Am. J. Bot. Plant Physiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (148)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Koch, M.
Right arrow Articles by Mitchell-Olds, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Koch, M.
Right arrow Articles by Mitchell-Olds, T.
Agricola
Right arrow Articles by Koch, M.
Right arrow Articles by Mitchell-Olds, T.
(American Journal of Botany. 2001;88:534-544.)
© 2001 Botanical Society of America, Inc.

Molecular systematics of the Brassicaceae: evidence from coding plastidic matK and nuclear Chs sequences1

Marcus Koch 2, 4, Bernhard Haubold 3 and Thomas Mitchell-Olds 3

2Institute of Botany, University of Agricultural Science, Gregor-Mendel-Str. 33, A-1180 Vienna, Austria; and 3Max-Planck-Institute for Chemical Ecology, Department of Genetics and Evolution, D-07745 Jena, Germany

Received for publication November 23, 1999. Accepted for publication April 25, 2000.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Phylogenetic relationships were inferred using nucleotide sequence variation of the nuclear-encoded chalcone synthase gene (Chs) and the chloroplast gene matK for members of five tribes from the family Brassicaceae to analyze tribal and subtribal structures. Phylogenetic trees from individual data sets are mostly in congruence with the results from a combined matK-Chs analysis with a total of 2721 base pairs, but with greater resolution and higher statistical support for deeper branching patterns. The analysis indicates that tribes Lepidieae, Arabideae, and Sisymbrieae are not monophyletic. Among taxa under study four different lineages each were detected in tribes Arabideae and Lepidieae, interspersed with taxa from tribes Sisymbrieae, Hesperideae, and Brassiceae. It is concluded that tribe Brassiceae might be the only monophyletic group of the traditional tribes. From our data we estimated several divergence times for different lineages among cruciferous plants: 5.8 mya (million years ago) for the Arabidopsis–Cardaminopsis split, 20 mya for the Brassica–Arabidopsis split, and ~40 mya for the age of the deepest split between the most basal crucifer Aethionema and remaining cruciferous taxa.

Key Words: Brassicaceae • chalcone synthase • convergent evolution • maturase K • molecular clock • molecular phylogenetics


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Although the family Brassicaceae includes several agronomically important species such as Brassica cultivars and the molecular model plant Arabidopsis thaliana, phylogenetic relationships are poorly understood. The family by itself, comprising ~340 genera and 3350 species, is well defined by flower architecture and is separated from other families of the Capparales. Several authors have tried to provide a natural system to divide the family of Brassicaceae into tribes (Hayek, 1911 ; Schulz, 1936 ; Janchen, 1942 ; Al-Shehbaz, 1984 ). In the system of Hayek (1911) the Brassicaceae were divided into ten tribes; Schulz (1936) recognized 19 different tribes, and, finally, Janchen (1942) tried to present a comprehensive "natural" system with 15 tribes. These studies were based on a small number of morphological characters such as fruit shape, position of the embryo and cotyledons. However, most of the characters considered, e.g., fruit properties (Mummenhoff, Franzke, and Koch, 1997 ; Koch, Mummenhoff, and Hurka, 1999 ), are subject to convergent evolution, at least on the tribal and subtribal level (Hedge, 1976 ; Al-Shehbaz, 1984 ). Only tribes Brassiceae, Thelypodieae, and Lepidieae have been long thought to be natural groups (Zunk, Mummenhoff, and Hurka, 1999 ). However, Koch, Bishop, and Mitchell-Olds (1999) showed that some taxa from tribe Lepidieae (e.g., Capsella rubella) are closer to taxa from tribe Arabideae than to Lepidieae. It has also been suggested that some taxa from tribe Thelypodieae should be included in tribe Lepideae (Zunk et al., 1996 ), because they were closely related to genus Thlaspi from tribe Lepidieae. There are only a few molecular studies of the phylogeny of crucifers on a higher taxonomic level. These studies mostly focused on tribe Brassicinae and its subtribes, which is thought to be the most likely natural tribe (e.g., Warwick and Black, 1997 and references therein; Francisco- Ortega et al., 1999 ). More detailed phylogenetic information is available for tribes Lepidieae and Arabideae and their subtribes (Zunk et al., 1996 ; Koch, Bishop, and Mitchell-Olds, 1999 ; Koch, Haubold, and Mitchell-Olds, 1999 ; Koch, Mummenhoff, and Hurka, 1999 ).

A first analysis using plastid DNA markers demonstrated that tribal relationships might be highly artificial (Price, Palmer, and Al-Shehbaz, 1994 ). Taxa from four tribes (eight taxa) were analyzed using rbcL sequence variation, and five tribes (19 taxa) were considered using chloroplast DNA restriction site variation. These analyses demonstrated (1) closer affinities of tribe Thelypodieae to Sisymbrieae and Brassiceae and to some taxa of tribe Lepidieae as shown by Zunk et al. (1996) , and (2) the highly artificial status of tribes Arabideae and Sisymbrieae.

Galloway, Malmberg, and Price (1998) sequenced the nuclear-encoded arginine decarboxylase locus (Adc) and the plastid F subunit of NADH dehydrogenase (ndhF) from 13 taxa covering a broader range of crucifers and found complete congruence of both gene trees, but taxa sampling was not sufficient to draw any conclusions about tribal relationships. Using alcohol dehydrogenase (Adh) sequence information, Miyashita et al. (1998) investigated a different set of 13 crucifers concentrating on tribe Arabideae and the genera Arabidopsis and Arabis. They concluded that Arabidopsis is not monophyletic and proposed that Arabidopsis should be split into several genera. These findings have been concluded first by O'Kane and Al-Shehbaz (1997) . In a comprehensive molecular analysis using sequence information of the internal transcribed spacer regions (ITS1 and ITS2) of the nuclear ribosomal DNA this hypothesis was confirmed (Al-Shehbaz, O'Kane, and Price, 1999 ; Koch, Bishop, and Mitchell-Olds, 1999 ), and it has been shown that tribe Arabideae is an unnatural group (Koch, Bishop, and Mitchell-Olds, 1999 ). In summary, these analyses provided some evidence that all tribes investigated in previous studies, i.e., Arabideae, Sisymbrieae, and even tribe Lepidieae, with the exception of tribe Brassiceae, are not natural groups as treated traditionally.

In the present study we test the hypothesis that tribes Lepidieae and Arabideae are polyphyletic. We further investigate the phylogenetic position of several taxa from the tribes Brassiceae, Sisymbrieae, and Hesperideae. For this purpose we used coding regions from the nuclear genome and the plastome. Chalcone synthase (Chs) is a nuclear gene that plays a central role in secondary metabolism (EC 2.3.1.74) of flavonoid biosynthesis. It has been shown to be extremely useful for phylogenetic analysis among cruciferous plants (Koch, Haubold, and Mitchell-Olds, 2000 ) and has provided high confidence in reconstructions, particularly at deeper nodes. A synonymous substitution rate (mutations per site per year) of 1.4 x 10-8 has been observed equally distributed along all branches (Koch, Haubold, and Mitchell-Olds, 2000) . The chloroplast matK gene encodes a maturase and is located within the trnK intron (Neuhaus and Link, 1987 ). It evolves nearly two to three times faster than rbcL (Johnson and Soltis, 1994, 1995 ), and matK sequence information data have been used successfully to resolve generic and even species-level relationships (Steele and Vigalys, 1994 ; Kron, 1997 ; Brochmann et al., 1998 ).

