CRPPA

Summary: The broad conserved-site catalytic assertion can be refined to the established CDP-ribitol reaction.

Reason: The IPR018294 conserved catalytic site is genuine, but human biochemical studies resolve the predominant physiological reaction as D-ribitol-5-phosphate cytidylyltransferase activity. Use GO:0047349 for this broad catalytic assertion; the independently retained GO:0070567 assertions also record measured pentose-phosphate substrate breadth.

Proposed replacements: D-ribitol-5-phosphate cytidylyltransferase activity

Summary: The actual ancestral cytosol assertion is inherited by human CRPPA.

Reason: PTN000440569 is an actual ancestor of human PTN002534301 in the 65-node PTHR43015 tree. The literal GO:0005829 IBD is grounded in human experimental PMID:26687144, and the ungapped target MSA exactly matches the 451-residue displayed product. Target self-evidence is legitimate ancestral grounding, not circularity. Tagged human-cell fractionation independently supports the compartment.

Summary: Tagged human CRPPA localizes to the cytosolic fraction.

Reason: Full PMID:26687144 Results and Figure 3A report cytosolic fractionation of inducible C-terminal Myc-His-tagged full-length human protein in HEK293, with Hsp90/calnexin/ATP5H controls. Immunostaining is corroborative. This is tested recombinant-expression localization, not an endogenous untagged distribution survey.

Summary: The UniProt subcellular-location source supports cytosol.

Reason: The actual SL-0091 mapping agrees with official human CRPPA localization and the tagged full-length HEK293 fractionation experiment in PMID:26687144. This source is not inferred solely from the absence of a signal peptide.

Summary: The curated Reactome reaction places the CRPPA catalyst in cytosol.

Reason: Reactome R-HSA-9940844 explicitly joins catalyst 9940820 to CRPPA dimer 9940689, whose two component entries are 9940850. That human protein resolves to UniProt A4D126 and GO:0005829. The reaction transfers cytidylyl to ribitol phosphate; the compartment assignment agrees with direct tagged human-cell localization.

Summary: The conserved-site isoprenoid-pathway inference does not describe the human enzyme.

Reason: IPR018294 maps a conserved bacterial-type cytidylyltransferase site to isoprenoid biosynthesis. PMID:26687144 instead establishes pentose-phosphate nucleotide-sugar synthesis and unchanged dolichol species/levels in two independently generated HAP1 knockout clones. Its basal in-vitro MEP turnover is acknowledged; a minor substrate reaction does not establish operation of the absent mammalian MEP pathway. The source mapping, not the historical alias, is the defect.

Summary: The supported glycoprotein-biosynthesis inference can be specified to the established O-mannosyl-glycan pathway.

Reason: All 67 current ARBA00027560 condition sets were evaluated against official target features and lineage. Only branch 26 matches, through FunFam 3.90.550.10:FF:000080 and Euteleostomi, and it emits GO:0009101. This broad source assertion remains valid. Current ontology ancestry places GO:0035269 beneath GO:0009101, and direct human donor-synthesis experiments plus the existing experimental glycan-pathway assertions establish this more informative process. CRPPA synthesizes the activated CDP-ribitol donor used in elongation of the mannose-initiated alpha-dystroglycan glycan; it does not itself catalyze the initial mannose transfer or the downstream FKTN/FKRP transfer. Refine the review to GO:0035269 while preserving the literal rule output and positive transfer status.

Proposed replacements: protein O-linked glycosylation via mannose

Summary: The metazoan ancestral glycosylation assertion is grounded in human and zebrafish experiments.

Reason: PTN001452245 is an actual ancestor of the human target, with a literal GO:0035269 IBD citing human A4D126 and ZFIN:ZDB-GENE-061110-16. Complete current donor records include human IMP studies and the zebrafish IMP from PMID:22522421, with MO1 directly reducing IIH6-reactive alpha-dystroglycan and both morpholinos corroborating structural, sarcolemma and interaction phenotypes. Human biochemical work identifies the performed step as activated CDP-ribitol donor synthesis. The current term includes further glycan elongation, not only the first mannose transfer. No target-path loss record or target-specific functional divergence was found.

Summary: Human patient-cell complementation supports CRPPA participation in the O-mannosyl glycan pathway.

