|Home | About | Journals | Submit | Contact Us | Français|
Mutations in presenilin 2 are rare causes of early onset familial Alzheimer’s disease. Eighteen presenilin 2 mutations have been reported, although not all have been confirmed pathogenic. Much remains to be learned about the range of phenotypes associated with these mutations. We have analysed our unique collection of 146 affected cases in 11 Volga German families, 101 who are likely to have the same N141I mutation in presenilin 2 (54 genotyped confirmed). We have also assessed the detailed neuropathologic findings in 18 autopsies from these families and reviewed the world’s literature on other presenilin 2 mutations; presenting a novel mutation that is predicted to lead to a premature truncation codon. Seven presenilin 2 mutations reported in the literature have strong evidence for pathogenicity whereas others may be benign polymorphisms. One hundred and one affected persons, with sufficient historical information from the Volga German pedigrees (N141I mutation), had a mean onset age of 53.7 years ± 7.8 (range 39–75) and mean age at death of 64.2 years ± 9.8 (range 43–88). These figures overlap with and generally fall between the results from the subjects in our centre who have late onset familial Alzheimer’s disease or mutations in presenilin 1. Seizures were noted in 20 (30%) of 64 subjects with detailed medical records. Two mutation carriers lived beyond age 80 without developing dementia, representing uncommon examples of decreased penetrance. Two persons had severe amyloid angiopathy and haemorrhagic stroke. Eighteen cases had detailed histopathology available and analysed at our institution. Braak stage was five or six, amyloid angiopathy and neuritic plaques were common and more than 75% had Lewy bodies in the amygdala. TAR DNA-binding protein-43 inclusions were uncommon. In addition, a 58-year-old female with a 2 year course of cognitive decline and no family history of dementia has abnormal fludeoxyglucose-positron emission tomography imaging and a novel 2 base pair deletion in presenilin 2 at nucleotide 342/343, predicted to produce a frame-shift and premature termination. We conclude that mutations in presenilin 2 are rare with only seven being well documented in the literature. The best studied N141I mutation produces an Alzheimer’s disease phenotype with a wide range of onset ages overlapping both early and late onset Alzheimer’s disease, often associated with seizures, high penetrance and typical Alzheimer’s disease neuropathology. A novel premature termination mutation supports loss of function or haploinsufficiency as pathogenic mechanisms in presenilin 2 associated Alzheimer’s disease.
Four genes have been confirmed to play an important role in the pathogenesis of various forms of Alzheimer’s disease. Apolipoprotein E alters the risk for, and the age at onset of, the common late onset type of Alzheimer’s disease. Mutations in the amyloid precursor protein gene on chromosome 21 were first reported in 1991 and more than 30 pathogenic mutations have been described (Goate et al., 1991). Mutations in the presenilin 1 gene (PSEN1) on chromosome 14 were first reported in 1995 and more than 170 mutations have been described, making this the most common cause of autosomal dominant early onset Alzheimer’s disease (Sherrington et al., 1995; Larner and Doran, 2006). Mutations in presenilin 2 (PSEN2) on chromosome 1 were first described in 1995 and only 18 potentially pathogenic mutations have subsequently been reported, making this the least common genetic cause of Alzheimer’s disease (Levy-Lahad et al., 1995; Rogaev et al., 1995). The discovery of mutations in these genes has been fundamental in advancing our understanding of Alzheimer’s disease. However, because mutations in PSEN2 are rare there has never been an extensive evaluation of them or the phenotype associated with them.
The purpose of the present study is to fill this gap in our knowledge by (i) reviewing the world’s literature and assessing the likelihood of pathogenicity associated with each reported PSEN2 mutation; (ii) reporting a new frame-shift truncating mutation in PSEN2; and (iii) describing the clinical and pathological phenotype of the largest collection of subjects with a PSEN2 mutation (N141I) in eleven families with the Volga German genetic founder background.
We performed a PubMed search of all articles relating to mutations in PSEN2. These reports were analysed for number of families, number of affected generations, number of affected subjects, segregation of mutation with disease and review of the functional studies of several mutations reported by Walker et al. (2005).
