Background & Aims: In the gastrointestinal tract, prostaglandins are implicated as important mediators of normal physiological processes. Prostaglandin G/H synthase (PGHS) is the first enzyme leading to the formation of prostaglandins. Two forms exist: the constitutive PGHS-1 and the inducible PGHS-2 isoforms. The purpose of this study was to examine the expression of PGHS-1 and -2 in gastrointestinal tissues. Methods: PGHS-1 and -2 expression and activity were examined in rat, dog, monkey, and human gastrointestinal tracts by immunoblot and biochemical assays. Results: PGHS-1 but not PGHS-2 protein was identified in all gastrointestinal tissues. PGHS-1 protein varied throughout the gastrointestinal tracts; interspecies dif- ferences were also noted. Immunohistochemical studies showed PGHS-1 staining of rat endothelial cells in all gastrointestinal regions; PGHS-2-specific staining was noted in a subset of macrophages in 3 of 22 rats examined. Elevated activity was shown in tissues expressing greater concentrations of PGHS-1 protein. Indomethacin, a nonsteroidal anti-inflammatory drug that inhibits both isoforms, inhibited prostaglandin synthesis, whereas NS-398, a selective PGHS-2 inhibitor, showed little or no inhibition of prostaglandin synthesis in gastrointestinal tissues. Conclusions:These results indicate that prostaglandins produced in normal gastrointestinal tissue and required for normal physiological functioning are derived from the PGHS-1 isoform.

A method for preparing skin biopsies for cryosectioning was developed to accurately obtain samples from specific areas of the dermis, while minimizing contamination with epidermal tissue. Routine preparation of 6mm punch biopsies from freshly excised, full-thickness skin produced contraction and folding of the edges of the biopsy prior to mounting for snap-freezing and cryosectioning. Sample orientation was ruined, and cryosections were heterogeneous with respect to dermal structures and/or to dermal and epidermal layers. Biopsy artifacts were prevented by prefreezing skin over dry ice prior to taking biopsies. The biopsies were held frozen on dry ice until they were mounted on cryostat pegs with flattened, frozen OCT surfaces; then they were snap-frozen in chilled OCT in an isopentane bath cooled with liquid nitrogen. The method for determining skin level homogeneity of cryosections consisted of taking 10 mu m cryosections for histology between sections sampled for drug level analysis. The histological sections were fixed in 5% acetic acid in methanol and stained with hematoxylin and eosin to define the skin layers and structures associated with each sample for analysis. Histological sections from prefrozen skin had fewer processing artifacts, and dermal cryosections free of epidermal contamination were dramatically increased compared to the routine procedure.

Effective methods exist for separating epidermis from dermis for many species; however, a simple and effective skin separation method for non-human primates is not available. This investigation describes an easy and reliable method for separating epidermis from dermis in Rhesus monkeys. Skin was shaved and washed prior to necropsy. Skin samples were placed on cardboard and then in Whirl-Pak bags, frozen on dry ice and stored at -70 degrees C. Just prior to the separation procedure, Whirl-Pak bags were returned to dry ice storage. Immediately after removal from dry ice, each closed Whirl-Pak bag was placed into a waterbath maintained between 60 and 67 degrees C. After 2 minutes, the Whirl-Pak bag was removed from the waterbath, opened and the skin surface of the application site was gently scraped with a scalpel blade to remove the epidermis. Effectiveness of removal was verified by histologic examination of the remaining dermal samples.

Immunohistochemical localization of osteopontin, a phosphorylated acidic glycoprotein, was compared in adult rat femur fixed in 4% paraformaldehyde at 4° C for 48 h and demineralized at 4° C in ethylenediaminetetraacetic acid (EDTA), modified Jenkin's solution, or 15% formic acid, until radiographs indicated demineralization was complete. Formic acid was also evaluated at room temperature. EDTA solution (15 days) resulted in intense staining of osteocytes, periosteal osteoclasts and osteoblastic cells in osteonal bone. Osteoblasts were negative in the periosteum. No megakaryocyte staining was present; however, occasional neutrophils in the bone marrow were non-specifically stained. Demineralization in modified Jenkin's solution (16 days) showed osteopontin localization in bone matrix, hypertrophic and articular chondrocytes, and osteocytes. In cortical bone, almost all cement lines demarcating osteons showed very dense labeling. In the bone marrow, occasional megakaryocytes were immunopositive and neutrophils were non-specifically stained. Jenkin's produced non-specific staining of skeletal muscle and connective tissue. Formic acid demineralization (14 days, 4° C) resulted in osteopontin expression in osteoblasts, osteocytes, osteoclast precursors, bone matrix, osteoid, cement lines, and chondrocytes; osteoclasts, although present in very low numbers, were also positive. More labeled osteoblasts could be identified compared to Jenkin's demineralization. Also more intense non-specific staining of the bone marrow neutrophils was obtained than with Jenkin's. Harsh, rapid demineralization with formic acid (4 days, room temperature) produced a loss in antigenicity demonstrated by a reduction in staining intensity not experienced with the 4° C protocol; however, osteopontin was still localized in bone matrix and hypertrophic zone chondrocytes. These results indicate that demineralization is compatible with retention of immunoreactive osteopontin in adult rat bone. Both EDTA and formic acid demineralization produce excellent immunostaining and are preferred over the modified Jenkin's solution to minimize background levels of non-specific staining.

A commercially available mouse monoclonal antibody to human platelet glycoprotein IIIa was used to demonstrate sequestration of platelets in hepatic biopsies obtained from baboons following intravenous infusion of echistatin, a novel fibrinogen receptor antagonist derived from the venom of the snake Echis carinatus. Biopsies of liver and spleen were taken prior to administration of echistatin. The hepatic biopsies were either snap-frozen in Freon-22/liquid nitrogen or fixed in 10% neutral buffered formalin. Biopsies of spleen were snapfrozen. During infusion of echistatin (2.3 µg/kg/min), circulating platelet counts decreased from 331000/mm3 to 167000/mm3. Selective sequestration within the liver was confirmed using whole body gamma camera imaging to demonstrate 111Indium-oxine labeled platelet accumulation within the liver during the thrombocytopenic episode. Hepatic biopsies were again taken and either snap-frozen in Freon-22/liquid nitrogen or fixed in 10% neutral buffered formalin. Biopsies of spleen and inguinal lymph node were also snap-frozen. Platelet rich plasma smears, included as positive controls, dewaxed paraffin sections, and cryosections of liver, spleen, and lymph node were stained with monoclonal mouse anti-human platelet glycoprotein IIIa using an avidin biotinylated peroxidase complex (ABC) technique. Prior to infusion of echistatin, platelet staining within the liver was minimal. After echistatin infusion, hepatic cryosections showed prominent platelet staining within hepatic sinusoids. No localization was shown in lymph node, however, the spleen showed prominent platelet staining both before and after echistatin infusion. Platelet rich plasma smears were intensely positive. No prominent platelet staining was observed in formalin-fixed, paraffin-embedded material. Thus, this immunocytochemical technique may help localize platelets in cryosections of tissues from baboons and other primate species.