Novel Prion Protein Gene Mutation in an Octogenarian With Creutzfeldt-Jakob Disease FREE

THE TRANSMISSIBLE spongiform encephalopathies (TSE; also known as prion diseases) are rare neurodegenerative disorders that in humans occur in 3 different epidemiological settings: sporadic, familial, and horizontally transmitted (usually iatrogenic). Creutzfeldt-Jakob disease (CJD), Gerstmann-Sträussler-Scheinker syndrome, and fatal familial insomnia constitute the most common distinguishable vertically transmitted phenotypes. These disorders follow autosomal dominant inheritance patterns and are caused by a range of mutations in the prion protein gene (PRNP) on chromosome 20.^{1 - 28} Surveillance studies report the familial varieties to constitute around 10% to 15% of all human TSE,²⁹ but this may be an underestimate of the heredofamilial forms. Systematic PRNP analyses of CJD patients without a definite family history of a similar neurological disorder not infrequently disclose causal mutations, suggesting that camouflaged or unrecognized family histories may be relatively common.^{3 ,13 ,15 ,18 ,27 - 28}

In this context, we report a novel PRNP threonine-to-alanine codon 188 mutation in an 82-year-old woman with clinical and pathological features typical of CJD who had no known family history of neurological disease. The patient was discovered during routine monitoring activities of the Australian Creutzfeldt-Jakob Disease Case Registry. We postulate that the threonine-to-alanine substitution at codon 188 of PRNP is causally linked to spontaneous spongiform degeneration in humans, and we hypothesize that the mutation may have reduced penetrance or reduced pathogenetic capacity, explaining the advanced age at symptom onset, and this is perhaps why it has not been previously reported.

Two separate samples of brain tissue obtained postmortem were snap frozen and homogenized under liquid nitrogen. A portion of the resulting frozen tissue powder was added to polymerase chain reaction (PCR) buffer to which proteinase K was added. Protein digestion was allowed to proceed overnight at 56°C, after which the samples were boiled.

The open reading frame (ORF) of PRNP was amplified based on the method and primers previously described.² Five microliters of the amplified DNA was added to 1 µL of shrimp alkaline phosphatase and 1 µL of exonuclease I, then incubated at 37°C for 15 minutes followed by 80°C for 15 minutes. Direct DNA sequencing of the PRNP ORF was accomplished by cycle sequencing according to the manufacturer's instructions (ThermoSequenase; Amersham, Cleveland, Ohio). Deoxyribonucleic acid from 2 separate brain pieces was isolated, amplified by PCR, and sequenced on 2 separate occasions to validate findings.

The DNA sequence obtained was then compared with the PRNP sequence deposited in GenBank (http://www.ncbi.nlm.nih.gov; accession number M13899). The adenine-to-guanine transversion detected at nucleotide 611 in the patient was not found in 63 controls, nor was it listed in GenBank.

Postmortem examination was limited to the brain. The brain was fixed in 15% formal saline solution for 2 weeks prior to sectioning. The brainstem and cerebellum were removed by cutting at the level of the midbrain, and the cerebral hemispheres were then sectioned in the coronal plane at 10-mm intervals. The cerebellum was sectioned in the sagittal plane, and the brainstem was sectioned at 3-mm intervals in the horizontal plane. Blocks of tissue were taken from the frontal, parietal, temporal, and occipital cortices, as well as from the thalamus, basal ganglia, midbrain, pons, medulla, cerebellar vermis, and lateral hemisphere. These samples were placed in 95% formic acid for 1 hour prior to paraffin embedding and 10-µm sectioning for routine staining.

