Tao Sang

Phylogeny and Biogeography of Paeonia (PAEONIACEAE)

DISSERTATION

chapter 3

EVOLUTION OF CHLOROPLAST DNA INTERGENIC SPACERS AND PHYLOGENETIC IMPLICATIONS IN PEONIES (PAEONIA. PAEONIACEAE)

ABSTRACT. Efforts have been recently undertaken to search and sequence rapidly evolving regions in the chloroplast genome for phylogenetic information at lower taxonomic levels. The psbA-trnH and trnL(UAA)-trnF(GAA) intergenic spacers of chloroplast DNA have been sequenced for 32 species of Paeonia (Paeoniaceae). Patterns, rates, and phylogenetic information of mutations in the two intergenic spacers are compared with the rapidly evolving matK coding region. Nucleotide substitutions occur slowly and homoplasiously in the trnL-trnF intergenic spacer, but rapidly and least homoplasiously in the psbA-trnH intergenic spacer, suggesting that the latter may serve as a new and better phylogenetic marker at the intrageneric level. Insertions and deletions, which occur less frequently than nucleotide substitutions in both intergenic spacers, provide reliable phylogenetic information. In the psbA-trnH intergenic spacer, short sequences bordered by long inverted repeats can frequently undergo inversions that are often homoplasious mutations and thus are not useful in phylogenetic analysis. Understanding the nature of mutations in DNA sequences, therefore, is critical for interpreting them appropriately in phylogenetic reconstructions. A short cpDNA intergenic spacer alone may not provide enough synapomorphic characters to group closely related species whose relationships may be assessed by rapidly evolving sequences of multiple coding and noncoding regions of cpDNA.

INTRODUCTION

Restriction site variation of chloroplast DNA has served as a major source of evidence for plant phylogenetic reconstructions at lower taxonomic levels, particularly the intrageneric level, over the past ten years (Sytsma and Gottlieb, 1986; Palmer et al., 1988; Wendel and Albert, 1992; Olmstead and Palmer, 1994). Recently, efforts have been undertaken to search and sequence rapidly evolving regions in the chloroplast genome for phylogenetic uses at lower taxonomic levels (Johnson and Soltis, 1994; Steele and Holsinger, 1994; Gielly and Taberlet, 1994). Noncoding regions, including introns and intergenic spacers, are potentially good candidates because they are under less functional constrain and thus evolve more rapidly (Clegg et al., 1994; Gielly and Taberlet, 1994). A few noncoding regions have been examined to assess intrafamilial and intrageneric relationships, including intergenic spacers of at pB-rbcL. rbcL-psaI. .trnT(UGU)-trnL(UAA) and trnL(UAA)-trnF(GAA), and the trnL(UAA) intron (Taberlet et al., 1991; Golenberg et al., 1993; Morton and Clegg, 1993; Bohle et el., 1994; Ham et al., 1994; Manen et al., 1994; Mes and Hart, 1994). These noncoding regions provided relatively good resolution for intergeneric relationships in a variety of plant families, but their applications at the intrageneric level seem to be less effective (Bohle et al., 1994). Therefore, the question of whether sequences of cpDNA noncoding regions can serve as significant sources of evidence for phylogenetic reconstructions at the intrageneric level needs to be addressed further.

