GitBucket
Newer
zeynalig
<?xml version="1.0" ?> <tei> <teiHeader> <fileDesc xml:id="_1471-2148-8-1"/> </teiHeader> <text xml:lang="en"> <note place="headnote">BioMed Central <lb/></note> <page>Page 1 of 8 <lb/></page> <note place="footnote">(page number not for citation purposes) <lb/></note> <front>BMC Evolutionary Biology <lb/> Open Access <lb/> Research article <lb/> Prezygotic reproductive isolation between Saccharomyces cerevisiae <lb/> and Saccharomyces paradoxus <lb/> Calum J Maclean* and Duncan Greig <lb/> Address: The Galton Laboratory, University College London, 4 Stephenson Way, London, NW1 2HE, UK <lb/>Email: Calum J Maclean* -calum.maclean@ucl.ac.uk; Duncan Greig -d.greig@ucl.ac.uk <lb/>* Corresponding author <lb/> Abstract <lb/> Background: Matings between different Saccharomyces sensu stricto yeast species produce <lb/>sexually sterile hybrids, so individuals should avoid mating with other species. Any mechanism that <lb/>reduces the frequency of interspecific matings will confer a selective advantage. Here we test the <lb/>ability of two closely-related Saccharomyces sensu stricto species to select their own species as <lb/>mates and avoid hybridisation. <lb/> Results: We set up mate choice tests, using five independently isolated pairs of species, in which <lb/>individual germinating spores were presented with the opportunity to mate either with a <lb/>germinating spore of their own species or with a germinating spore of the other species. For all five <lb/>strain pairs, whether a S. cerevisiae or S. paradoxus occupies the role of "chooser" strain, the level <lb/>of hybridisation that is observed between the two species is significantly lower than would be <lb/>expected if mates were selected at random. We also show that, overall, S. cerevisiae exhibited a <lb/>stronger own-species preference than S. paradoxus. <lb/> Conclusion: Prezygotic reproductive isolation is well known in higher organisms but has been <lb/>largely overlooked in yeast, an important model microbe. Here we present the first report of <lb/>prezygotic reproductive isolation in Saccharomyces. Prezygotic reproductive isolation may be <lb/>important in yeast speciation or yeast species cohesion, and may have evolved to prevent wasted <lb/>matings between different species. Whilst yeast has long been used as a genetic model system, little <lb/>is known about yeast in the wild. Our work sheds light on an interesting aspect of yeast natural <lb/>behaviour: their ability to avoid costly interspecific matings. <lb/></front> <body>Background <lb/> The biological species concept defines a species as an <lb/>interbreeding group that is reproductively isolated from <lb/>other such groups [1]. Species are isolated by barriers that <lb/>either prevent fertilisation between species (prezygotic <lb/>barriers) or those that allow fertilisation but make the <lb/>resulting hybrid sterile or inviable (postzygotic barriers) <lb/>[2] (for a review see [3]). <lb/>Mating in yeast occurs through the fusion of haploid gam-<lb/>etes. When starved, diploid Saccharomyces yeast cells pro-<lb/>duce haploid spores by meiosis. Each diploid cell <lb/>produces four dormant and resilient haploid spores, two <lb/>spores of each mating type (a and α). When nutrients <lb/>become available again the spores germinate to become <lb/>metabolically active gametes. Gametes of both mating <lb/>types produce attractive pheromones used to signal to the <lb/>other mating type. Gametes of different mating-types fuse, <lb/></body> <front>Published: 7 January 2008 <lb/> BMC Evolutionary Biology 2008, 8:1 doi:10.1186/1471-2148-8-1 <lb/>Received: 15 June 2007 <lb/>Accepted: 7 January 2008 <lb/>This article is available from: http://www.biomedcentral.com/1471-2148/8/1 <lb/>© 2008 Maclean and Greig; licensee BioMed Central Ltd. <lb/>This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), <lb/>which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. <lb/> BMC Evolutionary Biology 2008, 8:1 <lb/> http://www.biomedcentral.com/1471-2148/8/1 <lb/></front> <page>Page 2 of 8 <lb/></page> <note place="footnote">(page number not for citation purposes) <lb/></note> <body>producing diploid zygotes that can reproduce asexually by <lb/>mitosis until nutrients are exhausted again [4,5]. <lb/> Saccharomyces sensu stricto species are postzygotically iso-<lb/>lated. Diploid F1 hybrids are formed by fusion of gametes <lb/>from different species. These hybrids can reproduce asex-<lb/>ually by mitosis, but spores produced by meiosis are invi-<lb/>able, failing to germinate and form gametes [6]. Thus F1 <lb/>hybrids are viable but sexually sterile. Several recent inves-<lb/>tigations have examined possible causes of this hybrid ste-<lb/>rility, and concluded that sequence and chromosomal <lb/>differences between the species are major contributors [7-<lb/>9]. Two Saccharomyces species, S. cerevisiae and S. para-<lb/>doxus, have been found to occupy the same natural habitat <lb/>(oak trees and associated soils) [10], providing the oppor-<lb/>tunity for hybridisation. Kuehne et al [11] have recently <lb/>shown that the North American and Eurasian S. paradoxus <lb/> isolates represent two distinct groups. Within each group <lb/>the strains are highly related (indicating a large breeding <lb/>population) and have distributions spanning their respec-<lb/>tive land masses [11]. The population structure of S. cere-<lb/>visiae is not so clear, perhaps because human <lb/>domestication of the species overshadows their natural <lb/>biogeography [12]. Yeast hybrids can be formed in the <lb/>laboratory but wild F1 hybrids, containing a full genome <lb/>from both S. paradoxus and S. cerevisiae, have not been <lb/>described [12]. Several reports have, however, shown <lb/>introgression of genes between the two species, indicating <lb/>that interspecific mating can occur in the wild [12-14]. <lb/>Given that hybrids are sexually sterile, the ability to avoid <lb/>hybridisation may be favoured by natural selection. <lb/>In a recent paper Murphy et al. [15] failed to find prezy-<lb/>gotic reproductive isolation between species from sympat-<lb/>ric natural populations of S. cerevisiae and S. paradoxus. <lb/> Murphy et