PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Neurobiol Aging. Author manuscript; available in PMC 2012 May 1.
Published in final edited form as:
PMCID: PMC2888951
NIHMSID: NIHMS121635

Age-associated cerebral atrophy in mouse lemur Primates

Abstract

We assessed the regional brain atrophy in mouse lemur primates from 4.7Tesla T2-weighted magnetic resonance images. Thirty animals aged from 1.9 to 11.3 years were imaged. Sixty one percent of the 23 animals older than 3 years involved in the study displayed an atrophy process. Cross sectional analysis suggests that the atrophy follows a gradual pathway, starting in the frontal region then involving the temporal and/or the parietal part of the brain and finally the occipital region. Histological evaluation of five animals selected according to various stages of atrophy suggested that extracellular amyloid deposits and tau pathology can not explain by themselves this atrophy and that intracellular amyloid deposition is more closely linked to this pathology. This study suggests that most of the age-related atrophy occurring in mouse lemurs is caused by one clinical, evolving, pathological process. The ability to follow this pathology non invasively by MRI will allow to further characterize it and evaluate its relationship with neuropathological lesions that are involved in human diseases such as Alzheimer.

Keywords: Alzheimer's disease, amyloid, animal models, atrophy, MRI, Tau

1. Introduction

In humans, cerebral atrophy occurs in the course of aging (Dekaban and Sadowsky, 1978, Good et al., 2001) and also in many age-related neurodegenerative diseases (Valk et al., 2002). The location and rate of atrophy depends on underlying pathological processes (Tisserand et al., 2004). For example, during Alzheimer's disease (AD) the atrophy starts in temporal areas and progressively reaches all the brain regions (Baron et al., 2001). In frontotemporal dementia, the atrophy is more prominent in the frontal lobes and anterior temporal regions (Bocti et al., 2006).

Animal models are critical to improve our understanding of age-related neurodegenerative diseases and to develop new treatments against these diseases. To date, most of the animal studies in the field of AD are based on transgenic mice (Delatour et al., 2006, Lee et al., 2005). Although very useful, these models do not reproduce all the aspects of human pathologies. In particular their cerebral atrophy is not similar to that found in AD patients (Delatour et al., 2006). Spontaneous models can be an alternative to transgenic animals. However, as for transgenics, the characterization of these models is a critical step prior to their use in therapeutic trials.

Mouse lemurs (Microcebus murinus) are small (about 12 cm, 100 g) primates with a mean life span of 5 years and a maximum life span of 12 years in captivity (Perret, 1997). Their brain weighs approximately 1.7 g (Bons et al., 1998). It measures 23 mm from the tip of the olfactory bulbs to the caudal end of the medulla, and has a maximum width of 18 mm (Le Gros Clark, 1931). Previous studies have shown that some aged mouse lemurs can present with some of the lesions that characterize AD such as extracellular amyloid deposits (Bons et al., 1991) or altered tau proteins (Bons et al., 1994, Delacourte et al., 1995). They also display age-related iron accumulation in the brain (Dhenain et al., 1998) and cerebral atrophy (Dhenain et al., 1997). Such alterations remind human cerebral aging. In a pilot study, which concerned only the temporo-parietal area, we described in a series of 30 lemurs an atrophy process starting at an age between 5 and 8 years and evolving rapidly once initiated (Dhenain et al., 2000). In a more recent study in a limited series of 12 animals, we reported that this atrophy process shows regional specificity (Dhenain et al., 2003). In these two studies, cerebral atrophy was detected by measuring cerebrospinal fluid (CSF) volume, an index of atrophy. Despite these first studies, the cerebral pathologies displayed by the mouse lemurs are still poorly characterized. One important question is the heterogeneity of the alterations within the animal population: do most of the animals with atrophic and neuropathological alterations suffer from the same disease or are they presenting with many different clinical entities?

Thus, the first aim of the current study was to evaluate by MRI, whether or not the atrophy process observed was consistent in all animals and could reflect different stages of a single pathological process. This in-vivo study was based on the automatic segmentation of MRI from mouse lemur brains and on the automated quantification of CSF volumes in various pericortical regions. As in our previous studies, CSF volumes in various regions surrounding the cortex were used as indexes of regional atrophies. Results showed focal, multi-focal and generalized atrophy patterns. These patterns were consistent with a progression of the atrophy starting from sharply demarcated regions toward a more generalized process encompassing the whole brain. In addition, neuropathological alterations were assessed in subgroups of aged lemurs selected according to different stages of localized and generalized atrophy. Intracellular and extracellular accumulations of amyloid, slight tau pathology as well as increased gliosis were found in various animals. Interestingly, the atrophied animals were those with the most severe neuropathology suggesting that aging in mouse lemurs can be associated with a progressive and consistent pattern of neurono-glial alterations and cerebral atrophy.

2. Material and Methods

2.1. Subjects

Thirty mouse lemurs aged from 1.9 to 11.3 years (young adult (from 1.9 to 2.5 years old, n=7), middle-aged animals (from 3 to 5.9 years old, n=11), and aged animals (6.4 to 11.3 years, n=12), born and raised in a laboratory breeding colony at Brunoy (France, agreement n°962773), were involved in the MRI study (Table 1). The brain of 5 of these animals, aged from 6.4 to 11.2 years, presenting with various atrophy levels (Table 2) were evaluated postmortem. The delay between MRI and death of these animals was less than one year.