Immunocytochemistry was performed to detect prion protein (PrP) in the representative brain regions (including occipital cortex, basal ganglia, and cerebellum) after hydrolytic autoclaving to enhance antigen retrieval.³⁰ The well-characterized monoclonal antibodies 3F4 (Senetek PLC, Napa, Calif) and 6H4 (Prionics AG, Zurich, Switzerland) were used, as well as a departmental polyclonal antibody (PrP-725) raised against a synthetic peptide equivalent to the amino acid sequence corresponding to codons 89 to 106 of PrP. Additional immunodetection of β-amyloid plaques (using the 1E8 antibody [departmental antibody] raised against a synthetic peptide fragment corresponding to amino acids 17 to 25 of β-amyloid), τ (Dako Corp, Carpinteria, Calif), and ubiquitin (Dako Corp) was undertaken. Relevant positive controls were routinely included for all sections, and immunoreactivity was completed using a secondary antibody conjugated to horseradish peroxidase with 3,3′-diaminobenzidine as chromagen.

The method was performed as described previously³¹ with the following modifications. Briefly, 50 µL of cerebrospinal fluid (CSF) was combined with 10 µL of sample buffer (final concentrations: 5% glycerol, 1% 2-mercaptoethanol, 1% sodium dodecyl sulphate, and trace bromophenol blue) and boiled for 10 minutes before separation on denaturing 15% polyacrylamide Tris hydrochloride gels with a 4% stacking component (Bio-Rad Laboratories Inc, Hercules, Calif). After resolution for approximately 40 minutes at 200 V, transfer was effected to a polyvinylidene difluoride membrane (Immobilon-P; Millipore, Bedford, Mass) using 100 V for 1 hour. Filter blocking was for 60 minutes using 0.5% Tween 20 (Sigma-Aldrich, St Louis, Mo) in a 5% (wt/vol) skim milk Tris hydrochloride buffer solution. Primary immunodetection was at room temperature for 1 hour with a 1:1000 dilution of the anti–14-3-3 β antibody (Santa Cruz Biotechnology, Santa Cruz, Calif) followed by incubation (at 1:2000) with a horseradish peroxidase–conjugated polyclonal antirabbit antibody (Amersham PLC, Buckinghamshire, England) before enhanced chemiluminescent detection (Amersham PLC) as per the manufacturer's instructions. Positive and negative controls and colored protein molecular weight markers (Rainbow; Amersham) were included to aid orientation and verify results.

The patient was 82 years old when admitted to the hospital with confusion following a fall at home. During the 2 years prior to admission, family members had recognized minor forgetfulness that did not compromise her ability to live alone and cope independently. There was a remote personal medical history of ischemic heart disease and osteoarthritis but no neurological or ophthalmological surgical procedures.

Preliminary inpatient investigations, including routine biochemical and hematological parameters as well as a computed tomographic head scan, did not reveal an explanation for her confusion. During the first 2 weeks following admission, neuropsychological examinations confirmed disorientation (for day and date), dyslexia, reduced attention, dyscalculia, left-right disorientation, poor short-term memory, and visual-spatial difficulties. Cognition and mood were noted to fluctuate. The initial salient neurological findings were myoclonus affecting the left arm, generalized hyperreflexia, bilateral extensor plantar responses, and frontal release phenomena in the form of a pout reflex. An initial electroencephalogram revealed changes suggesting the possibility of nonconvulsive status epilepticus, but there was no clinical improvement with varying combinations of anticonvulsant therapy using phenytoin, gabapentin, and carbamazepine. During the ensuing weeks, the patient's neurological status progressively declined, evidenced by probable visual hallucinations, cortical blindness, reduced verbal output, and worsening generalized hypertonia. Brain imaging with 2 further computed tomographic scans and a magnetic resonance imaging scan revealed only minor cerebral atrophy consistent with the patient's age. Repeated electroencephalograms eventually showed generalized, synchronous 0.5- to 1-Hz periodic sharp-wave complexes, emphasized over the right hemisphere, typical of sporadic CJD.³² Cerebrospinal fluid analysis revealed a very minor elevation of protein to 0.62 g/L (normal, <0.4 g/L) and minimal pleocytosis, with 10 white blood cells per microliter; using Western immunoblot analysis, the 14-3-3 β protein was detected (Figure 1). The patient died approximately 2 months after admission (total illness duration of approximately 4 months) with a firm clinical diagnosis of CJD.