Understanding of evolutionary mode and tempo of cpDNA noncoding regions is essential to their appropriate phylogenetic applications. Rates of nucleotide substitutions appear to vary considerably among different noncoding regions as well as among different plant groups. Substitution rates in the trnL(UAA) intron and the trnL(UAA) -trnF(GAA) intergenic spacer were found to be more than three times faster than the rbcL coding region (Gielly and Taberlet, 1994), whereas rates in the rbcL-psaI intergenic spacer are not greater than the synonymous rate of the rbcL (Morton and Clegg, 1993). Sequence divergence in the trnL intron vary from 0 to 6% among different plant genera (Gielly and Taberlet, 1994). Insertions/deletions (indels) occur frequently in noncoding regions (Clegg et al., 1994), but their phylogenetic values are unclear. It has been suggested that they are primarily synapomorphic characters in certain groups but homoplasious in others-(Golenberg et al., 1993; Morton and Clegg, 1993; Ham et al., 1994; Mes and Hart, 1994). In the present study, two intergenic spacers, trnL(UAA) - trnF(GAA) and psbA-trnH,, in the large single copy region of cpDNA have been sequenced for 32 species of Paeonia. The trnL-trnF intergenic spacer is probably the most frequently used noncoding region of cpDNA in phylogenetic studies (Bohle et al., 1994; Gielly and Taberlet, 1994; Ham et al., 1994; Mes and Hart, 1994). The psbA-trnH intergenic spacer was chosen as a phylogenetic marker in this study because it was suggested to be an evolutionarily plastic region that could tolerate many indels (Aldrich et al., 1988). Higher rates of mutations have been detected in the psbA-trnH intergenic spacer among species of peonies than in the trnL-trnF intergenic spacer, suggesting that the former should be more useful than the latter for phylogenetic studies at the intrageneric level. Phylogenetic reconstruction from mutations in the psbA-trnH spacer is compared with the phylogeny recently generated from the matK coding region, the most rapidly evolving coding regions found so far in the chloroplast genome (Olmstead and Palmer, 1994; Sang et al., in prep.). Phylogenetic information yielded by the two intergenic spacers and the matK coding region is also compared to assess phylogenetic potential of rapidly evolving regions in the chloroplast genome. Paeonia (Paeoniaceae, Paeoniales), occurring in several disjunct areas of the Northern Hemisphere, contains approximately 35 species of shrubs and perennial herbs in three taxonomic sections, Paeonia. Moutan, and Oneapia (Stern, 1946; Pan, 1979; Tzanoudakis, 1983). As described in the previous chapters, studies using sequences of internal transcribed spacers (ITS) of nuclear ribosomal DNA and the matK gene revealed complex reticulate evolution in section Paeonia (Sang et al., 1995a, in prep.). Therefore, phylogenies based on cpDNA, which is usually maternally inherited in angiosperms, do not represent the species phylogeny of Paeonia.

MATERIALS AND METHODS

Forty accessions of 32 Paeonia species were sequenced. For most species, fresh leaves used as sources of DNA were collected from natural populations in Bulgaria, China, Greece, and Spain. The remaining species were collected from The Royal Botanic Gardens, Kew (Table 1) . Total DNA was isolated from leaf tissues using the CTAB method (Doyle and Doyle, 1987), and purified in CsCl/ethidium bromide gradients. Double-stranded DNAs of the trnL-trnF and psbA-trnH intergenic spacers were amplified by 30 cycles of symmetric PCR (Sang, 1995b).

The amplification products were purified by electrophoresis through 1.0% agarose gel followed by use of Bioclean (U. S. Biochemical). Purified double-stranded DNAs were used for sequencing reactions employing Sequenase Version 2.0 (U. S. Biochemical), deoxyadenosine S'-^a-[35S]thio]triphosphate, and the same primers used in PCR (Sang et al., 1995b). The sequencing reaction products were separated electrophoretically in 6% acrylamide gel with wedge spacers for 3 hr at 1500 V. After fixation, gels were dried and exposed to Kodak XAR x-ray film for 18-48 hr. DNA sequences were aligned manually.

The primers designed for amplifying the trnL-trnF intergenic spacer are: the forward primer (trnLf), 5'-AAAATCGTGAGGGTTCAAGTC-3'; and the reverse primer (trnFr) , 5'-GATTTGAACTGGTGACACGAG-3' . The reverse primer (trnFr) is almost the same as the primer f of Taberlet et al. (1991) except for one more nucleotide at the 5' end of the trnFr. The forward primer (trnLf) is designed 10 nucleotides further away from the 3' end of the trnL 3' exon than the primer e of Taberlet et al. (1991) in order to read sequences closer to the 5' end of the intergenic spacer when it is also used as a sequencing primer. The primers designed to amplify the psbA-trnH intergenic spacer are: the forward primer (psbAf) , 5' -GTTATGCATGAACGTAATGCTC-3' complementing nucleotide 608-587 of the B strand of the tobacco cpDNA sequence (Shinozaki et al., 1986); and the reverse primer (trnHr), 5'-CGCGCATGGTGGATTCACAATC-3', corresponding to nucleotide 28-49 of the B strand of the tobacco cpDNA sequence (Shinozaki et al., 1986). The primers are designed in conserved regions of psbA and trnH genes among different dicot families. Sequences of psbA gene are compared between tobacco (Shinozaki et al., 1986) and Brassica napus of Brassicaceae (Genebank No. M36720). Sequences of trnH gene are compared among tobacco, Helianthus annus of Asteraceae (Genebank No. X60428), and Arabidopsis thaliana of Brassicaceae (Genebank No. X79898).

