Animals

Experiments were performed in adult male Wistar rats, weighing 300–320 g at the beginning of the study. Rats were housed in cages under controlled temperature (21 ± 1°C) and humidity (55 ± 10%), with a 12-h light/12-h dark cycle (light between 8:00 AM and 8:00 PM). Food and water were available ad libitum during all experiments. Animal work was performed following the local legislation (Decret 214/1997 of July 30th by the ‘Departament d’Agricultura, Ramaderia i Pesca de la Generalitat de Catalunya’) under the approval of the Ethical Committee of the University of Barcelona (CEEA), and in compliance with European legislation.

Surgery

Rats were anaesthetized under 4% isofluorane in a mixture of 30% O₂ and 70% N₂O, and then maintained under 1.5% isofluorane. They were allowed to breath spontaneously throughout surgery. Body temperature was monitored using a rectal probe and maintained between 36.5°C and 37.5°C. Animals were randomly selected for BCCAo (n=17) or sham (n=13) surgery. A ventral midline incision was made to expose the common carotid arteries, which were carefully separated from the vagus nerves. Both common carotid arteries were tied off with two 3-0 sutures (Suturas Aragó, Barcelona, Spain) before the bifurcation of the internal and the external carotids. The first carotid to be occluded, either right or left, was alternated throughout the experiment. Sham animals underwent the same surgical procedure except for the occlusion of the two carotids.

Magnetic Resonance Imaging

MRI experiments were conducted on a 7.0 T BioSpec 70/30 horizontal animal scanner (Bruker BioSpin, Ettlingen, Germany), equipped with an actively shielded gradient system (400 mT/m, 12 cm inner diameter). The receiver coil was a 4-channel phased-array surface coil for the rat brain. Animals were placed in supine position in a Plexiglas holder with a nose cone for administering anaesthetic gases (1.5% isofluorane in a mixture of 30% O₂ and 70% CO₂) and were fixed using a tooth bar, ear bars and adhesive tape. Tripilot scans were used to ensure accurate positioning of the head in the isocenter of the magnet.

For the quantification of the vascular remodeling time-course, time-of-flight (TOF) angiography with a three-dimensional (3D) Fast Low Angle Shot (FLASH) method was acquired before the surgery (pre-occlusion) and 2 h, 24 h, 10 days and 3, 7 and 12 weeks after BCCAo (n=10) or sham-operation (n=6). The scan parameters were as follows: echo time (TE) = 2.5 ms, repetition time (TR) = 15 ms, field of view (FOV) = 35 x 35 x 50 mm, matrix size = 256x 256 x 128 pixels, resulting in a spatial resolution of 0.137 x 0.137 x 0.39 mm.

At the same time points, brain lesions were evaluated longitudinally by T2 mapping of coronal slices acquired with a multislice-multi-echo (MSME) sequence by applying 16 TEs, from 11 to 176 ms, TR = 4764 ms, slice thickness = 1 mm, number of slices = 18, FOV = 40 x 40 mm, and matrix size = 256x 256 pixels, resulting in a spatial resolution of 0.156 x 0.156 mm in 1.00 mm slice thickness.

For registration purposes and longitudinal evaluation of other structural alterations that were not observable in T2 relaxometry maps, high resolution 3D Modified Driven Equilibrium Fourier Transform (MDEFT) images were acquired at the same time points, as above. The scan parameters were as follows: TE = 3.5 ms, TR = 4000 ms, 8 segments and the same geometry as T2 maps except for slice thickness, which was 0.5 mm.

To study microsctructural changes, Diffusion Tensor Imaging (DTI) was performed 7 weeks after BCCAo or sham surgery, using an echo planar imaging DTI sequence with TR = 14500 ms, TE = 30.85 ms, four segments, b-value = 1000, 126 diffusion directions, five B0 images, FOV = 22.23 x 22.23 x 17.92 mm, matrix size = 72x 72 x 58 pixels, resulting in an isometric spatial resolution of 0.309 x 0.309 x 0.309 mm and an acquisition time of 2 h 6 min.

