| Alzheimer's disease(AD) is a high prevalence neurodegenerative disease accompanied bygradual and irreversible behavioral and cognitive impairments. Brain lesions observedduring the course of AD involve two main aspects: extracellular amyloid-beta(Aβ)deposition as senile plaques and intracellular tau accumulation forming neurofibrillarytangles and promoting cytoskeletal disorganization. Current research on AD is largelyguided by a dominant physiopathogenic hypothesis, the so-called amyloid cascade theory.Regularly commented on and amended, this model posits accumulation of Aβin thebrain, as a key primary event that determines the onset of other brain alterations(e.g.synaptic and neuronal death), finally leading to the clinical stage of dementia. Supportingthe amyloid cascade hypothesis, early-onset familial forms of AD are associated withmutations in different genes(Amyloid Precursor Protein(APP) and Presenilins 1&2,(PS1&2)) involved in the biosynthesis of Aβ. Dysfunction of these genes is logicallythought to compromise the normal catabolism of APP resulting in exaggerated Aβproduction. Increased Aβproduction and parenchymal amyloid plaques are indeeddescribed in transgenic mice overexpressing one or more of these mutated genes. Thesemice subsequently develop neuropathological alterations and behavioural impairmentsmimicking AD phenotype.The exact impact of brain Aβaccumulation on clinical symptoms remains to-date difficultto decipher, both in AD patients and in animal models of the disease. Clinico-pathologicalcorrelative analyses have led to mitigated conclusions and it is now considered thatpre-plaques Aβassemblies are the most deleterious species while aggregated insolubledeposits have a reduced pathogenicity. In addition, the relationship between Aβaccumulation and brain dysfunction might be indirect and mediated by secondaryneuropathological alterations. For instance, white matter anomalies are described in ADpatients, presumably in association with cerebro-vascular impairments, and in transgenicmice modeling brain amyloidosis. They can be detected through conventionalpostmortem neuropathological examination, but also in vivo by means of dedicatedtechniques such as diffusion tensor imaging and it has recently been proposed that white matter defects are potent diagnostic biomarkers for AD. The functionalconsequences of altered white matter are obvious: Disconnection of neural networks occurfollowing fibres loss, leading to diaschesis and cortical disorganized activity. Also,disruption of myelin in the CNS white matter might have deleterious effects on neuronalcommunication by altering propagation of action potentials and increasing brain energyexpenditure.Importantly, white matter anomalies might reflect either loss of fibers and/or demyelinationbut deciphering between the two processes could sometimes be hazardous. In particular, ithas been demonstrated that AD patients show concurrent decreased axonal densities andmyelin breakdown.It is clear, especially from experimental studies in AD animal models, that axonal pathologycan be driven by Aβdeposition in the brain therefore possibly bridging the gapbetween amyloidosis and behavioural/cognitive impairments.The aim of the present study was to further evaluate white matter integrity in a doubleAPPxPS1 mouse transgenic model with aggressive Aβ-related pathology. Twoaxonal bundles(corpus callosum and anterior commissure) that show significant alterationsin AD patients were selected and analyzed: axonal densities were quantified by meansof anti-neurofilament immunostainings and myelin integrity was evaluated byhistochemistry using a gold chloride staining that provides high contrast and spatialresolution. Age-dependent anomalies were detected in the white matter of APPxPS1mice and we further assessed the relationship between these alterations and Aβdeposition.Methods1) Three groups of new-born rats, each group contained 5 rats, which were 7 days,14days and 21 days old. The other 13 Alzheimer's disease transgenic rats, contained 5APP/PS1 mol/Lice and PS1 mol/Lice. Following decapitation the brains were extractedand treated. The brains were cut by the microtome. The brain tissue was stained byimproved gold chloride myelin staining. Observe the staining results by lightmicroscope directly, and quantitatively analyze staining by optical density.2) Adjecent series of APP/PS1 brain tissure was stained myelin with gold chloride and axon with anti-neurofilament M145 antibody, we took corpus callosum and anteriorcommissure as representative regions of white matter. We measured the ROD(relativeoptical density) of representative regions, and quantitatively analyse myelinationdensity and axon density.3) Evaluation of the extracellular amyloid load was only performed on the 24-months-oldAPP/PS1 mice as almost no Congo red positive deposits were observed in youngdouble transgenic mice.Plaques loads were quantified using computer-based thresholding methods. Scanswere prepared using Photoshop CS2 to outline selected regions of interest. Imageswere