167 Axonal Loss in Narcolepsy Type 1 reference value within the whole section. This allowed us to correct for variance in overall colour intensity between tissue sections. We hereafter exported all sections to ImageJ (version 1.53a) and extracted the mean colour intensities of the ROIs and the 95% maximum intensity reference values through pixel intensity histograms. Intensity ratios were calculated separately for each ROI by dividing the mean colour intensity of the ROI by the 95% maximum intensity value of the corresponding section. Besides these quantification analyses, we qualitatively inspected each section for the possible presence of axonal swelling or tortuousness, amyloid plaques and neurofibrillary tangles. Axonal injury The SMI312 staining was used to quantify the percentage of area stained with phosphorylated neurofilament chains (all nine ROIs were assessed). We utilized an in-house developed pixel classifier script (previously described by Frigerio et al. [252], for an example see Appendix A, Supplementary Figure 1) that identified phosphorylated neurofilament medium and heavy chains in axons of all nine ROIs (Table 1, Figure 1). Myelin integrity Colour intensity analyses were implemented on the two myelin integrity stainings (PLP and LFB) in the same way as described above for the Bielschowsky silver staining. Myelin stainings were also qualitatively checked for areas of demyelination or myelin swelling (defined as an enlarged spherical shape of myelin with clear inner presence of axonal staining) [242]. Statistical analysis Statistical analyses were performed in RStudio 2022.07.2+576. Matching of subjects was compared separately between the three narcolepsy type 1 donors and the two control groups using Student’s t-tests were used for continuous and ordinal data after testing for normality with the Shapiro-Wilk test. The Fisher exact test was used to compare sex distributions between groups. 6
RkJQdWJsaXNoZXIy MjY0ODMw