The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

© The Histochemical Society, Inc.

0022-1554/05/$3.30

929

ARTICLE

Volume 53(8): 929–940, 2005Journal of Histochemistry & Cytochemistry

http://www.jhc.org

Expression of Cellular Prion Protein in the Frontal and Occipital Lobe in Alzheimer’s Disease, Diffuse Lewy Body Disease,and in Normal Brain: An Immunohistochemical Study

Payam Rezaie, Charlie C. Pontikis, Lance Hudson, Nigel J. Cairns, and Peter L. Lantos

Department of Biological Sciences, Faculty of Science, The Open University, Milton Keynes, United Kingdom (PR); Departments of Neuropathology and Neuroscience, Institute of Psychiatry, King’s College London, London, United Kingdom (PR,CCP,LH,NJC,PLL); and Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri (NJC)

SUMMARY

Cellular prion protein (PrP

c

) is a glycoprotein expressed at low to moderatelevels within the nervous system. Recent studies suggest that PrP

c

may possess neuroprotec-tive functions and that its expression is upregulated in certain neurodegenerative disor-ders. We investigated whether PrP

c

expression is altered in the frontal and occipital cortexin two well-characterized neurodegenerative disorders—Alzheimer’s disease (AD) and dif-fuse Lewy body disease (DLBD)—compared with that in normal human brain using immu-nohistochemistry and computerized image analysis. The distribution of PrP

c

was furthertested for correlation with glial reactivity. We found that PrP

c

was localized mainly in thegray matter (predominantly in neurons) and expressed at higher levels within the occipitalcortex in the normal human brain. Image analysis revealed no significant variability in PrP

c

expression between DLBD and control cases. However, blood vessels within the white mat-ter of DLBD cases showed immunoreactivity to PrP

c

. By contrast, this protein was differen-tially expressed in the frontal and occipital cortex of AD cases; it was markedly overex-pressed in the former and significantly reduced in the latter. Epitope specificity ofantibodies appeared important when detecting PrP

c

. The distribution of PrP

c

did not corre-late with glial immunoreactivity. In conclusion, this study supports the proposal that re-gional changes in expression of PrP

c

may occur in certain neurodegenerative disorders suchas AD, but not in other disorders such as DLBD.

(J Histochem Cytochem 53:929–940, 2005)

P

rion protein

(PrP), also designated as CD230, ex-ists in both membrane-bound and soluble secretoryforms (Lehmann et al. 1999). It is encoded by a genelocated on human chromosome 20 and comprises 253amino acid residues with two potential

N

-linked gly-cosylation sites. Cellular prion protein (PrP

c

) can befound on the surface of most cells as a glycophospho-lipid-anchored protein. At least two differing trans-membrane forms have been described. PrP

c

is notablyexpressed by neurons within the central nervous sys-tem (Piccardo et al. 1990; Bendheim et al. 1992;

Taraboulos et al. 1992; McLennan et al 2001), partic-ularly in the hippocampus, neocortex, spinal motorneurons, and cerebellar Purkinje cells (Salès et al.1998; Jansen et al. 2001). Modest amounts of PrP

c

arealso expressed in glial cells within the brain and spinalcord, in peripheral tissues, and in leukocytes.

PrP

c

can undergo posttranslational modification toform the disease-associated or “scrapie” isoform (PrP

Sc

)(Harris 1999). After the production of the abnormalisoform, PrP

Sc

is deposited extracellularly as plaquesor intracellulary as fibrils or prion rods. This confor-mational change of the protein causes PrP

Sc

to becomepartially resistant to proteases and represents the keyevent in the pathogenesis of a group of infectious neu-rodegenerative disorders termed transmissible spongi-form encephalopathies or prion diseases, the mostcommon being Creutzfeldt-Jakob disease (CJD) in hu-

KEY WORDS

PrP

c

immunohistochemistry

neurodegenerative diseases

image analysis

Correspondence to: Payam Rezaie, PhD, Department of Biologi-cal Sciences, Faculty of Science, The Open University, Walton Hall,Milton Keynes, MK7 6AA, United Kingdom. E-mail: p.rezaie@open.ac.uk

Received for publication October 15, 2004; accepted January20, 2005 [DOI: 10.1369/jhc.4A6551.2005].

The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

930

Rezaie, Pontikis, Hudson, Cairns, Lantos

mans (Budka 2003). Clinically, the human disorderspresent with dementia, cerebellar dysfunction, andextrapyramidal signs. Neuropathological hallmarksconsist of the deposition of PrP

Sc

, neuronal loss,spongiform vacuolation, astrocytosis, and microglialactivation. It is thought that prion diseases are a con-sequence of either a loss of normal function or a gainin neurotoxic function of the host PrP (Jansen et al.2001; Martins et al. 2001).

Although much is known of the conversion of PrP

c

to PrP

Sc

, the normal functions of PrP

c

are still unclear(Martins et al. 2001; Lasmezas 2003). Structurally,PrP

c

contains several distinct domains: an

N

-terminalsignal peptide, a series of five proline- and glycine-richoctapeptide repeats, a highly conserved central hydro-phobic segment, and a

C

-terminal hydrophobic regionthat acts as a signal for the glycophospholipid anchor(Harris 1999). PrP

c

may exist in non-, mono-, or di-glycosylated forms, and evidence suggests that the re-gional heterogeneity in PrP

c

glycoforms within thenervous system may contribute to the selective target-ing of neuronal populations by PrP

Sc

(DeArmond et al.1999; Beringue et al. 2003). The prevalence of PrP

c

atsynapses (Kitamoto et al. 1992b) has led to the idea ofa possible function in modulating synaptic transmis-sion (Collinge et al. 1994). Other putative roles in-clude an involvement in ligand uptake, as a chelatoror carrier of metal ions; in signal transduction andregulation of intracellular calcium; in cell adhesionand interaction with extracellular matrix componentssuch as laminin; in the promotion of neurite out-growth and neuronal differentiation; in the modula-tion of neurotransmitter metabolism; in the long-termsurvival of cerebellar Purkinje cells; and in sleep.Although mice devoid of PrP develop normally andshow minimal deficits (Cohen et al. 1994), they ap-pear to be more susceptible to seizures induced byconvulsant agents (Walz et al. 1999). The absence ofPrP

c

also confers susceptibility to oxidative stress orapoptosis (Kuwahara et al. 1999; White et al. 1999;Wong et al. 2001; Roucou et al. 2004). These findingshave fueled the notion that PrP

c

exerts a neuropro-tective function against apoptotic or oxidative stressmechanisms and may regulate the survival of neurons(Roucou et al. 2004).

Very few studies have addressed the expression ofPrP

c

and its involvement in human neurodegenerativediseases other than the prion diseases (Hansen et al.1989; Hainfellner et al. 1998; Esiri et al. 2000; Ferreret al. 2001; Voigtländer et al. 2001; Kovacs et al.2002b). Some studies have suggested that the ex-pression of PrP

c

may be upregulated in certain neuro-degenerative disorders (Esiri et al. 2000; Voigtländeret al. 2001). This would be consistent with a possibleneuroprotective role for this protein (Roucou et al.2004). However, experimental overexpression of PrP

may itself produce pathological alterations (includingfocal vacuolation of the central nervous system, skele-tal muscles, and peripheral nerves) in transgenic micethat harbor a high copy number of PrP transgenes(Westaway et al. 1994). Furthermore, transgenic miceharboring high copy numbers of wild-type mouse PrP

c

spontaneously develop neurological dysfunction in anage-dependent manner (Telling et al. 1996; Perrier etal. 2002), and accumulation of PrP

c

in the cytoplasmappears to be neurotoxic in transgenic mice overex-pressing PrP

c

in the cytosol (Ma et al. 2002).Moreover, the expression of certain forms of host

PrP

c

could alter the susceptibility to prion diseases orbe associated with particular neurodegenerative pro-cesses (DeArmond et al. 1999). For example, the ex-pression of a truncated form of PrP [32-121/134], de-void of the

N

-terminal region in PrP knockout mice,specifically causes ataxia and death of cerebellar gran-ular neurons (Shmerling et al. 1998). Separately, thetoxicity induced by PrP

Sc

should (according to theprion hypothesis) depend on the basal levels of PrP

c

expressed within the nervous system.In the present study, we examined whether the im-

munohistochemical expression of PrP

c

was altered inthe frontal and occipital lobe in two well-character-ized neurodegenerative disorders that share certainclinico-pathological characteristics with CJD: Alzhei-mer’s disease (AD), and diffuse Lewy body disease(DLBD) (also referred to as “dementia with Lewybodies”), and in normal human control cases. DLBDin particular, is considered the second most commonform of dementia in the elderly, after AD (McKeith etal. 1996; Rezaie et al. 1996; Hansen 1997; Förstl1999; Luis et al. 1999; McKeith and O’Brien 1999).We further assessed whether the distribution of PrP

c

correlated with glial reactivity in these cases.

