A nonsense mutation in PRNP associated with clinical Alzheimer's disease

Here, we describe a nonsense haplotype in PRNP associated with clinical Alzheimer's disease. The patient presented an early-onset of cognitive decline with memory loss as the primary cognitive problem. Whole-exome sequencing revealed a nonsense mutation in PRNP (NM_000311, c.C478T; p.Q160*; rs80356711) associated with homozygosity for the V allele at position 129 of the protein, further highlighting how very similar genotypes in PRNP result in strikingly different phenotypes.

The application of this technology to the study of small families and individual cases with different forms of dementia has also resulted in the association of unexpected molecular causes to different clinical phenotypes (for a review see, Guerreiro et al., 2014). For example, TREM2 homozygous mutations, known to be the cause of Nasu-Hakola disease, were recently found to also cause frontotemporal dementia with no associated bone phenotypes (Guerreiro et al., 2013c); homozygous mutations in ATP13A2 (a gene known to cause Kufor-Rakeb) and GRN (where heterozygous mutations cause frontotemporal dementia) were identified in families with neuronal ceroid-lipofuscinosis (Bras et al., 2012a;Smith et al., 2012). Exome sequencing has not only allowed the identification of the genetic causes of disease in cases that otherwise would have never been screened for mutations in the implicated genes because of their atypical phenotypes, but has also uncovered common biological pathways between different clinical entities (Bras et al., 2012b).
Here, we describe one more of these cases: a patient clinically diagnosed with AD found by exome sequencing to harbor a nonsense mutation in the PRNP gene.

Methods
When genetic tests for APP, PSEN1, and PSEN2 revealed no mutations, the patient was included in a whole-exome sequencing study. Genomic DNA was prepared according to Illumina's TruSeq Sample Preparation v3 (Illumina, CA, USA) and capture was performed with Illumina's TruSeq Exome Enrichment according to the manufacturer's instructions. Sequencing was performed in Illumina's HiSeq2000 using 100 bp paired-end reads. Sequence alignment and variant calling were performed against the reference human genome (UCSC hg19) using bwa (Li and Durbin, 2009) and reads processed according with the Genome Analysis Toolkit best practices (McKenna et al., 2010). Variants were called using UnifiedGenotyper and recalibrated using VQSR, both tools from the GATK. Finally, variants were annotated using snpEff (Cingolani et al., 2012). The PRNP mutation was confirmed by Sanger sequencing using standard methodology.

Results
The analysis of the exome sequencing data confirmed the absence of pathogenic mutations in APP, PSEN1, and PSEN2. Additionally, no coding variants were found in the dementia associated genes APP, PSEN2, GRN, TREM2, or PLD3. The patient was found to carry the PSEN1 (NM_000021) p.E318G and the MAPT (NM_001123066) p.Q230R variants.
Further inspection of the 9423 coding variants found (445 of which were novel), revealed a nonsense mutation in PRNP (NM_000311, c.C478T; p.Q160*; rs80356711) associated with homozygosity for the V allele at position 129 of the protein.
The patient was followed in the Mayo Clinic and presented an early-onset of cognitive decline at 38 years with memory loss as the primary cognitive problem, but also showing an impulsive behavior on her neuropsychological assessment. Her mother had a similar problem, also of early onset (no DNA was available for testing). Her maternal grandparents lived long and were said not to be affected. Her brother and daughter were also unaffected at the time of evaluation. She had temporary diarrhea, which was thought to be related to the introduction of Aricept, and her positron emission tomography scan showed left frontal hypometabolism. The patient was diagnosed with clinical AD and no neuropathologic assessment was possible.

Discussion
The mutation here described (p.Q160*) has been previously reported in 2 other cases (Table 1) diagnosed with an Alzheimer-like dementia. The first case did not have a detailed clinical description and no pathologic findings were reported (Finckh et al., 2000). The second case was deeply phenotyped, and neuropathologic evaluation showed abundant limbic and neocortical neuritic plaque-like structures and neurofibrillary tangles consistent with a neuropathologic diagnosis of AD. Immunohistochemical studies also demonstrated PrP immunopositive deposits (Jayadev et al., 2011).
In the literature, 6 different mutations in PRNP leading to a premature truncation of the protein can be found (Table 1). None of these cases was initially diagnosed with a prion disease. In fact, the proband's mother in the report by Jayadev et al. (2011) was also neuropathologically diagnosed as AD before immunochemical studies were performed.
Recently, Mead et al. (2013) described an unusual phenotype associated with a novel nonsense mutation in PRNP. The affected members of this family carried the p.129V-163* PRNP truncation haplotype and developed autonomic failure with chronic diarrhea and peripheral polyneuropathy in adulthood.
The different truncating mutations in PRNP appear to have some common features namely: prolonged clinical courses, atypical for prion diseases, severe neurofibrillary tangle pathology, and high levels of cerebral amyloidosis. However, it is remarkable that the simple removal of an extra 3 amino acids on the same haplotype (V129 background), consistently results in a very different phenotype: truncated PRNP at amino acids 160 or 163 present with a clear hippocampal involvement or an autonomic defect, respectively.

Disclosure statement
The authors declare no competing financial or personal interests that can influence the presented work.