Pantoate-derived coenzyme A biosynthesis

The module models eight reaction activities in either complete route and keeps molecular functions on leaf annotons. It intentionally has no module-level cellular location: bacterial and archaeal enzymes are commonly cytosolic, whereas eukaryotic isoforms and fused CoA synthases can occupy different compartments. Family selectors identify established catalytic implementations, while the required molecular function remains the decisive criterion. Related reductase families should not satisfy the pantoate step without substrate-specific evidence. Taxon-specific precursor sources, paralogs, and pathway-map spillover belong in pathway-instance reviews rather than this module. ALIGN with YeastPathways (2026-10): the ten PANTOSYN2-PWY reactions map onto the module, and the two-reaction PWY-5760-1 frame is folded in as the polyamine-derived beta-alanine supply variant (FMS1 then ALD2/ALD3); a PanD variant was added alongside it so the optional supply part is cross-taxon. The YeastCyc catalyst lists are correct, but YeastCyc lists CAB3, SIS2 and VHS3 as alternative PPCDC enzymes; the decarboxylase step was re-modeled with an architecture variant set so the budding-yeast Cab3-Sis2/Vhs3 heterotrimer (PMID:19915539) is distinct from homo-oligomeric PPCDCs, including the S. pombe Hal3-thymidylate synthase fusion protein (PMID:23962284). A standalone fungal PPAT (Cab4-type) variant was added because yeasts encode the PPAT and dephospho-CoA kinase as separate proteins rather than as fused CoA synthase. Compartments: both GO-CAMs place the late CoA steps in the cytosol; the PomBase model puts ecm31 and pan5 in mitochondria, and the S. cerevisiae ECM31 gene review likewise places Ecm31 in mitochondria, whereas YeastPathways places it in the cytosol; Cab1 is a separate cytoplasmic, rate-limiting kinase outside the CoA-synthesizing protein complex (GO:1990143, Cab2, Cab3, Cab4, Cab5 with Sis2/Vhs3), and Cab5 sits on the cytosolic face of the mitochondrial outer membrane. Because fungal and bacterial compartments differ, these locations are recorded here rather than as module-level context. GO-CAM disagreements: YeastPathways types FMS1 as N(1)-acetylpolyamine oxidase (GO:0052903) instead of spermine oxidase (GO:0052901) and ALD2/ALD3 as generic GO:0004029 instead of GO:0102244; PomBase 678073a900002636 assigns dephospho-CoA kinase activity to cab4, a PPAT-only protein, and includes acetyl-CoA synthesis and mitochondrial CoA transport, which are outside the module. PAN5 deletion is tolerated in S. cerevisiae (PMID:19266201), so a second ketopantoate reductase may exist in yeast. Exemplar and family decisions (2026-11): M. tuberculosis CoaX (P9WPA1) was replaced as a type III PanK exemplar by B. subtilis CoaX (P37564), because Mtb CoaX is a pseudopantothenate kinase that regulates depupylation of PanB rather than phosphorylating pantothenate (PMID:39585979); the PTHR34265 family review records the six sequence-identical MTBC CoaX entries as member exceptions. The CoaD selector was narrowed from PTHR21342 to PTHR21342:SF1 because SF0 holds archaeal NMN adenylyltransferases. S. pombe hal3 is selected as a gene product because PANTHER places the Hal3-thymidylate synthase fusion in the thymidylate synthase family PTHR11548. PAINT nodes PTN000470402, PTN000466854, PTN000480612 and PTN000482076 (also proposed in PR #5076) were verified against the local PAINT files and added. PR #5076's PANTHER terms for Fms1 (PTHR10742:SF410) and Ald2/Ald3 (PTHR43720) were not adopted: both are functionally mixed groups named for other enzymes (see the PTHR43720 family review), so these selectors rely on the required molecular function and representative members. GO has no terms for pantoate kinase (EC 2.7.1.169) or 4-phosphopantoate--beta-alanine ligase (EC 6.3.2.36) activity; the archaeal annotons use GO:0016773 and GO:0016881 as the closest existing terms. The TK2141 and TK1686 gene reviews MODIFY those generic rows to NTR replacements and carry the proposed terms as proposed_molecular_function, so the validator reports FUNCTION_UNREVIEWED for pok_activity (GO:0016773, the proposed term's parent) and REVIEW_DISAGREEMENT for pps_activity; the GO ids are kept here because leaving the annotons ungrounded only replaces those two warnings with eight NO_FUNCTION_ID warnings. Three further module-validator warnings remain by design: PTHR12280 pantothenate kinase activity is residue-determined because PANK4 pseudokinases share SF20 with Cab1 and ptk1, and the PTHR10695 review leaves GO:0004595 UNRESOLVED at family level because PPAT-domain and kinase-only members share SF46; in both cases the family reviews state that the module's annotons are correct for their representative members.

Connections