Materials and Methods

Tissue Samples

Frozen brain tissues were obtained from the MRC LondonNeurodegenerative Diseases Brain Bank, (King’s College Lon-don, London, UK). All materials were obtained with informedconsent and approval of the local ethical committees (Table1). Thirty cases were selected for this study. Of these cases,10 were neuropathologically diagnosed with “pure” ADwithout Lewy bodies according to the Consortium to Estab-lish a Registry for Alzheimer’s Disease criteria (Mirra et al.1991) (mean age 70.2 years, range 58–92 years), 10 were di-agnosed with DLBD (mean age 75.2 years, range 58–92years), and 10 were neuropathologically diagnosed as nor-mal, with no history of neurological or psychiatric disorders(mean age 79.3 years, range 66–89 years). The DLBD cases(Table 1) were further classified as “pure” forms if there wasno Alzheimer-type pathology (plaques or tangles) or if theyfulfilled the criteria for a Lewy body variant of AD in whichplaques and tangles were also present (McKeith et al. 1996).

The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

PrP

c

Expression in AD, DLBD, and Normal Brain

931

Further clinical data (e.g., concerning the cognitive status ofthese cases) were not available. Frozen tissues from the fron-tal and occipital lobe were available from all cases for thisstudy. Frontal blocks included Brodmann areas 6/8 (agranu-lar-intermediate frontal cortex) and 24/32 (cingulate cor-tex). Occipital blocks included the calcarine sulcus (Brod-mann area 17/striate cortex). Serial sections were cut fromeach block on a cryostat at

�

20

�

m thickness and stored at

�

70C before immunohistochemistry. Sections from the cer-ebellum of a case of variant CJD (vCJD) (embedded in par-affin wax) were used as positive controls.

Immunohistochemistry

Optimal working dilutions for each of the antibodies target-ing PrP were predetermined on positive control vCJD sec-tions as well as test sections to obtain minimal backgroundreactivity. Immunohistochemistry using antibodies to PrP:3f4 [1:100 dilution; Senetek

, Napa, CA

(Kascsak et al.

1987)], sp40 [1:500 (Lantos et al. 1992)], 12f10 [1:200(Krasemann et al. 1996)], and F89/160.1.5 [Ab3, 1:1300 di-lution; CN Biosciences, Nottingham, UK (O’Rourke et al.1998; Van Everbroeck et al. 1999)] was performed accordingto a standard three-step immunoperoxidase method (ABC-HRP protocol; Dako, Ely, Cambridgeshire, UK), with 3,-3

�

diaminobenzidine (DAB) as chromogen. The putative bind-ing regions of these antibodies to human PrP are shown inFigure 1. Immunohistochemistry was performed on duplicatesections (processed simultaneously in entire batches) per an-tibody reagent (i.e., all frontal or occipital cortical samples as-sayed on the same occasion) to minimize interassay variability.

Briefly, cryostat sections were mounted onto silane-coatedslides and allowed to dry for 1 hr at room temperature be-fore being transferred to a 37C drying oven for a further 1hr. Sections were then immersed in a solution of methanolcontaining 2.5% of a 30% hydrogen peroxide solution for 1hr. Following this, they were briefly rinsed with water,placed in phosphate-buffered saline (PBS) solution (pH 7.6)and incubated for 1 hr with lysis buffer solution consistingof Hank’s balanced salt solution supplemented with 1% ofeach of the following: 1 M magnesium chloride, 1 M cal-cium chloride, bovine serum albumin, and Tween 20. Non-specific binding was blocked by incubating sections for 1 hrwith normal serum diluted 1:10 in PBS [normal rabbit serumfor monoclonal mouse antibodies Ab3, 3f4, 12f10; normalswine serum (Dako) for polyclonal antibody sp40]. Primaryantibody solution (diluted in a 1:200 solution of normal se-rum) was applied to slides and incubated overnight at roomtemperature in a humidity chamber. Sections were washedin PBS (two changes, 5 min each wash) and incubated for 90min with biotinylated secondary antibody [rabbit anti-mouse IgG (Dako) for monoclonals, swine anti-rabbit IgG(Dako) for polyclonals] diluted 1:200 in PBS. After this, thesecondary antibody solution was discarded, and sectionswere washed a further two times with PBS and incubated for90 min with ABC-HRP complex (Dako) prepared 1 hr be-fore use. After two more washes in PBS, sections were time-reacted with DAB (the positive control sections were used asa reference for intensity of staining), counterstained lightlywith Harris’s hematoxylin solution or methyl green solution,dehydrated through graded alcohols, cleared in xylene, andmounted with glass coverslips using DPX mountant (Merck;Lutterworth, UK). Slides were processed as entire batches tolimit variability in immunoreactivity for subsequent imageanalysis. Additional immunodetection of astrocytes [glial fibril-lary acidic protein (GFAP), 1:1000 dilution (Dako)] and mi-croglia [CD68/clone PG-M1, 1:200 dilution (Dako); majorhistocompatiblility complex (MHC) class II/clone CR3-43,1:100 dilution (Dako)] were also assessed qualitatively fol-lowing the standard immunohistochemical procedure alreadydescribed [overnight incubation of sections with the primaryantibody at room temperature, ABC method (Dako)].

Controls

Sections from the cerebellum of a case of vCJD served aspositive controls for detection of PrP. Negative controls in-cluded the following: preincubating test frozen sections with5 and 10

�

g/ml proteinase K, 4 M guanidinium thiocyanate(30 and 60 min), or formic acid (5 and 10 s)—treatmentsknown to abolish detection of host cellular PrP (Bell et al.

Table 1

Control, DLBD, and AD cases investigatedin this study

Case GenderAge

(years) Cause of death

PMdelay(hr)

Normal human brain1 M 58 Myocardial infarction 232 F 63 Myocardial infarction 343 M 64 Acute pulmonary edema 484 M 66 Ischemic heart disease 675 F 68 Ischemic heart disease 426 M 69 Congestive heart failure 247 M 70 Peritonitis 378 M 75 Myocardial infarction 349 M 77 Myocardial infarction 96

10 F 92 Myocardial infarction 27DLBD

11

a

M 58 Bronchopneumonia 1912 M 71 Bronchopneumonia 5413 M 72 Shock/hypothermia 2614

a

M 73 Bronchopneumonia 615 M 76 Septicemia 3016 M 77 Bronchopneumonia 917 M 77 Bronchopneumonia 3118

a

M 78 Bronchopneumonia 719 F 78 Bronchopneumonia 1020

b

F 92 Cardiovascular disease (?) 8AD

21 F 66 AD 2122 M 68 Bronchopneumonia 2723 M 78 Coronary artery occlusion 2524 F 80 Bronchopneumonia 925 F 80 AD 526 F 81 Bronchopneumonia 2427 F 81 Bronchopneumonia 2928 M 84 Septicemia 429 M 86 Ischemic heart disease 4830 F 89 Bronchopneumonia 17

a

“Pure” DLBD.

b

DLBD with additional cardiovascular disease.Unless indicated otherwise, all DLBD cases had concomitant AD pathology(Lewy body variant of AD) according to the Newcastle Criteria for classifica-tion of DLBD (McKeith et al. 1996).

The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

932

Rezaie, Pontikis, Hudson, Cairns, Lantos

1997). Additional controls included preincubation of testand vCJD sections with normal blocking serum, or nonspe-cific mouse IgG alone (omitting incubation with primary an-tibody).