Postmortem examination of the brain showed no macroscopic abnormalities. Microscopic assessment confirmed changes typical of a spongiform encephalopathy. The changes were most severe in the occipital cortex and in particular, in the striate (visual) cortex (Figure 2). There were moderate neuronal loss and severe gliosis and spongiosus, mainly involving the middle and lower layers of the occipital cortex. The vacuoles were generally small (10-20 µm) but some were as large as 40 µm. The changes were reasonably uniform throughout the occipital cortex but were more focal and less severe in all other cortical regions examined. Occasional senile plaques with β-amyloid cores were noted in the cerebral cortex, but the numbers were only commensurate with age. The thalamus and basal ganglia showed mild diffuse spongiform change with relatively little gliosis or neuronal loss. The cerebellum showed only very minor spongiosis but no obvious gliosis or neuronal loss. No plaques were seen in the cerebellum. The brainstem was normal. Despite the pathognomonic histological appearance, immunohistochemical staining for PrP was repeatedly negative with the 3 different antibodies in all sections from the cerebral cortex, basal ganglia, hippocampus, and cerebellum. Western immunoblot analysis for protease-resistant PrP could not be performed because of the unavailability of fresh-frozen brain tissue. Auxiliary immunostaining showed frequent diffuse β-amyloid plaques and very occasional neuritic plaques in areas of the cerebral cortex, basal ganglia, and hippocampus. Neurofibrillary tangles were restricted to the hippocampus, where ubiquitin was colocalized with the τ-positive neurofibrillary changes. However, because of the age of the patient, these neuropathological features were insufficient to substantiate a Consortium to Establish a Registry for Alzheimer's Disease (CERAD)³³ diagnosis of Alzheimer disease.

In the absence of any hint of familial prion disease, a son requested PRNP genotyping posthumously to assuage his remote concerns of this possibility. Polymerase chain reaction amplification and direct sequencing of the PRNP ORF revealed an adenine-to-guanine substitution at nucleotide 611, causing alanine to replace threonine at codon 188 (T188A). The patient was homozygous for methionine at codon 129.

Unfortunately, because of issues relating to the patient's illegitimate birth, no information was available concerning the medical history of the father or his family, and a protective code of secrecy was adopted by informed matrilineal relatives so that direct sourcing of information from the other surviving maternal relatives was not possible. Working within this constraint, the only history of neurological disease in the patient's mother was sudden death from a stroke at age 75 years (verified on death certificate). Among the 11 maternal half-siblings of the proband, 7 were dead, with the family's understanding of cause of death as "heart" or "stroke" in 5 (ages at death ranged from 48 to 67 years), motor neurone disease in 1 (died at age 65 years), and a motor vehicle accident in 1 (died at age 50 years). The 4 living siblings (ages ranged from 58 to 68 years) were all well. The index case had 3 children, aged 52, 60, and 65 years, who were alive and well. At the time of this report, the proband's children all declined PRNP genotyping, and because of the code of secrecy adopted by the informed family members, none of the other relatives was offered screening for the T188A mutation.

As summarized in Table 1 and Table 2, to the best of our knowledge, this T188A mutation has not been described previously in association with human TSE or as a normal PRNP polymorphism.^{1 - 28 ,34 - 37} We propose that this mutation is causally associated with the pathologically verified spongiform encephalopathy seen in our patient, but we await confirmation of its etiological significance with future reports. Nevertheless, the threonine normally found at codon 188 of PRNP is highly conserved across a broad range of mammalian species³⁸ (probably lying within an α helix³⁹ ), offering further circumstantial support to the likely significance of any mutation at this site. However, the alanine substitution appears to be a relatively conservative primary sequence change, and despite its proximity, it does not appear to directly alter important native conformational determinants, such as the N-linked glycosylation motifs or the disulfide bridge.³⁹ The potentially modest structural impact of the mutation may relate to the advanced age of our patient at symptom onset, which is less typical for PRNP mutations.⁴⁰ This substitution may have less inherent capacity to facilitate the postulated conformational change of wild-type PrP to the relatively protease-resistant, disease-associated conformations thought to underlie the pathological etiology of PRNP mutations.

