Hippocampal sclerosis of aging, a prevalent and high-morbidity brain disease

Abstract

Hippocampal sclerosis of aging (HS-Aging) is a causative factor in a large proportion of elderly dementia cases. The current definition of HS-Aging rests on pathologic criteria: neuronal loss and gliosis in the hippocampal formation that is out of proportion to AD-type pathology. HS-Aging is also strongly associated with TDP-43 pathology. HS-Aging pathology appears to be most prevalent in the oldest-old: autopsy series indicate that 5-30 % of nonagenarians have HS-Aging pathology. Among prior studies, differences in study design have contributed to the study-to-study variability in reported disease prevalence. The presence of HS-Aging pathology correlates with significant cognitive impairment which is often misdiagnosed as AD clinically. The antemortem diagnosis is further confounded by other diseases linked to hippocampal atrophy including frontotemporal lobar degeneration and cerebrovascular pathologies. Recent advances characterizing the neurocognitive profile of HS-Aging patients have begun to provide clues that may help identify living individuals with HS-Aging pathology. Structural brain imaging studies of research subjects followed to autopsy reveal hippocampal atrophy that is substantially greater in people with eventual HS-Aging pathology, compared to those with AD pathology alone. Data are presented from individuals who were followed with neurocognitive and neuroradiologic measurements, followed by neuropathologic evaluation at the University of Kentucky. Finally, we discuss factors that are hypothesized to cause or modify the disease. We conclude that the published literature on HS-Aging provides strong evidence of an important and under-appreciated brain disease of aging. Unfortunately, there is no therapy or preventive strategy currently available.

Fig. 1

Histopathology of HS-Aging: hematoxylin and eosin (H&E) findings. a Low-power photomicrograph of hippocampal formation of a woman who died at 88 years of age with dementia and HS-Aging pathology. Even at low power one can appreciate that the hippocampus has atrophy and areas of neuropil rarefaction (blue arrows) in dentate granule area, CA1, and subiculum. b For comparison sake, the brain of a man who died at age 77 with dementia and end-stage (Braak VI) AD, with dentate granules (DG), CA1, and subiculum labeled. Note that even in AD the hippocampus (at the same scale as in a) is somewhat larger and without the neuropil rarefaction. c In individuals with neither AD nor HS-Aging, such as this 71-year-old cognitively intact male, with Braak stage I pathology, the hippocampal neuropil appears homogenously pink and nondisrupted on an H&E stain. d Other features of HS-Aging are shown in this medium-power photomicrograph of the boxed area from a. Even at this magnification, the disruption of the normal hippocampal architecture can be observed, along with thickened medium-sized blood vessel (green arrow). e At higher magnification, from the hippocampus of a woman who died with dementia at age 91 with HS-Aging pathology, this blood vessel profile (green arrow) shows the arteriolosclerosis and thickened multilumen blood vessel profiles that can accompany HS-Aging pathology. Scale bars a–c =1 mm, d =500 μm, e =150 μm

Fig. 2

Histopathology of HS-Aging: Phospho-TDP-43 and phospho-tau immunohistochemical findings. Observations in brain sections from the same case are as shown in composite in Fig. 3. These sections (dentate granule cells, a, b; CA1, c, d) have been stained with hematoxylin (stain nuclei and some cell contour blue) and counter-stained with brown chromagen. Sections a and c are stained for phospho-TDP-43 (clone 1D3, Millipore). Note that in the dentate granule cells, there are immunoreactive inclusions in cell bodies (red arrows), whereas there are prominent neuritic (narrow non-tapering nerve cell processes) TDP-43+ staining in the CA1 area, in addition to some cellular staining (intranuclear inclusion, green arrow). Phospho-tau staining (PHF-1, a gift from Dr Peter Davies) has features that do not map well onto Braak staging. For example, there are relatively numerous phospho-tau-positive dentate granule inclusions, whereas CA1 shows very little phospho-tau immunoreactivity. Scale bars a, b =30 μm; c, d =50 μm

Fig. 3

Histopathology of HS-Aging: Composite low-power figure depicts the distribution of pathology localized with multiple pathological immunomarkers. Sections were analyzed from the brain of a man who died cognitively impaired at age 92 years; autopsy showed HS-Aging and Braak stage II pathologies. An Aperio ScanScope XT with Genie™ image recognition software was used to highlight the positive immunoreactivity. The top portion shows the composite results of three nearly consecutive sections stained for phospho-tau (blue), phospho-TDP-43 (red); and GFAP (yellow). Labeled are dentate granule cell layer (DG, shown in green in top portion), CA1, and subiculum (bottom). Green arrows show same abnormally enlarged Virchow-Robin space on both top and bottom figures. This representative case shows that HS-Aging brains have a multifaceted pathological picture that includes TDP-43 pathology, astrocytosis, tauopathy, and vascular profiles that are aberrant in comparison to that which would be observed in younger control individuals. Future work is required to identify the truly specific, and clinical disease-driving, feature(s) of HS-Aging pathology

