PilU (PA0396) is a 382-residue cytoplasmic PilT-like ATPase of the GspE / secretion NTPase superfamily, encoded immediately downstream of the main type IV pilus retraction ATPase gene pilT. It has the Walker A/B, Asp box and His box motifs of the family, hydrolyses ATP in vitro and forms a hexameric motor ring. PilU localizes only to the piliated cell pole and acts as an accessory retraction ATPase that works in a PilT-dependent manner, improving pilus retraction, particularly under high load. pilU mutants are hyperpiliated and lose twitching motility but still retract pili well enough to remain sensitive to pilus-specific phages; they also have elevated cAMP, linking PilU to pilus-dependent surface sensing, and reduced virulence in animal models. The requirement for PilU in twitching differs between species and can be bypassed by some pilin and PilY1 alleles.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0005524 ATP binding | IEA GO_REF:0000002 | ACCEPT | Summary: InterPro2GO mapping from IPR006321 (PilT/PilU family). Reason: PilU has a canonical Walker A P-loop, hydrolyses ATP in vitro, and mutation of its Walker A residue abolishes function. ATP binding is a true and mechanistically necessary property of this motor ATPase. Supporting Evidence: PMID:18174131 Mutation of conserved WA or WB residues in any of the three ATPases abrogated twitching motility PMID:18174131 Assay of purified 6xHis-tagged PilB, PilT and PilU from P. aeruginosa showed that all three proteins have ATPase activities in vitro. |
| GO:0005737 cytoplasm | EXP PMID:15659660 Disparate subcellular localization patterns of Pseudomonas a... | ACCEPT | Summary: Functional YFP-PilU fusions are cytoplasmic and localize only to the piliated pole. Reason: Direct localization of a complementing fluorescent fusion in P. aeruginosa; this is where the ATPase acts, at the cytoplasmic face of the pilus machinery. Supporting Evidence: PMID:15659660 PilB and PilT fusions localized to both poles, while PilU fusions localized only to the piliated pole. |
| GO:0005737 cytoplasm | IEA GO_REF:0000044 | ACCEPT | Summary: UniProt subcellular-location mapping to cytoplasm, from the curated experimental location in the entry. Reason: Same claim as the EXP row; the UniProt location it maps from is itself experimental (PubMed:15659660). Supporting Evidence: PMID:15659660 Fluorescent fusions to two of the three ATPases (PilT and PilU) were functional, as shown by complementation of the corresponding mutants. |
| GO:0016888 DNA endonuclease activity, producing 5'-phosphomonoesters | IBA GO_REF:0000033 | REMOVE | Summary: PAINT IBA from PTN000766390, whose DNA endonuclease IBD has a single seed: the PseudoCAP IDA on this protein from PMID:18174131. Reason: The only seed is a PseudoCAP row that keyed GO:0016888 where the paper reports ATP hydrolysis (GO:0016887, an adjacent GO identifier). No member of the PilT/PilU family is known to be a nuclease, PilU has no nuclease domain, and nothing in the family supports DNA cleavage. The IBD therefore rests on a keying error and the IBA is an over-propagated wrong term, not a debatable node placement. Propagation Review Root cause: SOURCE BAD Failure modes: SOURCE MISCITATION Sources checked: PANTHER:PTN000766390 · PTN000766390 SOURCE BAD Sole IBD seed is the PseudoCAP IDA on G3XCX3 (PMID:18174131), which keyed GO:0016888 for GO:0016887 ATP hydrolysis activity. UniProtKB:G3XCX3 · PilU (PA0396) SOURCE BAD Seed row's paper reports ATPase activity of purified PilU, not nuclease activity. Supporting Evidence: PMID:18174131 Assay of purified 6xHis-tagged PilB, PilT and PilU from P. aeruginosa showed that all three proteins have ATPase activities in vitro. PMID:41055394 PilU is a hexameric motor ATPase with unique features. |
