We used a polydonal antibody and a mixture of three monoclonal antibodies (MAb), all recognizing the protein core of the s m a l l dermatan sulfate proteoglycan (DSPG) (known as PG-II or decorin) derived from human skin fibroblasts, to immunolocalize this molecule in the characteristic lesions in Alzheimer's brain. All antibodies demonstrated positive decorin immunoStaining in both the amyloid deposits of neuritic plaques (NPs) and the filamentous structures within neurofbrillaty tangles (NFTs). Unlike heparan sulfate proteoglycans (HSPGs), which tend to be evenly distributed throughout NPs containing amyloid fibrils, decorin was pri-

Previous histochemical studies have suggested a close temporal relationship between the deposition of highly sulfated glycosaminoglycans (GAGs) and amyloid during experimental AA amyloidosis. In the present investigation, we extended these initial observations by using specific immunocytochemical probes to analyze the temporal and ultrastructural relationship between heparan sulfate proteoglycan (HSPG) accumulation and amyloid deposition in a mouse model of AA amyloidosis. Antibodies against the basement membrane-derived HSPG (either protein core or GAG chains) demonstrated a virtually concurrent deposition of HSPGs and amyloid in specific tissue sites regardless of the organ involved (spleen or liver) or the induction protocol used (amyloid enhancing factor + silver nitrate, or daily azocasein injections). Polyclonal antibodies to AA amyloid protein and amyloid P component also demonstrated co-localization to sites of HSPG deposition in amyloid sites, whereas no positive immunostaining was observed in these locales with a polyclonal antibody to the protein core of a dermatan sulfate proteoglycan (known as "decorin"). Immunogold labeling of HSPGs (either protein core or GAG chains) in amyloidotic mouse spleen or liver revealed specific localization of HSPGs to amyloid fibrils. In the liver, heparan sulfate GAGs were also immunolocalized to the lysosomal compartment of hepatocytes and/or Kupffer cells adjacent to sites of amyloid deposition, suggesting that these cells are involved in HSPG production and/or degradation. The close temporal and ultrastructural relationship between HSPGs and AA amyloid further implies an important role for HSPGs during the initial stages of AA amyloidosis.

We generated a monoclonal antibody (Mab) against a large chondroitin sulfate proteoglycan (CSPG) isolated from bovine aorta. This Mab (941) immunoprecipitates a CSPG synthesized by cultured monkey arterial smooth muscle cells. The immunoprecipitated CSPG is totally susceptible to chondroitinase ABC digestion and possesses a core glycoprotein of Mr approximately 400-500 KD. By use of immunofluorescence light microscopy and immunogold electron microscopy, the PG recognized by this Mab was shown to be deposited in the extracellular matrix of monkey arterial smooth muscle cell cultures in clusters which were not part of other fibrous matrix components and not associated with the cell's plasma membrane. With similar immunolocalization techniques, the CSPG antigen was found enriched in the intima and present in the medial portions of normal blood vessels, as well as in the interstitial matrix of thickened intimal lesions of atherosclerotic vessels. Immunoelectron microscopy revealed that this CSPG was confined principally to the space within the extracellular matrix not occupied by other matrix components, such as collagen and elastic fibers. These results indicate that this particular proteoglycan has a specific but restricted distribution in the extracellular matrix of arterial tissue.

Ultrastructural localization of growth hormone in rat anterior pituitary and of muscle-specific actin in rabbit arterial smooth muscle cells was accomplished with a post-embedment procedure using colloidal gold. Plastic sections (2 microns) were mounted on slides, deplasticized, immunostained with immunoglobulin-colloidal gold particles, re-embedded in Epon, and sectioned for electron microscopy. This procedure enabled light and electron microscopic localization of these intracellular antigens on the same section. Positive immunostaining was demonstrated with this procedure with a muscle-specific actin antibody which previously failed to localize antigenic sites by EM. The procedure described yielded staining of high specificity, with minimal background and well-preserved ultrastructure. This re-embedding technique is useful in situations where problems with post-embedding EM immunostaining exist and where correlative LM and EM immunostaining is essential.

We localized tissue antigens on ultra-thin sections by deplasticizing the sections while on the grid, incubating in primary antiserum followed by immunoglobulin-conjugated colloidal gold, and ultimately re-embedding in dilute Epon. This procedure permitted ultrastructural localization of tissue antigens that were previously masked by the embedding plastic surrounding tissue components. In addition, replacement ofthe plastic matrix on the thin section after immunostaining prevented development ofthe drying artifacts that occur in unsupported tissue sections. Optimal preservation of components in the tissue sections was achieved