For cerebral perfusion measurements 12 weeks after BCCAo, a T2 * sensitive version of the Echo Planar Imaging (EPI) sequence was used, TE = 18.7 ms, TR = 160 ms, slice thickness = 1 mm, number of slices = 18, FOV = 30 x 30 mm, matrix size = 234x 256 pixels, resulting in a spatial resolution of 0.234 x 0.234 mm in 2.00 mm slice thickness, and 2100 repetitions were recorded consecutively within a total acquisition time of 5 min 36 s. Thus, the recording time per volume was 0.16 s. Gadolinium bolus track experiments were performed in 3 coronal planes. The first 2 min (750 volumes) were used as baseline, then a bolus of 0.5 mmol/kg of gadodiamide (Omniscan™, GE Healthcare Bio-Sciences, S.A., Madrid, Spain) was injected into the femoral vein over a 20-s interval (125 volumes), and the experiment finished with an additional 3 min 36 s of recording (1350 volumes).

A scheme of the timeline for the whole experimental design is shown in Figure S1.

Image processing

The images resulting from TOF angiography were subjected to a post-processing procedure performed with specific software for scientific image processing and visualization (Amira, Visage Image, Inc). The procedure consisted of several steps, namely signal intensity correction (for the signal decay induced by the surface coil), brain masking, skull stripping, background removal, semi-automatic segmentation of the vertebro-basilar artery, right and left middle cerebral arteries, and azygos-pericallosal artery and, finally, skeletonization of the segmented arteries. The images acquired after BCCAo or sham-operation were affine-registered to their respective pre-occlusion scan so that the same brain mask and artery segmentation could be applied to all time points with minor manual editing. Subsequently, arterial length and arterial tortuosity were measured from the artery skeleton. For vertebro-basilar arteries, the ratio between the hypothetical minimum length and the true length was expressed as the arterial tortuosity. A third parameter related to vascular remodeling was calculated from the Maximal Intensity Projection (MIP) of the cropped base of the brain (Figure S2). The ratio between the area occupied by the projected arteries and the area of the brain projection were calculated for all acquired time points in sham and BCCAo animals.

T2 maps and MDEFT images were reconstructed with Paravision 5.0 software (Bruker Biospin, Etlingen, Germany) and custom-made programs written in Matlab (The MathWorks, Inc., Natick, MA, USA). The volume of focal lesions was estimated in the T2 maps after manual delineation of the region showing increased T2 values (range used: 20-150 ms). The areas were then integrated to calculate the volume. Moreover, the presence and number of hyperintense areas in the striatum and the cortex on MDEFT images were determined for each animal at each acquired time point. All analyses were performed blinded to the treatment and by the same experimenter in order to avoid inter-individual differences.

DTI images were studied using a voxel based analysis (VBA) approach as previously described in [14]. Briefly, all rat brains were registered to a reference brain by means of an affine registration that maximized mutual information of fractional anisotropy (FA) volume followed by an elastic warping based on diffeomorphic demons [15] available in MedINRIA 1.9.4 software. Registered volumes were smoothed with a Gaussian kernel of 3 x 3 x 3 voxels with a standard deviation of one voxel to compensate for possible misregistrations. Voxel-wise statistical test (Mann-Whitney U test) was performed, obtaining the voxels with a statistically significant different distribution of diffusion- related parameters between sham and BCCAo animals. Note that the results may be biased by this choice since VBA requires the definition of a reference brain. In order to avoid such bias and increase the reliability of the results, the VBA procedure was repeated taking all animals as template, and only the regions where differences appeared consistently among the different templates were considered. In this way, the variability produced by the arbitrary choice of the reference template was discarded.

For cerebral perfusion imaging, three regions of interest (ROI) were drawn in each hemisphere from T2*-EPI images (medial prefrontal cortex, caudate-putamen and retrosplenial cortex, Figure S3). The time-course of the signal intensity was calculated for each pixel, and the average values were calculated for each ROI. Data were processed with custom-made programs written in Matlab as previously described [16] with minor modifications. Briefly, in order to obtain relative brain perfusion parameters, we calculated the following: the maximum of the signal intensity curve; the time interval from reaching the gadolinium to the moment of the maximum concentration, or time to peak (TTP); the area under the curve (AUC), which provides an estimation of the relative cerebral blood volume (relCBV); and the estimated mean transit time, calculated as the full width of the curve at half maximum (FWHM). Relative CBF (relCBF) was obtained from the ratio relCBV/FWHM.