Over the past 25 years our Alzheimer’s disease Research Centre has evaluated 184 families with possible Alzheimer’s disease having a ‘German from Russia’ ethnic background. Forty-three were Black Sea Germans from the Odessa region and 124 were Volga Germans from the Saratov region. Since the discovery of the PSEN2 gene in 1995 we have determined that 11 of these families represent a genetic founder effect and carry the same N141I mutation in PSEN2 (Levy-Lahad et al., 1995). From personal examination or medical record review we have collective detailed clinical information on 101 affected persons in these 11 families. From these evaluations we determined age of onset, age of death, likely first symptom, presence of psychotic features (delusions and/or hallucinations) and occurrence of seizures. Age at onset was defined as in our previous studies (Bird et al., 1988). For comparison we have also used the onset age, death age and duration data collected in our centre from 23 families with PSEN1 mutations and 125 families each having three or more persons with late onset Alzheimer’s disease.
Twelve affected subjects had formal neuropsychological testing by a variety of examiners that usually included four or more of the following tests: Wechsler Memory Scale–Revised, Wechsler Adult Intelligence Scale, Reitan Test Battery, Boston Naming, Trail Making Test, Rey Figure Drawing, Wisconsin Card Sorting, Clock Drawing, Blessed Dementia Rating Scale, California Verbal Learning Test, Category Test, Controlled Oral Word Association, Stroop and Mini Mental State Examination.
Thirty-five brain autopsies were performed in these Volga German families. Twelve were performed at outside institutions and only reports were available. Slides from 23 autopsies were available for review at our institution and 18 had sufficient material for a more detailed neuropathological examination. The additional examination included immunohistochemical assessment for tau (AT8, Endogen, 1:250), alpha-synuclein (LB509, 1:1:1000 and syn 303, 1:500; generous gifts of J. Q. Trojanowski) and TAR DNA binding proteion (TDP43; Protein Tech, 1:2000) pathology as previously described (Leverenz et al., 2008). For each case, we determined the Consortium to Establish a Registry for Alzheimer Disease (CERAD) neuritic plaque scores and Braak staging for neurofibrillary tangles (Braak and Braak, 1991; Mira et al., 1991).
Pittsburgh compound-B PET brain imaging was performed on one subject using previously described methods (Lopresti et al., 2005).
Twenty-three DNA variants resulting in non-synonymous substitutions have been reported in the PSEN2 gene (http://www.molgen.ua.ac.be/ADMutations). Four of these are known normal polymorphisms. Three variants, R29H, L143H and A252T were found in a screening of 130 individuals from seven African populations belonging to the Centre d’Etude du Polymorphisme Humain-Human Genome Diversity Panel (Guerreiro et al., 2008). A P334R variant has been reported in a late onset Alzheimer’s disease family, though the change did not segregate with disease (Lleó et al., 2002a). Recently a R163H mutation was found in a Parkinson’s disease family carrying a de novo alpha-synuclein A53T mutation (Puschmann et al., 2009). The patient carrying the gene change had cognitive deficits although her unaffected mother also carried the mutation. Eighteen have been proposed as possible disease causing mutations (Table 1 and Fig. 1). Our analysis indicates that seven of these mutations are highly likely to be pathogenic. These seven have occurred in more than one family and/or have been shown to segregate with disease in a family. In addition, Walker et al. (2005) found that five of these mutations produce an excessive amount of Aβ peptide in a cell culture system. Eleven other DNA variants are less likely to be pathogenic, although this remains a possibility. These 11 variants have occurred only in a single sporadic case, have not been shown to segregate with disease in a family and four of them showed no abnormality in the cell culture experiments of Walker et al. (2005). Also, in two families having multiple affected members with late onset AD we were able to show that the R171W “mutation” was not segregating with the disease and is unlikely to be pathogenic.
