Long-term monitoring of chronic demyelination and remyelination in a rat ischemic stroke model using macromolecular proton fraction mapping

Abstract

Remyelination is a key process enabling post-stroke brain tissue recovery and plasticity. This study aimed to explore the feasibility of demyelination and remyelination monitoring in experimental stroke from the acute to chronic stage using an emerging myelin imaging biomarker, macromolecular proton fraction (MPF). After stroke induction by transient middle cerebral artery occlusion, rats underwent repeated MRI examinations during 85 days after surgery with histological endpoints for the animal subgroups on the 7th, 21st, 56th, and 85th days. MPF maps revealed two sub-regions within the infarct characterized by distinct temporal profiles exhibiting either a persistent decrease by 30%–40% or a transient decrease followed by return to nearly normal values after one month of observation. Myelin histology confirmed that these sub-regions had nearly similar extent of demyelination in the sub-acute phase and then demonstrated either chronic demyelination or remyelination. The remyelination zones also exhibited active axonal regrowth, reconstitution of compact fiber bundles, and proliferation of neuronal and oligodendroglial precursors. The demyelination zones showed more extensive astrogliosis from the 21st day endpoint. Both sub-regions had substantially depleted neuronal population over all endpoints. These results histologically validate MPF mapping as a novel approach for quantitative assessment of myelin damage and repair in ischemic stroke.

Keywords

Macromolecular proton fraction middle cerebral artery occlusion myelin magnetic resonance imaging histology

Introduction

Stroke is one of the most common causes of chronic disability and long-term physical and cognitive impairment that require extensive rehabilitation and support in daily routine.¹ Sudden interruption of cerebral blood flow causes a number of pathological events in a relevant brain territory including loss of neurons, cerebral edema, axonal death, demyelination, neuroinflammation, and formation of the glial scar. As a result, partial destruction of the neural network is manifested as functional impairment. At the same time, the nervous system has the potential to restore itself and compensate for the lost functions through spontaneous recovery due to neuroplasticity, axonal and synaptic remodeling, and regeneration of neural networks.^2–5

The mechanisms underlying spontaneous recovery after stroke is an area of extensive research. Human studies showed that early recovery is caused by edema reduction⁶ and penumbra reperfusion,⁷ whereas late recovery is associated with functional and structural brain plasticity.^6,8 Magnetic resonance tractography in patients showed visible reconstruction of fiber tracts connected to the lesion areas in 3 months after stroke.⁹ A number of MRI and histological studies demonstrated axonal regrowth¹⁰ and reorganization of neural networks and connectome¹¹ after experimental stroke in animal models.^11–14

The development of myelin in the post-ischemic brain may serve as an important indicator of functional reorganization or restoration of neural tissue after stroke. Reorganization of neural networks leads to changes in neuronal activity, which, in turn, may affects the thickness of the myelin sheath.¹⁵ Axons that have lost the myelin sheath can be remyelinated.¹⁶ Remodeling of damaged neural networks may lead to formation of new synaptic contacts and de novo axonal growth and myelination.^17,18 Therefore, accurate quantification of myelination is crucial for monitoring and predicting the effect of rehabilitation and neurorestorative therapies in stroke.

To date, there have been no conventional non-invasive imaging methods enabling quantitative assessment of remyelination in stroke. One previous study¹⁹ has demonstrated using the rat middle cerebral artery occlusion (MCAO) model that a recently proposed quantitative MRI method, fast macromolecular proton fraction (MPF) mapping,^20,21 enables accurate assessment of myelin damage in acute and sub-acute ischemic stroke, being in close agreement with histology. Earlier, MPF has demonstrated strong correlations with the myelin content in the normal rat and murine brain^22,23 and in the model of murine cuprizone-induced demyelination with subsequent remyelination.^23,24 Besides, pilot clinical studies of multiple sclerosis^25,26 and mild traumatic brain injury²⁷ have shown the feasibility of using MPF mapping for quantification of demyelination in humans.

The objective of this study was to evaluate the feasibility of quantitative assessment of demyelination and remyelination in ischemic stroke from the acute to late chronic stage using the fast MPF mapping method. To achieve this goal we performed long-term monitoring of the ischemic lesion evolution in the rat transient MCAO model using MPF maps and conventional MRI techniques with histological validation for a series of temporal endpoints.

Material and methods

Animals

Nine adult male Wistar rats (age 8 weeks, weight 240–270 g at the study entry) were used in this study. Normal conditions with a 12-hour dark–light cycle at a temperature of a 24 ± 2°C were maintained in the vivarium during the study. Food and water were provided ad libitum. All animal experiments were carried out under approval of the Bioethical Committee at the Pirogov Russian National Research Medical University and followed the National Institutes of Health Guide for Care and Use of Laboratory Animals. The methodology and results of animal studies were reported in accordance with the ARRIVE (Animal Research: Reporting in vivo Experiments) 2.0 guidelines.²⁸ The researchers devoted a great deal of effort to the minimization of animal suffering during the study.

Experimental design

Animals were divided into two groups randomly: the control group (n = 3) and the stroke group (n = 6). All animals underwent brain MRI scanning during the study. Imaging and perfusion of the control rats were carried out on the third day after sham surgery. Animals with stroke were imaged on the day before MCAO surgery, 3 hours after surgery, and then on the 1st, 3rd, 5th, 7th, 14th, 21st, 31st, 42nd, 56th, and 85th day after surgery. Subgroups of the ischemic rats were perfused on the 7th (n = 2), 21st (n = 1), 56th (n = 2), and 85th (n = 1) day after MCAO; brains were removed and frozen for further immunochemical and histological analyzes.

Surgical procedures

Focal brain ischemia was induced by temporary occlusion of the middle cerebral artery with silicone coated sutures (Doccol, USA) according to the modified MCAO model.²⁹ In brief, the filament was inserted into the circle of Willis through the external carotid artery to occlude the middle cerebral artery for 1.5 hours. Intraoperative MCAO monitoring was performed using MRI as detailed elsewhere.³⁰ The animals from the control group underwent sham surgery including the same procedures except for the artery occlusion. All surgeries were performed under inhalation anesthesia (1%–1.5% isoflurane in air) with atropine premedication (0.5 μl in 1 ml of solution intramuscularly). The temperature was maintained using an electric blanket at 37 ± 0.5°C.

MRI

Imaging was performed using a 7 Tesla small-animal MRI scanner (ClinScan; Bruker BioSpin, Germany). The animals were imaged under isoflurane anesthesia with monitoring of the body temperature, blood pressure, and heart rate.

