VCP/p97 (P55072) Bioinformatics Analysis Results

Checklist

[x] None of the scripts use hardcoded inputs or outputs (all take CLI arguments)
[x] Scripts tested on another input (NSF/P46459, a related type II AAA+ ATPase)
[x] Analyses completed as expected
[x] Direct results of the scripts are in results/ directory
[?] Conservation analysis uses pairwise alignment rather than full MSA (MUSCLE/MAFFT not available); this approximation may underestimate conservation in gapped regions
[?] Arabidopsis CDC48A (682 aa) and Gallus VCP (573 aa) sequences from UniProt appear truncated, which may skew conservation estimates downward
[x] Summary includes detailed provenance and justification

Overview

VCP (Valosin-Containing Protein, also known as p97 or CDC48 in yeast) is a type II AAA+
ATPase (EC 3.6.4.6) that forms a homohexameric ring and functions as a ubiquitin-selective
segregase. It extracts ubiquitinated proteins from membranes, chromatin, and protein
complexes for proteasomal degradation or other downstream processing. This analysis
examines its domain architecture, functional motifs, evolutionary conservation, disease
mutation distribution, and cofactor binding interfaces.

1. Domain Architecture and Functional Motifs

Script: analyze_domains.py
Results: results/VCP_motifs.tsv, results/VCP_domains.tsv, results/VCP_key_residues.tsv

Identified Domains

Domain	Start	End	Size (aa)	Evidence
N-domain	1	195	195	Inferred from Walker A position in D1
D1-ATPase	205	488	284	Walker A at 245-252
D2-ATPase	488	788	301	Walker A at 518-525
C-tail	789	806	18	Region after D2 ATPase domain

These boundaries are consistent with the known structure (PDB: 5FTK, 5B6C, etc.) and UniProt
annotations. The N-domain boundary at ~195 matches the known N-D1 linker region (aa 187-209).

Walker A/B Motifs

Motif	Domain	Positions	Sequence	Notes
Walker A (P-loop)	D1	245-252	GPPGTGKT	ATP binding; UniProt annotates binding at 247-253
Walker A (P-loop)	D2	518-525	GPPGCGKT	ATP binding; UniProt annotates binding at 521-526
Walker B	D1	300-305	IIFIDE	Contains catalytic D304/E305
Walker B	D2	573-578	VLFFDE	Contains catalytic D577/E578

Key Catalytic Residues

Residue	Domain	Role	Functional Importance
K251	D1	Walker A lysine	Essential for ATP binding. K251Q mutation impairs ERAD (UniProt MUTAGEN)
K524	D2	Walker A lysine	Essential for ATP binding. K524A impairs RNF19A activity (UniProt MUTAGEN)
E305	D1	Walker B glutamate	Catalytic base. E305Q + E578Q double mutant traps ATP-bound state
E578	D2	Walker B glutamate	Main catalytic base for ATP hydrolysis. E578Q increases CAV1/UBXN6 binding, impairs autophagy

The D2 domain is the primary catalytic ATPase. E578Q is the well-characterized "substrate trap"
mutation that locks VCP in an ATP-bound, high-affinity state for substrates.

2. Conservation Across CDC48/p97 Family

Scripts: fetch_orthologs.py, analyze_conservation.py
Results: results/VCP_pairwise_identity.tsv, results/VCP_conservation.tsv, results/VCP_motif_conservation.tsv

Pairwise Identity to Human VCP

Ortholog	Species	Identity
Q01853 (Vcp)	Mus musculus	100.0%
Q7ZU99 (vcp)	Danio rerio	96.8%
Q7KN62 (TER94)	Drosophila melanogaster	84.4%
P54811 (cdc-48.1)	Caenorhabditis elegans	80.2%
Q9P3A7 (cdc48)	Schizosaccharomyces pombe	73.5%
P25694 (CDC48)	Saccharomyces cerevisiae	72.0%
Q5ZL72 (VCP)	Gallus gallus	46.2% (partial sequence, 573 aa)
Q9SZJ3 (CDC48A)	Arabidopsis thaliana	44.4% (partial sequence, 682 aa)

VCP is extraordinarily conserved across eukaryotes. The 72% identity to S. cerevisiae CDC48
over ~800 aa (spanning >1 billion years of divergence) is among the highest conservation
levels observed for any eukaryotic protein, reflecting its essential and fundamental cellular
role.

