbe of a ®xed length (Hedenstierna et al. 2000), as foundin the P5b insertions (see Fig. 1D). Regardless of itsspeci®c role, upon establishment, the helix has appar-ently been under strong selective constraint and has notbeen lost in any of the 31 publicly available introns atposition 516 in the SSU rRNA that have been reportedfrom the Bangiales or from Aureoumbra.To determine whether other publicly available groupI introns, none of which apparently contain the P5binsertion described here, may share a close evolutionaryrelationship with the Bangiales and Aureoumbra84
  4. 4. 516 introns, we used the BLAST program (BLASTNver. 2.1.2; Altschul et al. 1997) to search for other intronsequences with signi®cant sequence identity with theAureoumbra intron. This analysis identi®ed introns intwo strains of the lobose amoeba, Acanthamoeba grini(GenBank S81337, U02540), with a 95% identity over a48-bp sequence. Intrigued by this result, both Acanth-amoeba intron sequences (strains ATCC30731, H37)were included in our intron alignment; and secondarystructure models were developed. The Acanthamoebaintrons also interrupt the SSU rRNA at the 516 site,although they each lack the insertion in the P5b helix(CRW Algal Introns). The Acanthamoeba introns,however, share many secondary structure features incommon with IC1 introns, such as the nine conservedhelical domains shown in Fig. 1A, and can conclusivelybe classi®ed in this group (for details, see Michel andWesthof 1990; Suh et al. 1999).To put the ®nding of the P5b insertion in perspective,of the approximately 1,214 publicly available group Iintrons (350 of which are in subgroup IC1), approxi-mately 772 of them occur in SSU and LSU rRNA; andnearly 600 of these introns interrupt 35 di€erent SSUrRNA sites. The SSU rRNA sites with the largestnumber of introns are 1506 (123 introns), 516 (92 in-trons), and 943 (90 introns). Approximately one-half ofthe 516 introns are in subgroup IC1, whereas the otherhalf are in the IE subgroup. In the entire group I introndata set, only the Bangiales and Aureoumbra IC1516 introns contain the P5b insertion. Thus, this inser-tion associates these two sets of introns and suggests thatthey are speci®cally related to one another. To resolvethe evolutionary relationships of these introns, we re-Table 1 List of organisms containing the inserted helix in the P5b region of the IC1 516 introns. GenBank accession numbers are listed athttp://www.ncbi.nlm.nih.gov/Organism Strain/isolate GenBank accession numberAureoumbra lagunensis U40258Bangia atropurpurea D88387B. fuscopurpurea AF342745Porphyra carolinensis AF133792P. dentata Isolate Koga Fukuoka AB013183P. haitanensis AB015795P. katadae Isolate Kawatana Yamaguchi AB013184P. leucosticta (Baltic) AF342746P. lilliputiana AF136424P. onoi AB015794P. pseudolinearis Isolate Tohaku Tottori AB013185Porphyra sp. Isolate GRB108 AF136420Porphyra sp. Isolate RAK049 AF136425Porphyra sp. Isolate Shimonoseki Yamaguchi AB013182Porphyra sp. Strain BRU107 AF136418, AF136419P. suborbiculata AB015796P. suborbiculata Isolate Kawatana Yamaguchi AB013180P. tenera Isolate Kawaura Kumamoto AB013176P. tenera Isolate Matsukawa Fukushima; strain Matsukawa-3 AB029883P. tenera Isolate Matsukawa Fukushima; strain Matsukawa-4 AB029884P. tenera Isolate Shinwa Kumamoto AB013175P. tenera Isolate Shinwa Kumamoto; strain Kawaura-4 AB029881P. tenera Isolate Shinwa Kumamoto; strain Shinwa-2 AB029879P. tenera Isolate Shinwa Kumamoto; strain Shinwa-3 AB029880P. tenera Strain T-1 D86237P. tenera Strain TU-3 D86236P. tenera Strain T8 AB000964P. tenuipedalis AB015797P. yezoensis D79976P. yezoensis Isolate Hakodate Hokkaido AB013177P. yezoensis Isolate Ogatsu Miyagi AB013178Fig. 1A±D Analysis of intron secondary structures. A Compara-tive secondary structure model for the group IC1 intron atposition 516 in the small subunit (SSU) rRNA of Bangiaatropurpurea (GenBank D88387), drawn according to the conven-tions of Cech et al. (1994). The 5¢ and 3¢ splice