WO2010029548A1

WO2010029548A1 - Method for identifying genetic loci invovled in hybrid vigor

Info

Publication number: WO2010029548A1
Application number: PCT/IL2009/000885
Authority: WO
Inventors: Eyal Fridman; Hadar Less; Imri Ben Israel; Eli Reuveni
Original assignee: Yissum Research Development Company Of The Hebrew University Of Jerusalem, Ltd.; Yeda Research And Development Company Ltd.
Priority date: 2008-09-11
Filing date: 2009-09-10
Publication date: 2010-03-18

Abstract

The invention provides a method for identifying one or more loci in a genome of an organism involved in overdominance of a phenotypic trait. a plurality of crosses are performed between pairs of founder organisms. An overdominance heterosis parameter (ODH) is calculated indicative of a deviation of the phenotypic value of the parents and the progeny. The progeny are divided based on their genotype at the locus into a homo-haplo-group (HoHG) consisting of progeny carrying two identical haplotype segments at the locus, and a hetero-haplo-group (HeHG) consisting of progeny carrying two different haplotype segments. Heterosis trait loci are identified showing a significant differencein the overdominance heterosis parameter ODH values between the HoHG group and the HeHG group.

Description

METHOD FOR IDENTIFYING GENETIC LOCI INVOVLED IN HYBRID

VIGOR

FIELD OF THE INVENTION

The present invention relates to methods and systems for quantitative genetic analysis.

BACKGROUND OF THE INVENTION

Genetic variation usually takes the form of a continuous phenotypic range due to the involvement of many alleles in a polygenic system together with environmental effects. Genetic studies exploring the molecular basis of quantitative variation, for example identifying and characterizing quantitative trait loci/nucleotides (QTL/QTN), either use bi-parental crosses (e.g. F2 backcrosses, recombinant inbreds, and introgression lines) or by performing association studies using a panel of individual lines (Flint-Garcia 2003, Takeda and Matsuoka, 2008).

The term "heterosis", or "hybrid vigor" is used to refer to the phenomenon in which a hybrid organism posseses an enhanced phenotype in compraison with its parental inbred lines. If the hybrid phenotype is enhanced compared to the mean of the two parental phenotypes, the heterosis is referred to as "midparent heterosis" (MPH). If the hybrid phenotype enhanced compared to both of the two parental phenotypes, the heterosis is referred to as "best parent heterosis" (BPH) (Falconer and Mackay 1996, Hochholdinger and Hoecker 2007). Identification of genes involved in hybrid vigor can be used in selecting parent organisms in hybrid crosses to generate hybrids having enhanced phenotypes in comparison to the parents. The identified genes can be cloned and transferred to elite varieties through standard transgenic methods, or by allele replacement of endogenous alleles, to generate organisms with enhanced phenotypes.

SUMMARY OF THE INVENTION

This present invention provides a method for identifying genetic loci in an organism involved in hybrid vigor for of one or more predetermined traits. Genetic loci involved in hybrid vigor are referred to herein as "heterosis trait loci" (HTL). The invention may be applied to any type of organism, including plants, especially crops, microorganisms, fungi, and animals.

In accordance with the invention, a plurality of crosses are performed between pairs of founder individuals. Each of the founders and the progeny of the crosses undergo a quantitative phenotypic characterization for one or more predetermined quantitative traits, which for a plant, for example, may be any one or more of the individual's biomass, growth rate, flowering time, and yield. For each cross that was performed, a parameter, referred to herein as the "over dominance heterosis" (ODH), explained in detail below, is calculated. The ODH parameter is a measure of the deviation of the phenotypic value of the progeny of the cross from that of its parents.

In each of one or more predetermined genetic loci, the crosses are grouped based on the genotype of the progeny in the locus into "homo-haplo-groups" (HoHGs) of crosses in which the progeny of the cross carry two identical haplotype segments, and "hetero-haplo-groups" (HeHGs) of crosses in which the progeny carry two different haplotype segments. A statistical test is applied to determine whether there is a significant difference in ODH values of the progeny individuals in the HoHG groups and the HeHG groups of the locus. A locus for which there is a significant difference in the ODH values of the progeny in at least one of the HeHG groups compared with the corresponding HoHG groups is a locus likely to be involved in hybrid vigor. Loci that show a significant difference in the ODH values of the progeny in the corresponding two HoHG groups compared with the HeHG group are identified as being heterosis trait loci. Furthermore, heterosis trait loci that show a non-significant difference in the ODH values of the progeny between the corresponding two HoHG are identified as being major heterosis trait loci.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to understand the invention and to see how it may be carried out in practice, embodiments will now be described, by way of non-limiting example only, with reference to the accompanying drawings, in which:

Fig. 1 shows the yeast genomic infrastructure for genetic association of heterogeneity with growth rate; Fig. 2A-3B shows spot dilutions of selected strains (Fig. 2A) and growth curves (Fig. 2B);

Fig. 3 shows a graph summarizing the inheritance of growth rate for 105 yeast hybrids; and Fig. 4 depicts a cumulative distribution function plot of an (A) overdominant and (B) underdominant HTL identified in yeast.

