AOBPreview originally published online on July 25, 2006
Annals of Botany 2006 98(4):857-868; doi:10.1093/aob/mcl167
Karyotype Variation, Evolution and Phylogeny in Borago (Boraginaceae), with Emphasis on Subgenus Buglossites in the Corso-Sardinian System
Dipartimento di Biologia Vegetale dell'Università, sez. Botanica Sistematica, Via G. La Pira 4, I-50121 Firenze, Italy
* For correspondence. E-mail selvi{at}unifi.it
Received: 1 March 2006 Returned for revision: 11 April 2006 Accepted: 12 June 2006 Published electronically: 25 July 2006
 |
ABSTRACT |
|---|
 TOPFOOTNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
• Background and Aims Karyological variation in the Mediterranean genus Borago and cytogeography of subgenus Buglossites in Corsica, Sardinia and the Tuscan Archipelago were investigated in combination with a molecular phylogenetic analysis aimed at elucidating relationships between subgenera and taxa with different chromosome features.
• Methods Karyotype analysis was performed on population samples of B. pygmaea, B. morisiana, B. trabutii and B. officinalis. Phylogenetic analyses were based on ITS1 nrDNA and matK cpDNA sequences.
• Key Results Four base numbers were found, x = 6, 8, 9 and 15, and three ploidy levels based on x = 8. In subgenus Buglossites the Sardinian endemic B. morisiana is diploid with 2n = 18, while B. pygmaea includes three allopatric cytotypes with 2n = 30 (Sardinia), 2n = 32 (southern Corsica) and 2n = 48 (central northern Corsica and Capraia). In subgenus Borago, the Moroccan endemic B. trabutii and the widespread B. officinalis have 2n = 12 and 2n = 16, respectively. Molecular data support the monophyly of Borago, while relationships in subgenus Borago remain unclear. Borago trabutii appears as the earliest divergent lineage and is sister to a clade with B. officinalis, B. morisiana and B. pygmaea. Subgenus Buglossites is also monophyletic, but no correspondence between ITS1 phylogeny and B. pygmaea cytotypes occurs.
• Conclusions Chromosome variation in Borago is wider than previously known. Two base numbers may represent the ancestral condition in this small genus, x = 6 or x = 8. An increase in chromosome number and karyotype asymmetry, a decrease in chromosome size and heterochromatin content, and the appearance of polyploidy are the most significant karyological changes associated with the divergence of the Buglossites clade. High ITS1 variation in the tetra- and hypotetraploid races of B. pygmaea suggests a multiple origin, while the lower polymorphism of the hexaploid race and its allopatric distribution in the northernmost part of the range is better explained with a single origin via union of unreduced and reduced gametes.
Key words: Boraginaceae, Borago, chromosomes, cytogeography, ITS1, karyotype, matK, molecular phylogeny, polyploidy
|
INTRODUCTION |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Borago L. is the ‘type’ genus of Boraginaceae tribe Boragineae Bercht. & J. Presl, a group of 16 genera native to the Old World and distributed in Asia, Africa and Europe (Hilger et al., 2005
). A recent phylogenetic analysis based on plastid and nuclear molecular markers suggested that Borago represents a monophyletic clade sister to Symphytum L., within a most basal taxon-group of the tribe (Hilger et al., 2004). In contrast to Symphytum, however, Borago is a small genus with only five species, four of which are restricted to the south-western Mediterranean basin in north-west Africa, Corsica, Sardinia and the Tuscan Archipelago. Only the common borage, B. officinalis L., is widespread beyond the boundaries of the Euro-Mediterranean region as a wild weed or cultivated as a garden plant, crop vegetable or pharmaceutical herb.
According to Gu
uleac (1928), Borago is split into two subgenera representing distinct evolutionary lineages, a continental one in North Africa and an insular one in Corsica and Sardinia. Subgenus Borago contains the type species B. officinalis plus B. trabutii Maire, endemic to the High and Anti-Atlas of Morocco, and B. longifolia Poir. endemic to northern Algeria and Tunisia. These three species share the erect habit and the showy cymes of blue flowers with rotate corollas and robust, exserted faucal scales and stamens, characters which represent adaptations for allogamic pollination by means of Hymenoptera. Borago officinalis is mainly an outbreeding species (Leach et al., 1990), although evidence of self-compatibility has also been established (Montaner et al., 2000). Subgenus Buglossites (Moris) Guul. is restricted to Corsica and Sardinia, as well as a single locality in the Tuscan Archipelago (Capraia Island). It comprises B. pygmaea (DC.) Chater & Greuter, endemic to Sardinia, Corsica and Capraia, and B. morisiana Bigazzi & Ricceri, described from the small island of San Pietro in south-western Sardinia (Bigazzi and Ricceri, 1992). Both species are hemicryptophytic and have decumbent stems bearing small flowers with white (B. morisiana) or pale blue (B. pygmaea), campanulate corollas with stamens hidden inside and no clear adaptations for entomophilous pollination. There are no published data about reproductive systems in these two species. Ecologically they are linked to constantly humid, often shady habitats, such as wet rocks, permanent springs, drippings, margins of rivulets and small stretches of riparian Alnus woods, from sea level to approx. 1000 m a.s.l. (Arrigoni, 1980; Gamisans and Marzocchi, 2005). Unlike species of subgenus Borago, which are exclusively myrmechocorus (cf. Hegi, 1975), seed dispersal in Buglossites is primarily by means of water and secondarily by means of ants, usually for short distances and within restricted biotopes (personal observation).
