Hybrid Aspen Expressing a Carbohydrate Esterase Family 5 Acetyl Xylan Esterase Under Control of a Wood-Specific Promoter Shows Improved Saccharification

Abstract

Fast-growing broad-leaf tree species can serve as feedstocks for production of bio-based chemicals and fuels through biochemical conversion of wood to monosaccharides. This conversion is hampered by the xylan acetylation pattern. To reduce xylan acetylation in the wood, the Hypocrea jecorina acetyl xylan esterase (HjAXE) from carbohydrate esterase (CE) family 5 was expressed in hybrid aspen under the control of the wood-specific PtGT43B promoter and targeted to the secretory pathway. The enzyme was predicted to deacetylate polymeric xylan in the vicinity of cellulose due to the presence of a cellulose-binding module. Cell-wall-bound protein fractions from developing wood of transgenic plants were capable of releasing acetyl from finely ground wood powder, indicative of active AXE present in cell walls of these plants, whereas no such activity was detected in wild-type plants. The transgenic lines grew in height and diameter as well as wild-type trees, whereas their internodes were slightly shorter, indicating higher leaf production. The average acetyl content in the wood of these lines was reduced by 13%, mainly due to reductions in di-acetylated xylose units, and in C-2 and C-3 mono-acetylated xylose units. Analysis of soluble cell wall polysaccharides revealed a 4% reduction in the fraction of xylose units and an 18% increase in the fraction of glucose units, whereas the contents of cellulose and lignin were not affected. Enzymatic saccharification of wood from transgenic plants resulted in 27% higher glucose yield than for wild-type plants. Brunauer-Emmett-Teller (BET) analysis and Simons' staining pointed toward larger surface area and improved cellulose accessibility for wood from transgenic plants compared to wood from wild-type plants, which could be achieved by HjAXE deacetylating xylan bound to cellulose. The results show that CE5 family can serve as a source of enzymes for in planta reduction of recalcitrance to saccharification.

Keywords: Populus; acetyl; acetyl xylan esterase; enzymatic saccharification; hybrid aspen; xylan.

FIGURE 1

Growth and transgene expression in selected WP:CE5 lines (11, 14B, and 14C) used in this study compared to wild type (WT). (A) Transgene transcript levels in developing wood determined by RT-qPCR; the expression relative to two reference genes was normalized to the levels observed in line 11. AXE specific activity in wall-bound (B) and soluble (C) protein fractions isolated from developing wood, determined by release of acetic acid from acetylated aspen wood powder used as substrate. Data in (A–C) are means ± SE, N = 3, biological replicates. (D) Height growth of transgenic lines during 8 weeks in the greenhouse. (E) Diameter of internodes 20 and 40 at week 8. (F) Average internode length at week 8. Data in (D–F) are means ± SE, N = 6 trees per transgenic lines and 23 trees per WT. P-values correspond to post-ANOVA Dunnett test (for individual lines) or contrast analysis comparing all transgenic lines to WT.

FIGURE 2

Diffuse reflectance Fourier-transform infrared (FTIR) spectra of wood of WP:CE5 transgenic (lines 11, 14B, 14C) and wild type (WT). Dotted lines show bands that are significantly different (≥50% correlation) in the transgenic lines compared to WT, according to OPLS-DA (orthogonal projections to latent structures – discriminant analysis) models using 1 + 1 (predictive + orthogonal) components. Model components are: R2X (cum) = 0.788, R2Y (cum) = 0.891, Q2 (cum) = 0.878. Score plot (A), loadings (B) and the corresponding average spectra (C).

FIGURE 3

Effects of WP:CE5 expression (lines 11, 14B, and 14C) on xylan acetylation in aspen wood. (A) Cell wall acetyl content determined by acetic acid release. (B) 2D qHSQC NMR spectra of extracted xylan showing signals from acetylated (green) and non-acetylated (blue) Xylp units in xylan, which were used for quantification in (C). X23-diacetylated Xylp; X2-Xylp monoacetylated at position 2; X3G2-Xylp acetylated at position 3 and glucuronosylated at position 2; numbers in parenthesis correspond to carbon number in Xylp; Xint – different internal Xylp signals as assigned by Grantham et al. (2017). (D) OLIMP analysis of acidic and neutral xylo-oligosaccharides (XOS) released by endoxylanase AaGH10. Data are means ± SE, N = 2 or more biological × 3 technical replicates for (A), 2 or more biological × 2 technical replicates for (C), and 2 biological replicates for (D). P-values in (A,C) correspond to post-ANOVA contrast analysis comparing all transgenic lines to WT.

FIGURE 4

Effects of WP:CE5 expression on saccharification. (A) Sugar yields of enzymatic hydrolysis without pretreatment. (B) Sugar yields of enzymatic hydrolysis after acid pretreatment. (C) Combined sugar yields of pretreatment liquid and enzymatic hydrolyzate. Data are means ± SE, N = 3 technical replicates of a pooled material from four trees. P-values correspond to ANOVA contrast analysis comparing all transgenic lines to WT.

FIGURE 5

Effects of WP:CE5 expression on wood nanostructure. Simons’ staining of non-pretreated (A) and acid-pretreated aspen wood (B). Data are means ± SE, N = 3 technical replicates. P-values correspond to ANOVA contrast analysis comparing all transgenic lines to WT.