Novel molecular tools to uncover the genetic architecture of hemp fibre quality

Hemp (Cannabis sativa L.) is a high-yielding fibre and environmentally-friendly crop that produces biomass for multiple purposes. The bio-composite applications of hemp fibres have received lots of attention in the last decades by governments and industry. These bio-composites have the capacity to replace non-renewable and unsustainable resources (e.g. synthetic and mineral fibres). However, the molecular knowledge of hemp fibre quality is limited. Consequently, breeding programs to increase the yield and quality of hemp fibre have mostly focused on conventional methods. It is envisioned the development of molecular markers associated to fibre quality, which are expected to accelerate the development of new hemp cultivars with high quality fibres.

The biochemical composition of the cell walls is a key factor defining the quality of hemp fibre. Nevertheless, prior to this thesis there were no high-throughput methods for biochemical characterization of many hemp accessions. Consequently, the analysis of many accessions was expensive and time consuming. In Chapter 2 of this thesis, we developed and optimized the throughput of five methods including cell wall extraction, biochemical composition of cell wall polysaccharides, quantification of lignin, quantifications of crystalline polysaccharides and stem morphology. The methods presented here revealed to be highly repeatable. In addition, the cell wall extraction was optimized to extract enough material for the complete characterization of the cell wall of hemp in a single run, while reducing the time of the entire analysis. Yet, wet biochemistry phenotyping methods are still costly and laborious. To overcome these limitations, high-throughput prediction models based on near-infrared spectroscopy were developed in Chapter 3. These models allowed phenotyping of large numbers of hemp accessions in a rapid and cost-efficient way.

Hemp fibre quality is a quantitative trait and little is known about its phenotypic variability, its heritability behaviour and its environmental component. In Chapter 3, we got insights into the variability of fibre quality within a hemp panel of 123 accessions. The plants were cultivated in three European locations with contrasting environments. The accessions were phenotyped for 30 traits relevant to fibre quality. The study showed extensive variability in all traits. In addition, the characterization of the phenotypic variation and heritability of the traits led to the identification of fibre quality traits, that potentially lead to high genetic gains when used in breeding programs. Relevant traits to maximize these genetic gains include most cell wall components (cellulose, hemicellulose and lignin contents), bast fibre content and flowering traits. Furthermore, all traits analysed showed significant G×E interactions. Yet, these interactions were small in important traits and they were not expected to interfere in breeding decisions. Moreover, all fibre quality traits showed an adaptive behaviour to environmental factors. Highly adaptive traits were pectin content, most agronomic traits and several fibre traits (fineness and decortication efficiency). Therefore, the influence of the environment on fibre quality should be considered while designing a breeding program for the fibre quality of hemp.

The development of updated breeding programs for hemp fibre quality requires the identification of QTLs. To this end, the hemp accession panel was genotyped using RAD-seq. A large set (>600.000) of SNPs was identified. The genotypes and the phenotypes were used to perform Genome-Wide Association Studies (GWAS) to get insights into the genetic architecture of highly heritable cell wall components and bast fibre content. Yet, the incomplete sequenced genome of hemp hindered the mapping. To go beyond these bounds, we first developed a methodology to identify independent QTLs (Chapter 4). The method used the collinearity between allelic frequencies of significant markers. The analyses identified 16 QTLs for fibre quality traits. These QTLs are excellent tools for the development of novel fibre cultivars of hemp. Furthermore, putative candidate genes for the biosynthesis and modification of monosaccharides, polysaccharides and lignin were detected in the QTLs. This study reported the first GWAS in hemp and the first QTLs for hemp fibre quality. Thus, it is a significant step towards the development of a marker-assisted selection for hemp fibre quality.

Flowering time underlies a shift in the carbon partitioning of hemp from producing biomass to produce flowers and seeds. This shift influences fibre quality. Sex determination is also known to influence fibre quality. Male plants die after flowering, while females remain alive until seed maturity. Meanwhile, monoecious plants are characterized by a uniform flowering and fibre production. It is well-known that genetic factors play a role in flowering time and sex determination but the major molecular mechanisms controlling these traits are not characterized in hemp. To gain insights into this knowledge gap, we reviewed the state-of-art of flowering time, sex determination and fibre quality in short-day plants and dioecious crops, respectively, and compare them with hemp (Chapter 5). In this review we proposed possible mechanisms for the genetic control of these complex traits in hemp.

In Chapter 6, we attempt to close the gap in molecular knowledge on flowering time and sex determination, using a GWAS approach. The study revealed six QTLs for flowering and two for sex traits. In addition, putative candidate genes were identified. Genes involved in the photoperiod, vernalization and endogenous flowering pathways were identified in the QTLs for flowering traits. Moreover, genes involved in the balance of phytohormones, such as gibberellic acid insensitive and auxin response factors were identified in a QTL for sex determination. Sex QTLs were associated to the stability of monoecious plants and the candidate genes support this association. This chapter provides additional tools to further develop molecular breeding for hemp.

Finally, the knowledge generated in this thesis was integrated in the light of hemp breeding. We provided high-throughput methods to improve the phenotyping. We provided deeper insights into the variability of hemp fibre quality and its relation to the environment. Complementary practices alternative to breeding for poorly heritable traits were discussed. In addition, we identified QTLs and molecular pathways of fibre quality traits. Overall, the collection of results from this thesis can be used to improve hemp fibre quality and expand the use of hemp as a feasible alternative to non-renewable and unsustainable materials.