Quantitative and qualitative analysis of cellular prion protein (PrP^C) expression in bovine somatic tissues

Abstract

The host encoded cellular prion protein (PrP^C) is an N-linked glycoprotein tethered to the cell membrane by a glycophosphatidylinositol (GPI) anchor.  Under certain conditions, PrP^C can undergo conversion into a conformationally-altered isoform (PrP^Sc) widely believed to be the pathogenic agent of transmissible spongiform encephalopathies (TSEs).  Understanding the tissue-specific expression of PrP^C is crucial considering that cells expressing high levels of PrP^C bear a risk for conversion and accumulation of PrP^Sc.  In the present study, fifteen bovine somatic tissues were analyzed for PrP^C expression by quantitative western blot and immunohistochemistry.  Quantitative western blot analysis revealed highest expression of PrP^C in cerebellum, obex and spinal cord. Intermediate levels were detected in thymus, intestine, nerve, heart and spleen, and lower levels in lung, muscle, kidney, lymph node, skin, pancreas and liver.  Immunohistochemical analysis detected intense cellular-specific PrPC staining in neurons, thymocytes and lymphocytes. PrP^C was also detected in the enteric wall, pancreatic islets of langerhans, myocardium, pulmonary alveolar sacs, renal glomeruli and dermal epithelial cells.  This study demonstrated the quantitatively varied, wide-spread, tissue- and cell-specific expression pattern of PrP^C in bovine somatic tissues. The importance of this study is to lay the foundation for understanding the tissue-specific expression of PrP^C and to consider the potential participation of more bovine tissues in the transmission of BSE infection.

Introduction

The cellular prion protein (PrP^C) is a 250-amino acid glycoprotein commonly found attached by a glycosylphosphotidylinositol (GPI) anchor to lipid rafts in the plasma membrane.1 Through a poorly understood process, PrP^C can undergo post-transcriptional conversion from a predominantly a-helical structure to a mainly β-sheet isoform (PrP^Sc).2,3 Substantial evidence indicates that PrP^Sc is the principal if not the only component of the agent causing TSEs.4–6 This group of neurodegenerative and infectious diseases affects several mammalian species including sheep (scrapie, the prototypical prion disease), humans (CJD and Kuru), deer (chronic wasting disease; CWD) and cattle (BSE).5,7 Although much is known about the role of PrP^Sc in TSEs, the normal function of PrP^C remains enigmatic. One line of investigation proposes that PrP^C is an antioxidant that directly or indirectly promotes detoxification of reactive oxygen species.8 Another hypothesis proposes that PrP^C has cytoprotective activity that blocks internal or environmental factors that initiate apoptosis.9 In addition, several authors have proposed that PrP^C participates in transmembrane signaling processes associated with cellular survival, replication and differentiation.10–12

Expression of endogenous PrP^C is essential for TSE pathogenesis. The requisite role of PrP^C in TSE pathogenesis has been unequivocally demonstrated in experiments using PrP^C knockout mice. PrP null mice (Prnp^0/0) do not express PrP^C and are completely resistant to experimental TSE infection.6 Hemizygous Prnp mice (Prnp^0/+) express roughly half the PrP^C of their wild-type counterparts.13 The interval from experimental TSE exposure to the onset of disease is nearly twice as long in hemizygous (170 days) vs wild-type (290 days) mice. Overexpression of PrP^C in transgenic mice leads to a shortening in the interval from TSE exposure to the onset of disease, with the degree of shortening proportional to the level of overexpression.14 Thus, PrP^C expression level, at least on a whole-animal basis, is a critical determinant of susceptibility to prion disease.

Comprehensive analyses of PrP^C distribution in various somatic tissues have been reported for hamsters,15 mice16 and sheep.17 These studies revealed that PrP^C is expressed in a wide range of tissues throughout the body. However, the level at PrP^C expressed was shown to vary widely between tissues.15,16 For example, brain and lymphoid tissues expressed relatively high levels of PrP^C, while expression in liver was comparatively less.16,17 In TSE-infected animals, infectious PrP^Sc is generally found in tissues that express high endogenous levels of PrP^C but not in tissues that express PrP^C at lower levels.7 Tissues that express relatively high levels of endogenous PrP^C are thus candidates for infection by TSEs. Furthermore, knowledge of qualitative and quantitative patterns of PrP^C expression may assist in understanding the role and function of PrP^C in healthy animals.

