Automatic section thickness determination using an absolute gradient focus function.

Quantitative analysis of microstructures using computerized stereology systems is an essential tool in many disciplines of bioscience research. Section thickness determination in current nonautomated approaches requires manual location of upper and lower surfaces of tissue sections. In contrast to conventional autofocus functions that locate the optimally focused optical plane using the global maximum on a focus curve, this study identified by two sharp 'knees' on the focus curve as the transition from unfocused to focused optical planes. Analysis of 14 grey-scale focus functions showed, the thresholded absolute gradient function, was best for finding detectable bends that closely correspond to the bounding optical planes at the upper and lower tissue surfaces. Modifications to this function generated four novel functions that outperformed the original. The 'modified absolute gradient count' function outperformed all others with an average error of 0.56 µm on a test set of images similar to the training set; and, an average error of 0.39 µm on a test set comprised of images captured from a different case, that is, different staining methods on a different brain region from a different subject rat. We describe a novel algorithm that allows for automatic section thickness determination based on just out-of-focus planes, a prerequisite for fully automatic computerized stereology.

Cognitive decline strongly correlates with cortical atrophy in Alzheimer's dementia.

Alzheimer's disease (AD) is characterized by progressive dementia and distinct neuropathology at autopsy. In order to test the relationship between dementia severity and loss of brain volumes, we prospectively documented the neurological/medical health of 26 male and 26 female controls and AD cases, and evaluated a subset of controls and AD cases using the Mini Mental State Examination (MMSE). At autopsy, Consortium to Establish a Registry for Alzheimer's Disease (CERAD) criteria confirmed diagnoses in 33 AD cases and 19 controls, and using unbiased stereology we quantified total volumes of cortical gray matter, subcortical grey matter including white matter, and forebrain. For ages of death between 50 to 100 years, controls showed minor cortical atrophy in the absence of cognitive decline. Cortical atrophy in AD cases was 20 to 25% greater than that in controls; AD patients dying at older ages showed less severe cortical atrophy than those dying at younger ages. Across all AD cases there was a strong correlation between cognitive performance on the Mini Mental State Examination and cortical volume loss. These findings confirm fundamental differences in the temporal patterns of cortical volume loss in aging and AD, and support cortical degeneration as the primary basis for cognitive decline in AD.

Stereological analysis of astrocyte and microglia in aging mouse hippocampus.

Recent evidence suggests neuroglia-mediated inflammatory mechanisms may stimulate neurodegenerative processes in mammalian brain during aging. To test the hypothesis that the number of microglia and astrocytes increase in the hippocampus during normal aging, unbiased stereological techniques were used to estimate total cell number in hippocampal subregions (CA1, dentate gyrus and hilus) of male C57BL/6J mice of different ages: 4-5 months, 13-14 months and 27-28 months. Immunocytochemical visualization for microglia and astrocytes were via Mac-1 and GFAP antibody, respectively. Estimates of total microglia and astrocyte number were assessed using the optical fractionator. No statistically significant age differences were found in the numbers of microglia or astrocytes in the hippocampal regions sampled. These findings suggest that age-related increases in the total numbers of hippocampal microglia and astrocytes is not causal for observed age-related increases in cytokine response.

Hippocampal neuron and synaptophysin-positive bouton number in aging C57BL/6 mice.

A loss of hippocampal neurons and synapses had been considered a hallmark of normal aging and, furthermore, to be a substrate of age-related learning and memory deficits. Recent stereological studies in humans have shown that only a relatively minor neuron loss occurs with aging and that this loss is restricted to specific brain regions, including hippocampal subregions. Here, we investigate these age-related changes in C57BL/6J mice, one of the most commonly used laboratory mouse strains. Twenty-five mice (groups at 2, 14, and 28-31 months of age) were assessed for Morris water-maze performance, and modern stereological techniques were used to estimate total neuron and synaptophysin-positive bouton number in hippocampal subregions at the light microscopic level. Results revealed that performance in the water maze was largely maintained with aging. No age-related decline was observed in number of dentate gyrus granule cells or CA1 pyramidal cells. In addition, no age-related change in number of synaptophysin-positive boutons was observed in the molecular layer of the dentate gyrus or CA1 region of hippocampus. We observed a significant correlation between dentate gyrus synaptophysin-positive bouton number and water-maze performance. These results demonstrate that C57BL/6J mice do not exhibit major age-related deficits in spatial learning or hippocampal structure, providing a baseline for further study of mouse brain aging.

