|Home | About | Journals | Submit | Contact Us | Français|
Twin studies assume exceptional relevance when we aim to analyze the brain pathologies associated with diagnosis of AD. It is now clear that multiple brain pathologies frequently co-occur in the same aged individual1, showing different levels of severity and anatomical distribution, and that cannot be linearly correlated with the specific cognitive manifestations observed in a single AD patient.2 Even MZ twins within a pair can show different degrees of clinico-pathologic variability.3–5 If we could assess autopsy-brains from clinically well-characterized MZ and DZ twins with or without dementia, and deceased during the same year, the epidemiological, clinical, pathologic, genetic, and molecular findings that could derive from these assessments would be potentially impressive. Although obtaining both brains from temporally-synchronized MZ and DZ twins, that is twins deceased at the same age and that received a brain autopsy, are events more unique than rare, we had such a chance.
We report cognitive diagnoses, neuropathologic findings, and semi-quantitative immunohistochemistry-based measurements, of a series of co-occurring pathologies found in multiple regions of the brain of two twin pairs, one identical (MZ) and one fraternal (DZ), deceased at the same age. The age at death for the older male MZ pair was 98 years; the age at death for the younger opposite sex DZ pair was 79 years. With respect to ApoE genotype, where the allele ApoE4 is the major genetic risk factor for AD known so far6, the MZ twins were heterozygotes for the ApoE2 allele (ApoE2/3), while DZ twins were both heterozygotes for the ApoE4 allele (ApoE3/4). The MZ twins were then heterozygotes for the rarest ApoE allele, the ApoE2, which is also known to have protective properties against AD.7
The aim of the study was to assess and report semi-quantitative neuropathologic findings of co-occurring brain pathologies in these two temporally-synchronized-at death, MZ and DZ twin pairs.
Both pairs of twins were initially part of the Swedish Twin Registry (STR).8 STR represents one of the oldest and largest twin registries available in the world. For a more detailed description of STR and its derivative studies, refer to previous publications.9
Selection criteria for this investigation were: a) brain autopsy for both twins of the same pair; b) one or both twins of the same pair diagnosed with dementia; c) death of both twins occurred during the same year. Among a total of 140 individual twin autopsies available, only 2 twin pairs met the above criteria.
Clinical diagnoses of dementia followed DSM-III-R10 clinical diagnosis and NINCDS/ADRDA criteria , reflecting the year in which diagnoses were made. AD diagnoses were also reviewed accordingly to NIA-AA clinical criteria.12 Furthermore, psychiatric symptoms were evaluated using the Neuropsychiatric Inventory for dementia (NPI).13
All four brains were stored and examined for standard neuropathologic assessments at the Brain Bank at Karolinska Institutet, Karolinska University Hospital at Huddinge, Stockholm, Sweden. Each autopsy was obtained in accordance with Swedish law. A detailed description of the methods, criteria, and antibodies used to assess these twin brains have been previously described.14 The following brain regions were selected for semi-quantitative assessments of each brain: middle frontal gyrus (MFG), medial temporal gyrus (MTG), anterior cingulate gyrus (ACG), posterior hippocampus (PH; at the level of the corpus geniculate lateral), amygdala (AMY), mesencephalon (MES; including the substantia nigra, SN), and pons (PONS; including the locus coeruleus, LC).