Combined data sets without conflictual phylogenetic signal have been used previously and provided more resolution and internal support for relationships than computation based on individual data sets (e.g., Soltis et al., 1993, 1996 ; Johnson and Soltis, 1994, 1995 ; Olmstead and Sweere, 1994; Xiang, Soltis, and Soltis, 1998 ). We used this approach to infer a robust phylogeny for 48 cruciferous taxa from five tribes.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 LITERATURE CITED
 
Taxa were choosen to comprise the whole set of species analyzed by Koch, Bishop, and Mitchell-Olds (1999) , focusing on molecular systematics and evolution of Arabis and Arabidopsis. In this analysis the position of a few taxa such as Capsella rubella, Turrits glabra, and Crucihimalaya himalaica were only weakly supported, and herein we tried to infer a more robust phylogeny. Several additional taxa from the tribes Brassiceae (Sinapis, Raphanus), Sisymbrieae (Sisymbrium), Lepidieae (Thlaspi, Cochlearia and Ionopsidium), Hesperideae (Matthiola), and Alliaria petiolata,which has been placed into different tribes (Arabideae, Sisymbrieae, or Lepidieae), were analyzed to investigate tribal relationships focusing on Lepidieae, Arabideae, and Brassicinae. Aethionema grandiflora served as an outgroup. This species has already been shown on the molecular level to be only distantly related to other crucifers (Galloway, Malmberg, and Price, 1998 ; Koch, Haubold, and Mitchell-Olds, 1999 ). However, traditionally Aethionema is placed in tribe Lepidieae (Hayek, 1911 ; Schulz, 1936 ; Janchen, 1942 ).

Plant material and DNA isolation
The species of Brassicaceae studied are listed in Table 1. DNA extractions were carried out as described in Koch, Haubold, and Mitchell-Olds (1999) . In previous studies a subset of the same accessions from this analysis were investigated using molecular markers (Koch, Bishop, and Mitchell-Olds, 1999 ; Koch, Haubold, and Mitchell-Olds, 2000 ). Sequences obtained for maturase K (matK) and chalkone synthase (Chs) genes from this study have been deposited in GenBank (see GenBank accession numbers given in Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Plant material, accession data, and GenBank accession numbers

 
Plastidic matK
Double-stranded DNAs of matK were amplified using the polymerase chain reaction (PCR) with trnK-710F* and trnK-2R* as PCR primers. Both primers were originally designed by Johnson and Soltis (1995) , but for the analysis presented here we optimized them using the entire DNA sequence from Sinapis alba as reference (GenBank X04826). Primer sequences are shown in Table 2. The resulting DNA fragments contained a 72-bp (base pair) fragment downstream from the matK gene (Sinapis alba as reference), the entire maturase K gene, and a 325-bp fragment upstream from the matK gene (Sinapis alba as reference). For further analysis only the matK gene including the ATG start codon and the stop codon was used. For each purified amplification product (PCR product purification kit, Boehringer Inc., Mannheim-Ingelheim, Germany) both strands were cycle-sequenced using the Taq DyeDeoxy Terminator Cycle Sequencing Kit (PE Biosystems, Inc., Foster City, U.S.A.) and the following primers: trnK-710F*, trnK-2R*, matK-1F, matK-16F, matK-312F, matK-495F, matK-495R, matK-1010F, matK-1010R, matK-1089R, and matK-1495R. Primer sequence data are provided in Table 2. This approach resulted in highly redundant sequence information and enabled us to obtain unambiguous sequences. To test the amount of infraspecific sequence variation we sequenced ten ecotypes from Arabidopsis thaliana, and different accessions from Arabidopsis petraea (three accessions), Arabis alpina (two accessions), and Arabis drummondii (two accessions) (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 2. Primers for matK amplification and sequencing (in 5'-3' direction)

 
Nuclear-encoded chalcone synthase loci
Chalcone synthase (Chs, E.C. number 2.3.1.74) participates in plant secondary metabolism by catalyzing the condensation of three molecules of malonyl-CoA and one molecule of p- coumaryl-CoA to yield chalcone, a precursor in the biosynthesis of flavonoids. Chs belongs to multigene families in plant taxa such as petunia, Ipomoea species, and legumes (Koes, Spelt, and Elzen, 1989 ; Durbin et al., 1995 ; Ryder et al., 1987 ; Wingender et al., 1989 ; An et al., 1993 ; Junghans, Dalkin, and Dixon, 1993 ; Howles, Arioli, and Weinman, 1995 ). However, in diploid taxa of Brassicaceae this gene has been successfully used for phylogenetic analysis and has been shown to be single copy (Koch, Haubold, and Mitchell- Olds, 2000) as in Arabidopsis thaliana (Forkman, 1993 ). Southern blot experiments (data not shown) strongly supported that there is only one orthologous copy of the gene among diploids analyzed in this study. Different gene copies indicating multiple paralogs for Chs have been reported for Sinapis (Durbin et al., 1995 ). However, there is strong evidence for ancient genome multiplication among Sinapis and its relatives (such as Brassica and Raphanus; Warwick and Black, 1991 ), leading to multiple Chs copies (Sadowski et al., 1996 ; Cavell et al., 1998 ).

Amplification, cloning, and sequencing of Chs
PCR amplification of the entire gene including the promoter region were performed using the primers CHS-FOR1 and CHS-REV5 designed by Koch, Haubold, and Mitchell-Olds (1999) . All PCR amplifications resulted in only one predominant band on an agarose gel. PCR products were purified from an agarose gel using the Boehringer PCR product purification kit and cloned either into the pGEM-T cloning vector (Promega, Mannheim, Germany) or into the TA kit pCR II cloning vector (Invitrogen, Groningen, The Netherlands). One clone each from three independent PCR reactions per single DNA preparation was cycle-sequenced as described for the matK gene to detect possible allelic variation. As sequencing primers we used the amplification primers, the vector-specific universal t7 forward and M13–48 reverse primers, and sequencing primers CHS- FOR2, CHS-REV2, CHS-FOR3, CHS-REV3, CHS-FOR4, and CHS-REV4 (Koch, Haubold, and Mitchell-Olds, 2000) .

