Genomic instability in an interspecific hybrid of the genus Saccharomyces: a matter of adaptability
Miguel Morard1,2
, Clara Ibáñez2, Ana C. Adam2, Amparo Querol2, Eladio Barrio1,2, Christina Toft
Ancient events of polyploidy have been linked to huge evolutionary leaps in the tree of life, while increasing evidence shows that newly established polyploids have adaptive advantages in certain stress conditions compared to their relatives with a lower ploidy. The genus Saccharomyces is a good model for studying such events, as it contains an ancient whole-genome duplication event and many sequenced Saccharomyces cerevisiae are, evolutionary speaking, newly formed polyploids. Many polyploids have unstable genomes and go through large genome erosions; however, it is still unknown what mechanisms govern this reduction. Here, we sequenced and studied the natural S. cerevisiae × Saccharomyces kudriavzevii hybrid strain, VIN7, which was selected for its commercial use in the wine industry. The most singular observation is that its nuclear genome is highly unstable and drastic genomic alterations were observed in only a few generations, leading to a widening of its phenotypic landscape. To better understand what leads to the loss of certain chromosomes in the VIN7 cell population, we looked for genetic features of the genes, such as physical interactions, complex formation, epistatic interactions and stress responding genes, which could have beneficial or detrimental effects on the cell if their dosage is altered by a chromosomal copy number variation. The three chromosomes lost in our VIN7 population showed different patterns, indicating that multiple factors could explain the mechanisms behind the chromosomal loss. However, one common feature for two out of the three chromosomes is that they are among the smallest ones. We hypothesize that small chromosomes alter their copy numbers more frequently as a low number of genes is affected, meaning that it is a by-product of genome instability, which might be the chief driving force of the adaptability and genome architecture of this hybrid.
Keywords- adaptation,
- genome instability,
- hybrids,
- resequencing,
- Saccharomyces cerevisiae ,
- Saccharomyces kudriavzevii
Polyploidy and allopolyploidy are gaining attention as important drivers of genome evolution, more specifically, their central role in plant and microbe domestication. In yeasts, hybridization is now used in industry to improve stress resistance and different relevant characteristics of the fermentative processes. Thanks to the soar of sequencing studies, it has become evident that industrial and domesticated yeasts harbour large genomic rearrangements such as polysomies and translocations. This is also the case for allopolyploid yeast. Genomic rearrangements can be important and frequent in these strains; however, the relation between them and selective pressures are not evident. In this study, we sequenced an allotriploid strain, VIN7, and show that its genome is unstable, and this instability can widen its phenotypic landscape. Moreover, we provide evidence that the genomic changes, most frequently observed, are the mirror of a high genomic instability, which could be an advantage in industrial and changing environments, due to their more neutral effects. This study provides more evidence of the link between domestication, genomic instability and hybridization as different sides of the same process : genomic flexibility to withstand ever-changing environments.
- All sequencing data generated in this study are available from the National Center for Biotechnology Information under BioProject PRJNA611499, BioSample accession number SAMN11349820, and SRA accession numbers SRR11301281 (Roche 454) and SRR9925222 (Illumina data).
- Genetic interactions were downloaded from http://thecellmap.org/costanzo2016/.
- Complexes were downloaded from the CYC2008 database – http://wodaklab.org/cyc2008/resources/CYC2008_complex.tab.
- Physical protein interactions were downloaded from the Saccharomyces Genome Database (SGD) – https://downloads.yeastgenome.org/curation/literature/interaction_data.tab.
- Genomes for determining chromosome copy number frequencies were downloaded from http://1002genomes.u-strasbg.fr/files/.
Duplication of the whole genome, either by self-genome duplication or hybridization between two phylogenetically related species, also referred to as autopolyploidy and allopolyploidy, respectively, have been linked to large evolutionary leaps in all kingdoms of life, i.e. plants [1–5], vertebrates [6–8], fungi [9–11]. In the case of hybridization, this is in part due to the inheritance of traits from both parents, although seldom in an accumulative way [12–14]. The divergence between the parental species has been linked with the ploidy of newly formed hybrids, with allopolyploids tending to have more diverged parents than homoploids [15]. Another important property of increasing ploidy of the genome is redundancy, which provides genetic buffering resulting in increasing genetic robustness [16]. Niche specialization for hybrids can occur rapidly, even in a few generations [17]. Despite the evidence for the power of this evolutionary mechanism, it comes with a high risk of failure, as newly formed hybrids exhibit low fertility, small population sizes and low variability, which increases their chance of being outcompeted in stable conditions and are often considered as evolutionary dead-ends [18].
Hybridization has played an important role in the domestication of different species. Agriculturally, human pressure on crop evolution has selected, for example, fruit size, yield, taste and evenness of maturation [19, 20]. Sequencing of cultivated crops, such as wheat, strawberries, mandarins, carrots, etc., have shown that they have undergone multiple hybridizations/crossings to become the fruits we know today [21]. Likewise, in fermentative conditions, where humans use microbes/yeasts in the production of alcoholic beverages, hybrid species have been isolated [22–24]. Here, the predominant hybrids are a cross between the good fermenter Saccharomyces cerevisiae and one of the cold-tolerant species of the genus, such as Saccharomyces kudriavzevii, Saccharomyces eubayanus or Saccharomyces uvarum [22–24]. Inheriting properties from both parents is indicative that the newly formed hybrid is often more cold-tolerant than S. cerevisiae strains and more ethanol-tolerant than the other parent species, e.g. S. kudriavzevii [25, 26]. Therefore, hybridization is shown to be an important domestication event in the genus Saccharomyces.