Caveat: The pairwise identity was computed using Biopython's global PairwiseAligner
(scoring: match=2, mismatch=-1, gap_open=-3, gap_extend=-0.5). The Gallus and Arabidopsis
sequences appear truncated in UniProt, leading to artificially lower identity values.

Conservation at Functional Motifs

Motif	Domain	Sequence	Mean Conservation	Fully Conserved (>90%)
Walker A	D1	GPPGTGKT	0.893	No
Walker A	D2	GPPGCGKT	0.812	No
Walker B	D1	IIFIDE	0.905	No
Walker B	D2	VLFFDE	0.897	No

All Walker A/B motifs show high conservation (81-90%), consistent with their essential
catalytic roles. The slightly lower D2 Walker A conservation (0.812) may reflect the
inclusion of partial sequences that do not extend fully into the D2 domain.

Overall Conservation Profile

Mean conservation across all 806 positions: 0.778
Positions with >90% conservation: 193/806 (24.0%)
Positions with 100% conservation (identical in all orthologs): 193/806 (24.0%)

3. Disease Mutation Clustering

Script: analyze_disease_mutations.py
Results: results/VCP_mutations.tsv, results/VCP_mutation_summary.tsv

Mutation Distribution by Domain

Domain	Size (aa)	Disease Mutations	Density (per 100 aa)
N-domain	195	13	6.67
N-D1 linker	~10	1	~10
D1-ATPase	284	4	1.41
D2-ATPase	301	1	0.33
C-tail	18	0	0.00

Key finding: Disease mutations are heavily concentrated in the N-domain (68% of all
mutations, density 6.67 per 100 aa) and the N-D1 interface region. This is consistent
with the well-established observation that IBMPFD/MSP1 mutations cluster at the interface
between the N-domain and D1 ATPase ring, disrupting the conformational communication
between these domains.

Hotspot Residues

R155: The most frequently mutated position, with 5 different amino acid substitutions
(C, H, L, P, S) all causing IBMPFD1. R155 is conserved at 85.7% across orthologs.
R155H shows increased ATPase activity and impaired autophagy.
R159: Mutated to G (FTDALS6) or H (IBMPFD1). Conserved at 87.5%.
R191: Mutated to Q in IBMPFD1/FTDALS6. Fully conserved (100%) across all orthologs
examined, indicating strong functional constraint. Mutations at this position abolish
K315 methylation by ASPSCR1.

Disease Associations

IBMPFD1 (Inclusion Body Myopathy with Paget Disease and Frontotemporal Dementia):
16 of 19 variants. Predominantly N-domain and N-D1 interface.
FTDALS6 (Frontotemporal Dementia and/or ALS): 3 variants spanning N-domain (R159G,
R191Q) and D2 (D592N).
CMT2Y (Charcot-Marie-Tooth Type 2Y): 2 variants in N-domain (G97E, E185K).

Conservation at Disease Sites

Mean conservation at disease mutation positions: 0.747 (compared to genome-wide mean of 0.778).
The disease sites are somewhat less conserved than average, which is surprising but may reflect
that the N-domain, where most mutations cluster, has both conserved structural residues and
more variable surface residues involved in cofactor binding specificity.

Notably, the most pathogenic mutations (R155, R159, R191) affect positions with above-average
conservation (0.857, 0.875, and 1.000 respectively), consistent with disruption of conserved
structural contacts.

