Wood represents the majority of the biomass on lands, and it constitutes a renewable source of biofuels and other bioproducts. However, wood is recalcitrant to bioconversion, meaning that feedstocks must be improved. We investigated the properties of wood that affect bioconversion, as well as the underlying genetics, to help identify superior biorefinery tree feedstocks. We recorded as many as 65 wood-related and growth traits in a population of European aspen natural genotypes. These traits included three growth and field performance traits, 20 traits for wood chemical composition, 17 traits for wood anatomy and structure, and 25 wood saccharification traits as indicators of bioconversion potential. We used statistical modelling to determine which wood traits best predict bioconversion yield traits. This way, we identified a core set of wood properties that predict bioprocessing traits. Several of these predictor traits showed high broad-sense heritability, suggesting potential for genetic improvement of feedstocks. Finally, we performed genome-wide association study (GWAS) to identify genetic markers for yield traits or for wood traits that predict yield. GWAS revealed only a few genetic markers for saccharification yield traits, but many more SNPs were associated with wood chemical composition traits, including predictors traits for saccharification. Among them, 16 genetic markers associated specifically with lignin chemical composition were situated in and around two genes which had not previously been associated with lignin. Our approach allowed linking aspen wood bioprocessing yield to wood properties and the underlying genetics, including the discovery of two new potential regulator genes for wood chemical composition.

Naturally fluorescent polymeric molecules such as collagen, resilin, cutin, suberin, or lignin can serve as renewable sources of bioproducts. Theoretical physics predicts that the fluorescence lifetime of these polymers is related to their chemical composition. We verified this prediction for lignin, a major structural element in plant cell walls that form woody biomass. Lignin is composed of different phenylpropanoid units, and its composition affects its properties, biological functions, and the utilization of wood biomass. We carried out fluorescence lifetime imaging microscopy (FLIM) measurements of wood cell wall lignin in a population of 90 hybrid aspen trees genetically engineered to display differences in cell wall chemistry and structure. We also measured the wood cell wall composition by classical analytical methods in these trees. Using statistical modeling and machine learning algorithms, we identified parameters of fluorescence lifetime that predict the content of S-type and G-type lignin units, the two main types of units in the lignin of angiosperm (flowering) plants. In a first step toward tailoring lignin biosynthesis toward improvement of woody biomass feedstocks, we show how FLIM can reveal the dynamics of lignin biosynthesis in two different biological contexts, including in vivo while lignin is being synthesized in the walls of living cells.

Trees constitute promising renewable feedstocks for biorefinery using biochemical conversion, but their recalcitrance restricts their attractiveness for the industry. To obtain trees with reduced recalcitrance, large-scale genetic engineering experiments were performed in hybrid aspen blindly targeting genes expressed during wood formation and 32 lines representing seven constructs were selected for characterization in the field. Here we report phenotypes of five-year old trees considering 49 traits related to growth and wood properties. The best performing construct considering growth and glucose yield in saccharification with acid pretreatment had suppressed expression of the gene encoding an uncharacterized 2oxoglutarate-dependent dioxygenase (2OGD). It showed minor changes in wood chemistry but increased nanoporosity and glucose conversion. Suppressed levels of SUCROSE SYNTHASE, (SuSy), CINNAMATE 4-HYDROXYLASE (C4H) and increased levels of GTPase activating protein for ADP-ribosylation factor ZAC led to significant growth reductions and anatomical abnormalities. However, C4H and SuSy constructs greatly improved glucose yields in saccharification without and with pretreatment, respectively. Traits associated with high glucose yields were different for saccharification with and without pretreatment. While carbohydrates, phenolics and tension wood contents positively impacted the yields without pretreatment and growth, lignin content and S/G ratio were negative factors, the yields with pretreatment positively correlated with S lignin and negatively with carbohydrate contents. The genotypes with high glucose yields had increased nanoporosity and mGlcA/Xyl ratio, and some had shorter polymers extractable with subcritical water compared to wild-type. The pilot-scale industrial-like pretreatment of best-performing 2OGD construct confirmed its superior sugar yields, supporting our strategy.

No abstract

A method to estimate the surface moisture content below the fibre saturation point that is a function of the surface temperature, wet-and dry bulb temperatures, equilibrium moisture content, and fibre saturation point was evaluated. The method is based on the premise that the surface temperature is solely influenced by the surface moisture content and the climate that the surface is exposed to. The prediction model contends that the surface moisture content is equal to the fibre saturation point when the surface temperature is equal to the wet bulb temperature, and equal to the equilibrium moisture content when the surface temperature is equal to the dry bulb temperature, with a linear interpolation between those two points. The model thus predicts that the average moisture content of a thin piece of veneer can be predicted with fairly good accuracy. Also, when drying boards in a fast changing climate, e.g. fan reversals in industrial kilns, the surface temperature and surface moisture content should change as abruptly as the climate does. Additionally, the surface moisture content should correlate to the known drying phases, with a consistently high surface moisture content during the capillary phase when the wet line is close to the surface, and a quickly decreasing surface moisture content when the wet line moves into the wood during the transition to the diffusion phase. The prediction model was tested in these three scenarios and the results suggest that the basic premise is reasonable, and that the method is useful for surface moisture content estimation.