The narrow range, ecological specialization and the low competitive ability of these two species are in contrast with the condition of the ubiquist, ruderal, xerophytic and annual weed of B. officinalis, which suggested a ‘derived’ character for the latter and a relict character of paleoendemic for subgenus Buglossites (Guuleac, 1928; Contandriopoulos, 1962; Pignatti, 1982; Bigazzi and Ricceri, 1992). Under this assumption and in the absence of more data, however, it was difficult to make hypotheses on the origin of the tetraploid chromosome complement reported for B. pygmaea (Strey, 1931; Contandriopoulos, 1962; Diana Corrias, 1980) compared with the diploid karyotype of B. officinalis (D'Amato and Marchi, 1983; Luque, 1989). The report 2n = 16 in B. morisiana (here shown to be erroneous) indicated the existence of variation within Buglossites, at least in ploidy level, and suggested a possible derivation of the tetraploid B. pygmaea through polyploidization events not involving the taxa of subgenus Borago (Bigazzi and Ricceri, 1992). As in many angiosperm groups, polyploidy and other types of chromosomal processes are a major driving force of species formation in several Boragineae genera, among which Symphytum, Cynoglottis, Pulmonaria, Nonea and others (Selvi and Bigazzi, 2002; Hilger et al., 2004). In consideration of the lack of knowledge on this subject, a comparative karyological and cytogeographical investigation was undertaken to understand better the role of polyploidy and other types of chromosomal variation in the little-known evolutionary history of Borago. This survey brought to light an unexpected range of inter- and intraspecific variation in terms of ploidy levels and haploid numbers, suggesting the use of molecular tools to elucidate relationships between taxa with different base numbers, using the matK region of the chloroplast genome, and between species and infraspecific cytotypes, using the more variable ITS1 sequences of the nuclear ribosomal DNA. The usefulness of both markers in genus- and species-level systematics and phylogeny of Boraginales has been widely demonstrated in recent studies (Diane et al., 2002; Winkworth et al., 2002; Hilger et al., 2004; Selvi et al., 2004). The results of this work are presented and discussed in this paper, as they may contribute to a better understanding of relationships in Borago and of the Buglossites phylogeography in the main insular system of the western Mediterranean.
|
MATERIALS AND METHODS |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Plant material
The only two native populations of B. morisiana known to date and 12 populations of B. pygmaea were sampled in the field by transplanting three to five young plants for each population into small pots that were then transferred to the Botanical Garden of the University of Florence, Italy. A more intensive sampling was neither possible nor advisable in view of the rarity and often small population size of these endemics. In the case of B. pygmaea, the five Sardinian populations analysed were from the three main distribution sub-areas on the island, i.e. in the south, centre and north (Arrigoni, 1980
). Likewise, the six Corsican populations were from distant sites in the southern, central and northern parts of the island. Finally, one (the only one existing today) was from Capraia Island in the Tuscan Archipelago. In addition, one accession of B. officinalis and one of B. trabutii from personal field collections obtained in central Italy and Morocco, respectively, were examined. The geographical origins of all the accessions are given in Table 1; voucher specimens were deposited in the Herbaria of Florence and Cagliari Universities (FI and CAG, respectively). Silica-gel dried samples of leaf tissue were prepared for each of these populations for molecular analyses.
|
Karyotype analysis
At least ten root tips were collected in early spring and/or autumn from each cultivated plant of B. morisiana and B. pygmaea (perennials), and from numerous germinating seeds in the case of the annuals B. officinalis and B. trabutii. In total, 30–50 root tips were examined for each accession. They were pretreated with 0·002 M 8-hydroxyquinoline or 0·05 % colchicine, 2·5 h at room temperature and then fixed overnight in ethanol : glacial acetic acid (3 : 1). When necessary, they were preserved in 70 % ethanol at 3–4 °C until preparation. The meristematic tissue was then thoroughly rinsed in distilled water, hydrolysed in 1 N HCl at 60 °C for 6–7 min, and stained in lacto-propionic orcein overnight (Dyer, 1979
). The meristems were finally dissected and squashed on glass slides in a drop of 45 % acetic acid. The use of this staining technique (O-banding; Sharma and Sen, 2002) was useful for revealing heterochromatic segments which appeared as deeply stained regions in intercalary, telomeric or pericentromeric position. Metaphase plates were examined with a Zeiss Axioscop light microscope under oil immersion (x100), and photographed with a Canon digital system. Karyotype formulas were determined on selected enlarged prints by carefully measuring the length of chromosome arms and satellites. The centromeric index was calculated as the long : short arm ratio to classify chromosomes following the system of Levan et al. (1964): m = metacentric (r = 1·00–1·69), sm = submetacentric (r = 1·70–2·99), st = subtelocentric (r = 3·00–6·99). SAT chromosomes were recognized based on the occurrence and position of secondary constrictions corresponding to the nucleolar organizing region(s). The approximate heterochromatin content for each plate was estimated as a percentage of the length of heterochromatic bands with respect to the total karyotype length, measured on enlarged prints of selected micrographs (Bigazzi and Selvi, 2001). The intrachromosomal asymmetry index (A1) was calculated according to Romero Zarco (1986), while the interchromosomal index (A2) was measured as the ratio standard deviation of chromosome length/mean chromosome length. Final karyotype formulas, chromosome size and asymmetry indexes for each accession resulted from the mean values of at least five individual karyotypes.
DNA extraction and amplification
Genomic DNA was extracted following a modified 2xCTAB protocol (Doyle and Doyle, 1990) using samples of tissue cut from leaves. Modification consisted of the purification of DNA using polyvinylpyrrolidone to eliminate the negative effects that the abundant polysaccharidic mucilage of Borago subgenus Buglossites had during subsequent amplification (Aljanabi et al., 1999). The extracted DNA was quantified after agarose gel electrophoresis (0·6 % w/v) in TAE buffer (1 mM EDTA, 40 mM Tris-acetate) containing 1 µg mL–1 of ethidium bromide by comparison with a known mass standard.
