1654Mol. Biol. Evol. 18(9):1654–1667. 2001᭧ 2001 by the Society for Molecular Biology and Evolution. ISSN: 0737-4038A Structural and Phylogenetic Analysis of the Group IC1 Introns in theOrder Bangiales (Rhodophyta)Kirsten M. Mu¨ller,*1 Jamie J. Cannone,† Robin R. Gutell,† and Robert G. Sheath**Department of Botany and Dean’s Ofﬁce, University of Guelph, Guelph, Ontario, Canada; and †Institute of Cellular andMolecular Biology, University of Texas at AustinOur previous study of the North American biogeography of Bangia revealed the presence of two introns insertedat positions 516 and 1506 in the nuclear-encoded SSU rRNA gene. We subsequently sequenced nuclear SSU rRNAin additional representatives of this genus and the sister genus Porphyra in order to examine the distribution,phylogeny, and structural characteristics of these group I introns. The lengths of these introns varied considerably,ranging from 467 to 997 nt for intron 516 and from 509 to 1,082 nt for intron 1506. The larger introns containedlarge insertions in the P2 domain of intron 516 and the P1 domain of intron 1506 that correspond to open readingframes (ORFs) with His-Cys box homing endonuclease motifs. These ORFs were found on the complementarystrand of the 1506 intron in Porphyra fucicola and P. umbilicalis (HG), unlike the 516 intron in P. abbottae, P.kanakaensis, P. tenera (SK), Bangia fuscopurpurea (Helgoland), and B. fuscopurpurea (MA). Frameshifts werenoted in the ORFs of the 516 introns in P. kanakaensis and B. fuscopurpurea (HL), and all ORFs terminatedprematurely relative to the amino acid sequence for the homing endonuclease I-Ppo I. This raises the possibilitythat these sequences are pseudogenes. Phylogenies generated using sequences of both introns and the 18S rRNAgene were congruent, which indicated long-term immobility and vertical inheritance of the introns followed bysubsequent loss in more derived lineages. The introns within the ﬂorideophyte species Hildenbrandia rubra (position1506) were included to determine relationships with those in the Bangiales. The two sequences of intron 1506analyzed in Hildenbrandia were positioned on a well-supported branch associated with members of the Bangiales,indicating possible common ancestry. Structural analysis of the intron sequences revealed a signature structuralfeature in the P5b domain of intron 516 that is unique to all Bangialean introns in this position and not seen inintron 1506 or other group IC1 introns.IntroductionMany eukaryotic genes have their coding regionsinterrupted by intervening sequences or introns. GroupI introns represent a family of RNA molecules with aspeciﬁc higher-order structure and the ability to catalyzetheir own excision by a common splicing mechanism(Cech 1990). Group I introns are divided into 11 sub-groups based on conserved primary- and secondary-structure elements (Michel and Westhof 1990) and havebeen reported in the 18S rRNA genes of numerous or-ganisms (Johansen, Muscarella, and Vogt 1996), such asfungi (Takashima and Nakase 1997), amoebae (De Jonck-heere 1994; Schroeder-Diedrich, Fuerst, and Byers 1998),and green algae (Wilcox et al. 1992; VanOppen, Olsen,and Stam 1993; Bhattacharya et al. 1994, 1996; Bhatta-charya 1998). Group I introns have also been reported inseveral red algae, including Hildenbrandia rubra (Raganet al. 1993) and particularly members of the order Ban-giales (Porphyra spiralis var. amplifolia and Bangia fus-copurpurea [as Bangia atropurpurea] [Oliveira and Ra-gan 1994]; Porphyra miniata, P. purpurea, P. linearis[Oliveira et al. 1995]; Bangia spp. [Mu¨ller et al. 1998]).Stiller and Waaland (1993) reported cryptic diversity ina number of species of Porphyra after ﬁnding large in-sertions in the nuclear SSU rRNA genes of a number oftaxa. Oliveira et al. (1995) speculated that these introns1 Present address: Department of Biology, University of Waterloo,Waterloo, Ontario, Canada.Key words: Group IC1 introns, Bangiales, phylogeny.Address for correspondence and reprints: Robert G. Sheath,Dean’s Ofﬁce and Department of Botany, University of Guelph,Guelph, Ontario, Canada N1G 2W1. E-mail: firstname.lastname@example.org be a means to differentiate among geographic en-tities within the genus Porphyra. During the course ofa study of the biogeography of Bangia in North Amer-ica, we reported the presence of two introns in positions516 and 1506 (Escherichia coli numbering) in the nu-clear SSU rRNA gene (Mu¨ller et al. 1998). Subsequentanalysis of additional nuclear SSU rRNA gene sequenc-es of Bangia and Porphyra revealed a rich source ofthese introns, which are also variable in occurrence. Var-iable occurrence of group I introns can be simply ex-plained by two models: intron insertion or intron dele-tion (Burke 1988). The ﬁrst model, intron insertion, ishypothesized to have begun with a gene devoid of in-trons followed by subsequent insertion of one or moreintrons (Burke 1988). The deletion model proposes agene initially containing one or more introns, afterwhich precise deletion of these introns occurs; that is,nonmobile introns are destined to be lost over time ifthey cannot reinfect homologous sites. Burke (1988)proposed that the variable occurrence of introns in genesmay be the result of a combination of these two models.High sequence similarity between introns is consistentwith descent from an ancestral intron that was initiallyacquired and vertically inherited (Schroeder-Diedrich,Fuerst, and Byers 1998). In addition, congruent intronand rRNA gene phylogenies would provide support forthis theory (Bhattacharya et al. 1994, 1996; Friedl et al.2000). The present study provides a large data set thatwill allow us to test these hypotheses in context withthe phylogeny of the rhodophyte order the Bangiales.Materials and MethodsCollections of Bangia and Porphyra collected formolecular analyses and obtained from GenBank for the
Group IC1 Introns in the Bangiales 1655present study are listed in table 1 Filaments collectedfor DNA analysis were cleaned of visible epiphytes, ba-ses were removed to prevent contamination, and thespecimens were stored atϪ20ЊC. Samples were groundin liquid nitrogen, and the DNA was extracted accordingto the protocol outlined by Saunders (1993) with mod-iﬁcations given in Vis and Sheath (1997). Ampliﬁcationand sequencing of the nuclear SSU rRNA gene for bothgenera are as outlined in Mu¨ller et al. (1998).Intron AmpliﬁcationThe intron in position 516 was ampliﬁed in thelarge fragment of the nuclear SSU rRNA gene as de-scribed in Mu¨ller et al. (1998) using the primers G02.1(5Ј-CGA TTC CGG AGA GGG AGC CTG-3Ј) andG15.1 (5Ј-CTT GTT ACG ACT TCT CCT TCC-3Ј).These fragments were then sequenced in both directionsusing internal primers ﬂanking the intron. The intron inposition 1506 was ampliﬁed using primers GO6 (5Ј-GTT GGT GGT GCA TGG CCG TTC-3Ј) and G07 (5Ј-TCC TTC TGC AGG TTC ACC TAC-3Ј) from Saun-ders and Kraft (1994) as follows: initial denaturation at95ЊC for 2 min, 35 cycles of denaturation for 1 min at93ЊC, primer annealing at 50ЊC for 1 min, and extensionfor 2 min at 72ЊC, followed by a ﬁnal extension time of3 min at 72ЊC. This intron was sequenced in both di-rections using an internal primer at the 5Ј end of theintron and the primer G07 (sequences for internal prim-ers can be obtained from R.G.S.). All products wereprepared for sequencing and sequenced using the pro-tocols outlined in Mu¨ller et al. (1998).RNA Extraction/cDNA Production in the BangialesRT-PCR was used to determine if introns were pre-sent in mature RNA or excised from all collections ofBangia and Porphyra listed in table 1 and the Bangiacollections utilized in Mu¨ller et al. (1998). RNA