al. [15] assayed species recognition using indi-<lb/>vidual mate choice trials: a single vegetative haploid cell <lb/>of known mating type was placed in contact with a con-<lb/>specific and a heterospecific vegetative cell of the opposite <lb/>mating type. The results showed that S. cerevisiae cells <lb/>mated with other S. cerevisiae cells more often than they <lb/>mated with S. paradoxus cells, as expected if a prezygotic <lb/>barrier existed. But, surprisingly, Murphy et al. [15] found <lb/>that S. paradoxus cells mated with S. cerevisiae cells (form-<lb/>ing sterile hybrids) more often than they mated with other <lb/> S. paradoxus cells. They explained this result by the obser-<lb/>vation that the mating propensity (the tendency or readi-<lb/>ness to mate) of S. cerevisiae gametes was higher than that <lb/>of S. paradoxus gametes and therefore focal cells of either <lb/>species were more likely to be able to mate with the more <lb/>willing S. cerevisiae gametes, regardless of whether or not <lb/>a hybrid zygote was produced. This difference in mating <lb/>propensity confounded the quantification of prezygotic <lb/>isolation between the species, and Murphy et al. [15] were <lb/>unable to detect prezygotic reproductive isolation. How-<lb/>ever they proposed that differences in mating kinetics <lb/>could potentially provide a prezygotic isolation barrier, <lb/>because fast maters would tend to mate with fast maters, <lb/>and slow maters with slow maters. Such a barrier can <lb/>evolve readily in the laboratory under artificial selection <lb/>against mating between genetically marked strains of the <lb/>same species [16]. <lb/>Although very little is known about yeast life history in <lb/>nature, it is likely that most mating occurs immediately <lb/>after germination, usually between members of the same <lb/>tetrad, and without any haploid cell division [17]. <lb/>Because wild yeast are naturally homothallic (see for <lb/>example [18] and [19]), and unfertilized gametes can <lb/>switch mating type after dividing to enable them to fuse <lb/>with their daughter cells, even isolated single spores <lb/>should yield diploid, not haploid cultures [4]. Murphy et <lb/>al. [15] prevented their strains from switching mating type <lb/>in culture by knocking out their HO genes with drug-<lb/>resistance markers, allowing the culture of clones of vege-<lb/>tative haploid gametes. However such clones of unferti-<lb/>lised gametes are not thought to occur in natural strains <lb/>with intact HO genes. Unfertilised gametes exist only <lb/>rarely and transiently, after spore germination but before <lb/>fertilisation by either a neighbouring germinated spore or <lb/>mating-type switched clone-mate. Thus the potential for <lb/>hybridization in nature is highest when spores from dif-<lb/>ferent species happen to be in contact at the time of ger-<lb/>mination. Such close contact between spores from <lb/>different species might occur if they occupy the same hab-<lb/>itat (e.g. the surface of oak trees), or if they are brought <lb/>together in the digestive tracts of species that eat yeast <lb/>[20]. Yeast-feeding insects, such as Drosophila, completely <lb/>digest vegetative yeast cells, but yeast spores are not <lb/>digested and are passed through the gut unharmed and <lb/>ready to germinate [21]. Therefore prezygotic reproduc-<lb/>tive isolation is likely to involve species differences in ger-<lb/>mination conditions or timing as well as in gamete fusion. <lb/>Here we present the results of mate choice assays using <lb/>wild type homothallic (HO) single spores of five S. cerevi-<lb/>siae and five S. paradoxus isolates from natural popula-<lb/>tions. All pairings were made between strains isolated <lb/>from the same continent (either North America or Eura-<lb/>sia). Strains used in three of the pairings were both iso-<lb/>lated from the same small woodland area and can be <lb/>considered sympatric [18,11] (see Methods for full strain <lb/>details). <lb/> Results <lb/> Both species avoid hybridisation <lb/> Individual mate choice assays were conducted by placing <lb/>two spores of one species (the chooser strain) in contact <lb/>with a single spore from the other species. All spores were <lb/>taken from different tetrads to ensure mating types were <lb/> <note place="headnote">BMC Evolutionary Biology 2008, 8:1 <lb/></note> <note place="headnote">http://www.biomedcentral.com/1471-2148/8/1 <lb/></note> <page>Page 3 of 8 <lb/></page> <note place="footnote">(page number not for citation purposes) <lb/></note> sampled randomly. Since it is impossible to determine the <lb/>mating types of spores before germination, deviations <lb/>from random mating were calculated based on probabil-<lb/>ity (see Methods and Figure 1). Hybrid and non-hybrid <lb/>zygotes were identified using species specific PCR. Mate <lb/>choice trials were carried out using five independent pairs <lb/>of S. cerevisiae and S. paradoxus strains (Table 1). <lb/>Figure 2 shows the proportion of matings that resulted in <lb/>hybrid zygotes for each of the five independent species <lb/>pairs. If mating was random with respect to species, then <lb/>on average 2/3 (66.67%) of the zygotes formed would be <lb/>hybrids (see Methods). Instead, we found that for every <lb/>pair, whether S. cerevisiae or S. paradoxus occupied the role <lb/>of chooser, significantly fewer hybrid zygotes were formed <lb/>than expected by random mating (full data in Figure 2 and <lb/>Table 2). <lb/> Both species in each pair had similar mating propensities <lb/> Many trials did not result in a zygote, either because all <lb/>three spores in a trial were the same mating type (this will <lb/>occur on average in 25% of trials – see Figure 1) or <lb/>because mating does not always occur even when trials do <lb/>contain gametes of both mating types. Differences in mat-<lb/>ing propensity between the species in each pair would <lb/>mean that a higher proportion of trials would result in <lb/>zygotes when the species with a high mating propensity <lb/>was "chooser" than when the species with a low mating <lb/>propensity was chooser. For each pair, we found no signif-<lb/>icant difference between the number of zygotes formed <lb/>when either strain was chooser (Table 2). <lb/> S. cerevisiae is choosier than S. paradoxus <lb/> Do spores always choose to mate with a member of their <lb/>own species, if available? If hybrids only form when there <lb/>is no mate of the same species available, then on average <lb/>only 1/3 of zygotes formed would be expected to be <lb/>hybrids (Figure 1). We tested whether there was significa-<lb/>tion deviation from this expectation (Table 2). Only one <lb/>of the five S. cerevisiae strains (Sc1) produced a signifi-<lb/>cantly higher proportion of hybrid zygotes than the 1/3 <lb/>expected if a S. cerevisiae strain always chose to mate with <lb/>a member of its own species. In contrast four of the five S. <lb/>paradoxus strains (Sp1, Sp2, Sp3 and Sp4) produced a sig-<lb/>nificantly higher number of hybrid zygote than the 1/3 <lb/>predicted under the "always choose own species" model. <lb/> Mate choice pairing possibilities <lb/> Figure 1 <lb/>Mate choice pairing possibilities. The eight possible <lb/>combinations when a spore of one species is placed against <lb/>two spores from another species. Grey and white circles <lb/>indicate species, arrows indicate cells with a choice of mates, <lb/>and a and α indicate the mating types of the spores. <lb/> a <lb/> a <lb/> α <lb/> α <lb/> a <lb/> α <lb/> a <lb/> α <lb/> a <lb/>a <lb/> α <lb/>α <lb/> a <lb/>a <lb/> α <lb/>α <lb/>α <lb/> a <lb/> α <lb/>α <lb/>α <lb/> a <lb/>a <lb/>a <lb/> Uninformative <lb/> No Mating <lb/>Informative <lb/> Table 1: Strain table <lb/> Pairing <lb/>Known as <lb/>Strain <lb/>Species <lb/>Isolated From <lb/>Location <lb/>Ref. <lb/>1 <lb/> Sc1 <lb/> Y55 <lb/> S. cerevisiae <lb/> Wine Grape <lb/>France a <lb/> [14] <lb/> Sp1 <lb/> N-17 <lb/> S. paradoxus <lb/> Oak tree <lb/>Tatarstan, Russia b <lb/> [29] <lb/>2 <lb/> Sc2 <lb/> SK1 <lb/> S. cerevisiae <lb/> Soil (Lab strain) <lb/>USA (exact location unknown) c <lb/> [14] <lb/> Sp2 <lb/> YPS138 <lb/> S. paradoxus <lb/> Soil beneath oak tree <lb/>Tyler Arboretum Media, PA, USA d <lb/> [30] <lb/>3 <lb/> Sc3 <lb/> YPS128 <lb/> S. cerevisiae <lb/> Soil beneath oak tree <lb/>Tyler Arboretum Media, PA, USA d <lb/> [30] <lb/> Sp3 <lb/> YPS145 <lb/> S. paradoxus <lb/> Soil beneath oak tree <lb/>Tyler Arboretum Media, PA, USA d <lb/> [30] <lb/>4 <lb/> Sc4 <lb/> YPS681 <lb/> S. cerevisiae <lb/> Oak tree <lb/>Buck Hill Falls, PA, USA e <lb/> [15] <lb/> Sp4 <lb/> YPS664 <lb/> S. paradoxus <lb/> Oak tree <lb/>Buck Hill Falls, PA, USA e <lb/> [15] <lb/>5 <lb/> Sc5 <lb/> YPS670 <lb/> S. cerevisiae <lb/> Oak tree <lb/>Buck Hill Falls, PA, USA e <lb/> [15] <lb/> Sp5 <lb/> YPS646 <lb/> S. paradoxus <lb/> Oak tree <lb/>Buck Hill Falls, PA, USA e <lb/> [15] <lb/>Table showing the strains used in each of the 5 species pairs. <lb/>For the purpose of our analysis, we considered pairs 1 and 2 to be allopatric and pairs 3, 4, and 5 to be sympatric. <lb/> a – Exact time of isolation unknown but believed to be between 1930 and 1960. <lb/> b – Exact time of isolation unknown – first published reference is from 1988 [31] <lb/> c – Exact time of isolation unknown – first published reference from 1974 [32] <lb/> d – Isolated in July 1999 <lb/> e – Isolated in 2000 <lb/> <note place="headnote">BMC Evolutionary Biology 2008, 8:1 <lb/></note> <note place="headnote">http://www.biomedcentral.com/1471-2148/8/1 <lb/></note> <page>Page 4 of 8 <lb/></page> <note place="footnote">(page number not for citation purposes) <lb/></note> W <lb/> e also noted that, in all five pairs, fewer hybrid zygotes <lb/>were formed when S. cerevisiae was chooser than when S. <lb/>paradoxus was chooser. This suggested prezygotic isolation <lb/>was stronger when S. cerevisiae was the chooser than when <lb/> S. paradoxus was chooser. <lb/>We wanted to determine the effect of two factors on the <lb/>strength of prezygotic reproductive isolation: the species <lb/>choosing, and whether the pair of species were isolated in <lb/>allopatry or sympatry (see Table 1). To do this we per-<lb/>formed a joint analysis by means of a Generalised Linear <lb/>Model (GLM). To accommodate the binary structure of <lb/>the data (hybrid mating vs. non-hybrid mating) we car-<lb/>ried out a binomial GLM with logit link function. We <lb/>included the factors 'species' (S. cerevisiae or S. paradoxus), <lb/> 'locality' (sympatric or allopatric) and their interaction. <lb/>The analysis was performed using the statistical package R <lb/>(version 2.4.1, R Development Core Team 2007) [22]. <lb/>The GLM showed that across sympatric and allopatric <lb/>matings, species differ significantly in the level of pre-<lb/>zygotic isolation ('species' term, p = < 0.0001). Thus, S. <lb/>cerevisiae is choosier than S. paradoxus. In addition, the <lb/>data showed that across species, prezygotic isolation <lb/>tended to be stronger in allopatric as compared to sympat-<lb/>ric matings. However, this effect was not statistically sig-<lb/>nificant ('locality' term, p = 0.07). Finally, the analysis <lb/>showed that both species were similar in the change in <lb/>discrimination between sympatric and allopatric individ-<lb/>uals of the other species (interaction term, p = 0.93). <lb/> Table 2: Mate choice information table <lb/> Pair <lb/> "chooser" <lb/>strain <lb/>Total <lb/>trials as <lb/>"chooser" <lb/>Total <lb/>zygotes <lb/>formed <lb/>Observed <lb/>number of <lb/>hybrid <lb/>zygotes <lb/>Expected <lb/>number of <lb/>hybrid <lb/>zygotes a <lb/> Hybrid zygote <lb/> χ 2 (p value) b <lb/> Always own <lb/>species χ 2 (p <lb/>value) c <lb/> Mating <lb/>propensity d <lb/> Mating <lb/>Prospensity χ 2 <lb/> (p value) <lb/>1 <lb/> Sc1 <lb/> 300 <lb/>115 <lb/>49 <lb/>76.67 <lb/>28.89 (<0.001) 4.882943 (0.0271) <lb/>51.11 <lb/>1.035 (0.3089) <lb/> Sp1 <lb/> 206 <lb/>69 <lb/>34 <lb/>46 <lb/>8.63 (0.0033) <lb/>8.628882 (0.0033) <lb/>44.66 <lb/>2 <lb/> Sc2 <lb/> 270 <lb/>88 <lb/>28 <lb/>58.67 <lb/>31.82 (<0.001) 0.035264 (0.8510) <lb/>43.46 <lb/>2.212 (0.1369) <lb/> Sp2 <lb/> 220 <lb/>85 <lb/>47 <lb/>56.67 <lb/>4.45 (0.0349) <lb/>19.4555 (<0.0001) <lb/>51.51 <lb/>3 <lb/> Sc3 <lb/> 250 <lb/>70 <lb/>22 <lb/>46.67 <lb/>31.42 (<0.001) 0.044643 (0.8327) <lb/>37.33 <lb/>0.164 (0.6857) <lb/> Sp3 <lb/> 344 <lb/>90 <lb/>42 <lb/>60 <lb/>15.31 (<0.001) <lb/>7.8125 (0.0052) <lb/>34.88 <lb/>4 <lb/> Sc4 <lb/> 210 <lb/>78 <lb/>25 <lb/>52 <lb/>40.51 (<0.001) 0.014423 (0.9044) <lb/>49.52 <lb/>0.0046 (0.9462) <lb/> Sp4 <lb/> 220 <lb/>80 <lb/>38 <lb/>53.33 <lb/>12.38 (<0.001) <lb/>7.876563 (0.005) <lb/>48.48 <lb/>5 <lb/> Sc5 <lb/> 240 <lb/>90 <lb/>26 <lb/>60 <lb/>56.13 (<0.001) 0.877974 (0.3488) <lb/>50.00 <lb/>2.229 (0.1355) <lb/> Sp5 <lb/> 260 <lb/>80 <lb/>33 <lb/>53.33 <lb/>22.13 (<0.001) 2.626563 (0.1051) <lb/>41.03 <lb/>Table showing the raw data obtained from the mate choice trials as well as the hybrid zygote and mating propensity χ 2 values. <lb/> a – The expected number of hybrid zygotes if no mating preference exists between the two species (2/3 total – see Methods) <lb/> b – χ 2 test on the number of hybrid zygotes observed compared to the number expected if there is no mate preference (Yates corrected) (see also <lb/>Figure 2). <lb/> c – χ 2 test on the number of hybrid zygotes observed compared to the 1/3 expected under the "always choose own species" model. (see Figures 1 <lb/>and Methods) <lb/> d – Percentage of mating trials producing zygotes after correcting for the 25% of trials that cannot form zygotes because all three spores are the <lb/>same mating type (see Methods).). <lb/> e – χ 2 test for the effect of species in pair on mating propensity (Yates corrected). <lb/> Graphs of % hybrid matings for each pairing <lb/> Figure 2 <lb/>Graphs of % hybrid matings for each pairing. Bar chart <lb/>showing the percentage of matings that resulted in hybrid <lb/>zygotes for the five species pairs. For each pair, the light grey <lb/>bar represents the result when S. cerevisiae (Sc) chose and the <lb/>dark grey bar represents the result when S. paradoxus (Sp) <lb/>chose. On each bar the numbers are the number of hybrid <lb/>zygotes formed over the total number analysed. Dashed lines <lb/>indicate the proportions of hybrids that would be expected if <lb/>the chooser always mated with its own species (33.33%), had <lb/>no preference (66.67%) and always mated with the other spe-<lb/>cies (100%). All strains formed significantly fewer hybrids than <lb/>would be expected by chance (*** = p < 0.001, ** = p < 0.01. <lb/>* = p < 0.05). (For full dataset see Table 2.) <lb/> % <lb/> H <lb/>Y <lb/>B <lb/>R <lb/>I <lb/>D <lb/>S <lb/> 50 <lb/> 100 <lb/> Always <lb/>choose other <lb/>species <lb/>No <lb/>Preference <lb/>Always <lb/>choose own <lb/>species <lb/> Sc1 <lb/> 49 <lb/>115 <lb/> Sc2 <lb/> 28 <lb/>88 <lb/> Sp4 <lb/> 38 <lb/>80 <lb/> Sc4 <lb/> 25 <lb/>78 <lb/> Sp3 <lb/> 42 <lb/>90 <lb/> Sc3 <lb/> 22 <lb/>70 <lb/> Sp2 <lb/> 47 <lb/>85 <lb/> Sp5 <lb/> 33 <lb/>80 <lb/> Sc5 <lb/> 26 <lb/>90 <lb/> 1 <lb/> 5 <lb/>4 <lb/>3 <lb/>2 <lb/> Sp1 <lb/> 34 <lb/>69 <lb/> 100 <lb/>50 <lb/> *** <lb/>*** <lb/>*** <lb/>*** <lb/>*** <lb/>*** <lb/>** <lb/>* <lb/>*** <lb/>*** <lb/> <note place="headnote">BMC Evolutionary Biology 2008, 8:1 <lb/></note> <note place="headnote">http://www.biomedcentral.com/1471-2148/8/1 <lb/></note> <page>Page 5 of 8 <lb/></page> <note place="footnote">(page number not for citation purposes) <lb/></note> Discussion <lb/> Closely related Saccharomyces species show prezygotic <lb/>reproductive isolation <lb/> We have demonstrated that the closely related species S. <lb/> cerevisiae and S. paradoxus show lower levels of interspecific <lb/>hybridisation than would be expected if mating between <lb/>the two were random with respect to species. To our knowl-<lb/>edge this is the first time that prezygotic reproductive isola-<lb/>tion has been demonstrated between Saccharomyces yeast <lb/>species. A previous report by Murphy et al [15] using simi-<lb/>lar mate choice trials showed that when S. cerevisiae occu-<lb/>pied the role of chooser interspecific mating was lower than <lb/>expected by random mating. However Murphy et al [15] <lb/>did not find prezygotic isolation because the result was <lb/>reversed when S. paradoxus was chooser – when offered a <lb/>choice S. paradoxus preferentially mated with S. cerevisiae <lb/> individuals, producing more hybrids than expected by <lb/>chance. This effect was caused by differences in mating pro-<lb/>pensity – S. cerevisiae haploid vegetative cells were both <lb/>faster maters and more likely to mate than S. paradoxus <lb/> cells. Thus mating tended to occur with the partner that was <lb/>more willing to mate, regardless of its species. This meant <lb/>that any prezygotic isolation was undetectable because it <lb/>was obscured by the large difference in mating propensity. <lb/>The protocol of Murphy et al [15] used strains that had <lb/>been genetically modified to grow as stable clones of veg-<lb/>etative gametes. This meant that only the fusion element <lb/>of the mating process was tested. In nature, yeast gametes <lb/>probably only exist immediately after spore germination, <lb/>so we used spores, rather than vegetatively-growing gam-<lb/>etes, in our assay, so that the whole mating process could <lb/>be tested. Murphy et al [15] kindly provided us with sam-<lb/>ples of the ancestors of their strains, and in our assay we <lb/>were able to detect significant prezygotic reproductive iso-<lb/>lation between them (pairings 4 and 5), as well as <lb/>between species in the other three pairs tested (pairings 1, <lb/>2, and 3). Our results can only be explained by the pres-<lb/>ence of prezygotic reproductive isolation. <lb/>Nevertheless we were still interested in whether the <lb/>strength (but not the direction) of the measured prezygotic <lb/>isolation was confounded by differences in mating propen-<lb/>sity. Unlike Murphy et al [15], we detected no differences <lb/>in mating propensity between the strains in each pair, but <lb/>we had limited power to detect small differences. We did, <lb/>however, determine that hybrids were formed less often <lb/>when S. cerevisiae was chooser than when S. paradoxus was <lb/>chooser. This observation could be caused by S. cerevisiae <lb/> having stronger preference for members of its own species <lb/>than S. paradoxus., i.e. S. cerevisiae is choosier. Mating with <lb/>the wrong species is equally bad for either species, so it is <lb/>not immediately obvious how such a