Table 1
Overview of animals involved in the study. Age during MR exam, age at death and survival are in years.
Table 2
Overview of neuropathological findings in the animals studied by MRI. The atrophy patterns are showed in the third column. The following columns correspond to intracellular amyloid deposits, extracellular deposits, CP13 and GFAP stainings in the cortex ...

2.2. MRI evaluation of cerebral atrophy

2.2.1. In vivo three dimensional MR imaging

Brain images were recorded according to previously published protocols (Dhenain et al., 2003) on a 4.7 Tesla Bruker Biospec 47/30 system by using a surface coil (diameter = 30 mm) actively decoupled from the transmitting birdcage probe (Bruker GmbH). Briefly, animals were pre-anesthetized with atropine (0.025mg/kg subcutaneously) and anesthetized by isoflurane. Respiration rate was monitored to insure animal stability until the end of the experiment. Body temperature of the mouse lemur was maintained by using water filled heating blanket. Three-dimensional inversion-recovery fast spin-echo images (IR-RARE) were recorded with an isotropic nominal resolution of 234 μm (TR/TE=2500/6 msec, TEw=45 msec, TI=200 msec, RARE-factor=16).

2.2.2. Image post-processing and segmentation protocol

MR images were processed as described in Figure 1: first, the sensitivity gradient resulting from surface coil acquisition was corrected by a phantom based method that we previously validated (Fig. 1A) (Dhenain et al., 2003). Briefly, images of a 2.5% agar and NaCl 0.09% phantom were recorded under the same conditions as the brain images. Time domain data were filtered by a sine-squared function to enhance signal to noise ratio. Then, the 3D matrix that yielded isointense images of the phantom was calculated and applied to brain images (code written under IDL 5.4, Research Systems Inc.).

Figure 1
Processing pipeline used in the current study. A. Correction of sensitivity gradient resulting from surface coil acquisition by using a phantom based method (Dhenain et al., 2003). B. Creation of a mask for the brain and brain extraction thanks to this ...

Intensity non-uniformity in brain images was further slightly corrected before to apply the segmentation protocols. First, the brain and surrounding CSF were automatically extracted from other head tissues by using the freely available BrainVisa software (http://brainvisa.info/download.html) (VipGetBrain and VipClosing routines; Fig. 1B) (Mangin et al., 1998). The two eyes were also extracted by using the same routines. The non-uniformity in MR images intensity was corrected by using an automatic method based on the minimization of the entropy within the images also by using the BrainVisa software (Fig. 1C; VipBiasCorrection routine, (Mangin, 2000)). This was performed in a three step manner: The correction was applied to the brain alone, then to the eyes, finally it was performed for the brain and the eyes that were put back together on the images (-Kregul 10.0 -Dimfield 3 -sampling 40.0 -Grid 2.00). Then the images of the brain and of the eyes were automatically segmented by using published segmentation algorithms (Fig. 1D) (Dhenain et al., 2003). Briefly, in a preliminary step, before the segmentation procedure, the bit depth of MR images was reduced from 32 to 8 and the high intensity artifactual voxels were removed (Dhenain et al., 2003). Then, the automatic statistical segmentation protocol based on a classification-maximization algorithm that we previously developed was applied to the images to detect the thresholds that separated white matter (WM), gray matter (GM), and fluid (Dhenain et al., 2003). This technique allows unsupervised partitioning of voxels into four pre-defined bins (WM, GM, fluid, and one 'background'). Briefly, the algorithm performed voxel segmentation in two steps: 1) First, for each bin it estimated the bin mean (μ), variance (σ2), and fractional probability (πk). Then an iterative optimization of voxel attribution to the various bins was based on the calculation of modified Mahalanobis distance 2) During the final segmentation step, each voxel in the original input volume was segregated into its final tissue bin destination using a Gaussian probability equation, with the final estimated bin parameters computed in step 1 as coefficients (see (Dhenain et al., 2003) for a more focused description of the method).

The segmented MRI maps from the different mouse lemurs were rotated and shifted so that all the maps had the same orientation (Fig. 1E; code written under IDL 5.4) (see (Dhenain et al., 2003)). The orientation corresponded to that of a reference mouse lemur brain that we previously used as a reference for MRI analysis (Dhenain et al., 2003).