Fig. 4

Human subiculum affected by HS-Aging pathology contains abundant detergent-insoluble glial fibrillary acidic protein (GFAP). For this experiment, tissue was isolated from subiculum of six different cases, three HS-Aging and three controls. The tissue was processed as previously described [43, 82] to isolate detergent-insoluble protein using a method that ultimately solubilizes proteins with 7 M urea. Coomassie Blue stained urea-polyacrylamide gel from the detergent-insoluble extract shows a ~50 kDa band that is accentuated in HS-Aging cases (three leftward lanes on the gel). Individual 50 kDa gel bands were excised for each of the six cases and the gel fragments were submitted separately for liquid chromatography–electrospray ionization mass spectrometric (LC-ESI-MS) proteomic analysis. Gel pieces were digested with trypsin, and LC-ESI-MS performed using a ThermoFinnigan LTQ. Resulting MS spectra were searched against human proteins in the Swiss-Prot database using the Mascot search engine (Matrix Science). In both the HS-Aging and control cases, the overwhelming proportion of this 50 kDa band was GFAP. Shown at the bottom right of the figure are the two proteins in this size range with the most peptide query matched reads, for each of the six samples. With caveats appropriate for comparison between two experimental groups comprising only three samples each, the cases with HS-Aging pathology had larger amount of GFAP peptide fragments, covering almost the entire span of the protein, than the controls (P< 0.03)

Fig. 5

Data related to HS-Aging epidemiology underscore the large, and increasing, public health impact of HS-Aging. Data from The Nun Study [73, 110] among research subjects with pathologic data (N = 526). The proportion of individuals with moderate or severe Alzheimer’s disease (AD; moderate or severe neuritic amyloid plaque pathology and Braak stages using two different threshold cutoffs, Braak IV and above and Braak V and above) are compared to the proportion of individuals with HS-Aging pathology. Note that a significant number of patients had both pathologies as would be expected. This is a birth cohort that had been followed for many years, incorporating a full spectrum of cognitive impairment, without many of the biases that are linked to dementia clinics, thus insights into the population-level epidemiology. Median age of this cohort is >90 years of age at death. b The late-life increase in HS-Aging pathology can be viewed in context of projected demographic increases in numbers of very old persons predicted by the U.S. Census Bureau. Source: http://www.census.gov/population/projections/data/national/2012/su mmarytables.html

Fig. 6

Neurocognitive changes provide a clinical feature that distinguishes cases with eventual HS-Aging pathology versus controls. Each data point represents an individual research volunteer. N = 43 HS-Aging cases, and N = 75 controls, matched for age, gender, education level, and APOE status with each of those parameters used as covariate. These 118 participants had a total of 966 yearly longitudinal assessments for an average of 8.2 assessments per participant. All were followed from nondemented cognitive state at baseline. Plots show the distribution of values for the ratios of test scores of word list delay (WLD): verbal fluency (VF) at baseline and at an examination 5.5–6.5 years prior to the patients’ death. The timepoint of ~6 years prior to death was selected because this usually was after symptom onset but before end-stage disease. All statistical analyses were performed using SAS/STAT^® 9.2 software. This figure is adapted from PT Nelson et al. [76] Brain, published by Oxford University Press, and is reproduced with permission

Fig. 7

Magnetic resonance images (MRIs) from individuals with eventual autopsy diagnosis of HS-Aging. This group of coronal MRIs from four individuals illustrates that HS-Aging pathology is often associated with co-morbid brain diseases. a A 96-year-old female patient with an acute stroke shortly before autopsy. Shown on the T1-weighted image are signs of acute cortical swelling in the medial and inferior right temporal lobe due to a right posterior cerebral artery stroke (arrowheads), subcortical white matter hypointensity from prior ischemia (arrow), and marked left hippocampal atrophy (double arrow). Extensive vascular disease, hippocampal sclerosis, and Alzheimer’s disease (Braak stage V) were found at autopsy. b T2-weighted image from 80-year-old female patient demonstrates atrophy and abnormally increased signal in the hippocampi (arrowheads). Autopsy 16 years later demonstrated AD pathology, Braak stage VI, and HS. c HS with Braak stage III pathology and stroke were found at autopsy 7 years after this scan demonstrates asymmetric right medial temporal atrophy (double arrows) and dilation of the right frontal horn (arrow), d T1-weighted scan from 86-year-old woman with slowly progressive memory loss and stroke demonstrates asymmetric left medial temporal atrophy (double arrows). Autopsy confirmed HS-Aging pathology