difference in choosi-<lb/>ness might evolve. One possibility is that the barrier <lb/>evolved in S. cerevisiae to prevent mating with another yeast <lb/>species more frequently encountered than S. paradoxus (the <lb/>barrier not providing complete prezygotic reproductive iso-<lb/>lation between the two species studied here). However, <lb/>another explanation, consistent with Murphy et al. [15], is <lb/>that the difference is caused by a greater mating propensity <lb/>in S. cerevisiae than in S. paradoxus, perhaps reflecting some <lb/>difference in the evolution of these two species. <lb/>How strong is prezygotic reproductive isolation in yeast? <lb/>Whilst hybrids form readily when no member of the same <lb/>species is present, our results show that many strains <lb/>(especially S. cerevisiae) have near perfect discrimination <lb/>when given choice of species, and hybrids only occur in <lb/>the 33% of matings in which no choice is available (see <lb/>Fig 1 and Results). Clearly, this ability could be important <lb/>for yeast in its natural environment. It is impossible, given <lb/>our current lack of knowledge of wild yeast ecology, to say <lb/>how often the situation of only three spores being in iso-<lb/>lated contact (as in our mate trials) occurs in its natural <lb/>environment. Due to the nature of the yeast tetrad it is <lb/>highly likely that a gamete will usually find itself in close <lb/>proximity to another gamete originating from the same <lb/>tetrad. But the digestion of yeast tetrads by insect vectors <lb/>releases spores from their tetrads, increasing inter-tetrad <lb/>mating [21]. Insects that feed on different yeast species are <lb/>therefore likely to increase the possibility of hybridisation <lb/>between different yeast species [20]. These mate choice tri-<lb/>als have allowed us to demonstrate that mating behaviour <lb/>can reduce hybridisation in Saccharomyces species. <lb/>One caveat that must be noted is that we have only looked <lb/>into the level of hybridisation between S. cerevisiae and S. <lb/>paradoxus isolates but not between strains of the same spe-<lb/>cies. It is possible that variation in the traits that result in <lb/>the prezygotic isolation between species may also result in <lb/>varying levels of hybridisation between different isolates <lb/>of the same species. Within species variation would be an <lb/>interesting avenue for further investigation, especially <lb/>between genetically distinct strains such as the geographi-<lb/>cally isolated S. paradoxus "groups" identified by Kuehne <lb/>et al [11] and Koufopanou et al [23]. <lb/> Did prezygotic reproductive isolation evolve by direct <lb/>selection? <lb/> Though we have demonstrated that these two species <lb/>show reduced levels of hybridisation we are not able to <lb/>say how or why it evolved. There are two possibilities. The <lb/>first is that natural selection has acted directly to reduce <lb/>the costly formation of sterile hybrids. The second is that <lb/>the prezygotic isolation is an indirect consequence of evo-<lb/>lution, whether by selection on another trait, or by genetic <lb/>drift, that happens to result in reduced hybridisation. A <lb/>laboratory study has already shown that prezygotic isola-<lb/>tion can evolve quickly between Saccharomyces in direct <lb/>response to selection against hybrids [16]. We also know <lb/> <note place="headnote">BMC Evolutionary Biology 2008, 8:1 <lb/></note> <note place="headnote">http://www.biomedcentral.com/1471-2148/8/1 <lb/></note> <page>Page 6 of 8 <lb/></page> <note place="footnote">(page number not for citation purposes) <lb/></note> that the two species are likely to encounter each other in <lb/>nature: strains of these two species can be isolated from <lb/>the same trees [10,18] and the two species exhibit very <lb/>similar phenotypic profiles [24]. <lb/>One way to address whether prezygotic isolation evolved <lb/>by direct selection is to compare the strength of isolation in <lb/>sympatric and allopatric species pairs. Sympatric pairs are <lb/>likely to interact more frequently than allopatric pairs, so if <lb/>prezygotic reproductive isolation evolved to reinforce the <lb/>species barrier it should be stronger in sympatric pairs than <lb/>in allopatric pairs. Coyne and Orr have shown this to be <lb/>true for a large number of Drosophila species pairs [[25,26], <lb/>for review see [3]]. In our assay we found that prezygotic <lb/>reproductive isolation was stronger in the three sympatric <lb/>species pairs than in the two allopatric species pairs; how-<lb/>ever the difference was not statistically significant. But with <lb/>only five pairs we have little statistical power and we antic-<lb/>ipate that a larger study may well find a strong effect of <lb/>sympatry on prezygotic reproductive isolation. Further, we <lb/>note that the members of the two pairs we designated as <lb/>allopatric do share the same continent, and recent work <lb/>has shown that yeast populations inhabiting the same con-<lb/>tinent recombine freely [11,23]. Dispersal between conti-<lb/>nents appears to be very low, so it would be very interesting <lb/>to measure prezygotic reproductive isolation (if any) <lb/>between yeast species isolated from different continents. <lb/> Possible prezygotic reproductive isolation barriers in yeast <lb/> There has been considerable debate over whether postzy-<lb/>gotic or prezygotic reproductive isolation barriers are most <lb/>important in maintaining distinct species [3]. As prezygotic <lb/>barriers act earlier the general consensus is that they repre-<lb/>sent a stronger barrier to gene flow between populations. If <lb/>no mating actually occurs in the first place genes cannot be <lb/>exchanged between species [3]. In nature species pairs often <lb/>exhibit both prezygotic and postzygotic barriers which over <lb/>time have accumulated to reproductively isolate the popu-<lb/>lations. It is however extremely hard to determine which <lb/>barrier evolved first to start the speciation process [3]. <lb/>The Saccharomyces sensu stricto yeasts are known to be <lb/>strongly postzygotically isolated from one another [7-9]. <lb/>The prezygotic reproductive isolation identified here <lb/>could represent a "work in progress", with the two species <lb/>still undergoing selection and evolution towards less <lb/>"leaky" barriers to hybridisation. <lb/>Several parts of the yeast life cycle can possibly act as <lb/>prezygotic barriers to stop, or reduce, hybridisation. One <lb/>possible barrier that is analogous to mate selection mech-<lb/>anisms in many higher organisms is pheromone recogni-<lb/>tion [3]. If each Saccharomyces species is able to identify <lb/>members of its own species by the pheromone it pro-<lb/>duces, avoiding hybridisation may be possible. This how-<lb/>ever seems unlikely to be the case as the peptide sequence <lb/>of the pheromone is conserved across the Saccharomyces <lb/> sensu stricto species [27]. Differences in mating kinetics <lb/>have been previously highlighted as a possible premating <lb/>barrier between Saccharomyces species. Leu et al [16] <lb/>showed experimentally that S. cerevisiae could evolve to <lb/>avoid mating with strains that would generate lethal com-<lb/>binations of genetic markers in the progeny. The reduc-<lb/>tion in harmful matings evolved because of changes in the <lb/>speed of mating, with faster maters mating with other <lb/>compatible fast maters, slow maters mating with other <lb/>compatible slow maters, thus reducing harmful mating. <lb/>Murphy et al [15] proposed that a similar mechanism <lb/>might work for wild yeast. They found that S. cerevisiae <lb/> strains had a higher mating propensity than the S. para-<lb/>doxus strains used in their experiments. They postulated <lb/>that as S. cerevisiae was more willing to mate and did so <lb/>quicker it would be possible that in certain natural situa-<lb/>tions all the fast and willing maters of one species could <lb/>mate together and all the slow and unwilling maters could <lb/>mate together, reducing the rate of hybrid formation. <lb/>The spores that we used, unlike vegetative haploid cells, <lb/>are dormant and do not produce pheromone making <lb/>them effectively invisible to gametes actively seeking a <lb/>mate. Vegetative yeast gametes up regulate the production <lb/>of pheromone early in the mating response. As yeast select <lb/>partners on the strength of the pheromone signal they <lb/>produce, S. paradoxus is more likely to choose the display-<lb/>ing S. cerevisiae as a mate (which being faster up regulates <lb/>pheromone quicker) [5]. This explains the asymmetric <lb/>mating preference observed by Murphy et al [15] in their <lb/>vegetative cell assays. When spores are used, if one species <lb/>germinates quickly and begins to mate before the other <lb/>species has germinated hybridisation can be reduced. For <lb/>example if S. cerevisiae is a faster germinator than S. para-<lb/>doxus in a S. paradoxus "choosing" mate choice trial the S. <lb/>cerevisiae spore germinates first and will not sense any <lb/>pheromone. The now metabolically active S. cerevisiae cell <lb/>will enter the cell cycle and begin vegetative asexual <lb/>growth, preventing it from mating until it has divided (~2 <lb/>h). If, during this time, the S. paradoxus cells germinate, <lb/>they will be more likely to mate with their own species. As <lb/>the majority of wild yeast are homothallic this may be par-<lb/>ticularly important because an unmated fast-germinating <lb/>individual is also likely to undergo mating-type switching <lb/>and subsequently mate with its own daughter cell, elimi-<lb/>nating it from the pool of potential mates. We propose, <lb/>therefore, that spore germination is an important prezy-<lb/>gotic barrier that has previously been overlooked in yeast. <lb/>Due to the length of time speciation takes to evolve <lb/>between species it is likely that, through selection, many <lb/>isolating barriers can evolve simultaneously [3]. Single <lb/>isolating barriers can be "leaky" allowing for some hybrid-<lb/> <note place="headnote">BMC Evolutionary Biology 2008, 8:1 <lb/></note> <note place="headnote"> http://www.biomedcentral.com/1471-2148/8/1<lb/></note> <page>Page 7 of 8 <lb/></page> <note place="footnote">(page number not for citation purposes) <lb/></note> isation between populations, whereas multiple prezygotic <lb/>barriers working in unison can act to reduce hybridisation <lb/>[3]. A combination of isolating barriers may be involved <lb/>in the observed prezygotic reproductive isolation between <lb/> Saccharomyces species. By developing the methods out-<lb/>lined here as well as those previously reported by Murphy <lb/>et al [15] it may be possible to tease apart the relative <lb/>importance of all parts of the Saccharomyces life cycle as <lb/>they affect reproductive isolation. <lb/> Conclusion <lb/> We have shown for the first time that the two closely <lb/>related (and often sympatric) yeast species S. cerevisiae <lb/> and S. paradoxus are prezygotically reproductively iso-<lb/>lated. Because hybrids are sexually sterile, the ability to <lb/>mate with the correct species could be an adaptation. <lb/>Our results contrast with those of an earlier study [15] that <lb/>failed to detect prezygotic isolation between vegetative <lb/>gametes from different species. Mating in wild yeast is <lb/>most likely to take place between newly germinated <lb/>spores, so differences in germination may allow the evo-<lb/>lution of prezygotic isolation barriers. With greater knowl-<lb/>edge of yeast life history, in particular exactly when in the <lb/>life cycle mating typically occurs, we will be able to more <lb/>fully understand the contribution of differences in germi-<lb/>nation and mating kinetics to reproductive isolation in <lb/>the wild. Clearly it is desirable that further experiments be <lb/>carried out to expand our knowledge of this important <lb/>model organism in its natural environment. <lb/> Methods <lb/> Strains, media and growth conditions <lb/> A list of the strains used in this work can be found in Table <lb/>1. All strains were wild-type and unmodified from their <lb/>original ancestral type. Strains were sporulated on potas-<lb/>sium acetate plates (2% potassium acetate, 0.22% yeast <lb/>extract, 0.05% glucose, 0.087% complete amino acid mix, <lb/>2.5% agar) for 4 days at 25°C. Mate choice assays were <lb/>conducted on YEPD plates (1% yeast extract, 2% bac-<lb/>topeptone, 2% glucose, 2.5% agar) at 30°C. <lb/> Mate choice assays <lb/> The ascus containing the yeast spores was digested using a <lb/>standard zymolyase protocol [28]. Using a tetrad dissec-<lb/>tion microscope, two spores of the chooser species were <lb/>taken from separate tetrads and placed in contact with each <lb/>other and a third spore of the other species on a YEPD plate <lb/>(see Fig. 1). This allowed the chooser strain the possibility <lb/>of mating with a member of its own species or another spe-<lb/>cies. All spores used came from different tetrads. <lb/>After incubation, unmated individuals were removed and <lb/>zygotes left in place. If the removed individual was found <lb/>to have been inviable the test was ignored because no <lb/>choice of mate was possible. Two forms of triad were used: <lb/>one in which the S. cerevisiae strain was chooser and the <lb/>other with S. paradoxus as chooser. This allowed the level <lb/>of reproductive isolation of both species to be investigated. <lb/>8 mating type combinations are possible when two spores <lb/>from one species are paired with one from another (Fig. <lb/>1). Of these 8 possibilities, 4 can provide information on <lb/>mate choice between Saccharomyces species (dashed box). <lb/>These four contain both an a and α spore from the chooser <lb/>species and a spore of the other species of either mating <lb/>type. When this occurs a chooser strain cell has a choice of <lb/>mate. The remaining possible combinations provide no <lb/>useful information. Two (solid box) represent triads where <lb/>mating occurs but there is no possibility for mate choice <lb/>between species. Only hybrids can be formed by mating <lb/>within such triads. The remaining two possible combina-<lb/>tions (dotted box) are those where all individuals within a <lb/>triad are of the same mating type and unable to mate. <lb/>If there is no preferential mating between the species, 2/3 <lb/>of the zygotes will be hybrid. This is because 1/3 of all <lb/>matings will always produce hybrids (solid box, Fig. 1) <lb/>and if there is no preference 1/2 of the informative triads <lb/>(2/3 – dashed box, Fig. 1) will produce hybrid zygotes (1/ <lb/>3 + (1/2 × 2/3) = 2/3). If, when offered choice, a spore <lb/>always mates with its own species 1/3 of zygotes will be <lb/>hybrid. Alternatively, if when presented with a choice of <lb/>mate a cell always chooses to mate with a member of <lb/>another species, all zygotes will be hybrid. Prezygotic iso-<lb/>lation can therefore be measured by deviation from the <lb/>proportion of interspecific matings expected if no prefer-<lb/>ence exists (2/3). <lb/> DNA extraction and PCR identification of hybrids <lb/> Matings are identified as being interspecific or intraspe-<lb/>cific by species specific PCR. DNA was extracted from the <lb/>colonies formed by each mating using a glass bead <lb/>method [28]. <lb/>Species specific primers were designed that only produce <lb/>an amplicon if the genome of a particular species is <lb/>present as a template. The primers were designed by align-<lb/>ing the reference genomes for the two species using the <lb/>fungal alignment viewer provided by the Saccharomyces <lb/>Genome Database [27]. Different amplicon lengths were <lb/>used for each species. S. cerevisiae primers produced an <lb/>amplicon of approximately 500 bp whilst those for S. par-<lb/>adoxus amplified approximately 300 bp of DNA. This sim-<lb/>ple design feature allowed for the quick and easy <lb/>determination of hybrids and allowed us to control for <lb/>the accidental use of an incorrect primer by contamina-<lb/>tion. Two sets of species specific PCR primers that amplify <lb/>a different region of the particular species genome were <lb/>used for each zygote to allow confirmation of results. <lb/>Some species specific primers did not work with some <lb/>strains of the correct species, presumably because the <lb/>sequence at a primer site was polymorphic. New primers <lb/>were designed in these cases. See Additional file 1: PCR <lb/> <note place="footnote">Publish with Bio Med Central and every <lb/>scientist can read your work free of charge <lb/> "BioMed Central will be the most significant development for <lb/>disseminating the results of biomedical researc h in our lifetime." <lb/> Sir Paul Nurse, Cancer Research UK <lb/> Your research papers will be: <lb/> available free of charge to the entire biomedical community <lb/>peer reviewed and published immediately upon acceptance <lb/>cited in PubMed and archived on PubMed Central <lb/>yours — you keep the copyright <lb/> Submit your manuscript here: <lb/> http://www.biomedcentral.com/info/publishing_adv.asp <lb/> BioMedcentral <lb/></note> <note place="headnote">BMC Evolutionary Biology 2008, 8:1 <lb/></note> <note place="headnote">http://www.biomedcentral.com/1471-2148/8/1 <lb/></note> <page>Page 8 of 8 <lb/></page> <note place="footnote">(page number not for citation purposes) <lb/></note> primer sequences for the sequences of the primers and the <lb/>strain pairs with which they were used.<lb/> </body> <back> <div type="annex">Authors' contributions <lb/> CJM developed the assay, carried out the experiment and <lb/>analysed the data. DG designed and supervised the exper-<lb/>iment. Both authors wrote the paper. <lb/> Additional material <lb/></div> <div type="acknowledgement">Acknowledgements <lb/> We're indebted to Max Reuter, who performed the GLM analysis. We <lb/>would like to thank Greg Hurst, Dave Rogers, Carl Smith and two anony-<lb/>mous reviewers for their comments on the work and help with the manu-<lb/>script. We thank H. Murphy and P. Sniegowski for discussing their work and <lb/>providing us with their strains. The work was supported by a Royal Society <lb/>Research Fellowship to DG and a NERC PhD studentship to CJM. <lb/></div> <listBibl>References <lb/> 1. <lb/>Mayr E: Animal Species and Evolution Cambridge: Harvard University <lb/>Press; 1963. <lb/>2. <lb/>Dobzhansky T: Genetics and the Origin of Species New York: Columbia <lb/>University Press; 1951. <lb/>3. <lb/>Coyne JA, Orr HA: Speciation Sunderland: Sinauer Associates; 