Analysis of the matK region was done for B. trabutii and for one accession of B. morisiana, B. pygmaea and Trachystemon orientalis (Table 1). The two primers matKF (5'-GAT TCG AAC CCG GAA CTA G-3') and matKR (5'-CGA TCA ACA TCT TCT GGA ATC-3') used in this study were obtained, respectively, from the two primers trnK3R and tK3MY2F used for Myosotis (Winkworth et al., 2002). The former was modified in the last four nucleotides and the latter in the second and eighth nucleotides, based on the published sequences of Borago officinalis (GenBank AJ429308
[GenBank]
) and Echium candicans L. f. (GenBank AF542610
[GenBank]
).
Analysis of ITS1 nuclear rDNA was done for B. trabutii, the two populations of B. morisiana and nine accessions of B. pygmaea representing the distribution range and the different cytotypes revealed by karyotype analysis (Table 1); the primers ITS4 and ITS5 of Baldwin (1992) were used.
For both matK and ITS1, PCR amplifications were performed in a total volume of 50 µL containing 5 µL of reaction buffer (Dynazyme II; Finnzyme, Espoo, Finland), 1·5 mM MgCl2, 20 pmol of each primer, 200 µM of each dNTP, 1 U of Taq DNA polymerase (Dynazyme II; Finnzyme) and 10 ng of template DNA. Reactions were performed in a Perkin-Elmer 9600 thermocycler (Perkin Elmer, Norwalk, CT, USA). Subsequently, 5 µL of each amplification mixture was analysed by agarose gel (1·5 % w/v) electrophoresis in TAE buffer containing 1 µg mL–1 ethidium bromide. The PCR reactions were purified from excess salts and primer with the PCR Purification Kit (Roche, Mannheim, Germany).
Automated DNA sequencing was performed directly from the purified PCR products using BigDye Terminator v.2 chemistry and an ABI310 sequencer (PE-Applied Biosystems, Norwalk, CT, USA).
Sequence alignment and analysis
The ITS1 sequences were checked for orthology to the sequences of Borago officinalis (GenBank AY383283
[GenBank]
), Anchusa officinalis (AY045710
[GenBank]
) and Trachystemon orientalis (AY383287
[GenBank]
), all of the tribe Boragineae. The two latter species were then used as outgroups for cladogram construction based on their position in the phylogeny of the tribe (Hilger et al., 2004). The matK sequences were aligned with the sequences of Borago officinalis (AJ429308
[GenBank]
) and Echium vulgare L. (AY092893
[GenBank]
); the latter and Trachystemon orientalis (original matK sequence obtained from plants cultivated in the Botanical Garden of Florence University, voucher B & S 99·023—FI, GenBank no. DQ657836
[GenBank]
) were then used as outgroup representatives in the matK tree.
Multiple alignments were performed with Multalin (Corpet, 1988), and then further examined and slightly modified manually. All characters were weighted equally, and character state transitions were treated as unordered. Gaps were coded and added at the end of the sequences according to the ‘simple gap’ coding method of Simmons and Ochoterena (2000). Neighbour-joining (NJ) and maximum parsimony (MP) methods were used to analyse the aligned sequences. NJ trees (Saitou and Nei, 1987) were obtained on the basis of a Kimura-2 parameter distance matrix. MP trees were calculated with PAUP* ver. 4.0 b (Swofford, 1998) through a heuristic search, adding sequences at random, with tree-bisection-reconnection branch swapping, MULTREES option on, ADDSEQ = random, ten randomized replicates. Accelerated transformation (ACCTRAN) optimization was used to infer branch lengths. Internal support to the branches was estimated by means of 50 % majority-rule bootstrap analysis (Felsenstein, 1985), with 1000 random addition replicates for ITS1 and 100 for matK.
|
RESULTS |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Karyotype analyses
The summary of somatic and haploid chromosome numbers, ploidy levels, karyotype formulas, mean chromosome lengths and approximate lengths of heterochromatic bands observed in the examined accessions is reported in Table 2.
|
Borago trabutii
All examined individuals (18) showed an invariable diploid complement of 2n = 2x = 12, based on x = 6 (Fig. 1A). This is the first karyological analysis of a native accession of this rare endemic of the Moroccan Atlas, and the first finding of such a low number in Borago. This observation is in contrast with an old report of 2n = 16 from non-native material (Contandriopoulos, 1962
), suggesting that a wider population analysis may be useful to ascertain the existence of infraspecific chromosomal races. However, occurrence of cytotypes with different base numbers in such a well-characterized endemism with little morphological variation, restricted range and isolated position is unlikely. The karyotype has a very low intrachromosomal asymmetry (Fig. 2) due to the presence of only metacentric chromosomes, two of which are provided with satellites (Fig. 3A). Chromosomes are distinctly larger than in B. morisiana and B. pygmaea but slightly smaller than in B. officinalis (see below). With respect to the latter species, a larger variation in size accounts for a higher interchromosomal asymmetry. Each homologous pair shows a distinct banding pattern visibile as deeply stained heterochromatic regions in pericentromeric, intercalary and telomeric positions. Representing approx. 25 % of the total karyotype length, heterochromatin is more abundant than in the two Buglossites species, but less than in B. officinalis (see below).
|
|
|
Borago officinalis
The accession examined confirmed that this widespread herb is diploid with 2n = 2x = 16 (Fig. 1B), as already known from previous reports from Italy and Spain (e.g. D'Amato and Marchi, 1983
; Luque, 1989). The base number is therefore x = 8, as in B. pygmaea (see below). However, the karyotype of this species is clearly closer to that of B. trabutii in terms of chromosome size class, asymmetry (Fig. 2), and heterochromatin content. In addition to the higher number of chromosomes, it differs from B. trabutii also by a higher A1 value due to the presence of two pairs of submetacentrics and six pairs of metacentrics, one of which is satellited (Fig. 3B). Interchromosomal asymmetry is instead lower. Borago officinalis shows the largest chromosomes in the genus (approx. 4·5 µm) and the highest content of constitutive heterochromatin (about 38 % of the total karyotype length). However, the banding pattern in this species is different from that observed in B. trabutii. In all the homologue pairs, heterochromatin is organized in large and compact regions in the pericentromeric position, while no telomeric and intercalary bands could be detected.
