Codistribution of Amyloid β Plaques and Spongiform Degeneration in Familial Creutzfeldt-Jakob Disease With the E200K-129M Haplotype

Background  Dominantly inherited Creutzfeldt-Jakob disease (CJD) represents 5% to 15% of all CJD cases. The E200K mutation in the prion protein (PrP) gene (PRNP) is the most frequent cause of familial CJD. Coexistent amyloid β (Aβ) plaques have been reported in some transmissible spongiform encephalopathies but to date have not been reported in familial CJD with the E200K mutation.

Objective  To characterize a family with CJD in which Aβ plaques codistribute with spongiform degeneration.

Design  Clinicopathologic and molecular study of a family with CJD with the E200K-129M haplotype.

Setting  Alzheimer disease research center.

Participants  Two generations of a family.

Main Outcome Measures  Clinical, biochemical, and neuropathologic observations in 2 generations of a family.

Results  In this kindred, 3 autopsied cases showed pathologic changes typical for the E200K-129M haplotype, including spongiform degeneration, gliosis, neuronal loss, and PrP deposition. Moreover, 2 of these cases (ages 57 and 63 years) showed numerous Aβ plaques codistributed with spongiform degeneration. APOE genotyping in 2 cases revealed that Aβ plaques were present in the APOE ε4 carrier but not in the APOE ε4 noncarrier. Two additional cases exhibited incomplete penetrance, as they had no clinical evidence of CJD at death after age 80 years but had affected siblings and children.

Conclusions  To our knowledge, this is the first description of Aβ plaques in familial CJD with the E200K mutation. The codistribution of plaques and CJD-associated changes suggests that PrP plays a central role in Aβ formation and that Aβ pathology and prion disease likely in fluence each other. The kindred described herein provides support that PrP^E200K may result in increased Aβ deposition.

Creutzfeldt-Jakob disease (CJD) is the most common human prion disease. Its clinical course is characterized by rapid onset and progression of dementia, myoclonus, and cerebellar, visual, pyramidal, and extrapyramidal dysfunction and invariably culminates in death, usually within a few months of onset.^1-4 Spongiform degeneration, gliosis, and neuronal loss are the major histopathologic hallmarks of the disease. Deposition of the scrapie prion protein (PrP^Sc), the pathologic conformational isoform of the normal cellular glycoprotein PrP (PrP^C), occurs in the brains of affected individuals.²

Sporadic CJD has a worldwide incidence of approximately 1 case per 1 million per year.⁵ Dominantly inherited familial CJD represents 5% to 15% of all CJD cases.⁶ Several mutations have been identified in the prion protein gene (PRNP) (176640). The E200K mutation results in a nonconservative substitution of lysine for glutamate at codon 200^2,7 and accounts for more than 70% of familial CJD.⁸ Disease phenotype, duration, and age at onset are further influenced by the methionine (M)/valine (V) polymorphism at codon 129 in PRNP.^2-4,9 The prevalence of the E200K mutation is especially high among those of Slovakian descent, including ancestors who migrated to Hungary.^2,8,10,11

We describe herein a non-Jewish family of Hungarian heritage with the E200K-129M haplotype, with 5 affected individuals in 2 successive generations who developed ataxia and rapidly progressive dementia. This kindred illustrates several novel and atypical features of CJD with respect to clinical presentation, disease duration, genetics, and pathologic features in which 2 distinct misfolded proteins coexist.

This family emigrated from Hungary (cases I-1, I-2, and II-1) to the United States between 1902 and 1910 (Figure 1). Five adults in 2 successive generations developed ataxic gait, followed by rapidly progressive dementia. Between December 1985 and July 2006 the proband (case II-10), a niece (case III-4), and a nephew (case III-5) were clinically examined by one of us (J.C.M.), and another niece (case III-8) was followed up by a local neurologist. Cases II-10, III-4, and III-8 underwent postmortem examinations.

Two months before her 63rd birthday, this woman developed an inability to walk and recognize family members and had a progressive memory decline. Examination 6 months after symptom onset revealed temporal disorientation, impaired learning and memory, and ataxic gait. Head computed tomography revealed left frontal focal atrophy, which was later confirmed on magnetic resonance imaging (MRI). Neither diffusion-weighted MRI nor cerebrospinal fluid (CSF) levels of the 14-3-3 protein were obtained. An electroencephalogram (EEG) was abnormal, showing frontal diffuse slowing with random sharp waves. The patient died at age 63 years, approximately 11 months after the onset of illness.

