Skip to content

Lactoferrin Shio-Koji Protease Stability: Computational Analysis (comp-005)

Status: Complete — 2026-05-05
Experiment folder: experiments/comp-005-lactoferrin-shio-koji-protease-stability/
Companion analysis: comp-001 (uricase)

Summary

AlphaFold pLDDT structural analysis of human lactoferrin (P02788, 710 aa) against three A. oryzae koji proteases (ALP, NPr, acid protease) under shio-koji conditions (17.5% NaCl midpoint, pH 4.5–5.0, 22°C).

Two verdicts:

Scope Verdict Max risk score Worst protease
Full sequence (incl. signal peptide) HIGH 0.388 NPr
Mature protein only (aa 20–710) MODERATE 0.188 ALP

The distinction is critical: all top-5 cleavage sites across all three proteases are in the signal peptide (aa 1–19, pLDDT 35–54). If A. oryzae signal peptidase cleaves the heterologous signal sequence (which it commonly does for secreted proteins), the operative risk for the mature protein is MODERATE. If signal peptide is retained, full-sequence verdict (HIGH) applies.

The §1.10 lactoferrin arm remains a feasibility gate — this analysis identifies where to look if degradation is observed, but the wet-lab result is the primary determination. Contrast with comp-001 (uricase), which produced a LOW verdict and reframed §1.10's uricase arm as a confirmation experiment.

Method

Same shared library as comp-001 (experiments/lib/protease_stability.py). Risk score per site:

risk_score = accessibility_weight × salt_residual_activity × ph_activity_factor

Accessibility weights: buried (pLDDT ≥ 80) = 0.1; partially exposed (pLDDT 65–80) = 0.4; exposed (pLDDT < 65) = 1.0.

pH activity factors: ALP conservatively 1.0 (outside active pH 6–12), NPr conservatively 1.0 (at edge of pH 5–9), acid protease 0.30 (Koaze et al. 1964 activity-vs-pH curve).

Structural context

Signal peptide (aa 1–19, MKLVFLVLLFLGALGLCLA): pLDDT 35–54 throughout — fully disordered. This is the native mammalian secretory signal; it is co-translationally cleaved in the ER in the native context. In A. oryzae, heterologous signal peptide processing is common but not guaranteed for all foreign sequences.

Inter-lobe linker (approx aa 432–445): pLDDT 68–81 — partially exposed. Connects the N-lobe and C-lobe. Less constrained than either lobe core, giving some accessibility to recognition sites in this window.

Lobe cores (majority of residues): pLDDT 88–99 — well-folded, comparable to uricase. Protease risk in the lobe cores is low.

Per-protease results

Protease Total sites Exposed (full) Exposed (mature) Partly exposed Max risk (full) Max risk (mature)
ALP (alkaline subtilisin) 492 21 3 4 0.188 0.188
NPr (neutral metalloprotease) 234 14 0 1 0.388 0.155
acid_protease (aspergillopepsin) 128 10 0 0 0.195 0.020

ALP's 3 mature-protein exposed sites (pLDDT < 65, outside signal peptide) drive the MODERATE verdict for mature lactoferrin. NPr has the highest full-sequence score (0.388) but zero exposed sites in the mature protein — its entire risk comes from signal peptide access.

Comparison with uricase (comp-001)

Feature Uricase (comp-001) Lactoferrin (comp-005)
Length 301 aa 710 aa (incl. signal peptide)
Mean pLDDT 97.1 95.0
Min pLDDT 80.5 35.8 (signal peptide)
% residues pLDDT > 80 100% 96.1%
Signal peptide None (mature form only) Present (uncertain processing)
Quaternary structure Homotetramer (buries additional surface) Bilobal monomer (exposed linker)
Full-sequence verdict LOW HIGH
Mature-protein verdict LOW MODERATE

The key structural difference: uricase has no disordered regions and no signal peptide in the analyzed sequence — it is all compact, homotetramer-ready structure. Lactoferrin's soft spots are the signal peptide and the inter-lobe linker.

Impact on §1.10 experimental framing

Unlike the uricase arm, the comp-005 result does not shift §1.10's lactoferrin arm from a feasibility gate to a confirmation experiment. The MODERATE mature-protein verdict (0.188 max, ALP) and the signal-peptide uncertainty together mean the wet-lab result is the primary determination.

If wet-lab §1.10 shows lactoferrin degradation while uricase survives: - Most likely cause: ALP attack on 3 mature-protein exposed sites, and/or NPr attack if signal peptide is retained. - First diagnostic: Western blot for characteristic ~40 kDa N-lobe + ~40 kDa C-lobe cleavage products (inter-lobe linker cleavage) vs. intact 80 kDa band. - Second diagnostic: N-terminal sequencing to determine whether the 19 aa signal peptide was processed.

If wet-lab shows lactoferrin stability comparable to uricase: - The mature-protein MODERATE verdict would be overstated — possibly because ALP's pH factor (conservatively 1.0) overcounts its activity at pH 4.5–5.0, where ALP is outside its active range. - The correct update: revise ALP pH factor down (to ~0.1–0.3) and rerun comp-005; expect LOW mature-protein verdict.

Key limitations

  • pLDDT ≠ SASA. Signal peptide pLDDT 35–54 reliably predicts disorder; inter-lobe linker pLDDT 68–81 indicates partial exposure but solvent-accessible surface area (SASA) calculation would sharpen the accessibility score. A structure-based SASA analysis could refine the linker risk.
  • Signal peptide processing uncertain. A. oryzae has signal peptidase activity, but cleavage efficiency for heterologous signal sequences varies. The two verdicts bracket the expected range.
  • Glycosylation not modelled. Human lactoferrin has three characterized N-glycosylation sites (N137, N478, N623). N-glycans sterically shield backbone near Asn, and A. oryzae expression may hyperglycosylate the protein further — potentially shielding surface-accessible mature-protein sites. This biases the MODERATE verdict toward overestimating risk.
  • P1/P1' rules only. Extended subsite specificity (P2–P4) not modelled; may over-count recognition sites.
  • ALP and NPr pH factors conservatively set to 1.0. True activity at pH 4.5–5.0 is lower — likely ~0.1–0.2 for ALP (outside range 6–12) and ~0.5–0.7 for NPr (at lower edge 5–9). Risk is conservatively overstated for both.
  • Iron-binding state not modelled. Apo-lactoferrin (iron-free) is more open than holo-lactoferrin; iron availability in shio-koji is uncertain. This analysis models the AlphaFold holo-like structure.

Cross-references