The literature on distribution and expression level of PrP^C protein in bovine tissues is sparse. While one study has characterized expression of PrP^C mRNA in cattle,18 no corresponding studies have been reported for the PrP^C protein. This is surprising in light of agricultural importance of BSE and its transmissibility to humans in the form of new, variant CJD (nvCJD).19 Quantification of PrP^C mRNA indicated concordance with PrP^C protein levels in the central nervous system (CNS), thymus, lung, kidney and muscle; whereas, no association was observed in lymph node, spleen and liver.18 A comprehensive assessment of endogenous and tissue-specific expression pattern of PrP^C is crucial in identifying tissue types that may be at risk for BSE infection and transmission. In the present study we sought to analyze and compare expression of PrP^C in fifteen bovine somatic tissues by quantitative western blot and immunohistochemistry. Computerized quantification of western blot bands showed that PrP^C was expressed in all tissues sampled. PrP^C was highly expressed in CNS and thymus and expressed at lower levels in the other tissues examined. Specific antibody staining revealed that PrP^C was expressed in a cell-specific manner in neurons, thymocytes, lymphocytes, keratinocytes, cardiomiocytes and pneumocytes. Wide spread expression of PrP^C in bovine tissues suggests an important biological function for this molecule and expands the variety of tissues that may be at potential risk for infection and transmission of PrP^Sc.

Results

Western blot.

Independent assays comparing the relative levels of PrP^C expression were performed on tissues from three different animals by western blot. PrP^C was detected in all tissue samples in this analysis. PrP^C showed three distinct migration bands corresponding to molecular weights of 35, 28 and 25 kDa (Fig. 1A). Although we did not specifically analyze bands for glycosylation, the 35, 28 and 25 kDa bands are likely associated with the di-, mono- and un-glycosylated forms of PrP^C, respectively.20 The presumptive di-glycosylated form (35 kDa) was the predominant form across all tissues, whereas the presumptive mono- (28 kDa) and un-glycosylated (25 kDa) bands showed more variable intensities. The 35 kDa band dominated in CNS tissues (obex, cerebellum and spinal cord) and thymus, whereas the same form was weakly expressed in pancreas and liver. In peripheral nerve, intestine, lung and heart the 35 kDa band was observed as a doublet. The 25 kDa band was undetectable in peripheral nerve and lymphatic tissues (lymph node and thymus). Immunoreactive GAPDH was observed at 30 kDa (Fig. 1A). Computerized quantification of PrP^C bands revealed the highest (p < 0.05) level of PrP^C expression in the cerebellum, obex and spinal cord (Fig. 1B). Intermediate PrP^C expression levels were found in thymus. Intestine, nerve, heart and spleen showed lower (p < 0.05) levels compared to CNS tissues, however, the lowest (p < 0.05) expression levels were observed in lung, muscle, kidney, lymph node, skin, pancreas and liver (Fig. 1B).

Immunohistochemistry.

In order to localize cellular expression of PrP^C within the tissues analyzed, immunohistochemistry was performed using the anti-PrP SAF-32 monoclonal antibody. Tissues from each of the three animals sampled were analyzed by immunohistochemistry. Results presented are typical for multiple experiments. No staining was observed in the negative control sections incubated with normal horse serum instead of SAF-32 antibody (shown in inserts and some specific sections).