Reason: PMID:22522420 identifies recessive human variants and shows rescue of functional alpha-dystroglycan glycosylation by wild-type CRPPA but not the patient exon-3-deletion construct. Its original POMT-activity interpretation is not treated as proof that CRPPA performs the first mannose-transfer reaction: later PMID:26687144 reports retained POMT activity in other deficient cells and establishes nucleotide-sugar synthesis. The pathway annotation remains supported by the genetic rescue plus the independently resolved CDP-ribitol reaction.

Summary: Human genetics and zebrafish knockdown support the same glycan-biosynthesis role.

Reason: The original human cohort has biallelic CRPPA variants and reduced IIH6-reactive alpha-dystroglycan in muscle. In zebrafish, two splice-blocking morpholinos targeting common transcript regions produce concordant structural and sarcolemma defects, with control morpholinos, splice tests and interaction experiments. The direct IIH6 glycan-loss measurements use MO1, including the FKTN/FKRP co-injection assays. These are model perturbations rather than purified human catalysis; the later human enzyme studies provide the performed donor-synthesis step. The direct human and model scopes remain distinct.

Summary: Recombinant human CRPPA forms a homodimer.

Reason: The human 43–451 recombinant crystal forms a homodimer through the extended arm; the original reports consistent solution gel filtration, with those solution data not shown. This is positive homotypic association evidence, not proof of endogenous cellular stoichiometry for every splice product. Retain the structural association as non-core: the available dimer experiments characterize the enzyme assembly but do not establish an additional primary molecular activity distinct from the catalytic donor-synthesis function.

Summary: Human recombinant CRPPA directly synthesizes CDP-ribitol.

Reason: Original PMID:26923585 Results/Figure 4 and enzyme Methods directly test His-tagged human ISPD purified from HEK293T. Ribitol5P plus CTP yields the HPLC/MS product shared with bacterial TarI; controls omit substrate and mutant comparisons control protein amounts. Rescue by supplied CDP-ribitol separately links this reaction to glycan production.

Summary: Human CRPPA residues 43–451 directly synthesize CDP-ribitol.

Reason: Original 27130732 directly tests purified N-terminal His8/TEV human 43–451 expressed in HEK293F. MS/MS detects CDP-ribitol only with enzyme, CTP and ribitol5P; free ribitol is inactive in that assay. This is direct target catalysis despite the truncated construct.

Summary: The specific catalytic IBD is inherited and directly corroborated in human CRPPA.

Reason: The exact GO:0047349 IBD resides at PTN001452245, on the authenticated human ancestral path. Its human A4D126 descendant evidence includes the direct 26687144, 26923585 and 27130732 enzyme studies. Self-evidence is valid experimental grounding; it is not counted as an independent replicated transfer. The exact human MSA and present biochemical activity support retention.

Summary: Human CRPPA produces chemically characterized CDP-ribitol.

Reason: Original PMID:26687144 human enzyme assays, chromatography, MS and NMR establish CDP-ribitol formation. Full-length 1–451 E. coli protein and 43–451 Sf9 protein have separately stated uses; the latter produces the verified ribitol product. Product identity is supported beyond a generic pyrophosphate readout.

Summary: The RHEA and EC sources specify the established ribitol-phosphate cytidylyltransferase reaction.

Reason: The exact source RHEA:12456/EC:2.7.7.40 pair matches the current GO:0047349 reaction and the authenticated UniProt catalytic-reaction entry. Its D-ribitol5P input and CDP-L-ribitol product naming do not imply an extra epimerase reaction. Direct human biochemical studies independently corroborate the mapping.

Summary: Human CRPPA has experimentally demonstrated activity toward multiple pentose phosphates.

Reason: Original PMID:26687144 demonstrates actual human activity toward ribose5P and ribulose5P as well as ribitol5P, with product characterization; PMID:27130732 independently tests ribose5P. GO:0070567 correctly accommodates this measured breadth. Do not replace every broad assertion with the ribitol-only child or infer an additional autonomous core from in-vitro promiscuity.

Summary: The broader family cytidylyltransferase mapping matches measured human activity.

Reason: IPR034683 is the broader IspD/TarI cytidylyltransferase family and currently maps to GO:0070567. Its description explicitly includes the human pentose-phosphate activity. This broad molecular-function inference is valid, unlike the narrower bacterial-pathway implication of IPR018294.