Phylogenetic analysis
Fifty-four samples were analyzed for sequence variation within the matK gene. This included ten ecotypes of Arabidopsis obtained from the stock centers. Insertions and deletions were treated as missing data. Characters and character states were weighted equally (Fitch parsimony). The analysis were run using PAUP 4.0* beta version (Swofford, 1999 ) under the following conditions: HEURISTIC, TBR, STEEPEST DESCENT. The bootstrap option of PAUP (1000 replicates) and a decay analysis (Donoghue et al., 1991) were used to assess relative support in the unweighted analysis. Weighted parsimony was performed according to the criterion of Albert, Chase, and Mishler (1993) after estimation of the transition/transversion ratio [using the commands in PAUP 4.0* beta version: DSET DISTANCE = ABS SUBST = TRATIO; SHOWDIST; SAVEDIST FORMAT = TABTEXT FILE = XXX]. An estimate of the phylogenetic signal present within the matK data was obtained using the Random Trees option of PAUP 3.1.1 (Swofford, 1993 ). The skewness (-g1) of the distribution of 10 000 random trees was evaluated.

Forty-four accessions were analyzed for sequence variation within the Chs gene. Three of them originated from databases (Matthiola incana, Raphanus sativus, and Sinapis alba). Twenty-eight Chs sequences (with two alleles for Olimarabidopsis pumila) have been reported previously (Koch, Haubold, and Mitchell-Olds, 2000) , and 13 sequences are new in this study. The same DNA material has been used as for the matK analysis. Both data sets comprise the same taxa with the exception of Cochlearia pyrenaica (only matK), Ionopsidium prolongoi (only matK), Raphanus sativus (only Chs). Phylogenetic analyses were performed in the same way as for the matK gene.

After comparing tree topologies from Chs and matK separately, both data sets were combined to a total data matrix. Parsimony analysis were performed in the same way as for the separate data matrices. Additionally, a distance method was used to compare the impact of different phylogenetic analysis approaches on the tree topology. Phylogenetic distances were computed using Kimura's two-parameter model (Kimura, 1980 ), and the resulting distance matrix was subjected to the neighbor-joining algorithm as implemented in PHYLIP (Felsenstein, 1995 ). One thousand bootstrap samples were analyzed to assess the significance of nodes on the original neighbor-joining tree.

Rates of synonymous (Ks) and nonsynonymous (Ka) substitutions were calculated according to Li's method (1993) as implemented in the li93 program (Wolfe, 1993 , University of Dublin, ftp://acer.gen.tcd.ie/pub/khwolfe/li93).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 LITERATURE CITED
 
The matK-phylogenetic analysis
The length of matK sequences varied from 1509 bp (Sisymbrium irio) to 1521 bp and the sequences were aligned by hand with a total alignment of 1521 bp. Two amino acid insertions were observed for the two Arabis alpina accessions at nucleotide positions 154–159 (GAT-GCC corresponding to the amino acids aspartate and alanine) not found in the remaining taxa. A deletion of two amino acids at nucleotide positions 262–267 has been observed in Arabis blepharophylla, Sinapis alba, and Sisymbrium irio. The total G/C content is 31.1%. Among the 11 Arabidopsis accessions, including the ten different ecotypes, no variation could be detected. The same is true for Cochlearia danica vs. C. pyrenaica and R. amphibia vs. Rorippa palustris. Two accessions of Arabis drummondii differed at three nucleotide positions (0.19%), two accessions of Arabis alpina at three nucleotide positions (0.19%), and three accessions of Arabidopsis lyrata [= Cardaminopsis petraea] at 3–11 nucleotides (0.19–0.72%).

Of the 1521 bp of the aligned matK gene 1037 sites were invariable (68.2%); of the remaining 484 variable sites 220 sites (14.5%) were potentially parsimony informative. The overall sequence distance values go up to a maximum of 8.3% including the outgroup Aethionema grandiflora; without the outgroup highest sequence distance values are ~6.6%.

Fitch parsimony resulted in one most parsimonious tree (MPT) of 824 steps [consistency index (CI) = 0.72, retention index (RI) = 0.75] (Fig. 1). Estimation of the transition/transversion ratio provides a value of 1.09 ({sigma}n = 0.24) with a minimum between Arabis alpina accession from Europe and Africa (0.4) and a maximum between Olimarabidopsis pumila and Boechera divaricarpa (2,09) and within Arabidopsis lyrata from Germany and the USA (2.66). This value is comparable to a transition-transversion ratio of 1.21 found among matK sequences within Cornales (Xiang, Soltis, and Soltis, 1998 ). Weighted parsimony using a transition transversion value of 1.09 for character state weighting yielded the same tree topology as unordered Fitch parsimony (Fig. 1). Characteristic for the matK analysis is that basal branching patterns are only weakly supported and mostly exhibit bootstrap values below 50%, corresponding to a DECAY value of +1.



View larger version (40K):
[in this window]
[in a new window]
  		Fig. 1. Single most parsimonious tree (MPT) for the analysis of the matK gene. Bootstrap values in percent (above branches) from 1000 replicates and decay values (below branches) are indicated. Information about tribal assignments according to Janchen (1942) is given on the right margin

 
The overall ratio of synonymous vs. nonsynonymous nucleotide changes (Ks/Ka) is 2.6. This is comparable to other plastid coding regions such as ndhF (Ks/Ka = 2.6; Galloway, Malmberg, and Price, 1998 ). A summary of plastidic Ks/Ka values is given in Wolfe, Sharp, and Li (1989) for a Solanaceae/ Brassicaceae comparison. They found values for the chloroplast genes psbK (4.8), atpE (6.6) and orfK (= matK, 2.8). Based on molecular clock assumptions within cruciferous plants (Koch, Haubold, and Mitchell-Olds, 2000) we estimated for matK sequences of cruciferous plants a synonymous substitution rate of 1.0–3.0 x 10-9. We tested this estimation with data provided by Yang et al. (1999) . On the base of the mitochondrial nad4 gene they calculated a date of divergence of 14.5–20.4 mya for Brassica vs. Arabidopsis. Our corresponding Ks value of matK between the Arabidopsis thaliana clade and representatives from a clade (see Fig. 1) consisting of Brassica relatives (Fourraea alpina vs. Arabidopsis thaliana) is 0.056, which yields an estimate for the mutation rate of 1.4– 1.9 x 10-9 mutations per site per year using the date of divergence of 14.5–20.4 mya. Previous studies produced different estimates (4 x 10-10 in Paeonia; Sang, Crawford, and Stuessy, 1997 ), which might reflect different molecular clocks in different lineages among flowering plants.

The Chs-phylogenetic analysis
The chalcone synthase sequences varied between 1176 bp in Arabis alpina accessions and 1200 bp in Ionopsidium abulense. All deletions and insertions were between nucleotide positions 4–28 at the 5'end in the first exon. The single intron, located between nucleotide positions 207 and 208, was deleted completely for all subsequent analysis. The total G/C content of the coding region is 53.8%. Intraspecific variability due to allelic variation has been found in Arabis alpina (0.48%), Aubrieta deltoidea (0.42%), Arabidopsis lyrata (0.76%), and Arabis drummondii (0.51%). From tetraploid Olimarabidopsis pumila we isolated two different Chs genes, which are probably alleles of duplicated loci because of a sequence divergence of 2.0%, which is significantly higher than the intraspecific allelic variation described above (Koch, Haubold, and Mitchell-Olds, 2000) . Of the 1200 bp of the aligned Chs gene 683 sites were invariable (56.9%). From the remaining 517 variable sites 380 sites (31.7%) were potentially parsimony informative. The overall sequence distance values go up to a maximum of 22.9% including the outgroup Aethionema grandiflora. Without the outgroup the highest sequence distance values are ~16.5%.