junctions are markedwith arrows, as are the locations of the pairing segments P1±P9 andthe internal guide sequence (IGS). A portion of P2.1 marked a isshown as the primary structure and has not been folded in thismodel. B The secondary structure model of the Aureoumbralagunensis intron (U40258). C Comparison of the P5a±c domains inthe 516 introns in B. fuscopurpurea (AF342745), Porphyra leucost-icta (AF342746), and Tetrahymena thermophila (J01235). DAlignment of the P5b region, including the insertion. Base-pairingsin helices are identi®ed in the boxed regions; and the insertion isidenti®ed by the bold boxes. The T. thermophila large subunitrRNA group I intron is used as the reference for thesecomparisons. These secondary structure ®gures, including theT. thermophila, Acanthamoeba spp, and all the red and brown algalintron diagrams we have determined for sequences that are publiclyavailable, can be accessed at our comparative RNA web site(http://www.rna.icmb.utexas.edu/) and at http://www.rna.icmb.ut-exas.edu/PHYLO/ALGINT/b85
  5. 5. constructed the phylogeny of the 516 introns, other SSUrRNA introns, and the host cells that contain them,using more traditional phylogenetic methods.Test for mutational saturation of intronsBefore embarking on phylogenetic analyses with thegroup I intron sequences, we wanted to ®rst establishthat they encoded a signi®cant phylogenetic signal. Todo this, a data set of 62 sequences that spanned the di-versity of IC1 and IE intron SSU rRNA insertion sites(see Figs. 2, 3) was used to determine the extent of su-perimposed substitutions. Uncorrected distances wereplotted versus those corrected with the HKY-85 modelto detect mutational saturation with regard to trans-versions and transitions (see Daugbjerg and Andersen1997; Lopez et al. 1999). In this analysis, the bending ofthe line at higher, uncorrected distances indicates mu-tational saturation (Moritz et al. 1992). The plots pre-sented in Fig. 2 suggest that the group I introns encode asigni®cant phylogenetic signal. Transversions show aminor level of mutational saturation, whereas transi-tions become more rapidly saturated as intron sequencedivergence increases. One obviously cannot comparethese group I introns to ``workhorse molecular markerssuch as rRNA coding regions, in terms of utility across abroad phylogenetic spectrum. The introns often have 4±10 times greater rates of sequence divergence (Bhat-tacharya et al. 1996b). Nonetheless, these sequencescontain a phylogenetic signal that can be used to esti-mate a group I intron evolutionary tree.Phylogenetic analysis of introns and host cellsTo test the monophyly of the Bangiales and Aureoumbra516 introns developed by the comparison of the intronsecondary structure features, we ®rst estimated a phy-logeny of the IC1 and IE introns from a wide variety ofalgae and protists. The preliminary distance analysis,which included 322 IC1 and 101 IE introns in nuclearSSU and LSU rRNA demonstrated a speci®c evolu-tionary relationship between the Acanthamoeba, Aureo-umbra, and red algal 516 introns within the IC1 cluster(data not shown). This group was distinct from all otherIC1 introns and all other publicly available introns atthe 516 site, i.e., those in Naegleria spp and the IE516 introns. Knowing this, we reduced the data set to62 SSU rRNA introns for a more detailed analysis.The distance tree (HKY-85 model) inferred from thereduced data set of group I introns provides strongsupport (96% bootstrap value) for the monophyly of theAcanthamoeba, Bangiales, and Aureoumbra IC1 intronsat position 516, the early divergence of the Acanth-amoeba introns within this clade (86% bootstrap sup-port), and the distant evolutionary relationship of theseintrons to the 516 introns in the IE group (Fig. 3). TheNaegleria spp. 516 introns are positioned at the base ofthe IC1 516 intron clade without bootstrap support. Itis, therefore, unclear whether the Naegleria introns aremonophyletic with the other 516 introns in subgroupIC1. Interestingly, the IC1 subgroup also includes the1506 SSU rRNA intron in the Bangiales and one mem-ber of the derived red algal subclass Florideophycidae(Hildenbrandia rubra; Ragan et al. 1993). The BangialesFig. 