DETAILED DESCRIPTION OF THE INVENTION

As explained above, the method of the invention for identifying heterosis trait loci involves calculating an overdominance heterosis parameter. The ODH parameter is a measure of the extent to which the average phenotype of the progeny of a cross deviates from the phenotypes of one or both of the parents. The ODH parameter of a cross may be obtained by first calculating an ODH parameter for each individual progeny of the cross, referred to herein as a "replicate" of the cross", and then calculating the ODH of the cross of as the average ODH of the replicates of the cross.

In one preferred embodiment, the ODH parameter of a cross Px X Py (a cross between founder x and founder y) is calculated as follows. The mean value of the low parent and the mean value of the high parent is found by determining the mean phenotypic value of the founder lines. The low mean is referred to herein as "the mean value of the low parent" and is denoted by "Pl ". The high mean determined is referred to herein as "the mean value of the high parent" and is denoted by "P 2 ". The ODH parameter is then determined as follows:

(1) If the phenotypic value of replicate r of the cross, Fl_xyr, is greater than or equal P2_xy, then ODH parameter of the replicate r, 0DH_xyr, is calculated using the algebraic expression: ODHχ_yr =(Fl_xyr -P2_xy)/P2_xy.

(2) If the phenotypic value of the replicate r, Fl_xyrj less than or equal to Pl_x^ then 0DH_xyr is calculated using the algebraic expression: 0DH_xyr =(Fl_xyr -

(3) If the phenotypic value of the replicate r, Fl_xyr, is between Pl_xy and P2_xy, then ODH_xyr=0. The mean ODH_xy for the cross is then calculated for all the replicates of the

∑0DH_xyr cross Px X Py using the algebraic expression ODH _xy = — , where R is the

R total number of replicates of the cross.

The one or more predetermined loci may be selected by applying a variable- sized sliding window (VSW; Guo et al. 2009) to define multi-SNP haplotype segments in the founder lines across the genome. The size of the sliding window in each genomic region, defined by the local haplotype diversity (Guo et al. 2009) is determined to obtain a distribution of haplotypes segments (alleles) in which at least two alleles are shared by three founder lines or more. The genotype of the parental lines {PL..Pi} and their hybrids is then defined at a loci set L= { 1 , ... ,1} with haplotype segments, or alleles, indicated as A={1,..., a} [the number of alleles in each locus vary]. The different allele combinations at a particular locus can be described as Ci_xy (_x={_li..._ja}; y={i,...,a})_> for a diploid organism (x and y being the two haplotypes in trans, each originating from a different parent). In a preferred embodiment, the set of founder individuals is selected from a pool of individuals based on one or more genetic criteria. The genotype of the candidates may be obtained, for example, through targeted sequencing/resequencing of predetermined loci throughout the genome followed by sequence alignment and determining the different haplotypes in each of the predetermined loci. Genetic criteria used to select the founder individuals from the pool of candidate individuals may include any one or more of the following criteria: i) For each of one or more predetermined genetic loci, the population of selected founder lines should preferably contain a mean allele number per locus that is a predetermined fraction of the number founder individuals to be selected. Preferably, the predetermined fraction is between 20-25 % of the number founders to be selected. For example, if 20 founders are to be selected, the mean allele number should preferably be 4-5 alleles per locus. ii) The allele frequencies among the predetermined loci should be as equal as possible. iii) The founders are preferably selected from different branches of a phylogenetic tree in order to reduce linkage disequilibrium between non-linked loci. The selected set of founder organisms may undergo a more detailed genotyping than was initially applied to the entire pool of candidates. This genotyping may involve, for example, genotyping technologies such as SNP arrays or genome resequencing by next-generation sequencing (NGS) technologies (Huang et al. 2009). The phenotypic characterization of the parents and progeny is preferably performed by a formula which parameterizes a phenotype of a hybrid in relation to its parents. For example, midparent heterosis (MPH) or best parent heterosis (BPH) may be used.