Estrogens modulate experimentally induced apoptosis of granule cells in the adult hippocampus.

Estrogens are known to have broad effects on neuronal plasticity, but their specific role in neuronal cell death has not been determined. In the present study, we investigated the effects of beta-estradiol on an experimental model of apoptosis of granule cells of the dentate gyrus, i.e., apoptosis induced by intraventricular injection of the microtubule polymerization inhibitor colchicine. Cell death was characterized with multiple methods, including TUNEL and DNA electrophoresis. Nonrandom digestion of DNA was observed within 8-10 hours after colchicine injection, followed by condensation and fragmentation of granule cell nuclei and extensive anterograde degeneration of mossy fibers/terminals in 2 days. We compared the outcomes of the above-described manipulation in ovariectomized or sham-operated rats and animals treated daily with beta-estradiol or vehicle. Animals were lesioned with colchicine or vehicle 2 weeks after ovariectomy or sham operation. Beta-estradiol or vehicle was administered for 1 week prior to lesion and was continued for a further 2 weeks. Total numbers and densities of granule cells in different animal groups were counted by stereology in various anteroposterior levels of the hippocampus. Our results show that ovariectomy intensifies colchicine-induced granule cell apoptosis, which is ameliorated by exogenous beta-estradiol. In doses that ameliorate the adverse effect of ovariectomy, exogenous beta-estradiol appears to have no effect of preventing granule cell death in animals with intact ovaries; i.e., an estrogen excess is not more neuroprotective than physiological levels of these hormones. Taken together, our results indicate that estrogen deprivation increases the vulnerability of hippocampal neurons to injury and may predispose to neurological diseases occurring after menopause.

Effects of estrogen and raloxifene on neuroglia number and morphology in the hippocampus of aged female mice.

Hormone replacement therapy with the gonadal steroid estrogen or synthetic agents such as raloxifene, a selective estrogen receptor modulator, may affect cellular function in brains of postmenopausal women. In vitro studies suggest that 17beta estradiol and raloxifene can alter the microglial and astrocyte expression of immuno-neuronal modulators, such as cytokines, complement factors, chemokines, and other molecules involved in neuroinflammation and neurodegeneration. To directly test whether exogenous 17beta estradiol and raloxifene affect the number of glial cells in brain, C57BL/6NIA female mice aged 20-24 months received bilateral ovariectomy followed by s.c. placement of a 60-day release pellet containing 17beta estradiol (1.7 mg), raloxifene (10 mg), or placebo (cholesterol). After 60 days, numbers of microglia and astrocytes were quantified in dentate gyrus and CA1 regions of the hippocampal formation using immunocytochemistry and design-based stereology. The results show that long-term 17beta estradiol treatment in aged female mice significantly lowered the numbers of astrocytes and microglial cells in dentate gyrus and CA1 regions compared with placebo. After long-term treatment with raloxifene, a similar reduction was observed in numbers of astrocytes and microglial cells in the hippocampal formation. These findings indicate that estrogen and selective estrogen receptor modulators can influence glial-mediated inflammatory pathways and possibly protect against age- and disease-related neuropathology.

Length measurement: new developments in neurostereology and 3D imagery.

Quantification of linear biological structures has important applications in neuroscience; for example, the length of neurotransmitter-specific axonal innervation or length of dendritic processes within particular brain structures. Until recently, however, there have been practical limitations in the application of stereological tools for the unbiased estimation of object length on tissue sections. The recent development of efficient new approaches allows for the wider application of theoretically unbiased sampling and estimation techniques that are devoid of the assumptions and models of earlier methods. In this review, we outline the historical background and recent advances in the estimation of total length for biological objects on tissue sections, including a practical method to estimate the length of cholinergic fibers using newly developed methods. These newer methods also take advantage of three-dimensional image datasets and virtual probes, techniques that may have wider application in quantitative morphometry.

Length measurement: new developments in neurostereology and 3D imagery.

Quantification of linear biological structures has important applications in neuroscience; for example, the length of neurotransmitter-specific axonal innervation or length of dendritic processes within particular brain structures. Until recently, however, there have been practical limitations in the application of stereological tools for the unbiased estimation of object length on tissue sections. The recent development of efficient new approaches allows for the wider application of theoretically unbiased sampling and estimation techniques that are devoid of the assumptions and models of earlier methods. In this review, we outline the historical background and recent advances in the estimation of total length for biological objects on tissue sections, including a practical method to estimate the length of cholinergic fibers using newly developed methods. These newer methods also take advantage of three-dimensional image datasets and virtual probes, techniques that may have wider application in quantitative morphometry.