Neural tissues. Among tissues analyzed, the greatest levels and most widely distributed PrP^C immunostaining was observed in the nervous system. PrP^C labeling in the cerebellum was confined to the gray matter and appeared homogenous and diffuse on neuron bodies and the neuropil (Fig. 2A). At the cellular level, immunoreactivity for PrP^C was present in unmyelinated fibers, cells of the granular layer (GL), and stellate and basket cells of the molecular layer (ML) (Fig. 2B). Purkinje cells observed in all the extensions of the central layer showed intense PrP^C staining (Fig. 2B). Similarly, immunoreactivity for PrP^C was intense in neuronal bodies of the solitary tract nucleus in the obex (Fig. 2D). Glia cells, presumably astrocytes and oligodendrocytes observed around neurons showed moderate levels of PrP^C labeling (Fig. 2E). Immunopositivity in cerebellum and obex was supported by the lack of reactivity in the negative controls (Fig. 2C–F and inserts). Immunoreactivity for PrP^C was analyzed in the thoracic portion (Pars thoracalis) of the spinal cord (Fig. 2G). Despite the presence of immunoreactive tracts in the white matter (WM), the majority of the staining was confined to the gray matter (GM). Analysis of PrP^C distribution in peripheral nerves was performed in transverse sections obtained from the sciatic nerve (Fig. 2H). PrP^C labeling was restricted to neural fibers contained in nerve fascicles. No reactivity for PrP^C was observed in the connective tissue forming the perineurium (P).

Lymphoreticular tissues. Lobules in the cortex (Cx) of the thymus were intensely labeled for PrP^C (Fig. 3A). Observation with higher magnification revealed a cell-specific staining associated with thymocytes in the cortical area (Fig. 3B and C). Less intense immunoreactivity for PrP^C was detected in epithelial cells located in the medulla (M) (Fig. 3A). The intense, widely-distributed immunoreactivity observed in the thymus contrasted with a scattered staining detected in the spleen (Fig. 3D). PrP^C-positive cells with the appearance of myeloid dendritic cells (DCs) were located in the perilymphoid zones of the red pulp (RP) immediately adjacent to nodules of white pulp (WP). Higher magnification showed cell-specific staining presumably associated to myeloid DCs (Fig. 3E and F). Mesenteric lymph nodes showed cellular PrP^C staining located in germinal centers (GC) and lymphocytes coronas (LC) of secondary lymphoid follicles (LF) in the cortical area (Fig. 3G). PrP^C-positive cells in the lymph node were presumably lymphocytes of the B lineage and follicular DCs (Fig. 3H and I).

Digestive tissues. Immunohistochemical analysis for PrP^C was performed in the ileum section of the intestine. Staining was intense and restricted to presumably neural cells present in the lamina propia between intestinal crypts, between the inner and outer layers of the muscularis, and parallel to the inner circular muscular layer (Fig. 4A and B). Clusters of parasympathetic ganglion cells associated with the myenteric plexus showed intense PrP^C staining throughout the extension of the ileum sections (Fig. 4C). In pancreas, PrP^C labeling was restricted to the islets of Langerhans (Fig. 4D and E). No staining was observed in the exocrine pancreatic tissue. Liver tissue showed no detectable immunoreactivity PrP^C (Fig. 4F).

Striated and cardiac muscle. PrP^C immunoreactivity for was not observed in the skeletal muscle sections (Fig. 5A). PrP^C labeling was detected in small, unidentified structures located adjacent to cardiomyocytes (Fig. 5B) but no immunolabeling was observed in cardiomyocytes proper.

Lung, kidney and skin. PrP^C labeling in the lung was mainly associated to the alveolar wall (W) (Fig. 6A). In the kidney, PrP^C positive staining was observed in proximal convoluted tubules (CT) and medullary collecting ducts (Fig. 6B). Renal glomeruli (Gl) showed strong PrP^C labeling associated with extraglomerular mesangial cells (EMC) and moderate labeling in podocytes and endothelial cells (Fig. 6B and C). The skin tissue sample displayed PrP^C labeling in keratinocytes localized in the epidermis and outer sheaths of the hair follicle (Fig. 6D). Furthermore, horizontal sections at the level of the dermis allowed the detection of PrP^C staining in sebaceous glands (Fig. 6E).