| GO:0016888 DNA endonuclease activity, producing 5'-phosphomonoesters | IDA PMID:18174131 Functional role of conserved residues in the characteristic ... | MODIFY | Summary: PseudoCAP IDA (2019-03-11) citing Chiang et al. 2008. The reference is correct for PilU, but the paper reports ATPase activity of purified PilU, not DNA endonuclease activity. Reason: The experiment behind this row is real and directly assays PilU: the abstract states that purified His-tagged PilU has ATPase activity in vitro. The term is wrong. GO:0016888 is the identifier next to GO:0016887 ATP hydrolysis activity, which PseudoCAP assigned from the same paper on the same date to PilB (P22608), so this is an off-by-one term keying error, not a misreading of the paper. The PMID itself is correct (it is not a mis-keyed neighbour of PMID:18174132, which is the unrelated ccmC paper). The annotation should be repointed to GO:0016887. Proposed replacements: ATP hydrolysis activity Supporting Evidence: PMID:18174131 Assay of purified 6xHis-tagged PilB, PilT and PilU from P. aeruginosa showed that all three proteins have ATPase activities in vitro. |
| GO:0043107 type IV pilus-dependent motility | IBA GO_REF:0000033 | ACCEPT | Summary: PAINT IBA from PTN000766390, sole-seeded by the PseudoCAP IMP on this protein (PMID:7854122). PilU appearing in its own WITH/FROM is expected for a seed. Reason: The seed is sound (see the IMP row), and PilU itself does part of the work of twitching as an accessory retraction motor. The node is a small PilU clade (13 leaves, 9 taxa). The term holds across most of it, though the need for PilU in twitching varies: P. stutzeri pilU mutants still twitch and Neisseria PilU has no clear twitching role, so the IBA is a reasonable but not universal clade-level claim. Supporting Evidence: PMID:7854122 The other three transposon mutations were found to have occurred within a new gene located directly downstream of pilT. PMID:31525185 Finally, we show that Pseudomonas aeruginosa PilU also functions as a PilT-dependent retraction ATPase PMID:11466271 Insertional inactivation did not affect piliation, twitching motility, or PO4 infection but reduced transformation to about 10%. |
| GO:0043107 type IV pilus-dependent motility | IMP PMID:7854122 Characterization of a gene, pilU, required for twitching mot... | ACCEPT | Summary: PseudoCAP IMP from Whitchurch and Mattick 1994: transposon insertions in pilU abolish the spreading-colony twitching phenotype; mutants are hyperfimbriate but remain phage-sensitive. Reason: Twitching loss in pilU mutants, with retained but hyper-abundant pili, fits PilU being a retraction motor that contributes to the retraction step of twitching; later work shows directly that P. aeruginosa PilU acts as a PilT-dependent retraction ATPase. This is participation in the process, not just a requirement for it. Supporting Evidence: PMID:7854122 Characterization of a gene, pilU, required for twitching motility but not phage sensitivity in Pseudomonas aeruginosa. PMID:7854122 Like pilT mutants, the pilU mutants were hyperfimbriate PMID:31525185 Finally, we show that Pseudomonas aeruginosa PilU also functions as a PilT-dependent retraction ATPase |
Loading supporting content…
Download this section (compressed HTML)Q: GO has no process term for type IV pilus retraction (or disassembly), so a retraction motor such as PilU can only be annotated to type IV pilus-dependent motility. Should a "type IV pilus retraction" term be added under pilus organization?
Q: How does PilU contribute to retraction mechanistically: does it form stacked or mixed assemblies with PilT hexamers at the platform protein PilC, or act sequentially, and why is it dispensable in P. stutzeri and with some PilA or PilY1 alleles?
Experiment: Measure ATP hydrolysis by purified PilU and PilT, alone and mixed, with Walker B and His box variants of each, and relate the rates to retraction force and speed measured in single-pilus assays of the matching P. aeruginosa mutants. This tests whether PilU ATPase activity drives retraction directly or only boosts PilT.
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)