KRT19 (Keratin, type I cytoskeletal 19; K19, CK19) - curation notes

UniProt: P08727; HGNC:6436; 400 aa; type I (acidic) keratin of the intermediate filament (IF) family.

Deep research status

• just deep-research-falcon human KRT19 was launched in parallel with just fetch-gene-pmids.
  Status recorded at the end of this file. Literature was gathered independently from the
  cached publications and PubMed (PubMed MCP), so the review does not depend on it.

Structure

• Smallest keratin; uniquely lacks the C-terminal non-helical tail domain
  PMID:2448790.
• Domain layout: head (1-79), IF rod (80-391) with coils 1A/1B/2 and linkers [file:human/KRT19/KRT19-uniprot.txt "This keratin differs from all other IF proteins in lacking the C-terminal tail domain."].
• Highly conserved with the bovine ortholog (89% identity), suggesting strong structural constraint
  PMID:2448790.

Core molecular role: structural subunit of keratin IFs

• Obligate heteropolymer with type II keratins (mainly KRT8 in simple epithelia)
  [file:human/KRT19/KRT19-uniprot.txt "Forms intermediate filaments by heterodimerizing with the type II keratin KRT8, providing mechanical integrity to epithelial cells"];
  [file:human/KRT19/KRT19-uniprot.txt "Heterodimers composed of one type I and one type II keratins; forms parallel coiled-coil heterodimers"].
• High-throughput Y2H (HuRI, PMID:32296183; CCSB PMID:16189514, 25416956) recovers many type II keratins
  (KRT1-6, KRT8, KRT71-86 etc.) as binary partners - consistent with the promiscuous type I/type II
  rod-domain coiled-coil pairing. Non-keratin Y2H hits (ABI2, HGS, EXOC8, CARD9, etc.) are not
  biologically validated.

Expression

• Simple and ductal epithelia, basal keratinocytes in hair follicle outer root sheath, and others
  [file:human/KRT19/KRT19-uniprot.txt "Expressed in a defined zone of basal keratinocytes in the deep outer root sheath of hair follicles."].
• Marker of germinative epidermal layers in fetal skin, declining with maturation
  PMID:23377137.
• Widely used marker of cholangiocytes / ductal plate hepatoblasts PMID:19185580
  and of carcinomas / circulating tumour cells.
• Higher in ER-positive vs ER-negative breast cancer lines PMID:10037815
• this is differential expression between lines, not a demonstrated response to estrogen.

Striated muscle role (secondary, tissue-specific)

• K8/K19 concentrate at costameres; dystrophin's ABD binds K19 directly and specifically
  PMID:16000376.
• K19-null mice: mild myopathy, costamere disruption, subsarcolemmal gap with mitochondria
  PMID:17971417.
• K19/desmin double-null comparisons PMID:21209367;
  passive mechanics PMID:22287836.

Signalling-scaffold reports (cancer cell lines; non-core)

• K19 bridges tTG (TGM2) and Src in SKBR3 cells PMID:20080707.
• K19 interacts with beta-catenin/RAC1 and modulates NUMB/NOTCH signalling in breast cancer cells
  PMID:27345400.
  Context-dependent opposing effects in colon vs breast cancer PMID:30650643.
• K19 interacts with cyclin D3 and supports MCF7 proliferation PMID:31601969.
• These are single-lab, cancer-cell findings; treated as non-core / not annotated as new GO terms.

Annotation review decisions (summary)

• Core: structural constituent of cytoskeleton; keratin filament / IF cytoskeleton; IF organization.
• Muscle terms (costamere, sarcolemma, Z disc, structural constituent of muscle) kept as non-core.
• Epidermis-specific IBA (structural constituent of skin epidermis), epithelial cell differentiation,
  response to estrogen (IEP), DGC membership and apicolateral PM flagged as over-annotations.
• protein binding: REMOVE for uninformative non-keratin hits; MODIFY to protein heterodimerization
  activity (GO:0046982) for type II keratin partners; MODIFY to protein-macromolecule adaptor
  activity (GO:0030674) for the TGM2 bridging report.

Deep research status (update)

• Falcon deep research completed successfully: KRT19-deep-research-falcon.md. Its synthesis agrees
  with this review: K19 is structural (not an enzyme), pairs with type II keratins (especially K8),
  and CK19 positivity is a lineage readout rather than evidence that K19 drives differentiation
  [file:human/KRT19/KRT19-deep-research-falcon.md "a CK19-positive stain does not by itself show that K19 caused the observed differentiation"].
• Additional leads from the falcon report (not yet cached/verified, so not used for annotation):
  K8/K19 filaments scaffold cytoplasmic HNRNPK in MDA-MB-231 cells (Fallatah et al. 2023,
  doi:10.1186/s12860-023-00488-z); extracellular K19-CXCL12 coat formed by TG2 in pancreatic cancer
  mediating immune exclusion (Yan et al. 2023); Krt18/Krt19 complementary roles in muscle injury (Muriel et al.).