Quantitative Image Analysis of PrP

c

Reactivity

Image analysis was conducted to assess the relative levels ofexpression of PrP

c

in the following brain areas from allcases: frontal gray matter, to include the agranular/interme-diate frontal and cingulate cortex (Bodmann areas 6/8, 24/32); frontal white matter; occipital striate cortical gray mat-ter (Brodmann area 17); and corresponding occipital whitematter. Quantitative image analysis has been previously em-ployed to assess PrP immunoreactivity and associated pa-thology in studies on CJD (MacDonald et al. 1996; vonEitzen et al. 1998). The profile of immunoreactivity for eachantibody was assessed quantitatively (mean percentage ofimmunoreactive product per defined area) according to es-tablished protocols (Al-Sarraj et al. 2004; Pontikis et al.2004). The analysis was performed blind with respect toneuropathological details of the cases investigated on dupli-cate sections in the brain regions specified above. Field selec-tion was randomly determined by the image analysis pro-gram controlling the movement of the microscope stage(Image-Pro Version 4.2 and Optimas version 6.2; Media Cy-bernetics, Silver Spring, MD).

A survey of the immunoreacted tissue sections was per-formed by two independent operators (PR, CP) to verify spe-cific immunoreactivity in duplicate sections subsequently pro-cessed to quantitative image analysis. Briefly, nonoverlappingred-green-blue (RGB) images were digitally captured at ran-dom within the defined brain areas, providing a systematicsurvey throughout each region for each case. Images werecaptured via a live color video camera (JVC, 3CCD, KYF55B)mounted onto a Zeiss Axioplan microscope with a

�

10 ob-

jective and neutralizing gray filter (Zeiss; Welwyn GardenCity, Hertfordshire, UK). All parameters including the lampintensity, video camera setup, and microscope calibrationwere held constant. The optimal segmentation of immunore-active profiles was analyzed with the Optimas image analy-sis program (Media Cybernetics), using a previously describedsemi-automated thresholding method based on the opticaldensity of the immunoreactive product (Al-Sarraj et al. 2004;Pontikis et al. 2004). Each RGB image was processed using ablue-band filter to assess DAB reactivity (in brown), and thethreshold was selected to define foreground immunostaining.Foreground immunostaining was accurately defined accord-ing to averaging the highest and lowest immunoreactivitieswithin the sample population for each antibody [measuredon a scale of 0 (100% transmitted light) to 255 (0% transmit-ted light) for each pixel]. A semi-automated histogram-basedprotocol and a manual RGB color cube–based protocol (spec-ified in the image analysis programs) were employed to deter-mine the optimal segmentation (threshold setting) for each an-tibody, rated by two independent investigators (PR, CP). Oncethis optimal segmentation was selected, the chosen thresholdsetting was then applied as a constant to all subsequent im-ages analyzed for this antibody. Furthermore, the specificityof the detection method was also verified manually by moni-toring the analysis as it progressed, per region, per case. Theimmunoreactive profiles were discriminated according to theirintensity (a property that reflects the intensity of the anti-gen–antibody immunoreacted complex), to determine thespecific immunoreactive area (the mean pixel gray value wasobtained by subtracting the total mean gray value from thenonimmunoreacted value per defined field). Each field mea-sured 475

�

m wide, with a height of 320

�

m, the total areaassessed for each region corresponding to 19

�

12.8 mm.Macros were subsequently recorded to transfer the data to aspreadsheet. The data

were analyzed using the Statistical Pack-

Figure 1 Binding sites of the antibodies used in this study to the human prion protein (PrP). The putative binding regions of antibodies tothe human PrP (amino acids 1-253) are indicated. The full-length amino acid sequence of human PrP comprises a signal peptide sequence(AA1-23), octapeptide repeat sequences (AA51-95) for five Cu2� binding sites, a highly conserved hydrophobic central region in cellular PrP(PrPc) that is neurotoxic in “scrapie” PrP (PrPSc) (AA106-126), proteolytic cleavage sites (AA111-112,**), an �-helical region (AA144-154), twoglycosylation sites (AA181 and AA197), and a glycophospholipid anchor (AA229-253) that can be enzymatically cleaved. Of the antibodiestested, F89/160.1.5 (Ab3) was raised against synthetic peptide residues 146–159 of bovine PrP (O’Rourke et al. 1998), and recognizes resi-dues 135–145 in humans (Van Everbroeck et al. 1999). 12f10 binds to helix region 2 of the human PrP, residues 142–160 (Krasemann et al.1996). 3f4 was raised against synthetic peptide Met-Lys-Hist-Met of human PrP (Kascsak et al. 1987). The polyclonal antibody sp40 wasraised against sheep PrP residues 219–232, and recognizes residues 220–229 in man (Lantos et al. 1992). From the figure it can be seen thatF89/160.1.5 (Ab3) and 12f10 map at either end of a sequence within the core of the protein [overlapping region indicated in bold (gsdy)].The sp40 binding region is near the C terminus, whereas the 3f4 binding region is near the N terminus.

The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

PrP

c

Expression in AD, DLBD, and Normal Brain

933

age for Social Sciences (SPSS v.11) program (SSPS Inc.; Chi-cago, IL). Data were plotted as the mean percentage area ofimmunoreactivity per field

�

SEM for each brain region(Figure 3). Statistical significance was assessed using Mann-Whitney

U

and Wilcoxon

Z

tests. Linear regression andSpearman correlation tests were used to assess the relation-ship between PrP

c

expression, postmortem delay, and age. A

p

value of 0.05 or less was considered significant.

Qualitative Analysis of Glial Reactivity

A simple graded scoring system was adopted for visual inter-pretation of PG-M1, MHC class II, and GFAP immunoreac-tivities, using a standard light microscope with a

�

20 objec-tive. The scoring was as follows: 0, no reactivity; 1, weak/mild reactivity on a few scattered cells; 2, moderate reactiv-ity; 3, intense reactivity on numerous cells. For cases of AD,plaque-associated glial reactivity was further assessed ac-cording to the following graded scores: 0, no reactivity; 1,weak/mild reactivity on one or two plaque-associated cells;2, moderate reactivity surrounding plaques; 3, intense reac-tivity on many cells associated with plaques. The results arepresented in Table 2.

Results

Microscopic Analysis

The four antibodies demonstrated variability in theirdetection of host PrP

c

. F89/160.1.5 (Ab3) and 12f10showed similar and consistent immunoreactivity pro-files for PrP

c

. By contrast, 3f4 showed variable and in-consistent reactivity, whereas sp40 was least effectiveat detecting PrP

c

in all samples (data not shown). Nev-

ertheless, all four antibodies proved to be reliablewhen detecting PrP

Sc

on positive control sections ofvCJD cerebellum. In our study, pretreatment with 10

�

g/ml proteinase K (Figure 2G), 4 M guanidiniumthiocyanate (60 min), and formic acid (10 s)—treat-ments known to denature PrP

c

—abolished the detec-tion of PrP

c

, thus confirming specificity of the anti-bodies. In most of the cases examined, the expressionof PrP

c

was confined to the deeper cortical layers andwas particularly notable as a distinctly visible diffuseband at the boundary between the gray and whitematter (Figures 2A–2C). This pattern of immunoreac-tivity was suggestive of an association with projec-tions of thalamocortical fibers within these regions.Microscopically, PrP

c

expression was predominantlycellular (it was mainly present in neurons) (Figure 2F)but was also diffuse within the neuropil of the graymatter (Figures 2D, 2H, and 2L). By contrast, andwith the exception of DLBD cases, immunoreactivitywas very weak and diffuse when present within thewhite matter (Figures 2E and 2I). We further notedthat PrP

c

was weakly associated with blood vessels ofcontrol and AD cases (Figures 2J and 2K). However,this was markedly upregulated on blood vessels, par-ticularly within the white matter in the majority ofDLBD cases (Figures 2M–2O).