Borago morisiana
The two populations examined (MP and ML, the only ones known to date) are geographically isolated from each other, as one is from a coastal habitat in the Island of San Pietro (south-western Sardinia) and the other from the rugged and mountainous Laconi region in the central part of the island (Fig. 4). The discovery of this rare endemic in the latter locality is recent and still not published.
|
Both populations showed an identical chromosome complement consisting of 2n = 2x = 18 based on x = 9 (Fig. 1D). The karyotype is formed by four pairs of metacentrics, three of submetacentrics and two pairs of subtelocentrics, one of which is distinctly smaller and provided with satellites on the short arms (Fig. 3C). The intrachromosomal asymmetry is considerable, due to the prevalence of subtelo- and submetacentric chromosomes with respect to metacentric chromosomes (Fig. 2). Interchromosomal asymmetry is also relatively high due to the variation in chromosome length (1·9–2·5 µm). Heterochromatin was visible as deeply stained regions mainly in intercalary and pericentromeric positions, but in its content is always <20 % of the total karyotype length.
Borago pygmaea
The populations examined revealed remarkable differences in somatic karyotypes and even base chromosome numbers. Three allopatric cytotypes were recognized, and their geographical distribution is shown in Fig. 4.
The five Sardinian accessions (SI, SN, LA, SD and CL) are all characterized by a complement with 2n = x = 30 based on x = 15 (Fig. 1E). The karyotype consists of nine pairs of metacentrics, one of which was provided with satellites, and six pairs of submetacentrics (Fig. 3D). No subtelocentrics are present. The A1 asymmetry is lower than in B. morisiana (Fig. 2), whereas mean chromosome size, A2 asymmetry and mean length of heterochromatic bands are comparable.
The three populations from south Corsica (AU, CG and AL) possess an invariable tetraploid complement of 2n = 4x = 32, based on x = 8 (Fig. 1F). This differs from that of the Sardinian accessions by the presence of eight pairs of metacentrics, seven pairs of submetacentrics, one of which is satellited, and one pair of subtelocentrics (Fig. 3E). Consequently, the average A1 asymmetry is slightly higher, whereas for the A2 value, mean chromosome size and heterochromatic bands are comparable with the Sardinian populations.
The three accessions from central and northern Corsica (GT, BI and OL) and the one from the western coast of Capraia (CP) are characterized by an invariable hexaploid complement of 2n = 6x = 48, based on x = 8 (Fig. 1C). Their karyotype consists of 12 pairs of metacentrics and ten pairs of submetacentric chromosomes, one of which with satellites on the short arms, and two pairs of subtelocentrics. Average intra- and interchromosomal asymmetry values are higher than in the other accessions (Fig. 2). Mean chromosome size is comparable, whereas the percentage of heterochromatin is slightly lower.
Phylogenetic analyses
Region matK
Sequences have been deposited in GenBank-EMBL-DDB and can be retrieved using the numbers in Table 1 (except for Trachystemon orientalis, accession in Materials and methods). The aligned sequences (available from the authors in NEXUS format upon request) are 531 bp long, including codified gaps (16 positions). Ingroup variation is low, with only 28 variable positions in at least one accession. In the phylogenetic analysis, 461 positions are constant, 52 are variable but parsimony non-informative and 18 are parsimony-informative.
The single most-parsimonious tree (length = 77; consistency index = 0·95; retention index = 0·80), topologically identical to the neighbour-joining tree, is shown in Fig. 5. Borago forms a strongly supported clade (94 % BS), with B. trabutii appearing as the earliest divergent lineage and sister to the other members of the genus. This species does not cluster with B. officinalis, which differs in eight 1-bp substitutions and 12 1-bp deletions. Borago officinalis is, in turn, sister to B. morisiana and B. pygmaea with which it forms a well-supported clade (90 % BS). The monophyly of subgenus Buglossites is also strongly supported (91 % BS). Borago pygmaea differs from B. morisiana in only two deletions (one of 2 bp and one of 1 bp) and one 1-bp insertion.
|
Region ITS1
Sequences have been deposited in GenBank-EMBL-DDB (accession numbers in Table 1). The aligned sequence data set (available in NEXUS format) is 282 bp long. Ingroup variation is remarkable, with 115 positions (approx. 41 %) being variable in at least one accession. Adding the two outgroups for the purposes of the phylogenetic analysis, 131 characters are constant, 83 parsimony-uninformative and 68 (= 24.1 %) parsimony-informative. The most-parsimonious phylogeny (L = 208, CI = 0·89, RI = 0·82), which resulted in being topologically identical to the strict consensus and neighbour-joining trees, is shown in Fig. 6. Optimized branch lengths and bootstrap values >50 % are shown, and somatic chromosome numbers are superimposed for each accession examined.
|
With respect to the outgroups, Borago is monophyletic and strongly supported (100 % BS). Borago trabutii is again the early divergent lineage and sister to the rest of the ingroup. However, there is no evidence for the monophyly of subgenus Borago due to the position of B. officinalis which does not cluster with B. trabutii but is sister to the group of B. morisiana/B. pygmaea. Together with the latter, B. officinalis forms a well-supported clade (91 % BS), but the monophyly of subgenus Buglossites is even more strongly corroborated (98 %).