Absolute number and size of pigmented locus coeruleus neurons in young and aged individuals.

Significant loss of noradrenergic neurons of the locus coeruleus in aging and Alzheimer's disease has been reported. The interpretation of these analyses, however, is problematic because of the model- and assumption-based nature of conventional sampling and estimation techniques. In the present study, unbiased stereological methods were used to estimate the total number and mean cell volume of pigmented neurons of the locus coeruleus in the brains of young and aged nondemented persons. No side-to-side differences are seen, and there is no change in pigmented cell number or size in the locus coeruleus of nondemented older persons as compared with that of young individuals. In light of previous studies that show severe locus coeruleus cell loss in Alzheimer's disease, these data support further critical investigations into the possible protective role of noradrenaline in normal cognitive functions and emphasize the importance of avoiding methodological bias in quantitative neuroanatomical studies.

What counts in brain aging? Design-based stereological analysis of cell number.

The advent and implementation of new design-based stereological techniques allows the quantification of cell number without the assumptions required when obtaining areal densities. These new techniques are rapidly becoming the standard for quantifying cell number, particularly in aging studies. Recently, studies using stereological techniques have failed to confirm earlier findings regarding age-associated neural loss. This newly emerging view of retained cell number during aging is having a major impact on biogerontology, prompting revaluation of long-standing hypotheses of age-related cell loss as causal for age-related impairments in brain functioning. Rather than focus on neuronal loss as the end-result of a negative cascade of neuronal injury, research has begun to consider that age-related behavioral declines may reflect neuronal dysfunction (e.g., synaptic or receptor loss, signal transduction deficits) instead of neuronal death. Here we discuss design-based stereology in the context of age-related change in brain cell number and its impact on consideration of structural change in brain aging. Emergence of this method of morphometrics, however, can have relevance to many areas of gerontological research.

Empirical assessment of synapse numbers in primate neocortex.

Reliable methods are needed to assess the impact of synaptic loss on brain function. In this empirical study we demonstrate a novel and efficient method using immunocytochemistry (ICC) and modern stereological techniques to quantify synapses in neocortex of adult primates (Macaca fascicularis). Systematic-uniform-random sections through forebrain from two 10-year-old monkeys were immunostained for estimation of synaptophysin-immunoreactive (synaptophysin-IR) presynaptic boutons (synapses). Adjacent sections were stained with cresyl violet for estimation of total number of neuronal cell bodies. The unbiased Cavalieri method was used to estimate total forebrain and neocortical volumes to a high level of precision (coefficient of error (CE) < or = 0.10)). Synapse-to-neuron ratios varied from 860 in frontal cortex to 2300 in parietal-temporal cortex. The combination of Cavalieri and optical disector methods provided a direct means of estimating approximately 1.25 trillion (x 10(12)) total synaptophysin-immunopositive boutons and approximately 1.01 billion (x 10(9)) cell bodies in neocortex, with low CEs (0.12). Time required to make precise estimates of total neocortical and forebrain volumes and total numbers of synapses and neurons in neocortex was approximately 2-3 h per case from stained sections. The approach is a direct and efficient technique to quantify total synapse and neuron numbers within a defined brain structure.

Stereological estimation of total microglia number in mouse hippocampus.

Microglia are brain cells of considerable interest because of their role in CNS inflammatory responses and strong association with neuritic plaques in Alzheimer's disease (AD). In the present study, immunocytochemistry was combined with unbiased stereology to estimate the mean total number of microglia in dentate gyrus and CA1 regions of the mouse hippocampus. Systematic-uniform-random (SUR) sections were cut through the hippocampal formation of male C57BL/6J mice (n = 7, 4-5 months) and immunostained with Mac-1, an antibody to the complement subunit 3 receptor (CR3). The total number of Mac-1 immunopositive cells was determined using the optical fractionator method. The mean total number of microglia in the mouse dentate gyrus was estimated to be 20,300 (CV = 0.21) with a mean coefficient of error (CE) = 0.09. The mean total number of microglia in the mouse CA1 was estimated to be 43,200 (CV = 0.24) with a CE = 0.11. Comparison of total number estimates, derived from fraction- or volume-based methods, supported stereological theory regarding the equivalence of the two techniques. The time required to determine total microglia number in both hippocampal sub-regions was approximately 6 h per mouse from stained sections. The combination of immunocytochemistry and stereology provides a reliable means to assess microglia number that can easily be adopted for studies of transgenic and lesion-based models of aging and neurodegenerative diseases.