Quantitative Image Analysis of PrP

c

Expression

The quantitative analysis of PrP

c

immunoreactivity de-tected with Ab3 (F89/160.1.5), 12f10, and sp40 isshown graphically in Figure 3. There were significantdifferences in the expression of PrP

c

(

p

�

0.05) betweengray and white matter and between occipital and fron-tal cortex in cases of AD and DLBD and in controls.Immunoreactivity with 3f4 antibody was variable andinconsistent within the frontal and occipital cortex ofthe cases examined (data not shown). Of the four anti-bodies, sp40 was least effective at detecting PrP

c

inall samples analyzed (Figure 3C). By contrast, F89/160.1.5 (Ab3) and 12f10 demonstrated consistency indetecting PrP

c

in all samples (Figures 3A and 3B). Thesimilarity in immunoreactivity profiles demonstratedwith F89/160.1.5 (Ab3) and 12f10 may be associatedwith an overlap in the binding region within the coreof the human PrP recognized by these two antibodies(Figure 1). This was not shared by 3f4, which recog-nizes four amino acid residues (containing the cleav-age site for PrP) closer to the

N

-terminal of the PrP.Similarly, sp40, which recognizes a sequence nearthe

C

-terminal, was far less effective at immunohisto-chemical detection of PrP

c

at optimal dilutions. Thefact that all four antibodies proved to be highly effec-tive at detecting PrP

Sc

on positive control sections sug-gests that the differential detection was more likely tobe related to the epitope specificity of these antibodies,associated with regional variations in the tertiary

Table 2

Qualitative analysis of PG-M1 (CD68) and MHC class II immunoreactivity in the frontal and occipital cortex in casesof DLBD and AD and in normal controls

PG-M1 MHC

Control DLBD AD Control DLBD AD

FGM 1 (1) 1 (2) 1 (1) 1 (2) 1 (1) 1 (2)FWM 1 (2) 1 (2) 2 (2) 2 (2) 2 (3) 3 (3)OGM 1 (1) 1 (2) 1 (1) 1 (1) 1 (1) 1 (2)OWM 1 (1) 1 (2) 2 (3) 2 (2) 2 (2) 3 (3)FGMpa — — 2 (2) — — 3 (3)OGMpa — — 2 (2) — — 2 (3)

The following graded scoring system was used for qualitative assessment ofimmunoreactivity: 0, no reactivity; 1, weak/mild reactivity on a few scatteredcells; 2, moderate reactivity; 3, intense reactivity on numerous cells. For casesof Alzheimer’s disease, plaque-associated glial reactivity in the frontal andoccipital cortical gray matter was further assessed according to the followinggraded scores: 0, no reactivity; 1, weak/mild reactivity on one or two plaque-associated cells; 2, moderate reactivity surrounding plaques; 3, intense reac-tivity on many cells associated with plaques. Data are presented as medianscores, with maximum scores per region indicated in brackets (

n

10 cases foreach patient group). Examination of PG-M1 and MHC class II immunoreactivi-ties revealed no marked differences between control and DLBD brains. Bycontrast, immunoreactivity for both markers was higher in AD cases: particu-larly in association with perivascular reactivity within the white matter andspatially distributed focal clusters of reactivity within the gray matter, consis-tent with plaques. Both PG-M1 and MHC class II revealed the presence of nu-merous activated microglia and perivascular cells. FGM, frontal cortical graymatter; FWM, frontal white matter; OGM, occipital gray matter; OWM, occip-ital white matter; FGMpa, plaque-associated cellular immunoreactivity in thecortical gray matter of the frontal lobe; OGMpa, plaque-associated cellularimmunoreactivity in the cortical gray matter of the occipital lobe.

The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

934 Rezaie, Pontikis, Hudson, Cairns, Lantos

structure (i.e., glycosylation) of PrPc, although this re-quires further investigation.

Quantitative image analysis of the immunoreactiveprofiles detected using F89/160.1.5 (Ab3) and 12f10confirmed that the expression of PrPc was predomi-nantly confined to the gray matter, with greater levels(3- to 8-fold) detected in the occipital than frontal cor-tex of normal control and DLBD cases (Figure 3A,B).Expression of PrPc did not differ significantly betweenDLBD and control cases in any of the areas examined.By contrast, both F89/160.1.5 (Ab3) and 12f10 de-tected markedly higher levels of PrPc (p�0.05) withinfrontal cortical gray matter in AD and lower levelswithin the occipital cortex compared with DLBD andcontrol cases. A slightly higher level of expression wasalso detected within the frontal cortical gray matter ofAD cases with sp40 (Figure 3C), reflecting to a muchlesser degree the results obtained using F89/160.1.5(Ab3) and 12f10 in this region. Although linear re-

gression and correlation analyses showed no statisti-cally significant relationship between PrPc expression,postmortem delay or age, a general reduction in theexpression of PrPc (detected using 12f10 and F89/160.1.5) was noted with advancing postmortem delay(data not shown). There was no significant correlationof PrPc levels with increasing age. However, the num-ber of cases may have been limiting, and a greater co-hort of cases is required to verify these observations.

Qualitative Analysis of Glial Markers

No significant qualitative differences in the expressionof PG-M1 and MHC class II (markers that detect mi-croglia) were noted between cases with DLBD andcontrols (Table 2). Immunoreactivity for GFAP (amarker for astrocytes) was very weak, diffuse, and in-consistent on frozen sections, and was therefore notassessed qualitatively. Although there was an upregu-lated microglial response in AD (mainly perivascular

Figure 2 Patterns of PrPc immunoreactivity in the frontal and occipital cortex, and immunohistochemical expression of cellular prion pro-tein (PrPc) in normal aged brain and in Alzheimer’s disease (AD) and diffuse Lewy body disease (DLBD) 12f10 (A,B) and 3f4 (C) immunoreac-tivity in the frontal cortex (A,C) and occipital cortex (B). PrPc expression is localized within deeper cortical layers in the majority of cases, par-ticularly at the boundary between gray and white matter within the cingulate gyrus (A,C) as well as in the occipital cortex (B). (A) and (B)are representative sections immunostained for PrP from a control case, and (C) is from a case with AD. F89/160.1.5 (Ab3) immunohistochem-istry in the frontal and occipital cortex of a normal aged brain (D–G), a case of AD (H,I) and in DLBD (L,M). (J,K,N,O) 12f10 immunohis-tochemistry. Immunoreactivity within the gray matter shown in (D,H,L) was more intense and widespread compared with that within thewhite matter (E,I,M), which showed weak background reactivity in control and AD cases. Higher power images of PrPc expression within thegray matter (F) demonstrate specific cellular and pericellular reactivity, associated with neurons. This staining within the gray matter can beabolished by preincubation with 10 �g/ml proteinase K (G), 4 M guanidinium thiocyanate, or formic acid, confirming the detection of hostPrPc. Within the white matter and to a lesser extent the gray matter (J,K) of control and AD cases, PrPc is weakly associated with blood ves-sels. By contrast, vessel-associated expression of PrPc was upregulated within the white matter in the majority of DLBD cases (M–O). How-ever, this expression was mainly associated with small- to medium-sized vessels intermittent throughout the white matter. PrPc reactivity as-sociated with these vessels did not colocalize with GFAP or MHC class II reactivity in DLBD cases. Abbreviations: CC, corpus callosum; CG,cingulate gyrus; GM, gray matter; nBr, normal brain; PK, proteinase K; WM, white matter. Bars: A–C �0.5 cm; D,E,G,H,I,L,M �125 �m;J,K,N,O �35 �m; F �20 �m.

The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

PrPc Expression in AD, DLBD, and Normal Brain 935

reactivity within the white matter and plaque-associ-ated clusters of cells within the gray matter) (Table 2),the expression of PrPc did not visibly correlate withPG-M1 or MHC class II immunoreactivity. In particu-lar, PrPc expression was not obviously associated withplaques in AD (as determined on the basis of morpho-logical characteristics), which were clearly investedwith microglia. However, further assessments of thecolocalization of PrPc with -amyloid protein are re-quired to clarify this relationship.

DiscussionIn view of the recent reports suggesting that PrPc isupregulated in certain neuropathological conditions(Esiri et al. 2000), including in AD (Ferrer et al. 2001;Voigtländer et al. 2001), we examined the expressionof PrPc in frozen sections from the frontal and occipi-tal lobe in AD and DLBD—two well-characterizedneurodegenerative disorders that share certain clinicaland pathological characteristics with CJD (DeArmond1993; Budka et al. 1995; Hsiao 1997; Haïk et al. 2000;Tschampa et al. 2001; Castellani et al. 2004), and innormal control cases. This is significant because un-like routine paraffin-embedded materials, the tissuesin our study had not been previously subjected tolong-term fixation in aldehyde-based fixatives (nota-bly formalin) and antigenicity was therefore wellpreserved by comparison. Our results indicate thatthe epitope specificity of antibodies directed at PrPc

clearly has an impact not only on the detection of thisprotein in the human nervous system but also on thesubsequent interpretation of the data obtained in bothnormal and pathological conditions. It is possible thatthe use of frozen versus paraffin-embedded materialand fixation of tissues (as well as a host of other fac-tors such as postmortem delay, age, and duration ofillness, where applicable) will also have some bearingon the results obtained in different laboratories usingan immunohistochemical approach.