Within Buglossites, no clear correspondence between the clades and cytotypes of B. pygmaea can be detected. The hypotetraploid (2n = 30) accession of this species from northern Sardinia (CL) shows several autapomorphic indels in the sequence and appears sister to the rest of Buglossites, failing to cluster with the two other accessions from southern and central Sardinia with the same chromosome complement (LA and SI, 2n = 30). These form a small group with one of the two Corsican tetraploid accessions (CG), though this relationship is not supported by bootstrap values. The other Corsican tetraploid population (AL) is external to this group, confirming the lack of relationship with either chromosome complement or geographical location in the group of 2n = 32 and 30 races. The mixed Corsican/Sardinian accessions of tetra- and hypotetraploid B. pygmaea (CG, SI and LA) form a moderately supported lineage (69 %), together with a monophyletic, sister clade (79 %) consisting of the two accessions of B. morisiana (51 %) and the hexaploid populations of B. pygmaea. The two northernmost accessions of the latter from Capraia Island (CP) and the Corsican ‘finger’ (OL) cluster together with 60 % BS support, but their relationships to the two more southern hexaploids (GT and BI) remain unresolved.
|
DISCUSSION |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
According to the scanty literature available, Borago is characterized by the base x = 8, with B. officinalis, B. trabutii, B. longifolia and B. morisiana being diploid (Contandriopoulos, 1962
; D'Amato and Marchi, 1983; Luque, 1989; Bigazzi and Ricceri, 1992) and B. pygmaea tetraploid (Strey, 1931; Contandriopoulos, 1962; Diana Corrias, 1980). However, this study shows that the reports for B. morisiana and probably B. trabutii are wrong, while it reveals the existence of four base numbers, x = 6, x = 8, x = 9, x = 15, and three ploidy levels (di-, tetra- and hexaploid) based on x = 8. Chromosomal variation in this small genus is therefore much wider than previously known (Fig. 7), suggesting that karyological rearrangements have been a major driving force for the evolution and radiation of the taxa into present-day ranges. This is not surprising, as comparable levels of variation are known to occur in related Boragineae groups, such as Symphytum (Gadella et al., 1983), Pulmonaria (Sauer, 1975) and Nonea (Selvi and Bigazzi, 2002; Bigazzi and Selvi, 2003).
|
Relationships between primary base numbers
As in other plant groups, a major issue of evolutionary significance concerns the relationships between the base numbers (Grif, 2000
; Stace, 2000). Two possible scenarios can be considered in the light of the present karyological data and molecular cladograms, the first of which assumes the number 2n = 12 of the early divergent lineage of B. trabutii as the ancestral condition in Borago. The base x = 6 is the lowest in Boraginaeae, and is known only for a few other taxa with plesiomorphic morphological characters like the species of Brunnera (Britton, 1951; Al-Shehbaz, 1991; Bigazzi and Selvi, 2001). Under this scenario, the backcrossing of unreduced (n = 12) and regular (n = 6) gametes may have resulted in the formation of a triploid ‘bridge’ with 2n = 18 and in the origin of the new base x = 9 found in B. morisiana. Descending aneuploidy by tandem centric fusion between two homologous pairs in 2n = 18 Buglossites populations could explain the origin of 2n = 16 and x = 8, at least in the Corso-Sardinian system.
More plausible, however, is the hypothesis of x = 8 as the ancestral base conserved in both subgenera Borago (B. officinalis) and Buglossites (B. pygmaea), as well as in most taxa of the outgroup genera Anchusa, Echium (Strey, 1931; Britton, 1951) and, possibly, the paleopolyploid Trachystemon orientalis (2n = 56; Markova and Ivanova, 1970; Coppi et al., 2006). This hypothesis would imply a dysploid origin for both x = 9 in the Buglossites clade, possibly by chromosome fission, and x = 6 in the B. trabutii clade, by progressive descending aneuploidy. At the diploid level, the latter type of rearrangement does not imply a loss of the requisite genetic material but simply a decrease in the number of chromosomes as separate linkage groups (Jackson and Casey, 1980; Levin, 2002). Although no karyological evidence for the origin of x = 6 by descending aneuploidy can be provided here, it is known that this mechanism has played an important role in speciation processes of other Boragineae genera such as Nonea (Selvi and Bigazzi, 2002; Bigazzi and Selvi, 2003) and Pulmonaria (Sauer, 1975), as well as in other dicot groups [i.e. Vicia (Hollings and Stace, 1974), Genista (Sañudo, 1979), Minuartia (Favarger, 1999) and Houstonia (Church, 2003)]. The hypothesis of x = 8 as the ancestral condition in Borago and related groups would be supported by a confirmation of 2n = 16 in B. longifolia, the other North African endemic of subgenus Borago (Contandriopoulos, 1962).
The tetra- and hypotetraploid races of B. pygmaea
In B. pygmaea, the apparent extinction of diploid progenitors with 2n = 16, already hypothesized by Contandriopoulos (1962) and here corroborated, may have been concomitant with ancestral autopolyploidization events which have triggered the formation of the tetraploid race via sexual union of unreduced gametes. This is a prime mechanism of increase in ploidy level and chromosome number in many angiosperm groups (Thompson and Lumaret, 1992; Bretagnolle and Thompson, 1995; Leitch and Bennett, 1997; Ramsey and Schemske, 1998; Burton and Husband, 2001; Levin, 2002), not least in other taxa of Boragineae (Bigazzi and Selvi, 2001; Selvi and Bigazzi, 2002). The achievement of the polyploid condition in B. pygmaea has possibly enhanced its competitive ability, for example, in allowing the high rate of selfing that was observed in cultivated plants (personal observation). The breakdown of mechanisms that reduce inbreeding in diploids is believed to be a prime biological feature that conveys a competitive advantage to natural polyploids over diploid progenitors (Thompson and Lumaret, 1992). Furthermore, the increased heterozygosity and allelic diversity which occur in sexual autopolyploids as a consequence of genome multiplication is acknowledged as another factor leading to the formation of new adaptive gene combinations (Brochmann et al., 1992; Soltis and Soltis, 1993, 1995, 2000; Otto and Whitton, 2000; Wendel, 2000; Levin, 2002). Finally, the successful establishment, persistence and spread of autotetraploid B. pygmaea has possibly been enabled by the small size of the populations, in which genetic and environmental effects, as well as effects due to the increased mating that occurs among closely related individuals, may cause a marked increase in the frequency of 2n gamete formation (Thompson and Lumaret, 1992). Under these conditions the diploid progenitors may have become extinct as a consequence of ‘the minority cytotype exclusion’ rule which is believed to occur in local plant populations (Levin, 1975).