Induction of cortical cholinergic hypofunction and memory retention deficits through intracortical AF64A infusions.

Ethylcholine mustard aziridinium ion (AF64A), an irreversible inhibitor of high-affinity choline uptake on cholinergic nerve terminals, appears to selectively decrease presynaptic cholinergic markers after intracerebral injection. To restrict AF64A's action to cholinergic terminals within the frontoparietal (FP) cortex, the present study utilized multiple-site cortical infusions of the agent. Following an extensive histological analysis, a dose of 1 nmol AF64A/1 microliter was selected for determining AF64A's effects on acetylcholinesterase (AChE) staining, cortical cholinergic/non-cholinergic markers, and passive avoidance behavior. Adult rats given two infusions of AF64A into the right FP cortex had reduced AChE staining throughout 75% of the ipsilateral FP cortex at 10 days following infusion, thus suggesting an extensive cortical diffusion of the agent; minimal non-specific damage was seen (totalling only 4% of the ipsilateral FP cortex for both infusion sites) and no effects on AChE staining were observed in the striatum or hippocampus. Three weeks after bilateral AF64A infusions into the FP cortex (two injections on each side), significant frontal cortex deficits were observed in high-affinity choline uptake, acetylcholine synthesis, acetylcholine release, and hemicholinium-3 binding compared to vehicle-infused controls. However, choline acetyltransferase activity within the anterior cortex did not appear to be consistently affected by AF64A infusion. Cortical glutamic acid decarboxylase activity, as well as cortical monoaminergic markers, and neuropeptide levels were also unaffected. Moreover, animals that received bilateral AF64A infusions and were tested two weeks afterwards showed marked memory retention deficits during both the 24-h and 48-h postshock trials of passive avoidance testing. These results indicate that cortical AF64A infusion induces a specific, long-term cholinergic hypofunction of presynaptic markers within the cortex, resulting in a significant long-term memory impairment. Since the primary cholinergic innervation to the FP cortex, originating in the nucleus basalis of Meynert, appears to become dysfunctional (but not totally degenerative) in Alzheimer's disease, cortical AF64A infusions may closely reflect this cholinergic dysfunction by 'functionally' eliminating cortical cholinergic terminals.

Atrophy of cholinergic neurons within the rat nucleus basalis magnocellularis following intracortical AF64A infusion.

The rat nucleus basalis magnocellularis (nBM) was morphometrically analyzed following multiple intracortical AF64A infusions. At 3 weeks post-infusion, brains were histochemically double-stained for acetyl-cholinesterase and Nissl substance following diisopropylfluorophosphate pretreatment. Intracortical AF64A induced significant atrophy, but not degeneration, of nucleus basalis cholinergic cell bodies. These results suggest that retrograde cellular atrophy is associated with inhibition of presynaptic high-affinity choline transport on cortical terminals of nBM cholinergic neurons.

Intracortical AF64A: memory impairments and recovery from cholinergic hypofunction.

The long-term effects of intracortical AF64A (ethylcholine mustard aziridinium ion) treatment on presynaptic cortical cholinergic markers and cognitive function in the rat were investigated. Two 1.0 microliter infusions of AF64A (1 nmole/microliter) or vehicle were placed bilaterally into the fronto-parietal cortex. At 3 weeks postinfusion, AF64A-treated animals were found to be deficient in passive avoidance memory retention. During weeks 4 through 10, AF64A-treated animals were markedly deficient in the extinction (memory) phase, but not the acquisition (learning) phase of 2-way active avoidance behavior. Cortical acetylcholine synthesis and high-affinity choline uptake were significantly decreased in AF64A-treated animals at 24 hours, 3 weeks, and 10 weeks following infusion. At 6 months after AF64A treatment, however, cortical cholinergic markers were not reduced compared to controls. These data indicate that memory deficits are observed during a period of cortical cholinergic hypofunction induced by cortical AF64A infusions and that a recovery from such hypofunction occurs by 6 months after these infusions.

The effects of age and lipopolysaccharide (LPS)-mediated peripheral inflammation on numbers of central catecholaminergic neurons.