MPF maps were obtained using the single-point synthetic reference method.^20,21,31 In brief, the method is based on reconstruction of MPF maps from three source spoiled gradient-echo images with magnetization transfer (MT), T₁, and proton density (PD) contrast weightings as the two-step procedure. In the first step, PD and R₁=1/T₁ maps are computed from the T₁- and PD-weighted images according to the two-point variable flip angle method.³² In the second step, an MPF map is reconstructed from an MT-weighted image and PD and R₁ maps by voxel-wise iterative solution of the pulsed MT equation²⁰ with a calculated synthetic reference image for data normalization.^20,21 Whole-brain 3D MPF maps were obtained with spatial resolution of 182 × 182 × 400 μm³ (matrix 192 × 192 ×128, field-of-view (FOV) = 35 × 35 × 51 mm³) and the total acquisition time of about 20 min. The MPF mapping protocol included the following sequences: MT-weighted with repetition time (TR) = 29 ms, echo time (TE) = 2 ms, flip angle (FA) = 10°, and scan time 9 min 21 s; Т₁-weighted with TR/TE= 16/2 ms, FA= 16°, and scan time 5 min 9 s; and PD-weighted with TR/TE =16/2 ms, FA = 3°; and scan time 5 min 9 s. Gaussian off-resonance saturation pulse was applied in the MT-weighted sequence with offset frequency of 1.8 kHz, effective FA = 500°, and duration of 9.984 ms.

ADC maps were obtained using a diffusion-weighted single-shot spin-echo echo-planar sequence with TR/TE = 5000/22 ms, three diffusion directions, 14 b-values in a range of 0–2000 s/mm², in-plane resolution 400 × 400 μm² (matrix 80 × 52, FOV = 32 × 21 mm²), slice thickness 1.0 mm, and scan time 3 min 25 s. T₂-weighted images were acquired using a turbo-spin-echo sequence with TR/TE = 7290/47 ms, echo train = 12, echo spacing 11.9 ms, in-plane resolution 117 × 117 μm², (matrix 256 × 180, FOV = 30 × 21 mm²), slice thickness 0.8 mm, and scan time 3 min 32 s. All images were obtained in the coronal plane.

Histology and immunohistochemistry

The rats were transcardially perfused on the 7th, 21st, 56th, or 85th day after surgery with PBS and 4% formaldehyde sequentially. Perfusions were performed under chloral hydrate anesthesia (450 mg/kg, intraperitoneally). Extracted brains were fixed in 4% paraformaldehyde for 3 days followed by cryoprotection in 10% and 20% sucrose sequentially. After cryoprotection, tissues were frozen in liquid nitrogen and stored at –80°C for further histological processing. Coronal brain sections (10 µm thickness) were prepared using an HM525 cryostat (Thermo Fisher Scientific, Germany). Brain sections for histological and immunohistochemical analysis were defined by visual matching to MPF maps in a range from +1.20 mm to +0.60 mm from bregma according to the rat brain atlas.³³

Brain sections were stained histologically with Luxol Fast Blue (LFB) (Bio-Optica, Milano, Italy) for the myelin assessment and immunohistochemically for the detection of axons (NF-H), mature neurons (NeuN), astrocytes (GFAP), neuroblasts (DCX), oligodendrocyte precursor cells (OPCs) (NG2), and myelination (MBP) with the following primary antibodies:

mature neurons (NeuN+ cells): Rabbit anti-NeuN, 0.5:100, ABN78, Millipore

myelin basic protein (MBP+ fibers): Goat anti-MBP, 1:100, sc-25665, Santa Cruz

axons (NF+ fibers): Rabbit anti-NF-H, 1:1000, ab8135, Abcam

astrocytes (GFAP+ cells): Rabbit anti-GFAP,1:100, ab7260, Abcam

young neurons (DCX+ cells): Goat anti-DCX, 1:100, sc-8066, Santa Cruz

OPC (NG2+ cells): Rabbit anti-NG2,1:100, AB5320, Millipore

Donkey antibodies anti-rabbit AlexaFlour 488 and anti-goat AlexaFlour 594 (#711-545-152 and #705-585-147; dilution 1:500; Jackson Immuno-research Laboratories, USA) were used as secondary antibodies. After labeling, brain slices were covered with mounting medium (Vector Laboratories, USA) with DAPI (4′,6-diamidino-2-phenylindole) and then examined using an Axio Imager Z2 microscope (Carl Zeiss, Germany) with AxioVision 4.8 software. Micrographs of the whole-brain sections were obtained to compare them with MRI. LFB-stained sections were photographed with 1× magnification, and sections with fluorescent labeling were photographed with 10×, 20×, and 40× objectives.

Image processing and analysis

MPF, R₁, and PD maps were reconstructed according to the single-point synthetic-reference method^20,31 using custom C++ language software available at https://macromolecularmri.org/. ADC maps were reconstructed using the MRI scanner’s software. MR images and parametric maps were analyzed using ImageJ 1.52 (National Institutes of Health, USA) and ITK-SNAP 3.8.0 (US National Library of Medicine) software. The ischemic lesion was identified within the caudoputamen according to its hypointensity on MPF maps and hyperintensity on T₂-weighted images on the first day after MCAO. Separate zones were defined within the lesion based on distinct time courses of their evolution on MPF maps. Specifically, it was observed that separate portions of the lesion exhibited either a continuous trend of a postoperative decrease (DE zone) or a decrease followed by recovery (RE zone). To place a region of interest (ROI) for MPF measurements in each zone, the MPF map sections at the same anatomical location were compared between the first day after MCAO and the day of endpoint. The RE zones were defined according to a visible increase in MPF between these time points, whereas the DE zones were defined by a reduced or visibly unchanged MPF. Then MPF values were recorded using a square ROI with the 3 × 3 voxels area (546 × 546 µm²) centered in each zone. For each ROI within the lesion, a matched ROI in a similar anatomic location within the contralateral hemisphere was also analyzed. The same ROIs were applied to the MPF maps obtained at other time points, T₂-weighted images, and R₁, PD, and ADC maps. Three consecutive slices were analyzed for T₂-weighted images and MPF, R₁, and PD maps, and two consecutive slices – for ADC maps.

The micrographs were analyzed using ImageJ software. Number and morphology of mature neurons (NeuN), astrocytes (GFAP), neuroblasts (DCX), OPCs (NG2), myelination (MBP and LFB), and axonal density (NF) were assessed in three brain sections for each animal. The ROIs with locations corresponding to those defined according to the MPF maps were placed on the micrographs of whole brain sections. For sham-operated animals, ROIs were placed approximately in the center of the caudoputamen. LFB optical density (LFB OD) was measured in 200 × 200 μm² ROIs from the intensity of the red channel on RGB images with background correction as previously described.^19,22,23

Myelination according to MBP staining and axonal density according to NF staining were quantified using the Otsu threshold method³⁴ in the ImageJ implementation. The measurements were taken from a 200 × 200 μm² ROI using semi-automatic ‘Analyze Particles’ function providing the parameters of object size and percentage of labeled total area. Percentages of the total area of detected objects were used as surrogate measures of axonal density³⁵ and myelination. Additionally, the object size was used to quantify microstructure of the myelinated and unmyelinated axonal bundles.