The seven mutations that are likely to be pathogenic have occurred only in a single family in five instances, in two families with the T122P mutation and in 12 families with the N141I mutation. These seven mutations have all been associated with early onset Alzheimer’s disease with a range of 39–75 years. The A85V mutation in one family was associated with somewhat later onset (60–73 years), but overlaps with the other six mutations. The S175C mutation in a newly reported family is also likely to be pathogenic (Piscopo et al., in press).
Figure 1 shows the location of the seven pathogenic mutations in the PSEN2 protein.
A 59-year-old female has had a three year course of slowly progressive memory loss. Her family history was negative for dementia, but she knew relatively little about her extended family. Her mother died of cancer at the age of 45 and she knows nothing about her biological father. Formal neuropsychological testing revealed significant impairments in visual spatial skills and executive function. A fludeoxyglucose PET scan showed decreased uptake in the parietal and temporal lobes bilaterally. She was taking both donepezil and memantine. DNA sequencing of the amyloid precursor protein and PSEN1 genes was normal. However, sequencing of the PSEN2 gene (Athena Diagnostics, Inc; Worcester, MA) revealed a novel two base pair deletion of GA from nucleotide positions 342 and 343 in exon 5 leading to a frame-shift and premature termination codon in exon 6 (K115E fx10; Fig. 1).
The first mutation reported in PSEN2 was N141I (Levy-Lahad et al., 1995). This mutation was identified in a collection of familial Alzheimer’s disease pedigrees studied in our centre since 1983 and having the common ethnic background of being Germans from Russia. We have evaluated more than 180 such families from a large region in Russia and the former USSR including Black Sea Germans and Volga Germans. The N141I mutation was found in 11 Volga German families all originating from three small adjacent villages southwest of Saratov, Russia (Bird et al., 1988).
The clinical phenotype associated with the N141I mutation is summarized in Table 2. We have historical information on 101 affected persons in the 11 families. The number of affected people per family varies from 2 (BE family) to 17, 20 and 26 in the HD, HB and R families, respectively. Fifty-four of these affected persons have had the N141I mutation confirmed by DNA genotyping. The mean age of onset is 53.7 years with a wide range of 39 to 75 years. Mean age at death is 64.2 years with a mean disease duration of 10.6 years. Detailed medical records were available for review in 64 affected persons. The first symptom was almost always a problem with memory, thinking or orientation. The disease was slowly, relentlessly progressive and usually ended in a mute, rigid, bed ridden state. Pyramidal signs and parkinsonian features were not seen early in the disease.
We have compared this PSEN2 phenotype with the other families ascertained in our centre with PSEN1 mutations and those having late onset familial Alzheimer’s disease (Table 3). The PSEN1 families had an early mean age of onset of 45.5 years, but with a wide range (28–79 years). Other reported mutations in PSEN1 have also been associated with a wide range in age of onset. The latest recorded ages of onset in PSEN1 were 84 and 88 years in a family with the Met239Val mutation (Sherrington et al., 1995). The late onset Alzheimer’s disease families from our centre had a mean age of onset of 71.5 years, and a very wide range (45–91 years). The Volga German PSEN2 families have a mean onset in between these two groups (mean 53.7, range 39–75 years). Note that only three (1.4%) of the PSEN1 cases had an onset after age 60 and these three cases all came from a single family with the A79V mutation known to be associated with a later age of onset (Brickell et al., 2007). The Volga German PSEN2 families had 16 (16%) subjects with onset ≥60 years. Disease duration was shorter in the PSEN1 cases (mean 8.4 years) whereas the duration was similar between the Volga German PSEN2 cases (10.6 years) and the of late onset Alzheimer’s disease (10.2 years).
The largest known family with a mutation in PSEN2 is shown in Fig. 2. This is the ‘R’ family with 26 affected persons, an early mean onset age (49.5 years), high penetrance and a known phenocopy (see below).
Of the 64 affected persons with detailed medical records, 31% (20/64) had mention of one or more seizures. The seizures were recurrent and frequent in two subjects. Thirty-three percent (21/64) had mention of hallucinations, delusions or ‘psychotic’ features. Twenty-three persons were treated with a cholinesterase inhibitor and/or memantine. No systematic study of treatment effects was possible. There were only anecdotal reports of both beneficial and negative treatment results, but more than half of these families (13/23) reported some possible ‘benefit’.