Parkinson's disease (PD), an age-related movement disorder, is characterized by severe catecholaminergic neuron loss in the substantia nigra pars compacta (SN(PC))-ventral tegmental area (VTA) and locus coeruleus (LC). To assess the stability of these central catecholaminergic neurons following an acute episode of severe inflammation, 6 to 22 month old C57/Bl6 mice received a maximally tolerated dose of lipopolysaccharide (LPS) followed by euthanasia 2 hours later to assay peak levels of peripheral and central cytokines; and, 14 weeks later for computerized stereology of tyrosine hydroxylase-immunopositive (tyrosine hydroxylase-positive [TH+]) neurons in the SN(PC)-VTA and LC. Two hours after LPS, cytokine levels varied in an age-related manner, with the greatest peripheral and central elevations in old and young mice, respectively. Severe inflammation failed to cause loss of TH+ neurons in SN(PC)-VTA or LC; however, there was an age-related decline in these TH+ neurons in LPS-treated and control groups. Thus, unknown mechanisms in the B6 mouse brain appear to protect against catecholaminergic neuron loss following an acute episode of severe inflammation, while catecholaminergic neuron loss occurs during normal aging.

Nerve growth factor increases the size of intracortical cholinergic transplants.

The effects of continuous intracortical mouse Nerve Growth Factor on fetal rat basal forebrain transplants in denervated adult rat neocortex were investigated. Enzyme-linked immunoassay (ELISA) was used to measure the time course of endogenous NGF protein production in neocortex, hippocampus, and basal forebrain in a cohort of animals receiving unilateral ibotenic acid (IBO) lesions of the nucleus basalis magnocellularis (nBM). A second cohort of IBO-nBM lesioned animals received transplants of fetal basal forebrain followed by two to four weeks of continuous NGF or cytochrome-C infusion into the ipsilateral frontoparietal neocortex. To study the effects of abnormally high NGF doses on transplanted and host tissue, the cumulative dose of intracortical NGF was on the order of micrograms, compared with maximum picogram levels of neocortical NGF produced following IBO-nBM lesions. A four-fold increase in transplant size, and greater cell and fiber densities were observed in NGF-treated compared with NGF-untreated transplants. No adverse histological effects of long-term, high-dose NGF treatment were observed on transplanted basal forebrain or host neocortical tissue. These data indicate that cholinergic-rich mammalian brain tissue and intrinsic host tissue can be stimulated by high doses exogenous NGF without obvious deleterious effects.

Design-based estimation of surface area in thick tissue sections of arbitrary orientation using virtual cycloids.

Surface area is a first-order stereological parameter with important biological applications, particularly at the intersection of biological phases. To deal with the inherent anisotropy of biological surfaces, state-of-the-art design-based methods require tissue rotation around at least one axis prior to sectioning. This paper describes the use of virtual cycloids for surface area estimation of objects and regions in thick, transparent tissue sections cut at any arbitrary (convenient) orientation. Based on the vertical section approach of Baddeley et al., the present approach specifies the vertical axis as the direction of sectioning (i.e. the direction perpendicular to the tissue section), and applies computer-generated cycloids (virtual cycloids) with their minor axis parallel to the vertical axis. The number of surface-cycloid intersections counted on focal planes scanned through the z-axis is proportional to the surface area of interest in the tissue, with no further assumptions about size, shape or orientation. Optimal efficiency at each x-y location can be achieved by three virtual cycloids orientated with their major axes (which are parallel to the observation planes) mutually at an angle of 120 degrees. The major practical advantage of the present approach is that estimates of total surface area (S) and surface density (SV) can be obtained in tissue sections cut at any convenient orientation through the reference space.

Stereological length estimation using spherical probes.

Lineal structures in biological tissue support a wide variety of physiological functions, including membrane stabilization, vascular perfusion, and cell-to-cell communication. In 1953, Smith and Guttman demonstrated a stereological method to estimate the total length density (Lv) of linear objects based on random intersections with a two-dimensional sampling probe. Several methods have been developed to ensure the required isotropy of object-probe intersections, including isotropic-uniform-random (IUR) sections, vertical-uniform-random (VUR) slices, and isotropic virtual planes. The disadvantages of these methods are the requirements for inconvenient section orientations (IUR, VUR) or complex counting rules at multiple focal planes (isotropic virtual planes). To overcome these limitations we report a convenient and straightforward approach to estimate Lv and total length, L, for linear objects on tissue sections cut at any arbitrary orientation. The approach presented here uses spherical probes that are inherently isotropic, combined with unbiased fractionator sampling, to demonstrate total L estimation for thin nerve fibres in dorsal hippocampus of the mouse brain.