In our study, we found that of the four epitope-spe-cific antibodies used to detect PrPc, the ones that pro-duced consistent labeling were targeted toward thecentral region of the protein [F89/160.1.5 (Ab3) and12f10] (Figure 2). By contrast, both 3f4 (which bindsto an epitope closer to the N terminus) and sp40(which binds to an epitope close to the C terminus)were far less useful in detecting PrPc in the cases exam-ined, even though all four antibodies proved to be reli-able when detecting PrPSc on positive control sectionsfrom the cerebellum of a case with variant CJD. Pre-treatment with proteinase K, 4 M guanidinium thiocy-anate or formic acid abolished PrP immunoreactivityon all sections apart from the positive vCJD control,demonstrating that the staining on sections repre-sented cellular PrP. The variability in staining and

Figure 3 Quantitative analysis of PrPc expression in the frontaland occipital cortex in cases of AD and DLBD and in normal brain.Data are presented as the mean percentage immunoreactivity �SEM (based on analysis of 400 fields on duplicate sections from 10cases per group). Expression of PrPc was significantly higher withinthe cortical gray matter (p�0.05) than in the white matter in allcases examined. Expression in DLBD did not differ significantlyfrom controls in all brain regions examined. However, there wereapparent differences in the detection of PrPc between the antibod-ies used. F89/160.1.5 (Ab3) and 12f10 displayed similar and consis-tent profiles of immunoreactivity in AD and DLBD and in controlbrain, demonstrating higher reactivity within the occipital cortexof DLBD and control cases than within the frontal cortex. By con-trast, both F89/160.1.5 (Ab3) and 12f10 detected significantlyhigher levels of PrPc (p�0.05) expression within frontal corticalgray matter in AD and lower levels within the occipital cortex com-pared with DLBD and control cases. sp40 was least effective at de-tecting PrPc in all samples analyzed. Even so, there was a slightlyhigher expression detected within the frontal cortical gray mattercompared with DLBD and control cases, reflecting the results ob-tained using F89/160.1.5 (Ab3) and 12f10. Differences in the levelsof detection and regional distribution of PrPc may either reflectvariation in the glycosylation patterns of this protein and/orepitope specificity of the antibodies. FG, frontal gray matter; FW,frontal white matter; OG, occipital gray matter; OW, occipitalwhite matter.

The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

936 Rezaie, Pontikis, Hudson, Cairns, Lantos

poor immunoreactivity that has been obtained previ-ously using 3f4 (Esiri et al. 2000; Salès et al. 1998), aswell as in our study, may either be due to fixation andprocessing of samples, masking of the antibody bind-ing region that depends on the folding of the PrP, orpossibly the normal proteolytic cleavage of PrP by ly-sosomotropic amines at amino acid residues 111 (his-tidine) and 112 (methionine) (Figure 1) (Chen et al.1995; Lehmann et al. 1999). It is recognized thatexcessive cross-linking of proteins which occurs fol-lowing prolonged exposure to aldehydes, as well asconventional paraffin wax histological preparations,markedly affect the antigenicity of proteins. There-fore, detection will depend not only on the antibodyemployed, but also on the extent of protein cross-link-ing within specimens. Because the use of frozen sec-tions should, in theory, prove more reliable, our find-ings point to enzymatic cleavage of the protein as adeterminant for the variability in detection with the3f4 antibody.

PrPc occurs in several conformational states in situwithin the human brain (Salès et al. 1998). However,the regional variability in topological forms of PrPc

(i.e., secreted cleavage products, transmembrane orcytosolic compartmentalized fragments) have notbeen investigated. Use of epitope-specific antibodies incombination with other techniques such as Westernblot and electrophoretic analysis would aid in identi-fying these forms. Due to the limited source of fro-zen material, these studies could not be pursued inthe present investigation. Moreover, considering thatWestern blot examination would be reliant on thenumber of freeze/thaw cycles to which a tissue is ex-posed and that PrPc is readily degraded, the use of thistechnique would be restricted in this respect. Further-more, we were limited in our investigation to examin-ing the frontal and occipital cortices, because thesewere available for all cases. Although these selectedbrain regions are known to be targeted in neurodegen-erative diseases (including prion diseases), it would beimportant to examine more fully the topographicaldistribution of PrPc expression in cases of AD andDLBD and in normal controls.

In our samples, cellular and diffuse PrPc immunore-activity were detected within the cortical gray matter(predominantly associated with neurons), and in a dis-tinct band along the border between gray and whitematter in the majority of cases investigated. Becausethe neocortex is a major target for CJD pathology(Jansen et al. 2001), and the primary structure of thehost PrP influences the distribution of abnormal PrP inthe nervous system (Kitamoto et al. 1992a), this pat-tern of PrPc expression could in theory influence thesynthesis of PrPSc regionally within the brain. Tosupport this, we previously noted that some cases di-agnosed with sporadic CJD showed PrPSc deposits

within the deeper cortical layers (IV–VI), which corre-lated with microglial reactivity in the same region(Rezaie and Lantos, unpublished observations). Simi-lar findings have been reported by other groups (vonEitzen et al. 1998; Verghese-Nikolakaki et al. 1999),and it is possible that the deposition of PrPSc in theseareas precedes glial activation (Rezaie and Lantos2001).

Our findings of PrPc expression confined largely toneurons within the cerebral cortex in normal brainsgenerally support recent reports (Salès et al. 1998; Esiriet al. 2000; Moya et al. 2000; McLennan et al. 2001;Voigtländer et al. 2001). However, there is some dis-crepancy between our results and reports of the distri-bution of PrPc message within the neocortex of thenormal human brain in paraffin-sectioned materials.In situ hybridization studies by Jansen and colleagues(Jansen et al. 2001) found PrPc mRNA mainly in up-per cortical neurons as well as in cerebellar Purkinjecells, and McLennan and colleagues (2001) demon-strated higher levels of PrP mRNA in fixed paraffinhuman tissue in layers II (external granular layer) andIII (pyramidal cell layer) compared with layers V (gan-glionic layer) and VI (fusiform layer). However, theseauthors could not detect PrP mRNA within the sub-cortical white matter; likewise, Salès et al. (1998)found very limited PrPc staining within the white mat-ter of the human brain. Other studies have found nospecific laminar distribution of the PrP message orprotein in the cortex (McLennan et al. 2001; Kovacset al. 2002b). These differences could possibly be ex-plained by differences between the transcription andtranslation of PrPc or by the state of tissues analyzed(frozen as in this study versus paraffin-embedded tis-sues in the other studies). Our immunohistochemicalfindings of minimal PrPc expression within the whitematter of the normal brain are consistent with thesestudies. We also noted minimal expression in thewhite matter of AD cases in our study. By compari-son, immunoreactivity confined to blood vessels in thewhite matter was a finding specific to cases of DLBD.

PrPc has been detected on the surface of human en-dothelial cells (human umbilical vein endothelium,Simak et al. 2002; Starke et al. 2002; human mi-crovascular endothelium, Starke et al. 2002,2003b)maintained in culture, using immunohistochemical andflow cytometric methods. Endothelial cells of bothmacrovascular and microvascular origin maintainedin vitro are also capable of releasing PrPc slowly andconstitutively into the tissue culture medium (Simak etal. 2002; Starke et al. 2002), particularly followingproapoptotic stimulation (Simak et al. 2002). RT-PCRanalysis has further confirmed the presence of PrPc

mRNA in these cultured endothelial cells. Further-more, Starke et al. (2002,2003a) have proposed thatendothelial cells could be a likely source for the solu-

The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

PrPc Expression in AD, DLBD, and Normal Brain 937

ble plasma pool of PrPc. However, it has been ques-tioned whether endothelial cells in situ also expressdetectable levels of PrPc under normal conditions (Si-vakumaran 2003; Sivakumaran et al. 2003). Specifi-cally, Sivakumaran and colleagues (2003) could notdetect PrPc in paraffin sections of human umbilicalcord vessels or adult saphenous vein and aorta usingthe 3f4 antibody. Although these authors noted theexpression of PrPc by neurons in the human cerebralcortex (used as a positive control in their study), theypresented no further data regarding PrPc expressionon blood vessels in the brain. In this study, we foundweak and minimal PrPc expression on blood vessels inthe brain of control and AD cases. A separate study inrodents attributed blood vessel–associated PrP immu-noreactivity to astrocytic foot processes terminatingon these vessels (Verghese-Nikolakaki et al. 1999).However, due to the inconsistent staining patterns forGFAP on our frozen tissue samples, we could not de-termine whether vessel-associated PrP reactivity wasassociated with astrocytic GFAP expression in ourstudy. For this reason, this particular issue remains in-conclusive at present.