At age 61 years, this woman developed progressive difficulty with walking and balance and began repeating statements and stories. Five months later, she had difficulty taking medications correctly, preparing meals, managing household finances, and identifying family members. Eight months after the onset of symptoms, the patient was no longer ambulatory and had no spontaneous conversation. She required full care for dressing and grooming and was incontinent. Her CSF was sampled and was positive for the 14-3-3 protein. Brain MRI with diffusion imaging revealed restricted diffusion signal abnormalities involving the periventricular white matter, bilateral basal ganglia, and thalamus, as previously described in individuals with the E200K-129M haplotype.^12,13 An EEG was abnormal, with excessive generalized diffuse slowing but without periodic activity. The patient died at age 62 years, 13 months after her initial presentation.

At age 52 years, this man presented with gait and balance difficulty and with reduced spontaneous speech. Examination revealed hesitant speech, with errors in auditory comprehension, and unsteady gait. Four months after symptom onset, he was noted to have no spontaneous conversation, poor recall, temporal and geographic disorientation, and wide-based unsteady gait. No involuntary movements were observed. An EEG was abnormal, with mild generalized slowing without periodic waveforms. His CSF was sampled and was positive for the 14-3-3 protein. Magnetic resonance imaging revealed moderate diffuse cortical atrophy. The patient died at age 54 years, approximately 17 months after the onset of gait difficulty.

This woman came to medical attention 6 months after symptom onset at age 56 years. The history indicated deteriorating memory function, with inability to perform accustomed activities and with language dysfunction. The examination was notable for paucity of speech and for difficulty in auditory comprehension. Hypertonicity and bradykinesia were noted in the right upper and lower extremities, and her gait was apraxic and ataxic. The EEG was abnormal, with moderate generalized slowing (left posterior hemispheric predominance) without periodic waveforms. The CSF 14-3-3 protein assay was positive. Magnetic resonance imaging revealed asymmetric increased diffusion in the cortex of the left frontal and posterior parietal lobes and the basal ganglia (Figure 2A and B). Subsequently, she became mute and developed action myoclonus. The patient died at age 57 years, approximately 7 months after the onset of illness.

A brother (case II-7) of the proband (case II-10) died at age 60 years. History from relatives indicated that he had rapidly progressive dementia with features similar to those of the proband.

A sister (case II-2) of the proband (case II-10) died at age 81 years after a stroke. Before death, there were no known motor, coordination, or cognitive deficits. Two of her children (cases III-4 and III-5) developed CJD.

Another sister (case II-4) of the proband (case II-10) died at age 82 years of sepsis. There were no known motor or coordination deficits or cognitive decline before death. One of her children (case III-8) developed CJD.

Paraffin sections from formalin-fixed blocks of cerebral cortex were obtained from 3 family members who underwent autopsy (cases II-10, III-4, and III-8). These were processed using a standard battery of histologic stains, including hematoxylin-eosin, Luxol fast blue, periodic acid–Schiff, and Bielschowsky (silver) to evaluate spongiosis, gliosis, and neuronal loss.

In all cases, immunohistochemistry was performed on deparaffinized, rehydrated, formic acid–pretreated sections. For standard and double PrP and amyloid β (Aβ) immunostaining, sections were treated with hydrochloric acid and microwaved for antigen retrieval. For PrP immunostaining, sections were probed with 3F4 monoclonal antibody (1:3000 dilution¹⁴), incubated with polymer/horseradish peroxidase (EnVision G/2 double-stain system rabbit/mouse; DAKO, Carpinteria, California), and visualized with diaminobenzidine tetrahydrochloride.^3,15 For Aβ immunostaining, sections were probed with 4G8 monoclonal antibody (1:3000 dilution¹⁴), incubated with polymer alkaline phosphatase, and visualized with fast blue BB salt, hemisalt (zinc chloride) (Sigma, St Louis, Missouri), or permanent red substrate (EnVision G/2 double-stain system rabbit/mouse, DAKO). For apolipoprotein E (ApoE) immunohistochemistry, sections were pretreated with citrate buffer and hydrogen peroxide. Sections were incubated with ApoE4 monoclonal antibody (1:15 000 dilution; MBL International Corporation, Woburn, Massachusetts).

Genomic DNA was extracted from frozen brain tissue obtained from cases III-4 and III-8 at autopsy.¹⁶ The PRNP coding region was amplified and sequenced as previously described.^3,4,17 Sequence analysis confirmed the presence of the E200K mutation and excluded other mutations. The codon 129 genotype was determined by digestion of amplified DNA with NspI restriction endonuclease.⁴ Genomic DNA from case II-10 could not be obtained because of a lack of archival frozen tissue.