Reversal learning and set-shifting behavioral tasks

A separate group of animals was used for behavioral testing (sham n= 7, BCCAo n=7).

Tissue preparation, histochemical and immunohistochemical studies

After the last MRI acquisition (12 weeks), animals were anesthetized with isofluorane and transcardially perfused with saline followed by 4% paraformaldehyde in 0.1M phosphate-buffered saline (PBS) (pH=7.4). Afterwards, the brains were removed and the basilar (BA) and middle cerebral (MCA) arteries were isolated. Brains were postfixed overnight in the same fixative at 4 °C, dehydrated, and embedded in paraffin. The brains were cut (5 µm) in coronal slices at different levels from Bregma (Bregma +3.7, +0.2, and -3.14) following the atlas of Paxinos and Watson (1996) and using a microtome (Leica, Microsystems, Wetzelar, Germany). Cellular damage was assessed by hematoxylin and eosin (HE) staining. Klüver-Barrera Luxol fast blue (Luxol) staining was used to evaluate myelin damage. Immunohistochemistry against glial fibrillary acidic protein (GFAP, a specific marker of astroglial cells) and against Iba-1 (a specific marker of reactive microglia/macrophages) was carried out to reveal signs of glial reaction. The rabbit polyclonal antibodies against GFAP and Iba-1 (Dako, Dakopats, USA) were diluted at 1:500 in PBS containing 0.3% Triton-X100 and 0.2% gelatin. The immunocytochemical procedure has been described elsewhere [20]. Briefly, sections were rehydrated and incubated overnight with the corresponding primary antibody in a humidified chamber at 4 °C and then incubated for 1 h at room temperature with the appropriate biotinylated second antibodies, diluted 1:200 (goat anti-rabbit IgG) (Vector Laboratories, USA). All sections were incubated with avidin-biotin-horseradish peroxidase complex 1:100 (ABC kit, Vectastain, Vector Laboratories) for 1 h. Peroxidase labeling was visualized with a solution containing 0.025% diaminobenzidine (Sigma Aldrich) and 0.03% H₂O₂ (Sigma Aldrich) in PBS. In each experiment, a tissue section was also processed in parallel without the primary antibody as a control for non-specific staining.

For quantitative analysis, images of the corpus callosum (CC) and of the optic tract were taken though an optical microscope with a x 40 objective connected to a digital camera, and the analyses were performed using NIH ImageJ software. Myelin fiber density was determined using a threshold that was fixed for all images, and values were expressed as percent density of myelin fibers per field. The number of Iba1-immunopositive cells was counted in fields of 0.075 mm². The quantification was performed in at least 3 representative fields per region and rat, and the numbers were averaged and the values expressed as the mean ± SEM.

The dissected arteries were post-fixed with 4% PFA for 45 min and washed in three changes of PBS. After cleaning, vessels were placed overnight in PBS containing 30% sucrose. They were then transferred to a cryomold (Bayer Química Farmacéutica, Barcelona, Spain) containing Tissue Tek OCT embedding medium (Sakura Finetek, Europe, The Netherlands) for 20 min and frozen in liquid nitrogen. Frozen transverse sections (14µm) of BA and MCA were cut onto gelatin-coated slides and air-dried for at least 90 min. After blockade, sections were incubated with a rabbit polyclonal antibody against collagen I/III (1:30; Calbiochem, Pacific Center Court, San Diego, CA, USA) in PBS containing 2% bovine serum albumin (BSA) for 1 h at 37°C in a humidified chamber. After washing, rings were incubated with the secondary antibody, a donkey anti-rabbit IgG conjugated to CyTM3 (1:200; Jackson Immunoresearch Laboratories Inc., West Grove, PA, USA) for 1 h at 37°C in a humid chamber. After washing, sections were stained with the nuclear dye Hoechst 33342 (0.01 mg/ml; Sigma Aldrich) for 15 min followed by 3 washes in PBS. Immunofluorescent signals were viewed using an inverted Leica TCS SP2 confocal laser scanning microscope with an oil immersion lens (x20). CyTM3-labeled antibody was visualized by excitation at 568 nm and detection at 600-700 nm. The specificity of collagen I/III immunostaining was evaluated by omission of the primary antibody and processed as above. Under these conditions, no staining was observed in the vessel wall in any experimental condition. Quantitative analyses of collagen I/III fluorescence and nuclei number were performed with MetaMorph 4.6 Image Analysis Software (Universal Imaging, Molecular Devices, Downingtown, PA, USA). The intensity of fluorescence per area was calculated in three rings of each animal, and the results were expressed in arbitrary units. To evaluate the number of smooth muscle cells/ring the autofluorescence of the elastic laminae was used to identify the media. In addition, the nuclear shape of the different cell types, visualized by Hoesch 33342, confirms the presence of smooth muscle cells in the media (see Figure S4). The number of nuclei was counted in three rings of each animal, and the results were expressed as cell density. All measurements were conducted in a blind fashion.

