| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | SEARCH RESULT |
|
|
|
|||||||
|
|||||||||
What's this? |
Systematics |
2Université Jean Monnet, Faculté des Sciences et Techniques, Laboratoire de Biotechnologies Végétales Appliquées aux Plantes Aromatiques et Médicinales (LBVPAM) EA 3061, 23 rue du Docteur Paul Michelon, F-42023 Saint-Etienne Cedex 02, France; 3Université Joseph Fourier, Laboratoire de Biologie des Populations d'Altitude (LBPA), UMR CNRS 5553, BP 53, F-38041 Grenoble Cedex 09, France
Received for publication March 19, 2002. Accepted for publication July 12, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: AFLP • DNA fingerprinting • genetic diversity • hybridization • Lamiaceae • Mentha • polyploid
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
), cytological (Ruttle, 1931
; Heimans, 1938
; Morton, 1956
; Sharma and Bhattacharyya, 1959
; Harley, 1967
, 1972
; Harley and Brighton, 1977
; Singh and Sharma, 1986
), and chemical (Lawrence, 1978
) markers. Harley and Brighton (1977)
published a critical review of the chromosome numbers in relation to the taxonomy of the genus. They estimated that there are probably only 25 species and rather fewer hybrids. They recognized five sections (Audibertia, Eriodontes, Pulegium, Preslia, Mentha) on the basis of basic chromosome numbers and morphological features. There is no problem of identification for the first four sections because no example of natural interspecific hybridization exists. The fifth section, Mentha, includes five species: M. suaveolens Ehrh., M. longifolia (L.) Hudson, M. spicata L., M. arvensis L., and M. aquatica L. Chromosome counts for this section suggest a basic number of x = 12 with a range of numbers from diploid to octoploid. Plants of this section have a vigorous rhizome system as a means of spreading and dispersal and are self-compatible, and a high level of outbreeding is assured due to gynodioecy.
Natural interspecific hybridization occurs with high frequency in section Mentha, both in wild populations and in cultivation (Fig. 1). Most hybrids are sterile or subfertile, but vegetative propagation enables them to persist. Complex hybrid populations may arise, and if they are subfertile, may cross with parental or nonparental species. This leads to a large diversity of chromosomes numbers (24–120), and much of the taxonomy of section Mentha has been complicated by hybridization, by a high morphological polymorphism, as well as polyploidy and vegetative propagation. The best known hybrids are M. x piperita (peppermint) and M. spicata L. (native spearmint), which are intensively cultivated for their essential oils. Mentha x piperita results from a cross between M. aquatica and M. spicata; M. spicata is the hybrid between M. suaveolens and M. longifolia (Harley and Brighton, 1977
; Fig. 1). The great variability of M. spicata led several workers to establish a subdivision of this hybrid, and two subgroups were described based upon two features. Cytological studies (Ruttle, 1931
; Morton, 1956
) led to the conclusion that two M. spicata cytotypes exist, with 2n = 36 and 2n = 48 chromosomes, respectively. According to the cytotype implied in the cross with M. aquatica, two M. x piperita cytotypes result, with 2n = 66 or 2n = 72 chromosomes, respectively (Fig. 1). Moreover, morphological and chemical data divide M. spicata into two different subgroups according to the presence or absence of nonsecreting trichomes and the essential oil composition. Wild M. spicata is nearly always hairy, like its diploid parents, and can contain other terpenes that are found commonly in its diploid progenitors. Selected by man as an aromatic plant, M. spicata became glabrous with a characteristic odor due to carvone and menthone as the prevailing terpenes. Mentha spicata plants, introduced and distributed throughout the world, are often found as garden escapes. According to Lebeau (1974)
, it was essential to distinguish two M. spicata subspecies, M. spicata subsp. spicata and M. spicata subsp. glabrata, with and without nonsecreting trichomes, respectively, for the following reasons: (1) the presence or absence of nonsecreting trichomes led to a different aspect, (2) wide difference in perfume, and (3) difference in habitats. Mentha spicata propagates almost entirely by vegetative means. Lebeau (1974)
and Harley and Brighton (1977)
described some individuals of M. spicata that were close in appearance to its progenitor diploid species. They noted that M. spicata segregates parental characters in its progeny by selfing, which was impossible to distinguish from the hybrids it often forms with either M. suaveolens or M. longifolia. In some cases, such hairy M. spicata plants were confused with M. longifolia.