then processed with ImageJ freeware(Rasband, W. S., Image J, U. S. NationalInstitutes of Health, Bethesda, Maryland, USA, http://rsb.info.nih.gov/ij/, 1997-2006)using a dedicated macrocommand that extracts amyloid deposits from backgroundtissue. Regional amyloid loads were expressed as percent of tissue surface stained bythe Congo red dye that corresponds to the proportion of plaques volume according toDelesse principle(Delesse, 1847). Amyloid loads were evaluated 1) in the whole brain, 2) in the rostral isocortex that is richly innervated by axons passing through theanterior corpus callosum, and 3) at the level of the two fibre tracts that wereinvestigated in the present study(anterior commissure, corpus callosum). Amyloidloads in the different regions were analyzed correlatively with the axon density andmyelin density in the corresponding regions and the atrophy of corpus callosum andanterior commissure.4) Intraneuronal Aβimmunoreactivity was only quantified in the young APPxPS1 micebecause intracellular Aβstaining is absent in the aged mice of this line. Twosections/animals were selected at the level of the rostral cortex and a semi-quantitativeanalysis, based on a four points scale, was performed to evaluate levels of neuronalimmunostainings: 0: no obvious positive staining; 1: weak intracellular staining; 2:moderate staining; 3: strong staining.5) Staining of degenerating neurons was performed using the Fluoro-Jade B dye with aslightly modified protocol. Glass-mounted sections were passed through absoluteethanol and 75% ethanol followed by a 1 minute rinse in distilled water. Tissue wasthen incubated in 0.06% potassium permanganate solution for 15 minutes with slight agitation and rinsed before staining in Fluoro-Jade B(Histo-Chem., Jefferson, AR;0.001% solution prepared in 0.1% acetic acid; 30 minutes at RT). After extensiverinsing in distilled water, sections were dehydrated, cleared in xylen and coverslipped.ResultsMyelin staining with improved gold chloride staining7 days of rats are completely deficient of myelin staining. Quantitative analysis ofmyelination confirmed that myelination of 21 days old rats was higher than that of 14 daysold rats.Pathological myelin staining was observed in APP/PS1 group. Myelination of APP/PS1mice was obviously lower than that of PS1 mol/Lice. There is conspicuous demyelinationin APP/PS1 group.Altered volumes of fibre tracts in APPxPS1 miceGold chloride myelin staining, as compared to standard stains(e.g, HE or Nissl stains), wasallowed to precisely outline the area of the corpus callosum. In particular, delineating thecorpus callosum from adjacent white matter tracts(e.g. cingulate bundle, dorsal fornix, anddorsal hippocampal commissure) was greatly facilitated on myelin-stained sections. Laterallimits of the corpus callosum and borders of external capsula were identified by ahorizontal to vertical shift in fibre orientation. Also, the anterior commissure was easilyidentified and outlined from gold chloride-stained sections.The size of the anterior commissure was similar in 2-months old PS1 mice andage-matched APPxPS1 transgenics(t(13)=-0.58, P＞0.05). On the contrary, the callosal sizewas significantly decreased in young APPxPS1 as compared to controls(t(13)=3.501,p＜0.005). Subregional analysis indicated a significant reduction in the size of the rostralcorpus callosum of young APPxPS1 mice (t(13)=3.743, p＜0.005) while there was nodifferences between genotypes in the surface area of the posterior corpus callosum(t(13)=0.136, P＞0.05).With aging, a significant increase in white matter volumes was observed in PS1 control mice(corpus callosum: t(12)=3.858, p＜0.005; anterior commissure: t(12)=4.275,p＜0.005). This phenomenon was clearly not evidenced in the double APPxPS1 transgenics:in this genotype, the size of the corpus callosum remained constant between 2 and 24months(t(12)=1.850, P＞0.05) and surface area of the anterior commissure evenundergoes atrophy with aging(t(12)=2.284, p＜0.05). As a consequence, strong differencesbetween genotypes were observed in 24-months old mice with APPxPS1 transgenicsshowing, in comparison to PS1 controls, decreased white matters surface areas. This wasobserved at the level of the anterior commissure (t(11)=6.388, p＜0.0001) and of thecorpus callosum(total: t(11)=4.653, p＜.001; anterior: t(11)=5.404, p＜0.0005). The onlyposterior part of the corpus callosum did not show significant atrophy in old APPxPS1 mice(t(11)=1.492, P＞0.05).Potentiation of axonal loss in old APPxPS1 miceAxonal densities in the corpus callosum and anterior commissure were quantified usingROD analysis of neurofilament immunostainings.In 2-months old mice axonal densities were similar in both genotypes whatever the fibretract considered(all p＞0.35). With aging, a severe decrease in neurofilament staining wasobserved, both in APPxPS1 and PS1 mice, testifying for axonal loss in the corpus callosumand anterior commissure(all p＜0.0001). Age-related reduction of neurofilament stainingwas however largely more pronounced in old APPxPS1 mice