Why the cellulosome?

The cellulosome is a large extracellular multi-enzyme complex produced by anaerobic cellulolytic bacteria for efficient degradation of plant cell wall polysaccharides.

It is one of the most efficient natural systems for degrading crystalline cellulose and lignocellulosic biomass.

Motivations for study:

• Biofuels / biorefinery — consolidated bioprocessing for cellulosic ethanol
• Industrial enzymes — understanding synergistic enzyme action
• Protein engineering — modular architecture enables "designer cellulosomes"
• Basic science — model for supramolecular enzyme complexes
• Synthetic biology — engineering organisms for biomass conversion

Cellulosome architecture

Three functional classes of components:

1. Scaffoldins — non-catalytic proteins containing cohesins that organize the complex
2. Catalytic subunits — enzymes containing dockerins that bind to cohesins
3. Cell surface anchors — proteins that attach the cellulosome to the cell wall

Key assembly logic (cohesin–dockerin):

• Type I dockerin (on enzymes) binds Type I cohesin (on the CipA scaffoldin)
• Type II dockerin (on CipA) binds Type II cohesin (on anchor proteins, e.g. SdbA / OlpB)

This modular system lets many enzymes be organized on one scaffold, then anchored to the cell surface.

The approach: AI-assisted gene review

• Apply the AI gene review framework to systematically review existing GO annotations against the literature and bioinformatics.
• Each gene gets a structured review YAML; annotations are kept, modified, removed, or flagged as over-annotated.

Primary model: Acetivibrio thermocellus (formerly Clostridium thermocellum)

• UniProt code ACET2 (taxonomy ID 1515)
• The organism in which cellulosomes were discovered (as "cellulose-binding factor")
• Best-characterized system; ~79 cellulosomal genes
• Thermophilic (~60 °C) and industrially relevant

Second model: Clostridium cellulovorans (CLOCL) — mesophilic, produces both cellulosomal AND free cellulases (57 cellulosomal genes).

Genes reviewed — A. thermocellus (1/2)

Priority 1 — Scaffoldins and anchors (architectural, non-catalytic):

Priority 2 — Key cellulases (GH-family enzymes):

Genes reviewed — A. thermocellus (2/2)

Priority 3 — Hemicellulases and other activities:

Priority 4 — Additional cellulosomal proteins:

All 17 A. thermocellus genes reviewed and validated.

Second model — Clostridium cellulovorans

Mesophilic complement to thermophilic A. thermocellus; produces both cellulosomal and free cellulases.

• CbpA scaffoldin: 9 cohesin domains and 4 SLH domains.
• Gene cluster: cbpA–exgS–engH–engK–hbpA–engL–manA–engM–engN.
• All 5 genes reviewed and validated.

Finding: scaffoldins are non-catalytic — fix the annotations

Scaffoldins (cipA, cipB, sdbA, cbpA) organize the complex but do not catalyze cellulose degradation. Several IEA annotations were incorrectly propagated:

• GO:0004553 (hydrolase activity) — incorrectly propagated from dockerin domains to non-catalytic scaffoldins → flagged
• GO:0030246 (carbohydrate binding) — incorrectly attributed to cohesin domains, which bind dockerins (proteins), not carbohydrates → flagged
• For CbpA: GO:0071555 (cell wall organization) marked for REMOVAL

Recommended additions to capture true function:

• GO:0043263 (cellulosome) — cellular component
• GO:0044575 (cellulosome assembly) — biological process
• GO:1990308 / 1990309 / 1990311 / 1990312 — specific cohesin–dockerin domain binding

Finding: cellulases — right activity, sharpen the context

Cellulases (celS, celA, celC, celD, celK, celX) — annotations mostly correct. Recommendations:

• Add GO:0043263 (cellulosome) as cellular component
• Add GO:1990311 (type-I cohesin domain binding) for the dockerin function
• Mark overly general process terms (polysaccharide catabolic, carbohydrate metabolic) as over-annotated, in favor of specific GO:0030245 (cellulose catabolic process)

EngK / EngL (GH9, C. cellulovorans):

• Accepted GO:0008810 (cellulase activity), GO:0030245 (cellulose catabolic process)
• Added GO:0043263 (cellulosome), GO:1990311 (type-I cohesin domain binding)

Finding: cellulosomal vs free — EngO is the contrast case

EngO (Q6DTY2) is a GH9 endoglucanase that is NOT part of the cellulosome:

• Lacks a dockerin; instead carries its own CBM for substrate targeting
• Functions as a free secreted enzyme that complements the cellulosome
• Added GO:0030248 (cellulose binding) for its CBM function
• No cellulosome annotations, unlike the dockerin-bearing EngK / EngL

This distinction — dockerin-anchored vs. free CBM-targeted — is exactly what GO annotations must capture correctly.

HbpA (Q9RGE7) — accessory scaffoldin: GO:0000272 (polysaccharide catabolic) marked for REMOVAL (non-catalytic); CBM-style GO:0030246 marked UNDECIDED (possible artifact); added type-I dockerin binding + cellulosome assembly.

Challenges

• Modular, multi-domain proteins — a single protein mixes catalytic domains, dockerins/cohesins, CBMs, and SLH anchors; annotations get cross-propagated between domains incorrectly.
• Complex assembly — cohesin–dockerin interactions are protein–protein, not protein–carbohydrate; easy to mis-annotate as carbohydrate binding.
• GO representation of multi-enzyme complexes — capturing "cellulosome" as a component and "cellulosome assembly" as a process, distinct from the catalytic activities.
• Data quality — several literature-derived UniProt IDs mapped to the wrong organism (e.g. Q9RGE9 = cysN from C. jejuni; Q9L3K5 = cheA from B. cereus; P23661 = celB from R. albus) and had to be removed.

Status and future directions

Status — complete:

• 17 A. thermocellus genes reviewed and validated
• 5 C. cellulovorans genes reviewed and validated (after pruning mis-mapped IDs)
• Systematic corrections: removed mis-propagated catalytic / carbohydrate-binding terms on scaffoldins; added cellulosome, cellulosome assembly, and specific cohesin–dockerin binding terms

Future directions:

• Expand to Ruminococcus flavefaciens (FD-1) — the most complex cellulosome described, 223 dockerin-containing proteins, unique cohesin–dockerin binding modes
• Deeper review of remaining C. cellulovorans proteins still lacking validated UniProt entries