wasextracted following the procedure previously outlinedfor DNA extraction. Following this, 12 l of extract wasplaced in a clean tube, to which 3 l of 25 mM MgCl2and 1 l of DNAase was added. This mixture was thenleft at room temperature for 2 h, after which 5 l wasloaded in a 0.5% agarose gel to determine if all DNAhad been digested within the extract. If this was the case,the extract was then immediately placed in the freezerat Ϫ20ЊC. RT-PCR using Titan by Boehringer Mann-heim was used to amplify the RNA. The primers G02.1and G09 (5Ј-ATC CAA GAA TTT CAC CTC TG-3Ј)and G15.R (5Ј-GGA AGG AGG AGT CGT AAC AAG-3Ј) and G07 were used to amplify the region containingthe intron in positions 516 and 1506, respectively (Saun-ders and Kraft 1994; Mu¨ller et al. 1998). Following theprotocols for RT-PCR outlined by Boehringer Mann-heim, the reaction consisted of master mix 1 (1 l ofRNA template; 20 M of each primer; 40 mM of dNTP;2.5 l DTT solution; dH2O) and master mix 2 (14 ldH2O; 5 ϫ RT-PCR buffer with MgCL2; 1 l enzymemix). These two master mixes were combined andplaced in a 0.5-l thin-walled PCR tube on ice. Ampli-ﬁcation was performed as follows: initial denaturationat 94ЊC for 2 min; 10 cycles of denaturation at 94ЊC for30 s, primer annealing at 50ЊC for 30 s, and extensionfor 2 min at 68ЊC; 25 cycles of denaturation at 94ЊC for30 s, primer annealing at 50ЊC for 30 s, extension for 2min at 68ЊC, plus cycle elongation for 5 s for each cycle;and a ﬁnal extension of 7 min at 68ЊC. Sequencing ofthe ampliﬁed regions was carried out as previouslyoutlined.Alignment of Nuclear SSU rRNA Genes and GroupIC1 IntronsThe nuclear SSU rRNA gene and group IC1 intronsequences in the present study were incorporated intolarge alignments that spanned a large taxonomic rangein order to ensure optimum alignment. The new se-quences were manually aligned with the alignment ed-itor AE2 (developed by T. Macke; see Larsen et al.1993) on the basis of sequence similarity and a previ-ously established eukaryotic secondary-structure model(Gutell 1993). The alignments were then subjected to aprocess of comparative sequence analysis (Gutell et al.1985). This process consisted of searching for compen-sating base changes using computer programs developedwithin the Gutell Laboratory (University of Texas atAustin, http://www.rna.icmb.utexas.edu/; discussed inGutell et al. 1985) and using the subsequent informationto infer additional secondary-structural features. This re-ﬁned alignment was reanalyzed and the entire process wasrepeated until the proposed structures were entirely com-patible with the alignment. Secondary-structure diagramswere generated with the computer program XRNA (de-veloped by B. Weiser and H. Noller, University of SantaCruz). Individual secondary-structure diagrams will beavailable at http://www.rna.icmb.utexas.edu/, and align-ments can be obtained from K.M.M.Analysis of Nuclear SSU rRNA Gene and Group IIntron SequencesAll analyses on both the nuclear SSU rRNA geneand the group I introns were carried out using only well-aligned regions of the sequences. Parsimony analyses onthe two introns were carried out with PAUP 3.1.1 (Swof-ford 1993) with a heuristic search under the constraintsof random sequence addition (100 replicates), steepestdescent, and tree bisection-reconnection (TBR) branchswapping. The data were then subjected to bootstrap re-sampling (1,000 replicates). Analyses of the intronswere carried out as follows: (1) the alignment gaps weretreated as missing data, and (2) the alignment gaps werecoded as independent single evolutionary events (inser-tion or deletion) based on secondary-structure models(Damberger and Gutell 1994). Group IC1 introns inChlorella ellipsoidea (GenBank accession number A:X63520, B: D13324) were used as outgroups in the phy-logenetic analyses. Parsimony trees could not be deter-mined for the nuclear SSU rRNA gene due to the largenumber of taxa and limited computational capacity;however, neighbor-joining trees were calculated for thisdata set. PHYLIP (Felsenstein 1993) was used to con-struct neighbor-joining trees for both the introns and the
Group IC1 Introns in the Bangiales 1657Table 1ContinuedCOLLECTION COLLECTION INFORMATION/REFERENCEGENBANK ACCESSION NOS.18S rRNA Intron 516 Intron 1506Porphyra rediviva . . . . . . . . . . . . . . . . . . . . . . Mu¨ller et al. (unpublished) for 18S rRNA gene AF175544 — AF172600Porphyra spiralis var. amplifolia B . . . . . . . . Oliveira and Ragan (1994) — — L26175P. spiralis var. amplifolia D . . . . . . . . . . . . . . Oliveira and Ragan (1994) — — L26176P. spiralis var. amplifolia R . . . . . . . . . . . . . . Oliveira and Ragan (1994) L26177 — L26177Porphyra suborbiculata (KY) . . . . . . . . . . . . . Kunimoto et al. (unpublished) AB013180 AB013180 —Porphyra tenera (KK) . . . . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) AB013176 AB013176 AB013176P. tenera (SK) . . . . . . . . . . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) AB013175 AB013175 AB013175P. tenera (T-1) . . . . . . . . . . . . . . . . . . . . . . . . . Yamazaki et al. (unpublished) D86237 D86237 —P. tenera (TU-3) . . . . . . . . . . . . . . . . . . . . . . . Yamazaki et al. (unpublished) D86236 D86236 —P. tenera (T8) . . . . . . . . . . . . . . . . . . . . . . . . . . Yamazaki et al. (unpublished) AB000964 AB000964 —Porphyra torta . . . . . . . . . . . . . . . . . . . . . . . . . Mu¨ller et al. (unpublished) for 18S rRNA gene AF175552 AF172579 ?Porphyra umbilicalis (HF) . . . . . . . . . . . . . . . Ragan et al. (1994) L26202 — —P. umbilicalis (HG) (HG ϭ Hallig Gro¨de,Germany) . . . . . . . . . . . . . . . . . . . . . . . . . . . Mu¨ller et al. (unpublished) for 18S rRNA gene AF175549 — AF172602P. umbilicalis (NJ) . . . . . . . . . . . . . . . . . . . . . . Mu¨ller et al. (unpublished) for 18S rRNA gene AF175553 AF172573 —P. umbilicalis (NM) . . . . . . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) AB013179 — —Porphyra yezoensis . . . . . . . . . . . . . . . . . . . . . Yamazaki et al. (unpublished) D79976 D79976 —P. yezoensis (HH) . . . . . . . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) AB013177 AB013177 —P. yezoensis (OM) . . . . . . . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) AB013178 AB013178 AB013178P. yezoensis (OG-1) . . . . . . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) (ITS) — ? AB017078P. yezoensis (OG-4) . . . . . . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) (ITS) — ? AB017081P. yezoensis narawaensis . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) (ITS) — ? AB017083P. yezoensis (NA-2) . . . . . . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) (ITS) — ? AB017075P. yezoensis (NA-4) . . . . . . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) (ITS) — ? AB017076Porphyra sp. (Brest) . . . . . . . . . . . . . . . . . . . . Mu¨ller et al. (unpublished) for 18S rRNA gene AF175548 — AF172601Porphyra sp. (Marseille) . . . . . . . . . . . . . . . . . Mu¨ller et al. (unpublished) for 18S rRNA gene AF175546 — AF172597Porphyra sp. (SHH) . . . . . . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) — ? AB017077Porphyra sp. (SY) . . . . . . . . . . . . . . . . . . . . . . Kunimoto et al. (unpublished) AB013182 AB013182 AB013182Porphyra sp. (Wales) . . . . . . . . . . . . . . . . . . . . Mu¨ller et al. (unpublished) for 18S rRNA gene AF175554 AF172574 —NOTE.