2.2.3. Regional evaluation of CSF volumes

In the present study, we focused on the CSF voxels that were localized between the frontal lobe and the occipital part of the brain. The most frontal slice taken into account was the first slice where the olfactory bulb was not detected anymore. CSF pixels were classified (Fig. 1F) according to their region of origin by using the following nomenclature: First, the ventricles were classified into three regions: the third ventricle (3V), the lateral ventricles (LV), and the external part of the temporal horns of the lateral ventricles (ETH). CSF from all the peri-encephalic spaces were categorized into various regions, called RPES (Regions of the Peri-Encephalic Spaces) as previously described (Dhenain et al., 2003). The RPES corresponded to regions that were directly surrounding the cortex (Fig. 2). They were identified as 1) the frontal (medio-dorsal) and parietal cortex (FPcort), 2) the frontal ventral region (Fvcort), 3) the temporal cortex (Tcort), 4) the secondary parieto-temporal cortex including the CSF localized in the interhemispheric fissure in the slab where the parieto-temporal cortex was defined (PTcort), 5) the occipital cortex (Ocort), and 6) the interhemispheric fissure in the most anterior parts of the brain (IHfiss). The method used to define the RPES has been previously reported (Dhenain et al., 2003). Briefly, RPES definition was based on the use of reference landmarks that could be easily identified on MR images and on the cytoarchitectural atlas of the mouse lemur brain by Le Gros Clark (Le Gros Clark, 1931). Because it is relatively straightforward to associate a MR slice to a given section from the atlas, the first landmarks were the sections from the atlas (Le Gros Clark, 1931). They were used to divide the brain into five coronal slabs (Fig. 2). In the second step, easily visible anatomical landmarks such as sulci or fissures, shape, and curvature of the white matter and of the brain, were used to define several “preliminary” RPES (pre-RPES) within each slab. For example, the Sylvian fissure separated temporal and parietal pre-RPES in the medium 1 slab. Seventeen pre-RPES were defined to classify the CSF from all the peri-encephalic spaces (see (Dhenain et al., 2003) for further description of the method). The criteria used to separate each slab and pre-RPES within each slab were meticulously noted and reported in an intranet web site (data not shown). Pre-RPES from different slabs were then assembled to define the final RPES.

Figure 2
Colored transparent layers corresponding to the RPES, pasted on a lateral view of the brain. Letters correspond to various slabs that were used to define brain regions (A: Anterior 1; B: Anterior 2; C: Medium 1; D: Medium 2; E: Posterior 1; F: Posterior ...

Thanks to this method, the voxels categorized as CSF by the automatic segmentation were manually labeled and classified as belonging to one of the ventricle regions or pre-RPES (Fig. 1F). CSF painting was performed by two evaluators (MD and EC). Once all images from all the animals had been labeled, we double-checked that the voxel attribution to the different ventricle regions and pre-RPES was consistent across animals and between the two evaluators. Then, the number of voxels belonging to each region was automatically counted (`Display' freeware (http://www.bic.mni.mcgill.ca/software/Display/Display.html)). A total brain atrophy index was also calculated by summing up CSF volumes from all the RPES and ventricles and a total cortical atrophy index was computed by adding CSF volumes from all regions surrounding the cortex. Intraclass correlation coefficients were calculated to evaluate the measures performed by the two evaluators before and after the control of the consistency of CSF labeling. The obtained coefficients ranged from 0.98 to 1 for the nine evaluated RPES and was equal to 0.97 and 1 for the total cortical atrophy and total atrophy indexes, respectively.

2.2.4. Staging of the atrophy in individual animals

For individual studies, regional atrophy from cortical areas (FPcort, Fvcort, Tcort, PTcort and Ocort) was quantified for each animal with a three grade scale (0: no atrophy; +: moderate atrophy, and ++: severe atrophy). To define the thresholds between the atrophy grades, animals younger than 3 years were taken as reference. The thresholds were Mcsf + (C × SDcsf) where Mcsf and SDcsf were the mean and standard deviation for CSF estimate in a given brain area for young adults. C was a constant equal to 2.58 (1% confidence interval of a Gaussian population) for the threshold between 0 and + levels, and 5 (arbitrary value) for the threshold between + and ++ levels. The atrophy was then represented on brain maps that showed the spatial location of the atrophy with a color code (0: blue; +: pink; ++: red).

2.3. Neuropathology

2.3.1. Tissue processing and histochemical staining

The mouse lemur brains were removed and fixed in 4% formalin after the natural death of animals. The brains were split into their right and left hemispheres. The right hemisphere from each animal was sectioned on the coronal plane on a cryostat (40-μm-thick sections) whereas the left hemisphere was used for paraffin sectioning (6-μm-thick sections) and sectioned following a sagittal plane. Right hemispheres were plunged 24 hours in a 15% and then 30% sucrose solution for cryoprotection. They were then frozen and sliced into 40-μm-thick coronal sections on a freezing Microtome. Slices were then stored at -20°C into a storage solution (Glycerol 30%, Ethylene Glycol 30%, and Phosphate Buffer 0,1M). Then, floating sections were rinsed with 0.1 M phosphate buffer and mounted on Superfrost plus glass slides before being stained for amyloid-β and Glial Fibrillary Acidic Protein (GFAP) detection. For anti-Aβ (biotinylated mouse monoclonal 4G8 antibody, Biovalley) and anti-GFAP (Rabbit Anti-GFAP, DAKO) immunohistochemistry, the brain samples were pretreated with hydrogen peroxide (0.3%). Then they were bathed into a Phosphate Buffer's (0.1M) Saline's (0.9%) and Triton's (SIGMA 100X, 0.2%) solution (PBS+Tx) with Normal Goat Serum (NGS) (Sigma G6767, 4.5%). Then, during two days, the samples were incubated into a solution of PBS+Tx with NGS (3%) and the primary antibody (concentrations of 1/1000 and 1/5000 for anti-amyloid and anti-GFAP, respectively). Then the samples were bathed for one hour into a solution PBS+Tx with NGS (3%) with the 1/250 biotinylated anti-mouse and anti-rabbit secondary antibodies (VECTOR BA-92000 and BA-1000 respectively). Next, for amyloid staining, the signal was amplified by using Avidine-Biotine-Vectastain Complex standard (ELITE PK 6100 0.225%). For anti-GFAP staining, the antibody was first bathed in a Tyramine Biotin Reagent bath (Blast PC 2815-0897 1:1000) before being amplified by Avidine-Biotine-Vectastain Complex (0.1125%). Final reaction made use of diaminobenzidine (DAB) nickel (VECTOR, SK-4100) or DAB for two minutes as chromogen (gray-black or brown stains, respectively). Negative controls were performed by omitting the primary antibody in the procedure. Positive controls for Abeta immunochemistry were performed using formalin-fixed samples taken from the superior temporal gyrus of human Alzheimer samples received in 1994 in the R. Escourolle Laboratory of La Salpêtrière Hospital (Paris, France). Aβ studies were also performed using automated immunostaining protocols based on Ventana Discovery XT technologies. To this aim, 40-μm-thick coronal sections, mounted on Polysine glass slides, were pretreated by an automated heat induced epitope retrieval procedure (tris/borate/EDTA buffer, pH 8.0, 8 minutes at 95°C followed by 4 minutes at 100°C) and were incubated with primary antibody (biotinylated mouse monoclonal antibody 4G8, 1/250, Sigma) during 60 minutes. A 32 minutes incubation with secondary antibody (biotinylated anti-mouse antibody, 1/2000, VECTOR BA-92000) was performed before the disclosure using the ABC method with DAB as the chromogen (brown stain). Finally, the sections were counterstained with hematoxylin (Ventana Medical Systems, 760-2021) and a bluing reagent (Ventana Medical Systems, 760-2037).

