The Cytoskeletal Elements MAP2 and NF-L Show Substantial Alterations in Different Stroke Models While Elevated Serum Levels Highlight Especially MAP2 as a Sensitive Biomarker in Stroke Patients

Abstract

In the setting of ischemic stroke, the neurofilament subunit NF-L and the microtubule-associated protein MAP2 have proven to be exceptionally ischemia-sensitive elements of the neuronal cytoskeleton. Since alterations of the cytoskeleton have been linked to the transition from reversible to irreversible tissue damage, the present study investigates underlying time- and region-specific alterations of NF-L and MAP2 in different animal models of focal cerebral ischemia. Although NF-L is increasingly established as a clinical stroke biomarker, MAP2 serum measurements after stroke are still lacking. Therefore, the present study further compares serum levels of MAP2 with NF-L in stroke patients. In the applied animal models, MAP2-related immunofluorescence intensities were decreased in ischemic areas, whereas the abundance of NF-L degradation products accounted for an increase of NF-L-related immunofluorescence intensity. Accordingly, Western blot analyses of ischemic areas revealed decreased protein levels of both MAP2 and NF-L. The cytoskeletal alterations are further reflected at an ultrastructural level as indicated by a significant reduction of detectable neurofilaments in cortical axons of ischemia-affected areas. Moreover, atomic force microscopy measurements confirmed altered mechanical properties as indicated by a decreased elastic strength in ischemia-affected tissue. In addition to the results from the animal models, stroke patients exhibited significantly elevated serum levels of MAP2, which increased with infarct size, whereas serum levels of NF-L did not differ significantly. Thus, MAP2 appears to be a more sensitive stroke biomarker than NF-L, especially for early neuronal damage. This perspective is strengthened by the results from the animal models, showing MAP2-related alterations at earlier time points compared to NF-L. The profound ischemia-induced alterations further qualify both cytoskeletal elements as promising targets for neuroprotective therapies.

Keywords: Atomic force microscopy; Biomarker; Cerebral ischemia; MAP2; NF-L; Stroke.

Fig. 1

Immunofluorescence labeling of MAP2 and NF-L at different time points in the mouse model of MCAO: the reduction of MAP2-related immunoreactivity (green) identifies the infarct border (dashed line) in the cortex of mice at all time points of transient and permanent ischemia. The NF-L-related immunoreactivity (red) is increased in the infarct area, with a clear increase after 24h and 72h, but no apparent increase after 4h-t. Scale bars: 200μm

Fig. 2

a The fluorescence intensity of MAP2 is regularly reduced, including early time points, but did not reach statistical significance after 4h of transient ischemia. b NF-L-related fluorescence intensities in the infarct area are significantly and progressively increased, but not yet after 4h of transient or permanent ischemia. Sample sizes: each group, n=5–6; except for 72h-p: n=1–2. Data are presented as mean values; error bars indicate the standard error. *p<0.05, **p<0.01, ***p<0.001

Fig. 3

a The illustration indicates regions of interest (ROIs) used for quantifications of MAP2- and NF-L-related immunofluorescence intensities after 24h of transient and permanent MCAO in mice. b When compared to the contralateral hemisphere, the MAP2-related immunofluorescence intensity was significantly reduced throughout the infarcted cortex, c whereas the increase of NF-L-related immunoreactivity was only significant in ROIs of the somatosensory cortex. Neither MAP2- nor NF-L-related immunofluorescence intensities differed significantly between transient and permanent MCAO. The dashed line represents the infarct border. Sample sizes in b and c: n=6 for all ROIs. Data are presented as mean values; error bars indicate the standard error. *p<0.05, **p<0.01, ***p<0.001

Fig. 4

a, b Western blot analyses of MAP2 and NF-L in mice after 4 and 24 h of transient and permanent MCAO: MAP2 levels decrease in stroke tissue throughout the applied models, although without reaching significance. c–f) Western blot was performed with a polyclonal “pab” NF-L antibody (c, e, f) and a monoclonal antibody “mab” (d). Thereby a decrease of NF-L in infarcted tissue was measured, which was most pronounced after 24h-p, but failed to reach statistical significance. NF-L levels after 4h-t remained mostly unaltered. e, f NF-L degradation products of 40-65 kDa are significantly increased after 24h-p. n=5–6. Data are presented as mean values; error bars indicate the standard error. **p<0.01, ***p<0.001

Fig. 5

a Electron microscopy images show cross sections of cortical axons after 4h and 24h of transient ischemia in mice and unaffected contralateral regions. For illustration, transversally sectioned axons are transparently highlighted in yellow. The regular distribution of axonal neurofilaments is lost (insets) 4h and 24h after tMCAO. An ischemia-induced cellular edema is indicated by a less electron dense cytoplasm in the highlighted axons and insets as well as in non-axonal structures (asterisks). Please note that the images and insets also contain other components of the axonal cytoplasm which were not considered for the analysis of neurofilament density. b The density of detectable neurofilaments is decreased in ischemia-affected cortical axons 4h and 24h after tMCAO. n=5. Data are presented as mean values; error bars indicate the standard error. *p<0.05. Scale bars: 1 μm

Fig. 6

The elastic modulus of infarcted brain tissue, an indicator of tissue elasticity, was measured in a cortex of a 24h-p mouse by AFM indentation. a Upper left image: Overview of the measured brain slice illustrates the localization of the measurements. The ischemia-affected tissue was identified by areas of FITC-albumin extravasation (green/white). To enhance the visibility of the contralateral hemisphere on behalf of figure generation, the image was acquired with fluorescence and transmitted light. Lower left image: Overview of MAP2- (green) and NF-L (red)-labeled coronal brain slice captured with epifluorescence microscopy to illustrate that the AFM measurements comply with the infarct border visualized by immunofluorescence labeling. Right image: Higher magnification of the inset (upper left image) showing an overlay of the infarct border and the heat map (blue corresponds to softer and red to stiffer regions) of the AFM measurements. b Over a span of 1100μm (positions 1–11), measurements revealed decreased values of tissue elasticity in central regions (1, 3, 4, 5), whereas peripheral regions (7, 8, 9, 10, 11) exhibited an elevated elastic modulus. Each region includes 100 individual measurements. Data are presented as mean values; error bars indicate the standard error. Scale bar: 500 μm

Fig. 7

MAP2 and NF-L serum levels of patients suffering from stroke at day 1 and day 3 after stroke onset: patients were divided into groups depending on a stroke intervention or b the size of the ischemic lesion. Patients with stroke (with and without intervention) have significantly higher serum concentrations of MAP2 than patients suffering from TIA. Significantly higher MAP2 serum levels are also found in patients suffering from larger infarcts compared to no ischemic lesion. Significant differences between day 1 and day 3 are only found after a lacunar stroke (MAP2 and NF-L). a and b: n=12–30. Bars represent mean values; dots represent individual values; error bars indicate the standard error. *p<0.05, **p<0.01, ***p<0.001