Human ALPK3 is HGNC:17574 / UniProt Q96L96. The exact normal seed contained ten PENDING assertions. All source terms, evidence codes, original references, ordinary qualifiers and supporting entities are preserved. Current UniProt is sequence version 3, 1705 residues. Older papers use longer historical constructs and residue numbering; those coordinates are reported as the authors' coordinates and are not silently projected onto the current sequence. The historical alias MAK must not be confused with the separately approved human MAK gene.
The normal Falcon attempt, with the configured perplexity-lite fallback and installed client override, failed at DNS resolution and produced no report. Manual primary reading therefore supplies the research. The ordinary seed-publication command found no PMIDs because the seed references are GO_REF records. One finite additional normal fetch for the ten primary records below finished with exit 1 after 308.779 seconds: 0/10 records cached, all blocked by DNS resolution. The local command/output record documents that failure; it does not claim that the primary web route failed. Primary web access and a normal CLI cache fetch are distinct routes. The absence of canonical publication files remains a DRAFT gate even where the original article was read externally. No source text, raw annotation, provider output or cache is synthesized or overwritten.
PMID:36321451 DOI 10.1161/CIRCULATIONAHA.122.059688, original PMC article, Methods and Results/Figures 2–6. Human isogenic cardiomyocyte catalytic-cleft variants largely preserve the phosphoproteome, whereas loss of ALPK3 changes it substantially. Overexpression likewise fails to give a broad substrate phosphoproteomic signature. These support the pseudokinase interpretation, but are cellular rather than universal negative biochemical tests. Figure 6 uses transient ALPK3-FLAG in human cardiomyocytes, with a nuclear-envelope and M-band pool. HEK293T co-IP is positive for MYOM2, negative for MYOM1 despite colocalization; do not describe both as directly bound. Comparable cardiac marker maturation under the study's differentiation protocol is not proof that ALPK3 never contributes to subsequent cellular development.
PMID:39196058 DOI 10.1038/s44161-023-00219-9, original PMC article, recombinant-expression, global-kinase and targeted-kinase Methods plus Extended Data Figure 9. The positive preparation is an isolated His-tagged domain, reported residues 1577–1828, expressed in BL21(DE3), purified on Ni-NTA, washed and dialyzed. The global assay treats ALPK3-mutant human cardiomyocyte lysates with FSBA to inhibit endogenous kinases, removes FSBA, then compares added domain with water. The targeted assay uses SQSTM1-HA immunoprecipitated from HEK293FT cells, active versus heat-denatured domain and lambda-phosphatase controls. These are meaningful positive controls, but the target is an immunoprecipitate rather than a fully purified two-protein system. The T269/S272 antibody does not independently assign serine versus threonine specificity. The human endogenous-tagging and organoid experiments establish M-band function independently of this catalytic interpretation.
PMID:38048395 DOI 10.1161/CIRCULATIONAHA.123.065993, original PMC research letter, complete main text and Figure 1. Mouse ALPK3 domain 1339–1608 does not label heat-inactivated H9c2 or human cardiomyocyte lysates containing SQSTM1, whereas mouse TRPM7 is active. Mouse K1420R knock-in retains normal cardiac function and the measured SQSTM1 phosphosignal. These are substantive counterexperiments. They are not a negative test of every full-length human state, and failure to reproduce does not prove that the positive preparation was contaminated.
The three catalytic rows remain UNDECIDED. ATP binding is independently UNDECIDED: kinase inactivity does not imply lack of ATP binding, while an ATP-containing phosphotransfer reaction does not by itself measure affinity. No catalytic residue-loss claim is asserted against current coordinates. The curated human reaction is inferred from ALPK1/Q96QP1, not an ALPK3-specific human assay. The serine-specific source retains this actual donor identity.