In contrast to the diploid B. morisiana, the newly formed tetraploid B. pygmaea may have easily radiated across the Corso-Sardinian system because of the land connections between the two islands which occurred for the best part of the last 30 Myrs (Cherchi and Montadert, 1982; Speranza et al., 2002).
The origin and establishment of x = 15 has probably represented a second step in the karyological evolution of B. pygmaea. Aneuploid reduction again provides a plausible explanation for the formation of this hypotetraploid race, although the reasons for its current distribution restricted to Sardinia are obscure. Karyotype analysis suggests a secondary origin via a tandem centric fusion between the pair of SAT submetacentrics and the pair of subtelocentrics of the 2n = 32 race, with formation of the two SAT metacentrics and loss of subtelocentrics in the 2n = 30 complement. This kind of rearrangement by descending aneuploidy leading to the secondary formation of higher base numbers is common in polyploid complexes, e.g. Artemisia (James et al., 2000; Stace, 2000).
The hexaploid race of B. pygmaea
Backcrossing between 2n and regular n gametes in 2n = 32 populations followed by minor rearrangements in the SAT submetacentric and subtelocentric units and an increase in intrachromosomal asymmetry is a likely mechanism for the cryptic formation of the hexaploid cytotype (32 unreduced: 2n = 16m + 12sm + 2smSAT + 2st; 32 reduced: n = 8m + 6sm + 1smSAT+ 1st 
hexaploid: 24m + 18sm + 2smSAT + 4st). In this case, there is evidence that SAT chromosomes are frequently involved in karyotype changes of Buglossites, possibly in relation to the necessity of maintaining only one functional nucleolar organizing region per genome. Several polyploid plants have only one nucleolar organizing region because those resulting from genome multiplication that are in excess and not functional are ‘suppressed’ or silenced in some way (Stace, 2000).
The apparent lack of hybrid zones between the tetraploid and hexaploid races in Corsica is likely to depend on historical and ecological causes, as in other cases of geographical separation between cytotypes with different ploidy levels (Thompson and Lumaret, 1992; Levin, 2002; Thompson, 2005). Also in the light of the monophyly of the 2n = 48 accessions suggested by the phylogenetic analysis, a model of single-event origin along the northern limit of the tetraploid range is, in this case, more plausible than an alternative one of multiple events. The latter often results in mixed polyploid complexes with no clear distributional patterns of the cytotypes as a consequence of directional or balanced selection (Petit et al., 1999; Soltis and Soltis, 1999; Weiss et al., 2002). In B. pygmaea, the geographical segregation may be the result of a northward, unidirectional spread of the hexaploid race successively maintained by ecological factors limiting the range of the tetraploid to the north but not that of the more tolerant hexaploid. Ecological differentiation is acknowledged as a key factor for maintaining cytotypes in an allopatric condition (Thompson and Lumaret, 1992; Levin, 2002; Husband, 2004). Comparative autoecological analyses of the tetra- and hexaploid races could be useful in understanding the role of edaphic factors in the substantial correspondence between the range of the hexaploid and the so-called ‘schistose sector’ of north-eastern Corsica formed mainly by schistose and volcanic rocks (Contandriopoulos, 1962; Thompson, 2005).
The occurrence of B. pygmaea on the wild, western coast of Capraia Island is also worth mentioning. In both karyotype and ITS1 sequence, this small population is identical to the northernmost accession examined from Corsica (OL), which strongly supports a relationship with the populations in northern Corsica and a relatively recent event of disjunction. The lack of adaptative traits for long-distance seed dispersal in B. pygmaea and the volcanic origin of Capraia, which emerged from the sea in the Upper Miocene (7·3–6·0 Myrs ago; Borelli and Groppelli, 2002) about 31 km at the north-east of the Corsican cape, pose the question of how and when the hexaploid race successfully reached the island.
Phylogeny of Borago
Sequence variation in the matK and ITS1 regions indicate that the genus Borago and subgenus Buglossites are both monophyletic lineages, whereas relationships in subgenus Borago still remain unclear. To this purpose, further analyses should make use of additional markers and also include the North African endemic B. longifolia. Nonetheless, the present reconstruction indicates that the split of Buglossites from Borago has been karyologically paralleled by an increase in chromosome number and karyotype asymmetry, a decrease in chromosome size and heterochromatin content, and the appearance of polyploidy.
Relationships between the two endemic Buglossites species are obscured by the high infraspecific ITS1 variation in the polyploid B. pygmaea compared with the diploid B. morisiana. The identical sequences and chromosome complements found in the two isolated populations of the latter species support the hypothesis of an ancient origin and a condition of relict paleo-endemic with no morphological variation and extremely reduced, fragmented distribution (Bigazzi and Ricceri, 1992). On the other hand, chromosomal and ITS1 variation indicate that active evolutionary forces have worked in the case of B. pygmaea. Polymorphism in the ITS1 region is especially found in the 2n = 32 and 2n = 30 accessions (12·8 % of variable positions) possibly due to the occurrence of paralogue sequences in the multiple copies of the genome, a well-known feature of several polyploid complexes (Mansion et al., 2005; Stace, 2005). In particular, different paralogues may be the result of recurrent events of polyploidization, as in other autopolyploids with multiple origins (Soltis and Soltis, 1999). The lower ITS1 variation in the hexaploid cytotype may appear therefore unexpected, although this is likely to be related to its more recent, monophyletic origin in central northern Corsica as discussed above. The sister position that this race takes with respect to B. morisiana in the cladogram appears difficult to explain, and parallel evolution of the ITS1 repeats in the hexaploid cytotype and B. morisiana cannot be ruled out as a possible cause of this unexpected relationship.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
This paper is dedicated to the memory of our beloved friend and colleague Massimo Bigazzi, who started this work with enthusiasm but passed away prematurely on 19 April, 2006. F.S. and A.C. wish to thank Gianluigi Bacchetta (Cagliari) for his help during field work and invaluable information on the distribution of B. morisiana and B. pygmaea. Two anonymous referees provided useful suggestions on the first version of the manuscript. Research grants from the Italian Ministry of Scientific Research (MIUR) and University of Firenze are acknowledged.
