Human ALX4 is HGNC:450 / UniProt Q9H161, a 411-residue paired-like homeodomain protein. The exact seed6 baseline contains 29 PENDING annotations and 13 reference identities, with no alternative-product objects. The fresh preflight found the canonical and PFM2/FPP/PFM/KIAA1788 alias directories, histories and task branch absent on main, and no matching open PR. All machine-supplied source terms, evidence codes, qualifiers, original references and donor entities are retained.
The genuine normal Falcon attempt with perplexity-lite fallback and the installed client override failed DNS and produced no report. Manual primary research supplies the assessment. The seed-publication fetch found all five existing records. One additional normal fetch for 9847249, 11696550, 22829454, 38481039 and 40410151 terminated naturally with exit 1, 0/5 records and DNS failures; receipt tmp/ALX4-full-audit/additional-normal-fetch-receipt.json binds the durable output. The shared 38262408 fetch belongs to the ALX1 owner and is not duplicated. Primary web access below is distinct from this normal cache route. Six publication caches remain required, so the review remains DRAFT.
PMID:11137991 (DOI) is the intended human ALX4 haploinsufficiency paper. Its cached abstract identifies a paired-related homeodomain transcription factor; it is not used as a substitute for later reporter or localization experiments. The original full body was not recovered. PMID:11106354 (DOI) independently reports mutations in two of three tested human families with skull ossification defects. These historical NAS assertions have independent functional corroboration below; their clinical phenotypes alone do not establish a particular biochemical activity.
PMID:22829454 (DOI), original indexed author manuscript, Methods/Results: human constructs in control human calvarial osteoblasts and P3 luciferase reporters measure transcriptional activation. Variant behavior is specific to these conditions. It is not assumed to match every later reporter host or DNA arrangement.
PMID:40410151 (DOI), original publisher full article, Results/Methods: purified mouse ALX4 regions 209–274 and 169–303 are explicitly identical to human in the corresponding regions. Those structural/biophysical preparations are separate from full-length human HA-ALX4 tested in HEK293T reporters. Cooperative P3 binding, P3/P4 comparisons, ITC and interface/DNA-contact mutants distinguish altered cooperativity from monomer affinity or expression/localization. The V241R interface perturbation and R218Q DNA-contact phenotype support the measured molecular mechanism. K211E/R272P effects must remain assay-context-specific when compared with the earlier osteoblast study. The genomic analysis reuses the human 2024 chromatin data; it is not a second independent chromatin experiment.
PMID:38262408 (DOI), published original PMC10872279, Methods/Results/Figures 5–6: human H9-derived cranial neural crest cells, three ALX4 frameshift clones and ALX1 degradation reveal regional mesenchymal programs and changes in TWIST1 occupancy. Structure/EMSA and cooperative DNA occupancy establish positive assembly on DNA. The Methods explicitly identify human TWIST1 101–170, TCF4 565–624 and ALX4 210–277 fragments; bacterial expression is a separate experimental-host fact. The lack of a stable interaction in DNA-minimizing solution extracts does not negate a DNA-associated transcription complex. These data do not assert an obligatory soluble full-length complex. The published version is used; the earlier preprint was only a discovery route.
PMID:9847249 (DOI), primary PubMed and original-paper introduction: positive Alx4–Cart1 heterodimeric DNA binding, reporter output and mouse double-mutant genetics support cooperation. Complete construct Methods were not recovered, so those biochemical preparations are not silently labeled human.
PMID:11696550 (DOI), original author-uploaded article, Methods/Results: COS1 co-IP and deletion mapping identify the LEF1 HMG domain and an Alx4 N-terminal interaction region. C33A mouse N-CAM promoter reporters show dose-dependent activation/repression; a DNA-binding-defective Alx4 construct fails to modulate the reporter. Overexpression in primary mouse mammary stromal cells lowers N-CAM. Murine LEF1 is specified, whereas the earlier referenced Alx4 cDNA species is not re-established by the retrieved Methods. These are positive specific-partner experiments, not evidence that all regulation is activating or that N-CAM promoter regulation itself is neuron development.
PMID:19692347 (DOI), original author-uploaded article, complete main Methods/Results/Figures read externally while the canonical record remains abstract-only. Human R265X families and patient/control skin histology show altered epidermal and hair-follicle differentiation. The cached abstract states: “Hair follicle-like structures were present but showed altered differentiation.” Follicles initiate; the annotation need not imply that ALX4 initiates all follicle formation. Tagged wild-type and mutant human ALX4 localization is measured in COS7, while a separate HEK293T experiment addresses protein stability. The available antibody did not detect endogenous protein in either control or patient osteoblasts. That limitation is not interpreted as loss of nuclear residence. Altered junctional beta-catenin does not isolate a direct adhesion or Wnt function.