The number of mature and young neurons, astrocytes, and OPCs were assessed as the count of marker-positive cells co-localized with DAPI within an ROI of 100 × 200 μm² and normalized to 1 mm² area. One operator (MSK) analyzed MRI data. Other operators (AAK or TVA) performed visual matching between MRI and histology and analyzed histological data. All operators were blinded to information about the time between surgery and endpoint.

Statistical analysis

All statistical analyzes were performed using Statistica 10.0 (StatSoft Inc, USA) software. The percentage changes (PC) in the measured MRI, histological, and immunohistochemical parameters were calculated as PC = (Ref – Obs)/Ref * 100, where Obs is an observed parameter value at a particular time point, and Ref is a reference value. For MRI parameters, a ROI measurement at the corresponding anatomical location in the baseline maps obtained before surgery was taken as a reference value. As the observed values, the MPF, ADC, and R₁ measurements were taken in the native form, whereas for the T₂-weighted signal and PD, the ratios of the measurement in the lesion to that in the contralateral hemisphere were used. For histological and immunohistochemical parameters, a reference value was taken from a similar ROI in the contralateral hemisphere. PC in parameters were compared using the mixed-model analysis of variance in the ‘Variance Estimation and Precision’ module of Statistica software. In these analyses, the Restricted Maximum Likelihood Estimation method was used and the mixed model included five factors:

fixed hemisphere factor (ipsilateral or contralateral, for MPF,ADC, and R₁ only);

fixed zone factor nested in the hemisphere factor (DE zone, RE zone);

fixed time factor (for MRI: before surgery, 3 hours, and 1, 3, 5, 7, 14, 21, 30, 42, 56, 85 days after surgery; for histology and immunochemistry: sham surgery, 7, 21, 56, and 85 days after MCAO surgery);

random animal factor;

random slice factor nested in the animal factor.

The interaction of the factors time × zone was assessed by post-hoc pairwise tests with Fisher least significant difference (LSD) correction for multiple comparisons. For each model, normality of the distribution of standardized residuals was assessed using the Shapiro–Wilk test. No significant departures from the normal distribution were identified. All tests were two-tailed with significance level of p < 0.05.

Results

MPF maps reveal two distinct tissue evolution zones in the ischemic lesion

An example of longitudinal MRI monitoring of the ischemic lesion evolution in a single animal using MPF, R₁, PD, and ADC mapping and T₂–weighted imaging is illustrated in Figure 1. A summary of temporal lesion evolution on MPF maps in all studied animals is shown in Supplementary Figure S1. Temporal trajectories of MRI parameters evolution across all animals and time points are plotted in Figure 2 as the mean percentage changes in the parameter measurements in the lesion and contralateral hemisphere relative to the baseline. The lesions located in the caudoputamen were clearly seen in all animals from the first day after MCAO as regions of hypointensity in MPF, R₁, and ADC maps and hyperintensity on T₂-weighted images (Figure 1). The corresponding values of MPF, R₁, ADC, and T₂-weighted signal ratio were significantly different from the contralateral hemisphere (p < 0.001), whereas PD ratio showed a trend of increase, which did not reach significance (Figure 2). From the fifth day after surgery, two zones with different dynamics within the ischemic lesion were observed in the MPF maps (Figures 1, 2(a) and Supplementary Figure S1). One of these zones demonstrated visible hypointensity (Figure 1 and Supplementary Figure S1) and a significant MPF decrease (p < 0.001, Figure 2(a)) relative to the contralateral hemisphere, which has been maintained throughout the entire observation period. Another zone showed the most prominent hypointensity on the days 1–3 after MCAO (about 30%–40% relative to the baseline) followed by an increase in MPF values. As mentioned above, these regions were defined as the DE and RE zones for subsequent analysis. In the RE zones, the differences relative to the contralateral hemisphere diminished with time and became non-significant from the 30th day after surgery. The difference in MPF values between DE and RE zones increased from the first day after MCAO and remained significant (p < 0.001) from the first day until the end of observation. No significant changes relative to the baseline were detected for the contralateral regions matching the DE and RE zones over the entire course of observation.

Figure 1.

An example longitudinal observation of the ischemic lesion evolution in the same animal on MPF maps (top row), T₂-weighted images (second row), ADC maps (third row), PD maps (fourth row), and R₁ maps (bottom row) of the rat brain. Images were obtained before, 3 hours after, and then on the 1st, 3rd, 5th, 7th, 14th, 21st, 31st, 42nd, and 56th day after MCAO. Red and green arrows indicate the DE and RE zones within the ischemic lesion, respectively.

Figure 2.

Dynamic profiles of the mean percentage changes in MPF (a), T₂ signal ratio (b), ADC (c), PD ratio (d), and R₁ (e) relative to the pre-surgical baseline assessed over all time points and animals in the DE and RE zones identified within the ischemic lesions (red and green curves, respectively) and their contralateral matching regions (blue and violet curves). Error bars correspond to the standard deviations. Color-coded stars indicate significant differences (*p < 0.05, **p < 0.01, ***p < 0.001) between ipsilateral and contralateral measurements for the DE (red) and RE (green) zones and between the DE and RE zones (black).

ADC, R₁, and PD maps and T₂-weighted images showed different patterns of the lesion evolution in both DE and RE zones (Figures 1 and 2(b) to (e)). T₂-weighted signal ratio in both zones exhibited marked elevation followed by return to nearly normal values from the 21st day after MCAO (Figure 2(b)). The DE zone had significantly higher T₂-weighted signal ratio as compared to the RE zone in the early period of the lesion evolution (days 3–14, p < 0.01–0.001). These distinctions weakened at the later observation points (Figure 2(b)). The time course of ADC changes demonstrated an initial decrease followed by an increase (Figure 2(c)). The most notable differences in ADC values between the DE and RE zones were observed at the delayed time points (days 42–85, p < 0.01–0.001, Figure 2(c)) and were associated with hyperintensity of the DE zone and nearly normal ADC in the RE zone. During the earlier period, the RE zones showed slightly larger ADC in the interval from the third to seventh day with marginally significant differences (Figure 2(c)). The PD ratio demonstrated a trend of increase during the acute stroke phase, but the changes were subtle and statistically non-significant (Figure 2(d)). R₁ showed a significant transient decrease, which was observed from 3 hours till the 21st day after MCAO (Figure 2(e)). Both PD ratio and R₁ had no significant differences between the DE and RE zones at any time point.