Discussion

We present here a quantitative and qualitative evaluation of PrP^C expression in 15 bovine somatic tissues. PrP^C expression levels varied widely between tissues and as expected were greater in the CNS and thymus and less in the other tissues examined. The broad tissue distribution of PrP^C expression pattern and wide variation in PrP^C expression level between tissues, together with its conserved molecular structure suggest an important biological function for this protein. However, apart from its central role in prion disease, the normal function of PrP^C is unknown.

Transmission of TSEs, including BSE, can occur through consumption of PrP^Sc-contaminated feedstuffs.21 Ingested PrP^Sc enters the host via the alimentary canal where it subsequently induces conversion of host-encoded PrP^C to PrP^Sc. The mechanism of conversion is poorly understood but is thought to require intermolecular contact between infectious PrP^Sc and host-encoded PrP^C.6 Newly-converted, host-derived PrP^Sc subsequently drives further conversion and replication of endogenous PrP^C.7 While initial conversion and replication may occur in alimentary tissues, neuropathogenesis becomes manifest only after a long incubation period (>2 years in BSE), where PrP^Sc replication reaches the CNS presumably in anterograde fashion through peripheral nerves.7

Evaluation of tissues from TSE-infected animals and humans indicate that the affected tissues correspond to those that express relatively high levels of endogenous PrP^C. For example, in BSE-infected cattle PrP^Sc has been found in myenteric plexi, peripheral neurons, spinal cord and brainstem.22,23 In our quantitative analysis each of these tissues was shown to express relatively high levels of PrP^C. On the other hand, skeletal muscle, spleen, mesenteric lymph nodes, and liver from those BSE-infected animals have not been found to contain detectable PrP^Sc by immunohistochemistry or western blot22 nor have they been shown to be infectious in highly sensitive bioassays in mice.24 Similarly, our quantitative analysis revealed that these tissues express relatively less PrP^C compared to neuronal tissues prone to BSE infection. These observations are consistent with the view that a component of prion infection depends on endogenous PrP^C expression level.6,14

To protect consumers of US beef from BSE, the United States Department of Agriculture has banned the consumption of tissues known to be infectious for BSE in affected animals.22 These specified risk materials (SRM) include brain, skull, eyes, trigeminal ganglia, spinal cord, central portions of the thoracic, lumbar and sacral vertebrae and dorsal root ganglia of cattle 30 month of age and older, and tonsils and distal intestinal ileum of all cattle.25 Our data covered a subset of these tissues (brain, spinal cord and ileum). PrP^C was abundantly expressed in brain and spinal cord. The ileum expressed a somewhat lesser amount than these tissues. Nevertheless, immunohistochemistry revealed intense, localized PrP^C staining in myenteric plexi within the ileal wall. The myenteric plexi lie in close proximity to the intestinal lumen and potentially to any ingested PrP^Sc.23 Ileal myenteric plexi are thus candidate sites for initial conversion of endogenous PrP^C to PrP^Sc during BSE transmission. Overall, SRM tissues correspond to tissues in our study that express relatively high levels of PrP^C.

Our immunohistochemical analysis of the obex showed intense PrP^C immunoreactivity in the neuropil, neuroglia and neuronal bodies of the solitary tract nucleus. This area of the medulla oblongata is commonly used for BSE diagnosis due to consistent early accumulation of PrP^Sc.26 Neuronal bodies expressing increased levels of PrP^C seem particularly sensitive to PrP^Sc neurotoxicity, which may explain the intense neuronal degeneration of the obex during PrP^Sc infection.26,27

Among non-neural tissues examined here, the greatest levels of PrP^C expression were detected in the thymus. Our immunohistochemical analysis showed PrP^C labeling in the thymic cortex associated with thymocytes. Previous studies in mice have described similar immunodetection in epithelial cells of the mouse thymic medulla and cortex.16 Expression of PrP^C is regulated during lymphocyte development in both thymus and bone marrow.28,29 Studies in mice described a rapid accumulation of PrP^Sc in the thymus after intracerebral scrapie inoculation, suggesting an active role of this organ in replication of PrP^Sc.30 While we could find no reports in the literature that evaluated the presence of PrP^Sc or infectivity in the thymus of BSE-infected cattle, the high level of PrP^C expression observed here suggests the thymus as a potential candidate for BSE infection.

