Development of a thermophilic coculture for corn fiber conversion to ethanol

Conversion of lignocellulosic biomass to biofuels is challenging due to recalcitrance, typically requiring thermochemical pretreatment and large cellulase/hemicellulase inputs. Generation 1.5 concepts seek to upgrade corn ethanol plants by converting the fiber fraction (about 10% of kernel mass), potentially boosting ethanol yield by up to 13% per bushel and enhancing DDGS protein value. Clostridium thermocellum, a moderate thermophile with strong cellulolytic capability, has outperformed fungal cellulases in biomass solubilization and operates advantageously at elevated temperatures. However, it cannot ferment most hemicellulose, notably the complex, highly substituted corn fiber glucuronoarabinoxylan (GAX), which is particularly recalcitrant. The study aims to: (1) evaluate C. thermocellum performance on corn fiber and in coculture with hemicellulose-utilizing thermophiles; (2) identify organisms and enzymes capable of deconstructing recalcitrant corn GAX; (3) transfer key GAX-degrading activities into a thermophilic ethanologen to improve hemicellulose utilization and ethanol production.

Prior work shows C. thermocellum can be 2–4× more effective than commercial fungal cellulases on various feedstocks and is suited for consolidated bioprocessing. Corn fiber’s major hemicellulose, GAX, varies structurally and is notably recalcitrant due to dense and unusual substitutions (e.g., L-galactose, α-xylose), necessitating multiple synergistic enzymes. Gut Bacteroides spp. deploy Sus-like systems for complex glycan catabolism and have been studied on corn xylan. Rogowski et al. characterized Bacteroides ovatus xylanolytic enzymes active on corn GAX. Various thermophilic xylanolytic bacteria (e.g., Caldanaerobius polysaccharolyticus) possess endoxylanases and accessory enzymes but still struggle with corn fiber GAX. Recent reports indicate hemicellulose hydrolysates can inhibit C. thermocellum, motivating cocultures that efficiently remove these products.

• Feedstocks: Wet-milled corn fiber (WMCF) and destarched corn bran (DCB) prepared by amylase treatments, washes, drying, and milling.
• Organisms: C. thermocellum M1570; T. saccharolyticum M1442; T. thermosaccharolyticum LL1548; additional thermophiles and a thermophilic consortium (55 °C) screened; isolation of LL1355 and LL1354 from the consortium.
• Fermentations: pH-controlled bioreactors (1–2 L, 55 °C, pH 6.95, 200 rpm) and serum bottle cultures. Solids loadings typically 10–20 g/L (screening) and 40 g/L (bioreactors). Sequential cocultures adjusted to pH 6.25 for T. saccharolyticum. Duration generally 5 days unless noted.
• Analytical methods: Residual solids and soluble carbohydrates quantified via two-step acid hydrolysis (Quantitative Saccharification, QS) and HPLC (Aminex HPX-87H, RI detection). Broth carbohydrates characterized by MALDI-TOF MS and 1D/2D NMR; size-exclusion chromatography fractionation for neutral/acidic oligosaccharides.
• Organism screening: Growth on Monoculture Broth (spent C. thermocellum broth) to assess utilization of solubilized hemicellulose-derived oligosaccharides.
• Isolation and sequencing: Enrichment on TS-Coculture Broth; colony isolation; 16S rRNA identification; whole-genome sequencing (JGI) and CAZyme annotation via dbCAN.
• Enzyme discovery: From LL1355, 27 CAZyme ORFs cloned into E. coli (pEXP-5-NT), expressed, and screened by pNP assays and monosaccharide release from coculture broths. Six enzymes selected: α-D-xylosidase (GH31), α-L-galactosidases (GH95: 687, 697), β-D-xylosidase (GH120), and two α-L-arabinofuranosidases (GH43: 1120, 996). Purification by His-tag affinity; activities characterized on pNP substrates and on purified GAX oligosaccharide mixtures by NMR/MS.
• Enzyme synergy tests: Individual and combinations of LL1355 enzymes and crude cell-free extracts (CFEs) added to TT- or TS-Coculture Broths to quantify monosaccharide release and improved utilization by T. thermosaccharolyticum.
• Heterologous expression in T. thermosaccharolyticum: Construction of plasmids expressing single, dual (operons with α-Xylp_1211 paired), and four-gene operons (including α-Xylp_1211 plus combinations of β-Xylp_1710, α-Araf_1120/996, α-L-Galp_697). Adaptation transfers on coculture broths; assessment of carbohydrate utilization.
• Coculture bioreactors with added enzymes/CFE: 40 g/L fiber-enriched corn bran fermentations comparing C. thermocellum + T. thermosaccharolyticum with/without LL1355 CFE and with C. thermocellum + LL1355. Bottle tests with five purified enzymes (1 mg/g substrate) assessing ethanol titers.