|
) and cultivar identity (Fenwick and Ward, 2001
) based on RAPD markers have been undertaken in Mentha species. The complex systematics of section Mentha led us to choose amplified fragment length polymorphism (AFLP), based on the selective polymerase chain reaction (PCR) amplification of restriction fragments from a total digest of genomic DNA (Vos et al., 1995
), to assess genetic diversity. It combines the specificity of whole-genome restriction fragment analysis and the selectivity of high stringency PCR amplification without prior knowledge of primer target sequences. A large number of markers can be obtained quickly compared to restriction fragment length polymorphism (RFLP), randomly amplified polymorphic DNA (RAPD), or simple sequence repeat (SSR) (Powell et al., 1996
; Pejic et al., 1998
; Garcia-Mas et al., 2000
). Because of its high repeatability and resolution, the AFLP technique has been used successively to assess genetic diversity, for example, in Brassica juncea (Srivastava et al., 2001
) and to resolve phenetic relationships in Dactylorhiza (Orchidaceae) (Hedrén, Fay, and Chase, 2001
) and in Lactuca sensu lato (s.l.) (Lactuceae, Asteraceae) (Koopman, Zevenbergen, and Van den Berg 2001
). The goal of the present study was to genotype diploid and polyploid species of section Mentha by AFLP to (1) assess the relationships among accessions of an extended collection representing legitimate species and the hybrids M. x piperita and M. spicata; and (2) check the existence of two M. spicata groups with molecular markers.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Flow cytometry measurement
Nuclei were prepared from leaf tissue and stained according to Galbraith et al. (1983)
and Marie and Brown (1993)
. About 4 cm2 of Mentha leaves and 2 cm2 of Petunia hybrida or Lycopersicon esculentum used as reference were chopped using a razor blade in 1 mL of modified Galbraith buffer (1% Triton X-100, 10 mmol/L metabisulfite, and 1% polyvinylpyrrolidone [PVP] 10 000). After filtration through a 30-µm nylon mesh, 10 mg/mL DNase-free RNase A (Boehringer, Paris, France) and 50 µg/mL propidium iodide (Sigma, Saint-Quentin Fallavier, France) were added. After an incubation for 20 min at room temperature, the fluorescence intensity was measured by flow cytometry (Elite cytometer, Beckman-Coulter, Roissy, France) at 488 nm. For each sample two nuclei preparations were analyzed, and for M. x piperita 38, eight repetitions were performed, randomly among all the other individuals. The DNA nuclear content was calculated by the ratio of the 2C peak positions, Mentha/reference, on the intensity fluorescence histograms. Lycopersicon esculentum or Petunia hybrida were chosen as references according to (1) the chromosome number of each Mentha species and (2) after pre-examination of one individual per species. Their 2C values were, respectively, 1.99 and 2.85 picograms (pg).
AFLP procedure
DNA extraction was performed using the DNeasy Plant Mini Kit (Qiagen, Courtaboeuf, France) according to manufacturer's instructions using 20 mg of dried leaf material. DNA concentration was assessed by fluorimetry with the PicoGreen double strand DNA quantification Kit (Molecular Probes, Eugene, Oregon, USA). Genomic DNA was digested with EcoRI and MseI, and the restricted fragments were ligated with EcoRI and MseI adaptators (Vos et al., 1995
) in the same reaction. Briefly, 5.5 µL (100 ng) extracted DNA was added to a 5.5 µL digestion-ligation solution containing T4 Buffer 1x, 50 mmol/L NaCl, 50 µg/mL bovine serum albumine, 1 unit (U) MseI, 5 U EcoRI, 1 U T4-ligase, 0.9 µmol/L MseI adapter, and 0.9 µmol/L EcoRI adapter. This resulting reaction mixture was then incubated at 37°C for 2 h and subsequently diluted tenfold in purified water. Preselective PCR by primers having one selective nucleotide each, namely E + 1 and M + 1 (Vos et al., 1995