Vessel morphometry

Morphometric assessment of the lumen, media and vessel areas was performed in coronal sections of the dissected arteries using MetaMorph 4.6 Image Analysis Software, as previously described [21]. To determine the luminal area, the cross-sectional area enclosed by the internal elastic lamina was corrected to a circle by applying the form factor l2/4π to the measurement of the internal elastic lamina, where l is the length of the lamina. The vessel area was determined by the cross-sectional area enclosed by the external elastic lamina corrected to a circle by applying the same form factor (l2/4π) to the measurement of the external elastic lamina. Internal and external diameters were calculated from luminal and vessel areas, respectively. Structural parameters were calculated as follows: cross-sectional area (CSA) = le – li; wall thickness (WT) = (De - Di)/2 and wall/lumen (W/L) ratio = WT/Di, where le is the vessel area, li the luminal area, De the external diameter, and Di the internal diameter. Mean values were obtained from three sections of each dissected artery from BCCAo (n=6) and sham (n=5) animals.

Statistical analysis

The length and tortuosity of the arteries and vascular network quantification were analyzed using a two-way analysis of variance (ANOVA), with treatment (BCCAo- or sham-operated animals) as the ‘between’ factor and time (days) as the ‘within-subject’ factor. When significant overall interactions were found, further analyses of partial interactions were carried out. Post-hoc analyses for intergroup comparisons were performed with Fisher’s least significant differences (LSD) test when the initial p value was significant. All data were analyzed using Statistica software (StatSoft Inc., France). Differences were considered significant when P< 0.05. All results are expressed as mean ± SEM. Data from arteries and brain histology and immunohistochemistry, and gadolinium bolus track measurements and behavior were analyzed using an unpaired Student’s t-test and two-way ANOVA, respectively. In all cases, data analysis was carried out using GraphPad Prism 4 Software. A value of P< 0.05 was considered significant.

Tissue alterations assessed with MRI

The presence of acute ischemic focuses was evaluated with T2 relaxometry mapping 24 h after BCCAo or sham-operation. In 70% of the rats, BCCAo induced unilateral focal acute damage manifested as areas of increased T2 values located in the striatum with widely variable sizes ranging from 1.7 mm³ to 27.13 mm³. Acute lesions tended to be more frequent in the right (6 out of 10) than in the left (3 out of 10) striatum, but differences were not statistically significant (Chi-square test = 1.82, P> 0.05). Figure S5 shows an incidence map of all ischemic lesions observed 24 h after BCCAo. The same lesions were observed as hypointense zones in the 3D MDEFT T1-weighted images. Coronal sections of two representative BCCAo animals illustrating the largest and smallest tissue lesion are shown in Figure 1A and 1B, respectively, at several time points and with the MRI modalities used in this study. In the MDEFT images, acute damage was observed as a hypointense zone while adjacent areas and secondary injuries were seen as hyperintense zones (Figure 1). The size of the acute MDEFT hypointense areas decreased with time and was progressively substituted by hyperintense signals that surrounded the hypointense core. MDEFT provided a contrast superior than that obtained in T2-weighted images and maps for the progressive evolution of the tissue lesion.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Longitudinal multimodal MRI.

