Human protein complex interfaces reveal candidate assembly roles for initiator methionine excision

Key results

1. 01

   ShownN-terminal positions in the seven MetAP-permissive residue-2 classes (Ala, Cys, Gly, Pro, Ser, Thr, Val) are less often buried at assembly interfaces than nonpermissive ones. The association holds after adjusting for assembly size and after boundary filtering.

   37.5% vs 47.5% reach ≥25% burial (Fisher odds ratio (OR) 0.662, p = 9.5 × 10⁻⁴; 696 vs 427 proteins); median burial 1.4% vs 20.1% (Mann–Whitney p = 5.9 × 10⁻⁴); log chain count-adjusted OR 0.652 (95% CI 0.509–0.837); first-observed subset of 845 proteins, adjusted OR 0.669 (0.506–0.886); OR 0.528 in assemblies of ≥12 chains, 0.686 (p = 0.17) in the smallest. · MetAP-permissive classes engage interfaces less…; Abstract

2. 02

   ShownDespite the overall shift, a sizeable subset of permissive-class proteins buries the selected N-terminal residue in at least one modeled assembly. This subset is a candidate list for testing whether processing affects assembly.

   261 of 696 permissive proteins reach ≥25% burial. 175 remain after dropping observations whose selected residue is not the first modeled one. Of the 261, 147 are measured at the removal-annotation-selected residue (residue 2) and 114 at modeled methionine (109 MET, 5 MSE); 126 are engaged in ≥75% of their extracted observations. · A model-based candidate set and its…; Recurrence within selected observations versus…

3. 03

   SuggestedComplex-level enrichment is fragile: proteasome enrichment weakens after boundary filtering, after which no curated complex passes catalogue-wide correction. A histone-symbol group stays enriched, but several maxima come from histone tail peptides bound to recognition proteins, not nucleosomes.

   Proteasome/assembly prefix group 12/15 engaged (OR 6.94, Benjamini–Hochberg (BH) q = 0.0114) fell to 7/10 (OR 3.96, q = 0.202); Complex Portal 26S proteasome (CPX-5993) 10/13 (q = 0.0309) became q = 0.481 (original Benjamini–Yekutieli q = 0.147); histone-symbol group 9/11 (q = 0.0216) then 9/9 (q = 0.00180), e.g. H3(1–15) on the ATRX ADD domain (PDB 3QLA). · Group-level associations depend on membership…; Histone-tail recognition and spliceosomal…

4. 04

   SuggestedPartner-resolved geometry of proteasome α subunits places modeled N-terminal positions in different partner environments. PSMA7 Ser2 faces either PSMA2 or PSMA5, the PSMA6 Met1 backbone contacts the assembly chaperone PSMG1, and PSMA4 Ser2 is packed among several α subunits.

   In 22 observations (seven entries): PSMA7 in 6KWY/Q approaches PSMA2 at 2.92 Å (17.3% pair-only burial) vs PSMA5 at 3.27 Å in 8YIX/C (77.2%); PSMA6 Met1 in pre-13S 8TM4/G approaches PSMG1 at 3.27 Å (98.2% backbone burial from PSMG1 alone); five complete PSMA4 Ser2 observations show 67.6–90.7% total burial, with PSMA2 the largest single occluder (37.0–43.0%), whereas pre-13S 8TM4/B shows only 2.8%. · Proteasome terminal evidence and structural…; Geometry

5. 05

   ShownAmong permissive-class proteins, the potential to form a Met-Φ degron if methionine is retained shows no detectable association with interface engagement. Degron formation and interface shielding are therefore treated as separable candidate mechanisms.

   352 degron-forming proteins (Ala or Val at residue 2) vs 344 other permissive proteins: mean burial 26.3% vs 24.4%; 38.6% vs 36.3% at ≥25% burial (OR 1.10, 95% CI 0.81–1.50; Fisher p = 0.58); class-by-degron interaction after size adjustment, likelihood-ratio p = 0.89. · Conditional-degron potential and interface…; Classification