Dissecting key biomass traits through traditional QTL mapping, exploring novel functional genomics approaches, and investigating self-incompatibility in Miscanthus

Gifford, Justin Michael

Permalink

https://hdl.handle.net/2142/89108 Copy

Description

Title

Dissecting key biomass traits through traditional QTL mapping, exploring novel functional genomics approaches, and investigating self-incompatibility in Miscanthus

Author(s)

Gifford, Justin Michael

Issue Date

2015-12-02

Director of Research (if dissertation) or Advisor (if thesis)

Juvik, John A

Doctoral Committee Chair(s)

Juvik, John A

Committee Member(s)

• Brown, Patrick J
• Moose, Stephen P
• Sacks, Erik J

Department of Study

Crop Sciences

Discipline

Crop Sciences

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Biofuel
• Cellulosic ethanol
• Miscanthus
• quantitative trait locus (QTL) mapping
• Virus induced gene silencing
• Barley stripe mosaic virus
• Self-incompatibility

Abstract

It has long been recognized that relience on fossil fuels for the world’s energy needs is unsustainable, yet few attractive alternatives have been developed. Establishing an alternative energy source for the transportation sector has been particularly challenging as the vast majority of the current alternative energy technologies result in production of electricity as opposed to liquid fuel. The one major exception to this is biofuels. While corn ethanol has grown rapidly over the last decade, the dedication of such a high percentage of corn production to fuel has put a strain on food prices. For this reason, dedicated lignocellulosic crops and the technology to efficiently produce ethanol from lignocellulosic feedstocks have moved into focus. Miscanthus, a genus of perennial grasses that can produce high yields in temperate environments, has been identified as a potential substrate for ethanol production. Until recently, most of the effort put into Miscanthus breeding over the last century has focused on developing it as an ornamental crop. As breeding goals shift to increasing yield and adapting Miscanthus to local environments, breeders are in need of information regarding the genetic architecture of the traits integral to yield and adaptability. To address this, a four year quantitative trait locus (QTL) study was conducted on a full-sib, F1 mapping population of M. sinensis segregating for flowering time, height, leaf width, and yield using a genetic map consisting of 846 segregating SNP and SSR markers. In total, 78 QTLs with LOD scores above the genome-wide, permuted threshold equivalent to a P-value of 0.05 were identified across 13 traits. Forty of the 78 QTLs were detected in multiple years, and the power to detect QTLs appeared to peak in the third year of growth. Both the use of spring emergence and vigor rating as a covariate to account for variation related to differences in establishment increased the power to detect QTLs in the two year establishment period. Finally, a dry period in the middle of the 2012 growing season suggested that yield declines may be due to a decrease in tiller diameter. Despite the rapid expansion of genetic information pertaining to the Miscanthus genus over the last decade, a rapid, highly efficient functional genomics tool remains elusive. Limited success has been achieved with traditional bombardment and Agrobacterium-mediated transformation, but the low efficiencies of these protocols in combination with the genotype specificity and lengthy time requirements make these techniques imperfect. Virus induced gene silencing (VIGS) may present an ideal alternative. This technique is rapid and robust to polyploidy, but it is untested in Miscanthus. Here we attempted to adapt an existing VIGS vector with a large monocotyledonous host range for use in Miscanthus. Barley stripe mosaic virus (BSMV) has been reported to infect over 240 species within the Poaceae family. This vector was previously modified to include a ligation independent cloning (LIC) site to aid in cloning small plant gene fragments into the virus. In theory, during viral replication, double-stranded RNA will trigger the post-transcriptional gene silencing response, resulting in down regulation of viral transcripts and the cloned plant gene. In order for this process to work, the virus must be able to infect Miscanthus. Five Miscanthus, two maize, and two sorghum genotypes were screened for susceptibility to BSMV to no prevail. Simultaneously, a number of gene fragments were attempted to clone into the LIC site. After extensive trouble shooting, it was determined that a deletion in the original LIC BSMV clone was preventing successful integration of gene fragments. Two strategies are weighed moving forward: test additional genotypes of Miscanthus or acquire one or more additional VIGS vectors to screen Miscanthus for susceptibility. Neither method is guaranteed to work. The decision comes down to the amount of resources and effort available to dedicate towards developing a VIGS system for Miscanthus. Screening additional genotypes with the current BSMV VIGS vector will require the fewest resources but may have a smaller chance of success. Beyond the need for a functional genomics tool, the adoption of Miscanthus as a dedicated bioenergy crop is impeded by, among other things, the necessity to establish fields via clonal propagation. Plants grown from seed segregate widely, resulting in a decrease in plot yield. Further, seed established plots increase the invasive potential of this non-native genus of grasses. Both of these concerns could be alleviated by developing hybrid seed from inbred lines. In order to accomplish this goal, the strong self-incompatibility (SI) mechanism acting in Miscanthus would need to be manipulated. To better understand this mechanism, SI relationships between full siblings of two biparental populations were utilized to map the locus/loci responsible for self-recognition. A single locus was found to be segregating in both mapping populations, and conserved synteny among the grasses does not suggest that this locus corresponds to the S- or Z-loci found to operate in self-recognition of the Pooideae subfamily of Poaceae. Using rice and sorghum as bridge species, this region appears to be homologous to the T-locus, previously believed to only have two alleles: a functional allele and a non-functional allele. Increasing evidence suggests that sorghum has a deletion in the region homologous to the mapped SI locus in Miscanthus, explaining its lack of SI. Despite only detecting a single locus in each mapping population, an investigation of pollen tube growth in a more diverse set of crosses revealed two separate crosses in which 75% of the pollen was compatible. This can only occur in a two locus SI system, suggesting that at least one more locus is responsible for SI in Miscanthus. In light of these new revelations, the model in which the common ancestor of all grasses had a four locus SI system is supported. After the split of the Pooideae and Panicoideae subfamilies, these four SI loci underwent distinct evolutionary pathways, resulting in two related but distinct SI systems.

Graduation Semester

2015-12

Type of Resource

text

Permalink

http://hdl.handle.net/2142/89108

Copyright and License Information

Copyright 2015 Justin Gifford

Owning Collections