Despite the reported involvement of the spleen in TSE pathogenesis, this organ expressed relatively small amounts of PrP^C in cattle. Most cell types and stromal elements did not stain for PrP^C and staining was restricted to a few cells in the marginal zone of the white pulp. Sheep orally inoculated with BSE agent revealed deposition of PrP^Sc in the same marginal zone of the white pulp where myeloid DCs positive cells were observed in our study.31 Despite these findings, other reports have failed to detect PrP^Sc deposition in the spleen of BSE-infected cattle.21,22

In the intestine, we detected PrP^C immunostaining in neural cells occupying the lamina of enteric crypts and in neurons located within intestinal muscularis as well as in the parasympathetic myenteric plexus. BSE transmission under natural circumstances is widely believed to occur via the oral route.21 Following experimental oral inoculation, PrP^Sc accumulation is first observed in the gut and in gut-associated lymphoid and neural tissues.21 Thus, initial conversion and amplification of PrP^Sc likely occurs in PrP^C-expressing intestinal tissues. Our results concur with others that bovine intestinal tissues express PrP^C,32 and could thus serve as a primary substrate for initial PrP^C to PrP^Sc conversion during infection.33

Our analysis of mesenteric lymph nodes revealed PrP^C-positive cells in germinal centers and surrounding areas of the lymphoid follicles. The location of these cells coincides with that of B lymphocytes and follicular DCs.34,35 Both cell types play direct and indirect roles in PrP^Sc pathogenesis. In mice orally inoculated with PrP^Sc, follicular DCs are critical for PrP^Sc replication and accumulation in mouse lymphoid tissues and participate in neuroinvasion.36 PrP^C is expressed in bovine follicular DCs37 and could thus in theory play a role in BSE pathogenesis.

Previous studies have reported expression of PrP^C in murine16 and bovine38 pancreatic tissue restricted to a subset of cells in the islets of Langerhans, suggesting a role for PrP^C in islet function. However, PrP^Sc has not been detected in the pancreas of BSE-infected cattle.22

Evidence of expression of PrP^C in bovine skeletal muscle has important implications for the potential transmission of BSE through the consumption of beef products. Our data showed low (4.2% relative to levels in the cerebellum, p < 0.05) levels of PrP^C in bovine skeletal muscle. This reduced level suggests a low propensity for PrP^C to PrP^Sc conversion and PrP^Sc accumulation even in BSE-infected animals. This observation is consistent with the lack of observed BSE infectivity of skeletal muscle from BSE-infected cattle.21,22

In the kidney PrP^C was detected in glomeruli, convoluted tubules and medullary collecting ducts. These observations confirm and extend the results of a previous report that indicated the presence of PrP^C in bovine glomeruli.39 Scrapie-infected hamsters and CJD patients have been shown to excrete PrP^Sc in urine and PrP^Sc has been detected in the collecting tubules of scrapie-infected hamsters that shed PrP^Sc in urine.40 While in principle the presence of PrP^C in the bovine kidney may render it susceptible to prion infection, PrP^Sc has not been detected in the urine of BSE-infected cattle21 and to our knowledge, kidneys from BSE-infected cattle have not been evaluated for the presence of PrP^Sc or BSE infectivity.

Our study showed expression of PrP^C in bovine skin where it was observed in epidermal keratinocytes. PrP^C was also expressed in sebaceous glands located in horizontal tissue sections of the skin at the level of the dermis and in cells surrounding the inner and the outer root sheet of hair follicles. PrP^C has been reported in human skin as well.41 TSE infectivity and PrP^Sc has been described in the skin of affected rodents and sheep.42 Most PrP^Sc was associated with neurons innervating the skin and the skin thought to have been infected centrifugally via neurons. To our knowledge skin from BSE-infected cattle has not been evaluated for PrP^Sc or infectivity.