Fitch parsimony yielded five MPTs 1886 steps in length [CI = 0.43, RI = 0.68, RC = 0.29]. In contrast to matK analysis more basal branching patterns of the strict consensus tree (Fig. 2) have higher statistical support, mostly indicated by higher DECAY values of +2 to +4. However, bootstrap support above 50% is rare along those branches (Fig. 2). Estimation of the transition/transversion ratio gave a value of 1.47 ({sigma}n = 0.42) with a minimum of 0.40 (Aubrieta deltoidea vs. Cardamine amara) and a maximum of 2.68 (Olimarabidopsis pumila vs. Aubrieta deltoidea). Analyzing the Adc locus among cruciferous plants, Galloway, Malmberg, and Price (1998) estimated a comparable transition transversion ratio of 1.34. Weighted parsimony using a transition transversion ratio of 1.47 resulted in five MPTs differing in topology from the fitch parsimony (CI = 0.42, RI = 0.63, RC = 0.26). The strict consensus tree of the weighted parsimony approach (Fig. 3) differs from the strict consensus tree obtained by the Fitch parsimony approach in the relative positions of Lepidium campestre and Matthiola incana. However, the strict consensus tree obtained by the weighted parsimony approach is more similar to the MPT from matK sequence data regarding the positions of Matthiola incana and Lepidium campestre.



View larger version (43K):
[in this window]
[in a new window]
  		Fig. 2. Strict consensus tree from five MPTs for the Chs gene using unweighted Fitch parsimony. Bootstrap values in percent (above branches) from 1000 replicates and decay values (below branches) are indicated. Information about tribal assignments according to Janchen (1942) is given on the right margin

 


View larger version (41K):
[in this window]
[in a new window]
  		Fig. 3. Strict consensus tree from five MPTs for the Chs gene using weighted parsimony with a transition:tranversition ratio of 1:1.47. Bootstrap values in percent (above branches) from 1000 replicates are indicated. Information about tribal assignments according to Janchen (1942) is given on the right margin

 
The ratio of synonymous vs. nonsynonymous nucleotide changes (Ks/Ka) is 15.3. Similar values were obtained for the nuclear-encoded arginine decarboxylase (Ks/Ka = 13.4 for Adc1 and Ks/Ka = 13.8 for Adc2) (Galloway, Malmberg, and Price, 1998 ). In a comparison of nuclear genes from Brassicaceae vs. Solanaceae similar values were obtained for gapC (21.5), gapA (22.8), EPSPs (15.1), chalcone synthase (>25), and acetolactate synthase (>25), summarized in Wolfe, Sharp, and Li (1989) . Recently, we estimated a synonymous substitution rate for Chs sequences from cruciferous plants of ~1.4 x 10-8 substitutions per site per year (Koch, Haubold, and Mitchell-Olds, 2000) . Previous estimates differed by a factor of 2.5–1.75 [Wolfe, Sharp, and Li (1989) : 4–5.6 x 10-9 substitutions per site per year; Durbin et al. (1995) : 8 x 10-9].

Comparison of the phylogeny based on matK sequence data (Fig. 1) with the phylogeny based on Chs sequence data (Figs. 2, 3) revealed several differences. Single taxa such as Lepidium campestre, Mathiola incana, Crucihimalaya himalaica, Microthlasi perfoliatum, and Capsella rubella were placed at different positions. But for these taxa there is only in the matK or in the Chs phylogentic tree significant phylogenetic signal (as indicated by high bootstrap or decay values) for a particular position. This holds also true for the relative position of a group of taxa comprising Thlaspi and Brassica (Sinapis, Raphanus, Sisymbrium) relatives. In the matK analysis this group became a sister to Eurasian Arabis (Fig. 1), which is also true for Fitch parsimony analysis of the Chs data (Fig. 2). In contrast, weighted parsimony analysis of the Chs data placed this group outside a European Arabis/Aubrieta clade. Several major clades can be recognized in both data sets: (1) Cochlearia and Ionopsidium as a monophyletic clade, only distantly related to the remaining ingroup taxa, (2) Barbarea, Cardamine, and Rorippa, (3) a Arabis/Aubrieta group, (4) Arabidopsis thaliana and its relatives, (5) North American Arabis (these taxa should be united under the genus Boechera [Löve and Löve, 1975 ]), and (6) Turritis and Olimarabidopsis.

The Chs-matK combined data-phylogenetic analysis
The combined data set with 2721 bp comprises 1720 invariable characters (63.2%), 383 unique nucleotide changes (14.1%), and 618 parsimony-informative characters (22.7%). In the case of Raphanus sativus (matK sequence data were missing) and Crucihimalaya wallichii (Chs sequence data were missing) we treated missing data as unknown characters. In the case of Cochlearia pyrenaica and Rorippa palustris, matK sequences were identical to the corresponding Cochlearia danica and Rorippa amphibia matK sequences, respectively, and the Chs gene was not sequenced. Fitch parsimony resulted in nine MPTs (CI = 0.51, RI = 0.70) with a tree length of 2729 steps, 19 steps longer than the sum of minimum steps for matK (824 steps) and Chs (1886 steps) data sets. The strict consensus tree is shown in Fig. 4.



View larger version (43K):
[in this window]
[in a new window]
  		Fig. 4. Strict consensus tree from nine MPTs for the combined data set of the matK and the Chs gene using unweighted fitch-parsimony. Bootstrap values in percent (above branches) from 1000 replicates and decay values (below branches) are indicated. Information about tribal assignments according to Janchen (1942) is given on the right margin

 
The distance neighbor-joining tree (Fig. 5) demonstrates that relative branch length is largely in congruence with the parsimony tree (Fig. 4). Both approaches differ only in the relative positions of Fourraea alpina and Microthlaspi perfoliatum. However, the position of Matthiola incana is only weakly supported as shown for the individual data sets (Figs. 1–3), but for Lepidium campestre much greater confidence values for the phylogenetic position were obtained (Figs. 4, 5).