The left hemisphere was embedded in paraplast and cut into 6-μm-thick serial sagittal sections. The sections were tested for the presence of Tau pathology using the CP13 antibody specific for phosphorylated serine 202 (generous gift of P. Davies) (Andorfer et al., 2003). Sections were dewaxed in toluene and a graded series of alcohol solutions and were incubated with formic acid (80%) for 15 min. They were then incubated with 10% goat serum in Tris-buffered saline for 30 min and then with CP13 antibodies (1/200) during 4 days at 4°C. Immunological complexes were detected by sequential application of a 1/100 dilution of biotinylated immunoglobulin and horseradish peroxidase labeled avidin (Vector), for 30 min each. Immunoreactivity was detected with 0.005% DAB (Sigma). Standard immunohistochemical controls were performed, including (1) omission of the primary antibody, (2) use of irrelevant secondary antibodies, and (3) use of preimmune serum. No labeling was observed for any of the various controls. As all the controls were negative, the immunoreactivity observed was considered to be specific for the antibody used.

The severity of intracellular amyloid, extracellular amyloid, tau, and GFAP loads were quantified in the cortex and ventral hippocampus, using the following semi-quantitative scale: 0=no staining; +=slight staining; ++=moderate staining; +++=strong staining (Fig. 3). This evaluation was performed by two researchers who were blind to the age and atrophy levels of animals. Intraclass correlation coefficients for the various measures were equal to 0.66 and 0.71 for the intracellular cortical and hippocampal amyloid, 0.91 and 0.81 for the cortical and hippocampal GFAP, and 1 for the CP13 staining.

Figure 3
Anti-amyloid labeling (4G8) in the hippocampus (a-b-c-d) of mouse lemurs aged of 8.8, 6.4, 8.0, and 11.2 years, respectively. No staining (a), slight staining (b), moderate staining (c) and strong staining (d). Scale bars represent 100 μm.

3. Results

3.1. Regional location of the atrophy process

Fig. 4 displays typical MR images from one non atrophied and one atrophied mouse lemur brains (5.5 and 8.8 year old, respectively, see Table 1). Maps of CSF accumulations in regions surrounding cortical regions were calculated on the basis of these images. These maps revealed different patterns of atrophy in the animals involved in this study (Fig. 5). First, some animals did not show CSF increase in any of the cortical regions. This was the case for 100%, 54% and 25% of the young, middle-aged, and old animals, respectively. These animals were classified as stage 0 animals. In all other animals there was some degree of atrophy, which followed a typical sequential pattern that permitted the partitioning of these animals into 4 different stages of atrophy. Stage 1 animals (n=3) displayed a focal atrophy that involved only the frontal medio-dorsal and parietal (FPcort) cortical regions. Stage 2 animals (n=5) had an atrophy process that in addition to the FPcort, led to CSF accumulation in another cortical region adjacent to the FPcort. For 4 of these animals, the additional region was the parieto-temporal cortex (PTcort) (stage 2a). It was the temporal cortex (Tcort) for one animal (stage 2b). In stage 3 animals (n=1) the atrophy process involved FPcort, PTcort, and Tcort regions. Finally, in stage 4 animals (n=5), CSF increases were found all around the cortical regions evaluated, i.e. FPcort, PTcort, Tcort, and occipital cortex (Ocort) (n=5, stages 4a and 4b). One of these animals also displayed an increased CSF level around the frontal ventral (Fvcort) cortical region. Notably, four out of these five “stage 4” animals were blind. Three of these animals had bilateral mature cataract; one also had a severe bilateral keratitis whereas the last one had a severe bilateral keratitis leading to an opaque deposit on the whole cornea, its lens being no longer visible under slit lamp examination (data not shown). None of the other animals involved in the study, including one stage 4 animal, displayed detectable ophthalmologic alterations leading to blindness (ophthalmologic examination performed by a trained veterinarian by using an ophthalmoscope and a slit lamp after instillation of atropine in the eye (data not shown, but see (Beltran et al., 2007)).