than in aged PS1 controls.Decreased axonal densities in old APPxPS1 mice was further confirmed in the differentsub-regions of the corpus callosum(all p＜0.0001) and also at the level of the anteriorcommissure(t(10)=4.16; p＜0.005).Abnormal myelination in old APPxPS1 miceROD analysis of gold chloride stainings in 2-months old mice indicated comparable myelindensities in PS1 and APPxPS1 transgenics(corpus callosum: t(13)=0.318, P＞0.05;anterior commissure: t(13)=1.277, P＞0.05).The myelination of the anterior commissure was not affected by aging(PS1: t(12)=1.292,P＞0.05; APPxPS1: t(12)=0.556, P＞0.05) and myelin densities in this fibre tract were similar in 24-months old PS1 and APPxPS1 mice (t(11)=0.679, P＞0.05). On the otherside, an increase of the myelination of the corpus callosum was observed when comparing2-months and 24-months old PS1 mice (t(12)=2.823, p＜0.05). Noticeably, myelinationbuild up with progressive aging was observed in control animals in the rostral corpuscallosum(t(12)=4.171, p＜0.005) but not in its posterior part(t(12)=0.461, P＞0.05).Contrarily to PS1 mice, such age-dependent callosal myelination was not observed inAPPxPS1 mice (t(12)=0.7, P＞0.05) and consequently decreased myelin staining wasevidenced in 24-months old APPxPS1 mice when compared to PS1 age-matched controls(total corpus callosum: t(11)=3.332, p＜0.01). Differences between genotypes were furtherconfirmed at the level of the anterior corpus callosum: (t(11)=3.512, p＜0.005) while nodifference between PS1 and APPxPS1 mice was evidenced in more caudal regions of thecorpus callosum (t(11)=1.9; P＞0.05).Qualitative examination of myelin stained sections was then performed in old APPxPS1mice. No evidence of myelin breakdown(debris) was found in the large myelinated bundlesof the corpus callosum. However, in comparison to control animals, myelin appeared to befragmented in the isocortex and the hippocampus of APPxPS1 mice. Myelin material wasoften detected under the form of small tortuous segments with bead like varicosities. Thesemorphological anomalies, absent in young APPxPS1 mice, were found at the vicinity of Aβaggregation sites but also in the parenchyma in areas distant from plaques.Relationship with Aβpathology and neurodegenerationCongo red positive aggregates were detected and quantified in the anterior commissure(mean load=2.8%; min=1.6%; max=6%) and in the corpus callosum(mean load=2.2%;min=1.6%; max=2.7%) of old APPxPS1 mice. Correlative analysis did not revealsignificant associations between local amyloid loads in fibre tracts and white matteranomalies(decreased axonal densities and myelination: all p＞0.111). Also there were nocorrelations between morphology of the corpus callosum/anterior commissure and totalbrain or cortical amyloid loads (all p＞0.196).In addition to amyloid plaques loads, intraneuronal Aβwas semi-quantitatively assessed inthe frontal cortex of young APPxPS1 mice. As expected from previous observations, positive labeling was detected using the 4G8 antibody in a subset of cells. Staining wasmainly observedin deep cortical layers(Ⅴ) involving a distinctive band of large pyramidalcells. However, there were no associations between levels of intracellular Aβthat maysignificantly vary from one animal to the other(mean=7.8; min=4.5; max=11.5) andaxonal and myelin markers(all p＞0.119).The Fluoro-Jade B dye was used to assess neurodegeneration in APPxPS1 mice but nopositive neurons were detected in the studied animals(data not showed). In particular nodegenerating neurons were observed in the cortical layers with high densities of Aβpositiveneurons(see above). Only the core of amyloid deposits and surrounding degeneratingdystrophic neurites as well as reactive astrocytes were detected with Fluoro-Jade B in oldAPPxPS1 mice.Conclusions1) Improved myelin staining is rapid,simple,sensitive and stable, can be used for bothqualitative and quantitative analysis of myelination.2) Gold chloride myelin staining allowed outlining precisely the shape of the corpuscallosum and measuring the size, which give a simple and delicate method to analyzequantitatively the atrophy of the corpus callosum.3) There was conspicuous atrophy in the fiber tracts in the anterior brain of old APP/PS1mice, which is related with both axon loss and demyelination. However, thepotentiation of axon loss may be the first reason for the atrophy of fiber tracts.4) The amyloid loads of the whole brain and the regions of interest didn't correlate withthe anormalies of fiber tracts, including axon loss and demyelination. That mean,extracellular amyloid beta doesn't have a clear pathogenicity.5) Intracellular amyloid beta deposits are mostly located in the cortical layer V. the cells inCC, also the origins of callosal afferents, have been traced back to cell bodies in layerV. One may hypothesize that, early during aging, APPxPS1 mice accumulate Aβin asubset of cortical neurons that later become dysfunctional, develop axonal pathology(loss of neurofilament immunoreactivity and of myelin) and eventually die.
|
PDF Full Text Request