APOE (OMIM 107741) genotyping was performed using a single-nucleotide polymorphism (SNP) genotyping assay (ABI Real-Time TaqMan; Applied Biosystems, Foster City, California). Briefly, genomic DNA was used for allelic determination of SNP 112 (rs 429358, ϵ4 allele) and SNP 158 (rs 7412, ϵ2 allele) in the APOE gene. Both SNP assays were performed separately in 2 plates for the same samples, and the results were combined for genotypes. Fifty nanograms of DNA was combined with a 1× final concentration of polymerase chain reaction (PCR) mix (TaqMan Universal PCR Master Mix, Applied Biosystems), with a 0.5× final concentration of primers for SNP 112 and SNP 158. Primers for both SNPs were tagged with fluorescent dyes (VIC and FAM; Applied Biosystems). Real-time PCR was performed, and results were tabulated independently for each SNP. Genotypes were combined for the APOE genotype of an individual.

Brain tissue homogenates (10% wt/vol) from frozen brain tissue were prepared in lysis buffer (100mM NaCl, 10mM EDTA, 0.5% NP-40, 0.5% sodium deoxycholate, and 100mM Tris [pH 8.0]). For proteinase K (PK) (Sigma) digestion of PrP, brain homogenates were incubated with PK (100 μg/mL) for 1 hour at 37°C. Digestion was stopped by adding PMSF (phenylmethylsulfonyl fluoride) at a final concentration of 3mM.

Proteinase K–treated and untreated samples were boiled in an equal volume of 2× sodium dodecyl sulfate sample buffer (6% sodium dodecyl sulfate, 5% β-mercaptoethanol, 20% glycerol, 4mM EDTA, and 125mM Tris-hydrochloride [pH 6.8]) for 10 minutes. Proteins were separated using 15% Tris-hydrochloride gels (BioRad, Hercules, California). Proteins were transferred from gels to polyvinylidene fluoride membrane (Immobilon-P; Millipore Corporation, Bedford, Maine) at 70 V for 2 hours at room temperature. Membranes were incubated with blocking buffer (5% nonfat milk in Tris-buffered saline Tween 20) for 1 hour and were then probed with 3F4 monoclonal antibody (1:40 000 dilution) to human PrP residues 109 through 112.¹⁴Finally, membranes were incubated with a horseradish peroxidase–conjugated goat anti-mouse antibody (1:3000 dilution) for 1 hour. The PrP bands were visualized on film (Eastman Kodak, Rochester, New York) using enhanced chemiluminescence (ECL Plus; GE Healthcare, Piscataway, New Jersey) as described by the manufacturer. Western blots were analyzed using commercially available software (Image Acquisition and Analysis LabWorks 4.0; UVP Inc, Upland, California).

For cases II-10, III-4, and III-8, gross examination revealed normal weight for each brain (range, 1300-1430 g). The cerebral hemispheres appeared full, with normal sulcal and gyral patterns (Figure 2C). The cerebellum was grossly normal in these cases (Figure 2D). In all cases, histochemical examination revealed pathognomonic findings for CJD, with spongiform degeneration, neuronal loss, and gliosis throughout the hippocampus, neocortex, basal ganglia, and thalamus (Figure 3A). There was absent to minimal spongiform change noted in the cerebellum for 2 cases (II-10 and III-8). In case III-4, the cerebellum showed a moderate degree of spongiosis in the molecular layer and atrophy and gliosis in the dentate nucleus and the cortex, especially at the level of the vermis.

Prion protein deposition, primarily in diffuse synaptic form and to a lesser extent in clusters of PrP aggregates (Figure 3B), was evident throughout the brain, with sparing of the cerebellum in 2 cases (II-10 and III-8). In case III-4, PrP deposition was present in the cerebellum in granule and plaquelike structures. Moreover, in 2 cases (II-10 and III-8), there was a population of Aβ plaque that codistributed with spongiform degeneration in the hippocampus, neocortex, basal ganglia, and thalamus but not in the cerebellum (Figure 3C-E). Double immunostaining revealed codistribution of Aβ plaques with PrP immunostaining in every region examined where Aβ and PrP abnormal immunostaining occurred (Figure 3E).