4. Cofactor Binding Interfaces

Script: analyze_cofactor_interfaces.py
Results: results/VCP_cofactor_interfaces.tsv, results/VCP_interface_summary.tsv

Interface Mapping from Mutagenesis Data

Domain	Cofactors Identified	Mutagenesis Sites
N-domain	DERL1, NPLOC4	7
D1-ATPase	RHBDD1, VCPKMT	9
D2-ATPase	CAV1, RHBDD1, RNF19A, SOD1, UBXN6, ZFAND1	2

N-domain Cofactor Interfaces

NPLOC4 (Npl4) binding: Residues 52-55 (FRGD motif) and Y110 are critical. FRGD->ARGA
combined with Y110A abolishes NPLOC4 interaction. The UFD1-NPLOC4 heterodimer is the
primary cofactor complex that recruits VCP to ubiquitinated ERAD substrates.

DERL1 (Derlin-1) binding: Multiple N-domain residues contribute: R113-H115, F131,
L140, H183. These map to a surface patch on the N-domain that contacts the ER membrane
retrotranslocation channel. D179 does not affect DERL1 binding, providing a negative
control.

D1 Domain Interfaces

VCPKMT (VCP lysine methyltransferase): K315 is the methylation site. K315L/Q/R
abolishes methylation, while neighboring residues (K312, R313, E314, T316, H317, G318)
are dispensable. IBMPFD mutations R95G, R159H, and R191Q (in the N-domain) abolish
K315 methylation enhancement by ASPSCR1, connecting disease pathology to post-translational
modification.

D2 Domain and C-tail Interfaces

E578Q (Walker B mutant): This substrate-trap mutation increases interaction with CAV1,
UBXN6, and ZFAND1 while impairing autophagy. This demonstrates that the ATP hydrolysis
cycle in D2 is directly coupled to cofactor release.

PIM motif (aa 802-806): The C-terminal PIM motif mediates binding to PUL domain-
containing cofactors (PLAA/UBXN6). The C-terminal region (797-806) is the documented
UBXN6 interaction site.

Disease Mutations and Cofactor Interfaces

The overlap between IBMPFD mutation sites and cofactor binding interfaces is notable:
- R95G (N-domain): Decreased interaction with CAV1 and UBXN6
- R155H (N-domain): Decreased interaction with CAV1 and UBXN6
- A232E (N-D1 interface): Decreased interaction with CAV1 and UBXN6
- R159H, R191Q (N-domain): Abolish ASPSCR1-mediated K315 methylation enhancement

This pattern supports the model that IBMPFD mutations disrupt the N-domain conformational
cycle, which in turn affects cofactor recruitment and release kinetics.

Methods and Provenance

Tools and Versions

Python 3.x with Biopython for sequence parsing and pairwise alignment
Regex-based motif scanning for Walker A (GxxxxGK[TS]) and Walker B ([VILMF]{4}DE) motifs
UniProt REST API for ortholog sequence retrieval
UniProt flat file parsing for variant and mutagenesis annotations

Data Sources

Human VCP sequence: UniProt P55072 (entry version 248, 2026-01-28)
Ortholog sequences: UniProt REST API (Q01853, P25694, Q9P3A7, Q7KN62, P54811, Q9SZJ3, Q5ZL72, Q7ZU99)
Disease variants and mutagenesis data: UniProt P55072 flat file FT annotations

Limitations

Conservation analysis uses pairwise alignment rather than full MSA. A proper MUSCLE or MAFFT
alignment would provide more accurate column-wise conservation scores, especially in
regions with insertions/deletions.
Two ortholog sequences (Gallus gallus and Arabidopsis thaliana) appear truncated in UniProt,
which reduces the effective number of orthologs for conservation calculations at C-terminal
positions.
The disease mutation analysis is limited to variants annotated in UniProt. ClinVar and HGMD
may contain additional pathogenic variants not included here.
Cofactor interface mapping is based solely on mutagenesis data in the UniProt flat file.
Structural analysis of co-crystal structures (e.g., PDB 5B6C for VCP-NPLOC4) would provide
more comprehensive interface residue identification.

Reproducibility

All scripts accept command-line arguments and can be rerun via:

just run-vcp

Individual steps can also be run independently -- see the justfile for recipes.
Scripts were validated on human NSF (P46459), another type II AAA+ ATPase, confirming
that the Walker A/B detection and domain assignment logic generalizes correctly.