Histological features of the tissue evolution zones identified according to MPF maps

Figure 3 illustrates the correspondence between MPF maps and histological patterns of myelination, axonal density, and astrogliosis in the same animals for all study endpoints. Magnified views of LFB-, MBP-, NF-, NeuN-, NG2-, and DCX-stained sections are presented in Figure 4 and quantitative analyses of PC in histological variables in the DE and RE zones relative to the contralateral hemisphere are summarized in Figure 5. Both LFB and MBP showed good overall agreement with MPF maps in delineation of the ischemic lesion (Figure 3). MPF maps and histology indicated clear distinctions between the DE and RE zones on the 21th, 56th, and 85th days after MCAO (Figures 3 and 4). For the seventh day endpoint, slight visual distinctions between the zones were seen on MPF maps, while the lesion had a rather uniform appearance in LFB- and MBP-stained sections (Figures 3 and 4). Quantitative analysis revealed significant demyelination in the lesion on the seventh day without significant distinctions between the DE and RE zones in both LFB OD (Figure 5(a)) and MBP-labeled area (Figure 5(b)). For subsequent endpoints, significant differences between the DE and RE zones in both measures of myelination were observed. Myelin content increased from the 7th to 56th day and then remained similar on the 85th day in the RE zone, whereas the DE zone appeared steadily demyelinated with a minor trend of increase in the myelin content at the later endpoints (Figures 3 to 5(a) and (b)).

Figure 3.

Matched whole-brain MPF maps and histological sections stained with LFB and immunofluoresce-labeled with MBP and GFAP corresponding to the 7th, 21st, 56th, and 85th day endpoints after MCAO. Red and green arrows indicate the DE and RE zones within the ischemic lesions, respectively.

Figure 4.

Magnified views (200× magnification) of the histological sections immunofluorescence-labeled with NF-H (2nd column), MBP (3rd column), co-localized NF-H and MBP (4th column), GFAP (5th column), NeuN (6th column), and co-localized NG2 and DCX (7th column) corresponding to either the DE and RE zones within the ischemic lesion in the animals studied on the 7th (2nd row), 21st (3rd row), 56th (4th row), and 85th (5th row) day after MCAO or similar control locations in a sham-operated animal studied on the 3rd day after surgery (1st row). The locations of magnified regions are labeled in the reference LFB-stained sections (1st column). In all immunofluorescence images, the blue channel depicts the nuclei labeled with DAPI.

Figure 5.

Quantitative histology profiles of the mean percentage changes relative to the contralateral hemisphere in LFB optical density (a), MBP-positive total area (b), NF-positive total area (c), MBP-positive object size (d), NF-positive object size (e), NeuN-positive cell count (f), DCX-positive cell count (g), NG2-positive cell count (h), and GFAP-positive cell count (i) calculated for the DE and RE zones of the ischemic lesion defined according to the MPF maps. The data correspond to the groups of animals studied after sham surgery (n = 3) and on the 7th (n = 2), 21st (n = 1), 56th (n = 2), and 85th (n = 1) day after MCAO. Error bars correspond to the standard deviations. Color-coded stars indicate significant differences (*p < 0.05, **p < 0.01, ***p < 0.001) between ipsilateral and contralateral measurements for the DE (red) and RE (green) zones and between the DE and RE zones (black).

Axonal area according to NF labeling was significantly reduced in both zones on the seventh day endpoint with a significantly larger amount of spared axons in the RE zone as compared to the DE zone (Figures 4 and 5(c)). Visual inspection of co-localized MBP- and NF2-labeled sections (Figure 4) indicated that nearly all persistent axons in the RE zone were unmyelinated. From the 21st to 56th day, the area occupied by axons in the RE zone showed a dramatic increase (Figure 4) and significantly exceeded that in sham-operated animals on the 56th and 85th day endpoints (Figure 5(c)), similar to the MBP-labeled area (Figure 5(b)). Axonal regrowth from the 7th to 85th day endpoint was accompanied by a gradual increase in the proportion of myelinated axons and formation of compact fiber bundles, as seen in the co-localized MBP- and NF-labeled sections (Figure 4). Quantitatively, the observed axonal remodeling in the RE zone is reflected by a prominent increase in the size of detected objects in both MBP- and NF-labeled sections from the 7th to 85th day endpoint (Figure 5(d) and (e)). In contrast, the DE zone had a significantly smaller axonal area and object size relative to the RE zone with a slight increase over all study endpoints (Figure 5(c) to (e)). While minimal axonal regrowth was seen at the late endpoints (56th and 85th days), the new axons were loosely distributed and did not form compact fiber bundles (Figure 4).

Ischemic lesions demonstrated massive neuronal loss (>90%) in both DE and RE zones on the seventh day endpoint (Figures 4 and 5(f)). For the later endpoints, the population of mature neurons remained substantially depleted, although the RE zone showed a small but significant increase in the NeuN-positive cell count on the 56th and 85th days as compared to the DE zone (Figure 5(f)). At the same time, the RE zone showed a progressive increase in the population of neuronal precursors and immature neurons according to DCX staining, which significantly exceeded that in both sham-operated animals and the DE zone of the same animals on the 21st, 56th, and 85th days after MCAO (Figures 4 and 5(g)).

The DE and RE zones also exhibited substantial distinctions in the patterns of oligodendrogenesis (Figures 4 and 5(h)) and astroglial proliferation (Figures 4 and 5(i)). In the DE zone, the number of NG2-positive cells showed monotonic increase over all endpoints, while the RE zone demonstrated a transient increase with the maximum on the 21st day endpoint (Figure 5(h)). Astrogliosis was observed as an elevation of the count of GFAP-positive cells in both DE and RE zones on the seventh day endpoint (Figures 3 to 5(i)). On the later endpoints, the amount of astroglia gradually returned to the normal level in the RE zone, whereas the DE zone showed a dramatic increase on the 21st day followed by a decrease at subsequent endpoints (Figures 4 and 5(i)). For the 21st day and later endpoints, the zone of increased astroglial proliferation showed clear visual coincidence with the DE zone on MPF maps (Figure 3), and the amount of astroglia in this zone remained significantly elevated. For all endpoints, the majority of GFAP-positive cells in both DE and RE zones showed a typical morphologic features of activated astroglia including hypertrophic processes and enlarged cell bodies (Figure 4).

Discussion

This study demonstrates the capability of the new quantitative MRI method, fast MPF mapping, to monitor demyelination and remyelination in the ischemic stroke lesion during its evolution from the acute to late chronic stage. Longitudinal observation of the lesion evolution in the transient MCAO model using MPF maps revealed two sub-regions with distinct MPF temporal trajectories (identified above as DE and RE zones), which exhibited either sustained decrease or a transient decrease followed by recovery to nearly normal values. The key histological features of these sub-regions were identified as either chronic demyelination in the DE zones or demyelination followed by remyelination and reconstitution of fiber bundles in the RE zones. Besides the major distinctions in the patterns of myelin destruction and recovery, the DE and RE zones demonstrated substantial differences in other histopathological signatures of neural tissue damage and repair.