The following pathologic lesions were considered: β-amyloid diffuse-, neuritic-, and cored-plaques (Aβ-DP, Aβ-NP, Aβ-CP); hyperphosphorylated-tau tangles (tau-NFT), tau-threads (tau-th), tau-dystrophic neurites (tau-N), and tau-plaques (tau-P); alpha-synuclein positive Lewy Body (LB), and alpha-synuclein positive dystrophic neurites (LN); ubiquitin positive (ubiq) neuronal cytoplasmatic inclusions (ubiq-NCI), ubiq-neuropil threads (ubiq-th), and ubiq-dystrophic neurites (ubiq-N); protein p62 positive (p62) neuronal cytoplasmatic inclusions (p62-NCI), p62-neuropil threads (p62-th), and p62-dystrophic neurites (p62-N); TDP43 positive (TDP43) neuronal cytoplasmatic inclusions (TDP43-NCI), TDP43-neuropil threads (TDP43-th), and TDP43-dystrophic neurites (TDP43-N); phosphorylated-TDP43 (pTDP43-NCI), pTDP43-neuropil threads (pTDP43-th), and pTDP43-dystrophic neurites (pTDP43-N). As for pathologic criteria, each brain was evaluated using the “practical approach” of NIA-AA pathologic criteria for AD.15
The semi-quantitative system used was defined as: no lesions= negative (N), ≥2 lesions= sparse (S), <2–6> lesions= moderate (M), and >6 lesions= frequent (F). For NCIs, number of lesions were referred to the number of neurons containing at least one cytoplasmatic inclusion independently on its shape (i.e. spheroidal or lentiform). For statistical purposes the frequency-based semi-quantitative system was transformed in a numerical system with N=0, S=1, M=2, F=3.
After transformation to numerical values (0–3) for each brain region and immunostain, all lesions were aggregated per type of pathology:
Neuropil threads (th) and dystrophic neurites (N) TDP43 and pTDP43 pathology were also evaluated separately to determine possible different p-/TDP43 pathology subtypes as previously proposed.16 Cumulative scores (sum of all single numerical scores obtained for each brain region, and for each immunostain) are shown in Figure 1.
Main demographic, zygosity, attained years of education, clinical diagnoses, age of dementia-onset, interval between dementia-onset and death, and neuropsychiatric findings are summarized in Table 1a and 1b. Autopsy-confirmed diagnoses (CERAD criteria)11 were: probable AD (probAD) for TwinA1 and TwinA2, and definite AD (def AD) for TwinB1 and TwinB2. None of the twins presented clinical or autopsy-diagnosis of: PD, DLB, VaD, PDD or FTD. Levels of AD neuropathologic change (NIA-AA pathologic criteria-based) were “intermediate” for all four twins. All twins presented psychiatric symptoms (Table 1b).
For the ApoE2/3 MZ pair, the intra-pair difference between twins’ ages at dementia-onset was 7 years. Both twins showed marked TDP43 and pTDP43 pathology, defined as TDP43-pathology of type D16, with greater pathology in the twin with earlier onset. The twin with later onset (and shorter survival with dementia) showed greater tau pathology in particular, but also a higher β-amyloid cumulative score compared to his twin brother. Other co-occurring pathologies were minimal. For the ApoE3/4 DZ pair, there was marked β-amyloid and tau pathology in both twins, with both cumulative scores higher in the twin with the slightly later onset, as well as substantial co-occurring brain pathologies, ubiq in particular. In contrast, there was very little TDP43 and pTDP43 pathology in either DZ twin. As for the anatomical distribution, the only cerebral regions constantly positive to both TDP43 and pTDP43 pathologies, as for other pathologies (i.e. p62, ubiq), were PH and AMY.
Figure 1 shows time-scales, cognition-deaths intervals, and cumulative scores-histograms, for all pathologies in both twin pairs. Table 2a, b, c (supplementary material) shows single pathologic score for all types of pathology in each cerebral region considered in this study.
To the best of our knowledge, clinico-pathologic correlations with semi-quantitative immunohistochemistry-based neuropathologic analyses on twins deceased at the same age has never been described before.