2 Plots of uncorrectedpair-wise substitutions ver-sus values corrected formultiple hits (HKY-85model; Hasegawa et al.1985) for the group I intronsincluded in the phylogenyshown in Fig. 3. Transitionsand transversions are studiedseparately. In this analysis,the bending of the line athigher uncorrected distancesindicates mutational satura-tion (Moritz et al. 1992)Table 2 Base-pair frequencies for the inserted helix in P5b found inthe red and brown algae. Data are presented for the 31 publiclyavailable sequences containing this inserted helix (30 red algae,1 brown alga, see Table 1). Base-pairs refer to the ``insertionnumbers shown in Fig. 1D. For each base-pair, ± represents acombination which is not present at that base-pair. In the Notes, A,B, and C refer to the marked base-pairs in Fig. 1DBase-pair G:C C:G A:U U:A G:U U:G Notes03:18 96.8% ± 3.2% ± ± ± A04:17 16.1% ± 71.0% ± 12.9% ±05:16 38.7% ± 6.5% 3.2% 51.6% ± B06:15 ± 96.8% ± ± ± 3.2%07:14 96.8% ± ± 3.2% ± ± C08:13 3.2% ± 96.8% ± ± ±86
  6. 6. are postulated to be the sister group of the Florid-eophycidae (Oliveira and Bhattacharya 2000; MuÈ lleret al. 2001b), a hypothesis that is supported by thepresence of the 1506 intron in only these red algae. Wepresume here that the common ancestor of the Bangi-ales/Florideophycidae likely contained the 1506 intron.Bootstrap-distance analysis of the 516 introns with theGTR+G+I model provided similar results to theHKY-85 model. The monophyly of the IC1 introns atposition 516 of Acanthamoeba, Bangiales, and Aureo-umbra was strongly supported (96%), although thereFig. 3 Phylogeny of IC1 and IE introns interrupting the nuclearSSU rRNA of di€erent eukaryotes. This tree was built withneighbor-joining, using a matrix of corrected distances (HKY-85model). Bootstrap values (1,000 replications) are shown at thebranches. The values shown below the branches result from adistance bootstrap analysis (1,000 replications), using theGTR+G+I model. Only values ³50% are shown. The rRNAintron insertion sites are indicated. The tree has been rooted on thebranch leading to the IE introns. GenBank accession numbers areshown (in parentheses) for each intron sequence. The group IC1introns at position 516 that contain the P5b insertion are shown inbold87
  7. 7. was again no support found for the inclusion of theNaegleria spp. 516 introns within this clade.Given this phylogenetic framework for the group Iintrons at SSU rRNA site 516, we studied the phylogenyof the host cells containing these introns. Analysis of theSSU rRNA in red algae and stramenopiles showed thattaxa containing the 516 intron are evolutionarily dis-tantly related (Fig. 4). Recent phylogenetic analyses ofmore extensive data sets also suggested that the red algaeshare a most recent common ancestry with the greenalgae and land plants (Burger et al. 1999; Moreira et al.2000). None of the green algae and land plants, however,contain an IC1 intron at position 516. The strameno-piles are often allied with alveolate protists, Acanth-amoeba spp. are members of the Mycetozoa (a sistergroup of fungi/animals), and Naegleria (Heterolobosea)is a sister group of the Euglenozoa (e.g., Baldauf et al.2000; Van de Peer et al. 2000). There is, therefore, nodirect evidence to support a close evolutionary rela-tionship between red algae, brown algae, acanthame-bids, and Naegleria spp. The most parsimoniousexplanation for the close relationship of the IC1516 introns, whereas the host cells containing them arenot closely related, is that the intron was independentlyinserted into the homologous SSU rRNA site in theBangiales, Aureoumbra, Acanthamoeba, and Naegleria(i.e., introns late). An observation that is consistent withthis hypothesis is that, of 373 stramenopile SSU