Formal neuropsychological testing was performed on 12 affected persons, usually 1–4 years after onset of symptoms (Supplementary Table). All individuals demonstrated memory loss with relatively spared motor function. Verbal fluency, spatial perception, calculations and executive functions were also impaired, especially as measured by the Wechsler Memory Scale–Revised and Trail Making Tests. Subjects often did well on the Boston Naming Test early in the disease.
One person in the R family (V-34) had Pittsburgh compound B imaging obtained at age 52, after about 2 years of mild symptoms. At the point this subject was imaged, Pittsburgh compound B retention was very typical of that previously reported in patients with late onset Alzheimer’s disease (Klunk et al., 2004), with frontal, temporal and parietal/precuneus cortices and the caudate displaying especially prominent retention (Fig. 3). Since this subject was already symptomatic when scanned, it is not possible to know whether some or all pre-symptomatic PSEN2 mutation carriers would show the striatal-predomiant pattern of Pittsburgh compound B retention reported for pre-symptomatic PSEN1 mutation carriers (Klunk et al., 2007). However, in the symptomatic V-34 case, there was no indication of striatal predominance.
Clinical penetrance of the mutation was high (>95%). Two cases of decreased penetrance by the ninth decade were noted. One subject in the HB family died at age 80 of cancer without any cognitive difficulties noticed by the family. This person had an affected mother, two affected children and he carried the N141I mutation. Another individual in the H family died at age 89 of cancer and also had no cognitive problems according to the family. This person had an affected father and three affected children with the N141I mutation. He himself had not been genotyped.
There have been three phenocopies of the common form of Alzheimer’s disease in these families. One was in the R family (IV-2, Fig. 2). This individual was at risk for the disease and had the onset of a progressive dementia at age 62 years. This onset was later than the other affected persons in the R family (although within the range of the total Volga German family collection), he did not carry the N141I mutation and had an Apolipoprotein E genotype of ε4/ε4. Another phenocopy occurred in the W family in which an 83-year-old male had onset of mild dementia at age 79, did not have the N141I mutation (ApoE 3/3) and neuropathology showed mild to moderate signs of Alzheimer’s disease. A third phenocopy was a female in the HD family with onset of progressive dementia at age 75, death at age 83 and she did not carry the PSEN2 mutation (ApoE 3/4).
Nikisch et al. (2008) reported a small family with early onset Alzheimer’s disease in modern Germany with the N141I mutation. Together we have demonstrated that this family shares the same chromosome 1q haplotype with the Volga German pedigrees and speculated that Alzheimer's original patient (Auguste D.) may have had this mutation (Yu et al., in press).
Eighteen brain autopsies had sufficient material available for a more detailed histopathological evaluation (Table 4). In 17 of these 18 cases, the Braak stage was five or six and the CERAD plaque score was C, fulfilling pathologic criteria for Alzheimer’s disease (Braak and Braak, 1991; Mann et al., 1997; Mira et al., 1991). Alpha-synuclein staining inclusions were common in the amygdala (14/18, 78%) but much less common in the substantia nigra (7/16, 44%) and the neocortex (3/17, 18%). Only three cases (3/13, 23%) had TDP43 inclusions or neurites in the amygdala, and none had TDP43 pathology in dentate gyrus, parahippocampal gyrus or frontal cortex. Three cases (3/13, 23%) had evidence of hippocampal sclerosis.
One person in the HB family did not meet all criteria for Alzheimer’s disease, having a Braak stage of only three. This individual has been reported separately (Nochlin et al., 1998). She had onset of a progressive dementing process similar to other family members but died due a haemorrhagic stroke; she showed severe amyloid angiopathy at autopsy. Another affected member of the H family is living at age 56 with dementia and a haemorrhagic stroke. These two cases emphasize that the amyloid angiopathy associated with this mutation can sometimes be severe.
