miPEP155 / P155 (UniProt C0HMA1, gene symbol MIR155HG) — curation notes

Folder uses the repo's <HOST>__<ACC> convention for alternative-ORF peptides: the peptide
has its own UniProt entry (C0HMA1, P155_HUMAN) but UniProt files it under the host gene
symbol MIR155HG (HGNC:35460), the miR-155 host gene / BIC lncRNA. It must not be merged
with the host locus folder genes/human/MIR155HG/.

What it is

• 17 aa peptide (MEMALMVAQTRKGKSVV), PE1 (evidence at protein level), no domains, no TM,
  1879 Da. UniProt feature PRO_0000459474 covers residues 1..17 (the whole chain).
• Encoded by a 54-bp sense sORF ("ORF1") inside the human MIR155HG transcript, which is
  annotated as non-coding and is the primary transcript for miR-155
  PMID:32671205.
• Endogenous existence established three ways in the primary paper: CRISPR/Cas9 C-terminal
  EGFP knock-in at the ORF1 stop codon in HEK293T
  PMID:32671205,
  a peptide-specific antibody, and LC-MS of antibody immunoprecipitates
  PMID:32671205.
  Independently reproduced: the spliced pri-miR-155 transcript expresses miPEP155 in HeLa
  PMID:33810468.

Function (Niu et al. 2020, Sci Adv, PMID:32671205 — verified via PubMed/UniProt; journal is

Sci Adv, not Nat Commun)

• MIR155HG is induced in antigen-presenting cells in inflamed (psoriatic) dermis, not at
  steady state PMID:32671205.
• Target identification in human monocyte-derived DCs: biotin-P155 pull-down + silver
  stain + LC-MS + immunoblot identified a ~73 kDa partner as HSC70 (HSPA8)
  PMID:32671205;
  competition with free peptide established specificity. Also confirmed in human
  THP-1-derived DCs PMID:32671205.
• Binding is domain-selective (ATPase/NBD, not the SBD) and inhibits HSC70 ATPase activity
  PMID:32671205.
  ATPase assay used recombinant human HSPA8 protein plus synthetic peptide
  PMID:32671205.
• Downstream consequences were measured in mouse bone marrow-derived DCs: weakened
  HSC70–HSP90 and HSC70–LAMP2A co-IP, less OVA delivery to LAMP2A+ lysosomes, reduced
  MHC class II, reduced OT-II CD4+ T cell proliferation
  PMID:32671205,
  PMID:32671205.
• Therapeutic effect of exogenous synthetic peptide in two mouse models (MOG-EAE,
  imiquimod psoriasis-like skin inflammation)
  PMID:32671205.

Species caveat (important)

The peptide-coding sequence is not present in the mouse genome per the primary paper
PMID:32671205.
Mouse BMDC and in vivo work therefore tests the human peptide acting on mouse HSC70,
which is essentially identical to human HSC70 in both domains. A second group describes the
peptide as "extremely well conserved in primates and partially conserved in mice"
PMID:33810468,
so conservation outside primates is contested. Practical consequence for GO: the molecular
interaction (HSPA8 binding, ATPase inhibition) is demonstrated with human proteins, while
the cell-biological outcome (antigen trafficking, MHC-II, T cell priming) rests mainly on
mouse cells treated with the human peptide. The UniProt/GOA IDA rows are defensible, but
all of them derive from exogenously supplied synthetic peptide at 25 µM, not from
loss of function of the endogenous peptide. No peptide-specific knockout or knockdown
(which is hard here: the ORF sits inside the miR-155 primary transcript) has been reported.

What it does NOT do

Unlike plant miPEPs and human miPEP133, miPEP155 does not autoregulate its own host
pri-miRNA or the activity of miR-155
PMID:33810468;
instead it appears functionally antagonistic to miR-155
PMID:33810468.
This is a useful negative result: no transcription-regulation GO terms should be inferred
for this peptide. A third paper reports that MIR155HG promotes NK-cell proliferation
"without reliance on its derived miR-155 and micropeptide P155" (PMID:40486832, Acta Pharm
Sin B 2025) — i.e. a host-RNA function explicitly separable from the peptide; not cached and
not used as support for any annotation here.

Annotation decisions

• GO:0030544 Hsp70 protein binding (IPI, WITH UniProtKB:P11142) — ACCEPT. Informative MF,
  not bare protein binding; HSPA8 is the correct partner and the pull-down/LC-MS/PLA
  evidence is in human cells.
• GO:0002605 negative regulation of dendritic cell antigen processing and presentation
  (IDA) — ACCEPT, core. Direction and cell type are right; caveat recorded that readouts
  were in mouse BMDCs with exogenous human peptide.
• GO:0005737 cytoplasm / GO:0005634 nucleus (IDA, and the GO_REF:0000044 SubCell
  mappings SL-0086/SL-0191) — ACCEPT for cytoplasm (where HSC70 and the CMA/MHC-II
  trafficking machinery operate); nucleus kept but flagged as non-core, since it rests on
  immunofluorescence of FITC-synthetic peptide colocalising with antibody-detected
  endogenous peptide in HEK293T and no nuclear activity is known
  PMID:32671205.
  The two IEA rows are UniProt SubCell mappings of the same primary evidence — correct, and
  not circular in the sense CLAUDE.md reserves for IBA chains, but redundant with the IDA rows.
• NEW GO:0042030 ATPase inhibitor activity — proposed. Participation test is satisfied at
  the molecular level: the peptide itself binds the HSC70 nucleotide-binding domain and the
  measured ATP hydrolysis of recombinant human HSPA8 falls in its presence. This is the
  mechanistically informative MF, and it is what makes the BP term intelligible; the
  existing Hsp70 protein binding row records only the physical interaction.
• Not proposed: GO:0002587 negative regulation of antigen processing and presentation of peptide antigen via MHC class II. The MHC-II specificity is real but the GOA row
  GO:0002605 already captures the regulated process in the right cell type, and adding an
  overlapping sibling would be over-annotation on evidence from one paper. Raised as a
  suggested question instead.
• Not proposed: any chaperone-mediated autophagy term. The peptide inhibits a chaperone
  that performs that process; impairing someone else's machine is regulation of it, and the
  dedicated negative-regulation BP term already annotated covers the demonstrated outcome.