), was performed using 3 µL of digested-ligated DNA added to a 22-µL mixture containing Taq Buffer 1x, 150 µmol/L MgCl2, 16 µmol/L of each dNTP, 0.2 µmol/L EcoRI primer (E/A), 0.2 µmol/L MseI primer (M/C or M/G), and 0.5 U AmpliTaq Polymerase (Applied Biosystems Perkin Elmer, Courtaboeuf, France). Preamplification was performed using the following temperature profile: 2 min at 72°C, 25 cycles of 30 s at 94°C, 30 s at 56°C, and 2 min at 72°C, followed by one cycle for 10 min at 72°C. The preamplified DNA was then diluted in a ratio of 1 : 20 in purified water, and 5 µL of the diluted mixture were used as template for selective amplification and added to a 20-µL mixture containing Taq Buffer 1x, 250 µmol/L MgCl2, 16 µmol/L of each dNTP, 0.04 µmol/L fluorescent EcoRI primer (E/AGA, E/AGT, E/ATC, E/ATG), 0.2 µmol/L MseI primer (M/CAA, M/CAC, M/CAG, M/CAT, M/CTA, M/CTC, M/CTG, M/CTT or M/GAC, M/GTC) (Vos et al., 1995
) and 1 U AmpliTaqGold Polymerase (Applied Biosystems Perkin Elmer). The preamplified DNA was amplified using one cycle of 10 min at 95°C, 36 cycles of 30 s at 94°C, 1 min at 65°C, and 1 min at 72°C, followed by one cycle for 10 min at 72°C. The annealing temperature was reduced every cycle by 0.7°C, and after 12 cycles it reached the annealing temperature of 56°C. This temperature was maintained for the subsequent 23 cycles. For gel electrophoresis, the amplified product (4 µL) was mixed with a 3.5-µL mixture containing formamide, loading buffer, and internal lane standard (GeneScan 500 Rox, Applied Biosystems Perkin Elmer), evaporated to a final volume of 3 µL and denaturated at 95°C for 2 min prior to loading on 64-lane 5% Long Ranger polyacrylamide gels on an automated DNA sequencer (ABI 377, Applied Biosystems Perkin Elmer).
Primer pair screening
Sixteen primer pairs were tested on four individuals in the selective PCR. Thirteen combinations gave too many bands for reliable analysis. A total of three AFLP primer combinations (E-AGT/M-CAG, E-AGT/M-GTC, and E-ATC/M-GTC) were chosen according to the number of fragments amplified, the level of polymorphism exhibited among closely related accessions, and the quality of the peaks produced on electrophenograms. Reliability of the technique was checked as following; two independent DNA extractions were performed for two accessions that were submitted to the whole AFLP procedure.
Phenetic analysis
AFLP electrophenograms were analyzed with GeneScan Analysis 3.1 (Applied Biosystems Perkin Elmer). All peaks that could be unambiguously read on the electrophenograms were treated as individual dominant loci and scored as either present (1) or absent (0) across all 62 accessions of Mentha for each primer-pair combination. Genetic similarities, based on Nei and Li's formula (1979)
as GSXY = 2NXY(NX + NY), where NXY is the number of peaks shared in accessions X and Y, NX is the number of peaks in accession X, and NY is the number of peaks in accession Y, were calculated using the genetic distance estimation program of the TREECON software package (Van de Peer and De Wachter, 1994
). We chose Nei and Li's formula to calculate distances between accessions because this method does not take into account shared absent fragments (0–0) and so reduces the amount of homoplasy. The genetic distance was computed as GDXY = 1 – GSXY, and dendrograms were generated by using the unweighted pair group mean average (UPGMA) and neighbor-joining methods in TREECON. The robustness of UPGMA and neighbor-joining trees was evaluated by bootstrapping (1000 bootstrap replicates) using TREECON.
Parsimony analysis
Determination of phylogenetic signal in the data set and cladistic analysis were performed using PAUP version 4.0b8 (Swofford, 1999
). Phylogenetic signal was determined from the tree-length distribution of 100 000 trees, using the g1 statistic (Hillis and Huelsenbeck, 1992
). The Mentha data set contained >25 taxa and >50 variable characters, and therefore the critical value of –0.09 was used. Data sets that produce g1 values less than –0.09 are significantly more structured than are the random data (Hillis and Huelsenbeck, 1992
). An heuristic search comprised 1000 random-addition sequences, and tree bisection-reconnection (TBR) branch swapping was performed. Another search was run using successive weighting cycles. Characters were reweighted by the maximum value of the rescaled consistency index, and the search was conducted with 1000 random-addition sequences, followed by TBR branch swapping. Bootstrap values for the resulting tree were calculated using 1000 replicates.