Of the 18 autopsied cases, seven had a clinical description that included delusions, hallucinations or psychotic features. All seven had Lewy bodies in the amygdala and three had Lewy bodies in the frontal cortex (the only cases with this finding). Of the 18 autopsied cases, six had clinical evidence of seizures (two with frequent seizures). Only one of these cases had hippocampal sclerosis and it was neither of the two that had reports of recurrent seizures.
Three other families reported in the literature with a PSEN2 mutation have had neuropathological evaluation. All had typical neuritic plaques and neurofibrillary tangles associated with classic Alzheimer’s disease. A case with the M239I mutation showed that the neuritic plaques were most frequent in the amygdala (Finckh et al., 2000). A patient with the A85V mutation had unusually abundant and widespread cortical Lewy bodies (Piscopo et al., 2008). Siblings with the M239V mutation had Alzheimer’s disease pathology and ectopic neurons in the subcortical white matter often containing neurofibrillary tangles (Marcon et al., 2004). One of these patients presented with epilepsy.
Of the three genes known to cause familial early onset Alzheimer’s disease, mutations in the PSEN2 gene are the least common. Only seven of these mutations are well documented to be pathogenic, although pathogenicity in several others remains possible. For this reason, relatively little has been reported about the phenotype associated with mutations in PSEN2. This study attempts to provide this knowledge. By a coincidence of genetics and history the largest body of information regarding the PSEN2 phenotype comes from the Volga German families, which represent a founder effect for the N141I mutation.
It is unclear why mutations in PSEN2 are rarely identified. One possibility is that they are less likely to be recognized because of the relatively later age of onset in many cases. It is likely that with the advent of more readily available and less expensive genetic testing, the number of families identified with mutations in PSEN2 will increase. This is especially likely to be the case when more patients with onset of Alzheimer’s disease in the sixth and seventh decades are studied. Furthermore, defects in PSEN2 function may be covered by the normal function of its close homolog PSEN1. This phenomenon has been well illustrated in transgenic mice experiments. Mice with a knockout of PSEN2 are viable, but knockout of both PSEN2 and PSEN1 is lethal (Herreman et al., 1999). However, despite the overlap between presenilin 1 and presenilin 2, in vitro studies support unique roles of presenilin 2 in mammalian physiology. The two presenilins are under different transcriptional control during development and presenilin 2 has been shown to mediate cell signalling pathways not shared with presenilin 1 (Lee et al., 1996; Kang et al., 2006).
The novel mutation in PSEN2 (K115Efx10) reported in this study is important because, unlike all other reported PSEN2 mutations, it results in a frame-shift and a premature termination codon. A splice-site mutation leading to a frame-shift and premature termination codon has been reported in PSEN1 and hypothesized by the author to lead to haploinsufficiency (Tysoe et al., 1998). What remains to be determined is whether this shortened transcript is transcribed to yield a markedly truncated protein or, more likely, if it is instead subject to nonsense mediated decay (Holbrook et al., 2004). If indeed this mutation leads to a degraded transcript and haploinsufficiency, then such mutations in PSEN2 and PSEN1 could cause disease by a loss of function, similar to what has been described for progranulin gene mutations in patients with frontotemporal dementia (Baker et al., 2006; Yagi et al., 2008). PSEN1 and PSEN2 are part of the amyloid precursor protein gamma secretase complex and there is evidence that mutations in these proteins decrease the production of Aβ40 relative to Aβ42 and result in a greater proportion of the more toxic Aβ42 (De Strooper et al., 2007; Fluhrer et al., 2008).
The age of onset associated with mutations in PSEN2 is typically in the sixth decade (50s) but with a wide range from 39 to 75 years. This range overlaps with the generally earlier onset associated with PSEN1 mutations and the later onset seen in the more common form of Alzheimer’s disease. The reasons underlying the somewhat later mean age of onset and broader range in onset ages compared with mutations in PSEN1 remain unclear. Again, this could reflect partial replacement of PSEN2 function by normal PSEN1. Also, APOE genotype has been shown to influence age of onset in both PSEN1 and PSEN2 (Pastor et al., 2003; Wijsman et al., 2005). Marchani and colleagues (in press) have identified three additional genetic loci modifying age at onset in the Volga German PSEN2 families (1q23, 7q33 and 17p13). However, once the disease process has begun there is great overlap in duration of the clinical phenotype between the two genes, although the earliest onset PSEN1 cases tend to have a more rapid decline. Early and progressive defects in memory and executive functions are common with relative sparing of naming function in persons with PSEN2 mutations.