Western blot analysis of frozen brain tissue samples from cases III-4 and III-8 was conducted by the National Prion Disease Pathology Surveillance Center, Cleveland, Ohio. The analysis demonstrated the presence of the abnormal PK–resistant prion protein (PrP^Sc), characterized by underrepresentation of the unglycosylated PrP^Sc (21 kDa), consistent with familial CJD with the E200K mutation (Figure 4).⁴

Genetic analysis of cases III-4 and III-8 confirmed the presence of a known familial CJD mutation, E200K. The haplotype for these 2 representative cases was E200K-129M, while the codon 129 genotypes of the normal allele were 129V and 129M for cases III-4 and III-8, respectively. The genotype for case II-10 was unavailable. Given the Aβ burden demonstrated by immunohistochemistry, the APOE genotype was also determined. The APOE genotype for case III-4 was ε2/ε3 and for case III-8 was ε3/ε4. The APOE genotype for case II-10 was inconclusive when assessed by immunohistochemistry using an ApoE4-specific antibody.

Two cases (II-2 and II-4) who were likely E200K mutation carriers exhibited incomplete penetrance, as they had no cognitive or motor deficits before death. However, each had affected siblings and 1 or more affected children.

We describe herein a familial CJD kindred of Hungarian ancestry with the E200K-129M haplotype which features 5 affected cases in 2 successive generations. Distinctive aspects of this kindred include the type and distribution of CJD and Aβ pathologic findings and the apparent incomplete penetrance. The unusual pathologic finding in this family was the presence of numerous Aβ plaques in 2 of 3 cases autopsied (II-10 and III-8), suggesting that Aβ deposition is linked with CJD pathophysiology.

Cortical Aβ plaques are observed in Alzheimer disease (AD) and less commonly in clinically normal older adults.^18,19 However, in this kindred, the cases were all younger than 65 years. The Aβ plaques were unaccompanied by notable neurofibrillary pathology as would be characteristic of AD. The Aβ burden was not restricted to cortical areas and was found in subcortical regions such as basal ganglia. There have been reports of Aβ plaques in other transmissible spongiform encephalopathies such as sporadic and familial Gerstmann-Sträussler syndrome (GSS),²⁰ and iatrogenic CJD, as well as in older sporadic CJD cases^21-33 and in a case from a kindred with a PRNP gene insertion.³⁴ However, to our knowledge, there have been no reports of Aβ plaques in familial CJD with the E200K mutation.

Neither of the 2 cases herein with Aβ plaques had a history of traumatic brain injury, which has been proposed to result in increased Aβ deposition.³⁵ The APP (104760) and PS1 (104311) genes were not screened for mutations, as a review of the literature did not provide conclusive evidence of these mutations coexisting with PRNP mutations and contributing to CJD pathology.^36-38APOE genotyping was pursued because there is a strong association between the ε4 allele and increased Aβ deposition.^25,39 Case III-4, who lacked Aβ plaques, had an ε2/ε3 genotype, whereas case III-8, who had considerable Aβ burden, had an ε3/ε4 genotype. The APOE genotype for case II-10 could not be established because of limited archival clinical material.

The remarkable codistribution of plaques and CJD-associated changes in cases II-10 and III-8 suggests that this Aβ deposition is associated with CJD pathophysiology rather than an independent process. For instance, all 3 cases who underwent autopsy had evidence of spongiform changes and gliosis, which could be present in a region in the absence of Aβ plaques; however, the converse was not the case. This suggests that spongiform change is an upstream event and is required for Aβ deposition. Several lines of evidence indicate that the PRNP genotype and/or PrP^C influence the AD phenotype and that Aβ and PrP^C metabolism are interconnected: (1) The codon 129 methionine/valine polymorphism of the PRNP gene is a risk factor for AD.^40-43 (2) The occurrence of mature Aβ plaques has been reported in 2 familial forms of prion disease (in a study by Collinge et al³⁴ and in the present study). (3) Amyloid β–induced synaptic dysfunction is mediated through its binding to PrP^C.⁴⁴ (4) The formation of Aβ is increased in scrapie-infected mice and in the presence of PRNP pathogenic mutations.⁴¹ (5) The depletion of PrP^C or the presence of disease-associated mutant PrP in mouse N2a cells results in failed β-secretase inhibition, with resultant increase in Aβ levels.⁴² Together, these findings suggest that PrP^C plays a central role in Aβ formation and that Aβ pathology and prion disease likely influence each other. The kindred described herein provides support that PrP^E200K may result in increased Aβ deposition. The effect of the E200K mutation may be modulated by the codon 129 status on the normal allele.^40,42,43,45 For instance, PrP deposition patterns are different for 129MM vs 129MV.^2,46 In case III-4, the codon 129 status was MV and may have resulted in differential PrP deposition and subsequent lack of downstream Aβ deposition.⁴² Also, the APOE status of case III-4 was ε2/ε3, which may have further decreased the likelihood of Aβ deposition. Recognition of this possibility may encourage other investigators to assess potential Aβ deposition in cases with classic CJD pathology.