Principal coordinates analysis
Principal coordinates analysis (PCoA) was performed to visualize interspecies relationships on the genetic distance (Nei and Li, 1979
) matrix using the Ade-4 software (Thioulouse et al., 1997
).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
The AFLP polymorphism
Each primer combination generated an average of 40 bands per individual, ranging in size from 50 to 500 base pairs (bp). The percentage of polymorphic peaks for each primer combination varied from 50% to 60% (Table 1). Polymorphic peaks and individual-specific peaks were obtained among 62 accessions (Table 2). To analyze the genetic diversity of 62 Mentha accessions, we examined 67 polymorphic AFLP markers generated from three primer combinations.
|
|
Cluster analysis
The pairwise genetic distances of the accessions were calculated to cluster the data among the Mentha accessions using the unweighted pair-group method with an arithmetic average (UPGMA) algorithm and neighbor-joining method. The phenogram in Fig. 2a divided accessions into two major clusters: (I) M. suaveolens, M. spicata, and M. longifolia and (II) M. arvensis, M. x gracilis, M. aquatica, and M. x piperita. There were three distinct subgroups in cluster I: (1) M. suaveolens-M. spicata I, (2) M. spicata II, (3) M. longifolia and two subgroups in cluster II: (1) M. arvensis-M. x gracilis, (2) M. aquatica-M. x piperita (well supported with a 92% bootstrap value). The range of similarity values among M. spicata-M. suaveolens, M. spicata-M. longifolia, and M. suaveolens-M. longifolia ranged from 0.48 to 0.88, 0.27 to 0.71, and 0.11 to 0.38, respectively. The UPGMA analysis showed that M. arvensis accessions formed a sister group to the M. aquatica-M. x piperita cluster. Within the M. arvensis group, two subgroups are clearly defined: (1) M. arvensis 514, 515, var. piperascens 7C, and M. x gracilis (95% bootstrap value), and (2) M. arvensis 165, var. piperascens 101. The most well-known hybrid, M. x piperita (peppermint) showed a range of similarity coefficients, from 0.65 to 0.98, and was separated by UPGMA into two clusters. The first is composed of cultivated M. x piperita from France (M. x piperita 19 and 38) and Bulgaria (M. x piperita 110), while the second cluster is mainly composed of wild accessions from the United States. Mentha x piperita 93C from the Ukraine appeared as a sister group to the American one. Mentha x piperita showed the highest similarity with one of its diploid progenitors, M. aquatica (0.6–0.8), rather than with M. spicata (0.3–0.5). Mentha x piperita var. vulgaris 118 is clustered in the M. spicata II subgroup, while M. suaveolens 495 appeared as a sister group to the M. arvensis-M. aquatica-M. x piperita cluster. The neighbor-joining method (data not shown) led to the same results except M. arvensis var. piperascens 101 and 7C were clustered in M. spicata I.
|
Parsimony analysis
Two cladistic analyses were performed. The first was run with all accessions. The g1 statistic for this data set was –0.31, indicating significant phylogenetic signal. The parsimony analysis yielded 197 most-parsimonious trees of 282 steps (retention index [RI] = 0.80, consistency index [CI] = 0.23, rescaled consistency index [RC] = 0.19). The search with three successive weighting steps yielded 240 trees of 52 steps (RI = 0.91, CI = 0.45, RC = 0.41). As with the phenetic analysis, (1) the same two major groups (I and II) could be described (Fig. 2b), and (2) M. x piperita var. vulgaris 118 was clustered in M. spicata II subgroup, while M. suaveolens 495 appeared as a sister group to the M. arvensis-M. aquatica-M. x piperita cluster. However, three accessions, M. longifolia 18, M. longifolia 533, and M. asiatica, were found in cluster II while these accessions were found in cluster I in the phenetic analysis. With regard to the two M. spicata subgroups, M. spicata 32 was clustered in subgroup II, while this accession was clustered in subgroup I in the phenetic analysis. Mentha species were assigned to one cluster with the following bootstrap values: M. arvensis, 75%; M. aquatica-M. x piperita, 93%; and M. suaveolens-M. spicata, 69%. The second analysis was performed without misidentified accessions, hypothetical hybrids, and backcrosses (data not shown). The g1 statistic for this second data set was –0.33, indicating significant phylogenetic signal also. Parsimony analysis yielded 251 most-parsimonious trees with 207 steps (RI = 0.85, CI = 0.32, RC = 0.27). The search with three successive weighting steps yielded ten trees of 53 steps (RI = 0.94, CI = 0.58, RC = 0.55). With regard to this second cladistic analysis, we obtained the same results as those noted in the phenetic analysis. Mentha species were assigned to one cluster with high bootstrap values: M. arvensis, 100%; M. aquatica-M. x piperita, 99%; and M. suaveolens-M. spicata, 99%.