—The abbreviations UTEX, CCAP, and CCMP are for culture collections. Other abbreviations for taxa not noted were obtained from the GenBank ﬁleand represent those assigned by the original authors or the collector of the material.a Intron not present.b Intron not sequenced, or presence or absence unknown.nuclear SSU rRNA genes using a matrix of distance val-ues estimated according to the Kimura two-parametermodel (Kimura 1980) with a transition/transversion ratioof 2.0 and a single-category substitution rate, as well asthe Jukes and Cantor (1969) model (intron data setonly). These data sets were also subjected to bootstrapresampling. Maximum-likelihood analysis was carriedout using the puzzle function in PAUP (version 4.0 beta2a) as described by Strimmer and von Haeseler (1996)with 1,000 puzzling steps. The nuclear SSU rRNA genesequence analyses of the Bangiales were run using Er-ythrotrichia carnea (L26189) and Erythrocladia sp.(L26188) as outgroups. These taxa were determined tobe basal to the Bangiales and have been previously usedas outgroup taxa for this order (Ragan et al. 1994; Oliv-eira et al. 1995; Mu¨ller et al. 1998).ResultsSequence data for all intron sequences were sub-mitted to GenBank, and accession numbers are given intable 1. Presence and absence of introns within the nu-clear SSU rRNA gene Bangiales is depicted in ﬁgure 4.Letters and numbers behind taxonomic names refer tovarious collections that are listed in table 1, of whichsome represent standard state, provincial, or countrycodes, whereas others are arbitrary numbers or initialsof the author of the particular sequence. Due to the dif-ﬁculty in delineating species in Bangia (Mu¨ller et al.1998) and the lack of a deﬁnitive global key for Por-phyra species, identiﬁcation of species that are used inthe molecular analyses should be considered tentative.Thus, many of the analyses in the present study willfocus more on the relationship among the intron se-quences and nuclear SSU rRNA gene within the twogenera than on morphological species within each ge-nus, other than for characters that would not be disputed(e.g., monostromatic vs. distromatic). With respect toBangia, based on the ﬁndings in Mu¨ller et al. (1998),all freshwater collections (AT22, BI12, GL, IR, IT, andNL) should be classiﬁed as B. atropurpurea, while col-lections from marine locations have provisionally beengiven the name B. fuscopurpurea until they can be dif-ferentiated morphologically.Using RT-PCR it was determined that neither in-tron is present in the mature rRNA, indicating that theseintrons are excised (not shown). Thirty-three nuclearSSU rRNA gene sequences from Bangia and 52 fromPorphyra were examined for the presence of either in-tron in position 516 (intron 516) (E. coli numbering sys-tem) and position 1506 (intron 1506). Of these 52 Por-phyra specimens, there were 39 complete sequences ofthe nuclear SSU rRNA gene; the remainder consisted of
1658 Mu¨ller et al.FIG. 1.—Phylogenetic conservation of Rhodophyta (Bangiophy-ceae) group I (IC1) introns (18S rRNA positions 516 and 1506) su-perimposed onto the Bangia sp. (IR) group I (IC1) intron secondarystructure (18S rRNA position 1506). Positions with a nucleotide inϾ95% of sequences are shown: ACGU—95ϩ% conserved; acgu—90%–95% conserved; ●—80%–90% conserved; o—Ͻ80% conserved.Otherwise, the regions are represented by arcs. Positions shown insmall square boxes and arcs with dashed lines represent features thatdistinguish the 516 and 1506 introns (described in table 3). The exonsequences that ﬂank the intron are indicated with x’s.species for which the nuclear SSU rRNA gene could notbe obtained (difﬁculty in obtaining sequences, or samplewas not sequenced) but sequences of some introns wereavailable. Among the 33 collections of Bangia, 21 weredetermined to have intron 516 (64%), and 13 were ob-served to have intron 1506 (39%). Only 10 (30%) col-lections contained neither intron, and 11 (33%) collec-tions had both introns (ﬁg. 4). Within Porphyra, 20(40%) of 50 nuclear SSU rRNA gene sequences weredetermined to have intron 516, and 26 (52%) containedintron 1506. Only 14 (28%) Porphyra sequences hadeither intron, and 10 (20%) contained both introns (ﬁg.4).Structural Characteristics of Group IC1 Introns in theBangialesAll of the rRNA introns in Bangia and Porphyrawere determined to belong to the CI subgroup of thegroup I introns based on characteristic primary- and sec-ondary-structure features (ﬁg. 1) (Michel and Westhof1990; Damberger and Gutell 1994). The lengths of theseintrons within the Bangiales varied considerably, rang-ing from 467 to 998 nt for intron 516 and ranging from509 to 1,082 nt for intron 1506 (table 1). Figure 1 is aconsensus diagram of all of the group IC1 introns in theBangiales. Both introns were observed to have large in-sertions within the P1 (12–567 nt), P2 (62–553 nt), P5(3–91 nt), P6b (10–84 nt), and P9.2 (5–53 nt) domains(ﬁg. 1). The large insertions in the P1 and P2 domainswere found in introns 1506 and 516, respectively (ﬁg.1), whereas the other insertions (P5, P6b, and P9) weresimilar in size between both the 516 and the 1506 in-trons. The majority of the base pairs in the P4 and P7helices are highly conserved (Ͼ95%) between both setsof introns, whereas the remaining helices are less con-served. The single positions and base pair ‘‘signatures’’that distinguish the 516 and 1506 introns are shown assquare boxes and dashed lines in ﬁgure 1 and detailedin table 3. Approximately half of these nucleotides andthe 5Ј positions of the base pairs noted in table 3 (anddepicted in ﬁg. 1) are purines (A, G) in the 516 intronand are pyrimidines (C, U) in intron 1506 (positions 2,25, 26, 97:277, 216:257, 217:256, 262:312, 322:326,and 335:364 [relative to Tetrahymena thermophila]). Allof these positions are highly conserved (Ͼ95%) with theexception of base pair 97:277 (Ͼ90%, Ͻ95%). Position172 and the 322:326 base pair have no homologous po-sition with respect to T. thermophila. These three posi-tions are highly conserved in both introns (Ͼ95%). Fig-ure 2 highlights the differences between the 516 and1506 introns in the P5b and P8 domains. The P5b helix(ﬁg. 2a) is one of the most distinct features of the 516intron. This bifurcated helix, present in all of the Ban-giales 516 introns, replaces the typical single helix pre-sent in all IC1 introns, including the Bangialean 1506introns (ﬁg. 2a). The P8 domain (ﬁg. 2b) is also distinct,primarily on the basis of helix length. In the 516 versionof this structure, the 3-bp helix is capped by a tetraloop,while the 1506 structure has a longer helix capped by avariable-length hairpin loop.Haugen et al. (1999) determined that P. spiralis var.amplifolia (R) appeared to contain an open readingframe (ORF) corresponding to a 149-amino-acid His-Cys box protein within the P1 extension of the 1506intron, as does P. tenera (KK) within the P2 extensionof the 516 intron. Johansen, Embley, and Willassen(1993) noted that the His-Cys box motif is a hallmarkof nuclear homing endonucleases. Examination of thesedomains in the bangialean introns within the presentstudy revealed that this ORF is present in those taxacontaining large extensions within the P1 or P2 domainfor 1506 and 516 introns, respectively. Table 2 high-lights the sizes of the P2 domain (516 intron) and theP1 domain (1506 intron) for all collections examined inthe study and the presence or absence of the His-Cysbox motif. The size of the P2 domain in the 516 intronranged from 62 to 552 nt, with only those Ͼ414 nt con-taining an ORF and His-Cys box motif. This motif hasnot previously been reported in the 516 intron for thefollowing taxa: B. fuscopurpurea (Helgoland), B. fus-copurpurea (MA), P. abbottae, P. kanakaensis, and P.tenera (SK). The P1 domain in the 1506 intron rangedfrom 53 to 521 nt, and the His-Cys box motif was foundonly in those with larger insertions in this region: P.fucicola and P. umbilicalis (previously unreported).
