Fig 1.
ECM initiates in the olfactory bulb.
Upper row (blood-brain barrier disruption): Representative coronal 2D T1w- images at baseline (left column), after iv. Gadofluorine M (Gf-M, second column) and after Gadolinium-DTPA administration (third to fifth column); The high-molecular contrast agent Gf-M (bound to albumin) showed a multifocal extravasation pattern with predominance at the outer regions of the OB, when mice were clinically healthy = RMCBS 20 (small white arrows, second column). Before any moderate or severe clinical symptoms were evident (RMCBS 18) a diffuse extravasation pattern of the low-molecular contrast agent Gd-DTPA occurred and was mainly visible in the central region of the OB (parallel transparent arrows, third to fifth column). Middle row (edema): Representative 2D T2-weighted images (echo time = 22ms) showed normal signal at baseline (first column) and when multifocal Gf-M extravasation had already occurred in clinically healthy mice = RMCBS 20 (second column). When Gd-DTPA extravasation was evident in the central region of the OB, T2 signal also increased (parallel transparent arrows, third column) and continued to increase with progressive disease (parallel transparent arrows, forth and fifth column). Lower row (hemorrhage): Representative T2*-weighted images (echo time = 18ms) sensitively revealed microhemorrhages. At baseline (first column) and when multifocal Gf-M had already extravasated in clinically healthy mice = RMCBS 20 (second column) no microhemorrhages were detected. Before any moderate or severe clinical symptoms were evident (RMCBS 18) few microhemorrhages had occurred in the outer regions of the OB at similar locations of Gf-M extravasation (white arrowheads, third column). In moderate disease (RMCBS 14) microhemorrhages were still mainly located in the outer regions of the OB (white arrowheads, forth column). In severe disease (RMCBS 4) microhemorrhages had also occurred in the central regions of the OB (white arrowheads, fifth column). Mice of the two different groups are displayed. In the second column labelled "Gadofluorine M" one mouse of the second group is shown; in the columns denominated "Gd-DTPA" images of 2 mice of group 1 at different stages of severity are shown (baseline, RMCBS 14 and RMCBS 4 are from the same mouse, while RMCBS 18 is from another mouse).
Fig 2.
Edema spreads along the RMS in rostral-to-caudal direction.
A: Edema, seen as hyperintense (bright on T2*-weighted images) signal, started in the OB (white arrows in left column of A) and with progressive disease (from top to bottom) continued to spread from there along the RMS to the subventricular zone of the lateral ventricles. At high magnification zooms (right column with zoom region indicated in middle column) the involvement of the subcallosal subventricular zone and the dorsal migratory stream is apparent, but does not include the corpus callosum (“cc”) per se (arrowhead in focused magnifications and solid white single arrow on lower image in left column). Rostral-to-caudal extension was associated very closely with disease manifestation and progression both defined by decreasing RMCBS (RMCBS values are denoted on left of A). B: Schematic drawing of neurogenic regions in the mouse brain is presented. Neuroblasts migrate in the direction of the thick solid black curved arrow from the subventricular zone (SVZ) along the rostral-migratory-stream (RMS) to the olfactory bulb (OB) or along the dorsal-migratory-stream (DMS) above the hippocampus towards the occipital cortex (short curved thin black arrow pointing to right margin of image). C: T2 relaxation time increased in the RMS+OB early in the disease, when mice did not show evident clinical symptoms and was significantly elevated in progressive disease (RMCBS 15–10 and 9–0 each with p<0.05 versus baseline as indicated by asterisk, solid black line and circles indicating data for RMS+OB). A neuroanatomical landmark indicating transition to severe ECM is involvement of the dorsal-migratory-stream (DMS) showing significantly elevated T2 times in RMCBS 9–0 (p<0.05, asterisk). At this stage, also further spread into the external capsule and eventually into deeper brain structures and the brainstem occured (all <0.05 for severe ECM with RMCBS 9–0).
Fig 3.
Microglial cells are closely associated with the rostral migratory stream.
(A) On T2w images, the rostral migratory stream shows strongly hyperintense signal, indicative of vasogenic edema as a consequence of inflammatory BBBD. At the cribriform plate edema can also be appreciated (white arrow), emphasizing that the known anatomical connection of perivascular CSF spaces with the nasal lymphatics at the level of the cribriform plate is an important route of inflammatory infiltration in ECM (B). Inserted squares in panel (A) refer to the anatomical location of the co-immuofluorescent images in panel (C). Co-immunofluorescent staining of Iba-1 (green) and doublecortin (red) on sagittal brain sections through the RMS elbow in a representative control and an ECM mouse. In the ECM mouse, a clustering of activated microglial cells is visible along the RMS with a significantly increased number of activated microglial cells along the RMS (doublecortin positive regions) compared to a lower microglial cell number in the vicinity around the RMS in doublecortin negative regions. (D). Bars indicate 50μm.
Fig 4.
In severe disease edema extends to the brainstem.
A: T1 image relative variation of contrast by subtracting pre- from post -contrast images (after i.v. Gd-DTPA). Contrast enhancement indicative of BBBD is seen early on in the OB and anterior aspect of RMS (white transparent arrowheads in left column and second row: RMCBS 18, sagittal orientation). With increasing disease severity (third and fourth row: RMCBS 14 and RMCBS 4) contrast enhancement increases and includes the RMS completely (white transparent arrowheads left column, third and fourth row), extending towards the DMS (small white transparent arrows left column and middle column), deep grey matter and brain stem. Involvement of deep grey matter and brain stem can be better appreciated on coronal image sections (middle and right column, coronal orientation). B: Lacunarity values at the level of the thalamus and the brainstem are presented. During moderate (n = 5) and severe disease (n = 6) lacunarity values in the thalamus were not significantly different. In the brainstem, however, lacunarity increased significantly in severe disease compared to moderate disease.
Fig 5.
Severe disturbance of neuroblast chain migration along the RMS.
Immunofluorescent staining against doublecortin on sagittal brain sections of a representative control and an ECM mouse at the level of the RMS elbow and posterior RMS (illustrated in schematic drawing) as well as bar charts quantitatively assessing the linear migration pattern of neuroblasts (controls n = 3, moderate disease n = 5, severe disease n = 6). A linearity index was calculated to analyze morphological changes of neuroblast alignment (quotient of major axis length and minor axis length). Compared to controls, neuroblasts in ECM mice show severely altered morphology. The typical chain migration pattern and the linear appearance deviate when compared to controls and is more pronounced in the OB, but can also be detected in the posterior RMS (Bars = 50μm).
Fig 6.
T2* (A, coronal orientation), maximum-intensity-projection reconstructions of time-of-flight angiographic images (B, coronal orientation) and ADC maps (C, coronal orientation) at different RMCBS grades are presented (denoted on top). Microvascular pathology in ECM increases with disease severity and is paralleled by the occurrence of microbleeds (arrowheads, row A). It is also paralleled by increasing vessel density on T2* images consistent and by vessel rarification on TOF angiographic images. Both these latter findings are indicative of slow flow or microvascular sequestration in severe disease (row B). Ischemic infarctions, if present, were seen in very severe disease only and were confined to restricted regions of watershed/borderzone regions, which suffer first from low cerebral perfusion (arrows, row C). On ADC images edema was also seen underneath the corpus callosum and later along the external capsule as hyperintense signal alteration. Volumetric quantification of susceptibility vessel volumes (D) and microhemorrhage volumes (E) are displayed in the bar charts both plotted as a function of disease severity (RMCBS on x-axis).
Table 1.
Imaging protocol.