View larger version (38K):
[in this window]
[in a new window]
  		Fig. 5. Neighbor-joining distance tree for the matK-Chs combined data set. Bootstrap values in percent (above branches) from 1000 replicates are indicated. For nodes A–F dates of divergence have been calculated for both genes separately and are summarized in Table 3 . Information about tribal assignments according to Janchen (1942) is given on the right margin

 
Assumptions on divergence times
We used our estimates of the synonymous mutation rates for Chs (1.4 x 10-8) and matK (1.7 x 10-9) to date several pivotal points on the gene trees. For our estimates of several divergence times we used the simple equation Ks/2T = synonymous mutation rate (T = divergence time). The data for the pairwise Ks values are not shown and are available upon request. The divergence time of several nodes that are indicated in Fig. 5 have been calculated for both genes separately and are highly comparable (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 3. Date of divergence at different nodes indicated in Fig. 5 (mya = million years ago)

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
LITERATURE CITED
 
Phylogenetic relationships on the tribal level among the Brassicaceae
The results from this study as well as previous studies of phylogenetic relationships comprising Arabis/Arabidopsis relatives (Galloway, Malmberg, and Price, 1998 ; Koch, Bishop, and Mitchell-Olds, 1999 ; Koch, Haubold, and Mitchell-Olds, 2000 ) demonstrated close relationship among tribes Sisymbrieae, Lepidieae, and Arabideae. It has been shown that boundaries between tribes Lepidieae, Lunarieae, Sisymbrieae, Euclidieae, and Alysseae are highly artificial (Zunk, Mummenhoff, and Hurka, 1993 ; Price, Palmer, and Al- Shehbaz, 1994 ). In our analysis we also investigated taxa from tribe Brassiceae, which has been considered to be a natural group (Al-Shehbaz, 1984 ), a view upheld in a series of papers summarized in Warwick and Black (1997) . Not surprinsingly all representatives of this tribe were combined to one clade (Figs. 4, 5). Tribe Brassiceae is characterized by multiplied genomes, even in taxa regarded as to be diploid. Genome collinearity studies comparing the n = 5 genome of Arabidopsis thaliana with diploid Brassica genomes revealed up to three corresponding regions in diploid Brassica (Kowalski et al., 1994 ; Lagercrantz et al., 1995 ). One could speculate about polyploidization as one major force of speciation within this group. Extensive genome interaction might have led to reassembled alleles at different loci via recombination, gene duplication, and subsequent gene silencing. Therefore, it is important to find the orthologous loci. In the case of the Brassicaceae it is possible to use the genomic information of Arabidopsis thaliana to find single-copy genes and to demonstrate orthology via comparative genome analysis (Schmidt et al., 1999 ).

Remaining taxa analyzed in this study belong mostly to tribes Arabideae and Lepidieae, and include numerous species from Arabis and Arabidopsis. A comparative summary of tribal and subtribal classification of taxa analyzed in this study is provided in Table 4. Our phylogentic concept combined with previous phylogenetic analyses based on molecular markers showed that tribe Lepidieae as treated traditionally consists of at least three major lineages: (1) a group combining Cardaria, Coronopus, Stroganowia, Andrzeiowsia, and Hornungia, and the large genus Lepidium (Bowmann et al., 1999 ; Zunk, Mummenhoff, and Hurka, 1999 ), (2) taxa of Thlaspi senso lato representatives including genera such as Peltaria, Alliaria, Thlaspi, and Teesdalia (Zunk et al., 1996 ), and (3) taxa of Cochlearia and Ionopsidium (Koch, Mummenhoff, and Hurka, 1999 ). The phylogenetic trees (Figs. 4 and 5) support the recognition of these separate lineages. Additional lineages from this tribe consist of Capsella rubella (and closely related Neslia and Camelina; Zunk, Mummenhoff, and Hurka, 1999 ) and Aethionema grandiflora (outgroup).


View this table:
[in this window]
[in a new window]
  		Table 4. Tribal and subtribal classification according to Hayek (1911), Schulz (1936), and Janchen (1942)

 
The Arabideae were divided by Hayek (1911) into seven subtribes merging Arabidopsis, Arabis turrita, Cardaminopsis (= Arabidopsis lyrata, halleri, suecica), Cardamine, and Barbarea into subtribe Cardamininae and merging Arabis and Aubrieta into subtribe Arabidinae (Table 4). Our findings totally contradict this treatment. The genus Capsella was placed into tribe Lepidieae, and Hayek (1911) argued that Capsella is unrelated to Arabis and Arabidopsis. Different combinations and delineations of tribes and subtribes were presented by Schulz (1936) and Janchen (1942) , but none of these systems reflects a "true phylogenetic system." Janchen (1942) placed Arabidopsis in subtribe Arabidopsidinae of tribe Sisymbrieae and retained Cardamine (subtribe Cardamininae) and Arabis including Arabis turrita (subtribe Arabidinae) in tribe Arabideae. In addition, he separated Arabidopsis (tribe Sisymbrieae) from the closest relatives (former genus Cardaminopsis), which were merged into subtribe Arabidinae, tribe Arabideae. However, Janchen (1942) also presented a "phylogenetic network" indicating closer relationships of Sisymbrieae with Arabideae and Lepidieae.

Taking these lines of evidence into account, it becomes clear that a new classification of crucifers based on molecular data would be desirable.

Systematic relationships on the family level and the age of the Brassicaceae
Brassicaceae have been included in several analyses of angiosperm taxa (Soltis et al., 1997 ; Rodman et al., 1998 ). Usually the model taxa Arabidopsis thaliana and Brassica oleracea have been chosen. In all these analyses it has been shown that the Brassicaceae form a crown group in the clade of the order Capparales, characterized for example by the mustard-oil glucosides (glucosinolates) and possession of the hydrolytic enzyme myrosinase (ß-thioglucosidase). Both Arabidopsis and Brassica are part of our study, and the phylogenetic distance of ~20 mya since separation between both taxa is remarkable (Table 3). Some additional dates of divergence (Fig. 5, nodes A–F) have been calculated from both genes under study (Table 3). Considering the whole family, including the outgroup Aethionema grandiflora, the deepest split was calculated to be ~40 mya.

The closest known relative of the Brassicaceae that has been analyzed on the molecular level from order Capparales is the genus Cleome from the Capparaceae. Comparing phylogenetic trees based on rbcL and 18S rDNA data (Soltis et al., 1997 ; Rodman et al., 1998 ) regarding the sister relationship Cleome vs. Arabidopsis/Brassica it turned out that the two crucifers are well supported as a single clade. Furthermore, the number of nucleotide changes separating Arabidopsis/Brassica from Cleome is half the number of nucleotide changes along the Cleome branch. A similar result was presented by Galloway, Malmberg, and Price (1998) using ndhF sequence data and substituting Polanisia for Cleome as an outgroup. Polanisia is close to Cleome and belongs to the same subfamily Cleomoideae within the Capparaceae. Cleome has been shown to be present in the Cleome flora from the Upper Miocene in Europe (6–12 mya) (Negru, 1986 ). The oldest known records of the Capparaceae date back to the Upper Eocene (50 mya) (Mai, 1995 ). These lines of evidence agree with our hypothesis that the family Brassicaceae evolved as a crown group of the Capparaceae ~50 mya.