Group IC1 Introns in the Bangiales 1659FIG. 2.—Gallery highlighting differences between introns 516 and 1506 using consensus diagrams and individual secondary-structurediagrams. a, P5b region of intron 516, showing bangialean consensus and differences among speciﬁc taxa. b, P5b region of intron 1506, showinga bangialean consensus and differences speciﬁc among taxa. c, Consensus of the P8 regions of introns 516 and 1506 in the Bangiales. Forconsensus diagrams, positions with a nucleotide in Ͼ95% of sequences are shown: ACGU—95ϩ% conserved; acgu—90%–95% conserved;●—80%–90% conserved; o—Ͻ80% conserved. Otherwise, the regions are represented by arcs.Haugen et al. (1999) noted that the ORF within theP1 extension for intron 1506 was determined to be onthe complementary strand to that encoding the SSUrRNA, and this was also the case for the ORF in thepreviously unreported taxa (P. fucicola and P. umbili-calis). However, this was not true for the ORF in the P2extensions for the 516 introns, which were found on thesame strand as that encoding the SSU rRNA. Figure 3depicts an amino acid alignment of the His-Cys boxmotif region in the P1 domain of the 1506 intron andthe P2 domain of the 516 intron. As noted by Haugenet al. (1999), P. spiralis var. amplifolia (R) and the en-donuclease I-Ppo I were identical in 16 out of the 29amino acids. The two new additional amino acid se-quences, P. fucicola and P. umbilicalis (HG), were iden-tical in 13 amino acids to the I-Ppo I endonuclease.However, the amino-acid sequence for the His-Cys boxin the 516 introns was identical in only 9 or 10 aminoacids (ﬁg. 3). Haugen et al. (1999) also noted frame-shifts in the ORF of P. spiralis var. amplifolia (R) andB. fuscopurpurea (1). Frameshifts were noted in theORF of the 516 introns in P. kanakaensis and B. fus-copurpurea (HL); however, the remaining sequenceswere not found to have frameshifts, although they allterminated prematurely relative to the amino acid se-quence for the homing endonuclease I-Ppo I (ﬁg. 3).This raises the possibility that these sequences are pseu-dogenes; however, endonuclease activity in the Bangi-ales was not tested in this study and needs to be inves-tigated in further detail.Phylogenetic Analysis of IntronsFigure 4 depicts a neighbor-joining tree derivedfrom the analysis of the nuclear SSU rRNA genes (par-simony analyses were unobtainable due to the very longcomputation time required) upon which the presence(ϩ) and absence (Ϫ) of both introns have been mapped.From these analyses, there appear to be multiple lossesof both introns. For example, the ﬁrst major clade ofBangia and Porphyra at the bottom of the tree containsthree smaller clusters in which there have been at leasttwo losses of the 1506 intron along each lineage (ﬁg.4). This same trend is even more evident in the largerBangia-Porphyra clade, where taxa possessing an intronare intermingled with those that do not. There does notappear to be any consistent trend with respect to pres-ence or absence of either intron; however, it does appearthat the more derived a taxon is, the more likely theintron will be absent (e.g., P. miniata (CCAP 1379/2),P. amplissima, P. miniata, and P. miniata (NF)) (ﬁg. 4).Sequences of collections of B. atropurpurea were de-termined to have identical nuclear SSU rRNA gene se-quences, and all collections contained only intron 516.Interestingly, some collections of B. fuscopurpureathat were identical with regard to nuclear SSU rRNAgene sequences were variable with regard to possessionof either or both introns. For example, the collectionfrom Greece contained both introns 516 and 1506; how-ever, the other collection from Australia, with which itwas identical, did not contain the 1506 intron. In addi-
Group IC1 Introns in the Bangiales 1661Table 3Differences Between the 516 and 1506 Introns at SinglePositions or Base Pairs Using Tetrahymena thermophilaPosition NumberingT. thermophila Intron 516 Intron 15062 . . . . . . . . . . . . . .25 . . . . . . . . . . . . .26 . . . . . . . . . . . . .97:277 . . . . . . . . .98 . . . . . . . . . . . . .160 . . . . . . . . . . . .172a . . . . . . . . . . .205 . . . . . . . . . . . .216:257 . . . . . . . .217:256 . . . . . . . .262:312 . . . . . . . .272 . . . . . . . . . . . .281 . . . . . . . . . . . .304 . . . . . . . . . . . .313:413 . . . . . . . .322:326a . . . . . . .335:364 . . . . . . . .GAGA:UCAUUG:CG:CG:CUCGA:UG:CA:UUUUu:gUGCCC:GC:GC:GGUAG:CC:GC:GNOTE.—Each position number shown is the nearest T. thermophila nucleo-tide. Uppercase letters indicate that nucleotides are conserved in at least 95% ofthe sequences. Lowercase letters indicate conservation between 90% and 95%.a No homologous position exists in the T. thermophila sequence.FIG. 3.—Alignment of open reading frames (ORFs) coding forendonuclease-like amino acid sequences highlighting the His-Cys boxmotif from the 516 and 1506 introns in the Bangiales. A, 516 introns:Porphyra abbottae (P. abb.), Porphyra kanakaensis (P. kana.), Por-phyra tenera (KK) (P. ten. (KK)), P. tenera (SK) (P. ten. (SK)), Ban-gia fuscopurpurea (Helgoland) (B. fusc. (HL)), and B. fuscopurpurea(MA) (B. fusc. (MA)). B, 1506 introns: Porphyra spiralis var. ampli-folia (R), Porphyra umbilicalis (HG) (P. umb. (HG)), Porphyra fuci-cola, and B. fuscopurpurea (1) (B. fusc. (1)). Identical positions areindicated by dots, alignment gaps are indicated by dashes, and residuesinferred from reading frameshift corrections are shown in bold low-ercase letters. Conserved residues proposed to be directly involved inzinc binding (C100, C105, H110, C125, C132, H134, C138) and theactive site (H98, N119) of the I-PpoI endonuclease (Flick et al. 1998)are indicated. The His-Cys box motif and frameshifts for P. tenera(KK), P. tenera (SK), B. fuscopurpurea (1), and P. spiralis var. am-plifolia (R) were previously noted by Haugen et al. (1999).tion, the collection from Ireland (IR) contained both in-trons 516 and 1506, whereas the other collections (WAand TX), with which it was nearly identical (differingby 8 bp) contained neither intron. A similar pattern oc-curred in B. fuscopurpurea from Massachussetts (MA),which did not contain the intron 1506, yet it was presentin the other collections with nearly identical sequences(Helgoland and Nice, differing by 5 bp). Within Por-phyra, this was only seen in one case: Porphyra ye-zoensis (HH) did not contain intron 1506, whereas P.yezoensis (OM) did.Well-supported clades are along biogeographiclines with few exceptions, but the presence or absenceof introns does not appear to be consistent