Representative coronal section of two BCCAO animals (A, B) are shown at different time points. BCCAO induced variable degrees of focal acute ischemic lesions (see right striatum at 24 h in A and B) and delayed damage (see left striatum from day 10). For both animals (A, B), images in first row are T2-weighted (T2W), second row are T2 relaxometry map (T2R) and third row are 3D MDEFT T1-weighted (MDEFT) images. The box highlights MDEFT alterations at 12 weeks in the same animal shown in Figure 3 for histological damage. Color scale represents T2 relaxometry time (ms) for images shown in second row of panels A and S (T2R).

https://doi.org/10.1371/journal.pone.0074631.g001

Bilateral striatal increases in T2 values were not observed 24 h after BCCAo; however, in 20% of the rats, bilateral alterations became apparent at day 10. These delayed MRI alterations were manifested as hyperintensities in MDEFT from day 10 onwards, corresponding with decreased T2 values in regions showing no apparent T2 alterations at 24 h (see left striatum at 24 h in upper and middle row of Figure 1A). These lesions were therefore more delayed than the acute ischemic lesions. In addition, they showed distinct MRI features to those observed in the acute ischemic zones and were found in subcortical regions (striatum).

DTI identified alterations in multiple grey and white matter areas (Figure 2). Indeed, VBA showed a significant decrease in FA values in BCCAo compared to sham animals in structures such as the piriform, frontal and insular cortical regions, hippocampus, thalamus, and optical nerve and tract, while significant increase in FA values was observed in the lateral striatum (Figure 2, FA). Mean and radial diffusivities (Figure 2, MD and RD) were significantly reduced in the cingulum and increased in the lateral striatum of BCCAo versus sham rats. Furthermore, axial diffusivity (AD) was significantly increased in the striatum of BCCAo compared to sham rats, possibly as a result of the multiple focal lesions previously described in some animals (Figure 2, AD). The means of the four DTI indexes for the above-described areas are shown in Table S3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Statistical maps showing Diffusion Tensor Imaging (DTI) indexes.

Regions showing statistically significant differences of fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD) and radial diffusivity (RD), between sham and BCCAo animals 7 weeks after surgery. Background images are the corresponding averaged DTI index calculated for the sham group (left: coronal, middle: sagittal, and right: axial slices). Red areas (sham > BCCAo) have a significance of P <0.05 (Mann-Whitney U test). Blue areas (sham < BCCAo) have a significance of P <0.05 (Mann-Whitney U test). The slices show representative anatomical structures (see Results section for details).

https://doi.org/10.1371/journal.pone.0074631.g002

Immunohistochemical evaluation of brain tissue alterations

Relation of MRI changes and histological damage.

Necrotic zones (hematoxylin-eosin) with activated microglia/macrophages (Iba-1+) surrounded by an astroglial scar (GFAP+) (Figure 3A-D) were observed at 12 weeks in the striatum of all BCCAo rats previously showing an acute increase of T2 values (hypointensities in MDEFT images). Histological alterations were not detected in sham-operated rats. The anatomical distribution of the areas showing glial reaction matched those affected in MRI. In an attempt to relate hyper- and hypo-intensities in MRI images with the type of histological changes, we show the same rat in Figure 1A and Figure 3. The zone that was hypointense in the right striatum in MDEFT (Figure 1A) at 12 weeks (time of postmortem histological study) corresponds to a necrotic core that was surrounded by a strong glial reaction (Figure 3C-D, I-L) manifested as a hyperintense ring in MDEFT images (decreased T2 values in relaxometry maps) (e.g. right striatum in Figure 1A highlighted in the box). In addition to the acute MRI alterations, delayed MRI changes manifested as hyperintense signals in 3D MDEFT T1 images (e.g. left striatum in Figure 1A in the section highlighted in the box) (decreased T2 values in relaxometry maps) in the absence of acute T2 alterations were seen in some animals. These MRI changes corresponded to zones of abundant gliosis at 12 weeks. The presence of gliosis in zones lacking signs of acute infarction is compatible with progressive neuronal death and/or axonal degeneration subsequent to insufficient blood supply after BCCAo.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Histological evaluation of brain lesions 12 weeks after surgery.