The central role of hybridization of plants in human civilization has resulted in huge efforts in understanding the evolutionary consequences of this mechanism. It has been observed that newly formed polyploids go through large genomic changes after the initial ‘genomic shock’ of duplicating the genetic material. These changes include recombination between sub-genomes, gene/chromosomal losses and transcriptome rewiring [27–30]. The second face of hybrid evolution involves reshaping the large genomic redundancy through further gene loss or sub- and neo-functionalization of the retained genes [31, 32]. These early faces are thought to be a period of genomic instability and have been shown to vary in length, with examples spanning from hundreds to thousands of generations [33]. Long-term consequences of this can be seen in Saccharomyces, where approximately 30 % of the duplicated genes originated from an interspecific hybridization event, approximately 100-200 million years ago, have been retained [10, 34, 35].
The flexibility of the genomes of genus Saccharoymces is highlighted with the fact that they can form hybrids despite their high sequence divergence of up to 20 % [36], and very complex hybrids have been obtained, consisting of up to six parental species [37]; however, the most commonly observed number in naturally occurring hybrids is two to three parental species. As seen in plants, newly formed allopolyploid yeasts are unstable and undergo large genomic changes just after hybridization, which results in aneuploidies and chimeric chromosomes [22, 23, 38–44]. The stability of the sub-genomes is not equal, and in one study observing parental species found the least represented parent retained on average 50 % of the genome [23]; however, the range is large and the complexity of the resulting hybrids varies [22, 23, 45–49]. Furthermore, the stability of the hybrid can be influenced by environmental factors [50].
An interesting observation from looking at polyploids in S. cerevisiae is that they are more prone to aneuploidy than diploids or haploids [51]. In fact, in the strains sequenced to the date, aneuploidy is found in industrial strains that are often polyploid [51–53], which is also the case for allopolyploids of the genus, suggesting a link between genome stability, aneuploidy and hybridization in domestication. Furthermore, large difference of aneuploidy in a population has been coupled with phenotypic diversity, ultimately facilitating large phenotypic leaps [54, 55]. In general, it has been revealed that the genome structure is much less stable than what was classically thought, with increasing evidence that it could be an important aspect of evolution and domestication. However, there is still a lack of observations of how fast the genomic changes in unstable hybrids can occur, and if the genomic changes observed in polyploids and allopolyploids are the result of instability or are selected.
In this study, we sequenced the genome of VIN7, a natural S. cerevisiae × S. kudriavzevii allotriploid hybrid strain commercialized by Anchor Yeast as a dry yeast for winemaking. The genome of this strain was first described by comparative genomic hybridization [47] and then by sequencing in different studies [56, 57]. Intriguingly, the genome content reported was not coherent between these studies, because the origin of the strain was different: the derivative commercial dry yeast in the first study and the original strain in the other two. The differences in genome content could be the result of selective pressure or stochasticity and genome instability. We show here that the genome of VIN7 is not stable and that its instability influences the phenotype of the strain. Furthermore, we investigate different properties of the variable chromosomes that could influence the probability of a chromosome to be lost or retained and, hence, the shape of its genome architecture to illustrate the idea that the instability in itself could be the valuable trait of hybridization for industrial yeasts.
The hybrid yeast S. cerevisiae × S. kudriavzevii VIN7 used in this study was isolated from a commercial dry yeast sample provided by Anchor Yeast. The strain was re-hydrated and grown in GPY plates (4 % glucose, 0.5 % peptone, 0.5 % yeast extract and 2 % agar) at 25 °C overnight, from which a glycerol stock was prepared and stored at −70 °C. This glycerol stock was used as starting material for culture on a GPY plate, which was grown for 48 h.
The yeast strain isolates were cultivated in GPY medium (20 g glucose l−1, 5 g peptone l−1, 5 g yeast extract l−1), at 25 °C for 24 h, and DNA was isolated according to standard procedures [58] where Zymolate and SDS were used for lysing the yeast cells, followed by extraction of the DNA using potassium acetate.
The spot plate technique was used to examine the effect of temperature on the growth of the three sup-populations of VIN7 (dry yeast, glycerol stock and plate) on GPY agar. Starter yeast cultures were obtained by inoculating the yeast strains into 10 ml GPY medium and incubating overnight at 25 °C. Tubes were centrifuged at 4000 r.p.m. for 5 min. The supernatant was discarded and pellets were re-suspended in sterilized water. Cell suspensions were then prepared based on optical density determined using a spectrophotometer. Samples were diluted until an optical density of 1 was obtained at a wavelength of 600 nm. Subsequent 1 : 10 dilutions were carried out to prepare serially diluted samples. A 10 µl volume of each dilution was spotted onto GPY agar plates in triplicate. The plates were then incubated in a static incubator at 12 and 28 °C for 7 days. Data was recorded by photographing the spot plates.