based ongeographic locations. For example, the well-supportedclade (100% bootstrap) of B. fuscopurpurea containingprimarily collections from the Atlantic and Mediterra-nean (NC, NJ, Greece, Helgoland, Nice, MA [with theexception of the Paciﬁc samples from Australia andMexico]) all contained intron 516, except the samplefrom Mexico, and the samples from Massachusetts andAustralia did not contain intron 1506. Distromatic taxa(subgenus Diploderma) of Porphyra, P. miniata (CCAP1379/2, NF) and P. amplissima (2), contained neitherintron, whereas other species of Porphyra belonging tothe subgenera Diplastida and Porphyra were variable incontaining either intron.Introns occur frequently within the Chlorophyta,but within the Rhodophyta, Bangia, Porphyra, and Hil-denbrandia are the only genera currently known to con-tain introns (Ragan et al. 1993). This trend raises thequestion of whether the introns within Hildenbrandiaand the Bangiales might be related. Hence, ﬁgure 5 alsoincludes the two 1506 introns from H. rubra and onechlorophyte algae, C. ellipsoidea (A, B). This ﬁgure pre-sents the one most-parsimonious tree based on analysisof 1,025 parsimony-informative characters of well-aligned positions of the group IC1 intron in positions516 and 1506. This tree has a length of 4,589 steps anda consistency index (CI) (a measure of the amount ofhomoplasy exhibited by the set of characters: for nohomoplasy, CI ϭ 1) of 0.4988. This data set was alsoanalyzed by coding the gaps as single evolutionaryevents (e.g., 10 consecutive gaps are treated as one eventrather than 10 separate events), whereas in the previousanalysis, they were treated as missing data. This analysisdepicted little change in the topologies of the trees; infact, the resolution was much lower than when the gapswere treated as missing characters.
1662 Mu¨ller et al.FIG. 4.—Presence (ϩ) and absence (Ϫ) of group IC1 introns in positions 516 (ﬁrst symbol) and 1506 (second symbol) mapped on an 18SrRNA gene neighbor-joining tree. Lengths of branches in tree correspond to sequence divergence calculated using the Kimura (1980) two-parameter model.Figure 5 clearly depicts the 516 and 1506 intronsequences as two separate clades that are generally wellsupported (516: 100% maximum-parsimony [MP] boot-strap; 50% quartet puzzling values [QPS]; 100% neigh-bor-joining [NJ] bootstrap; 1506: 98% MP bootstrap;Ͻ50% QPS; 99% NJ bootstrap). The two introns in H.rubra are well associated with each other (100% MPbootstrap; 88% QPS; 95% NJ bootstrap) and are wellpositioned within the 1506 intron sequence clade of theBangiales (89% MP bootstrap; 89% QPS; 99% NJ boot-strap). The introns in P. spiralis var. amplifolia (position1506) are positioned on a branch that is basal to thecluster containing the 1506 intron sequences from theBangiales and H. rubra.Both the nuclear SSU rRNA gene NJ tree (ﬁg. 4)and the phylogenetic tree using intron sequences (ﬁg. 5)show similar trends. For example, the well-supported(100% bootstrap support; ﬁg. 5) cluster containing B.fuscopurpurea from New Jersey (NJ), Nice, Australia,Greece, North Carolina (NC), Helgoland, and Massa-chusetts (MA) is also present in the intron phylogenetictree for both the 1506 (100% MP bootstrap; 99% QPS;92% NJ bootstrap) and 516 introns (98% MP bootstrap;Ͻ50% QPS; 70% NJ bootstrap) (ﬁg. 5). In addition, thecollections from Ireland, Newfoundland (NF), and B.fuscopurpurea (2) also cluster together within the nu-clear SSU rRNA gene analysis (93% bootstrap) and theintron phylogeny for both intron insertion sites. This re-lationship is fairly well supported for the 1506 intronsequences (72% MP bootstrap; 94% QPS; 83% NJ boot-strap) but not for the 516 intron sequences (61% MPbootstrap; Ͻ50% QPS; Ͻ50% NJ bootstrap) (ﬁg. 5).Similarly, the 516 intron sequences within thefreshwater taxon B. atropurpurea from the British Isles(BI12), Netherlands (NL), Great Lakes (GL), Italy (IT),and Ireland (IR) are positioned in a well-supported clade(100% MP bootstrap; 57% QPS; 85% NJ bootstrap), asare the sequences of the nuclear SSU rRNA gene forthese collections. In addition, none of the freshwater col-lections contained intron 1506. The relationship of the
Group IC1 Introns in the Bangiales 1663FIG. 5.—The one most-parsimonious tree based on analysis of 1,025 phylogenetically informative characters of well-aligned positions ofthe group IC1 intron in positions 516 and 1506 in the order Bangiales (length ϭ 4,589, consistency index ϭ 0.4988). The ﬁrst number abovea branch represents bootstrap support (1,000 replicates) for maximum parsimony, the second value is the percentage of times that a particularcluster was found among the 1,000 intermediate steps (QPS) using quartet puzzling, and the ﬁnal value represents bootstrap resampling fordistance analysis (1,000 replicates). Numbers below branches represent Bremer indices or decay values.marine collection from the Virgin Islands (VIS7) wasunresolved (Ͻ55% bootstrap support) in both the intronanalyses, and it is only weakly associated (62% boot-strap) with one of the main clades in the NJ nuclearSSU rRNA gene tree.The relationship among the introns in Porphyra isnot as well resolved as that in Bangia. However, somegroupings are evident in all three analyses. For example,Porphyra sp. (SY) and P. pseudolinearis (TT) are close-ly associated in the nuclear SSU rRNA gene phylogeny(75% bootstrap), and the 516 intron sequences for thesetwo taxa are also well associated (99% MP bootstrap;Ͻ50% QPS; 99% NJ bootstrap) (ﬁg. 5). In addition, theclade consisting of P. tenera (T1, TU-3), P. umbilicalis(NJ), P. yezoensis, and P. yezoensis (HH, OM) is presentin the nuclear SSU rRNA gene phylogeny (although notwell supported), and this cluster is also seen for the 516introns for these taxa in ﬁgure 5 (100% MP bootstrap;97% QPS; 76% NJ bootstrap). A close relationshipamong Porphyra rediviva, Porphyra sp. (Brest), and P.umbilicalis (HG) is reﬂected in both the nuclear SSUrRNA gene tree (100% bootstrap) and the 1506 portionof the intron phylogeny, with the relationship betweenP. rediviva and Porphyra sp. (Brest) being well sup-ported (100% MP bootstrap; 100% QPS; 75% NJ boot-strap) and the 1506 intron sequence being only weaklyassociated with the previous clade (Ͻ50% MP boot-strap; 64% QPS; Ͻ50% NJ bootstrap). A grouping ofPorphyra 1506 intron sequences consisting primarily ofcollections from Japan (P. yezoensis (OM, OG-1, OG-4, NA-2, NA-4), P. yezoensis narawaenis, P. tenera(KK, SK), P. pseudolinearis (TT), Porphyra sp. (SHH))is moderately supported (82% MP bootstrap; 83% QPS;Ͻ50% NJ bootstrap). This trend is not, however, reﬂect-
