From a genetic point of view, the three main models that explain heterozygote advantage rely on dominance or deviation from additive effects in one or more loci Lippman and Zamir, The dominant complementation model proposes that dominant alleles from both parents reciprocally complement negative effects of recessive alleles, creating a hybrid with less recessive alleles than either parent Bruce, Another model explains how interactions between alleles of different loci, epistasis, which are created due to new combinations in the hybrid, contribute to its improved performance Crow, ; Comstock and Robinson, ; Comstock et al.
A third model explains heterosis by synergistic effects of over-dominant interactions between alleles created by a heterozygous state within one or more loci East, ; Shull, ; Crow, Additionally, a few more models, which do not necessitate dominance, explain heterosis by an overall improvement in hybrid's metabolism, energy production, and energy utilization that could result from gene dosage effects, polyploidy, changes in gene expression levels and changes in function of protein complexes Hedgecock et al.
The level of superiority varies considerably among hybrids and conditions, making heterosis a quantitative phenomenon. This phenomenon can often be found in reproductive fitness traits that are quantitative by themselves, like some of the production traits in plants and animals. Therefore, methods used to map loci contributing to heterosis were similar to those used in mapping quantitative trait loci QTL; Deutschbauer and Davis, ; Altshuler et al.
Unlike regular QTL mapping studies, mapping QTL contributing to heterosis is actually mapping the mode of inheritance of contributing loci rather than the contribution alone. It is conceivable that if heterosis shares similarities with other quantitative traits, the genetic basis underlying the superior performance of the hybrid will be composed of many loci, each with a minor or mild dominant effect due to interactions between alleles within and between loci.
As for many other quantitative traits, also for heterosis, loci with minor effects are harder to identify than loci with major effect. While many studies mapped loci contributing to heterosis, identifying over-dominant loci turned out to be a more difficult task. Consequently, there are only a few known cases, mostly in plants, of single genes that confer heterotic effect by an over-dominant interaction between their alleles Hua et al.
Over-dominant loci are interesting because of the nature of interaction between their alleles that confers a synergistic effect. Furthermore, for these known over-dominant loci, a major phenotypic contribution was measured, making such loci attractive and feasible to utilize for improving performance of hybrids in traits of interest.
On the other hand, some loci that were initially mapped as over-dominant were later studied further and turned out to be pseudo over-dominance loci. In pseudo over-dominance cases, two or more genetically linked genes, each having a dominance complementation effect but in repulsion phase, give the impression of one over-dominant QTL. The existence of pseudo over-dominance loci is used to argue against the prevalence estimates of true over-dominance loci that were obtained in some heterosis QTL studies Charlesworth and Willis, ; Schnable and Springer, We have constructed a collection of heterozygous hybrids by crossing 16 homozygous parental strains and measured their growth rate in five environmental conditions Shapira et al.
In this collection of hybrids, we identified a significant amount of best-parent heterosis. Thirty five percent of hybrid-condition combinations were heterotic, a large proportion that makes this collection suitable for studying the genetic basis of heterosis.
Analyses of phenotypic results provided evidence that all three genetic models, namely, dominant complementation, over-dominance, and epistasis, could be used to explain cases of heterosis found in this collection. Since the mechanisms underlying the synergistic effect of over-dominant loci are not well-understood, in this study we used the yeast model to identify such loci and characterize their contribution.
Baker ] was used unless other specific medium is mentioned. SC medium lacking uracil was used for selecting cells with intact URA3. The strains used in this study are diallel strains Shapira et al. Diploid cells were grown on YEPD plates in patches overnight and scraped off from the plates into sporulation solution [0. The PCR mix also contained: 1. PCR products were verified by electrophoresis on 1.
Primers used for both techniques are detailed in Supplementary Table 1. Replacing the native allele of a gene with an allele from another strain was done by the two-steps transformation method Storici et al. In the first step, by replacing the gene of interest with a cassette containing drug resistance and the URA3 gene homolog and in the second, by replacing the cassette with the alternative allele and selecting successful replacements on 5-FOA SC plates.
Successful integration into the genome was confirmed by lack of growth on uracil-lacking SC plates and by PCR using position specific primers.
The primers used in transformation procedures are detailed in Supplementary Table 1. In Shapira et al. In this study, these segregants were genotyped by HRM at five different genes to determine which parental allele each segregant had in each gene.
This genotypic information allowed assigning genotypes to the diploid segregants of the following populations. The doubling time DT values of the diploid segregants from our previous study were used in conjunction with the genotypic values obtained in this study for the analyses. The DT data is available from Shapira et al. A hybrid strain between S and SK1 containing HygB and G resistances in its HO locus was sporulated and 24 of its tetrads were dissected to obtain 96 haploid spores.
Each group contained several tens of strains according to the distribution of alleles in the chosen F1 haploid parents. For ADR1 and Win. The phenotypic and genotypic data of all strains measured in this study were stored in datasets according to the different experiments and these are given as Supplementary Data Sheet 2. Any experiment that measured growth of strains on more than one wells plate, included the same reference strain on each plate.