rRNAcoding regions that have been determined (organismnames and GenBank accession numbers available atCRW-algal introns), only the relatively late-divergingAureoumbra (Pelagophyceae) contains the IC1 516 in-tron (see Fig. 4). To explain this distribution in terms ofintrons early, one would need to invoke the non-parsi-monious hypothesis of widespread intron loss in allstramenopiles, except for the relatively late-divergingAureoumbra (see Fig. 4). Furthermore, the 516 intron islimited to and sporadically distributed in the Bangialeswithin the Rhodophyta; and it is only found in threeclosely related acanthamebids of about 61 nuclear SSUrRNA sequences of Acanthamoeba strains/species thatare currently available in GenBank (organism namesand GenBank accession numbers available at CRWAlgal Introns).Another fascinating possibility for the origin of theAureoumbra IC1 516 intron is through symbiotic geneFig. 4 Phylogeny of selected red algae (Rhodophytes) and stra-menopiles, based on SSU rRNA sequence comparisons. This treewas built with neighbor-joining, using a matrix of correcteddistances (HKY-85 model). Bootstrap values (1,000 replications)are shown above the branches on the left of the slash-marks. Thevalues shown on the right of the slash-marks result from a distancebootstrap analysis (1,000 replications) using the GTR+G+Imodel. Bootstrap values shown below the branches are derivedfrom an unweighted maximum parsimony analysis (1,000 replica-tions). Only values ³50% are shown. The Bangiales and Aureoum-bra are shown in bold. GenBank accession numbers are indicated(in parentheses) for each SSU rRNA sequence88
  8. 8. transfer. It is now generally believed that the plastid ofphotosynthetic stramenopiles has arisen through a singlesecondary (eukaryotic) symbiosis involving a red alga(Douglas et al. 1991; Bhattacharya and Medlin 1995,1998; Daugbjerg and Andersen 1997; Oliveira andBhattacharya 2000). It is, therefore, conceivable that theAureoumbra intron may trace its origin to an intron inthe rRNA gene of the red algal secondary symbiont. Theintron could have been transferred from the symbiontnucleus to that of the stramenopile prior to the elimi-nation of the red algal nuclear genome. However, anobservation that argues against this scenario is that, asstated above, only Aureoumbra (of the many nuclearSSU rRNA sequences that have been determined fromphotosynthetic stramenopiles) contains an IC1 intron atposition 516 in the SSU rRNA. If the intron was presentin the photosynthetic ancestor of all stramenopiles, thenit must have been lost in all of these taxa, except for therelatively late-diverging Aureoumbra.In conclusion, our analyses provide clear insights intothe origin of the IC1 introns at the 516 site of SSUrRNA. The host-cell phylogenetic data suggest that it isunlikely that this intron was vertically inherited from thecommon ancestor of the red algae, brown algae,acanthamebids, and Naegleria spp., because these lin-eages do not share a speci®c evolutionary relationship.This hypothesis is supported by the restricted distribu-tion of the intron in each of the host-cell lineages. Theintron phylogeny shows that the IC1 516 intron is evo-lutionarily distantly related to the IE 516 intron,supporting independent intron insertions into the ho-mologous SSU rRNA site. The intron tree also providesstrong support for the monophyly of the IC1 516 introncontaining the insertion in helix P5b within the Bangialesand Aureoumbra. This group also includes the Acanth-amoeba 516 intron that lacks the insertion. The Naegleriaspp 516 introns are only distantly related to this clade.Taken together, these data are consistent with thesecond scenario that the Bangiales and Aureoumbra IC1516 introns were acquired after the separation of the redand brown algal lineages. We suggest this intron hasbeen inserted independently into the SSU rRNA of theBangiales, into a single stramenopile brown alga,Aureoumbra, and into derived members of the