Notable histological findings in the RE zones included active axonal regrowth, remodeling of compact fiber tracts, proliferation of neuronal and oligodendroglial precursors, and resolution of reactive astrogliosis. Collectively, these features indicate active processes of brain tissue repair, which start between one and three weeks after ischemic injury. Axonal remodeling and remyelination observed in the RE zone well correspond to a transient increase in the amount of OPCs, which is a necessary condition to restore population of myelinating oligodendrocytes.^36,37 Temporary increase and subsequent resolution of reactive astrogliosis in this zone also can be viewed as a supporting factor enabling axonal regrowth.^38–40 At the same time, complete restoration of tissue microstructure was not achieved within the time-frame of our experiments, as evidenced by substantially depleted neuronal population and relatively loose fiber bundles observed at the late endpoints. As such, a longer observation time would be needed to investigate the possibility of complete neural tissue repair in the lesion sub-regions undergoing active remyelination. Relatively rapid axonal reorganization and remyelination in conjunction with slow recovery of neuronal population suggests that the primary route of tissue evolution in the RE zones is neuroplastic remodeling directed towards reconstitution of structural connectivity. This interpretation is in line with the evidences of post-stroke neuroplasticity obtained using diffusion tensor imaging (DTI) and functional MRI.^11–14 The signs of structural and functional brain connectivity reorganization in the rat MCAO model were observed in the timeframe from two weeks to several month after stroke^11–14 that is in good agreement with the timing of remyelination found in our study. Taken together, this and previous studies suggest that remyelination measured by MPF mapping can be viewed as a prospective biomarker of neural tissue plasticity in the post-ischemic brain.

The patterns of more severe and, possibly, irreversible tissue injury were histologically identified in the DE zones, which demonstrated persistent demyelination, almost complete neuronal loss, minimal signs of axonal sprouting, and formation of the glial scar. Nevertheless, the DE zones did not exhibit total tissue necrosis resulting in the development of a cystic lesion. The observed type of lesion evolution is probably typical for striatal infarcts in the transient MCAO model as opposed to cortical infarcts according to the literature.⁴¹ Our data also indicate that damaged neural tissue in the DE zones may have some potential for recovery, since subtle remyelination and axonal regrowth were observed at the late study endpoints. An increase in the amount of oligodendrocyte precursors observed on the 56th and 85th day endpoints is in line with this trend. While the observation timeframe of this study was rather long, it still seems to be insufficient to completely characterize the time course of the striatal ischemic infarct evolution in the MCAO model. Further studies with longer post-occlusion time intervals (perhaps up to one year⁴²) are warranted to fully investigate the potential of neural tissue regeneration in different zones of the ischemic lesion.

The two distinct zones defined according to the temporal trajectories of MPF are characterized by different profiles of some MRI parameters, particularly ADC and T₂-weighted signal. At the same time, other parameters (PD and R₁ = 1/T₁) did not reveal differences between these zones. The behavior of conventional imaging variables observed in this study was in good agreement with the literature data for the MCAO model with similar monitoring periods.^41,43,44 Particularly, T₁ and T₂ values were reported to return to normal values in about 2–3 weeks after ischemia^41,43 unless tissue undergoes total necrosis, which is typical for cortical infarcts but usually not seen in striatal lesions.^41,43 According to the literature, normalization of T₁ and T₂ values was associated with resolution of vasogenic edema, accumulation of iron-rich macrophages, and gliosis.⁴¹ PD changes in focal ischemia reported earlier,⁴³ similar to our data, were rather subtle with a maximal increase of about 10% on the second day after MCAO corresponding to the peak of edema. ADC is known to decrease in acute infarcts due to cytotoxic edema and subsequently increase as vasogenic edema develops.^44,45 Interplay between these factors typically results in the heterogeneous appearance of the ischemic lesion on ADC maps in the sub-acute period. Usually, ADC evolution has a transient period of pseudo-normalization, which is commonly viewed as a manifestation of the cytotoxic to vasogenic edema transformation.^44,45 We observed the pattern of ADC pseudo-normalization in the DE zone of the lesion, whereas the RE zone showed a recovery to normal values within the same timeframe. It is noticeable that these ADC changes preceded both the return of MPF in the RE zone to normal values and substantial axonal remodeling accompanied with re-myelination according to histology. On the other hand, ADC pseudo-normalization appeared coincident with the period of astroglial proliferation. Additionally, the previous study¹⁹ demonstrated strong positive correlation between ADC and the amount of activated microglia in acute and sub-acute stroke in rats. Earlier studies^46,47 also reported associations between ADC and glial markers in different ischemic stroke models. Taken together, these observations suggest that various glial responses may contribute into ADC normalization or pseudo-normalization besides water redistribution between the intra- and extracellular compartments. Long-term evolution of ADC in chronic infarcts also can be variable. High ADC values were reported for necrotic cortical infarcts, whereas a subtle elevation was found in the subcortical areas affected by ischemia.⁴³ An increase in ADC in the DE zone at late time points found in this study is likely caused by a reduced axonal density or other microstructural abnormalities described above.

From the practical standpoint, both T₂ and ADC are difficult to use as markers of tissue repair due to the effect of pseudo-normalization and heterogeneity of temporal profiles. While we found minor differences between the DE and RE zones in the time courses of ADC and T₂-weighted signal in the sub-acute phase of infarct (days 3–7 after MCAO), the distinctions were subtle and did not follow a consistent trend of separation, unlike that seen for MPF. These distinctions probably indicate a different extent and/or time of edema resolution. MPF measurements also can be affected by edema to some degree.¹⁹ Particularly, the difference in tissue water content may explain the discrepancy between the MPF maps and myelin histology on the seventh day endpoint, where no significant differences in the myelin loss between DE and RE zones were observed, while MPF values significantly diverged. It was demonstrated earlier¹⁹ that edema may explain up to 10% of a relative MPF decrease in the acute ischemic lesion. This estimate is in good agreement with the difference in MPF changes between the DE and RE zones observed for the seventh day endpoint. It also should be noted that for the later endpoints, MPF maps closely matched the patterns of demyelination and remyelination displayed by both LFB and MBP histology. From the physiological standpoint, our findings suggest that remyelination regions may be characterized by either less severe edema or its faster resolution. As such, more research is needed to evaluate the prognostic value of the persistence time and extent of edema in the ischemic lesion for tissue viability and recovery potential.

To the best of our knowledge, the presented study is the first application of the fast MPF mapping method^20,21 to chronic stroke. The previous reports provided extensive histological validation of MPF measured by this method as a myelin biomarker in the normal rat and murine brain,^22,23 a murine model of cuprizone-induced demyelination²³ and subsequent re-myelination,²⁴ as well as acute and sub-acute stroke phases in the rat MCAO model.¹⁹ This study revealed close correspondence between myelin histology and MPF at different stages of the chronic ischemic infarct evolution and the feasibility of monitoring post-stroke remyelination using fast MPF mapping. A practically important contribution of this study is the demonstration of the utility of MPF as a measure of myelin content in the setting of post-ischemic gliosis. Particularly, our results indicate that despite the presence of high cellular content and extracellular matrix associated with the glial scar, the heavily demyelinated DE zones had steadily low MPF values over the entire course of observation. Furthermore, a trend of slight MPF increase in the DE zones after the 30th day of longitudinal MRI monitoring appeared in consistence with subtle remyelination histologically confirmed on the 56th and 85th day endpoints. These findings substantiate insensitivity of MPF to gliosis and provide an additional evidence of MPF specificity as a myelin biomarker.