This kindred also demonstrates a novel genetic feature of incomplete penetrance. Cases II-2 and II-4 had no motor or cognitive impairment before death from acute illnesses at ages 81 and 82 years, yet each had 2 affected siblings and 1 or more affected children. Because the E200K mutation is dominantly inherited, these 2 cases likely were mutation carriers without CJD features because of incomplete penetrance. Ninety-six percent of E200K mutation carriers develop the clinical CJD phenotype if they live past 80 years⁵ except for those of Slovakian heritage (such as this kindred), among whom only 59% penetrance may be reached.^11,47

Another distinctive feature of this kindred is the clinicopathologic presentation. Presenting clinical features of the E200K-129M haplotype typically include cognitive abnormalities in up to 83% of patients and cerebellar signs in up to 55% of patients.² These initial features are followed by development of dementia in all patients, cerebellar signs in 79%, and myoclonus in 73% during the course of the disease.² The clinical presentation in the present family also features ataxia followed by dementia. Classically, ataxia associated with the E200K-129M haplotype is attributed to severe spongiform degeneration, gliosis, and neuronal loss in the cerebellum. However, in this kindred, the cerebellum was spared, with only moderate involvement in 1 case (III-4), a finding that is consistent with its E200K-129M haplotype.² Alternately, the gait disturbance observed in this family could be attributable to involvement of the spinal cord instead of the cerebellum. Cases of an amyotrophic form of CJD have been reported; however, motor signs consistent with motor neuron disease were observed in those instances.^48-51 Spinal cords were unavailable for analysis in the present family. In typical E200K-129M, the mean age at onset is 58 years, with a mean duration of 6 months.^2,11,52 Although the age at onset in the present kindred ranged from 52 to 62 years, the duration of illness was longer, ranging from 7 to 17 months. Other typical CJD features absent in this family were positive sharp waves on EEG and prominent brain atrophy.^2,52

To our knowledge, this is the first description of Aβ plaque pathology in familial CJD with the E200K mutation. The codistribution of Aβ deposition and spongiform degeneration in this family lends credence to the idea that these 2 synaptic proteins, amyloid precursor protein and PrP, may interact to result in disease.

Correspondence: Nupur Ghoshal, MD, PhD, Department of Neurology and Alzheimer's Disease Research Center, Washington University School of Medicine, 4488 Forest Park Ave, Ste 101, St Louis, MO 63108 (ghoshaln@neuro.wustl.edu).

Accepted for Publication: June 12, 2009.

Author Contributions:Study concept and design: Ghoshal, Perrin, Josephson, Sun, and Morris. Acquisition of data: Ghoshal, Cali, Perrin, Josephson, Sun, Gambetti, and Morris. Analysis and interpretation of data: Ghoshal, Cali, Perrin, Gambetti, and Morris. Drafting of the manuscript: Ghoshal, Josephson, Sun, and Morris. Critical revision of the manuscript for important intellectual content: Ghoshal, Cali, Perrin, Gambetti, and Morris. Obtained funding: Ghoshal, Perrin, Gambetti, and Morris. Administrative, technical, and material support: Ghoshal, Cali, Perrin, and Morris. Study supervision: Gambetti and Morris.

Financial Disclosure: None reported.

Funding Support: This work was supported in part by grant T32 NS007205 from the National Institutes of Health/National Institute of Neurological Disorders and Stroke (Drs Ghoshal and Perrin); by grant P01 AG14359 from the National Institutes of Health, by grant CDC UR8/ CCU515004 from the Centers for Disease Control and Prevention, by the Charles S. Britton Fund, and by the CJD Foundation (Dr Gambetti); and by grants P50 AG05681 and P01 AG03991 from the National Institutes of Health (Dr Morris).

Additional Contributions: C. Christopher Clark, MD, and Stuart Weiss, MD, provided clinical records. Richard Torack, MD, and Arie Perry, MD, provided neuropathologic evaluation. Alison Goate, DPhil, Joanne Norton, RN, Sumi Chakraverty, MS, and Pam Millsap, RN, provided genotyping assistance and pedigree development. Manuela Pastore, PhD, conducted initial Western blots and determined glycoform status. Yvonne Cohen, BS, Diane Kofskey, BS, and Deborah Carter, AT, HT ASCP, provided technical assistance.