|
FOOTNOTES |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
Deceased, 19 April 2006. 
|
LITERATURE CITED |
|---|
|
TOPFOOTNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSLITERATURE CITED |
|---|
-
Aljanabi SM, Forget L, Dookun A. 1999. An improved and rapid protocol for the isolation of polysaccharide-and polyphenol-free sugarcane DNA. Plant Molecular Biology 17: 1–8.
Al-Shehbaz I. 1991. The genera of Boraginaceae in the southeastern United States. Journal of the Arnold Arboretum, Supplementary Series 1: 1–169.
Arrigoni PV. 1980. Le piante endemiche della Sardegna 82—Borago pygmaea (DC.) Chater & Greuter. Bollettino della Società Sarda di Scienze Naturali 19: 331–335.
Baldwin BG. 1992. Phylogenetic utility of the internal transcribed spacers of nuclear ribosomal DNA in plants: an example from the Compositae. Molecular Phylogenetics and Evolution 1: 3–16.[CrossRef][Medline]
Bigazzi M, Ricceri C. 1992. Borago morisiana Bigazzi et Ricceri (Boraginaceae), a new species from Sardinia. Webbia 46: 191–202.
Bigazzi M, Selvi F. 2001. Karyotype morphology and cytogeography in Brunnera and Cynoglottis (Boraginaceae). Botanical Journal of the Linnean Society 136: 365–378.[CrossRef]
Bigazzi M, Selvi F. 2003. Chromosome variation in Anatolian species of Nonea Medik. (Boraginaceae) with special reference to endemics and N. persica. Caryologia 56: 509–519.
Borelli E, Groppelli G. 2002. Evoluzione geo-vulcanologica dell'Isola di Capraia. In: Morelli M, ed. L'isola di Capraia: progetto di un paesaggio insulare mediterraneo da conservare. Firenze: Alinea editrice, 26–39.
Bretagnolle F, Thompson JD. 1995. Gametes with the somatic chromosome number: mechanisms of formation and role in the evolution of polyploid plants. New Phytologist 129: 1–22.[CrossRef]
Britton D. 1951. Cytogenetic studies on the Boraginaceae. Brittonia 7: 233–266.[CrossRef]
Brochmann C, Soltis PS, Soltis DE. 1992. Recurrent formation and polyphyly of Nordic polyploids in Draba (Brassicaceae). American Journal of Botany 79: 673–688.[CrossRef]
Burton TL, Husband BC. 2001. Fecundity and offspring ploidy in matings among diploid, triploid, and tetraploid Chamerion angustifolium (Onagraceae): consequences for tetraploid establishment. Heredity 87: 573–582.[CrossRef][ISI][Medline]
Cherchi A, Montadert L. 1982. Oligo-Miocene rift of Sardinia and the early history of the Western Mediterranean basin. Nature 298: 736–739.[CrossRef]
Church SA. 2003. Molecular phylogenetics of Houstonia (Rubiaceae): descending aneuploidy and breeding system evolution in the radiation of the lineage across North America. Molecular Phylogenetics and Evolution 27: 223–238.[CrossRef][ISI][Medline]
Contandriopoulos J. 1962. Recherches sur la flore endemique de la Corse et sur ses origines. Annales Faculté des Sciences de Marseille 32: 1–354.
Coppi A, Bigazzi M, Selvi F. 2006. Chromosome studies in Mediterranean Boraginaceae. Flora Mediterranea 16 (in press).
Corpet F. 1988. Multiple sequence alignment with hierarchical clustering. Nucleic Acids Research 16: 10881–10890.
D'Amato G, Marchi P. 1983. Heterochromatin in Borago officinalis L. (Boraginaceae): comparison between Feulgen and Giemsa stained heterochromatic segments. Annali di Botanica (Roma) 41: 164–168.
Diana Corrias S. 1980. Numeri cromosomici per la Flora Italiana, 702. Informatore Botanico Italiano 12: 125–126.
Diane N, Förther H, Hilger HH. 2002. A systematic analysis of Heliotropium, Tournefortia and allied taxa of the Heliotropiaceae (Boraginales) based on ITS1 sequences and morphological data. American Journal of Botany 89: 287–295.
Doyle JJ, Doyle JL. 1990. Isolation of plant DNA from fresh tissue. Focus 12: 13–15.[Medline]
Dyer AF. 1979. Investigating chromosomes. London: Edward Arnold.
Favarger C. 1999. Contribution à la cytogéographie du Minuartia glomerata (M.Bieb.) Degen (Caryophyllaceae). Bullettin de la Societé Neuchâteloise des Sciences Naturelles 122: 27–33.
Felsenstein J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783–791.[CrossRef][ISI]
Gadella TWJ, Kliphuis E, Huizing HJ. 1983. Cyto- and chemotaxonomical studies in the sections Officinalia and Coerulea of the genus Symphytum. Botanica Helvetica 93: 169–192.
Gamisans J, Marzocchi JF. 2005. La flore endémique de la Corse. Aix-en-Provence: Edisud.
Grif V. 2000. Some aspects of plant karyology and karyosystematics. International Review of Cytology 196: 131–162.[ISI][Medline]
Guuleac M. 1928. Die monotypischen und artenarmen Gattungen der Anchuseae (Caryolopha, Brunnera, Hormuzakia, Gastrocotyle, Phyllocara, Trachystemon, Procopiania und Borago). Buletinul Facultalai Stiinte Cernãuti 2: 394–461.