1664 Mu¨ller et al.ed in the nuclear SSU rRNA gene tree, where the rela-tionship among many of the Japanese taxa is close butunresolved (Ͻ50% support). Many of the other relation-ships seen in the nuclear SSU rRNA gene NJ tree aresimilar to those in the intron trees, but most are not wellsupported by bootstrap analysis (Ͻ60%). The sequencedivergence in the intron at position 516 varied consid-erably, ranging from 0% to as high as 31.0%. The se-quence divergence for the intron in position 1506 washigher than that for the 516 intron (31.0% vs. 44.3%).Interestingly, the 1506 introns in P. spiralis var. ampli-folia (Oliveira and Ragan 1994) were quite distant fromall other Porphyra and Bangia taxa, ranging from 30.3%to 44.3% sequence divergence.DiscussionIntrons within the nuclear SSU rRNA gene frommembers of the Bangiales have been reported severaltimes within the past few years (Stiller and Waaland1993; Oliveira and Ragan 1994; Oliveira et al. 1995;Mu¨ller et al. 1998). The presence and absence of theseintrons presents us with an opportunity to examine lat-eral intron transfer events and the evolutionary historyof introns within the order Bangiales. Oliveira et al.(1995) speculated that these introns could also be usedto discern biogeographic or speciﬁc entities within thegenus Porphyra, as they determined that variants of thegroup IC1 intron were present in different geographicpopulations of this alga. Due to the considerable numberand variation (as high as 40%) in the introns in Por-phyra and Bangia, this system is well suited to address-ing some of these issues. In addition, group I intronsthat lack endonuclease coding regions appear to be non-mobile and provide a potentially valuable tool for trac-ing the evolutionary history of these sequences (Bhat-tacharya 1998).The similarity in phylogenies between the 516 and1506 introns and their nuclear SSU rRNA genes sug-gests that these introns have not been mobile during theevolution of the Bangiales. Introns have been docu-mented in the nuclear-encoded rRNA genes of differentNaegleria species. The majority of these occur at the516 position of the nuclear SSU rRNA gene. A moreextensive analysis of the Naegleria nuclear SSU rRNAgenes revealed that this intron is absent in the majorityof the species’ nuclear SSU rRNA genes (De Jonckheere1994). De Jonckheere (1994) hypothesized that sincethere was probably no selective advantage arising fromthe presence of these introns, they were eliminated inmany of the descendants. In the case of the Bangiales,it appears that after initial acquisition of the introns inthe ancestor of these taxa, there was subsequent verticalinheritance and evolution within the order (and species)and in some cases loss of the introns in either insertionsite, as the intron phylogenies are very similar to thatof the host gene. An alternative and less favorable hy-pothesis is that frequent lateral transfers within the ordermay have taken place, but this would result in intronphylogenies that would be widely discordant with thenuclear SSU rRNA gene phylogenies (Bhattacharya etal. 1994).The similarity among group I intron and nuclearSSU rRNA gene phylogenies has also been reported forother eukaryotic taxa. Takashima and Nakase (1997)noted that introns had been transferred horizontally todistinct insertion sites in the yeast-like fungus Tilletiop-sis ﬂava and then inherited vertically. Similarly, introns1506 and the nuclear SSU rRNA gene sequences sharesimilar phylogenies in members of the green algal orderZygnematales (Bhattacharya et al. 1994). Wilcox et al.(1992) also postulated that the distribution of intronswithin other green algal taxa could be explained by in-heritance of these sequences along with the gene. How-ever, VanOppen, Olsen, and Stam (1993) concluded thatnuclear group I introns in green algae have arisen sev-eral times and do not appear to be lineage-speciﬁc anddisagreed that the distribution of introns could be ex-plained by vertical inheritance within the green algae aspostulated by Wilcox et al. (1992). In the case of theBangiales, the latter scenario appears to be possible. Forinstance, the intron phylogenies appear to be in accor-dance with the gene phylogenies, indicating long-termimmobility and vertical inheritance.The evolutionary history of the Bangiales is quitelong, as bangiophyte taxa have been reported in the fos-sil record from the Proterozoic (1,200 MYA) (Butter-ﬁeld, Knoll, and Swett 1990), and the conchocelis stageof Porphyra has been reported in fossils up to 425 MYA(Campbell 1980). In terms of utilizing the introns withinthe Bangiales as speciﬁc or geographic markers as spec-ulated by Oliveira and Ragan (1994), the present studydoes not appear to indicate this possibility. In fact, anal-yses of the group I introns appear to provide furtherevidence for the synonymy of these two genera, as in-tron phylogenies are congruent with those of the nuclearSSU rRNA gene. This is not surprising, as other molec-ular studies using nuclear and chloroplast genes haveindicated that these taxa are not monophyletic genera(Mu¨ller et al. 1998). Thus, attempts at constructing aphylogenetic classiﬁcation within the order based onmorphology, life histories, and molecular analyses, and,in this case, intron sequences, have been problematic.The observation that similar group I introns are dis-tributed across different genomes and phylogeneticallydistant taxa is taken as evidence that intron transfer isrelatively common (Burke 1988). The clustering ofgroup I introns in phylogenetic trees from distinct in-sertion sites suggests that these introns may be tracedback to one or more events in which the introns wereinserted into rRNA and vertically inherited or lost (Bhat-tacharya et al. 1994; Bhattacharya 1998). Within theZygnematales, bootstrap analyses yielded little supportfor the single origin of all group I introns at differentinsertion sites within rRNA (Bhattacharya et al. 1994).It is difﬁcult to determine if this is the case with theintrons within the Bangiales due to the lack of a suitableoutgroup. Nonetheless, all of the 516 and 1506 intronsform two separate clades distinct from each other. Highsequence similarity between introns is consistent withdescent from an ancestral intron that was initially ac-