Principal coordinates analysis
Nei and Li's genetic similarity coefficients were used for principal coordinates analysis. The variance of the first two principal coordinates accounted for 55% of the total variation (Fig. 3). The first axis, accounting for 38% of the variation, separated the M. arvensis-M. x piperita-M. aquatica group from the M. suaveolens-M. spicata-M. longifolia group. The second axis accounted for 17% of the variation and distinguished distinct groups for each species. The PCoA results are in accordance with those of the phenetic and cladistic analyses. The M. asiatica and M. x gracilis accessions were clustered with the M. longifolia and M. arvensis groups, respectively. Mentha x piperita 118 and M. suaveolens 495 were tightly clustered with the M. spicata II and M. arvensis-M. aquatica-M. x piperita groups, respectively. The PCoA (Fig. 3) showed M. arvensis accessions dispersed except M. arvensis 514 and M. arvensis 515, which were tightly clustered. Mentha arvensis var. piperascens 101 and 7C occupied an intermediate position between the two major groups, M. aquatica-M. x piperita and M. suaveolens-M. spicata-M. longifolia. Mentha arvensis 165 was clustered with the M. x piperita group.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
by RAPD fingerprinting. Genetic diversity and genetic relationships were assessed among six Mentha taxa (M. arvensis, M. spicata, M. spicata cv. viridis, M. x piperita, M. x piperita cv. citrata, M. x gracilis Sole cv. cardiaca). Fenwick and Ward (2001)
were the first to evaluate genetic diversity within the most widely grown peppermint (M. x piperita) and Scotch spearmint (M. x gracilis) cultivars by random amplified polymorphic DNA markers. Their results indicate a high degree of genetic uniformity in mint cultivars grown for oil in the United States. Effectively, new cultivars are selected as either spontaneous or induced variants or "sports" of existing cultivars and then clonally propagated. The lack of genetic diversity could result, for example, in a poor disease resistance as demonstrated by the widespread occurrence of verticillium wilt (caused by Verticillium dahliae Kleb.) (Fenwick and Ward, 2001
). The AFLP method allowed us to assess genetic diversity between and within Mentha species. This assessment is fundamental because genetic diversity could be in future exploited through molecular approaches or plant breeding techniques to improve mint cultivars for disease resistance or to increase essential oil yield, for example. The major objectives of this study were to assess the phenetic relationships between species and hybrids of section Mentha and to have a better knowledge of the two M. spicata subgroups through use of molecular markers. For our purposes, the AFLP data will be discussed in the context of the classification described by Harley and Brighton (1977)
.
Genetic diversity in Mentha
Cluster I: M. suaveolens-M. longifolia-M. spicata
Based on inflorescence morphology, Linnaeus grouped these three species in the Spicatae. Mentha longifolia, a mountain plant, shows a great array of morphological diversity and has consequently been treated by some taxonomists as composed of numerous species, subspecies, varieties, and forms (Harley and Brighton, 1977
). For example, among our accessions, M. asiatica is synonymous with M. longifolia and is not a Mentha species. In our collection, eight of the M. longifolia accessions originate from Europe, one from Nepal (M. longifolia subsp. hymalaiensis 635), and the last from Syria (M. longifolia subsp. typhoides 20). Plants from Afghanistan eastwards with petiolate leaves have been referred to M. longifolia subsp. hymalaiensis, but according to Harley and Brighton (1977)
most of those plants show the character to a much lesser degree and are closer to "typical" M. longifolia. Our results confirm this assessment because M. longifolia subsp. hymalaiensis 635 has a range of genetic similarity coefficient between 0.66 and 0.81 compared to typical M. longifolia accessions. With regard to M. longifolia subsp. typhoides 20 (0.60–0.72), it seems that the situation is identical. Clustering of M. suaveolens and M. longifolia diploid accessions into strictly separate group was apparent, while the tetraploid, M. spicata, is separated into two distinct groups, one a sister group to M. suaveolens accessions and the second forming a distinct cluster. According to cladistic, and phenetic analyses, our results, based on molecular markers, showed that M. suaveolens, M. longifolia, and their hybrid M. spicata are closely related. The PCoA clearly separated the two progenitors along axis 2 and hybrids are in between.
