Expression of protease nexin-II in human dorsal root ganglia. A correlative immunocytochemical and in situ hybridization study

Expression of Kunitz protease inhibitor--containing forms of amyloid beta-protein precursor within vascular thrombi

BACKGROUND

The presence of patent neovessels within vascular occlusions in chronic thromboembolic pulmonary hypertension suggests that local mechanisms exist to regulate the coagulation system. This study investigated the expression of a potent inhibitor of Factor IXa and Factor XIa (ie, protease nexin-2/ amyloid beta-protein precursor, A beta PP) in the organized vascular occlusions harvested from patients with this disease.

METHODS AND RESULTS

Immunohistochemical analysis revealed intense immunoreactivity for A beta PP in the single layer of cells that line the neovessels. A positive signal was also detected by in situ hybridization analysis with the use of a 35S-UTP-labeled antisense riboprobe that recognizes the various alternatively spliced mRNA forms of this molecule. To identify the forms of A beta PP produced within the thrombi, total RNA was extracted from the thrombi, reverse transcribed, and subjected to amplification with the use of the polymerase chain reaction (PCR) and primers that flank the region encoding the alternatively spliced 56-amino acid Kunitz-type protease inhibitor (KPI) domain. The major PCR products consisted of 255 bp and 312 bp and corresponded to transcripts encoding this domain (ie, A beta PP751 and A beta PP770). In situ hybridization analysis with the use of a 35S-UTP-labeled antisense riboprobe complementary to the region encoding the KPI domain confirmed the presence of these mRNA species in nucleated cells lining the neovessels.

CONCLUSIONS

The expression of KPI-containing isoforms of A beta PP in thrombus endothelial cells may represent one mechanism utilized in this disease to shift the local hemostatic balance and preserve regional vessel patency.

Cell surface APP751 forms complexes with protease nexin 2 ligands and is internalized via the low density lipoprotein receptor-related protein (LRP)

The secreted isoforms of the amyloid precursor protein (APP) that contain the Kunitz domain are also known as protease nexin 2 (PN2). Normal proteolytic processing of transmembrane APP, which results in the majority of soluble PN2, cleaves within the Alzheimer's A beta peptide, precluding A beta formation. Recent data indicate that soluble PN2 is internalized by cells via the low density lipoprotein receptor-related protein (LRP), which binds multiple ligands including apolipoprotein E (apoE) [23]. However, soluble PN2 cannot contribute to amyloid accumulation, so we examined whether the unprocessed, transmembrane form of APP751 containing the intact A beta sequence would form complexes with a PN2 ligand, EGF binding protein (EGFBP), and be internalized by LRP. We found that the addition of EGFBP to cells overexpressing APP751 induced the internalization of this amyloidogenic form of APP. The 39 kDa LRP receptor associated protein (RAP), an antagonist for LRP, blocked the internalization of APP751/PN2, suggesting a common LRP-mediated internalization pathway for both soluble and transmembrane APP751/PN2 after protease complex formation. Previous work has shown that internalization of transmembrane APP can lead to the formation of amyloidogenic carboxyl-terminal fragments and increased secretion of the Alzheimer's A beta peptide. Our data suggest the protease ligands for PN2 may play an important role in altering APP processing pathways to favor amyloid formation, and that LRP may be a point at which the apoE and amyloid processing pathways intersect.

beta-Amyloid precursor protein (beta APP) in human gut with special reference to the enteric nervous system

The distribution of the beta-amyloid precursor protein (betaAPP) in the human gastrointestinal tract, from esophagus trough rectum, was studied using immunoblotting, as well as combined immunohistochemical and image analysis (optic microdensitometry) techniques. The study was focused on the enteric nervous system. betaAPP was detected by means of a monoclonal antibody (22C11), which recognizes all betaAPP isoforms as well as betaAPP-like proteins. Immunoblotting revealed two main protein bands, one corresponding to full-length betaAPPs (estimated molecular masses of approximately 97-115 kDa); the other corresponded to a protein with estimated molecular masses of 55 kDa. Specific betaAPP immunoreactivity (IR) was found in the submucous and myenteric plexuses localized in the supporting glial cells rather than in neurons. Differences were encountered neither in the localization nor in the intensity of immunostaining among different segments of the gastrointestinal tract. Moreover, no age-dependent changes were found. betaAPP IR was also regularly observed in blood vessels, primarily labelling endothelial cells. Our results provide evidence for the occurrence of betaAPP in human gastrointestinal tract of healthy people in both neuronal and nonneuronal tissues. Whether or not these findings have functional or clinical relevance remains to be clarified in future studies.