The average of that strain was used to estimate plate effect and normalize the growth of other strains allowing strain comparisons across plates and experiments. Haploid strains from SK1 and S genetic backgrounds were used in the two-steps construction of allele replacement strains see above for details on the procedure. In the first step, the original allele is deleted and replaced by a cassette containing a drug marker and in the second step the cassette is replaced with the introduced allele.
For the analysis of haploid strains, gene-deleted strains from the first step and allele-replaced strains from the second step were used.
For the analysis of diploids, nine genotypic groups were produced using the original and allele-replaced strains of both backgrounds. The nine groups were: three AEP3 genotypic groups, homozygotes for S allele, homozygotes for SK1 allele and heterozygotes in each of the SK1, S, and hybrid backgrounds. Doubling time DT -values of strains were measured and calculated as detailed in our previous work Shapira et al.
For statistical analyses that compared growth between strains, Log 2 DT-values were used. Heterotic value of each hybrid was calculated according to the following categories:.
If the hybrid Hy was faster than its faster parent FP , its heterotic value was calculated as:. If the hybrid was slower than its slower parent SP , its heterotic value was calculated as:. If the hybrid DT-value was within the range of the parents, its heterotic value was set to zero. Heterotic values of the hybrids were used in the genome-wide scan to identify candidate over-dominant loci.
Zeroing the value of all the hybrids that were within the range of the parental lines, left for the following scan only the phenotypic variability of those hybrids that exceeded the parental range.
Additive genetic deviation a —The difference between each of the homozygous parents and the mid-parent value:. Dominant genetic deviation d — deviation of the heterozygous hybrid Hy from its mid-parent value:. Degree of dominance values were used in the analysis of the hybrid-specific background loci, in the analyses of populations RBC1 and F2 , and in the analyses of deletion and allele-replacement strains.
The parental strains used for making hybrids were fully sequenced and the list of SNPs for each strain that was called in that study Liti et al. This list of polymorphism included only SNP genotypes and no other types of polymorphisms. Briefly, a sliding window algorithm was used to divide the SNP data of the 16 parental strains to genomic windows containing between two and nine haplotypes, with three to ten SNPs in each.
The genotype of hybrids was inferred from their known parental genotypes. For each genomic window, the hybrids were divided into genotypic groups: homozygotes for any of the haplotypes in the window and heterozygotes for any combination of haplotypes.
An example of such grouping in a window that contained two haplotypes and thus, three genotypic groups is given in Supplementary Figure 1A. In windows containing more than two haplotypes, only haplotypes found in three or more parents were considered for analyzing the phenotypic differences between local homozygotes and heterozygotes.
In every window with a significant difference, the three genotypic groups that yielded the most significant difference were chosen. Genomic windows in which the mean heterotic value of the heterozygous hybrids group was significantly larger than the means of both homozygous groups were considered as candidate loci with an over-dominant effect.
Finally, since multiple comparisons were carried out, to reduce the chances for false positives identification, we applied a Benjamini—Hochberg correction at an FDR level of 0. The set of hybrids in the heterozygous group varied among candidate windows. From the list of significant windows prior to the FDR correction, we chose 74 windows for which the specific SK1xS hybrid was included in the group of heterozygote hybrids.
This was a logical filtering step since over-dominant contribution is expected only from heterozygous loci. We then applied to these 74 windows a more stringent analysis to identify the more promising over-dominant loci in this hybrid background. Parental haplotypes were deduced from the sequences and hybrids were reassigned to three genotypic groups in each window based on the haplotype combination of their parents. Due to the differences between the SNP list and the full sequence alignments, changing a haplotype call for even one parent per window, changed the genotypic grouping of the hybrids and could have changed the results of the statistical tests.
In practice, grouping of hybrids in all 74 windows changed in the second analysis compared to the first scan. Therefore, using all sequence polymorphisms, not only SNPs, made this second analysis more stringent. Zeroing values for half of the hybrids gave more weight to the other half the heterotic hybrids and favored the identification of over-dominant loci in the initial analysis.
An example of the spread around the mean of each genotypic group is given in Supplementary Figure 1B. Based on the sequence alignments, from the list of 74 loci, 12 were filtered out since either SK1 and S shared the same haplotype, or the new haplotypes assignment created parental groups with less than three parents. The 62 remaining loci were reanalyzed to compare the three genotypic groups of the hybrids using both Kruskal—Wallis aparametric test and the more stringent Tukey—Kramer HSD parametric test.
The F2 population was used to evaluate the mode of inheritance of candidate loci. Secondly, we determined the mode of inheritance by comparing a - and d -values in each locus. The additive genetic deviation a was half the difference between means of the two homozygotes and the dominant genetic deviation d was the difference between the mid-parent and the mean of the heterozygous segregants.