Our autopsy findings from this unique set of same-age twin brains confirm previous larger studies supporting ApoE4 as an accelerating factor and ApoE2 as protective factor for AD . ApoE2 was associated indeed with dementia, but at a quite advanced age. At the same time, in the MZ twin pair, the interval between dementia onsets suggests that other modifying factors should be taken in account17 to better explain why their dementia onsets and clinical manifestations differed within the twin pair. Dementia in both MZ twins was associated with relatively low levels of amyloid, tau and all other considered co-occurring brain pathologies, except for p/TPD43 pathology, with frequent NCI among diffuse TDP43/pTDP43 positive nuclei in both members of the pair. As novelty, this last observation suggests a possible correlation between TDP-43 pathology and duration of disease. Intriguingly, a possible new modifying effect (protective) of ApoE2 on TDP-43 pathology in very old subjects could be hypothesized based on these novel clinico-pathologic observations. This possible correlation between TDP-43 pathology and duration of disease in very old subjects could acquire actually even higher relevance considering the discrepancy between duration of disease and other “classic” AD pathologies, like tau pathology, for example.
The association of ApoE2 with reduced amounts of various brain pathologies: β-amyloid, tau, LB, ubiq, and p62 pathology is coherent with previous neuropathologic observations on the relationship between ApoE2 vs ApoE4 allele and AD pathology, especially for amyloid.18 The TDP43 and pTDP43 findings may indicate differential positivity at different stages of disease, which could be also related to the ageing per se and not necessarily to a clinical form of FTD. Studies on TDP43/pTDP43 lesions, their anatomical distributions, age-related accumulations, and possible correlations to ageing-related neurodegenerative processes are largely missing.19 Relatively less accumulation of TDP43/pTDP43 pathology was seen in the other twin pair, who were ApoE4 heterozygotes and had a younger age of onset. Based on these rare twin cases, and with the limitation of the sample size, a differential TDP43/pTDP43 nuclear/NCI distribution pattern in AD, and in other neurodegenerative diseases as well, could be hypothesized. This new possible clinico-pathologic correlation (TDP-43 vs. duration of disease) needs to be confirmed by larger clinico-pathologic studies. These studies, however, should take also in account: the possible correlation between TDP-43 pathology accumulation and aging per se, the preferential deposition of TPD-43 pathology in specific brain regions, other co-occurring age-related brain pathologies, as well as possible modifying factors such as ApoE2 and other biological or environmental factors, which remain still unknown.
These clinico-pathologic findings on age-matched-at autopsy twins provide a rare and unusual example of the complexity of AD. Although genetic factors, especially ApoE genotype, may play an important accelerating role in the pathogenesis of dementia, and specifically for AD, other non-genetic factors are necessarily involved to explain the clinicopathologic variability observed among AD cases, even between identical twins with a genetic protective factor as ApoE2.
We cannot assume that these single pairs of MZ and DZ twins are representative of MZ and DZ twins in general, but together, the four cases offered a unique natural experiment and a rare investigative window into the variability of brain pathologies associated with AD.
Part a shows semi-quantitative scores for different forms of β-amyloid and tau lesions accumulation, as well as for vascular pathologies, assessed in multiple regions of the twin brains: middle frontal gyrus (MFG), middle temporal gyrus (MTG), anterior cingulate gyrus (ACG), posterior hippocampus (PH), amygdala (AMY), mesencephalon (MES, including the substantia nigra, SN), and the pons (PONS, including the locus coeruleus, LC). Part b shows Lewy pathology (bodies and neurites), ubiquitin-, and p62-neuronal cytoplasmic inclusions, neuropil threads and dystrophic neurites (ubiq-NCI, -th, -N; p62-NCI, -th, -N). Part c shows TDP/pTDP43 nuclear positivity, TDP-/pTDP43- neuronal cytoplasmic inclusions, neuropil threads, and dystrophic neurites (TDP-NCI, -th, -N; pTDP43-NCI, -th, -N). For all pathologies, cumulative scores are indicated. Cumulative scores are the sum of all partial scores obtained for each type of pathology on each cerebral region as based on their frequency.
The authors thank these twins and their family for their generous gift. The authors also thank the personnel from all Institutions involved in this study for their dedication to this project. We acknowledge funding from NIH Grants No. R01 AG08724, AG04563, AG10175, and AG08861. This study was supported by the Swedish Brain Power and the StratNeuro programme at Karolinska Institutet.