• Biomass solubilization:
  • C. thermocellum solubilized 95 ± 2% of carbohydrate from DCB at 10 g/L in 5 days; ~90% within 3 days.
  • Fungal cellulase (CTEC2) solubilized 23 ± 1% (20 mg/g) vs autoclaving alone 10%.
  • On WMCF, C. thermocellum achieved 91% (20 g/L, n=1).
• Hemicellulose utilization gaps:
  • After C. thermocellum digestion, T. saccharolyticum consumed only 38 ± 3% of non-glucose sugars; T. thermosaccharolyticum 45 ± 5%.
  • Most remaining carbohydrates were GAX oligosaccharides (DP 4–20) resistant to these partners.
• Structural insights:
  • Four major GAX side chains identified, including unusual residues L-Galp and α-D-Xylp; a complex structure with double-substituted xylose adjacent to GlcA (structure 4) persisted in coculture broths.
• Organism screening:
  • Bacteroides cellulosilyticus utilized 85% and Caldanaerobius polysaccharolyticus 83% of Monoculture Broth carbohydrate; a thermophilic consortium consumed 99%.
• Isolation of LL1355 (Herbinix spp.):
  • Consumed 75% of TS-Coculture Broth and 85% of Monoculture Broth carbohydrates.
  • Genome sequencing and CAZyme profiling enabled targeted enzyme selection.
• Enzyme activities (LL1355):
  • Six enzymes selected (GH31 α-xylosidase, GH95 α-L-galactosidases, GH43 α-L-arabinofuranosidases, GH120 β-xylosidase). α-Xylp_1211 and α-L-Galp_687 showed strong activities; α-Xylp_1211 removed α-D-Xylp from complex acidic oligosaccharide (structure 4).
  • Enzyme combinations exhibited synergy; five enzymes released 42 ± 3% of broth carbohydrate as monosaccharides (similar to LL1355 and B. cellulosilyticus CFEs at ~45%).
• Enzyme supplementation improves utilization:
  • Adding all five enzymes to T. thermosaccharolyticum increased carbohydrate utilization in Monoculture Broth to 78 ± 1% vs 53 ± 3% control; α-Xylp_1211 alone achieved 68 ± 2%.
  • LL1355 CFE further increased utilization to 89 ± 1% (suggesting additional endo-acting activities).
• Heterologous expression in T. thermosaccharolyticum:
  • Single α-Xylp_1211 expression improved utilization of TT-Coculture Broth from 8 ± 5% to 32 ± 2%.
  • Best four-gene strain (LL1703: α-Xylp_1211 + β-Xylp_1710 + α-Araf_996 + α-L-Galp_697) consumed 49 ± 7% of TT-Coculture Broth and 67 ± 2% of Monoculture Broth after 10 days (parent: 50 ± 2% on Monoculture Broth).
• Bioreactor cocultures at 40 g/L fiber:
  • C. thermocellum + T. thermosaccharolyticum: 33% carbohydrate solubilization; 33% GAX utilization.
  • • LL1355 CFE: 46% solubilization; 64% GAX utilization.
  • C. thermocellum + LL1355: 63% solubilization; 90% GAX utilization.
• Ethanol improvement:
  • Five-enzyme addition (1 mg/g) increased ethanol titer from 1.93 ± 0.02 to 2.40 ± 0.08 g/L on 8.8 g/L corn fiber (24% increase).

The study demonstrates that while C. thermocellum excels at solubilizing corn fiber, hemicellulose-derived GAX oligosaccharides limit overall carbohydrate conversion due to their structural complexity. Identifying and characterizing LL1355, capable of consuming most recalcitrant GAX, provided key debranching and exo-acting enzymes targeting unusual side chains (α-D-Xylp, L-Galp, Araf). Supplementation with these enzymes alleviated bottlenecks in T. thermosaccharolyticum, markedly increasing hemicellulose utilization and ethanol titers, and bioreactor cocultures pairing C. thermocellum with LL1355 or its CFE linked enhanced GAX utilization with greater fiber solubilization. Heterologous expression of LL1355 enzymes in T. thermosaccharolyticum improved utilization but underperformed relative to exogenous enzyme addition, implicating limitations in enzyme secretion and oligosaccharide transport. The findings address the core question by establishing that targeted accessory enzyme activities against corn GAX enable efficient thermophilic cocultures to capture hemicellulose sugars for ethanol production and that organismal traits (secretion, transport) are critical for transferring this capability into an ethanologen for consolidated bioprocessing.

C. thermocellum achieves >90% solubilization of corn fiber, outperforming fungal cellulases, but hemicellulose utilization is limited by recalcitrant GAX structures. Isolation of Herbinix spp. LL1355 and structural elucidation of corn GAX led to identification of six key enzymes. Enzyme supplementation to T. thermosaccharolyticum boosted GAX utilization (from 53% to 78%) and increased ethanol titers by 24%. Coculture with LL1355 or its CFE significantly improved GAX utilization and overall solubilization at high solids. Heterologous expression of selected enzymes in T. thermosaccharolyticum enhanced utilization but highlighted the need for effective secretion and transport systems. Future work should focus on engineering thermophilic ethanologens for extracellular secretion of GAX-degrading enzymes, enhancing oligosaccharide transport, incorporating endo-acting xylanases, and potentially engineering LL1355 for high-yield ethanol production to realize robust consolidated bioprocessing of corn fiber.

• Many LL1355 enzymes lacked secretion signals and were expressed intracellularly in T. thermosaccharolyticum; extracellular delivery was minimal, limiting in situ activity.
• T. thermosaccharolyticum may lack transporters for diverse GAX oligosaccharides, constraining benefits of intracellular enzymes.
• Oxygen sensitivity of several enzymes reduced activity; stability considerations apply.
• Difficulty cloning α-L-Galp_687 in multi-gene operons limited some constructs.
• Screening focused on debranching/exo-acting enzymes; insufficient endo-acting hydrolase activity likely constrained full deconstruction.
• Some broth assays may contain residual enzymatic activities from C. thermocellum, potentially confounding exact attributions.
• High-performing comparators (e.g., Bacteroides cellulosilyticus) are mesophilic or grow suboptimally at 55 °C, limiting direct coculture applicability.