Utility of nuclear coding genes for phylogenetic reconstructions
Previous studies mostly focused on chloroplast DNA (cpDNA) markers to generate data for phylogenetic reconstruction. These data sets were based either on cpDNA restriction site analysis (e.g., Mummenhoff and Koch, 1994 ; Koch, Mummenhoff, and Hurka, 1998 ), sequencing of coding regions such as rbcL, atpB, matK, ndhF, and rpl16 (e.g., Chase et al., 1993 ; Olmstead and Palmer, 1994 ; Olmstead and Sweere, 1994 ; Steele and Vilgalys, 1994 ; Hoot et al., 1997 ; Schnabel and Wendel, 1998 ) or noncoding regions such as the trnL spacer and intron regions (detailed listed in Small et al., 1998 ). In contrast, data sets from the nuclear genome were mostly obtained from high-copy ribosomal DNA regions, either the noncoding ITS (internal transcribed spacer) (e.g., Baldwin et al., 1995 ) or the coding 18S rDNA locus (Soltis et al., 1997 ). There are specific advantages associated with using nuclear or plastid genes. The main characteristic of the plastid markers is the absence of higher levels of recombination, making it possible to follow a single line of plastid transmission. In the Brassicaceae plastids are inherited maternally, whereas the nuclear rDNA markers are inherited biparentally and can therefore provide information on both parents while being greatly affected by concerted evolution (Buckler, Ippolito, and Holtsford, 1997 ). Nuclear single-copy genes are less influenced by concerted evolution and recombination, but have rarely been used for phylogenetic reconstruction among plants (Chs: Durbin et al., 1995 ; Adc: Miyashita, Innan, and Terauchi, 1996 ; Galloway, Malmberg, and Price, 1998 ; Adh: Gaut and Clegg, 1991 ; Hanfstingl et al., 1994 ; Gaut et al., 1996 ; Morton, Gaut, and Clegg, 1996 ; Innan et al., 1996 ; Sang, Donoghue, and Zhang, 1997 ; Charlesworth, Liu, and Zhang, 1998 ; Small et al., 1998 ). Demonstration of orthology is a prerequisite for computing adequate gene trees and corresponding species trees. If this is done either by southern hybridization, comparative genetic mapping data, or comparison of several gene trees for taxa under study (Small et al., 1998 ), nuclear coding sequences are a powerful tool to resolve phylogenetic relationships. In addition, hybridization and recombination among loci from different lineages can be tested and have been shown not only for Chs in Brassicaceae (Koch, Haubold, and Mitchell-Olds, 2000) but also for the Chs interlocus recombination in Ipomoea (Huttley et al., 1997 ) and Adh intragenic recombination in 17 Arabidopsis thaliana ecotypes (Innan et al., 1996 ).


View this table:
[in this window]
[in a new window]
  		Table 1. Continued

 

   FOOTNOTES
 
1 The authors thank D. Schnabelrauch and S. Ring for technical support and all institutes and colleagues providing seed and plant material. This work was supported in part by the Max-Planck-Gesellschaft, and by grants to TMO from the U.S. National Science Foundation (DEB-9527725). Back

4 Author for correspondence (Tel. ++43-(0)1-47654 3156, FAX +43-(0)1- 47654 4504, e-mail: koch{at}edv1.boku.ac.at) Back


   LITERATURE CITED
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Albert, V. A., M. W. Chase, and B. D. Mishler. 1993 Character-state weighting for cladistic analyis of protein-coding DNA sequences. Annals of the Missouri Botanical Garden 80: 752–766[CrossRef][ISI]

Al-Shehbaz, I. A. 1984 The tribes of cruciferae (Brassicaceae) in the southeastern United States. Journal of the Arnold Arboretum 65: 85–111

———, S. L. O'Kane, and R. A. Price. 1999 Generic placement of species excluded from Arabidopsis (Brassicaceae). Novon 9: 296–307

An, C., Y. Ichinose, T. Yamada, Y. Tanaka, T. Shiraishi, and H. Oku. 1993 Structure and organization of the genes encoding chalcone synthase in Pisum sativum. Plant Molecular Biology 21: 789–803

Baldwin, B. G., M. J. Sanderson, J. M. Porter, M. F. Wojciechowski, C. S. Campbell, and M. J. Donoghue. 1995 The ITS region of nuclear ribosomal DNA: a valuable source of evidence on angiosperm phylogeny. Annals of the Missouri Botanical Garden 82: 247–277[CrossRef][ISI]

Bowmann, J. L., H. Brüggemann, J.-Y. Lee, and K. Mummenhoff. 1999 Evolutionary changes in floral structure within the genus Lepidium L. (Brassicaceae). International Journal of Plant Sciences 160: 917–929[CrossRef][ISI][Medline]

Brochmann, C., Q.-Y. Xiang, S. J. Brunsfeld, D. E. Soltis, and P. S. Soltis. 1998 Molecular evidence for polyploid origins in Saxifraga (Saxifragaceae): the narrow arctic endemic S. svalbardensis and its widespread allies. American Journal of Botany 85: 135–143[Abstract]

Buckler, E. S., A. Ippolito, and T. P. Holtsford. 1997 The evolution of ribosomal DNA—divergent paralogues and phylogenetic implications. Genetics 145: 821–832[Abstract]

Cavell, A. C., D. J. Lydiate, I. A. P. Parkin, C. Dean, and M. Trick. 1998 Collinearity between a 30-centimorgan segment of Arabidopsis thaliana chromosome 4 and duplicated regions within the Brassica napus genome. Genome 41: 62–69[Medline]

Charlesworth, D., F. L. Liu, and L. Zhang. 1998 The evolution of the alcohol dehydrogenase gene family by loss of introns of the genus Leavenworthia (Brassicaceae). Molecular Biology and Evolution 15: 552– 559[Abstract]

Chase, M. W., et al. 1993 Phylogenetics of seed plants: an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80: 528–580[CrossRef][ISI]

Donoghue, M. J., R. G. Olmstead, J. F. Smith, and J. D. Palmer. 1992 Phylogenetic relationships of Dipsacales based on rbcL sequences. Annals of the Missouri Botanical Garden 79: 333–345

Durbin, M. L., G. H. Learn, G. A. Huttley, and M. T. Clegg. 1995 Evolution of the chalcone synthase gene family in the genus Ipomoea. Proceedings of the National Academy of Sciences, USA 92: 3338–3342

Felsenstein, J. 1995 PHYLIP: phylogeny inference package version 3.57c. University of Washington, Seattle, Washington, USA

Forkman, G. 1993 In J. B. Harborne [ed.], The flavonoids: advances in research since 1986, 537–564. Chapman and Hall, London, UK

Francisco-Ortega, J., J. Fuertes-Aguilar, C. Gomez-Campo, A. Santos-Guerra, and R. K. Jansen. 1999 Internal transcribed spacer sequence phylogeny of Crambe L. (Brassicaceae): molecular data reveal two Old World disjunctions. Molecular Phylogenetics and Evolution 11: 361–380[CrossRef][ISI][Medline]