A-D) Macroscopic images of coronal brain sections (at the level of Bregma +0.2) of a BCCAo rat showing bilateral alterations in the striatum. This animal corresponds to the same rat shown in the MRI image in Figure 1A. Sections show tissue structure (hematoxylin-eosin, HE) (A), myelinated fiber tracts (B), and astroglial (GFAP) (C) and microglial (Iba1) (D) reactions. Corresponding magnifications of the above images are shown in the middle row (C-L) for the areas illustrated with rectangles in the image (A). Alterations (arrowheads) are apparent in both hemispheres compared to the tissue of a sham-operated rat (M-P). Arrows in M-P point to healthy structures and cells. The features of the alterations in the left and right hemisphere of this BCCAo rat differ, with more prominent changes in the right striatum, thus illustrating the variability in the extent and severity of the damage after BCCAo. Q) Luxol staining of a brain section (Bregma -2.3) indicating with a rectangle the genus region of the corpus callosum shown in higher magnification in (R) and (S) for a sham-operated and BCCAo rat, respectively. BCCAo induces rarefaction with pallor of the fiber staining and vacuolization (arrowheads). T) Quantification of fiber density shows a significant reduction in the genus of the corpus callosum 12 weeks after BCCAo. Data are represented as mean ± SEM. ** P< 0.01 (Unpaired Student’s t-test). Scale bar a-d, q = 0.5 cm; e-p, r, s = 50 µm.

https://doi.org/10.1371/journal.pone.0074631.g003

Optic tract degeneration.

Twelve weeks after surgery, all BCCAo rats showed macroscopic signs of optic nerve damage with varying degrees of degeneration affecting the total or partial length of the tract. Figure 4A shows macroscopic images of representative perfused brains (ventral view) of sham-operated rats showing preserved optic tracts and of BCCAo rats showing complete and partial degeneration of the optic nerves before and after the optic chiasm. At the microscopic level, the optic tract of all animals that underwent BCCAo showed reduced fiber density and vacuolization (Figure 4B), astrogliosis (not shown), and microgliosis (Figure 4B). The quantitative analyses showed a statistically significant reduction in fiber density and increased number of microglia cells after BCCAo (P< 0.01) (Figure 4C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Signs of optic tract degeneration 12 weeks after surgery.

A) Macroscopic images of the optic tract (arrows) show reduce fiber volume in BCCAo rats compared to sham-operated rats. B) Histological evaluation of fiber density assed with Luxol staining of the optic tract (at the level of Bregma -2.3) and immunostaining to show microglia reactivity (Iba-1). C) Quantification of fiber tract density and number of Iba-1+ cells per field (objective x40) show significant increases in BCCAo rats (n=5) versus sham rats (n=3). Data are represented as mean ± SEM. ** P< 0.01 (Unpaired Student’s t-test). D) Reversal learning executive function as expressed by trials performed correctly to acquire each reversal test, white and black bars represent sham (n=4) and BCCAo animals (n=7) respectively. C) Set-shifting executive function as expressed by correct and incorrect responses during the 6 testing weeks in sham (n=4) and BCCAo animals (n=7). Asterisks indicate significant differences between correct and incorrect responses (Bonferroni test). ** P< 0.01.

https://doi.org/10.1371/journal.pone.0074631.g004

Behavioral alterations

Both sham and BCCAo groups performed the reversal learning task in a similar manner. There was no significant difference in the ability of the two groups to remember the previously learned discrimination at any time point (data not shown). Nor was there a difference in the number of trials required to reach the reversal learning acquisition criterion at any time tested (F_2,24 = 0.6718; ns) (Figure 4D). However, regarding the shift of animals from an egocentric spatial response strategy to a visual-cue discrimination strategy to obtain food reward (set-shifting task), the BCCAo rats were unable to achieve the acquisition criterion (Figure 4C). Two-way ANOVA showed time effect: F_5,40 = 6.40; P< 0.001, surgery effect: F_3,40 = 40.75; P< 0.01, and a significant interaction (F_5,40 = 33.87; P< 0.001). Indeed, after 4 weeks of set-shifting training sessions, sham animals started to give more correct than incorrect responses, and the number of correct responses was significantly different after 6 weeks (P< 0.01). Significant differences between groups were also observed 6 weeks after the beginning of the set-shifting training (P< 0.01).

