Fig. 2

CSF-tracer penetration in coronal slices 3 h after systemic LPS-injection (n = 14), compared with vehicle controls (n = 13). a Tracer intensity in coronal slices, at bregma 0 mm. FITC dextran, being a much smaller and not as easily fixed molecule compared to the previous tracer, appears blurrier in slice images. b Two-way ANOVA, BSA Alexa 647 mean pixel intensity in arbitrary units, MPI (a.u.), treatment difference P = 0.0002. At bregma 0 mm, tracer signal was significantly lower in LPS-mice, multiple comparison with Šidák correction. c Two-way ANOVA, area of FITC dextran MPI (a.u.), treatment difference P = 0.018. d Average BSA Alexa 647 tracer MPI (a.u.) per animal, normalized to vehicle mean. LPS 16% lower on average. Unpaired t test, 95% CI of difference = − 30 to − 2.8. e Average FITC dextran tracer MPI (a.u.) per animal, normalized to vehicle mean. LPS 18% lower on average, albeit not significantly so. Unpaired t test, 95% of difference = − 51 to 16. Mean pixel intensity in arbitrary units, MPI (a.u.). X-axis indicates distance to bregma. Means are shown as staples with error bars = SD (c, d), as well as mean with plotted values (e, f). Scale bars 2 mm and 500 μm. P values above comparative brackets. * indicate multiple comparison difference P < 0.05