Positive d -values were assigned when the mean growth of heterozygotes was faster than the mid-parent whereas negative d -values were assigned when slower. Larger d - than a -values indicate over-dominance. To compare between a - and d -values, each F2 diploid was assigned with either a - or d -values its difference from the overall mid-parent value if it was a homozygote or a heterozygote, respectively. We compared the mean a - and d -values for each locus using a t -test.
For the analysis of epistasis between pairs of candidate loci, we compared the d -values of a given locus with and without heterozygotes in another locus using a t -test. We sought to identify loci that show a non-additive mode-of-inheritance and therefore, have a non-additive effect on the phenotype and not just an effect due to substitution of alleles.
Each SNPs-containing genomic window defined a set of haplotypes for the haploid parental strains and based on the haplotypes of their parents, the hybrids were categorized as heterozygotes or homozygotes to one or the other haplotypes Supplementary Figure 1A.
Altogether, 50, windows were tested. The distribution of haplotype number per window ranged between two and twelve with a median of four haplotypes Supplementary Figure 2A. The window size ranged from 2 to 11, bp with a mean of bp and a median of 97 bp Supplementary Figure 2C.
These statistics reflect on the variability in polymorphism levels and on the extent of haplotype blocks in different regions of the genome. The size of a window was set computationally such that each haplotype in it was found in at least three parental strains. Therefore, the small median size and the small SNP number inside each window indicated high level of strain divergence and short spans of linkage disequilibrium.
In out of 50, windows that were tested, the mean heterosis value of the local heterozygous hybrids was significantly different from the means of both local homozygous hybrid groups Supplementary Data Sheet 1. These significant loci were divided into windows with an over-dominant effect and 11 with an under-dominant effect. Both over and under-dominant windows were distributed across the chromosomes Figure 1.
Figure 1. Genomic distribution of candidate over-dominant loci. X-axis denotes the position along the chromosome in Kb and Y-axis denotes the chromosome numbers. Each dot marks one of the windows that were found to be significant in the genome-wide scan. Green dots mark the 55 potential over-dominant windows in the SK1xS background and red dots the 12 statistically significant ones. Each version is called an allele.
For every gene, you inherit two alleles: one from your biological father and one from your biological mother. Together, these alleles are called a genotype.
If the two versions are different, you have a heterozygous genotype for that gene. For example, being heterozygous for hair color could mean you have one allele for red hair and one allele for brown hair. The relationship between the two alleles affects which traits are expressed. It means your biological parents contributed identical variants.
In this scenario, you may have two normal alleles or two mutated alleles. Mutated alleles can result in a disease and will be discussed later. This also affects which characteristics appear. In a heterozygous genotype, the two different alleles interact with each other. This determines how their traits are expressed. Commonly, this interaction is based on dominance.
Depending on how the dominant and recessive genes interact, a heterozygous genotype might involve:. In complete dominance, the dominant allele completely covers up the recessive one.
One example is eye color , which is controlled by several genes. An organism with a hybrid trait may contribute either a dominant or recessive allele.
In this way, an organism's offspring can be phenotypically different from its parents. For example, if both parents have a particular hybrid trait, the offspring can have a homozygous pairing of recessive alleles in that trait. In order to visualize the probabilities of pure or hybrid offspring, you can draw a diagram called a Punnett square. A Punnett square is a block of squares with one parent's alleles along the top of the diagram and another parent's alleles along the left of the diagram.
Represent a dominant allele with a capital letter and a recessive allele with a lowercase letter. In each square, write the combination of that specific row and column of alleles. For example, a Punnett square of a cross of two Pp organisms would yield PP in the top left square, Pp in the top right square, Pp in the bottom left square and pp in the bottom right square.
This particular cross could yield both pure and hybrid offspring. Serm Murmson is a writer, thinker, musician and many other things. This step is the dihybrid cross, and it is represented as:. Mendel observed that the F2 progeny of his dihybrid cross had a ratio and produced nine plants with round, yellow seeds, three plants with round, green seeds, three plants with wrinkled, yellow seeds and one plant with wrinkled, green seeds. From his experiment, Mendel observed that the pairs of traits in the parental generation sorted independently from one another, from one generation to the next.
Mendel chose to cross a pea plant that was homozygous and dominant for round RR , yellow YY seeds with a pea plant that was homozygous and recessive for wrinkled rr , green yy seeds, represented by the following notation: RRYY x rryy Organisms in this initial cross are called the parental, or P generation. This step is the dihybrid cross, and it is represented as: RrYy x RrYy Mendel observed that the F2 progeny of his dihybrid cross had a ratio and produced nine plants with round, yellow seeds, three plants with round, green seeds, three plants with wrinkled, yellow seeds and one plant with wrinkled, green seeds.
Further Exploration Concept Links for further exploration principle of independent assortment test cross gene allele dominant recessive genotype phenotype Principles of Inheritance.
0コメント