As explained above, a locus for which there is a significant difference in the ODH values of the progeny in the HoHG groups compared with the HeHG groups is a locus likely to be involved in hybrid vigor, and this is determined by applying a statistical test. The statistical test may be, for example, any one of the following statistical tests:

(i) Kruscal Wallis test (Sheshkin 1997) may be used between the HoHG and the HeHG groups using the mean ODH values of each one of the hybrids. Significant observed P values (P_OBS) is compared against the extreme (lowest) P value (P_MIN) °f 1000 iteration of permutation tests. The test is significant and is not obtained by chance when P_OBS < P_MIN | Alternatively, the null hypothesis is accepted that a significant value could be obtained by chance (ii) a pair-wise analysis between trios, i.e., hybrids belonging to HeHG to those belonging to the corresponding two HoHG (harbor the two alleles that constitute the HeHG genotype) is performed in the following manner:

1) for each trio, the entire set of ODH values was obtained;

2) one tailed Kolmogorov-Smirnov test (KS test, denoted as P_KS) between HoHG to HeHG gejn§ty]|es is performed by testing the null hypothesis that a population's CDF (Cumulative Distribution Function) from HeHG genotype is smaller or greater than each one of the two HoHG genotypes. 'Overdominance' is assumed in cases when P_KS < 0.05 under the 'smaller' tail option and P_KS > 0.05 under the 'greater' tail. 'Underdominance' is assumed in cases when Pκs> 0-05 under the 'smaller' tail and P_KS < 0.05 under the 'greater' tail.

In a preferred embodiment, an additional statistical test is applied to confirm the expectation of significant interaction between the two alleles in the locus by comparing the two HoHG genotypes, using a two tailed KS test, for example. Accepting the null hypothesis confirms the conclusion of intra-locus interaction and the identity of the locus as a heterosis trait locus.

EXAMPLE The opportunistic yeast Saccaromyces cerevisiae serves as a valuable model to dissect the genetic basis of quantitative and complex traits (Sinha et al. 2006). Recently, Liti et al. (2009) have presented one to four-fold coverage of the genome sequences of over seventy isolates of the baker's yeast S. cerevisiae and its closest relative, Saccharomyces paradoxus. These strains originated from diverse niches including isolates from soil, clinical tissues, wine, beer and many others, and therefore represent a repertoire of allelic diversity which is a key in associating genetic heterogeneity with quantitative fitness phenotype.

The method of the present invention was applied to the analysis of the genomic basis of overdominance in yeast with regard to the phenotype of growth rate (GR) of the yeast under temperature gradients.

Fig. 1 shows the yeast genomic infrastructure for genetic association of heterogeneity with growth rate. Haploid strains went through a process that included diploidization, introduction of two resistance cassettes (resistance to kanamycin and hygromycin), and sporulation to generate four different types of haploids (two mating types [a and α] X the two resistance cassettes). These served to generate a matrix of crosses that produced two of each homozygous parental line and four replicates of each reciprocal hybrid.

The homozygous strains and their hybrids were initially phenotyped for normal and hot temperature growth (Htg), at 30⁰C and 37°C, respectively, by the colony-size assay (Fig. 2A). Comparison of the different hybrids with their corresponding parent strains revealed that while for some of the trios (two parents and their hybrid offspring) a clear overdominant growth was observed, either in the size and the number of colonies, in others the hybrid showed either a mid-parent phenotype or a dominant mode of inheritance. For example, the hybrid of strains SlOOl and DBVPG6040 clearly showed overdominant growth over the homozygous diploids, as compared with the hybrid originating from crossing strains SlOOl to BC 187 which grew in a similar manner to its parent. The phenotype obtained by the colony-size assay was compared to that obtained by monitoring growth in liquid cultures. Overall, the growth rate of the parents and their hybrid offspring as obtained from the liquid culture experiments, correlated well with the results obtained in the colony-size assays, as can be seen in Figs; 2A and 2B.

In the next step, matrix of 15 homozygous genotypes and 105 hybrids was 5 arranged in 96-well plates, in a replicated trial, and were grown in an incubator-shaker at 37°C with a common reference (the common lab strain S288c). The liquid colonies were monitored for absorbance at OD595 and the normalized GR were obtained for each well against 8 replicates of the reference placed in all plates. Fig. 3 summarizes the results obtained in this experiment. The bar plots depict the average growth rate of 20 to

10 40 replicates of each genotype; for all 105 hybrids grown in the yeast diallele. The parent identity is depicted above and beside the matrix and the average growth rates observed in each trio (two parents and hybrids) is depicted as bar plots, including the mean ODH values and their standard deviations. Calculation of the growth rate was performed using a least squares fitting to the last 50 readings (250 min) which were

15 taken before reaching an OD₅₉₅=0.3.

In order to associate heterogeneity across the genome, a matrix of the haplotypes of each parent in a variable sliding window (VSW) was generated. As defined above, the criteria of obtaining at least one trio with HoHG represented by three independent founder lines was followed. Out of approximately 10,000 windows the statistical

20 analysis identified 381 overdominant and 67 underdominant HTLs (data not shown). Furthermore, performing an additional test to confirm the expectation of significant interaction between the two alleles (intra-locus interaction) in each of the HTLs by comparing the two HoHG genotypes in each of these loci delimited this list to only 10 and 3 major overdominant and underdominant HTLs, respectively.