generaAcanthamoeba and Naegleria. Finally, our data suggestnot only that lateral transfer may establish new intronlineages in di€erent organisms but also that, over time,these sequences may evolve into distinct secondarystructure variants (i.e., containing the P5b helix inser-tion), perhaps re¯ecting selection pressure to adapt tothe novel splicing environment. The IC1 introns at po-sition 516 in the SSU rRNA may, therefore, be a valu-able example of how novel intron groups are createdthrough the dual process of lateral transfer followed byrapid secondary structure evolution. In this regard, theP5b insertion has unambiguously marked the intron inthe Bangiales and Aureoumbra and has allowed us toidentify these homologous introns in evolutionarily di-vergent protists. Detailed analysis of group I intronsecondary structure promises, therefore, to provide animportant source of comparative data to elucidateintron origin and evolution.Acknowledgements D.B. was supported by grants from the CarverFoundation. R.R.G. was supported by research grant NIH-GM48207 and start-up funds from the University of Texas, Insti-tute for Cellular and Molecular Biology.ReferencesAbramovitz DL, Pyle AM (1997) Remarkable morphologicalvariability of a common RNA folding motif: the GNRAtetraloop±receptor interaction. J Mol Biol 266:493±506Altschul SF, Madden TL, SchaÈ €er AA, Zhang J, Zhang Z, MillerW, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a newgeneration of protein database search programs. Nucleic AcidsRes 25:3389±3402Baldauf S, Roger AJ, Wenk-Siefert I, Doolittle WF (2000) Akingdom-level phylogeny of eukaryotes based on combinedprotein data. Science 290:972±977Bhattacharya D (1998) The origin and evolution of protist group Iintrons. Protist 149:113±122Bhattacharya D, Medlin L (1995) The phylogeny of plastids: areview based on comparisons of small subunit ribosomal RNAcoding regions. J Phycol 31:489±498Bhattacharya D, Medlin L (1998) Algal phylogeny and the originof land plants. Plant Physiol 116:9±15Bhattacharya D, Surek B, RuÈ sing M, Damberger S, Melkonian M(1994) Group I introns are inherited through common ancestryin the nuclear-encoded rRNA of Zygnematales (Charophy-ceae). Proc Natl Acad Sci USA 91:9916±9920Bhattacharya D, Friedl T, Damberger S (1996a) Nuclear-encodedrDNA group I introns: origin and phylogenetic relationships ofinsertion site lineages in the green algae. Mol Biol Evol 13:978±989Bhattacharya D, Damberger S, Surek B, Melkonian M (1996b)Primary and secondary structure analyses of the rDNA group Iintrons of the Zygnematales (Charophyta). Curr Genet 29:282±286Burger G, Saint-Louis D, Gray MW, Lang BF (1999) Completesequence of the mitochondrial DNA of the red alga Porphyrapurpurea: cyanobacterial introns and shared ancestry of red andgreen algae. Plant Cell 11:1675±1694Cate JH, Gooding AR, Podell E, Zhou K, Golden BL, KundrotCE, Cech TR, Doudna JA (1996) Crystal structure of a group Iribozyme domain: principles of RNA packing. Science273:1678±1685Cech TR (1985) Self-splicing RNA: implications for evolution. IntRev Cytol 93:3±22Cech TR, Damberger SH, Gutell RR (1994) Representation of thesecondary and tertiary structure of group I introns. Nat StructBiol 1:273±280Damberger SH, Gutell RR (1994) A comparative database ofgroup I intron structures. Nucleic Acid Res 22:3508±3510Daugbjerg N, Andersen RA (1997) Phylogenetic analyses of therbcL sequences from haptophytes and heterokont algae suggesttheir chloroplasts are unrelated. Mol Biol Evol 14:1242±1251Doherty EA, Doudna JA (2000) Ribozyme structures and mecha-nisms. Annu Rev Biochem 69:597±615Doherty EA, Herschlag D, Doudna JA (1999) Assembly of anexceptionally stable RNA tertiary interface in a group I ribo-zyme. Biochemistry 38:2982±2990Douglas SE, Murphy CA, Spencer DF, Gray MW (1991) Molec-ular evidence that cryptomonad algae are evolutionary chima-eras of two phylogenetically distinct eukaryotes. Nature350:148±151Felsenstein J (1985) Con®dence intervals on phylogenies: anapproach using the bootstrap. Evolution 39:783±79189