Ethanol tolerance of the three sub-populations of VIN7 (dry yeast, glycerol stock and plate) was evaluated by performing growth essays in GPY medium with 10 % (v/v) ethanol. Growth was monitored by measuring the optical density at 600 nm in a SPECTROstar Omega instrument (BMG Labtech). Measurements were taken every 30 min for 43 h after a 20 s pre-shaking for all the experiments. All the experiments were carried out in sextuplicate. Growth parameters like starting optical density, maximum optical density, growth rate and area under the curve were calculated using the R package's Growthcurver [59].
The S. cerevisiae × S. kudriavzevii VIN7 genome was sequenced using 454 technology (shotgun with 550 bp read length and paired-end reads with 8 kb insert size) in combination with paired-end Illumina technology (600 bp insert size) on a HiSeq 2000 instrument. The Illumina reads were filtered using Sickle [60] with a minimum read length of 80 bp and minimum quality score of 30. A subset of 5000000 paired-end reads was used for the de novo assembly, giving a coverage of 22× on the haploid genome. The program sff_extract was used for extracting the 454 reads and to clip ends with low quality and/or adaptor sequence. The coverage of the 454 reads was 1.7× for the 157000 paired-ends and 7× for the 157000shotgun reads for the haploid genome. A de novo assembly was carried out using mira v 3.4.1.1 (https://sourceforge.net/projects/mira-assembler/) and GS De Novo Assembler (Roche/454 Life Sciences). The resulting assemblies were corrected and manually edited using Consed [61], contigs were concatenated or broken based on paired-end information. More specifically, fake reads generated from the corrected GS De Novo Assembler assembly (run with default parameters for heterozygotic mode and minimum read length of 25) were added to the main mira assembly (run without separating out long repeats and uniform read distribution). The determination of the recombination points was done manually though Consed. The scaffolds were aligned to S. cerevisiae reference strain S288C and S. kudriavzevii IFO 1802 with the all genome aligner MUMmer [62], and with this information the scaffolds were ordered into chromosome structure using an in-house script (https://github.com/evosysmicro/LabThings/blob/master/scripts/ultraScaf.pl).
Illumina reads mappings were done on the reference genomes of the S. cerevisiae S288C and S. kudriavzevii IFO 1802 strains by using bowtie2 [63] with default settings. Any potential secondary mappings were filtered away with Samtools [64], such that only uniquely mapped reads were kept for further analysis.
The annotation was carried out in three steps. (i) Annotation from the previously published VIN7 genome (pubVIN7) [56] was transferred to the new assembly using ratt [65]. (ii) Ab initio gene prediction was performed with Augustus [66] to detect possible unannotated genes in the pubVIN7. (iii) The result from the two first steps was carefully and manually checked using Artemis [67] and the pubVIN7 re-annotated as we detected two principal problems in the previous annotation: (a) the presence of indels (a usual problem with 454 sequencing technology) resulted in many coding sequences (CDS) being removed; (b) genes containing introns were not annotated previously or only one of the exons was present.
DNA from all yeast strains, hybrid VIN7, S. cerevisiae S288C and S. kudriavzevii CR85, was extracted following a previously described procedure [58]. DNA concentration and purity were determined with a NanoDrop ND1000 spectrophotometer (Thermo Fisher Scientific) and genomic DNA integrity was checked by electrophoresis in 0.8 % agarose gel.
Oligonucleotide primers for quantitative real-time PCR (Isogen Life Science) were species-specific and located in different chromosomes (III, VI, XI). We checked each chromosome with two pairs of primers (see Table 1) designed using the comparison between sequences from S288C and CR85 and the pubVIN7 [56].
Factors affecting chromosome aneuploidy. (a) CDF value heatmap of the characteristics analysed for each chromosome and its aneuploidy frequency. Values near to one indicate a higher than expected mean and near to zero a lower than expected mean. Aneup freq, aneuploidy frequency; PI intra, intra-chromosomal physical interactions; GI intra (+), positive intra-chromosomal genetic interactions; GI intra, intra-chromosomal genetic interactions; GI intra (−), negative intra-chromosomal genetic interactions; Expr GLY, expression on glycerol; Expr LAC, expression on lactate; Expr ETOH, expression on ethanol; Expr OXD, expression on oxaloacetate; PI mean, mean physical interactions; PI inter, inter-chromosomal physical interactions; GI (+), positive genetic interactions; GI inter (+), positive inter-chromosomal genetic interactions; GI, genetic interactions; GI (−), negative genetic interactions; GI inter, inter-chromosomal genetic interactions; GI inter (−), negative inter-chromosomal genetic interactions; Complexes (IN/Non), proteins being in a complex or not. (b) Aneuploidy frequency of the chromosomes compared to the chromosome size. (c) Less interacting chromosomes network. The arrow between chromosomes represents the chromosome with which the interaction tends to be neutral.
https://www.microbiologyresearch.org/content/journal/mgen/10.1099/mgen.0.000448
No hay comentarios:
Publicar un comentario