Figure 4
Horizontal (a, b) and coronal (c, d) images from a 5.5 year non atrophied animal as compared to a 8.8 year atrophied animal. White CSF surrounding cortical regions is obvious in the atrophied animal (arrow).
Figure 5
The maps of CSF accumulation in regions surrounding cortical regions revealed several stages of cerebral atrophy in the studied population of animals. The number of animals displaying each stage is showed underneath the maps. Red regions correspond to ...

General linear model was used to evaluate variables explaining survival rate in non blind mouse lemurs (which died of natural death). Age at MRI was associated with survival rate in mouse lemurs (F(1, 16) = 7.74, p < 0.05). However, the atrophy stage was not associated to the survival rate (F(4, 16) = 0.62, p > 0.05). This suggests that atrophy did not induce an increased mortality in lemurs.

3.2. Correlative studies between age and CSF levels

Correlative studies between age and CSF volumes in various brain regions were performed in a second step of the study. The four stage 4 blind animals with a high atrophy level (Fig. 6a) were excluded from these correlative analyses as their CSF volumes were greater than those of the other animals. First we looked at global atrophy indexes. Significant correlations were found between the age of the animals and the total brain atrophy index (CSF volumes from all the RPES and ventricles; r=0.49, Benjamini-Hochberg false discovery rateadjusted p value (BH-FDRp) < 0.05 (Benjamini and Hochberg, 1995)) (Fig. 6a) or the total cortical atrophy index (r=0.44, BH-FDRp < 0.05). By looking at the various regions of interest, we found a significant correlation between the age of the animals and the CSF volumes in the FPcort (r=0.49; BH-FDRp < 0.05) (Fig. 6b), IHfiss (r=0.55, BH-FDRp < 0.05) (Fig. 6c), PTcort (r=0.47, BH-FDRp < 0.05) (Fig. 6d), Tcort (r=0.44, BH-FDRp < 0.05) (Fig. 6e), third ventricle (3V) (r=0.49, BH-FDRp < 0.05), and external part of the temporal horns of the lateral ventricles (ETH) (r=0.47, BH-FDRp < 0.05) but not within the Fvcort, the Ocort (Fig. 6f), or the lateral ventricles (LV) (BH-FDRps > 0.05).

Figure 6
CSF volumes in brain regions from the mouse lemurs involved in this study. The volume of CSF surrounding all the regions of the periencephalic spaces and ventricles (a), frontal (medio-dorsal) and parietal cortices (b), the interhemispheric fissure (c), ...

3.3. Neuropathological studies

Two stage 0, one stage 2 and two animals with stage 4 atrophy were included in a detailed neuropathological analysis. An overview of the histological results is presented in the table 2.

Extracellular amyloid deposits were detected in the brain of two out of the three atrophied animals (Fig. 7a,b). In all 5 animals, whatever the atrophy level considered, intracellular amyloid deposition was observed in cortical and hippocampal regions either with manual (Fig. 3) or automatic staining methods (Fig. 7e,f). In the hippocampus, the intracellular amyloid load matched with the severity of atrophy (stage 0 animals had slight amyloid levels; stage 2 animals had moderate amyloid levels; stage 4 animals had high amyloid levels (Table 2)).

Figure 7
Amyloid pathology in mouse lemurs. Extracellular amyloid deposits in mouse lemur at low magnification (a) and higher magnification (b, arrows). c and d, show amyloid deposits stained with the same conditions in humans, and transgenic mouse model of AD, ...

CP13 positive neurons were detected in the hippocampus of three animals (Fig. 8a) with CP13 staining graded + in the hippocampus of the two most severely atrophied animals (Table 2) and graded + in one stage 0 animal. No CP13 staining could be detected in the cortex of any of the animals evaluated.

Figure 8
a: CP13 positive neurons (arrows) in an aged mouse lemur primate. b-e: anti-GFAP staining in stage 0 (b, c) and 4b (d, e) animals. The corpus callosum (asterix) was strongly stained in all the studied animals (b, d). Cortical staining (arrow) was not ...

GFAP immunocytochemical detection revealed a strong staining of white matter fibers such as the corpus callosum in all studied animals (Fig. 8b,d). In addition to white matter tracks, an increased GFAP staining was also observed in cortical areas (Fig. 8d) and hippocampus (Fig. 8e, Table 2), in some animals. The hippocampal labeling was more obvious in the animals displaying the highest atrophy (Table 2).

4. Discussion

In this study, we evaluated cerebral atrophy in a cohort of 30 mouse lemur primates aged from 1.9 to 11.3 years. Atrophy was quantified by estimating CSF volumes in regions surrounding the cortex. CSF pixels were determined by using an user-independent segmentation protocol that allowed to perform 3D segmented brain maps thanks to a Mahalanobis algorithm. In a previous study, we showed that results of the segmentation protocol applied to mouse lemurs are accurate and reproducible (Dhenain et al., 2003). The segmented maps of mouse lemur brains were then used to evaluate regional atrophy. In a preliminary step, all brain regions in which CSF was expected to be present were defined on reference atlases. Then the CSF voxels from the 3D segmented maps of the various animals were manually assigned to one of the defined areas by using morphological knowledge and reference atlases. Using these assays further increased the accuracy of the segmentation procedure. Thanks to this method, evaluations of CSF volumes by various evaluators led to high intraclass correlation coefficients (from 0.97 to 1).