Insensitivity of MPF to changes in the amount of various cell types in neural tissues is worth further consideration. In addition to the results of this study suggesting independence of MPF from astrogliosis, the previous study¹⁹ did not find correlations between MPF and the loss of axons and neurons, as well as microglial proliferation in acute and sub-acute ischemic stroke. The earlier study²² found that MPF does not correlate with the total cell count in the normal rat brain and glioma. At the same time, it is known that MPF does not provide one-to-one correspondence with the myelin content due to contributions from other macromolecular components in tissues. Previous studies^22,23 demonstrated that MPF has a close linear relationship with the myelin content characterized by a non-zero intercept, which reflects the background non-myelin macromolecular concentration. As discussed earlier,²⁵ the major non-myelin contribution into MPF is likely to arise from the plasma membranes, which represent complex spatially-ordered lipoprotein structures with a large concentration of the semisolid protons.⁴⁸ Based on the published MPF values and their regressions on the myelin density,¹⁹ the non-myelin contributions account for about 40%–60% of measured MPF in gray matter and 20%–30% in white matter. As such, changes in non-myelin MPF components theoretically could substantially affect MPF values in gray matter. However, the experimental data of this and previous¹⁹ studies do not support this expectation. The most probable explanation of this observation is a nearly constant density of plasma membranes in the course of the ischemic lesion evolution, which is maintained despite the dramatic changes in separate cell populations. More specifically, during the lesion development, vanishing neurons, axons, and oligodendrocytes are replaced by proliferating microglia followed by astroglia, thus preserving an average amount of cell membrane material per unit volume. In contrast, a contribution from destroyed myelin remains largely uncompensated and results in an observable MPF reduction. In good agreement with the estimated above myelin and non-myelin components of MPF in gray matter, the relative myelin loss of 60%–80% in the DE region of the striatal stroke lesion (Figure 4(a) and (b)) found in this study translates into an about 30%–40% reduction in MPF (Figure 2(a)). Accordingly, our data suggest that MPF adequately reproduces myelin content changes identified by histology and immunohistochemistry but has an about 50% narrower dynamic range. At the same time, it remains to be investigated whether MPF measurements reflect intact myelin or they also can be partially affected by myelin breakdown products, which may cause a similar reduction in the scale of quantitative changes.

The results of this and previous studies based on fast MPF mapping^{19–21,23–27,31,49–53} demonstrate that this method provides a convenient and reliable tool for preclinical and clinical studies of myelin damage and repair. The method is easily translatable to clinics as evidenced by several pilot clinical studies.^{25–27,49,50,52} It provides highly reproducible MPF measurements in humans⁵³ and animals.²³ Since MPF is independent of magnetic field strength,^31,51 MPF mapping enables straightforward comparison between quantitative data obtained using human and small-animal MRI equipment. The use of MPF mapping also can be of potential interest for multimodal cross-validation in preclinical and clinical studies of novel PET tracers allowing the assessment of demyelinated axons in the ischemic lesion.⁵⁴ An important practical advantage of the fast MPF mapping method for clinical stroke research is the feasibility of ultrafast acquisition protocols with the total scan time of a few minutes.^49,50,52

Post-stroke remyelination has been documented in a number of studies.^{16,18,36,55–58} However, the factors determining the capacity of brain tissue to remyelinate after ischemic injury and their regional distinctions are still poorly understood. Our observations suggest that the remyelination zone contains a larger amount of spared but demyelinated axons in the sub-acute infarct phase as compared to the zone of chronic demyelination. Remyelination of these axons may trigger more active OPC recruitment and axonal remodeling with subsequent myelination of newly formed axons. The difference in the axonal loss may be caused by less severe ischemic tissue damage in the RE zone during the infarct formation. An underlying reason can be related to regional distinctions in the microcirculatory environment and, therefore, local variations of the perfusion deficit within the infarct core. At the same time, it is rather unlikely that the DE and RE zones represent ischemic core and penumbra, since both portions of the lesion initially belong to the hypointense ADC region, which is classically defined as the infarct core.⁵⁹ These considerations suggest the need in further research involving perfusion imaging to investigate a relationship between the severity of ischemia and subsequent tissue remyelination potential.

This study has several limitations. First, it is based on a relatively small number of animals. To partially mitigate this limitation, we applied an advanced mixed modeling approach to obtain rigorous statistical estimates of the lesion evolution taking into account multiple observations per animal and unequal numbers of observations. Second, we used the T₂ signal ratio instead of quantitative T₂ mapping to characterize the lesion properties on T₂-weighted images. While our motivation was to reduce the scan time as much as possible in view of multiple scanning sessions per animal, the lesion evolution based on T₂ signal ratio appeared fairly consistent with the literature data based on quantitation of the T₂ relaxation time.^43,44 Similarly, PD was presented as a ratio, because absolute quantitation of this parameter requires rather cumbersome and time-consuming MRI acquisition procedures to correct for the effects of coil reception profile, T₂* decay, and field inhomogeneities.⁶⁰ These corrections were not applied, since uncorrected PD maps are used in the single-point MPF mapping method,^21,31 which was the primary focus of this study. Third, we did not apply B₀ and B₁ field corrections to MPF and other parametric maps due to the absence of dedicated pulse sequences on the used small-animal MRI scanner. However, B₀ inhomogeneity was recently shown to have a negligible effect on MPF measurements.⁵³ While B₁ errors may cause significant bias in R₁ and MPF values,⁵³ their effect is unlikely to be of practical importance for this study, because the animals were uniformly positioned within the scanner, and the serial measurements were obtained from identical anatomic locations. Fourth, the conclusions of this study were specifically derived for striatal ischemic lesions in the rat transient MCAO model. As such, the presented results should be used with caution in the interpretation of data obtained for different biological species, anatomic regions, and ischemic stroke models. Particularly, future studies are needed to evaluate relevance of our findings to stroke in humans and ischemic lesions in white matter. Finally, our interpretation of NG2-positive cells as OPCs should be considered with caution, since this marker is also expressed in pericytes.⁶¹ Accordingly, an increase in the NG2-positive cell count observed in this study may in part reflect vascular remodeling occurring in parallel with other restorative processes in the ischemic lesion.

In conclusion, this study histologically validated the fast MPF mapping method as a novel imaging approach for quantitative assessment of myelin damage and repair in ischemic stroke from the acute to late chronic stage. Our results demonstrate the technical feasibility of non-invasive monitoring of remyelination in the ischemic lesion and underscore the vital role of this process for post-stroke brain tissue recovery. The results of this study also suggest that remyelination measured by fast MPF mapping can be used as a surrogate quantitative biomarker of post-ischemic neuroplasticity for preclinical and clinical research.

