LINC-PINT (PINT87aa, UniProt A0A455ZAR2) – review notes

Scope

This folder is the UniProt entry A0A455ZAR2 (PNT87_HUMAN, "Transcriptional regulator
PINT87aa"), an 87-aa peptide. UniProt files it under the HGNC symbol of its host locus,
LINC-PINT (HGNC:26885), a long intergenic non-protein-coding RNA. There is no separate
HGNC symbol for the peptide, and the host is itself non-coding, so there is no competing
canonical protein for this symbol; the folder name is the plain host symbol.

Separate peptide-level from RNA-level evidence. Almost all LINC-PINT literature (miRNA
sponging, PRC2 recruitment, p53-induced transcript; dozens of cancer/inflammation papers)
concerns the RNA and is not evidence about PINT87aa. A PubMed search on 2026-10-03 for
PINT87aa, circPINT, PINT 87aa and LINC-PINT AND (circular OR circRNA) returned
only two primary papers on the peptide (PMID:30367041, PMID:33754036), plus reviews.
Both come from overlapping authors (Zhang N, Zhao K, Sun Yat-sen University); there is
no independent replication.

Origin: translated from a circular RNA

• Peptide is encoded by circPINTexon2 (circBase hsa_circ_0082389), a back-spliced circle of
  LINC-PINT exon 2. PMID:30367041
• Translation is cap-independent from an IRES within the circle. PMID:30367041
• The circle and the linear lncRNA are in different compartments: PMID:30367041
• Endogenous peptide detected with a custom antibody and IP–MS. PMID:30367041
  UniProt PE1 (protein sequence 1-7 and 37-56 by MS).
• Sequence is unusual: Arg-rich, with a run of seven Cys (CCCCCCC); no known domain other
  than Pfam PF21971 (LINC-PINT family, which is just this peptide). No PANTHER / PAN-GO.

Glioblastoma paper (PMID:30367041) – PAF1 complex

• Binds PAF1 (pull-down/IP–MS, co-IP, purified-protein binding to PAF1 aa 150–300).
  PMID:30367041
• Nuclear (RFP/GFP fusions; co-localises with PAF1).
  PMID:30367041
  PMID:30367041
  Note the tension: the coding circle is cytoplasmic (translation site) but the product is
  nuclear. Fine – small peptide, synthesised in cytoplasm.
• Reduces mRNA of PAF1 targets (CPEB1, SOX2, MYC) in brain tumour-initiating cells.
  PMID:30367041
• ChIP: co-occupies CPEB1 promoter with PAF1; overexpression/knockdown raise/lower PAF1 promoter occupancy.
  PMID:30367041
• Model (explicitly an assumption): PMID:30367041
• No DNA binding: PMID:30367041
• Important control separating RNA vs peptide: ASO knockdown of linear LINC-PINT did not
  change PAF1 targets. PMID:30367041
• Elongation per se was not measured (no Pol II pausing index / nascent RNA); GO:0034244
  (negative regulation of transcription elongation by RNA Pol II) would over-reach the data.
  GO:0000122 is the safer level.

HCC paper (PMID:33754036) – FOXM1

• Binds FOXM1 DNA-binding domain (docking + co-IP, GFP-only control negative; deletion of
  aa 30–36 abolishes binding). PMID:33754036
  PMID:33754036
• Does not change FOXM1 levels but lowers its targets incl. PHB2. PMID:33754036
• Downstream phenotypes from overexpression in HCC lines: senescence, G1 arrest, reduced
  PHB2-dependent mitophagy. PMID:33754036
  These are overexpression phenotypes in tumour lines; indirect, so not proposed as NEW.

Decisions

• GO:0000122 IMP – ACCEPT (core). Both papers point to repression of RNA Pol II target genes.
• protein binding (PAF1) – REMOVE: uninformative; PAF1 is an elongation-complex subunit,
  not a DNA-binding TF, so GO:0140297 does not fit; no "PAF1 complex binding" MF exists.
  Mechanism captured by the BP plus description.
• protein binding (FOXM1) – MODIFY to GO:0140297 DNA-binding transcription factor binding
  (FOXM1 is a forkhead DNA-binding TF; binding maps to its DBD).
• nucleus IDA and IEA – ACCEPT.
• Not proposing GO:0003714 transcription corepressor activity: plausible for the FOXM1
  arm but only overexpression data from one group; raised as a question instead.
• No NEW BP terms for senescence / mitophagy (indirect, overexpression).