PMID:11418590 DOI 10.1074/jbc.M100485200. Primary PubMed identity plus the original full article in an author-uploaded copy were read, including Methods and Figures 5–7. The cDNA is mouse Midori. HA/Myc-tagged localization is measured in COS cells; gain/loss experiments are stable sense/antisense mouse P19CL6 lines under DMSO differentiation. Beating foci, MF20 staining and myosin-heavy-chain protein support altered acquisition of cardiac features. Overexpression also increases death during differentiation; the proposed apoptosis mechanism is explicitly speculative, and no apoptosis NEW is made. Predicted nucleotide motifs and self-induced transcript abundance do not establish ATP binding or transcription-factor activity.
The MGI Alpk3 GO graph links nuclear IDA and cardiac differentiation/heart development IMP to J:71661, the Hosoda paper. It links cardiac muscle cell development IMP to J:222361, the knockout paper below. These are actual donor assertions. The two IBA rows still preserve PTN002931915; the exact ancestral IBD/tree was not recovered from local PANTHER data. A short donor list is not a weakness, and the target term is not rejected because node-level detail remains unreconstructed. The fetched PTHR47091 table contains Q96L96, but its unreviewed LLM description is not scientific evidence.
PMID:21441111 DOI 10.1177/0300985811402841. Primary PubMed abstract read; original full body not recovered. Mouse knockout cardiac imaging and microscopy show impaired contraction, intercalated-disc abnormalities and myofibril disarray. The normal Mendelian birth ratio and predominantly postnatal phenotype are retained; this does not directly establish a particular early patterning step.
PMID:26846950 DOI 10.1016/j.jacc.2015.10.093. Primary PubMed abstract read: five affected patients from three families, severe pediatric disease and myocardial/intercalated-disc histology. The clinical association corroborates function but does not independently demonstrate kinase or transcription-factor chemistry. The PubMed linked CommentIn is an editorial, not an erratum.
PMID:27106955 DOI 10.1093/eurheartj/ehw160. Primary PubMed abstract read; full body not recovered. Patient-derived and edited human stem-cell cardiomyocytes have abnormal sarcomeres, intercalated discs, electrical behavior and calcium handling. This is direct human cellular evidence but not evidence that ALPK3 transports calcium.
PMID:40135575 DOI 10.1161/JAHA.124.039464, original PMC article, Methods/Results and Figure 4. Mouse germline and inducible cardiac knockout, endogenous tagged localization, protein abundance and transcript comparisons support M-band maintenance in neonatal and adult hearts. The MuRF1 experiment transfects murine ALPK3 and MuRF1 into HEK293T; human host cells do not make these human constructs. ALPK3 relocalizes to MuRF1 puncta. This does not establish a ternary ligase-substrate bridge or ALPK3 ubiquitin-ligase chemistry. No endogenous nuclear signal was detected in this mouse model, limiting universality of the nuclear pool without disproving the independently observed human tagged nuclear-envelope pool.
PMID:41248220 DOI 10.1161/CIRCULATIONAHA.125.074632, original full PMC letter. Mouse miniALPK3 deletes residues 672–1044. Knock-in mice retain function, and neonatal tagged miniALPK3 gene delivery restores M-band organization and cardiac function. This deletes an internal region, not the alpha-kinase domain; rescue is not a catalytic-domain dispensability test.
PMID:42618778 DOI 10.1038/s44161-026-00843-1. Primary PubMed identity and publisher full Results and Methods read. The expressed transgene is full-length human ALPK3. Mutant mouse hearts and human cardiac organoids are separate experimental systems. Correct M-band targeting, restored sarcomeric structure/force and protein-abundance normalization strengthen the organizing/homeostasis role; they do not isolate direct kinase catalysis. The published paper supersedes the discovery-only preprint for this review; no additional preprint assertion is required.
Six ACCEPT, four UNDECIDED, zero NEW. The nucleus assertions retain their original breadth with tagged, developmental and species boundaries. The three developmental assertions have positive mouse donor evidence and independent human structural participation. GO:0055007 includes acquisition of specialized cardiac-cell features; GO:0055013 describes progression from cell formation to the mature state and is part of differentiation. Hence comparable early lineage-marker induction does not negate participation in organizing the developing contractile apparatus.