  9. 9. Friedl T, Besendahl A, Pfei€er P, Bhattacharya D (2000) Thedistribution of group I introns in lichen algae suggests thatlichenization facilitates intron lateral transfer. Mol PhylogenetEvol 14:342±352Gutell RR (1996) Comparative sequence analysis and the structureof 16S and 23S rRNA. In: Zimmermann RA, Dahlberg AE(eds) Ribosomal RNA: structure, evolution, processing, andfunction in protein biosynthesis. CRC Press, Boca Raton,pp 111±128Hasegawa M, Kishino H, Yano T (1985) Dating of the human±apesplitting by a molecular clock of mitochondrial DNA. J MolEvol 22:160±174Hedenstierna KO, Siefert JL, Fox GE, Murgola EJ (2000) Co-conservation of rRNA tetraloop sequences and helix lengthsuggests involvement of the tetraloops in higher-order interac-tions. Biochimie 82:221±227Kruger K, Grabowski PJ, Zaug AJ, Sands J, Gottschling DE, CechTR (1982) Self-splicing RNA: autoexcision and autocyclizationof the ribosomal RNA intervening sequence of Tetrahymena.Cell 31:147±157Lopez P, Forterre P, Philippe H (1999) The root of the tree of life inthe light of the covarion model. J Mol Evol 49:496±508Michel F, Westhof E (1990) Modelling the three-dimensional ar-chitecture of group I catalytic introns based on comparativesequence analysis. J Mol Biol 216:585±610Moreira D, Le Guyader H, Philippe H (2000) The origin of redalgae and the evolution of chloroplasts. Nature 405:69±72Moritz C, Schneider CJ, Wake DB (1992) Evolutionary relation-ships within the Ensatina eschscholtzii complex con®rm the ringspecies interpretation. Syst Biol 41:273±291MuÈ ller KM, Sheath RG, Vis ML, Crease J, Cole KM (1998) Bi-ogeography and systematics of Bangia (Bangiales, Rhodophyta)based on the rubisco spacer, rbcL gene and 18S rRNA genesequences and morphometric analyses. 1. North America.J Phycol 37:195±207MuÈ ller KM, Cannone JJ, Gutell RR, Sheath RG (2001a) A struc-tural and phylogenetic analysis of the Group IC1 introns in theorder Bangiales (Rhodophyta). Mol Biol Evol 18(8) (in press)MuÈ ller KM, Oliveira MC, Sheath RG, Bhattacharya D (2001b)Ribosomal DNA phylogeny of the Bangiophycidae (Rhodo-phyta) and the origin of secondary plastids. Am J Bot 88(8) (inpress)Nishida K, Suzuki S, Kimura Y, Nomura N, Fujie M, Yamada T(1998) Group I introns found in Chlorella viruses: biologicalimplications. Virology 15:319±326Oliveira M, Bhattacharya D (2000) Phylogeny of the Bang-iophycidae (Rhodophyta) and the secondary endosymbioticorigin of algal plastids. Am J Bot 87:482±492Posada D, Crandall KA (1998) Modeltest: testing the model ofDNA substitution. Bioinformatics 14:817±818Ragan MA, Bird CJ, Rice EL, Singh RK (1993) The nuclear 18Sribosomal RNA gene of the red alga Hildenbrandia rubracontains a group I intron. Nucleic Acids Res 21:3898RodrõÂguez F, Oliver JF, MarõÂn A, Medina JR (1990) The generalstochastic model of nucleotide substitution. J Theor Biol142:485±501Suh S-O, Jones KG, Blackwell M (1999) A group I intron in thenuclear small subunit rRNA gene of Cryptendoxyla hypophloia,an ascomycetous fungus: evidence for a new major class ofgroup I introns. J Mol Evol 48:493±500Swo€ord DL (2001) PAUP*: Phylogenetic analysis using parsi-mony, ver. 4.0b8. Smithsonian Institution, Washington, D.C.Turmel M, Coà te V, Otis C, Mercier J-P, Gray MW, Lonergan KM,Lemieux C (1995) Evolutionary transfer of ORF-containinggroup I introns between di€erent subcellular compartments(chloroplast and mitochondrion). Mol Biol Evol 12:533±545Van de Peer Y, Baldauf SL, Doolittle WF, Meyer A (2000) Anupdated and comprehensive rRNA phylogeny of (crown) eu-karyotes based on rate-calibrated evolutionary distances. J MolEvol 51:565±576Watanabe KI, Ehara M, Inagaki Y, Ohama T (1998) Distinctiveorigins of group I introns in the COXI genes of three greenalgae. Gene 213:1±7Woese CR, Winker S, Gutell RR (1990) Architecture of ribosomalRNA: constraints on the sequence of tetra-loops. Proc NatlAcad Sci USA 87:8467±847190

×