Manual literature synthesis for human PI16 (Q6UXB8)

Scope and provenance

This report was prepared manually from the project-cached UniProt record, GOA rows, PAINT trace, and primary publications after the configured automated providers failed on 2026-07-18 (Falcon/Edison HTTP 402; Perplexity-lite HTTP 401 insufficient quota). It is intentionally named deep-research-manual rather than after a provider.

Molecular identity and localization

PI16 is a CAP-superfamily glycoprotein with an N-terminal secretory signal and a C-terminal hydrophobic sequence compatible with GPI-anchor processing. Human PI16 was originally purified as a glycosylated PSP94/MSMB-binding protein from serum, with expression detected in prostate, testis, ovary, intestine, peripheral leukocytes, pituitary, and Leydig cells. The direct serum purification establishes a soluble extracellular pool PMID:15344909.

Human skin-homing CD8 T cells display PI16 on the plasma membrane through a GPI anchor; the protein is lost after restimulation through transcriptional down-regulation rather than demonstrated enzymatic shedding PMID:30365157. Mouse cardiac fibroblasts similarly expose Pi16 to the interstitium through a GPI anchor PMID:27539859. PI16 peptides and glycosylation sites have also been detected in human urine PMID:22171320. Thus, extracellular-region and cell-surface/plasma-membrane localizations are both supported and should not be treated as mutually exclusive.

Direct molecular activities

The strongest direct human biochemical evidence identifies PI16 as an extracellular endopeptidase inhibitor with at least two target classes:

1. In primary human coronary artery endothelial cells, laminar shear strongly induces PI16, while inflammatory cytokines reduce it. Recombinant PI16 inhibits recombinant MMP2 with an approximate 10 nM IC50 and binds a loop adjacent to the MMP2 active site. PI16 overexpression lowers secreted MMP activity and reduces endothelial-cell migration PMID:27996045. This supports GO:0008191 metalloendopeptidase inhibitor activity and GO:0010596 negative regulation of endothelial cell migration.
2. In human skin-homing CD8 T cells, inhibitor screening and pull-down experiments show that PI16 inhibits cathepsin K and may not bind other tested skin proteases PMID:30365157. Mouse cardiac experiments independently show recombinant Pi16 inhibits cathepsin K and reduces conversion of pro-chemerin to active chemerin PMID:27539859. This supports GO:0004869 cysteine-type endopeptidase inhibitor activity, with target specificity presently bounded to cathepsin K.

The two activities are more informative than the generic GO:0030414 peptidase inhibitor activity and provide the best-supported core molecular functions.

Biological roles and context

PI16 overexpression inhibits human coronary endothelial migration in vitro, consistent with reduced extracellular MMP activity PMID:27996045. In mouse myocardium, Pi16 suppresses cathepsin-K-dependent chemerin activation and limits cardiomyocyte growth [PMID:27539859; PMID:17909105]. Mouse loss-of-function also reduces nerve-injury-induced endothelial permeability, leukocyte influx, and neuropathic pain, placing fibroblast-derived Pi16 upstream of endothelial-barrier opening in that disease model PMID:32079726. These context-specific phenotypes are useful biological boundaries but do not by themselves establish additional human catalytic activities.

PAINT inference assessment

The cached PANTHER trace places GO:0060090 molecular adaptor activity at the broad PTHR10334 node PTN000036124 from only MGI:MGI:1916536, mouse Glipr1l1. Glipr1l1 functions in sperm adhesion/fusion, whereas PI16 is a long, glycosylated protease inhibitor expressed in fibroblasts, endothelium, T cells, serum, and urine. Because the inference crosses divergent CAP-family subfamilies and no PI16 study demonstrates adaptor activity, the IBA is an over-propagated family inference and should be removed. Direct MMP2 and cathepsin K inhibitor annotations replace it.

Evidence boundaries and open questions

• It is not known which PI16 isoforms or processing events determine stable GPI anchoring versus release into extracellular fluids.
• The inhibitor specificity panel is incomplete. Direct evidence supports MMP2 and cathepsin K, not all metalloproteinases or cysteine peptidases.
• The structural mechanism by which the CAP domain recognizes and inhibits two mechanistically distinct protease classes is unresolved.
• Whether the mouse cardiac, chemerin-processing, and neuropathic-pain mechanisms operate similarly in humans remains to be tested.
• The physiological significance of PSP94/MSMB binding and of high-throughput interaction candidates, including TNFRSF21, is not established.