Supplemental Material

sj-pdf-1-jcb-10.1177_0271678X211020860 - Supplemental material for Long-term monitoring of chronic demyelination and remyelination in a rat ischemic stroke model using macromolecular proton fraction mapping

Supplemental material, sj-pdf-1-jcb-10.1177_0271678X211020860 for Long-term monitoring of chronic demyelination and remyelination in a rat ischemic stroke model using macromolecular proton fraction mapping by Marina Yu Khodanovich, Ilya L Gubskiy, Marina S Kudabaeva, Darya D Namestnikova, Alena A Kisel, Tatyana V Anan’ina, Yana A Tumentceva, Lilia R Mustafina and Vasily L Yarnykh in Journal of Cerebral Blood Flow & Metabolism

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the Russian Science Foundation projects No. 14-45-00040 (MRI data acquisition), No. 18-15-00229 (histology and immunohistochemistry), and No. 19-75-20142 (data analysis). Software for MPF map reconstruction was provided under support of the NIH High-Impact Neuroscience Research Resource grant R24NS104098. MSK received salary support from the Russian Foundation for Basic Research, project No. 19-315-90119. AAK and VLY received salary support from the NIH grants R21NS109727 and R24NS104098.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Authors’ contributions

MYK and VLY designed the study. MYK performed statistical analysis and wrote the manuscript. DDN was responsible for the development of the animal model. ILG and DDN performed MRI studies. AAK, MSK, and TVA performed image analysis. AAK, MSK, TVA, and YAT performed tissue processing and immunohistochemical labeling. LRM performed histological processing and staining. VLY developed an MRI protocol and image processing software and critically revised the manuscript. All authors reviewed and approved the final version of the manuscript.

Supplemental material

Supplemental material for this article is available online.

ORCID iD

Marina Yu Khodanovich

References

Donkor

ES.

Stroke in the 21st century: a snapshot of the burden, epidemiology, and quality of life. Stroke Res Treat 2018; 2018: 3238165.

Balbinot

Schuch

CP.

Compensatory relearning following stroke: cellular and plasticity mechanisms in rodents. Front Neurosci 2019; 12: 1–22.

Hylin

Kerr

Holden

Understanding the mechanisms of recovery and/or compensation following injury. Neural Plast 2017; 2017: 1–12.

Bach-y-Rita

Theoretical and practical considerations in the restoration of function after stroke. Top Stroke Rehabil 2001; 8: 1–15.

Zhao

Willing

Enhancing endogenous capacity to repair a stroke-damaged brain: an evolving field for stroke research. Prog Neurobiol 2018; 163–164: 5–26.

Hallett

Plasticity of the human motor cortex and recovery from stroke. Brain Res Brain Res Rev 2001; 36: 169–174.

Barber

Davis

Infeld

, et al. Spontaneous reperfusion after ischemic stroke is associated with improved outcome. Stroke 1998; 29: 2522–2528.

Nudo

RJ.

Adaptive plasticity in motor cortex: implications for rehabilitation after brain injury. J Rehabil Med Suppl 2003; 35: 7–10.

Cheng

Schulz

Bönstrup

, et al. Structural plasticity of remote cortical brain regions is determined by connectivity to the primary lesion in subcortical stroke. J Cereb Blood Flow Metab 2015; 35: 1507–1514.

10.

Carmichael

Kathirvelu

Schweppe

, et al. Molecular, cellular and functional events in axonal sprouting after stroke. Exp Neurol 2017; 287: 384–394.

11.

Kalthoff

Kim

, et al. White matter reorganization and functional response after focal cerebral ischemia in the rat. PLoS One 2012; 7: e45629.

12.

Van Meer

MPA

Otte

van der Marel

, et al. Extent of bilateral neuronal network reorganization and functional recovery in relation to stroke severity. J Neurosci 2012; 32: 4495–4507.

13.

Jiang

Zhang

Chopp

MRI of stroke recovery. Stroke 2010; 41: 410–414.

14.

Dijkhuizen

van der Marel

Otte

, et al. Functional MRI and diffusion tensor imaging of brain reorganization after experimental stroke. Transl Stroke Res 2012; 3: 36–43.

15.

Gibson

Purger

Mount

, et al. Neuronal activity promotes oligodendrogenesis and adaptive myelination in the mammalian brain. Science 2014; 344: 1252304.

16.

Jia

Kamen

Pivonkova

, et al. Neuronal activity-dependent myelin repair after stroke. Neurosci Lett 2019; 703: 139–144.

17.

Zhang

Chopp

Gregg

, et al. Patterns and dynamics of subventricular zone neuroblast migration in the ischemic striatum of the adult mouse. J Cereb Blood Flow Metab 2009; 29: 1240–1250.

18.

Zhang

Chopp

Neurorestorative therapies for stroke: underlying mechanisms and translation to the clinic. Lancet Neurol 2009; 8: 491–500.

19.

Khodanovich

Kisel

Akulov

, et al. Quantitative assessment of demyelination in ischemic stroke in vivo using macromolecular proton fraction mapping. J Cereb Blood Flow Metab 2018; 38: 919–931.

20.

Yarnykh

VL.

Fast macromolecular proton fraction mapping from a single off-resonance magnetization transfer measurement. Magn Reson Med 2012; 68: 166–178.

21.

Yarnykh

VL.

Time-efficient, high-resolution, whole brain three-dimensional macromolecular proton fraction mapping. Magn Reson Med 2016; 75: 2100–2106.

22.

Underhill

Rostomily

Mikheev

, et al. Fast bound pool fraction imaging of the in vivo rat brain: association with myelin content and validation in the C6 glioma model. Neuroimage 2011; 54: 2052–2065.

23.

Khodanovich

Sorokina

Glazacheva

, et al. Histological validation of fast macromolecular proton fraction mapping as a quantitative myelin imaging method in the cuprizone demyelination model. Sci Rep 2017; 7: 12.

24.

Khodanovich

Pishchelko

Glazacheva

, et al. Quantitative imaging of white and gray matter remyelination in the cuprizone demyelination model using the macromolecular proton fraction. Cells 2019; 8: 1204.

25.

Yarnykh

Bowen

Samsonov

, et al. Fast whole-brain three-dimensional macromolecular proton fraction mapping in multiple sclerosis. Radiology 2015; 274: 210–220.

26.

Yarnykh

Krutenkova

Aitmagambetova

, et al. Iron-insensitive quantitative assessment of subcortical gray matter demyelination in multiple sclerosis using macromolecular proton fraction. AJNR Am J Neuroradiol 2018; 39: 618–625.

27.

Petrie

Cross

Yarnykh

, et al. Neuroimaging, behavioral, and psychological sequelae of repetitive combined blast/impact mild traumatic brain injury in Iraq and Afghanistan war veterans. J Neurotrauma 2014; 31: 425–436.

28.

Percie Du Sert

Hurst

Ahluwalia

, et al. The ARRIVE guidelines 2.0: updated guidelines for reporting animal research. J Cereb Blood Flow Metab 2020; 40: 1769–1777.