Group IC1 Introns in the Bangiales 1665quired and vertically inherited (Schroeder-Diedrich,Fuerst, and Byers 1998). In the protist Acanthamoeba,low sequence identity (highly variable) among intronsin four different positions suggests that the intron ac-quisition occurred independently at the four positionsafter divergence of the taxa within the tree (Schroeder-Diedrich, Fuerst, and Byers 1998). There is high se-quence identity (0% sequence divergence in some cases)among many of the introns sequenced within the Ban-giales; however, there is also considerable variability.For example, the 1506 introns in P. spiralis var. ampli-folia (B, D, R) differ from all other intron sequences by30%–44% sequence divergence. Despite this, P. spiralisvar. amplifolia is well supported as grouping with theremaining 1506 introns.Homing-Endonuclease Pseudogenes in the BangialesThe mobility of nuclear group I introns may dependon homing through the expression of intron-encodedhoming endonucleases (Belfort and Roberts 1997). TheHis-Cys box is one of several types of homing endo-nucleases that is restricted to group I introns (Johansen,Embley, and Willassen 1993). Haugen et al. (1999) de-scribed unusual P1 extensions within the nuclear SSUrRNA introns of Bangia and Porphyra. Their analysisshowed that the intron sequences contained His-Cys boxreading frames within the P1 extension and that thesemay represent homing-endonuclease pseudogenes. Theyalso demonstrated that the ORF in the 1506 intron of P.spiralis var. amplifolia occurred on the strand comple-mentary to that encoding the SSU rRNA gene and groupI intron. The present study also observed this phenom-enon for the P1 domain in the introns at position 1506for P. fucicola and P. umbilicalis (HG), both of whichalso contained the His-Cys box motif noted in P. spiralisvar. amplifolia. The introns in the 516 position that con-tained an ORF as well as a His-Cys box motif werewithin the P2 domain and not on the complementarystrand, as seen in the 1506 introns. Haugen et al. (1999)also noted frameshifts in the 3Ј end of the P. spiralisvar. amplifolia intron ORF which resulted in a truncatedC-terminal end and would explain the lack of cleavageactivity in the in vitro translated protein. Haugen et al.(1999) suggested that the truncation and frameshiftsseen in the Bangialean intron ORFs may be the resultof selection against endonuclease expression whichwould lethally cleave chromosomal DNA. Only two ofthe new sequences in this study had frameshifts, P. kan-akaensis and B. fuscopurpurea (Helgoland), and all se-quences were found to terminate prematurely. This mayindicate that these ORFs are no longer functional; how-ever, endonuclease activity has not been studied in thesesequences and needs to be investigated.Secondary-Structure Signatures of the Group IC1Introns in the BangialesIn 1987, Woese rationalized that the examinationof higher-order structure from a phylogenetic perspec-tive would enable us to examine the stages of evolutionof rRNA structure and aid in extrapolating the signiﬁ-cance of changes and variation in rRNA. Subsequently,Winker and Woese (1991) delineated the three primarydomains using structural characteristics of 16S rRNA.They deﬁned two types of signature characters: (1) ho-mologous positions whose compositions are (very near-ly) constant (present in at least 94% of the representativesequences) within each of the domains or kingdoms andare characteristic of at least one, and (2) nonhomologousstructures that characterize and distinguish the variousgroupings (e.g., structures that are present in one groupand absent in another and/or structural units that differstrikingly from one group to another). Within these twotypes there are different examples of phylogeneticallyconstrained elements ranging from single nucleotidesand base pairs to hairpin loops, noncanonical pairing,insertions, deletions, and combinations of these ele-ments, along with coaxial stacked helices or differencesin numbers between homologous structures (Gutell1992).In the present study, the Bangiales were observedto have sequence and structural elements, or ‘‘signa-tures,’’ that differentiate these introns from each otheras well as from other available intron sequences. Asnoted previously, the P5b domain in the 516 introns ofthe Bangiales contains a bifurcated helix that distin-guishes these introns from all other group IC1 introns(ﬁg. 2a). This element is not present in the P5b domainin the 1506 bangialean introns and is an example of anonhomologous signature. The variation in length in theP8 domain of the 1506 introns is also unique to theBangiales and is not seen in the bangialean 516 introns.In addition to these structural elements, the introns inthe Bangiales can be differentiated based on nonhomol-ogous and homologous single-nucleotide and base pairsignatures (table 3). For example, nucleotide 172 andbase pair 322:326 have no homologous positions withrespect to T. thermophila and can be considered non-homologous signatures deﬁning these introns. The twointrons can be differentiated based on nucleotide andbase pair signatures. For example, position 205 is a Uin intron 516 (conserved in Ͼ95% of the sequences) anda C in intron 1506 (conserved in Ͼ95% of the sequenc-es). These unique structural and sequence signaturesprovide further evidence that the introns in positions 516and 1506 are probably the result of a single (althoughseparate from each other) lateral transfer event and sub-sequent vertical inheritance. In addition, these structuralsignatures may provide a means for determining the pos-sible origin of the Bangialean 516 and 1506 introns ifan ancestral intron still exists. This information alsoyields a basis for utilizing introns to differentiate dif-ferent phyla/organisms or to determine the origin of in-trons based on structural characteristics.AcknowledgmentsThis research was supported by NSERC grant RGP0183503 to R.G.S., NIH grant GM 48207 to R.R.G., andan OGS Scholarship to K.M.M. Technical assistance inDNA sequencing by Angela Holliss is gratefully ac-knowledged. We thank John Stiller, Alison Sherwood,