Camouflaged or unrecognized family histories are not unique to TSE. The range of potential explanations offered for other autosomal dominant neurodegenerative disorders is also relevant to prion diseases.⁴¹ The most plausible explanation for our patient is probable patrilineal inheritance, with the disease unknown to the propositus and her family.

Acceptance of the pathoetiological relevance of the T188A mutation underscores potentially important epidemiological issues, especially for national TSE case registries. Using systematic genotyping, previous reports have also described the detection of recognized (V210I,^{15 ,27} E200K,²⁷ and D178N²⁸ ) and novel (V180I,³ M232R,³ R208H,¹³ and 96–base pair insert^{18 ,28} ) PRNP mutations in patients manifesting CJD in the absence of any known family history of neurological disease. The large, unselected series of Windl and colleagues²⁸ suggests a frequency of approximately 5% (3 of 67 cases). Hence, if the optimal epidemiological classification of prion disease is to be achieved in ascertained cases (especially in cases with incomplete or uncertain family histories), ethical implications notwithstanding, the best practice would appear to necessitate universal PRNP genotyping.

Detection of the 14-3-3 protein in CSF by Western immunoblot analysis has proven very useful in the premortem evaluation of patients with suspected sporadic CJD.^{31 ,42} Although a positive result in appropriately selected patients carries sensitivities and specificities of around 90%, it has proven a less reliable marker in familial prion diseases, although only small numbers of cases have been reported so far.^{42 - 43} The likelihood of a positive result appears to depend at least partly on the specific mutation. Virtually all symptomatic patients with the E200K mutation show positive immunoreactivity for the 14-3-3 β protein in CSF, whereas positivity rates in the D178N and P102L polymorphisms range from 20% to 50%.^{42 - 43} We believe that the 14-3-3 β protein CSF result in the patient described herein contributes to the expanding collective experience of the utility of this diagnostic test in familial TSE.

The consistently negative immunostaining results for PrP in all brain regions, despite using 3 antibodies recognizing different regions of the protein, are at variance with the usual neuropathological findings in CJD. We are unable to offer an explanation for this unexpected finding, but we are confident that it does not relate to a technical fault with tissue handling or preparation or to the antibodies used, as simultaneous control sections always stained positive with all 3 antibodies. Nevertheless, PrP immunostaining does not necessarily colocalize with routine pathological findings, and uncommonly, patients with sporadic CJD may show negative immunocytochemical results.⁴⁴ In addition, patients manifesting genetically determined spongiform encephalopathies, including fatal familial insomnia⁴⁵ and CJD³ (due to the V180I mutation; written communication from T. Kitamoto, February 24, 1999), may show an absence of immunostaining for PrP. Although in fatal familial insomnia, the correlating absence of immunocytochemically detectable PrP and spongiform change may be due to the relatively low intrinsic amounts of protease-resistant PrP in these negative regions⁴⁶ ; this appears to be a less likely explanation in the V180I CJD case reported, which showed typical spongiform degeneration in the cerebral cortex, similar to that in our patient.

Finally, the advanced age of our patient at symptom onset is also of interest. At a practical surveillance level, it emphasizes the heightened vigilance that must be practiced to avoid missing TSE cases falling outside the typical age range,⁴⁷ especially in the elderly, for whom alternative, more common, but incorrect diagnostic explanations may be entertained. The advanced age of a patient may also militate against the rigor of diagnostic evaluation usually afforded younger patients. Consequently, underrecognition of CJD in very elderly patients remains a continuing concern, and as illustrated by the patient described herein, a genetic basis for the disease is possible.

Accepted for publication December 1, 1999.

The Australian National Creutzfeldt-Jakob Disease Registry is funded by the Commonwealth Department of Health and Aged Care.

Corresponding author: S. Collins, MD, FRACP, National Creutzfeldt-Jakob Disease Registry, Level 5, Department of Pathology, University of Melbourne, Parkville, Victoria, Australia 3052 (e-mail: s.collins@pathology.unimelb.edu.au).