25 Fig. 4 shows the cumulative distribution function of an overdominant and underdominant HTL. The red and blue lines depict the distribution of the ODH values within hybrids belonging to the hetero-haplo-group (HeHG) and the two homo-haplo- groups (HoHG) in this locus, respectively.

Claims

CLAIMS:

1. A method for identifying one or more loci in a genome of an organism involved in overdominance of one or more predetermined traits, comprising:

(a) providing a plurality of founders, each founder being a homozygous individual of the organism;

(b) performing a plurality of crosses between pairs of founder organisms to produce progeny of each of the crosses;

(c) for each of the one or more predetermined traits calculating a phenotypic value for each of the founders and progeny, the phenotypic value being indicative of a phenotypic characterization of the founders and progeny with respect to the one or more predetermined traits;

(d) for each set of progeny and for each of the one or more predetermined traits, calculating an overdominance heterosis parameter (ODH), the overdominance heterosis parameter being indicative of a deviation of the phenotypic value of the trait and progeny from the phenotypic value of the trait of one or both of the parents of the progeny;

(e) for each of one or more loci, grouping the progeny based on their genotype at the locus into a homo-haplo-group (HoHG) consisting of progeny carrying two identical haplotype segments at the locus, and a hetero-haplo-group (HeHG) consisting of progeny carrying two different haplotype segments;

(f) applying one or more statistical tests to identify one or more heterosis trait loci, a heterosis trait locus being a locus showing a predetermined significant difference, as determined by the statistical test, in the overdominance heterosis parameter ODH values between the HoHG group and the HeHG group.

2. The method according to Claim 1 wherein calculation of the ODH parameter of the cross comprises calculating an ODH parameter for each replicate of the cross and then calculating the ODH of the cross of as an average ODH of the replicates of the cross.

3. The method according to Claim 2 wherein the ODH parameter of a replicate of the cross, 0DH_xyr, is calculated in a method comprising:

(a) if the phenotypic value of the replicate Fl_xyr, is greater than or equal to the mean value of the high parent, P2_xy, then 0DH_xyr, is calculated using the algebraic expression: 0DH_xyr =(Fl_xyr -P2_xy)/P2_xy;

(b) if the phenotypic value of the replicate Fl_xyr> less than or equal to the mean value of the low parent, Pl_xy, then ODH _xyr is calculated using the algebraic expression: ODH_xyr=(Fl_xyr -Pl_xy)/Pl_xy; and

(c) if the phenotypic value of the replicate, Flχy_r, is between Pl_xy and ¥2^, then ODH_xyr=0.

4. The method according to any one of the previous claims wherein the set of founder individuals is selected from a pool of candidate individuals based on one or more genetic criteria.

5. The method according to Claim 4 wherein the one or more genetic criteria is selected from:

(a) the set of selected founder lines contains a mean allele number per locus that is a predetermined fraction of a number founder individuals to be selected;

(b) allele frequencies among the predetermined loci are about the same; and

(c) the founders are selected from different branches of a phylogenetic tree.

6. The method according to Claim 5 wherein the one or more genetic criteria includes the criterion that the set of selected founder lines contains a mean allele number per locus that is a predetermined fraction of a number founder individuals to be selected, wherein the predetermined fraction is between 20-25 % of the number founders to be selected.

7. The method according to any one of the previous claims wherein the one or more predetermined loci are selected by applying a variable-sized sliding window to define multi-SNP haplotype segments in the founder lines across the genome.

8. The method according to Claim 4 further comprising determining a genotype of the candidate individuals.

9. The method according to Claim 8 wherein the genotype of the candidate individuals is determined by a method comprising targeted sequencing/resequencing of predetermined loci throughout the genome followed by sequence alignment and identifying different haplotypes in each of the predetermined loci.

10. The method according to any one of the previous claims further comprising determining a genotype of the founders.

11. The method according to Claim 10 wherein the genotype of the founders is determined by SNP arrays or genome resequencing by next-generation sequencing (NGS) technologies.

12. The method according to any one of the previous claims wherein the phenotypic characterization of the founders and progeny involves parameterizing a phenotype of a hybrid in relation to its parents.

13. The method according to Claim 12 wherein the parameterizing a phenotype of a hybrid in relation to its parents comprises calculating a midparent heterosis

(MPH), best parent heterosis (BPH) or overdominant heterosis (ODH).

14. The method according to any one of the previous claims wherein the statistical test is selected from

(a) a Kruscal Wallis test; and (b) a pair-wise analysis between hybrids belonging to the HeHG to those belonging to two corresponding HoHGs.

15. The method according to any one of the previous claims further comprising an additional statistical test to confirm a significant interaction between the two alleles in the HeHG in which the two HoHG genotypes are compared.

16. The method according to Claim 15 wherein the additional statistical test is a two tailed Kolmogorov-Smimov test to determine major HTLs.