LOXL3

Gap: The physiological human protein substrates and isoform-specific substrate preferences of LOXL3 are not established in vivo.

OPEN BIOLOGY RESIDUAL_SUBGAP

What is known: Recombinant human full-length LOXL3 and LOXL3-sv1 oxidize collagen and elastin with different preferences, and recombinant sv2 oxidizes collagen I; mouse/model work identifies fibronectin, but direct oxidation of endogenous human substrates has not been mapped.

Significance: Identifying native substrates is necessary to distinguish the conserved extracellular catalytic role from substrate assignments transferred from model organisms and to explain isoform-specific connective-tissue functions.

What would resolve it: Map LOXL3-dependent allysine sites in human extracellular matrices using isoform-selective knockout/rescue and quantitative carbonyl/allysine proteomics, followed by purified-enzyme validation of candidate substrates.

Provenance (the field's own admissions):

• "The recombinant LOXL3-sv1 protein showed a beta-aminopropionitrile-inhibitable amine oxidase activity toward elastin and collagen with substrate specificity." — PMID:17018530
• "We find that LoxL3 complexes with and directly oxidizes Fibronectin (FN), an ECM scaffold protein and integrin ligand enriched at the MTJ." — PMID:26954549

Gap: Whether human LOXL3 undergoes a required extracellular proteolytic activation step, including cleavage by BMP1-family proteases, is unknown.

OPEN BIOLOGYCURATION RESIDUAL_SUBGAP

What is known: Human LOXL3 contains a signal peptide and is secreted, but the reviewed record defines the chain after signal-peptide removal without a demonstrated LOXL3 propeptide; one predicted splice form lacks a potential BMP-1 cleavage site.

Significance: This determines whether secretion alone produces an active enzyme or whether tissue-specific proteolysis controls extracellular LOXL3 activity.

What would resolve it: Determine native N termini and cleavage products of endogenous secreted LOXL3 isoforms by N-terminomics and intact-mass analysis, then test activity before and after candidate-protease treatment and cleavage-site editing.

Provenance (the field's own admissions):

• "Recombinant LOXL3, expressed in HT-1080 cells, was secreted into the culture medium but was not detected by immunofluorescence staining in nuclei." — PMID:11284725
• "The putative resulting protein retained the region encoding the structural and functional elements of the amine oxidase but the second and fourth SRCR domains were truncated and the potential BMP-1 cleavage site was not present." — PMID:11386757

Gap: The biochemical reproducibility, physiological scope, and isoform dependence of the reported nuclear STAT3 deacetylation/deacetylimination activity remain unresolved.

OPEN BIOLOGY RESIDUAL_SUBGAP

What is known: One abstract-only human study assigns the major STAT3-directed activities to N-terminal SRCR repeats rather than the canonical C-terminal oxidase domain; sv1 lacks SRCR domains 1-3 and sv2 lacks SRCR domain 2.

Significance: Independent validation is needed before this unusual chemistry can be treated as a general intracellular molecular function of LOXL3 or applied across its structurally different isoforms.

What would resolve it: Independently reproduce STAT3-site chemistry with purified, cofactor-defined canonical LOXL3, sv1, sv2, isolated SRCR regions, and catalytic mutants; confirm products by site-resolved mass spectrometry and test endogenous isoform-specific rescue in human immune cells.

Provenance (the field's own admissions):

• "Surprisingly, Loxl3 N-terminal scavenger receptor cysteine-rich (SRCR) repeats, rather than the C-terminal oxidase catalytic domain, represent the major deacetylase/deacetyliminase activity." — PMID:28065600

Gap: The endogenous expression, secretion, localization, and exact database mapping of all reported human LOXL3 transcript forms are incompletely resolved.

OPEN BIOLOGYCURATION CC_DARK

What is known: UniProt names three isoforms; sv1 is experimentally characterized, the internal deletion of P58215-3 is structurally consistent with sv2, and an older exon-5/exon-8 EST form has neither demonstrated protein expression nor activity.

Significance: Correct product mapping is required to assign extracellular versus intracellular localization, substrate specificity, and SRCR-dependent activities without transferring evidence across non-equivalent products.

What would resolve it: Use long-read RNA sequencing, isoform-resolved endogenous proteomics, and isoform-specific secretion/localization assays in expressing human tissues, then reconcile each experimentally observed product with UniProt accessions.

Provenance (the field's own admissions):

• "A BLASTN search of the human EST database indicated the presence of ESTs, corresponding to alternative splice variants of LOXL3, that lacked exon 5 and exon 8." — PMID:11386757
• "Both LOXL3 isoforms were found similarly expressed in most melanoma cells tested (Fig. 1f). LOXL3Δ isoform probably corresponds to the recently identified splice variant LOXL3-sv2 [27]." — PMID:29229995