Sequence divergence between species for each cpDNA region was, calculated using DNADIST program of PHYLIP version 3.5c (Felsenstein, 1994). The Jukes-Cantor model was used for correcting possible multiple hits of nucleotide substitutions (Jukes and Cantor, 1969).

Mutations, including nucleotide substitutions and indels, were analyzed by unweighted Wagner parsimony using PAUP version 3.1.1 (Swofford, 1993). Indels were coded as binary characters in the analysis. The shortest trees were searched with TBR Branch Swapping of the heuristic method, and character changes were interpreted with the ACCTRAN optimization. Bootstrap analyses were carried out with 100 replications using TBR Branch Swapping of the Heuristic search (Felsenstein, 1985) . Section Oneapia of Paeonia was chosen as the outgroup for the cladistic analysis of the genus (Watrous and Wheeler, 1981; Maddison et al., 1984; Sang et al., 1995a).

RESULTS

The psbA-trnH intergenic spacer. Aligned sequences of the psbA-trnH intergenic spacer of Paeonia species are shown in Fig. 7. The sequences of two populations of P. lutea differ by one nucleotide substitution. For the remaining species, different accessions of each species have identical sequences. According to this straightforward alignment, the length of the intergenic spacer varies from 276 bp (P. mairei) to 318 bp (P. spontanea and P. szechuanica). A total of thirty-one variable sites (nucleotide substitutions) and thirteen indels are found among these species. However, two indels and several substitutions located in the regions between site 58 and 97 (Fig. 7) actually result from aligning inversions (see Discussion) . Since the inversions seem to occur quite frequently and can easily introduce homoplasy, they are not included in the phylogenetic analysis. Further, three substitutions in this region whose occurrence is facilitated by a special molecular structure are not included in calculating sequence divergence and reconstructing phylogeny (see Discussion). Consequently, a total of twenty-four variable sites and eleven indels are retained for estimating sequence divergence and reconstructing phylogeny.

Unweighted parsimony analysis of these mutations produced nine equally roost parsimonious trees. Length, consistency index (CI), and retention index (RI) of each of the nine trees are 36, 0.946 (0.913 excluding autapomorphies), and 0.981, respectively. A strict consensus tree is calculated from these equally most parsimonious trees (Fig. 8). Length, CI, and RI of the consensus tree are 37, 0.921 (0.875 excluding autapomorphies), and 0.971, respectively.

The trnL-trnF intergenic spacer. Because the forward primer is still very close to the 3' end of the trnL 3' exon, about 10 nucleotides at the 5' end of this intergenic spacer could not be read. Sequences of different accessions of the same species are identical. Length of the aligned sequences vary from 372 bp (P. obovata) to 404 bp (P. sterniana) for the species sequenced. Among all species, only nine variable sites and five indels were detected. Therefore, only descriptions of mutations instead of full sequences are given in Table 2. Because this spacer provides so little phylogenetic information, phylogenetic reconstruction was not performed.

Comparisons of sequence divergence and phylogenetic information from variable sites among the two intergenic spacers and the matK coding region are given in Table 3. Comparisons of phylogenetic information from indels and relative frequency of indels versus nucleotide substitutions between the two intergenic spacers are given in Table 4.

