Lecture 9: Hybrid Vigor (Heterosis) Michael Gore lecture notes Tucson Winter Institute version 18 Jan 2013
Breaking Yield Barriers for 2050 Phillips 2010 Crop Sci. 50:S-99-S-108
Hybrid maize is a modern marvel! Phillips 2010 Crop Sci. 50:S-99-S-108
Heterosis is one of the least understood biological phenomena that has been exploited by breeders to increase the productivity of domesticated species Springer and Stupar 2007 Genome Res. 17:264-275
Maize Traits with Heterosis Springer and Stupar 2007 Genome Res. 17:264-275
The Making of Corn Belt Dent Lancaster Surecrop OPV Reid Yellow Dent OPV Northern Flint, Modern CBD, Southern Dent Photo: http://thescientistgardener.blogspot.com/2010/12/maize-is-machine.html
Tracy and Chandler 2006 pp 219 233 Lessons from Corn Belt Dent Heterosis in maize has been known since the early 1900s Concept of heterotic patterns developed in the 1960s and 1970s Breeding for heterotic patterns has resulted in increased divergence between groups
Tracy and Chandler 2006 pp 219 233 Lessons from Corn Belt Dent H a : CBD heterotic patterns are not the result of historical or geographical influences Open-pollinated varieties and first cycle inbreds did not show heterotic patterns, thus markers would NOT have been helpful to identify heterotic groups
Tracy and Chandler 2006 pp 219 233 Lessons from Corn Belt Dent CBD heterotic patterns were created by breeders through trial and error In the 1940s, breeders started arbitrarily splitting the germplasm pool into groups (odd vs. even numbered lines) Genetic drift created initial divergence in allele frequencies, which was enhanced by selection
Tracy and Chandler 2006 pp 219 233
Genome-Wide Patterns in CBD Investigated 400 lines from over nearly a century of breeding with ~50k SNPs Steady increase in genetic differentiation and LD, allele frequencies in total population are mostly constant Modern heterotic groups are the product of divergence from a homogenous landrace (OPV) population van Heerwaarden et al. 2012 PNAS 109:12420-12425
Genome-Wide Patterns in CBD Detected very few signatures of directional selection Overall impact of directional selection on genome-wide patterns was limited van Heerwaarden et al. 2012 PNAS 109:12420-12425
Genome-Wide Patterns in CBD Minimal evidence for any single line disproportionately contributing favorable alleles Common alleles donated by a set of representative but few ancestral lines Selection and recombination of many common alleles important but what about genetic drift? van Heerwaarden et al. 2012 PNAS 109:12420-12425
Bernardo 2002 Breeding for Quantitative Traits in Plants pp 243-246 Genetics of Heterosis Dominance hypothesis masking of unfavorable recessive alleles in a heterozygote. Two or more loci are needed because the value of a heterozygote at a single locus (d>a) does not exceed the value of the superior parent. If true, it should be possible to obtain an inbred that performs equally as well as the best hybrid Overdominance hypothesis the heterozygote is superior over either homozygote. Only a single locus (d>a) is needed to achieve heterosis. Also, linkage is not needed to achieve heterosis. If true, it should NOT be possible to obtain an inbred that performs equally as well as the best hybrid
Bernardo 2002 Breeding for Quantitative Traits in Plants pp 243-246 Genetics of Heterosis Pseudo-Overdominance hypothesis repulsion phase linkage of loci that show partial or complete dominance The effects of two loci are difficult to separate if both are tightly linked. If we did not know that two loci comprise a single linkage block, we would incorrectly conclude that heterosis is due to overdominance. Pseudo-overdominance is similar to the two-locus dominance hypothesis, with the exception that repulsion phase linkage is required for pseudo-overdominance.
Birchler et al. 2006 PNAS 103:12957-12958 Genetic Models for Heterosis Complementation Slightly deleterious homozygous a, b, c alleles Allelic interactions Heterozygosity at the B locus with two functional alleles Repulsion Phase Linkage Superior A and B alleles create a superior phenotype from complementation
Hill and Robertson 1966 Genet. Res. 8:269-294 Hill-Robertson Effect HR effect linkage between sites under selection reduces the overall effectiveness of selection for finite natural populations Repulsion phase linkages among favorable alleles will reduce the effectiveness of selection
McMullen 2009 Science 325:737-740 Hill-Robertson Effect Favorable alleles have a higher chance of being in repulsion phase in the presence of low recombination If these favorable alleles exhibit dominance, then low recombination regions should be under high selective pressure to maintain heterozygosity
McMullen et al. 2009 Science 325:737-740 and Gore et al. Science 326:1115-1117 Pericentromeric Regions - within 10 cm on each side of the centromere position - the rest of the chromosome regions
McMullen et al. 2009 Science 325:737-740 and Gore et al. Science 326:1115-1117 Pericentromeric Regions Residual heterozygosity increased 30% in pericentromeric regions (P<0.0004) McMullen et al. 2009 Science - within 10 cm on each side of the centromere position - the rest of the chromosome regions
McMullen et al. 2009 Science 325:737-740 and Gore et al. Science 326:1115-1117 Pericentromeric Regions Residual heterozygosity and R are Residual inversely heterozygosity correlated (r 2 sp=0.35) increased 30% in Diversity pericentromeric (π) and gene regions density (P<0.0004) had no McMullen et al. 2009 Science association with residual heterozygosity. - within 10 cm on each side of the centromere position - the rest of the chromosome regions
NC Design III Design III is a mating design for partitioning the genetic variance into additive and non-additive effects Design III estimates the average level of dominance of genes affecting traits under investigation Random sample of F 2 individuals are separately backcrossed to each of the two inbred parents Comstock and Robinson 1948 Biometrics 4:254-266
QTL Mapping for Heterosis Analyzed backcross series separately Concluded overdominance (or pseudo-overdominance) is a major cause of heterosis for grain yield. Stuber et al. 2002 Genetics 132:823-839
QTL Mapping for Heterosis Re-analyzed Stuber et al. 2002 backcross series together with new statistical model Concluded dominance effects at multiple linked QTL are a major cause of heterosis for grain yield. Cockerham and Zeng Genetics 143:1437-1456
Fine Mapping of Heterosis QTL Overdominant QTL on chr 5 was dissected by NILs into two tightly, linked dominant effect QTL in repulsion phase. Provides evidence for pseudo-overdominance. Graham et al. 1997 Crop Sci. 37:1601-1610
Meta-QTL Analysis of Heterosis Concluded pseudo-overdominance is a major cause of heterosis in maize and no significant epistasis. Heterotic QTL for grain yield mapped near low R pericentromeric regions (i.e., likely repulsion phase) Schön et al. 2010 TAG 120:321-332