The occurrence of seizures in 31% of the Volga German cases is surprisingly high. The frequency of seizures occurring in common late onset Alzheimer’s disease is difficult to determine but has been reported to be 1.5–20% (Scarmeas et al., 2009). It is acknowledged that these numbers are imprecise because patients with Alzheimer’s disease are rarely observed carefully in the end stages of the disease (Risse et al., 1990). A higher frequency of seizures has also been reported with mutations in PSEN1, and certain mutations in that gene appear to have an especially high frequency of seizures (Larner and Doran, 2006; Palop and Mucke, 2009). Transgenic mice expressing the N141I mutation have been shown to have altered seizure thresholds in response to kainate, although results have varied, probably related to technical differences (Schneider et al., 2001; Schulte et al., 2009). The exact cause of the relatively frequent seizures is uncertain, but could relate to Aβ amyloid toxicity, cytoskeletal dysfunction associated with abnormal tau, cerebral vascular pathology, neurotransmitter dysfunction or a combination of these factors (Cook et al., 2005; Larner, 2009; Yand et al., 2004). We did not observe an association of hippocampal sclerosis with seizures in our cases.
Penetrance is very high (>95%) with mutations in PSEN2 just as with mutations in PSEN1. Occasional phenocopies are expected because of the common occurrence of the more general form of late onset Alzheimer’s disease. These factors play a role in the genetic counselling and genetic testing of these families. The relatively small experience with genetic testing in early onset Alzheimer’s disease families has been similar to that with Huntington’s disease (Steinbart et al., 2001).
Overall, neuropathological changes associated with PSEN2 mutations are those of typical Alzheimer’s disease with extensive neuritic plaque formation and neurofibrillary tangle accumulation. Of note is that amyloid angiopathy may be prominent and even produce haemorrhagic stroke (Nochlin et al., 1998). Consistent with previous reports in both sporadic and familial Alzheimer’s disease we also found very frequent alpha synuclein pathology in the amygdala (Leverenz et al., 2006, 2008; Tsuang et al., 2006). Neocortical alpha synuclein pathology was uncommon, but appeared to be associated with hallucinations and delusions. TDP43 pathologic change was uncommon in our cases, as has been reported in sporadic Alzheimer’s disease (Amador-Ortiz et al., 2007). Because of recent reports of Pittsburgh compound B binding in the striatum in cases with PSEN1 mutations we are in the process of evaluating this region in the PSEN2 subjects (Klunk et al., 2007).
This study has several weaknesses. Many of the family members were evaluated at different times, by different examiners, in different locations and using different techniques. Thus, records were often incomplete or mismatched. Detailed, systematic studies of clinical and neuropathological phenotypes could only be done on a representative selection of cases. Nevertheless, this remains the largest and most detailed study of families with mutations in PSEN2 and forms a solid basis for future investigations of this important type of Alzheimer’s disease.
Department of Veterans Affairs, National Institute on Aging/NIH (Grants Nos AG005136-24, P50 AG005133, R37 AG025516 and P01 AG025204).
Conflict of interest: GE Healthcare holds a license agreement with the University of Pittsburgh based on the PiB technology described in this manuscript. Dr Klunk is a co-inventor of PiB and, as such, has a financial interest in this license agreement. GE Healthcare provided no grant support for this study and had no role in the design or interpretation of results or preparation of this manuscript. All other authors have no conflicts of interest with this work and had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Dr Bird is on the Speakers Bureau of and receives licensing fees from Athena Diagnostics, Inc.
Supplementary material is available at Brain online.
The VQT family was ascertained by Dr J. Cummings. S. Watson assisted with analysis of the neuropsychological testing.