Cluster II: M. x piperita-M. aquatica-M. arvensis
This cluster is composed of species with larger genomes (hexaploid and octoploid) than the first one. Mentha x piperita, M. aquatica, and M. arvensis belong to the Spicatae, Capitae, and Verticillatae groups, respectively. With regard to the dichotomy noted in phenogram and cladogram, it seems that the section Mentha evolved into two lineages: (1) diploid species and (2) polyploid species. Progenitors of M. arvensis and M. aquatica have never been described. According to our results, two hypotheses could be assumed: (1) progenitors have evolved or disappeared or (2) polyploids have been subjected to hybridization and/or to successive polyploidization events and then progenitors have subsequently become difficult to find. According to Ruttle (1931)
, M. aquatica has an allopolyploid origin because 48 chromosome pairs were counted at meiotic mitosis in microsporocytes. Mentha arvensis, cultivated for its essential oils, shows a great natural diversity also revealed in the sampling used in this study. Two cytotypes exist, with 2n = 72 and 2n = 96 chromosomes, respectively. Moreover, Mentha arvensis could hybridize with all species of the section. For example, the cross between M. arvensis (2n = 72) and M. spicata (2n = 48) led to M. x gracilis L. According to the chromosome number counts of Harley and Brighton (1977)
, M. x gracilis is a pentaploid (2n = 60). Thus, M. arvensis contributes three of the five genomes of the hybrid, and it seems highly probable that M. x gracilis resembles M. arvensis, explaining why it is closely related to M. arvensis in the UPGMA, parsimony, and PCoA. Schürhoff (1929, in Ruttle [1931
]) described M. arvensis var. piperascens as a sterile hybrid between M. aquatica and M. arvensis. Harley and Brighton (1977)
noted it as an octoploid (2n = 96). According to our flow cytometry results, it seems that M. arvensis var. piperascens 101 and 7C are hexaploid (2n = 72). These accessions showed the highest similarity coefficients with other M. arvensis and M. spicata accessions while the similarity with M. aquatica is lower. According to UPGMA, parsimony, PCoA and flow cytometry analyses, it seems highly probable that M. arvensis var. piperascens 101 and 7C are the result of an hybridization event involving M. arvensis (2n = 96) and M. spicata (2n = 48) species (Fig. 1). This hypothesis is supported by neighbor-joining results, which clustered M. arvensis var. piperascens 101 and 7C in M. spicata I. Mentha arvensis 165 appeared as an hexaploid accession (2n = 84). According to the National Clonal Germplasm Repository, this accession gives almost no selfed seeds and is probably a male sterile due to the loss of a chromosome, but no count was done. In the PCoA, this accession is clustered in the M. aquatica-M. x piperita group, and the highest similarity coefficients were noted with M. arvensis, M. x piperita, and M. aquatica species. It seems that a hybridization event may have occurred between M. arvensis and M. x piperita or between M. arvensis and M. aquatica. The second hypothesis seems to be more probable than the first one because M. x piperita is a sterile hybrid. Similarity between M. arvensis 165 and M. aquatica and M. x piperita species could have resulted from common AFLP markers found in M. aquatica and its hybrid M. x piperita. In that case, M. arvensis 165 may be revisited as a M. arvensis var. piperascens. Finally, M. arvensis 514 and 515 appeared as tetraploid, while "typical" M. arvensis are hexaploid (2n = 72) or octoploid (2n = 96). These accessions are closely related with M. x gracilis in all our analyses. A great number of varieties have been described in M. arvensis, and M. x gracilis, one of its hybrids, shows a high variability. Thus, these species are difficult to identify leading to taxonomic complexity.
Close relationships between M. x piperita and M. aquatica
Mentha x piperita, which is intensively cultivated all over the world, results from a cross between M. spicata and M. aquatica. Actually, the latter is octoploid, while M. spicata is a tetraploid. Therefore, two-thirds of the M. x piperita genetic pool is composed of the M. aquatica genome. The PCoA showed that M. x piperita accessions are between its progenitors and as with UPGMA and parsimony analyses, M. x piperita is closer to M. aquatica than M. spicata. These results, based on molecular markers, confirm the two progenitors of M. x piperita, described on the basis of morphological and chemical features.
Genealogy of M. spicata: hybrid status and backcrosses
DNA content
Flow cytometry measurements are relatively consistent to previous cytological data. According to the basic chromosome number x = 12, it seems evident that the two M. spicata groups are characterized by 2n = 36 and 2n = 48 chromosomes, respectively. According to DNA content measurement by flow cytometry and the presence or absence of trichomes (data not shown), it seems that there is no relationship between these criteria and the two distincts groups of M. spicata resolved by UPGMA and principal coordinates analyses.