MRI longitudinal quantification of arteriogenesis after BCCAo

The temporal pattern of arterial changes after BCCAo was quantified in the main craniocerebral arteries: 1) the vertebrobasilar system, including both vertebral arteries, from its insertion at the dura mater, until the most anterior end of the basilar artery, right at its bifurcation with the posterior cerebral arteries; 2) the left and right middle cerebral arteries; and 3) the azygos-pericallosal artery, which corresponds to the intracerebral segment of the unified anterior cerebral arteries [22]. Notably, the length and tortuosity of the vertebrobasilar system were strongly altered after BCCAo (Fig. 5A). The results of the statistical analyses (two-way ANOVA) are shown in Table S1 for each artery. Significant effects of surgery (BCCAo vs. sham-operation or vs. pre-occlusion data from the same rats) were observed for both the length and tortuosity of the vertebrobasilar artery. Compared to pre-occlusion data, a significant increase in the vertebrobasilar length was found from day 23 after BCCAo (P< 0.01) (Fig. 5B). Compared to sham-operated rats, significant differences were also found at 7 and 12 weeks after BCCAo (P< 0.01). Similar results were observed for the increase in vertebrobasilar artery tortuosity (Fig. 5C). However, BCCAo did not affect the length or tortuosity of the middle cerebral and the azygos-pericallosal arteries (Figure S6). These results support the notion that the obstruction of blood flow in the carotid arteries promotes compensatory adaptive changes in the vertebrobasilar system, in agreement with previous reports [6,7]. Using MRI arteriography, here we demonstrate that arterial changes were already present one week after BCCAo, they persisted for at least 12 weeks, and they were most abundant in the vertebrobasilar tree. Significant differences were not detected in arterial branches of the internal carotid tree, although we cannot exclude small changes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Time course changes in vertebro-basilar artery length and tortuosity following BCCAo in rats.

A) Segmentation of the vertebro-basilar artery for representative sham and BCCAo rats for all the time points evaluated. The ratio between the hypothetical minimum length and the true length was expressed as the arterial tortuosity (bottom right picture) (B, C) Measurements of length and tortuosity of the vertebro-basilar arteries (see Methods). The minimal distance from the base of the basilar artery to the final points evaluated is schematically drawn (in green) on the bottom right image of panel a. Vertebro-basilar artery length (B) and tortuosity (C) progressively increased after BCCAo from day 10. A measure of the effect size was determined by calculating the eta squared (η2), which for the treatment effect over the length was 0.077 and 0.242 over the tortuosity, suggesting 7% of the effect observed in length and 24% of the effect observed in the tortuosity accounts for BCCA occlusion. White and black circles represent mean±SEM for sham (n=4) and BCCAo (n=6) groups, respectively. * P< 0.05, ** P< 0.01.

https://doi.org/10.1371/journal.pone.0074631.g005

In addition, after BCCAo, the presence of new arteries was detected, especially at the base of the brain (Figure 6A). The temporal characterization of the arteriovascular area revealed a significant effect of BCCAo versus sham-operation (or versus pre-occlusion data) that was time-dependent two-way ANOVA, time effect: F_5,40 = 16.160; P< 0.001, surgery effect: F_1,8 = 32.749; P< 0.001, and a significant interaction (F_5,40 = 16.295; P< 0.001). Subsequent post-hoc analyses revealed significant differences between the two experimental groups at all time points (P< 0.01 for all cases), and between pre-occlusion data and certain time points (10 days, 23 days, 7 weeks and 12 weeks after BCCAo) (Figure 6B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 6. Total arterial area progressively increases after BCCAo.