DISCUSSION

Inversions in the psbA-trnH intergenic spacer. Mutations in the region between nucleotide site 57 and 97 of the psbA-trnH intergenic spacer appear to be highly homoplasious. Based on the present alignment, P. rockii in subsection Vaginatae shares three substitutions at sites 95-97, and two indels(4 and 5) with subsection Delavayanae. In section Paeonia. four substitutions at sties 84, 86, 87, and 89 are also homoplasious according to the matK and ITS phylogenies (Fig. 9; Sang et al., 1995a, in prep.). Basically, three types of sequences in this region and surrounding regions can be recognized: type I occurs in section Oneapia,, and P. spontanea and P. szechuanica of subsection Vaginatae; type II occurs in subsection Delavayanae and P. rockii; and type III found in section Paeonia which further includes two sub-types i and ii (Fig 10A) .

A close examination of this region reveals a pair of long inverted repeats (Fig. 10A). Type I and II sequences can be converted into each other once the sequence bordered by the inverted repeats undergoes inversions. In type III sequence, an inversion of the sequence bordered by the inverted repeats in one sub-type sequence can give rise to the other sub-type. The homoplasious occurrence of the inversions suggests that inversions of short sequences bordered by inverted repeats can occur quite frequently. In contrast to some large inversions in cpDNA that provided reliable phylogenetic information at the higher taxonomic levels (Jansen and Palmer, 1987; Doyle et al., 1992; Raubeson and Jansen, 1992), short inversions in the intergenic spacer easily yield homoplasious information even at the interspecific level and thus should not be included in phylogenetic analyses.

The mechanism responsible for change between the type III sequence and the other two types is more complex. In the type I sequence, there is another pair of short inverted repeats in the region between the long inverted repeats (Fig. 10B) . Therefore, a stem-loop structure with two stems and two loops can be formed (Fig. 10B). The evolutionary changes that are likely to have occurred in the small loop between the two stems include deletion of the T and two transitions of A to G to match the two Cs so that a single longer stem of type III sequence could be formed. The two substitutions in this small loop are facilitated by this particular stem-loop structure, and thus should not be treated as regular substitutions in calculating sequence divergence in order to avoid overestimating rates of sequence divergence. Therefore, they were not taken into account in calculating sequence divergence or reconstructing phylogeny. Likewise, the two substitutions, A to C at site 76 and T to G at site 97, are at the corresponding positions of the inverted repeats, and should be considered as only one substitution for sequence divergence estimation and phylogenetic reconstruction.

Nucleotide substitutions. A comparison of average species pairwise sequence divergence in the two intergenic spacers and matK coding region (Table 2) indicates that the psbA-trnH intergenic spacer has much higher rates of nucleotide substitutions than the other two regions. The trnL-trnF intergenic spacer has lower substitution rates than the matK coding region, suggesting that higher substitution rates are not always expected in noncoding regions than in coding regions of cpDNA.

Distinguishing autapomorphic, synapomorphic, and homoplasious substitutions in the intergenic spacers and the coding region should enable comparisons of the quality of phylogenetic information yielded from these regions (Table 3) . The percentage of phylogenetically informative sites (the sites where substitutions are shared by two or more taxa) among the variable sites is similar for the three regions. The percentage of synapomorphic sites among informative sites is highest in the psbA-trnH intergenic spacer (92.3%), and lowest in the trnL-trnF intergenic spacer (66.7%). Therefore, the psbA-trnH spacer, which evolves most rapidly among the three regions and provides best synapomorphic information, should be a useful region for phylogenetic studies at the intrageneric level. The matK coding region, although evolving about twice as slowly as the psbA-trnH spacer, has over twice more synapomorphic sites than the intergenic spacer because it is about four times longer and contains ' fewer homoplasious sites. The matK coding region, therefore, may also be a good marker for phylogenetic studies at the intrageneric level. The most frequently used intergenic spacer, trnL-trnF, however, evolves most slowly and provides the most homoplasious phylogenetic information among these three regions, and thus its phylogenetic utility at the intrageneric level is questionable.