PMID:38481039 (DOI), primary abstract and indexed original Results/Figures 1–5: conditional mouse exon-2 deletion, comparison with lst-2J and Prrx1-Cre/Wnt1-Cre contexts support lineage-dependent skeletal and hair phenotypes. Complete Methods were not independently recovered. Mouse mesenchymal/cranial-neural-crest phenotypes are not relabeled as direct human assays or neuronal differentiation.
The current Human Protein Atlas ALX4 subcellular page reports supported nucleoplasm with HPA001903 in HEK293 and SiHa, with no staining in U2OS. The historical v16 nucleus label is distinguished from the current nucleoplasm wording. The compact core uses the independently supported nuclear location; the specific nucleoplasm row retains the source cell-line scope.
The neuron-development IBA remains UNDECIDED. Its exact PTN001216073 ancestral IBD and experimental chain were not reconstructed. The formal term concerns progression of a neuron, not all neural-crest derivatives. In the published 2024 study, the emphasized SOX10/SOX2/3/TUBB3 induction is associated with TWIST1 loss, while ALX perturbation changes regional facial mesenchymal identity. Neither this distinction nor the absence of a direct neuronal assay in the inspected papers proves evolutionary loss of an ancestral function. The PAINT assertion is preserved rather than replaced with an invented donor history.
PMID:28473536 (DOI): the immutable machine full-text flag is true, but the actual extraction contains only abstract and conclusion, with no ALX4-specific clone/motif supplement. The human methyl-SELEX design is visible and the seeded ALX4 IDA is retained with curator deference plus independent target DNA-binding evidence. No ALX4-specific methylation preference is inferred from a general homeodomain result. This is an atomic DNA-binding activity, compatible with the integrated transcription-factor core.
PMID:32296183 (DOI): actual main Methods and Results describe the intended human binary-interaction map. The pair-level IQUB/PARVG supplemental entries were not recovered. The immutable UniProt record nevertheless explicitly preserves both positive interactions, each with three experiments. Both generic protein-binding annotations remain UNDECIDED because their specific evidence cannot yet be adjudicated; they are not called false or converted into invented mechanistic binding functions.
All 29 rows are reviewed: 21 ACCEPT, five MODIFY and three UNDECIDED, with zero NEW. The five refinements resolve generic DNA binding, generic TF activity and generic transcription regulation to independently measured regulatory-region/Pol-II subtypes. The broader Pol-II TF/regulation terms remain useful because regulatory direction varies with dose, partner and promoter. Broad nucleus and organ development remain valid core-resolution assertions. The precise double-stranded DNA-binding aspect is retained at its assay resolution, without importing an unmeasured methylation preference.
PAINT nodes PTN004692309, PTN002803099 and PTN001216073 are preserved as ancestral assertions, with unresolved node reconstruction documented rather than judged from donor count. ALX4 Q9H161 occurring among the nucleus IBA descendants is legitimate target grounding, not circular evidence. The mouse O35137 donor and all supplied Ensembl identifiers remain unchanged. No local GO-CAM index entry for Q9H161/O35137 was found. No NEW term is proposed, and no redundant parent/child activity is manufactured.
One compact core integrates the measured activator activity, nuclear DNA-associated transcription complex and skeletal/hair development. Live official term checks covered GO:0001228, its regulatory-DNA aspect, GO:0005667, GO:0071837 and GO:0048666. Positive assembly is assessed in its DNA context, not against an invented requirement for stable solution stoichiometry.
The required authored citation census is eleven PMIDs: five seeded cached records plus six additional records above, including the shared ALX1-owned 38262408. Their eleven DOIs and the two additional PMC URLs identify these same works. Untouched UniProt sequence/proteomics bibliography, bibliography entries within articles and discovery-only search results are not asserted as independently consulted evidence. No genuine provider artifact exists, and no DOI-only or title-only work is substituted with an invented cache.
This later assessment supersedes the six-cache availability statements above. All eleven required PMIDs are now present as exact normal-fetch records. The six source26 records comprise abstract-only PMID:9847249 and PMID:11696550, plus scientific bodies for PMID:22829454, PMID:38262408, PMID:38481039 and PMID:40410151. The shared ALX1 record is byte-identical. Raw source metadata and all earlier history remain unchanged. Source identity, actual body access and biological interpretation are assessed separately; the original LEF1 author upload remains an external full-text route while its local cache remains abstract-only.