Our results support previous studies in cattle and other species that reveal PrP^C expression in a wide range of somatic tissues. In addition, our study reveals that PrP^C expression levels vary over a wide (>10 fold) range in the tissues examined here. The level of PrP^C expression in a particular tissue is important as it may determine its potential for prion infection. Moreover, the identity and the extent to which various cells and tissues express PrP^C may provide suggestions for further studies into the function of this unusual and enigmatic protein.

Materials and Methods

Tissue collection.

Bovine tissues were obtained from three, 13-month-old healthy Angus steers. Animals were slaughtered at an abattoir in the Department of Food Science and Technology at Virginia Tech. Samples of the following tissues were collected within 20 min of slaughter: cerebellum (hemisphere), obex (solitary tract nucleus), spinal cord (Pars thoracalis), peripheral nerve (sciatic), thymus (intrathoracic), spleen, lymph node (mesenteric), ileum, pancreas, liver (caudate lobe), skeletal muscle (semitendinosus), heart (ventricle), lung (cranial lobule), kidney and skin (flank). Samples for western blot were placed in a glass container and frozen on dry ice. For immunohistochemistry, samples were fixed in 10% formalin.

Western blot.

Tissue samples (<700 mg) were thawed and homogenized (10 w/v) in ice-cold lysis buffer (10 mM Tris, pH 7.4, 150 mM NaCl, 1% Triton-X-100, 1% deoxycholate, 0.1% SDS) using a pestle homogenizer (Fisher Scientific, Hampton, NH). Homogenates were centrifuged at 13,500 rpm for 5 min and the supernatants recovered. Total protein concentrations were determined in the raw homogenates using a Bicinchoninic acid (BCA) kit (Pierce, Rockford, IL) according to the manufacturer's instructions. Samples were prepared for SDS-PAGE by mixing 1:1 (v/v) with Laemmli buffer (BioRad Laboratories, Hercules, CA) and heated at 98°C for 5 min. Raw homogenate protein concentrations were used to calculate volumes of Laemmli-treated sample containing 20 µg of total protein, which were then loaded onto to each lane and separated by SDS-PAGE in 12% gels (BioRad Laboratories). Electrophoresis was performed under reducing conditions at 125 V for 60 min. Proteins were then transferred onto polyvynilidine fluoride (PVDF) membranes (Immobilon-FL; Millipore Corp., Billarica, MA) by electroblotting at 100 V for 1 h. Membranes were blocked by immersion in blocking buffer (LI-COR Corporation, Lincoln, NE) for 1 h with agitation. Membranes were probed simultaneously for PrP^C and a reference housekeeping protein, glyceraldehyde-3-phosphate dehydrogenase (GAPDH). PrP^C was probed with a mouse monoclonal antibody directed against amino acid sequence 59–89 located in the N-terminal octapeptide repeat region (SAF-32, 1:400; Cayman Chemical Company, Ann Arbor, MI). GAPDH was probed with a rabbit polyclonal anti-GAPDH antibody (FL-335, 1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA). Both primary antibodies were diluted in 0.1% Tween-20 in blocking buffer and incubated at room temperature for 1 h on a rotating platform. Membranes were then washed four times in 0.1% Tween-20 in 0.1 M phosphate-buffered saline at pH 7.4 (PBS) for 5 min each. To detect bound primary antibodies, membranes were incubated with fluorescent-tagged secondary antibodies against mouse and rabbit IgG (LICOR Biotechnology, IRdyes 700 and 800, respectively) diluted 1:5,000 in blocking buffer plus 0.1% Tween-20 for 30 min on a rotating platform. Pixel intensities of PrP^C and GAPDH-specific bands were detected and quantified as integrated intensity values using an Odyssey infrared imaging system (LI-COR Biotechnology). Expression of PrP^C in each tissue was normalized to GAPDH levels and calculated as a percentage of that in cerebellum, the tissue showing the highest expression level (Fig. 1).

Immunohistochemistry.