Evaluation of ventricular and periencephalic CSF increase in various animals indicated that some animals did not show any atrophy whereas atrophy, when present, followed a typical pattern: starting in the frontal cortex region, then involving the temporal and/or the parietal part of the brain and finally the occipital region. Such patterns of atrophy were found in 14 out of the 23 middle-aged and aged animals involved in the study (i.e. 61% of these animals). When studying the full cohort of 30 animals, a significant correlation between CSF levels and age was found in many cerebral regions. However, several middle-aged and aged animals were not atrophied. This indicates that aging by itself cannot predict atrophy in mouse lemurs and that the typical pattern of cerebral atrophy observed in some middle-aged and aged animals is part of a pathological process.

In a previous study that focused on temporal regions, we showed that cerebral atrophy can occur spontaneously with aging in mouse lemurs. We showed that the age of the beginning of the atrophy can vary in various animals and that once initiated, the atrophy process can evolve rapidly (Dhenain et al., 2000). However from that study, we could not identified whether the cerebral atrophy observed corresponded to various patterns or rather to one, typical pattern of evolution. The current study allowed us to identify a gradual pathway of atrophy suggesting that one pathological entity is responsible for most of the cerebral atrophy observed in mouse lemurs. This pathology leads to CSF accumulation first in the frontal region then in the temporal and/or parietal parts of the brain and finally the occipital region.

It is notable that we focused only on evaluation of CSF volumes. As in mouse lemurs the relationship between CSF volumes and regional gray or white matter atrophy has never been studied so far, future studies will be necessary to isolate the atrophied structures at the origin of the CSF increase in peri-cortical regions and notably to evaluate if the CSF increase is due to cortical atrophy or rather to an atrophy of subcortical structures or white matter.

Cerebral atrophy has been largely described in human cerebral pathologies or aging. Its detection in animal models of aging and/or of AD is more controversial. Age-related, MRI-detectable, severe atrophy has been rarely described in transgenic mouse models of AD and when reported is often associated with regions that are not involved in amyloid pathology such as white matter tracts (Delatour et al., 2006). Because of its severe age-related atrophy, the mouse lemur primate is an interesting model to evaluate age-associated cerebral atrophy.

At completion of the in vivo MR imaging study, we evaluated neuropathological alterations in five aged animals with stage 0, 2 and 4 atrophies. As it is often the case with mouse lemurs, the studied animals died quickly without obvious pathology (two animals lost a lot of weight during their last month of life). Tau pathology which was detected in two severely atrophied animals but also in one non atrophied mouse lemur cannot explain the cerebral atrophy. Diffuse extracellular amyloid deposits were also detected in two of the atrophied animals. However, this amyloid extracellular deposition was low (even in one stage 4 animal) as compared to what occurs in the brain of AD patients and/or transgenic mouse models of AD. Because of the lack/weakness of amyloid plaques detection in atrophied animals, our data suggest that the cerebral atrophy is not directly caused by extracellular amyloid deposition. On the other hand, our present data suggest a relationship between atrophy and intracellular amyloid deposition. Indeed, all the atrophied animals had high intracellular accumulation of β amyloid in the hippocampus. Intracellular Aβdeposition preceding extracellular deposition (Wirths et al., 2004) has been demonstrated in human AD (D'Andrea et al., 2001, Gouras et al., 2000), as well as in transgenic mouse models of AD (Casas et al., 2004, Oakley et al., 2006, Oddo et al., 2003, Wirths et al., 2001). This intracellular Aβaccumulation can lead to neuronal dysfunction, tau pathology, gliosis (Casas et al., 2004), apoptosis, and to cerebral atrophy (Christensen et al., 2008, Chui et al., 2001, Fernandez-Vizarra et al., 2004, Oakley et al., 2006, Wirths et al., 2004). Notably a recent study in a newly developed model of AD (APPT714I mouse model), characterized by intracellular accumulation of Aβ without extracellular Aβ deposits, has shown that intracellular amyloid deposition can lead to a MRI-detectable brain atrophy even in the absence of extracellular amyloid deposition (Van Broeck et al., 2008). To our knowledge, our study is the first description of a possible link between intracellular amyloid accumulation and cerebral atrophy, in non-transgenic animals that do not overexpress mutated forms of amyloid. Future studies on a larger number of mouse lemur primates will be necessary to further characterize the pathological alterations responsible for the atrophic process described here.

In conclusion, our data suggest that cerebral atrophy in mouse lemurs is mainly related to an evolving clinical entity that occurs approximately in 60 % of middle aged or aged lemurs. The pathology leads to CSF accumulation in frontal, then temporal and/or parietal part of the brain before involving in the last stage, the occipital part of the brain. Preliminary data also suggests that this pathology is associated to intracellular amyloid deposition and reactive astrocytosis. The ability to detect in-vivo by MRI the animals with pathological aging will open new perspectives for this model of normal cerebral aging either to select homogeneous populations in clinical studies or to follow-up the effect of drug treatments against pathological aging (Bons et al., 2006, Trouche et al., 2008).