One compact core records cardiac-cell development and the resolved M-band/nuclear-envelope context. No autonomous MF is forced into the optional slot. The evidence supports an organizing protein network, but a precisely demonstrated simultaneous bridge or stable complex scaffold is not yet isolated from partner associations and protein homeostasis. GO:0140378 and the adaptor distinction were considered, with the unresolved mechanism placed in questions. This is not a claim that structural participation is absent. Local GO-CAM index search found no Q96L96/ALPK3 entry. No NEW process is proposed from necessity, and no redundant parent/child developmental NEW is added.
The canonical authored PMID/DOI census is exactly the ten records above. Untouched UniProt bibliography and references internal to these papers are not represented as independently consulted evidence. No genuine provider output exists, so there is no provider-only bibliography. All ten citation identities and actual access routes are recorded in the temporary source census; cache availability is assessed separately. Before hosted recovery, supporting-reference pointers retain the original PMID but do not present external full passages as cached ordinary quotes or use this notes file as a self-validating substitute.
Full validation completed successfully on 2026-09-27 with the ten missing-reference warning and advisory notices for the M-band and nuclear-envelope core locations. These locations summarize independently measured contexts; a validator suggestion does not itself justify an additional NEW row. Rendering passed. The draft retains all ten original source objects, their qualifiers and donor identifiers, all six seeded reference identities, and both immutable raw files. The single organizing/developmental core deliberately leaves its optional molecular-function slot empty while the catalytic conflict remains unresolved.
An independent bounded consultation read the saved positive-assay Methods passages and recognized the meaningful FSBA, heat-denatured-domain and phosphatase controls, while noting that the SQSTM1 preparation is an immunoprecipitate and the T269/S272 readout does not isolate residue class. That consultation was not a second complete read of both original studies. The coordinator is separately reviewing the full draft. Required missing records remain exactly the ten primary PMIDs listed above; no provider artifact exists and no DOI-only or title-only work is silently counted as closed.
All ten required primary records are now present as unchanged normal-fetch records. The hosted recovery followed the single failed local attempt; no local retry, source synthesis or overwrite was used. Their actual primary identities, content and access limits were checked before use. The earlier missing-cache statements above document the initial draft and no longer describe current availability.
Four records remain abstract-only: PMID:11418590, PMID:21441111, PMID:26846950 and PMID:27106955. The externally read original Midori article remains distinguished from its abstract-only cache. Six records contain readable extracted bodies: PMID:36321451, PMID:38048395, PMID:39196058, PMID:40135575, PMID:41248220 and PMID:42618778. Both short Circulation research letters contain actual experimental text; a machine publication-type label does not turn their experiments into an editorial. This is not a claim that every linked supplementary file was recovered.
The actual records confirm the earlier source-specific synthesis. The positive domain assay in PMID:39196058 includes FSBA treatment/removal for the global lysate experiment and active-versus-heat-denatured domain plus phosphatase controls for immunoprecipitated SQSTM1. The T269/S272 antibody does not resolve serine from threonine. The negative mouse-domain/TRPM7-comparator experiment in PMID:38048395 and human cellular catalytic-cleft comparisons in PMID:36321451 are retained as substantive opposing evidence. Neither preparation is declared artifactual. The negative study's modeled ATP pocket is not a direct ligand-binding measurement. Accordingly the three catalytic assertions and ATP-binding assertion remain UNDECIDED.
The actual mouse knockout/FLAG and murine-construct Methods in PMID:40135575 preserve the distinction between protein species and HEK293T host. In PMID:41248220, the body describes an internal 1119-bp exon-6 deletion; the independently verified primary PubMed Figure caption specifies residues 672–1044. The alpha-kinase domain is not deleted. PMID:42618778 measures human-specific transgene peptides across the 1705-residue protein, mouse cardiac rescue and separate human organoid rescue. Its comparison with miniALPK3 is explicitly between different studies, models, capsids and delivery routes; it does not isolate the deleted region's mechanism or establish clinical efficacy.
