RHOJ (TCL, TC10-like) encodes a Cdc42/TC10-related Rho-family small GTPase that is expressed predominantly in endothelial cells. The canonical 214-residue product (UniProt Q9H4E5) undergoes rapid GDP/GTP exchange and hydrolyses GTP, is palmitoylated near its N-terminus and prenylated at a C-terminal CAAX box, associates with the plasma membrane, binds CRIB-domain effectors such as PAK kinases and WASP, and promotes F-actin-rich protrusions; it is required for endothelial lumen formation, migration and angiogenic vascular morphogenesis. The entry described here, G3V4H1, is a 153-residue alternative gene model translated from an Ensembl transcript annotated for nonsense-mediated decay. It is identical to the canonical product for residues 1-103, which carry the P-loop and switch I/II regions, and then continues into a 50-residue proline-rich tail read from the canonical 3' untranslated region. It therefore lacks the G4 and G5 guanine-base motifs, the C-terminal half of the Ras-like fold, the polybasic region and the CAAX prenylation site. No function has been demonstrated for this polypeptide, and it is not known to accumulate in cells.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0003924 GTPase activity | IEA GO_REF:0000120 | MARK AS OVER ANNOTATED | Summary: GTPase activity is directly established for the canonical RHOJ/TCL product (Q9H4E5), but this IEA is carried by G3V4H1, an NMD-transcript gene model whose Ras-like domain is truncated after residue 103. Reason: The function is true of the RHOJ gene. Full-length TCL hydrolyses GTP in vitro (PMID:10967094), and the whole Rho family shares GTPase activity (PAINT root node PTN000632671 of PTHR24072, whose IBD donors include Q9H4E5). The G3V4H1 entry receives the term only from InterPro/UniRule signature matches to its N-terminal 103 residues. Re-analysis shows that the product keeps the P-loop and switch I/II elements but lacks the G4 and G5 guanine-base motifs and the whole C-terminal half of the fold. It is translated from ENST00000557447, annotated for nonsense-mediated decay, and its tail is read from the canonical 3' UTR. An incomplete G-domain gives no basis for catalysis, so the term overstates what this product is known or likely to do. The gene-level GTPase function belongs on the canonical Q9H4E5 entry. REMOVE is not used, because no assay of this exact product exists and the term is correct for the gene. Supporting Evidence: PMID:10967094 In vitro, TCL shows rapid GDP/GTP exchange and displays higher GTP dissociation and hydolysis rates than TC10. file:human/RHOJ/RHOJ-bioinformatics/RESULTS-rereview.md G3V4H1 is the translation of Ensembl transcript ENST00000557447 (RHOJ-204), biotype nonsense_mediated_decay, not the Ensembl canonical transcript. file:human/RHOJ/RHOJ-bioinformatics/RESULTS-rereview.md | G4 (T/N)(K/Q)xD guanine-base motif | 133 TQID | absent | file:human/RHOJ/RHOJ-bioinformatics/RESULTS-rereview.md | G5 (C/S)SA(L/K) guanine-base motif | 175 CSAL | absent | |
| GO:0003925 G protein activity | IEA GO_REF:0000003 | MARK AS OVER ANNOTATED | Summary: This EC-mapping IEA follows from the ARBA-assigned name "small monomeric GTPase" (EC 3.6.5.2). It describes the canonical RHOJ product but is not supported for the truncated G3V4H1 gene model. Reason: G protein activity means acting as a nucleotide-regulated molecular switch that signals to effectors in its GTP-bound state. For canonical RHOJ this is supported by its GDP/GTP cycling (PMID:10967094). Plasma-membrane targeting and nucleotide loading depend on its N terminus and on a distal loop (residues 121-129) (PMID:27660391). G3V4H1 keeps the N terminus and switch regions but lacks the 121-129 loop region, the G4/G5 guanine-base motifs and the CAAX membrane anchor. It also comes from an NMD-targeted transcript. The EC number on the G3V4H1 record comes only from an ARBA rule, and an ARBA assignment is not independent evidence. The term is right for the gene but over-annotated for this product. Supporting Evidence: PMID:27660391 Chimeras of TCL and TC10 revealed amino acids 121-129 of TCL contributed to the differences in nucleotide loading. file:human/RHOJ/RHOJ-bioinformatics/RESULTS-rereview.md | C-terminal CAAX box | 211 CSII | absent | file:human/RHOJ/RHOJ-bioinformatics/RESULTS-rereview.md G3V4H1 retains the phosphate/Mg2+-binding elements (P-loop, switch I, switch II) but lacks the G4 and G5 motifs that contact the guanine base, the C-terminal half of the Ras-like fold (beyond codon 103), the polybasic region and the CAAX prenylation site. |
| GO:0005525 GTP binding | IEA GO_REF:0000120 | MARK AS OVER ANNOTATED | Summary: Canonical RHOJ binds GTP, but G3V4H1 keeps only the phosphate- and Mg2+-binding half of the nucleotide pocket. The guanine-base contacts (G4 TQxD and G5 SAL) are missing. Reason: In Rho/Ras GTPases, specific guanine-nucleotide binding needs the P-loop and switch residues that coordinate the phosphates and Mg2+. It also needs the G4 and G5 motifs on the beta5-alpha4 and beta6-alpha5 loops, which recognise the guanine base. G3V4H1 retains the first set and lacks the second, and with them the C-terminal half of the fold that holds them. Q9H4E5 nucleotide-binding feature 136 (in the G4 motif) has no counterpart in G3V4H1. Signature matches over the N-terminal 103 residues therefore do not show that this product binds GTP with the specificity the term implies. The term stays correct for the RHOJ gene and the canonical Q9H4E5 product. Supporting Evidence: PMID:10967094 In vitro, TCL shows rapid GDP/GTP exchange and displays higher GTP dissociation and hydolysis rates than TC10. file:human/RHOJ/RHOJ-bioinformatics/RESULTS-rereview.md G3V4H1 retains the phosphate/Mg2+-binding elements (P-loop, switch I, switch II) but lacks the G4 and G5 motifs that contact the guanine base, the C-terminal half of the Ras-like fold (beyond codon 103), the polybasic region and the CAAX prenylation site. file:human/RHOJ/RHOJ-bioinformatics/RESULTS.md | Binding site: | 136–136 | 0 / 1 | 0 | |
Loading supporting content…
Download this section (compressed HTML)Q: Does any G3V4H1 polypeptide (residues 1-103 of RHOJ plus a 3'-UTR-encoded proline-rich tail) accumulate in endothelial cells, or is the ENST00000557447 transcript removed by nonsense-mediated decay before appreciable translation? Proteomic support for this entry could come from peptides in the shared residues 1-103 rather than from the unique tail.
Q: Should the gene-level RHOJ GTPase annotations (GTPase activity, GTP binding, plasma membrane, angiogenesis-related processes) be curated on the canonical Q9H4E5 entry instead of on electronically annotated alternative gene models such as G3V4H1?
Experiment: Targeted mass spectrometry or immunoblotting with an antibody raised against the unique G3V4H1 C-terminal tail in HUVECs, with and without NMD inhibition (UPF1 knockdown or cycloheximide), to test whether the product is made at all.
Experiment: If the product is detectable, express recombinant G3V4H1 and measure mant-GTP binding and intrinsic GTP hydrolysis beside full-length RHOJ, to test whether the half G-domain binds guanine nucleotide.
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)