Acknowledgments

Acknowledgements

We thank S. Chahory, DVM, for ophthalmologic examination of the animals involved in this study. We thank P. Davies (Albert Einstein College of Medicine, Bronx, New York) for the generous gift of the CP13 antibodies. J.F. Mangin, J. Grenèche and H. Moreira are thanked for their help during image analysis. We are grateful to E. Kremer for carefully reading and correcting the manuscript. This work was supported by the Aging ATC 2002 (INSERM), the Fédération pour la Recherche sur le Cerveau 2003, the ACI Neurosciences 2004 (French Research Department), the France-Alzheimer association, and the National Institute on Aging (R01-AG020197).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Disclosure Statement Audrey Kraska: None

Olene Dorieux: None

Jean-Luc Picq: None

Fanny Petit: None

Emmanuel Bourrin: None

Evelyne Chenu: None

Andreas Volk: None

Martine Perret: None

Philippe Hantraye: None

Nadine Mestre-Frances: None

Fabienne Aujard: None

Marc Dhenain: None

References

  • Andorfer C, Kress Y, Espinoza M, de Silva R, Tucker KL, Barde YA, Duff K, Davies P. Hyperphosphorylation and aggregation of tau in mice expressing normal human tau isoforms. J. Neurochem. 2003;86(3):582–90. [PubMed]
  • Baron JC, Chetelat G, Desgranges B, Perchey G, Landeau B, de la Sayette V, Eustache F. In vivo mapping of gray matter loss with voxel-based morphometry in mild Alzheimer's disease. Neuroimage. 2001;14(2):298–309. [PubMed]
  • Beltran WA, Vanore M, Ollivet F, Nemoz-Bertholet F, Aujard F, Clerc B, Chahory S. Ocular findings in two colonies of gray mouse lemurs (Microcebus murinus) Vet. Ophthalmol. 2007;10(1):43–9. [PubMed]
  • Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. Series B Stat. Methodol. 1995;57(1):289–300.
  • Bocti C, Rockel C, Roy P, Gao F, Black SE. Topographical patterns of lobar atrophy in frontotemporal dementia and Alzheimer's disease. Dementia Geriatr. Cogn. Disord. 2006;21(5-6):364–72. [PubMed]
  • Bons N, Mestre N, Petter A. Senile plaques and neurofibrillary changes in the brain of an aged lemurian primate, Microcebus murinus. Neurobiol. Aging. 1991;13:99–105. [PubMed]
  • Bons N, Mestre N, Ritchie K, Petter A, Podlisny M, Selkoe D. Identification of amyloid beta protein in the brain of the small, short-lived lemurian primate Microcebus murinus. Neurobiol. Aging. 1994;15(2):215–20. [PubMed]
  • Bons N, Rieger F, Prudhomme D, Fisher A, Krause KH. Microcebus murinus: a useful primate model for human cerebral aging and Alzheimer's disease? Genes Brain Behav. 2006;5(2):120–30. [PubMed]
  • Bons N, Sihol S, Barbier V, Mestre-Frances N, Albe-Fessard D. A stereotaxic atlas of the grey lesser mouse lemur brain (Microcebus murinus) Brain Res. Bull. 1998;46(1, 2):1–173. [PubMed]
  • Casas C, Sergeant N, Itier JM, Blanchard V, Wirths O, van der Kolk N, Vingtdeux V, van de Steeg E, Ret G, Canton T, Drobecq H, Clark A, Bonici B, Delacourte A, Benavides J, Schmitz C, Tremp G, Bayer TA, Benoit P, Pradier L. Massive CA1/2 neuronal loss with intraneuronal and N-terminal truncated Abeta42 accumulation in a novel Alzheimer transgenic model. Am. J. Pathol. 2004;165(4):1289–300. [PubMed]
  • Christensen DZ, Kraus SL, Flohr A, Cotel MC, Wirths O, Bayer TA. Transient intraneuronal A beta rather than extracellular plaque pathology correlates with neuron loss in the frontal cortex of APP/PS1KI mice. Acta Neuropathol. 2008;116(6):647–55. [PubMed]
  • Chui DH, Dobo E, Makifuchi T, Akiyama H, Kawakatsu S, Petit A, Checler F, Araki W, Takahashi K, Tabira T. Apoptotic neurons in Alzheimer's disease frequently show intracellular Abeta42 labeling. J. Alzheimers Dis. 2001;3(2):231–9. [PubMed]
  • D'Andrea MR, Nagele RG, Wang HY, Peterson PA, Lee DH. Evidence that neurones accumulating amyloid can undergo lysis to form amyloid plaques in Alzheimer's disease. Histopathology. 2001;38(2):120–34. [PubMed]
  • Dekaban AS, Sadowsky D. Changes in brain weights during the span of human life: relation of brain weights to body heights and body weights. Ann. Neurol. 1978;4:345–56. [PubMed]
  • Delacourte A, Sautière P-E, Wattez A, Mourton-Gilles C, Petter A, Bons N. Biochemical characterisation of tau proteins during cerebral aging of the lemurian primate Microcebus murinus. C. R. Acad. Sci. 1995;III(318):85–9. [PubMed]
  • Delatour B, Guegan M, Volk A, Dhenain M. In vivo MRI and histological evaluation of brain atrophy in APP/PS1 transgenic mice. Neurobiol. Aging. 2006;27(6):835–47. [PubMed]