Thus it seems that the two M. spicata groups should not be defined based upon chromosome number or the presence/absence of nonsecreting trichomes but rather on the basis of molecular markers that cover the entire genome, such as AFLP markers. So, it seems evident that the M. spicata subdivision into two different subspecies, M. spicata subsp. spicata and M. spicata subsp. glabrata, by Lebeau (1974)
should be revisited. Two hypotheses could explain the presence of two M. spicata groups: (1) M. spicata often forms backcrosses with its diploid progenitors, indeed, hybrids like M. x villosa Hudson (M. spicata x M. suaveolens) and M. x villoso-nervata Opiz (M. spicata x M. longifolia) are difficult to distinguish from "typical" M. spicata and (2) M. spicata occasionally segregates parental characters in its progeny on selfing (Harley, 1967
). As a consequence, M. spicata plants are more and more closely related to M. suaveolens or M. longifolia during backcrossing or selfing. Harley found one specimen of M. x villosa in the southwest of England and said that it was remarkable for its very close similarity to M. suaveolens. It seems that AFLP methods allow us to distinguish clearly M. spicata from its progenitors and M. spicata accessions that are closer to M. suaveolens or to M. longifolia as well.
Detection of misidentified accessions
Mentha suaveolens 495 and M. x piperita 118 were not found in their respective species clade in the UPGMA and principal coordinates analyses. Mentha suaveolens 495 is a seedling selection from another M. suaveolens accession that is not in our collection and M. x piperita var. vulgaris 118 comes from England. Morphologically, these accessions do not look like M. suaveolens or M. x piperita. Cytologically, M. suaveolens 495 and M. x piperita 118 have 3.87 and 1.31 pg of DNA, respectively, while other accessions of M. suaveolens and M. x piperita have a range of DNA from 0.9 to 1 pg and from 2.3 to 3.1 pg of DNA, respectively. It seems probable that M. suaveolens 495 was misidentified and then mislabelled before sampling. This accession came from breeding experiments and particularly from open pollination. It became apparent that identification of M. x piperita 118 as an accession of M. x piperita was in error. This plant is probably the result of hybridization, and AFLP analysis provided some information about the origin of its genetic pool.
In conclusion, the present study of AFLP analysis of mint accessions supports current taxonomic classification. Our data set was analyzed phenetically as well as cladistically, and the main well-supported clusters of the trees were comparable for both types of analyses. Phenetic analysis using UPGMA based on Nei and Li's (1979)
genetic distance resulted in a highly resolved tree, while parsimony analysis showed phylogenetic signal and high bootstrap values. Finally, PCoA allowed us to easily visualize hybridization events. Actually, M. suaveolens, M. spicata, and M. longifolia species form a tight group and on the basis of AFLP markers, M. x piperita is closer to M. aquatica than M. spicata. The existence of polyploidy in section Mentha does not seem to represent a problem in AFLP analysis. The three primer combinations provided enough AFLP polymorphisms to resolve genetic diversity between and within species, allowing us to objectively identify species, hybrids, and accessions. This genetic variability could in future be exploited through molecular approaches for gene introgression in breeding programs to produce desired genotypes. While the number of mint accessions used for this study represent only a small sample of the available mint germplasm, the potential resolving power of AFLP analysis in a larger collection seems evident.
|
FOOTNOTES |
|---|
4 Author for reprint requests (Moja{at}univ-st-etienne.fr
)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Galbraith D. W. K. R. Harkins J. M. Maddox N. M. Ayres D. P. Sharma E. Firoozabady 1983 Rapid flow cytometry analysis of the cell cycle in intact plant tissues. Science 220: 1049-1051
Garcia-Mas J. M. Oliver H. Gómez-Paniagua M. C. de Vicente 2000 Comparing AFLP, RAPD and RFLP markers for measuring genetic diversity in melon. Theoretical Applied Genetics 101: 860-864[CrossRef]
Harley R. M. 1967 The spicate mints. Proceedings of the Botanical Society of the British Isles 6: 369-372
