ARHGAP11B: a GAP-dead RhoGAP that still has its arginine finger

All numbers below are produced by analyze_arhgap11b.py in this directory.
Re-run it to regenerate this file; do not hand-edit.

uv run python analyze_arhgap11b.py
uv run python analyze_arhgap11b.py --self-test

Subjects

The GTPase in the control structure is RHOA_HUMAN (P61586).

1. Where ARHGAP11B stops being ARHGAP11A

Colinearity proved (global alignment of the N-terminal regions opens no gaps): True.
No identity cutoff is chosen anywhere. The boundary is the single changepoint that maximises the between-segment sum of squares of the per-residue identity vector, and its significance is established by permuting that vector 2000 times: observed statistic 34.1653, best statistic over permutations 3.2891, empirical p = 5.00e-04. The script raises rather than reporting the argmax of noise if any permutation matches the observed statistic.

Derived divergence point: residue 220.

2. The arginine finger is retained

A control's own annotated arginine finger is aligned onto ARHGAP11B and scored as retained only if the aligned residue is an arginine and the aligned position lands on one of ARHGAP11B's own annotated Site features.

So the residue test that normally exposes a pseudo-enzyme returns retained here, for a protein that two independent experimental annotations record as GAP-dead. The hypothesis that ARHGAP11B lost catalysis by losing its arginine finger is not confirmed.

3. How much does the arginine-finger test discriminate?

All 66 reviewed (Swiss-Prot) human proteins carrying the PROSITE RhoGAP profile PS50238 were fetched. This is the Swiss-Prot subset of the family, not the whole family.

The screen flags 6 of 66. ARHGAP11B is not among them. But a count of flagged proteins is not a count of confirmed pseudo-enzymes, so each flagged entry is asked what UniProt itself concluded, and on what evidence:

Only 1 of the flagged entries carries an experimentally supported (ECO:0000269) UniProt statement that the domain is catalytically inactive for want of the arginine.

The sharper result is that residue identity and curated activity are decoupled in both directions within this family. ARHGAP11B keeps the arginine and is experimentally GAP-dead. RHG36_HUMAN (ARHGAP36) has a threonine at its own annotated arginine-finger position and UniProt nonetheless asserts GTPase activator activity for it — by similarity, ECO:0000250, which is the weakest thing UniProt says. That is the same class of defect as the one this review corrects, pointing the other way.

So the catalytic-residue screen could not have predicted ARHGAP11B's inactivity, and would not have been decisive even where it fires. A curation pipeline that gates a GAP-activity term on arginine-finger presence keeps the term on this protein.

4. What the truncation actually removes

The GAP:GTPase interface is computed from PDB 1TX4 (chain A = RHG01_HUMAN, chain B = RHOA_HUMAN), a transition-state mimic containing ALF, GDP, MG. Chain roles are proved, not assumed: chain A is 99.5% identical to Q07960 and chain B is 99.4% identical to P61586. Contacts are every GAP residue with an atom within 4.5 A of RhoA or of the nucleotide/metal/fluoride.

Sanity check on the geometry: the calculation recovers the control's own annotated arginine finger [282] among its contacts, as it must.

Those 25 contact positions are projected onto ARHGAP11A (global, control Rho-GAP domain vs paralog Rho-GAP domain) and split at the divergence point derived in section 1. The projection is accepted only because it is in register: the control's arginine finger (282) maps onto ARHGAP11A's own annotated arginine finger (87). Without that reciprocal anchor a global alignment of these two proteins lands the whole interface in the wrong half of ARHGAP11A, because ARHGAP1 carries its Rho-GAP domain at the C-terminus and ARHGAP11A at the N-terminus.

Interface positions ARHGAP11B does not have (ARHGAP11A numbering):

The truncation therefore removes 2 of 24 mapped GAP:GTPase interface positions (8%) while leaving the arginine finger in place. The catalytic residue survives; part of the surface that has to present it to the GTPase does not.

5. The UniProt Rho-GAP domain boundary

UniProt's Rho-GAP DOMAIN feature on ARHGAP11B runs 30 residues past the last residue that is homologous to ARHGAP11A, i.e. into the human-specific frameshift C-terminus. The annotated domain on the truncated paralog is therefore 11 residues longer than the annotated domain on the catalytically active parent. Any pipeline reading domain extent as evidence of a working RhoGAP will read this protein as more intact than the active one, not less.