Insertions/deletions. Of eleven indels in the psbA-trnH intergenic spacer, four are perfect (indels 8 and 11) or imperfect (indels 2 and 3) duplications or deletions of prior duplications of adjacent sequences (Fig. 7). Slipped-strand mispairing is most likely the mechanism responsible for this type of indels. Indels 7, 9, and 10, which are duplications or deletions of a portion of poly(T) tracks, may also result from slipped-strand mispairing. Since the probability of occurrence of further insertions or deletions increases as the track of repetitive nucleotide sequences gets longer (Streisinger and Owen, 1985; Golenberg et al., 1993), multiple indels roust have occurred at indels 7 and 9 to create the pattern of differential lengths of poly(T) tracks. In the trnL-trnF intergenic spacer, three of six indels are also portions of poly(T) tracks.

Indels in both intergenic spacers provide relatively reliable phylogenetic information (Table 4). The percentages of phylogenetically informative indels are 81.8% and 50% in the psbA-trnH and trnL-trnF intergenic spacers, respectively. As phylogenetic characters, indels do not conflict with each other or with nucleotide substitutions in either of the two intergenic spacers. Only indel 8 of the psbA-trnH intergenic spacer appears to be conflict with relationships in matK and ITS phylogenies (Figs., 7-9).

78 Therefore, indels in the intergenic spacers are likely to be reliable phylogenetic characters at the intrageneric level (Ham et al., 1994; Mes and Hart, 1994). However, they may not be reliable phylogenetic characters at higher taxonomic levels because the chance of superproposition of indels increases as divergence time increases (Morton and Clegg, 1993; Golenberg et al., 1993).

Phylogenetic reconstruction. On the phylogenetic tree obtained from sequences of the psbA-trnH intergenic spacer, each of the three sections of Paeonia forms a strongly supported monophyletic group (Fig. 9) . The species in subsection Vaginatae of section Moutan are grouped monophyletically. In section Paeonia. although two taxonomic subsections are recognized, it is not expected that species in each subsection would form a monophyletic group on any gene phylogeny because the section has undergone extensive reticulate evolution which made classification based on morphology very difficult (Stebbins, 1948; Sang et al., 1995a). The phylogenetic analysis of the psbA-trnH intergenic spacer resolved only two clades within section Paeonia. The clade containing 22 species supported only by indel 8 conflicts with relationships on the matK phylogeny. It is very unlikely that this clade reflects the species phylogeny, but just random deletions of the ATT duplication in P. emodi. P. sterniana, P. clusii and two subspecies of P. mascula (Figs. 7-9). Yet, because P. clusii and the two (the sites where substitutions are shared by two or more taxa) among the variable sites is similar for the three regions. The percentage of synapomorphic sites among informative sites is highest in the psbA-trnH intergenic spacer (92.3%), and lowest in the trnL-trnF intergenic spacer (66.7%). Therefore, the psbA-trnH spacer, which evolves most rapidly among the three regions and provides best synapomorphic information, should be a useful region for phylogenetic studies at the intrageneric level. The matK coding region, although evolving about twice as slowly as the psbA-trnH spacer, has over twice more synapomorphic sites than the intergenic spacer because it is about four times longer and contains ' fewer homoplasious sites. The matK coding region, therefore, may also be a good marker for phylogenetic studies at the intrageneric level. The most frequently used intergenic spacer, trnL-trnF, however, evolves most slowly and provides the most homoplasious phylogenetic information among these three regions, and thus its phylogenetic utility at the intrageneric level is questionable.

Apparently indels in these two intergenic spacers occur less frequently than nucleotide substitutions (Table 4). The ratios of indels to variable sites are 0.46 and 0.60 in the psbA-trnH and trnL-trnF intergenic spacers, respectively. The ratios of synapomorphic indels to synapomorphic sites are higher, i.e., 0.67 and 0.75 for the psbA-trnH and trnL-trnF intergenic spacers, respectively. The indels tend to be stabilized at the sectional and subsectional levels. Therefore, indels in the intergenic spacers are likely to be reliable phylogenetic characters at the intrageneric level (Ham et al., 1994; Mes and Hart, 1994). However, they may not be reliable phylogenetic characters at higher taxonomic levels because the chance of superproposition of indels increases as divergence time increases (Morton and Clegg, 1993; Golenberg et al., 1993).