2004. <lb/>4. <lb/>Herskowitz I: Life cycle of the Budding Yeast Saccharomyces <lb/> cerevisiae. Microbiol Rev 1988, 52:536-553. <lb/> 5. <lb/>Jackson CL, Hartwell LH: Courtship in Saccharomyces cerevisiae <lb/> – both cell-types choose mating partners by responding to <lb/>the strongest pheromone signal. Cell 1990, 63:1039-1051. <lb/> 6. <lb/>Naumov GI, James SA, Naumova ES, Louis EJ, Roberts IN: Three <lb/>new species in the Saccharomyces sensu strictocomplex: Sac-<lb/>charomyces cariocanus, Saccharomyces kudriavzevii and Sac-<lb/>charomyces mikatae. Int J Syst Evol Microbiol 2000, 50:1931-1942. <lb/> 7. <lb/>Chambers SR, Hunter N, Louis EJ, Borts RH: The mismatch repair <lb/>system reduces meiotic homeologous recombination and <lb/>stimulates recombination-dependent chromosome loss. Mol <lb/>Cell Biol 1996, 16:6110-6120. <lb/> 8. <lb/>Delneri D, Colson I, Grammenoudi S, Roberts IN, Louis EJ, Oliver SG: <lb/> Engineering evolution to study speciation in yeasts. Nature <lb/> 2003, 422:68-72. <lb/> 9. <lb/>Hunter N, Chambers SR, Louis EJ, Borts RH: The mismatch repair <lb/>system contributes to meiotic sterility in an interspecific <lb/>yeast hybrid. EMBO J 1996, 15:1726-1733. <lb/> 10. Naumov GI, Naumova ES, Sniegowski PD: Saccharomyces para-<lb/>doxus and Saccharomyces cerevisiae are associated with exu-<lb/>dates of North American oaks. <lb/> Can J Microbiol 1998, <lb/> 44:1045-1050. <lb/> 11. Kuehne HA, Murphy HA, Francis CA, Sniegowski PD: Allopatric <lb/>divergence, secondary contact, and genetic isolation in wild <lb/>yeast populations. Curr Biol 2007, 17:407-411. <lb/> 12. Liti G, Barton DB, Louis EJ: Sequence diversity, reproductive <lb/>isolation and species concepts in Saccharomyces. Genetics <lb/> 2006, 174:839-850. <lb/> 13. Barros Lopes M, Bellon JR, Shirley NJ, Ganter PF: Evidence for mul-<lb/>tiple interspecific hybridization in Saccharomyces sensu <lb/>stricto species. Fems Yeast Res 2002, 1:323-331. <lb/> 14. Liti G, Peruffo A, James SA, Roberts IN, Louis EJ: Inferences of evo-<lb/>lutionary relationships from a population survey of LTR-ret-<lb/> rotransposons and telomeric-associated sequences in the <lb/> Saccharomyces sensu stricto complex. Yeast 2007, 22:177-192. <lb/> 15. Murphy H, Kuehne H, Francis C, Sniegowski P: Mate choice assays <lb/>and mating propensity differences in natural yeast popula-<lb/>tions. Biol Lett 2006, 2:553-556. <lb/> 16. Leu JY, Murray AW: Experimental evolution of mating discrim-<lb/>ination in budding yeast. Curr Biol 2006, 16:280-286. <lb/> 17. Knop M: Evolution of the hemiascomyclete yeasts: on life <lb/>styles and the importance of inbreeding. Bioessays 2006, <lb/> 28:696-708. <lb/> 18. Sniegowski PD, Dombrowski PG, Fingerman E: Saccharomyces cer-<lb/>evisiae and Saccharomyces paradoxus coexist in a natural <lb/>woodland site in North America and display different levels <lb/>of reproductive isolation from European conspecifics. Fems <lb/>Yeast Res 2002, 1:299-306. <lb/> 19. Johnson LJ, Koufopanou V, Goddard MR, Hetherington R, Schafer SM, <lb/>Burt A: Population Genetics of the Wild Yeast Saccharomy-<lb/>ces paradoxus. Genetics 2004, 166:43-52. <lb/> 20. Pulvirenti A, Zambonelli C, Todaro A, Giudici P: Interspecific <lb/>hybridisation by digestive tract of invertebrates as a source <lb/>of environmental biodiversity within the Saccharomyces cer-<lb/>evisiae. Ann Microbiol 2002, 52:245-255. <lb/> 21. Reuter M, Bell G, Greig D: Increased outbreeding in yeast in <lb/>response to dispersal by an insect vector. Curr Biol 2007, <lb/> 17:R81-R83. <lb/> 22. R Development Core Team: R: A Language and Environment <lb/>for Statistical Computing. R Foundation for Statistical Com-<lb/>puting. Vienna, Austria 2007. <lb/>23. Koufopanou V, Hughes J, Bell G, Burt A: The spatial scale of <lb/>genetic differentiation in a model organism: the wild yeast <lb/> Saccharomyces paradoxus. Phil Trans R Soc Lond B 2006, <lb/> 361:1941-1946. <lb/> 24. Naumov GI, James SA, Naumova ES, Louis EJ, Roberts IN: Three <lb/>new species in the Saccharomyces sensu stricto complex: Sac-<lb/>charomyces cariocanus, Saccharomyces kudriavzevii and Sac-<lb/>charomyces mikatae. Int J Syst Evol Microbiol 2000, 50:1931-1942. <lb/> 25. Coyne JA, Orr HA: Patterns of speciation in Drosophila. Evolu-<lb/>tion 1989, 43:362-381. <lb/> 26. Coyne JA, Orr HA: "Patterns of speciation in Drosophila" revis-<lb/>ited. Evolution 1997, 51:295-303. <lb/> 27. Saccharomyces Genome Database 2007 [http://www.yeastge <lb/>nome.org/]. <lb/>28. Burke D, Dawson D, Stearns T: Methods in Yeast Genetics 2000 edi-<lb/>tion. Cold Spring Harbor Labratory Press; 2000. <lb/>29. Naumov G: Genetic basis for classification and identification <lb/>of the ascomycetous yeasts. Stud Mycol 1987, 30:469-475. <lb/> 30. Sweeney JY, Kuehne HA, Sniegowski PD: Sympatric natural Sac-<lb/>charomyces cerevisiae and S. paradoxus populations have dif-<lb/>ferent thermal growth profiles. Fems Yeast Res 2004, 4:521-525. <lb/> 31. Naumov GI, Nikonenko TA: New isolates of Saccharomyces par-<lb/>adoxus from oak exudates. Doklady Akademii Nauk SSSR 1988, <lb/> 7:84-87. <lb/> 32. Kane SM, Roth R: Carbohydrate metabolism during ascospore <lb/>development in yeast. J Bacteriol 1974, 118:8-14. <lb/></listBibl> <div type="annex">Additional file 1 <lb/> Table 1: PCR primer sequence table. Contains the sequences of the species <lb/>specific primers used to identify hybrid and non-hybrid matings. <lb/> Click here for file <lb/>[http://www.biomedcentral.com/content/supplementary/1471-<lb/>2148-8-1-S1.doc] <lb/></div> <note place="footnote">Publish with Bio Med Central and every <lb/>scientist can read your work free of charge <lb/> "BioMed Central will be the most significant development for <lb/>disseminating the results of biomedical researc h in our lifetime." <lb/> Sir Paul Nurse, Cancer Research UK <lb/> Your research papers will be: <lb/> available free of charge to the entire biomedical community <lb/>peer reviewed and published immediately upon acceptance <lb/>cited in PubMed and archived on PubMed Central <lb/>yours — you keep the copyright <lb/> Submit your manuscript here: <lb/> http://www.biomedcentral.com/info/publishing_adv.asp <lb/> BioMedcentral</note> </back> </text> </tei>