A) Maximal intensity projection (MIP) of the brain basal segment and arteries from representative sham (n = 4) and BCCAO (n = 6) rats at all the time points evaluated. B) Ratio between the brain surface and the vascular surface of the previous MIP calculated at all the time points evaluated. Data are represented as mean ± SEM. Black stars indicate significant differences versus the pre-occlusion data (Fisher’s LSD test), and white stars significant differences versus the sham-operated animals at corresponding time points. ★ P< 0.05; ★★ P< 0.01.

https://doi.org/10.1371/journal.pone.0074631.g006

Structural alterations of the basilar artery 12 weeks after BCCAo

The post-mortem morphometric study of the basilar artery (BA) 12 weeks after surgery showed a significant increase (P< 0.001) in vessel (sham: 1147.25 ± 90.42 µm, n=5; BCCAo: 2177.8 ± 102, n=6) and lumen (sham: 942.16 ± 78.78, n=5; BCCAo: 1929.9 ± 92.5 n=6) diameters, which lead to a greater (P< 0.001) cross-sectional area (CSA; Figure 7B) and a smaller (P< 0.05) wall/lumen ratio (sham: 95.9±5.01 µm, n=5; BCCAo: 123.98±6.8 µm; n=6). These changes were paralleled by an increase (P< 0.001) in the total number of smooth muscle cells per ring (Figure 7C). This observation suggests smooth muscle cell proliferation and is compatible with circumferential wall stress caused by artery occlusion [23]. Moreover, these changes were accompanied by a decrease (P< 0.05) in collagen I/III fluorescence intensity/µm² in BCCAo rats (35.1 ± 6.3, n=6 vs. 97.48 ± 16.9, n=5 in sham-operated respectively). Taken together, these findings are compatible with outward hypertrophic remodeling [24] of the BA after BCCAo. However, differences in the MCA were not statistically significant (see Figure S7).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 7. Structural evaluation of the basilar artery (BA), 12 weeks after BCCAo.

A) Representative coronal sections of the BA from a sham and a BCCAo rat. Red shows collagen immunostaining and blue shows the cell nuclei. B) Cross-sectional area and C) total number of smooth muscle cells (SMC)/ring calculated for the BA of sham (n = 5) and BCCAo (n = 6) group 12 weeks after surgery. Data are represented as mean ± SEM. Asterisks indicate significant differences between the 2 groups. ★★ P< 0.01, ★★★ P< 0.001 (Unpaired Student’s t-test).

https://doi.org/10.1371/journal.pone.0074631.g007

Dynamic susceptibility contrast imaging for evaluation of cerebral perfusion

Measures of relative CBF (relCBF) (Figure 8B) that were carried out by dynamic susceptibility contrast imaging demonstrated that cerebral hypoperfusion was not present 12 weeks after BCCAo. On the contrary, a subtle cerebral hyperperfusion was apparent at this time in the BCCAo group, since two-way ANOVA by type of surgery and brain region showed a significant effect of surgery (F_54,1=14.97; P< 0.01), whereas no regional effect (F_54,2=5.91; n.s) or interaction between these two factors (F_54,2=1.82; n.s) was observed (Figure 8B). The statistical analysis for maximum gadolinium-induced signal enhancement, TTP, relCBV (AUC) and FWHM are shown in Table S2. Briefly, significant effects of BCCAo were demonstrated in various structures for maximum gadolinium-induced signal enhancement, TTP and FWHM (respectively Figure 8C, D and F). Taken together, our results show that the hypoperfusion induced by BCCAo was transient since it was no longer observed 12 weeks after surgery.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 8. Dynamic susceptibility contrast imaging for evaluation of cerebral perfusion.

A) Scheme of the gadodiamide (Gd) bolus track and the evaluated perfusion-related parameters. The maximum (max) is the peak of signal intensity. Time to peak (TTP) is the period of time from gadodiamide injection to the peak of signal intensity. The area under the curve (AUC) provides an estimation of relative cerebral blood volume (relCBV). B) Relative cerebral blood flow (relCBF); C) maximum of signal intensity; D) time to peak; E) relative CBV (relCBV); F) full width at high maximum (FWHM) as an estimation of the mean transit time in several brain regions. White bars represent sham animals (n = 3) and black bars BCCAO animals (n = 7), 12 weeks after surgery. Black stars indicate significant differences between groups (Bonferroni test). ★ P< 0.05 CP: caudate-putamen; pCx: medial prefrontal cortex; rCx: retrosplenial cortex.

https://doi.org/10.1371/journal.pone.0074631.g008