Formalin-fixed tissues were embedded in paraffin and sectioned at 5–7 µm using a microtome (Historange, LKB, Bromma, Sweden). Tissue sections were mounted on adhesive slides (Newcomer Supply, Middleton, WI) and incubated overnight at 37°C. Mounted tissues were deparaffinized in xylene and dehydrated in serial alcohol solutions. Slides were subjected to an antigen unmasking protocol by autoclaving at 120°C for 5 min in an unmasking solution (Vector Laboratories, Burlingame, CA). Endogenous peroxidase was blocked by incubation in 3% hydrogen peroxide diluted in PBS for 30 min. Tissues were then rinsed two times in PBS and blocked in 2.5% horse serum for 15 min. PrP^C was specifically detected by overnight incubation at room temperature with primary antibody SAF-32 (1:400) diluted in 1.5% equine serum solution (Vector Laboratories). After two washes in PBS, bound primary antibody was detected using a horseradish peroxidase-tagged horse anti-mouse secondary antibody (Vector Laboratories) for 10 min at room temperature. Immune complexes were visualized using 3,3′-diaminobenzidine (DAB) substrate for 5 min or until the signal became visible. Probed sections were then counterstained with hematoxilyn and rehydrated in serial alcohol solutions. Sections were mounted with Permount mounting medium (Fisher Scientific) and coverslides. Neighboring sections were used for the following experimental controls: (1) positive control using obex tissue, (2) replacement of the primary antibody with non-immune serum, (3) omission of the secondary antibody, and (4) incubation with diaminobenzidine solution alone. Digital photos of tissue sections were obtained using bright microscopy (Olympus Vanox-T, Tokyo, Japan).

Data analysis.

Tissues for each of the three animals were run in three independent gels and quantified separately. Values of expression of PrP^C were transferred to a spreadsheet and then analyzed using SAS software (version 9.3.1, SAS Institute Inc., Cary, NC). Data was normalized to logarithmic scale in base 10 for normality and mean values for each animal were compared by one-way ANOVA. PrP^C expression values for individual tissues were compared to the tissue with the highest expression (cerebellum) using Dunnet's t-test. Significant differences (p < 0.05) between tissues were analyzed using Duncan's multiple comparison test.

Abbreviations

BSE	=	bovine spongiform encephalopathy
PrPC	=	cellular prion protein
CJD	=	Creutzfeldt-Jakob disease
PrPSc	=	scrapie prion protein
Prnp	=	prion gene
TSE	=	transmissible spongiform encephalopathy

Figures and Tables

Figure 1 Representative western blot for the expression of PrP^C in bovine tissues. (A) PrP^C displayed three distinct migration bands corresponding to molecular weights of 35, 28 and 25 kDa. PrP^C was detected in all tissues analyzed with greatest expression in CNS tissues (cerebellum, obex, spinal cord). GAPDH was used as control protein (30 kDa) and was present in all samples. (B) Cerebellum, obex and spinal cord showed the highest (p < 0.05) levels of immunoreactivity for PrP^C. Thymus expressed highest levels of PrP^C among non-neural tissues. Different superscripts letters indicate significant differences (p < 0.05). Ce, cerebellum; Ob, obex; SC, spinal cord; Th, thymus; In, intestine; Ne, nerve; He, heart; Sp, spleen; Lu, lung; Mu, muscle; Ki, kidney; LN, lymph node; Sk, skin; Pa, pancreas; Li, liver.

Figure 2 Expression of PrP^C in bovine neural tissues. Transverse tissue section incubated with SAF-32 antibody and stained using peroxidase. (A) PrP^C staining (brown) is intensely present in Purkinje cells (arrows) and cells of the molecular layer (ML) and granular layer (GL) in the cerebellum. Less immunoreactivity is observed in the white matter (WM). (B) Higher magnification shows intense staining in fibers of the ML, Purkinje cells (arrows) and neurons of the GL. (D) In the solitary tract nucleus of the obex, PrP^C is associated to neuronal bodies, neuropil and neuroglia. (E) Higher magnification shows labeling of PrP^C in neuronal bodies, appendixes and glial cells (arrow-heads). (G) PrP^C immunoreactivity in the spinal cord is confined to the gray matter (GM) with low intensity in the white matter (WM). (H) In the sciatic nerve, PrP^C staining is restricted to neural fibers associated in fascicles (F). No PrP^C labeling was observed in the perineurium (P). Inserts and figures C (cerebellum) and F (obex) represent serial sections incubated with non-immune horse serum instead of SAF-32 antibody (negative control).