  • Delatour B, Le Cudennec C, El Tannir-El Tayara N, Dhenain M. Transgenic models of Alzheimer's pathology: Success and caveat. In: Welsh EM, editor. Topics in Alzheimer's Disease. Nova Publishers; 2006. pp. 1–34.
  • Dhenain M, Chenu E, Hisley CK, Aujard F, Volk A. Regional atrophy in the brain of lissencephalic mouse lemur primates: measurement by automatic histogram-based segmentation of MR images. Magn. Reson. Med. 2003;50(5):984–92. [PubMed]
  • Dhenain M, Duyckaerts C, Michot J-L, Volk A, Picq J-L, Boller F. Cerebral T2-weighted signal decrease during aging in the mouse lemur primate reflects iron accumulation. Neurobiol. Aging. 1998;19(1):65–9. [PubMed]
  • Dhenain M, Michot J-L, Volk A, Picq J-L, Boller F. T2-weighted MRI studies of mouse lemurs: a primate model of brain aging. Neurobiol. Aging. 1997;18(5):517–21. [PubMed]
  • Dhenain M, Michot JL, Privat N, Picq JL, Boller F, Duyckaerts C, Volk A. MRI description of cerebral atrophy in mouse lemur primates. Neurobiol. Aging. 2000;21(1):81–8. [PubMed]
  • Fernandez-Vizarra P, Fernandez AP, Castro-Blanco S, Serrano J, Bentura ML, Martinez-Murillo R, Martinez A, Rodrigo J. Intra- and extracellular Abeta and PHF in clinically evaluated cases of Alzheimer's disease. Histol. Histopathol. 2004;19(3):823–44. [PubMed]
  • Gouras GK, Tsai J, Naslund J, Vincent B, Edgar M, Checler F, Greenfield JP, Haroutunian V, Buxbaum JD, Xu H, Greengard P, Relkin NR. Intraneuronal Abeta42 accumulation in human brain. Am. J. Pathol. 2000;156(1):15–20. [PubMed]
  • Le Gros Clark WE. The brain of Microcebus murinus. Proc. Zool. Soc. London. 1931:463–85.
  • Lee VM, Kenyon TK, Trojanowski JQ. Transgenic animal models of tauopathies. Biochim. Biophys. Acta. 2005;1739(2-3):251–9. [PubMed]
  • Mangin J-F, Coulon O, Frouin V. Medical Image Computing and Computer-Assisted Interventation — MICCAI'98 - Lecture Notes in Computer Science - 1496. Springer Verlag; 1998. Robust brain segmentation using histogram scale-space analysis and mathematical morphology; pp. 1230–41.
  • Mangin JF. IEEE Workshop. MMBIA, Hilton Head Island; South Carolina: 2000. Entropy minimization for automatic correction of intensity nonuniformity; pp. 162–9.
  • Oakley H, Cole SL, Logan S, Maus E, Shao P, Craft J, Guillozet-Bongaarts A, Ohno M, Disterhoft J, Van Eldik L, Berry R, Vassar R. Intraneuronal beta-amyloid aggregates, neurodegeneration, and neuron loss in transgenic mice with five familial Alzheimer's disease mutations: potential factors in amyloid plaque formation. J. Neurosci. 2006;26(40):10129–40. [PubMed]
  • Oddo S, Caccamo A, Shepherd JD, Murphy MP, Golde TE, Kayed R, Metherate R, Mattson MP, Akbari Y, LaFerla FM. Triple-transgenic model of Alzheimer's disease with plaques and tangles: intracellular Abeta and synaptic dysfunction. Neuron. 2003;39(3):409–21. [PubMed]
  • Perret M. Change in photoperiodic cycle affects life span in a prosimian primate (Microcebus murinus) J. Biol. Rhythms. 1997;12(2):136–45. [PubMed]
  • Tisserand DJ, van Boxtel MP, Pruessner JC, Hofman P, Evans AC, Jolles J. A voxel-based morphometric study to determine individual differences in gray matter density associated with age and cognitive change over time. Cereb. Cortex. 2004;14(9):966–73. [PubMed]
  • Trouche SG, Asuni A, Rouland S, Wisniewski T, Frangione B, Verdier JM, Sigurdsson EM, Mestre-Francés N. Antibody response and plasma Abeta1-40 levels in young Microcebus murinus primates immunized with Abeta1-42 and its derivatives. Vaccine. 2009;27(7):957–64. [PMC free article] [PubMed]
  • Valk J, Barkhof F, Scheltens P. Magnetic resonance in dementia; Springer-Verlag; Heidelberg Berlin New York: 2002. p. 353.
  • Van Broeck B, Vanhoutte G, Pirici D, Van Dam D, Wils H, Cuijt I, Vennekens K, Zabielski M, Michalik A, Theuns J, De Deyn PP, Van der Linden A, Van Broeckhoven C, Kumar-Singh S. Intraneuronal amyloid beta and reduced brain volume in a novel APP T714I mouse model for Alzheimer's disease. Neurobiol. Aging. 2008;29(2):241–52. [PubMed]
  • Wirths O, Multhaup G, Bayer TA. A modified beta-amyloid hypothesis: intraneuronal accumulation of the beta-amyloid peptide--the first step of a fatal cascade. J. Neurochem. 2004;91(3):513–20. [PubMed]
  • Wirths O, Multhaup G, Czech C, Blanchard V, Moussaoui S, Tremp G, Pradier L, Beyreuther K, Bayer TA. Intraneuronal Abeta accumulation precedes plaque formation in beta-amyloid precursor protein and presenilin-1 double-transgenic mice. Neurosci. Lett. 2001;306(1-2):116–20. [PubMed]