Franzke, A., K. Pollmann, W. Bleeker, R. Kohrt, and H. Hurka. 1998 Molecular systematics of Cardamine and allied genera (Brassicaceae): ITS and noncoding chloroplast DNA. Folia Geobotanica 33: 225–240[ISI]

Galloway, G. L., R. L. Malmberg, and R. A. Price. 1998 Phylogenetic utility of the nuclear gene arginine decarboxylase: an example from Brassicaceae. Molecular Biology and Evolution 15: 1312–1320[Abstract]

Gaut, B. S., and T. M. Clegg. 1991 Molecular evolution of alcohol dehydrogenase 1 in members of the grass family. Proceedings of the National Academy of Sciences, USA 88: 2060–2064[Abstract/Free Full Text]

———, B. R. Morton, B. McCaig, and M. T. Clegg. 1996 Substitution rate comparisons between grasses and palms: synonymous rate differences at the nuclear gene adh parallel rate differences at the plastid gene rbcL. Proceedings of the National Academy of Sciences, USA 93: 10274– 10279[Abstract/Free Full Text]

Hanfstingl, U., A. Berry, E. A. Kellogg, J. T. Costa, W. Rüdiger, and F. M. Ausubel. 1994 Haplotypic divergence coupled with lack of diversity at the Arabidopsis thaliana alcohol dehydrogenase locus: roles for balancing and directional selection. Genetics 138: 811–828[Abstract]

Hayek, A. 1911 Entwurf eines Cruciferensystemes auf phylogenetischer Grundlage. Beiheifte Botanisches Centralblatt 27: 127–335

Hedge, I. C. 1976 A systematic and geographical survey of the Old World cruciferae. In J. G. Vaughan, A. J. MacLeod, and M. B. Jones [eds.], The biology and chemistry of the cruciferae, 1–45. Academic Press, London, UK

Hoot, S. B., J. W. Kadereit, F. R. Blattner, K. B. Jork, A. E. Schwarzbach, and P. R. Crane. 1997 Data congruence and phylogeny of the Papaveraceae s.l. based on four data sets: atpB and rbcL sequences, trnK restriction sites, and morphological charcters. Systematic Botany 22: 575–590[CrossRef][ISI]

Howles, P. A., T. Arioli, and J. J. Weinman. 1995 Nucleotide sequence of additional members of the gene family encoding chalcone synthase in Trifolium subterraneum. Plant Physiology 107: 1035–1036[CrossRef][ISI][Medline]

Huttley, G. A., M. L. Durbin, D. E. Glover, and M. T. Clegg. 1997 Nucleotide polymorphism in the chalcone synthase-A locus and evolution of the chalcone synthase multigene family of common morning glory Ipomoea purpurea. Molecular Ecology 6: 549–558[CrossRef]

Innan, H., F. Tajima, R. Terauchii, and T. Miyashita. 1996 Intragenic recombination in the Adh locus of the wild plant Arabidopsis thaliana. Genetics 143: 1761–1770[Abstract]

Janchen, E. 1942 Das System der Cruciferen. Österreichische Botanische Zeitschrift 91: 1–21[CrossRef]

Johnson, L. A., and D. E. Soltis. 1994 matK DNA sequences and phylogenetic reconstructions in Saxifragaceae sensu stricto. Systematic Botany 19: 143–156

———, and ———. 1995 Phylogenetic inference in Saxifragaceae sensu stricto and Gilia (Polemoniaceae) using matK sequences. Annals of the Missouri Botanical Garden 82: 149–175[CrossRef][ISI]

Junghans, H., K. Dalkin, and R. A. Dixon. 1993 Stress response in alfalfa (Medicago sativa L.) 15. Characterization and expression patterns of members of a subset of the chalcone synthase multigene family. Plant Molecular Biology 22: 239–253[CrossRef][ISI][Medline]

Kimura, M. 1980 A simple method for estimating evolutionary rate of base substitutions through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16: 11–120[CrossRef][ISI][Medline]

Koch, M., J. Bishop, and T. Mitchell-Olds. 1999 Molecular systematics and evolution of Arabidopsis and Arabis. Plant Biology 1: 529–537

———, B. Haubold, and T. Mitchell-Olds. 2000 Comparative evolutionary analysis of the chalcone synthase and alcohol dehydrogenase loci among different lineages of Arabidopsis, Arabis and related genera (Brassicaceae). Molecular Biology and Evolution: accepted for publication

———, K. Mummenhoff, and H. Hurka. 1998 Molecular biogeography and evolution of Microthlaspi perfoliatum s.l. polyploid complex (Brassicaceae): chloroplast DNA and nuclear ribosomal DNA restriction site variation. Canadian Journal of Botany 76: 382–396[CrossRef]

———, ———, and ———. 1999 Molecular phylogenetics of Cochlearia L. (Brassicaceae) and allied genera based on nuclear ribosomal ITS DNA sequence analysis contradict traditional concepts of their evolutionary relationship. Plant Systematics and Evolution 216: 207–230[CrossRef][ISI]

Koes, R. R., C. E. Spelt, and P. J. M. van den Elzen. 1989 Cloning and molecular characterization of the chalcone synthase multigene family of Petunia hybrida. Gene 81: 245–257[CrossRef][ISI][Medline]

Kowalski, S. P., T.-H. Lan, K. A. Feldmann, and A. H. Paterson. 1994 Comparative mapping of Arabidopsis thaliana and Brassica oleracea chromosomes reveals islands of conserved organization. Genetics 138: 499–510[Abstract]

Kron, K. A. 1997 Phylogenetic relationships of Rhododendroideae (Ericaceae). American Journal of Botany 84: 973–980[Abstract]

Lagercrantz, U., J. Putterill, G. Coupland, and D. Lydiate. 1995 Comparative mapping in Arabidopsis and Brassica, fine scale genome collinearity and congruence of genes controlling flowering time. Plant Journal 9: 13–20

Li, W.-H. 1993 Unbiased estimation of the rates of synonymous and nonsynonymous substitutions. Journal of Molecular Evolution 36: 96–99[CrossRef][ISI][Medline]

Löve, A., and D. Löve. 1975 Nomenclatural notes on arctic plants. Botaniska Notiser 128: 497–523[ISI]

Mai, D. H. 1995 Tertiäre Vegetationsgeschichte Europas. Fischer, Jena, Germany

Miyashita, N. T., H. Innan, and R. Terauchi. 1996 Intra- and interspecific variation of the alcohol dehydrogenase locus region in wild plants Arabis gemmifera and Arabidopsis thaliana. Molecular Biology and Evolution 13: 433–436

———, A. Kawabe, H. Innan, and R. Terauchi. 1998 Intra- and interspecific DNA variation and codon bias of the alcohol dehydrogenase (Adh) locus in Arabis and Arabidopsis species. Molecular Biology and Evolution 15: 1420–1429[Free Full Text]

Morton, B. R., B. S. Gaut, and M. T. Clegg. 1996 Evolution of alcohol dehydrogenase genes in the palm and grass families. Proceedings of the National Academy of Sciences, USA 93: 11735–11739[Abstract/Free Full Text]