Our observation of increased PrPc expression inwhite matter blood vessels in DLBD was unexpected.The significance of this finding is presently unclear.Upregulation of PrPc could be associated in some waywith the vascular pathology that has been noted in thewhite matter of patients with DLBD (particularly thefrontal cortex) (Londos et al. 2000). Another possibil-ity could be related to an upregulation of haparan sul-fate proteoglycans on vascular endothelium, to whichPrPc binds (Warner et al. 2002), in DLBD. However,both of these suggestions are purely speculative atpresent and will require further investigation. It is alsoimportant to note that inflammatory mediators do notappear to alter the expression of PrPc by endothelialcells maintained in vitro (Simak et al. 2003). Onceagain, this observation will require further clarifica-tion in the context of AD and DLBD.

Our findings relating to AD also differ slightly fromthe observations previously noted (Ferrer et al. 2001;Voigtländer et al. 2001; Kovacs et al. 2002a); wecould not detect PrPc immunoreactivity specificallyat the periphery of senile plaques within the cortexof AD cases. Interestingly, Kovacs and colleagues(2002a) noted that out of a panel of 10 antibodies di-rected against PrP, 6H4 (raised against residues 144–152) and 12f10 failed to give this type of labeling atthe periphery of, or throughout, senile plaques. 3f4separately appears to produce variable results whendetecting PrP (Esiri et al. 2000; Moya et al. 2000).Therefore, the use of differing antibodies, fixation, orprocessing (frozen versus paraffin sections in particu-lar); the brain region examined [frontal and occipitalcortex (this study); hippocampus (Ferrer et al. 2001);

hippocampus, subiculum, entorhinal cortex, and tem-poral cortex (Voigtländer et al. 2001); hippocampus,temporal cortex, cerebellum (Kovacs et al. 2002a,b)];or detection of different forms of PrPc (Liu et al. 2001;Warner et al. 2002) may account for these differences.Nevertheless, we are in agreement with these authorsthat PrPc immunoreactivity is likely to be upregulatedregionally in AD (in this case, the frontal cortical graymatter), depending on the use of specific antibodies(i.e., F89/160.1.5 (Ab3) and 12f10 in our study). It isunclear at present whether the accumulation of PrPc

occurs as a result of increased synthesis, reduced deg-radation or vice versa.

Recent in vitro experiments suggest that neurotox-icity of the PrP is dependent on the presence of mi-croglia (for a review, see Rezaie and Lantos 2001) andastrocytes (Brown 1999). Microglia represent the resi-dent mononuclear phagocytes in the central nervoussystem (CNS), whose primary function is to respondto pathological insults by becoming activated to per-form phagocytosis, present antigen (via MHC class IImolecules), and release various inflammatory media-tors (Rezaie and Lantos 2001). Their activation corre-lates with deposition of PrPSc and occurs before neu-ronal cell death in vivo; microglia are located inregions of plaque formation, vacuolation, and accu-mulation of PrP in the form of diffuse deposits orplaques (von Eitzen et al. 1998). Given these observa-tions, we determined qualitatively whether glial reac-tivity corresponded spatially with the PrPc expressiondetected in DLBD or AD brains compared with con-trols. The results from the few studies that have inves-tigated the microglial response in DLBD have beensomewhat contradictory (Mackenzie 2000; Roze-müller et al. 2000; Shepherd et al. 2000; Mackenzie2001). However, we found that PG-M1 (CD68) andMHC class II immunoreactivities were similar withinthe frontal and occipital cortex of control and DLBDcases—there was no overt evidence for a microglial re-sponse in the frontal or occipital cortex of either“pure” DLBD cases or the LB variant of AD in ourstudy. This is in keeping with the results of a similarinvestigation by Rozemuller and colleagues (Roze-müller et al. 2000) and Shepherd and coworkers(2000), and may be taken together to further supportthe lack of an intrinsic inflammatory component tothis disorder. By comparison, we found moderate tointense PG-M1 and MHC class II reactivity in ADbrains, with clusters of microglia (corresponding tosites of AD plaques) evident throughout the corticalgray matter and intense, mainly perivascular reactivitywithin the white matter in all samples. Because PrPc

was widely expressed throughout the cortical graymatter in AD and did not colocalize with these focalclusters of microglia, it is more likely that microglialaggregates corresponded to the distribution of senile

The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

938 Rezaie, Pontikis, Hudson, Cairns, Lantos

plaques (identified on adjacent hematoxylin & eosin sec-tions; data not shown) containing -amyloid. There-fore, the expression of PrPc in our study did not cor-relate with glial reactivity per se but was insteadpredominantly associated with neurons in the neor-cortical gray matter.

In conclusion, this study has shown that the expres-sion of PrPc in neocortical gray matter does not differsignificantly between DLBD and control cases. How-ever, differential expression between the frontal andoccipital cortex were noted in AD, with higher levelsof PrPc detected in the frontal cortical gray mattercompared with DLBD cases or controls. Significantly,the frontal cortex and particularly the cingulate gyrus,are brain regions known to be primarily affected inneurodegenerative diseases. The varying levels of PrPc

noted in AD suggest that PrP is involved in some way inthis neurodegenerative disorder. Changes in the levelsof expression of PrPc could reflect a protective cellularresponse against oxidative stress (e.g., upregulation ofPrPc) or associate with susceptibility to neuronal dam-age (e.g., lower level of expression of PrPc) (Kim et al.2004; McLennan et al. 2004; Roucou et al. 2004).Whether these findings are directly related to the regionalspecificity in glycoform patterns and neuron-specific dif-ferences in PrPc glycosylation (Beringue et al. 2003) war-rants further critical assessment. Separately, it is wellrecognized that there is an inflammatory componentin AD, and this includes elevated microglial reactivityand expression of proinflammatory cytokines withinthe CNS (Eikelenboom et al. 2002). There are sub-stantial differences between patients with DLBD andAD in terms of the degree of brain inflammation;cases with DLBD have a much weaker inflammatoryresponse, if present. This observation introduces thequestion of whether an inflammatory response couldregulate the expression of PrPc or whether the func-tion of PrPc could in some way be associated with neu-roinflammation. Future studies should therefore notonly clarify the relative levels of expression of PrPc

within other brain regions known to be affected in AD(such as temporal and parietal cortex, hippocampus,thalamus, and basal ganglia) and in a larger cohort of“pure” cases of DLBD vs cases of DLBD with con-comitant AD pathology, but further assess the topo-graphical expression of this protein in other neurode-generative (e.g., Parkinson’s disease, motor neurondisease, Huntington’s disease) and neuroinflammatorydisorders (e.g., multiple sclerosis).

AcknowledgmentsWe thank Nadeem Khan (Coordinator, MRC London

Neurodegenerative Diseases Brain Bank, Department of Neu-ropathology, Institute of Psychiatry, King’s College London,London, UK), for providing tissue materials used in thisstudy. We are grateful to Dr. Steve Whatley and Prof. Brian

Anderton (Department of Neuroscience, Institute of Psychia-try, King’s College London, London, UK) for the generousgift of the sp40 antibody. 12f10 was a kind gift from Prof. G.Hunsmann, Drs. S. Krasemann and M. Groschup, and Prof.W. Bodemer (The German Primate Centre, Göttingen, Ger-many). This work was partially supported by the EuropeanUnion Biomed-2 Grant QLK2-CT-2001-01924 (“Strategiesfor the Prevention and Treatment of Prion Diseases”). Thework was in part presented as a poster at the 102nd meetingof the British Neuropathological Society.