Hegi G. 1975. Illustrierte Flora von Mitteleuropa 5 (3). Berlin: Paul Parey.
Hilger HH, Selvi F, Papini A, Bigazzi M. 2004. Molecular systematics of Boraginaceae tribe Boragineae based on ITS1 and trnL sequences, with special reference to Anchusa s.l. Annals of Botany 94: 201–212.
Hilger HH, Gottschling M, Selvi F, Bigazzi M, Langström E, Zippel E, et al. 2005. The Euro+Med treatment of Boraginaceae in Willdenowia 34—a response. Willdenowia 35: 43–48.
Hollings E, Stace CA. 1974. Karyotype variation and evolution in the Vicia sativa aggregate. New Phytologist 73: 195–208.
Husband BC. 2004. Chromosomal variation in plant evolution. American Journal of Botany 91: 621–625.
Jackson RC, Casey J. 1980. Cytogenetics of polyploids. In: Lewis WH, ed. Polyploidy: biological relevance. New York, NY: Plenum Press, 17–44.
James CM, Wurzell BS, Stace CA. 2000. A new hybrid between a European and a Chinese species of Artemisia (Asteraceae). Watsonia 23: 139–147.
Leach CR, Mayo O, Bürger RR. 1990. Quantitatively determined self-incompatibility. 2. Outcrossing in Borago officinalis. Theoretical and Applied Genetics 79: 427–430.[CrossRef]
Leitch IJ, Bennett MD. 1997. Polyploidy in angiosperms. Trends in Plant Science 2: 470–476.[CrossRef][ISI]
Levan A, Fredga K, Sandberg AA. 1964. Nomenclature for centromeric position on chromosomes. Hereditas 52: 201–220.
Levin DA. 1975. Minority cytotype exclusion in local plant populations. Taxon 24: 35–43.[Medline]
Levin DA. 2002. The role of chromosomal change in plant evolution. New York, NY: Oxford University Press.
Luque T. 1989. Estudio cariológico de Boraginaceae Españolas—IV. Boletim da Sociedade Broteriana 2: 211–220.
Mansion G, Zeltner L, Bretagnolle F. 2005. Phylogenetic patterns and polyploid evolution within the Mediterranean genus Centaurium (Gentianaceae–Chironieae). Taxon 54: 931–950.
Markova M, Ivanova P. 1970. Kariologische untersuchungen der vertreter der fam. Boraginaceae, Labiatae und Scrophulariaceae in Bulgarien. Izvestija na Botaniceskija Institut 21: 123–131.
Montaner C, Floris E, Alvarez JM. 2000. Is self-compatibility the main breeding system in borage (Borago officinalis L.)? Theoretical and Applied Genetics 101: 185–189.[CrossRef]
Otto SP, Whitton J. 2000. Polyploid incidence and evolution. Annual Review of Genetics 34: 401–437.[CrossRef][ISI][Medline]
Petit C, Bretagnolle F, Felber F. 1999. Evolutionary consequences of diploid–polyploid hybrid zones in wild species. Trends in Ecology and Evolution 14: 306–311.
Pignatti S. 1982. Flora d'Italia, Vol. 2. Bologna: Edagricole, 419.
Ramsey J, Schemske DW. 1998. Pathways, mechanisms, and rates of polyploid formation in flowering plants. Annual Review of Ecology and Systematics 29: 47–501.
Romero Zarco C. 1986. A new method for estimating karyotype asymmetry. Taxon 35: 526–530.[CrossRef]
Saitou N, Nei M. 1987. The neighbour-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4: 406–425.[Abstract]
Sañudo A. 1979. Chromosome variability in the Genisteae (Adans.) Benth. Webbia 34: 363–408.
Sauer W. 1975. Karyo-systematische Untersuchungen an der Gattung Pulmonaria (Boraginaceae). Bibliotheca Botanika 131: 1–85.
Selvi F, Bigazzi M. 2002. Chromosome studies in Turkish species of Nonea (Boraginaceae): the role of polyploidy and descending dysploidy in the evolution of the genus. Edinburgh Journal of Botany 59: 405–420.[CrossRef]
Selvi F, Papini A, Hilger HH, Bigazzi M, Nardi E. 2004. The phylogenetic relationships of Cynoglottis (Boraginaceae) inferred from ITS and trnL sequence variation. Plant Systematics and Evolution 246: 145–209.[CrossRef]
Sharma A, Sen S. 2002. Chromosome botany. Enfield, USA: Intercept.
Simmons MP, Ochoterena H. 2000. Gaps as characters in sequence-based phylogenetic analyses. Systematic Biology 49: 369–381.[CrossRef][ISI][Medline]
Soltis DE, Soltis PS. 1993. Molecular data and the dynamic nature of polyploidy. Critical Reviews in Plant Sciences 12: 243–273.
Soltis DE, Soltis PS. 1995. The dynamic nature of polyploid genomes. Proceedings of the National Academy of Sciences of the USA 92: 8089–8091.
Soltis DE, Soltis PS. 1999. Polyploidy: recurrent formation and genome evolution. Trends in Ecology and Evolution 14: 348–352.
Soltis PS, Soltis DE. 2000. The role of genetic and genomic attributes in the success of polyploids. Proceedings of the National Academy of Sciences of the USA 97: 7051–7057.
Speranza F, Villa IM, Sagnotti L, Florindo F, Cosentino C, Cipollari P, et al. 2002. Age of the Corsica-Sardinia rotation and Liguro-Provençal basin spreading: new paleomagnetic and Ar/Ar evidence. Tectonophysics 347: 231–251.[CrossRef][ISI]
Stace CA. 2000. Cytology and cytogenetics as a fundamental taxonomic resource for the 20th and 21th centuries. Taxon 49: 451–477.