Phylogenetic reconstruction. On the phylogenetic tree obtained from sequences of the psbA-trnH intergenic spacer, each of the three sections of Paeonia forms a strongly supported monophyletic group (Fig. 9) . The species in subsection Vaginatae of section Moutan are grouped monophyletically. In section Paeonia. although two taxonomic subsections are recognized, it is not expected that species in each subsection would form a monophyletic group on any gene phylogeny because the section has undergone extensive reticulate evolution which made classification based on morphology very difficult (Stebbins, 1948; Sang et al., 1995a). The phylogenetic analysis of the psbA-trnH intergenic spacer resolved only two clades within section Paeonia. The clade containing 22 species supported only by indel 8 conflicts with relationships on the matK phylogeny. It is very unlikely that this clade reflects the species phylogeny, but just random deletions of the ATT duplication in P. emodi, P. sterniana P. clusii. and two subspecies of P. mascula (Figs. 7-9) . Yet, because P. clusii and the two P. mascula subspecies share a nucleotide substitution in the tmL-trnF intergenic spacer (Table 2) , and P. emodi is most closely related to P. sterniana in matK phylogenies (Fig. the deletion may have occurred independently, once in the common ancestor of P. clusii and two subspecies of .P. mascula. and once in P. emodi and P. sterniana. The other clade containing P. lactiflora and P. xinjiangensis is concordant with the matK phylogeny. Resolution of relationships within section Paeonia is poor in the psbA-trnH spacer phylogeny because the intergenic spacer is too short to yield sufficient phylogenetic information for very closely related species

The trnL-trnF intergenic spacer provides very limited phylogenetic information (Table 2). Two indels distinguish section Oneapia from the other two sections. One indel and one nucleotide substitution serve as synapomorphies for section Paeonia. One substitution defines subsection Vaginatae. One substitution supports the sister relationship of P. lactiflora and P. xinjiangensis. as on the psbA-trnH spacer and matK phylogenies. One shared substitution by P. californica and P. rockii. however, is clearly homoplasious. Another apparent homoplasious substitution shared by eight species (OBO, STE, WIT, BAN, ARI, HUM, OFF, and PAR) may result from a synapomorphic substitution that defines the monophyletic group on the matK phylogeny (JAP, OBO, WIT, BAN ARI, HUM, OFF, AND PAR) followed by two independent reversal substitutions in P. japonica and P. sterniana (Table 2, 9) This explanation is in agreement with the matK and phylogenies where P. japonica is a sister group of JP. obovata,, and P. sterniana has the closest relationship to P. emodi.

Therefore, both intergenic spacers provide reliable phylogenetic information at the sectional and subsectional levels, but little resolution within section Paeonia. matK coding region, however, serves as a better phylogenetic marker for resolving close intrasectional relationships within this section (Fig. 9) . The only additional resolution obtained from the intergenic spacers is possibly monophyletic group consisting of P. clusii and two subspecies of P. mascula. Therefore, it seems necessary to sequence multiple rapidly evolving coding and noncoding regions to resolve very close interspecific relationships using cpDNA sequences.

In conclusion, the rapidly evolving psbA-trnH intergenic spacer with few homoplasious mutations can be a useful phylogenetic marker for assessing intrageneric relationships in plants. The frequently used trnL-trnF intergenic spacer, which evolves at on forth the rate of psbA-trnH spacer, has a higher percentage of homoplasy, may not be a good phylogenetic marker at the intrageneric level. A clear understanding of the nature of mutations in DNA sequences, such as inversions and substitutions facilitated by the stem-loop structure in the psbA-trnH intergenic spacer, is critical to appropriate interpretation of the mutations for phylogenetic reconstructions. A short cpDNA intergenic spacer alone may not provide enough synapomorphic mutations to resolve close interspecific relationships which may be assessed by sequencing multiple rapidly evolving coding and noncoding regions in the chloroplast genome.