The recovered mouse lineage paper now supplies actual Experimental Procedures as well as Results: the exon-2 conditional allele and global/lineage-specific Cre experiments distinguish mesenchymal, neural-crest and mesoderm contributions. Its Mesp1-associated skull ossification delay is transient. These findings corroborate developmental participation without transferring mouse lineage phenotypes into a direct human neuronal assay.
The 2012 human osteoblast reporter reports increased V7F/K211E output and reduced R272P output; the 2025 human HEK293T reporter gives different K211E/R272P behavior. These outcomes remain attached to their respective constructs, hosts and reporters. The 2025 purified mouse DNA-binding regions are explicitly identical to corresponding human sequence; they are separate from the full-length human reporter and microscopy constructs. Its genomic analysis reuses the published 2024 human chromatin study. In that study ALX4 knockout is combined with ALX1 degron experiments; ALX4 itself was not successfully degron tagged. No additional independent chromatin replication or fixed soluble complex is inferred.
Four nucleus rows previously attached a hair-development abstract passage from PMID:19692347. That passage supports the developmental interpretation, but not localization, and has been removed from those four evidence lists. The direct original nuclear-localization study remains cited with its previously inspected human-protein/COS7 scope. PMID:40410151 independently reports human HA-ALX4 nuclear staining in HEK293T, with the precise cached clause attached to these rows: “similar protein levels and nuclear localization of the wild type and V241R (V28R) ALX4 proteins in cell culture”. The unchanged core retains the hair-development passage for its appropriate developmental claim.
All 29 annotation source objects, actions (21 ACCEPT, five MODIFY, three UNDECIDED), 20 reference identities and the single core assignment are preserved; zero NEW assertions are introduced. The unresolved neuron-development ancestral chain and two pair-level interaction supplements remain scientific uncertainties, not missing normal-cache gates. No new DOI-only or title-only source is added. The complete citation census remains eleven PMIDs/eleven DOIs with no Reactome references or genuine provider report.
The official Human Protein Atlas ALX4 subcellular record, inspected directly on 2026-09-27, reports “Localized to the nucleoplasm.” It lists supported nucleoplasm with HPA001903 in HEK293 and SiHa and no staining in U2OS. The nucleoplasm annotation now cites this record directly as HPA:ENSG00000052850. These are the database's antibody-based cell-line annotations; the images were not independently reanalyzed. RNA-expression values in the table are not staining intensities. The original GO_REF source is intact, and the core retains the broader nuclear location.
The two HuRI protein-binding annotations remain UNDECIDED. The main article's assay class and UniProt/IntAct NbExp=3 pair records are available, but the pair-specific supplementary experiments and their functional interpretation have not been independently adjudicated. The review policy explicitly retains UNDECIDED for inaccessible or unadjudicable evidence, and the user instructions prohibit removing an experimental annotation on incomplete evidence. No specific molecular function can be inferred from these records alone. Typical IQUB/PARVG compartment descriptions do not establish mutual exclusion with ALX4 and are not evidence that either reported pair is false. Generic binding remains uninformative; that observation does not remove the stated evidence-access limit.
The accepted animal organ development term describes core developmental participation at a broad level. The current official ontology places hair follicle development within organ development through part_of ancestry via skin development. Bone development is_a animal organ development and part_of skeletal system development; skeletal-system development as a whole is not thereby an organ-development subtype. This clarifies the earlier imprecise parent wording. The molecular regulatory evidence and human craniofacial/skin findings remain the biological basis. Broadness and peripheral involvement are different judgments: this retained umbrella does not add an independent core claim or assign every organ phenotype to ALX4.
The ARBA double-stranded DNA-binding row now distinguishes its substrate aspect from the generic sequence-specific DNA-binding parent refined to regulatory-region binding. It is supported by the existing direct target evidence and does not count the electronic assertion as another assay or infer a methyl-CpG preference. The neuron-development IBA remains UNDECIDED: a different PAINT node identifier alone does not establish an incorrect ancestral placement, and the unresolved historical chain has not been recovered. Neural-crest mesenchymal and TWIST1-perturbation results are not silently converted into ALX4 neuron-development evidence. Repeated source-scope detail in otherwise unchanged rows remains available for independent reading.
All 29 original source objects, ordinary qualifiers, donors, 20 original references, two published histories and the compact core are preserved. One direct HPA database reference is added; all eleven PMID caches remain unchanged. No annotations are added, and the action tally remains 21 ACCEPT, 5 MODIFY and 3 UNDECIDED. These three scientific uncertainties do not reopen the closed source-cache gate; review status remains COMPLETE. This follow-up has a separate append-only curation session.