Harley R. M. 1972 Notes on the genus Mentha (Labiatae). Botanical Journal of the Linnean Society 65: 250-253
Harley R. M. C. A. Brighton 1977 Chromosome numbers in the genus Mentha L. Botanical Journal of the Linnean Society 74: 71-96
Hedrén M. M. F. Fay M. W. Chase 2001 Amplified fragment length polymorphisms (AFLP) reveal details of polyploid evolution in Dactylorhiza (Orchidaceae). American Journal of Botany 88: 1868-1880
Heimans J. 1938 Chromosomes in the genus Mentha. Chronica Botanica 4: 389-390
Hillis D. M. J. P. Huelsenbeck 1992 Signal, noise and reliability in molecular phylogenetic analyses. Journal of Heredity 83: 189-195
Khanuja S. P. S. A. K. Shasany A. Srivastava S. Kumar 2000 Assessment of genetic relationships in Mentha species. Euphytica 111: 121-125[CrossRef][Web of Science]
Koopman W. J. M. M. J. Zevenbergen R. G. Van den Berg 2001 Species relationships in Lactuca s. l. (Lactuceae, Asteraceae) inferred from AFLP fingerprints. American Journal of Botany 88: 1881-1887
Lawrence B. M. 1978 A study of the monoterpene interrelationships in the genus Mentha with special reference to the origin of pulegone and menthofuran. Ph.D. dissertation. Groningen University, Groningen, Netherlands
Linnaeus. 1767 Systema Naturae. Stockholm. Tomus II: 391-392
Lebeau J. 1974 Nouvelles mises au point dans le genre Mentha. Natura Mosana 27: 109-141
Malinvaud E. 1880 Simple aperçu des hybrides dans le genre Mentha. Bulletin de la Société Botanique de France 27: 332-347
Marie D. S. C. Brown 1993 A cytometric exercise in plant DNA histograms, with 2C values for seventy species. Biology of the Cell 78: 41-51[CrossRef][Web of Science][Medline]
Morton J. K. 1956 The chromosome numbers of the British Menthae. Watsonia 3: 244-252
Nei M. W. H. Li 1979 Mathematical model for studying genetic variation in terms of restriction endonucleases. Proceedings of the National Academy of Sciences, USA 76: 5269-5273
Pejic I. P. Ajmone-Marsan M. Morgante V. Kozumplick P. Castiglioni G. Taramino M. Motto 1998 Comparative analysis of genetic similarity among maize inbred lines detected by RFLPs, RAPDs, SSRs, and AFLPs. Theoretical Applied Genetics 97: 1248-1255[CrossRef]
Powell W. M. Morgante C. Andre M. Hanafey J. Vogel F. Tingey A. Rafalski 1996 The comparison of RFLP, RAPD, AFLP and SSR (microsatellite) markers for germplasm analysis. Molecular Breeding 2: 225-238
Ruttle M. L. 1931 Cytological and embryological studies of the genus Mentha. Gartenbauwissenschaft 44: 428-468
Sharma A. K. N. K. Bhattacharyya 1959 Cytological studies on different species of Mentha with special reference to the occurrence of chromosomal biotypes. Cytologia 24: 198-212
Singh T. P. A. K. Sharma 1986 Mentha—taxonomic status as interpreted through cytology, genetics and phytochemistry. Indian Journal of Genetics 46: (Supplement) 198-208
Srivastava A. V. Gupta D. Pental A. K. Pradhan 2001 AFLP-based genetic diversity assessment amongst agronomically important natural and some newly synthesized lines of Brassica juncea. Theoretical Applied Genetics 102: 193-199[CrossRef]
Swofford D. L. 1999 PAUP*: phylogenetic analysis using parsimony (*and other methods), version 4b8. Sinauer, Sunderland, Massachusetts, USA
Thioulouse J. D. Chessel S. Dolédec J. M. Olivier 1997 ADE-4: a multivariate analysis and graphical display software. Statistics and Computing 7: 75-83[CrossRef][Web of Science]
Van de Peer Y. R. De Wachter 1994 TREECON for Windows: a software package for the construction and drawing of evolutionary trees for the Microsoft Windows environment. Computer Application in the Biosciences 10: 569-570
Vos P. R. Hogers M. Bleeker M. Reijans T. van de Lee M. Hornes A. Frijters J. Pot J. Peleman M. Kuiper M. Zabeau 1995 AFLP: a new technique for DNA fingerprinting. Nucleic Acids Research 23: 4407-4414
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Lihova, J. Kucera, M. Perny, and K. Marhold Hybridization between Two Polyploid Cardamine (Brassicaceae) Species in North-western Spain: Discordance Between Morphological and Genetic Variation Patterns Ann. Bot., June 1, 2007; 99(6): 1083 - 1096. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. K. Shasany, M. P. Darokar, S. Dhawan, A. K. Gupta, S. Gupta, A. K. Shukla, N. K. Patra, and S. P. S. Khanuja Use of RAPD and AFLP Markers to Identify Inter- and Intraspecific Hybrids of Mentha J. Hered., September 1, 2005; 96(5): 542 - 549. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | SEARCH RESULT |