Literature CitedAl-Sarraj S, Mohamed S, Kibble M, Rezaie P (2004) Subdural he-

matoma (SDH): assessment of macrophage reactivity within thedura mater and underlying hematoma. Clin Neuropathol 23:62–75

Bell JE, Gentleman SM, Ironside JW, McCardle L, Lantos PL, DoeyL, Lowe J, et al. (1997) Prion protein immunocytochenistry-UKfive centre consensus report. Neuropathol Appl Neurobiol 23:25–35

Bendheim PE, Brown HR, Rudelli RD, Scala LJ, Goller NL, WenGY, Kascsak RJ, et al. (1992) Nearly ubiquitous tissue distribu-tion of the scrapie agent precursor protein. Neurology 42:149–156

Beringue V, Mallinson G, Kaisar M, Tayebi M, Sattar Z, JacksonG, Anstee D, et al. (2003) Regional heterogeneity of prion pro-tein isoforms in the mouse brain. Brain 126:2065–2073

Brown DR (1999) Prion protein peptide neurotoxicity can be medi-ated by astrocytes. J Neurochem 73:1105–1113

Budka H (2003) Neuropathology of prion diseases. Br Med Bull 66:121–130

Budka H, Aguzzi A, Brown P, Brucher JM, Bugiani O, Gullotta F,Haltia M, et al. (1995) Neuropathological diagnostic criteria forCreutzfeldt-Jakob disease (CJD) and other human spongiformencephalopathies (prion diseases). Brain Pathol 5:459–466

Castellani RJ, Perry G, Smith MA (2004) Prion disease and Alzhei-mer’s disease: pathogenic overlap. Acta Neurobiol Exp 64:11–17

Chen SG, Teplow DB, Parchi P, Teller JK, Gambetti P, Autilio-Gambetti L (1995) Truncated forms of the human prion proteinin normal brain and in prion diseases. J Biol Chem 270:19173–19180

Cohen FE, Pan KM, Huang Z, Baldwin M, Fletterick RJ, PrusinerSB (1994) Structural clues to prion replication. Science 264:530–531

Collinge J, Whittington MA, Sidle KC, Smith CJ, Palmer MS,Clarke AR, Jefferys JG (1994) Prion protein is necessary for nor-mal synaptic function. Nature 370:295–297

DeArmond SJ (1993) Alzheimer’s disease and Creutzfeldt-Jakobdisease: overlap of pathogenic mechanisms. Curr Opin Neurol6:872–881

DeArmond SJ, Qiu Y, Sanchez H, Spilman PR, Ninchak-Casey A,Alonso D, Daggett V (1999) PrPc glycoform heterogeneity as afunction of brain region: implications for selective targeting ofneurons by prion strains. J Neuropathol Exp Neurol 58:1000–1009

Eikelenboom P, Bate C, van Gool WA, Hoozemans JJM, RozemullerJM, Veerhuis R, Williams A (2002) Neuroinflammation in Alz-heimer’s disease and prion disease. Glia 40:232–239

Esiri MM, Carter J, Ironside JW (2000) Prion protein immunoreac-tivity in brain samples from an unselected autopsy population:findings in 200 consecutive cases. Neuropathol Appl Neurobiol26:273–284

Ferrer I, Blanco R, Carmona M, Puig B, Ribera R, Rey MJ, RinaltaT (2001) Prion protein expression in senile plaques in Alzhei-mer’s disease. Acta Neuropathol 101:49–56

Förstl H (1999) The Lewy body variant of Alzheimer’s disease: clin-ical, pathophysiological and conceptual issues. Eur Arch PsychiatClin Neurosci 249:S64–67.

The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

PrPc Expression in AD, DLBD, and Normal Brain 939

Haïk S, Brandel JP, Sazdovitch V, Delasnerie-Lauprêtre N, Peoc’hK, Laplanche JL, Privat N, et al. (2000) Dementia with Lewybodies in a neuropathologic series of suspected Creutzfeldt-Jakobdisease. Neurology 55:1401–1404

Hainfellner JA, Wanschitz J, Jellinger K, Liberski PP, Gullotta F,Budka H (1998) Coexistence of Alzheimer-type neuropathologyin Creutzfeldt-Jakob disease. Acta Neuropathol 96:116–122

Hansen LA (1997) The Lewy body variant of Alzheimer disease. JNeural Transm Suppl 51:83–93

Hansen LA, Masliah E, Terry RD, Mira SS (1989) A neuropatho-logical subset of Alzheimer’s disease with concomitant Lewybody disease and spongiform change. Acta Neuropathol 78:194–201

Harris DA (1999) Cellular biology of prion diseases. Clin MicrobiolRev 12:429–444

Hsiao KK (1997) From prion diseases to Alzheimer’s disease. J Neu-ral Transm Suppl 49:135–144

Jansen GH, Vogelaar CF, Elshof SM (2001) Distribution of cellularprion protein in normal human cerebral cortex—does it have rel-evance to Creutzfeldt-Jakob disease? Clin Chem Lab Med 39:294–298

Kascsak RJ, Rubenstein R, Merz PA, Tonna-DeMasi M, Fersko R,Carp RI, Wisniewski HM, et al. (1987) Mouse polyclonal andmonoclonal antibody to scrapie-associated fibril proteins. J Virol61:3688–3693

Kim BH, Lee HG, Choi JK, Kim JI, Choi EK, Carp RI, Kim YS(2004) The cellular prion protein (PrPc) prevents apoptotic neu-ronal cell death and mitochondrial dysfunction induced by serumdeprivation. Brain Res Mol Brain Res 124:40–50

Kitamoto T, Doh-Ura K, Muramoto T, Miyazono M, Tateishi J(1992a) The primary structure of the prion protein influences thedistribution of abnormal prion protein in the central nervous sys-tem. Am J Pathol 141:271–277

Kitamoto T, Shin RW, Doh-ura K, Tomokane N, Miyazono M,Muramoto T, Tateishi J (1992b) Abnormal isoform of prion pro-teins accumulates in the synaptic structures of the central ner-vous system in patients with Creutzfeldt-Jakob disease. Am JPathol 140:1285–1294

Kovacs GG, Head MW, Hegyi I, Bunn TJ, Flicker H, HainfellnerJA, McCardle L, et al. (2002a) Immunohistochemistry for theprion protein: comparison of different monoclonal antibodies inhuman prion disease subtypes. Brain Pathol 12:1–11

Kovacs GG, Voigtländer T, Hainfellner JA, Budka H (2002b) Dis-tribution of intraneuronal immunoreactivity for the prion pro-tein. Acta Neuropathol 104:320–326

Krasemann S, Groschup MH, Harmeyer S, Hunsmann G, BodemerW (1996) Generation of monoclonal antibodies against humanprion proteins in PrP0/0 mice. Mol Med 2:725–734

Kuwahara C, Takeuchi AM, Nishimura T, Haraguchi K, KubosakiA, Matsumoto Y, Saeki K, et al. (1999) Prions prevent neuronalcell-line death. Nature 400:225–226

Lantos PL, McGill IS, Janota I, Doey LJ, Collinge J, Bruce MT,Whatley SA, et al. (1992) Prion protein immunocytochemistryhelps to establish the true incidence of prion diseases. NeurosciLett 147:67–71

Lasmezas CI (2003) Putative functions of PrPc. Br Med Bull 66:61–70Lehmann S, Milhavet O, Mange A (1999) Trafficking of the cellular

prion protein. Biomed Pharmacother 53:39–46Liu T, Zwingman T, Li R, Pan T, Wong B-S, Petersen RB, Gambetti

P, et al. (2001) Differential expression of cellular prion protein inmouse brain as detected with multiple anti-PrP monoclonal anti-bodies. Brain Res 896:118–129

Londos E, Passant U, Brun A, Gustafson L (2000) Clinical Lewybody dementia and the impact of vascular components. Int JGeriatr Psychiatry 15:40–49

Luis CA, Mittemberg W, Gass CS, Duara R (1999) Diffuse Lewybody disease: clinical, pathological and neuropsychological re-view. Neuropsychol Rev 9:137–150

Ma J, Wollmann R, Lindquist S (2002) Neurotoxicity and neurode-generation when PrP accumulates in the cytosol. Science 298:1781–1785

MacDonald ST, Sutherland K, Ironside JW (1996) A quantitativeand qualitative analysis of prion protein immunohistochemicalstaining in Creutzfeldt-Jakob disease using four anti prion pro-tein antibodies. Neurodegeneration 5:87–94

Mackenzie IR (2000) Activated microglia in dementia with Lewybodies. Neurology 55:132–134

Mackenzie IR (2001) Cortical inflammation in dementia with Lewybodies. Arch Neurol 58:519–520

Martins VR, Mercadante AF, Cabral ALB, Freitas ARO, CastroRM (2001) Insights into the physiological function of cellularprion protein. Braz J Med Biol Res 34:585–595