Mummenhoff, K., and M. Koch. 1994 Chloroplast DNA restriction site variation and phylogenetic relationships in the genus Thlaspi sensu lato (Brassicaceae). Systematic Botany 19: 73–88

———, A. Franzke, and M. Koch. 1997 Molecular data reveal convergence in fruit characters used in the classification of Thlaspi s. l. (Brassicaceae). Botanical Journal of the Linnean Society London 125: 183– 199[CrossRef]

Negru, A. G. 1986 Mèoticekaja flora Severo-Zapadnogo Prièernomorja. Pp. 1–157. Schtiinca, Kischinev, Russia

Neuhaus, H., and G. Link. 1987 The chloroplast tRNALys (UUU) gene from mustard. Current Genetics 11: 251–257[CrossRef][ISI][Medline]

O'Kane, S., and I. A. Al-Shehbaz. 1997 A synopsis of Arabidopsis (Brassicaceae). Novon 7: 323–327[CrossRef][ISI]

Olmstead, R. G., and J. D. Palmer. 1994 Chloroplast DNA systematics: a review of methods and data analysis. American Journal of Botany 81: 1205–1224[CrossRef][ISI]

———, ———, and J. A. Sweere. 1994 Combining data in phylogenetic systematics: an empirical approach using three molecular data sets in the Solanaceae. Systematic Botany 43: 467–481

Price, R. A., J. D. Palmer, and I. A. Al-Shehbaz. 1994 Systematic relationships of Arabidopsis: a molecular and morphological perspective. In E. M. Meyerowitz and C. R. Sommerville [eds.], Arabidopsis, 7–19. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, USA

Rodman, J. E., P. S. Soltis, D. E. Soltis, K. J. Sytsma, and K. G. Karol. 1998 Parallel evolution of glucosinolate biosynthesis inferred from congruent nuclear and plastid gene phylogenies. American Journal of Botany 85: 997–1006[Abstract]

Ryder, T. B., S. A. Hedrick, J. N. Bell, X. Liang, S. D. Clouse, and C. J. Lamb. 1987 Organization and differential activation of a gene family encoding the plant defense enzyme chalcone synthase in Phaseolus vulgaris. Molecular General Genetics 210: 219–233

Sadowski, J., P. Gaubier, M. Delseny, and C. F. Quiros. 1996 Genetic and physical mapping in Brassica diploid species of a gene cluster defined in Arabidopsis thaliana. Molecular General Genetics 251: 298– 306

Sang, T., D. J. Crawford, and T. F. Stuessy. 1997 Chloroplast DNA phylogeny, reticulate evolution, and biogeography of Paeonia (Paeoniaceae). American Journal of Botany 84: 1120–1136[CrossRef]

———, M. J. Donoghue, and D. Zhang. 1997 Evolution of alcohol dehydrogenase genes in peonies (Paeonia): phylogenetic relationships of putative nonhybrid species. Molecular Biology and Evolution 4: 994– 1007

Schnabel, A., and J. F. Wendel. 1998 Cladistic biogeography of Gleditsia (Leguminosae) based on ndhF and rpl16 chloroplast gene sequences. American Journal of Botany 85: 1753–1765[Abstract/Free Full Text]

Schmidt, R., A. Acarkan, M. Koch, and M. Roßberg. 1999 A strategy for comparative physical mapping. In L.W.D. van Raamsdonk and J.C.M. den Nijs [eds.], Proceedings of the VIIth Symposium of the International Organization of Plant Biosystematics, Hugo de Vries Laboratory, Amsterdam, The Netherlands

Schulz, O. E. 1936 Cruciferae. In A. Engler and K. Prantl [eds.], Die natürlichen Pflanzenfamilien, 17b, 227–658. Engelmann, Leipzig, Germany

Small, R. L., J. A. Ryburn, R. C. Cronn, T. Seelanen, and J. F. Wendel. 1998 The tortoise and the hare: choosing between noncoding plastome and nuclear Adh sequences for phylogenetic reconstructions in a recently diverged plant group. American Journal of Botany 85: 1301–1315[Abstract/Free Full Text]

Soltis, D. E., R. K. Kuzoff, E. Conti, R. Gornall, and K. Ferguson. 1996 matK and rbcL sequence data indicate that Saxifraga (Saxifragaceae) is polyphyletic. American Journal of Botany 83: 371–382[CrossRef][ISI]

———, D. R. Morgan, A. Grable, P. S. Soltis, and R. Kuzoff. 1993 Molecular systematics of Saxifragaceae sensu stricto. American Journal of Botany 80: 1056–1081[CrossRef][ISI]

———, et al. 1997 Angiosperm phylogeny inferred from 18S ribosomal DNA sequences. Annals of the Missouri Botanical Garden 84: 1–49

Steele, K. P., and R. Vilgalys. 1994 Phylogenetic analysis of Polemoniaceae using nucleotide sequences of the plastid gene matK. Systematic Botany 19: 126–142

Swofford, D. L. 1993 PAUP: phylogenetic analysis using parsimony. Version 3.0. University of Illinois Natural History Survey, Champaign, Illinois, USA

———. 1999 PAUP*: phylogenetic analysis using parsimony (*and other methods). Version 4.0b2. Sinauer, Sunderland, Massachusetts, USA

Warwick, S. L., and L. D. Black. 1997 Phylogenetic implications of chloroplast DNA restriction site variation in subtribe Raphaninae and Cakilinae (Brassicaceae, tribe Brassiceae). Canadian Journal of Botany 75: 960–973

Wingender, R., H. Röhrig, C. Höricke, D. Wing, and J. Schell. 1989 Differential regulation of soybean chalcone synthase genes in plant defense, symbiosis and upon environmental stimuli. Molecular General Genetics 218: 315–322

Wolfe, K. H., P. M. Sharp, and W.-H. Li. 1989 Rates of synonymous substitutions in plant nuclear genes. Journal of Molecular Evolution 29: 208–211[CrossRef]

Xiang, Q.-Y., D. E. Soltis, and P. S. Soltis. 1998 Phylogenetic relationships of Cornaceae and close relatives inferred from matK and rbcL sequences. American Journal of Botany 85: 285–297[Abstract]

Yang, Y. W., K. N. Lai, P. Y. Tai, and W. H. Li. 1999 Rates of nucleotide substitution in angiosperm mitochondrial DNA sequences and dates of divergence between Brassica and other angiosperm lineages. Journal of Molecular Evolution 48: 597–604[CrossRef][ISI][Medline]

Zunk, K., K. Mummenhoff, and H. Hurka. 1993 Chloropast DNA restriction site variation in the Brassicaceae. Plant Molecular Evolution Newsletter 3: 40–44

———, ———, and ———. 1999 Phylogenetic relationships in tribe Lepidieae (Brassicaceae) based on chloroplast DNA restriction site variation. Canadian Journal of Botany, 77: 1504–1512