29.

Longa

Weinstein

Carlson

, et al. Reversible middle cerebral artery occlusion without craniectomy in rats. Stroke 1989; 20: 84–91.

30.

Gubskiy

Namestnikova

Cherkashova

, et al. MRI guiding of the middle cerebral artery occlusion in rats aimed to improve stroke modeling. Transl Stroke Res 2018; 9: 417–425.

31.

Naumova

Akulov

Khodanovich

, et al. High-resolution three-dimensional macromolecular proton fraction mapping for quantitative neuroanatomical imaging of the rodent brain in ultra-high magnetic fields. Neuroimage 2017; 147: 985–993.

32.

Deoni

SCL

Rutt

Peters

TM.

Rapid combined T1 and T2 mapping using gradient recalled acquisition in the steady state. Magn Reson Med 2003; 49: 515–526.

33.

Paxinos

Watson

The rat brain in stereotaxic coordinates. San Diego, CA, USA: Academic Press, 2007.

34.

Otsu

A threshold selection method from gray-level histograms. IEEE Trans Syst Man Cybern 1979; 9: 62–66.

35.

Kneynsberg

Collier

Manfredsson

, et al. Quantitative and semi-quantitative measurements of axonal degeneration in tissue and primary neuron cultures. J Neurosci Methods 2016; 266: 32–41.

36.

Tanaka

Nogawa

Suzuki

, et al. Upregulation of oligodendrocyte progenitor cells associated with restoration of mature oligodendrocytes and myelination in peri-infarct area in the rat brain. Brain Res 2003; 989: 172–179.

37.

Bonfanti

Gelosa

Fumagalli

, et al. The role of oligodendrocyte precursor cells expressing the GPR17 receptor in brain remodeling after stroke. Cell Death Dis 2017; 8: e2871.

38.

Faulkner

Herrmann

Woo

, et al. Reactive astrocytes protect tissue and preserve function after spinal cord injury. J Neurosci 2004; 24: 2143–2155.

39.

Anderson

Burda

Ren

, et al. Astrocyte scar formation aids central nervous system axon regeneration. Nature 2016; 532: 195–200.

40.

Baaklini

Rawji

Duncan

, et al. Central nervous system remyelination: roles of glia and innate immune cells. Front Mol Neurosci 2019; 12: 1–18.

41.

Wegener

Weber

Ramos-Cabrer

, et al. Temporal profile of T2-weighted MRI distinguishes between pannecrosis and selective neuronal death after transient focal cerebral ischemia in the rat. J Cereb Blood Flow Metab 2006; 26: 38–47.

42.

Modo

Beech

Meade

, et al. A chronic 1 year assessment of MRI contrast agent-labelled neural stem cell transplants in stroke. Neuroimage 2009; 47: T133–T142.

43.

van der Zijden

van der Toorn

van der Marel

, et al. Longitudinal in vivo MRI of alterations in perilesional tissue after transient ischemic stroke in rats. Exp Neurol 2008; 212: 207–212.

44.

Knight

Dereski

Helpern

, et al. Magnetic resonance imaging assessment of evolving focal cerebral ischemia: comparison with histopathology in rats. Stroke 1994; 25: 1252–1261.

45.

Sotak

CH.

The role of diffusion tensor imaging in the evaluation of ischemic brain – a review. NMR Biomed 2002; 15: 561–569.

46.

Weber

Chan

Nie

, et al. Sensitivity of diffusion MRI to perilesional reactive astrogliosis in focal ischemia. NMR Biomed 2017; 30: 1–8.

47.

Rudrapatna

Wieloch

Beirup

, et al. Can diffusion kurtosis imaging improve the sensitivity and specificity of detecting microstructural alterations in brain tissue chronically after experimental stroke? Comparisons with diffusion tensor imaging and histology. Neuroimage 2014; 97: 363–373.

48.

Bloom

Holmes

Mountford

, et al. Complete proton magnetic resonance in whole cells. J Magn Reson 1986; 69: 73–91.

49.

Yarnykh

Prihod’ko

Savelov

, et al. Quantitative assessment of normal fetal brain myelination using fast macromolecular proton fraction mapping. AJNR Am J Neuroradiol 2018; 39: 1341–1348.

50.

Korostyshevskaya

Savelov

Papusha

, et al. Congenital medulloblastoma: fetal and postnatal longitudinal observation with quantitative MRI. Clin Imaging 2018; 52: 172–176.

51.

Anisimov

Pavlova

Pirogov

, et al. Three-dimensional fast single-point macromolecular proton fraction mapping of the human brain at 0.5 Tesla. Quant Imaging Med Surg 2020; 10: 1441–1449.

52.

Korostyshevskaya

Prihod’ko

Savelov

, et al. Direct comparison between apparent diffusion coefficient and macromolecular proton fraction as quantitative biomarkers of the human fetal brain maturation. J Magn Reson Imaging 2019; 50: 52–61.

53.

Yarnykh

Kisel

Khodanovich

MY.

Scan–rescan repeatability and impact of B0 and B1 field nonuniformity corrections in single-point whole-brain macromolecular proton fraction mapping. J Magn Reson Imaging 2020; 51: 1789–1798.

54.

Guehl

Ramos-Torres

Linnman

, et al. Evaluation of the potassium channel tracer [8F]3F4AP in rhesus macaques. J Cereb Blood Flow Metab. Epub ahead of print 22 October 2020. DOI: 10.1177/0271678X20963404.

55.

Sozmen

Rosenzweig

Llorente

, et al. Nogo receptor blockade overcomes remyelination failure after white matter stroke and stimulates functional recovery in aged mice. Proc Natl Acad Sci U S A 2016; 113: E8453–E8462.

56.

Anan’ina

Kisel

Kudabaeva

, et al. Neurodegeneration, myelin loss and glial response in the three-vessel global ischemia model in rat. Int J Mol Sci 2020; 21: 6246.

57.

Marin

Carmichael

ST.

Mechanisms of demyelination and remyelination in the young and aged brain following white matter stroke. Neurobiol Dis 2019; 126: 5–12.

58.

Itoh

Maki

Lok

, et al. Mechanisms of cell–cell interaction in oligodendrogenesis and remyelination after stroke. Brain Res 2015; 1623: 135–149.

59.

Kidwell

Alger

Saver

JL.

Beyond mismatch: evolving paradigms in imaging the ischemic penumbra with multimodal magnetic resonance imaging. Stroke 2003; 34: 2729–2735.

60.

Venkatesan

Lin

Gurleyik

, et al. Absolute measurements of water content using magnetic resonance imaging: preliminary findings in an in vivo focal ischemic rat model. Magn Reson Med 2000; 43: 146–150.

61.

Zheng

Chopp

Chen

Multifaceted roles of pericytes in central nervous system homeostasis and disease. J Cereb Blood Flow Metab 2020; 40: 1381–1401.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.66 MB