Exact ordinary cached excerpts now directly accompany all annotation and core supporting-reference pointers, with cache-specific full-text flags. The ten original annotation objects, six ACCEPT/four UNDECIDED decisions and one organizing/developmental core are preserved. No NEW assertion or catalytic molecular function is added. The complete authored/notes PMID and decoded DOI census remains these ten records; the five explicitly named PMC records and the newly available PMC13561881 all map to those same publications. No provider artifact exists. References merely listed inside source bibliographies are not represented as independently read evidence, and no new DOI-only or title-only source is silently substituted.
The prior draft manifest and history remain immutable. Publication can now be reviewed with the source-cache gates closed; the YAML retains DRAFT while the scientific catalytic questions and two existing core-location advisories remain explicit. This status is not a claim of missing publication records.
This section supersedes the earlier zero-NEW count and the two-location advisory state. The ten original GOA objects and their six ACCEPT/four UNDECIDED decisions remain preserved, with one NEW M-band IDA proposal. All ten consulted primary records remain available; the only location advisory retained is the contextual nuclear envelope.
The four uncertain activity/binding rows now quote the observed increase at 500 phosphosites after recombinant ALPK3-domain addition in PMID:39196058, including the preceding lysate-inhibitor context. The prior control-tube sentence established assay design but did not itself demonstrate a positive result. The cardiac-cell-development row now quotes sarcomere disorganization and loss of M-band MYOM1 from the same paper. These attachments now show the result relevant to each claim. The ATP-binding row still distinguishes a phosphotransfer assay from a direct binding measurement; the serine-specific rows still distinguish the combined SQSTM1 T269/S272 signal from an isolated serine-specific result.
The new M-band location is supported directly by microscopy of two endogenously tagged human ALPK3 reporter lines in PMID:39196058, with obscurin, α-actinin and MYBPC3 landmarks; PMID:40135575 independently corroborates the location in endogenous-tagged mouse neonatal and adult cardiomyocytes. The official GO definition and parents for GO:0031430 were checked: this is a cellular anatomical structure at the sarcomeric thick-filament midline, part of the A band. No existing source CC assertion is its ancestor or descendant. No generic sarcomere parent is added. This proposal is based on directly observed localization, not merely on absence from GOA or a core-location validator advisory.
A separate nuclear-envelope NEW would duplicate the existing nucleus branch: official GO:0005635 is part_of GO:0005634 nucleus. The specific envelope context remains in the core and reasons rather than adding a redundant descendant. The IBA, IEA and ISS nucleus rows now distinguish PAINT provenance, mouse ortholog transfer and the SL-0191 vocabulary mapping. The human endogenous-tag study PMID:39196058 also lacked a detected nuclear signal; the positive human envelope observation in PMID:36321451 is specifically from transiently tagged cells. Retaining that observed context does not assert universal nuclear residence.
The review suggestion to mark ALPK1-derived serine kinase ISS as over-annotated was evaluated against actual Results, not titles. Q96QP1 is the correct ALPK1 paralog donor. The positive domain assay remains substantive, as do the negative human cellular and mouse domain experiments. The latter structural comparison is to TRPM7 and does not independently demonstrate loss of an ALPK1-like activity. Paralogy is therefore a reason to require target-specific adjudication, not proof of failure. UNDECIDED records conflicting accessible evidence as permitted by the action definition; it is not limited to unavailable publications. The four uncertain assertions ask different questions (broad catalysis, ATP binding, reaction mapping and paralog-derived serine specificity), now with source-specific explanations.
Two UniProt bibliography entries, PMID:28630369 and PMID:30046096, were not independently consulted and support no authored finding, annotation or biological conclusion here. They are explicitly outside the ten-record evidence census rather than silently claimed as reviewed or unavailable. Their presence in the immutable UniProt bibliography does not create additional reviewed evidence. Existing clinical sources and direct cell/biochemical studies support the present scope; no claim about the content of those two unconsulted papers is made.