McKeith IG, Galasko D, Kosaka K, Perry EK, Dickson DW,Hansen LA, Salmon DP, et al. (1996) Consensus guidelines forthe clinical and pathologic diagnosis of dementia with Lewy bod-ies (DLB): report of the consortium on DLB international work-shop. Neurology 47:1113–1124

McKeith I, O’Brien J (1999) Dementia with Lewy bodies. Aust N Z JPsychiatry 33:800–808

McLennan NF, Rennison KA, Bell JE, Ironside JW (2001) In situhybridization analysis of PrP mRNA in human CNS tissues.Neuropathol Appl Neurobiol 27:373–383

McLennan NF, Brennan PM, McNeill A, Davies I, Fotheringham A,Rennison KA, Ritchie D, et al. (2004) Prion protein accumula-tion and neuroprotection in hypoxic brain damage. Am J Pathol165:227–235

Mirra SS, Heyman A, McKeel D, Sumi SM, Crain BJ, BrownleeLM, Vogel FS, et al. (1991) The Consortium to Establish a Regis-try for Alzheimer’s disease (CERAD). Part II. Standardization ofthe neuropathologic assessment of Alzheimer’s disease. Neurol-ogy 41:479–486

Moya KL, Sales N, Hassig R, Creminon C, Grassi J, Di Giamberar-dino L (2000) Immunolocalization of the cellular prion proteinin normal brain. Microsc Res Tech 50:58–65

O’Rourke KI, Baszler TV, Miller JM, Spraker TR, Sadler-Riggle-man I, Knowles DP (1998) Monoclonal antibody F89/160.1.5defines a conserved epitope on the ruminant prion protein. J ClinMicrobiol 36:1750–1755

Perrier V, Kaneko K, Safar J, Vergara J, Tremblay P, DeArmond SJ,Cohen FE, et al. (2002) Dominant-negative inhibition of prionreplication in transgenic mice. Proc Natl Acad Sci USA 99:13079–13084

Piccardo P, Safar J, Ceroni M, Gajdusek DC, Gibbs CJ (1990) Im-munohistochemical localization of prion protein in spongiformencephalopathies and normal brain. Neurology 40:518–522

Pontikis P, Cella CV, Parihar N, Lim MJ, Chakrabarti S, MitchisonHM, Mobley WC, et al. (2004) Late onset neurodegeneration inthe Cln3�/� mouse model of juvenile neuronal ceroid lipofusci-nosis is preceded by low level glial activation. Brain Res 1023:231–242

Rezaie P, Cairns NJ, Chadwick A, Lantos PL (1996) Lewy bodiesare located preferentially in limbic areas in diffuse Lewy bodydisease. Neurosci Lett 212:111–114

Rezaie P, Lantos PL (2001) Microglia and the pathogenesis ofspongiform encephalopathies. Brain Res Rev 35:55–72

Roucou X, Gains M, LeBlanc AC (2004) Neuroprotective functionsof prion protein. J Neurosci Res 75:153–161

Rozemüller AJ, Eikelenboom P, Theeuwes JW, Jansen Steur EN, deVos RA (2000) Activated microglial cells and complement fac-tors are unrelated to cortical Lewy bodies. Acta Neuropathol100:701–708

Salès N, Rodolfo K, Hässig R, Faucheux B, Di Giamberardino L,Moya KL (1998) Cellular prion protein localization in rodentand primate brain. Eur J Neurosci 10:2464–2471

Shepherd CE, Thiel E, McCann H, Harding AJ, Halliday GM(2000) Cortical inflammation in Alzheimer’s disease but not de-mentia with Lewy bodies. Arch Neurol 57:817–822

Shmerling D, Hegyi I, Fischer M, Blattler T, Brandner S, Gotz J, Ru-licke T, et al. (1998) Expression of amino-terminally truncatedPrP in the mouse leading to ataxia and specific cerebellar lesions.Cell 93:203–214

Simak J, Holada K, D’Agnillo F, Janota J, Vostal JG (2002) Cellular

The

Jour

nal o

f His

toch

emis

try

& C

ytoc

hem

istr

y

940 Rezaie, Pontikis, Hudson, Cairns, Lantos

prion protein is expressed on endothelial cells and is releasedduring apoptosis on membrane microparticles found in humanplasma. Transfusion 42:334–342

Simak J, Holada K, Vostal JG (2003) Expression of cellular prionprotein on vascular endothelial cells: more evidence than contro-versies. Transfusion 43:680–681

Sivakumaran M (2003) The expression of prion protein (PrPc) byendothelial cells: an in vitro culture-induced artefactual phenom-enon? Br J Haematol 121:673–674

Sivakumaran M, Frost C, Ryan P (2003) Expression of cellularprion protein on vascular endothelial cells. Transfusion 43:678–680

Starke R, Drummond O, MacGregor I, Biggerstaff J, Gale R, Camil-leri R, Mackie I, et al. (2002) The expression of prion protein byendothelial cells: a source of the plasma form of prion protein?Brit J Haematol 119:863–873

Starke R, Harrison P, Drummond O, MacGregor I, Mackie I,Machin S (2003a) The majority of cellular prion protein releasedfrom endothelial cells is soluble. Transfusion 43:677–678

Starke R, Harrison P, Gale R, Mackie I, Drummond O, MacGregorI, Machin S (2003b) Endothelial cells express normal cellularprion protein. Br J Haematol 123:372–373

Taraboulos A, Jendroska K, Serban D, Yang SL, DeArmond SJ,Prusiner SB (1992) Regional mapping of prion proteins in brain.Proc Natl Acad Sci USA 89:7620–7624

Telling GC, Haga T, Torchia M, Tremblay P, DeArmond SJ,Prusiner SB (1996) Interactions between wild-type and mutantprion proteins modulate neurodegeneration in transgenic mice.Genes Dev 10:1736–1750

Tschampa HJ, Neumann M, Zerr I, Henkel K, Schröter A, Schulz-Schaeffer WJ, Steinhoff BJ, et al. (2001) Patients with Alzheimer’sdisease and dementia with Lewy bodies mistaken for Creutzfeldt-Jakob disease. J Neurol Neurosurg Psychiat 71:33–39

Van Everbroeck B, O’Rourke KI, Cras P (1999) Immunoreactivity

of the monoclonal antibody F89/160.1.5 for human prion pro-tein. Eur J Histochem 43:335–338

Verghese-Nikolakaki S, Michaloudi H, Polymenidou M, GroschupMH, Papadopoulos GC, Sklaviadis T (1999) Expression of theprion protein in the rat forebrain—an immunohistochemicalstudy. Neurosci Lett 272:9–12

Voigtländer T, Klöppel S, Birner P, Jarius C, Flicker H, Verghese-Nikolakaki S, Sklaviadis T, et al. (2001) Marked increase of neu-ronal prion protein immunoreactivity in Alzheimer’s disease andhuman prion diseases. Acta Neuropathol 101:417–423

von Eitzen U, Egensperger R, Kösel S, Grasbon-Frodl EM, Imai Y,Bise K, Kohsaka S, et al. (1998) Microglia and the developmentof spongiform change in Creutzfeldt-Jakob disease. J NeuropathExp Neurol 57:246–256

Walz R, Amaral OB, Rockenbach IC, Roesler R, Izquierdo I, Caval-heiro EA, Martins VR, et al. (1999) Increased sensitivity to sei-zures in mice lacking cellular prion protein. Epilepsia 40:1679–1682

Warner RG, Hundt C, Weiss S, Turnbull JE (2002) Identification ofthe heparin sulfate binding sites in the cellular prion protein. JBiol Chem 277:18421–18430

Westaway D, DeArmond SJ, Cayetano-Canlas J, Groth D, Foster D,Yan Torchia M, Carlson GA, et al. (1994) Degeneration of skel-etal muscle, peripheral nerves, and the central nervous system intransgenic mice overexpressing wild-type prion proteins. Cell 76:117–129

White AR, Collins SJ, Maher F, Jobling MF, Stewart LR, Thyer JM,Beyreuther K, et al. (1999) Prion protein-deficient neurons reveallower glutathione reductase activity and increased susceptibilityto hydrogen peroxide toxicity. Am J Pathol 155:1723–1730

Wong BS, Liu T, Li R, Pan T, Petersen RB, Smith MA, Gambetti P,et al. (2001) Increased levels of oxidative stress markers detectedin the brains of mice devoid of prion protein. J Neurochem 76:565–572