1666 Mu¨ller et al.Ron Deckert, Blair Brace, Wilson Freshwater, Mary Ko-ske, and John West for help in collecting or providingsamples for this study.LITERATURE CITEDBELFORT, M., and R. J. ROBERTS. 1997. Homing endonucle-ases: keeping the house in order. Nucleic Acids Res. 25:3379–3388.BHATTACHARYA, D. 1998. The origin and evolution of protistgroup I introns. Protist 149:113–122.BHATTACHARYA, D., T. FRIEDL, and S. DAMBERGER. 1996. Nu-clear-encoded rDNA group I introns: origin and phyloge-netic relationships of insertion site lineages in green algae.Mol. Biol. Evol. 13:978–989.BHATTACHARYA, D., B. SUREK, M. RU¨ SING, S. DAMBERGER,and M. MELKONIAN. 1994. Group I introns are inheritedthrough common ancestry in the nuclear-encoded rRNA ofZygnematales (Charophyceae). Proc. Natl. Acad. Sci. USA91:9916–9920.BURKE, J. M. 1988. Molecular genetics of group I introns:RNA structures and protein factors required for splicing—a review. Gene 73:259–271.BUTTERFIELD, N. J., A. H. KNOLL, and K. SWETT. 1990. Abangiophyte red alga from the proterozoic of Arctic Canada.Science 250:104–107.CAMPBELL, S. E. 1980. Palaeoconchocelis starmachii, a car-bonate boring microfossil from the Upper Silurian of Po-land (425 million years old): implications for the evolutionof the Bangiaceae (Rhodophyta). Phycologia 19:25–36.CECH, T. R. 1990. Self-splicing of group I introns. Annu. Rev.Biochem. 59:543–568.DAMBERGER, S. H., and R. R. GUTELL. 1994. A comparativedatabase of group I intron structures. Nucleic Acids Res.22:3508–3510.DE JONCKHEERE, J. F. 1994. Evidence for the ancestral originof Group I introns in the SHU rDNA of Naegleria spp. J.Eukaryot. Microbiol. 41:457–463.FELSENSTEIN, J. 1993. PHYLIP (phylogeny inference package).Version 3.5c. Distributed by the author, Department of Ge-netics, University of Washington, Seattle.FLICK, K. E., M. S. JURICA, R. J. MONNAT JR., and B. L.STODDARD. 1998. DNA binding and cleavage by the nuclearintron-encoded endonuclease I-PpoI. Nature 394:96–101.FRIEDL, T., A. BESENDAHL, P. PFEIFFER, and D. BHATTACHAR-YA. 2000. The distribution of group I introns in lichen algaesuggests that lichenization facilitates intron lateral transfer.Mol. Phylogenet. Evol. 14:342–352.GUTELL, R. R. 1992. Evolutionary characteristics of 16S and23S rRNA structures. Pp. 243–309 in H. HARTMAN and K.MATSUNO, eds. The origin and evolution of the cell. WorldScientiﬁc, N.J.———. 1993. Collection of small subunit (16S- and 16S-like)ribosomal RNA structures. Nucleic Acids Res. 21:3051–3054.GUTELL, R. R., B. WEISER, C. R. WOESE, and H. F. NOLLER.1985. Comparative anatomy of 16S-like ribosomal RNA.Prog. Nucleic Acid Res. Mol. Biol. 32:155–216.HAUGEN, P., V. A. R. HUSS, H. NIELSEN, and S. JOHANSEN.1999. Complex group-I introns in nuclear SSU rDNA ofred and green algae: evidence of homing-endonucleasepseudogenes in the Bangiophyceae. Curr. Genet. 36:345–353.JOHANSEN, S., T. M. EMBLEY, and N. P. WILLASSEN. 1993. Afamily of nuclear homing endonucleases. Nucleic AcidsRes. 21:4405.JOHANSEN, S., D. E. MUSCARELLA, and V. M. VOGT. 1996.Insertion elements in ribosomal DNA. Pp. 89–110 in R. A.ZIMMERMANN and A. E. DAHLBERG, eds. Ribosomal RNA.Structure, evolution, processing, and function in proteinbiosynthesis. CRC Press, New York.JUKES, T. H., and C. R. CANTOR. 1969. Evolution of proteinmolecules. Pp. 21–132 in H. N. MONROE, ed. Mammalianprotein metabolism. Academic Press, New York.KIMURA, M. 1980. A simple method for estimating evolution-ary rate of base change substitutions through comparativestudies of nucleotide sequences. J. Mol. Evol. 16:111–120.LARSEN, N., G. J. OLSEN, B. L. MAIDAK, M. J. MCCAUGHEY,R. OVERBEEK, T. J. MACKE, T. L. MARSH, and C. R. WOESE.1993. The ribosomal database project. Nucleic Acids Res.21:3021–3023.MICHEL, F., and E. WESTHOF. 1990. Modeling of the three-dimensional architecture of group I catalytic introns basedon comparative sequence analysis. J. Mol. Biol. 216:585–610.MU¨ LLER, K. M., R. G. SHEATH, M. L. VIS, T. J. CREASE, andK. M. COLE. 1998. Biogeography and systematics of Ban-gia (Bangiales, Rhodophyta) based on the Rubisco spacer,rbcL gene and 18S rRNA gene sequences and morphomet-ric analyses. 1. North America. Phycologia 37:195–207.OLIVEIRA, M. C., J. KURNIAWAN, C. J. BIRD, E. L. RICE, C.A. MURPHY, R. K. SINGH, R. R. GUTELL, and M. A. RAGAN.1995. A preliminary investigation of the order Bangiales(Bangiophycidae, Rhodophyta) based on sequences of nu-clear small-subunit ribosomal RNA genes. Phycol. Res. 43:71–79.OLIVEIRA, M. C., and M. A. RAGAN. 1994. Variant forms of agroup I intron in nuclear small-subunit rRNA genes of themarine red alga Porphyra spiralis var. amplifolia. Mol.Biol. Evol. 11:195–207.RAGAN, M. A., C. J. BIRD, E. L. RICE, R. R. GUTELL, C. A.MURPHY, and R. K. SINGH. 1994. A molecular phylogenyof the marine red algae (Rhodophyta) based on the nuclearsmall-subunit rRNA gene. Proc. Natl. Acad. Sci. USA 91:7276–7280.RAGAN, M. A., C. J. BIRD, E. L. RICE, and R. K. SINGH. 1993.The nuclear 18S ribosomal RNA gene of the red alga Hil-denbrandia rubra contains a group I intron. Nucleic AcidsRes. 16:3898–3839.SAUNDERS, G. W. 1993. Gel puriﬁcation of red algal genomicDNA: an inexpensive and rapid method for the isolation ofpolymerase chain reaction-friendly DNA. J. Phycol. 29:251–254.SAUNDERS, G. W. and G. T. KRAFT. 1994. Small-subunit rRNAgene sequences from representatives of selected families ofthe Gigartinales and Rhodymeniales (Rhodophyta). 1. Evi-dence for the Plocamiales ord. nov. Can. J. Bot. 72:1250–1263.SCHROEDER-DIEDRICH, J. M., P. A. FUERST, and T. J. BYERS.1998. Group I introns with unusual sequences at three sitesin nuclear rRNA genes of Acanthamoeba lenticulata. Curr.Genet. 34:71–78.STILLER, J. W., and J. R. WAALAND. 1993. Molecular analysisreveals cryptic diversity in Porphyra (Rhodophyta). J. Phy-col. 29:506–517.STRIMMER, K., and A. VON HAESELER. 1996. Quartet puzzling:a quartet maximum-likelihood method for reconstructingtree topologies. Mol. Biol. Evol. 13:964–969.SWOFFORD, D. L. 1993. PAUP: phylogenetic analysis usingparsimony. Version 3.1.1. Illinois Natural History Survey,Champaign, Ill.TAKASHIMA, M., and T. NAKASE. 1997. A phylogenetic anal-ysis of three group I introns found in the nuclear small
Group IC1 Introns in the Bangiales 1667subunit ribosomal RNA gene of the ballistoconidio genousanamorphic yeast-like fungus Tilletiopsis ﬂava. Genes Gen-et. Syst. 72:205–214.VAN OPPEN, M. J. H., J. L. OLSEN, and W. T. STAM. 1993.Evidence for independent acquisition of group I introns ingreen algae. Mol. Biol. Evol. 10:1317–1326.VIS, M. L., and R. G. SHEATH. 1997. Biogeography of Batra-chospermum gelatinosum (Batrachospermales, Rhodophyta)in North America based on molecular and morphologicaldata. J. Phycol. 33:520–526.WILCOX, L. W., L. A. LEWIS, P. A. FUERST, and G. L. FLOYD.1992. Group I introns within the nuclear-encoded small-subunit rRNA gene of three green algae. Mol. Biol. Evol.9:1103–1118.WINKER, S., and C. R. WOESE. 1991. A deﬁnition of the do-mains Archaea, Bacteria and Eucarya in terms of small sub-unit ribosomal RNA characteristics. Syst. Appl. Microbiol.14:305–310.WOESE, C. R. 1987. Bacterial evolution. Microbiol. Rev. 51:221–271DAVID IRWIN, reviewing editorAccepted April 10, 2001