Figure 3 Expression of PrP^C in bovine lymphatic tissues. (A) PrP^C-specific labeling is greatest in the cortex (Cx) of the thymus and moderate in the medulla (M). (B and C) Higher magnification in the cortex area shows PrP^C positive (arrows) and negative (arrow-heads) thymocytes. (D) In the spleen, staining for PrP^C was observed in perilymphoid zones surrounding nodules of white pulp (WP) (RP, red pulp). (E and F) Higher magnification shows presumably PrP^C positive myeloid DCs (arrow) in the spleen. (G) PrP^C-specific labeling is associated with lymphoid follicles (LF) present in the cortex area of the mesenteric lymph node. (H and I) Higher magnification evidence specific staining for PrP^C presumably associated with lymphocytes (arrow) surrounding the lymphocyte corona (LC) and germinal centers (GC). Inserts represent serial section incubated with non-immune horse serum instead of SAF-32 antibody (negative controls).

Figure 4 Expression of PrP^C in bovine digestive tissues. (A) Sagittal tissue section of the ileum shows PrP^C-specific labeling in lamina propia (*), muscularis (arrowhead) and myenteric plexus (arrow). (B) Higher magnification of neurons positive for PrP^C in the lamina propia (arrow in B and * in A). (C) PrP^C is highly expressed in parasympathetic ganglion cells forming the myenteric plexus (arrow in Fig. A and C). (D) PrP^C positive staining is restricted to the endocrine pancreas in the islets of langerhans. (E) Higher magnification shows specific PrP^C-positive pancreatic endocrine cells. (F) No PrP^C staining was observed in the liver tissue after incubation with SAF-32 antibody. Inserts represent serial section incubated with non-immune horse serum instead of SAF-32 antibody (negative control).

Figure 5 Expression of PrP^C in bovine striated and cardiac muscle. (A) No immunoreactivity for PrP^C was detected in semitendinosus striated muscle after incubation with SAF-32 antibody. (B) PrP^C labeling was observed in unidentified structures outside the cardiac muscle fibers; labeling was not observed in cardiac muscle cells of the myocardium. Inserts represent serial section incubated with non-immune horse serum instead of SAF-32 antibody (negative control).

Figure 6 Expression of PrP^C in bovine lung, kidney and skin. (A) PrP^C-specific labeling was observed associated with the alveolar wall (W) (arrow) (Alveolar Sacs, S; Pulmonary Artery, PA). (B) In kidney, PrP^C immunoreactivity is associated with glomeruli (Gl), proximal convoluted tubules (CT) and collecting ducts in the medulla. (C) Higher magnification of renal glomeruli shows strong PrP^C staining in extraglomerular mesangial cells (EMC). (D) PrP^C staining in the skin is associated with epidermal cells in hair follicles (arrow). (E) Staining was also present in sebaceous glands (SG) located in the dermis. Inserts represent serial section incubated with non-immune horse serum instead of SAF-32 antibody (negative control).

Acknowledgements

Our sincere thanks to Dr. William Huckle, Dr. Ludeman Eng, Dr. Xiang-Jin Meng, Dr. Jill Sible and Dr. Mary Lynn Johnson for their guidance throughout this study. We would also like to thanks Kathy Lowe and Shireen Hafez for their expert advice on IHC studies. This research was supported by NIH grant R21-NS045908 from the National Institute of Neurological Disease and Stroke. Additional funding was provided by the Virginia-Maryland Regional College of Veterinary Medicine.