Fig 2: C21orf91 ortholog overexpression accelerates morphological maturation of OPCs. OPCs were co-transfected with a GFP expression (for visualization, green) and an empty control vector (con, white bars) or a C21orf91 ortholog (RGD1563888) overexpression construct (ovex, black bars). Overexpression was confirmed using qRT-PCR (A) and immunostaining for C21orf91 (red, anti-C21orf91 antibody from Sigma-Aldrich) protein after 2 days of differentiation (B,C). Squares show the blow-ups of representative transfected cells which indicate that overexpressing cells in (C) exhibited enriched C21orf91 protein abundance. Additionally, their distinct morphological maturation is highlighted (compare cell morphologies in B,C) which was further assessed using process- and branch-dependent categories (low-high), indeed revealing an accelerated morphological maturation (D). Cell nuclei were labeled by DAPI (blue). Scale bars: 100 μm. Data are shown as mean values (±SEM) deriving from n = 3 experiments (A) and n = 8 experiments (D). Statistical significance was calculated using Mann-Whitney U two-tailed test (A) and Student’s two-sided, unpaired t-test (B): *p ≤ 0.05, **p ≤ 0.01. OPCs, despite their primarily designated development into myelinating oligodendrocytes (OL), can also differentiate into astrocytes (E).

Fig 3: Expression of human C21ORF91 gene. (A) The 50 tissues and cell types with the highest C21ORF91 expression according to Genevestigator measured with RNAseq and expressed as the log(2) of TPM (transcripts per kilobase million) mean values. Cortical myelinating oligodendrocytes and hippocampal astrocytes are the cell types, and corpus callosum and spinal cord are the tissues that are most enriched for C21orf91. (B) The 50 tissues and cell types with the highest C21ORF91 expression according to Genevestigator measured with microarrays and expressed as relative mean expression values. Corpus callosum, diencephalon, and cerebral white matter are the nervous system structures most enriched for C21orf91. (C) Heat map visualization of microarray expression data from six adult donors for the four available probes for C21ORF91 from the Allen Brain Atlas. Indicated with arrows are the structures with the highest expression. Data is displayed as z-scores. Probe 1–4 correspond to C21ORF91 sequence A_23_P211015, A_24_P125839, CUST_5965_PI416261804 and CUST_5966_PI416261804. (D) Schematic sagittal and inferior view of the human brain indicating the regions where C21ORF91 gene is expressed according to Genevestigator and Allen Brain Atlas. (E) Single-nucleus RNAseq data from Allen Brain Atlas expressed as count per million (cpm) trimmed means. MCA stands for multiple cortex area set of experiments and M1 for primary motor cortex experiment. L4–6, L6, L3–4, L2–6, L3–6, L4 refer to cortex areas. Oligo stands for oligodendrocytes, Exc for excitatory neuron, and Inh for inhibitory neuron. (F) C21ORF91 expression profile depicts a wave-like pattern, progressively increasing after birth to the age of 23 years. The expression is monitored for different brain structures at different ages. RNAseq data expressed as the mean of RPKM values. Ages vary from prenatal [8–37 post-conception weeks (pcw)] to postnatal stages from 4–10 months and 1–40 years. Brain structures were grouped as follows: allocortex (red) = hippocampus; basal ganglia (light blue) = striatum; neural plate (green) = 19 structures including frontal cortex and cerebellum; thalamus (dark blue) = dorsal thalamus, mediodorsal nucleus of thalamus. Data from the ventricular zone are not included. (G) Coexpression network of genes correlating with C21ORF91 expression and each other generated from BrainEXP database. Nodes are colored according to protein function: orange: myelination, purple: oligodendrogenesis, yellow: dendrite spine morphogenesis, green: other expression evidence in the nervous system, and gray: other functions.

Fig 4: C21orf91 ortholog overexpression results in aberrant OPC differentiation creating hybrid cells coexpressing astroglial and oligodendroglial markers. OPCs were co-transfected with a GFP expression (for visualization, green) and an empty control vector (con, white bars) or a C21orf91 ortholog (RGD1563888) overexpression construct (ovex, black bars). After 2 days (2d) of differentiation and C21orf91 overexpression, a significantly increased number of cells expressing GFAP (cyan) with oligodendroglial markers (red) Olig2 (A), Sox10 (B), and CC1 (C) were observed, representative images of which are shown in (A’–C’). Additional aberrant coexpression was also observed for GFAP (cyan) and Nkx2.2 (red; D) and for GLAST (cyan) with Olig2 (red; E). Arrows point at double-positive hybrid cells, arrowheads in (D) point to “normal” cells exhibiting a GFAP-positivity in the expected absence of oligodendroglial markers such as Nkx2.2. Scale bars: 100 μm. Data are shown as mean values (±SEM) deriving from n = 5–7 experiments. Mann-Whitney U two-tailed test: *p ≤ 0.05, **p ≤ 0.01.

Fig 5: C21orf91 ortholog overexpression leads to myelination failures and abnormal GFAP coexpression in myelin-producing oligodendrocytes. Immunofluorescence analysis of co-transfected OPCs with a GFP expression vector (for visualization, green) and an empty control vector (con, white bars) or a C21orf91 ortholog (RGD1563888) overexpression construct (ovex, black bars) revealed that after 3 days (3d) of differentiation the proportion of cells with strong MBP expression was increased upon C21orf91 overexpression (A). While control transfected cells exhibited MBP vesicle-like signals (red; B) C21orf91 overexpression resulted in stronger MBP signals (bold arrows) and myelin protein presence in cell processes (C). Thin arrows point at cells with weak MBP expression. Cell nuclei were labeled by DAPI (blue). To assess whether the observed elevated levels of produced MBP protein correlate with the establishment of functional myelin, transfected OPCs were transplanted onto and cultured for 10 days on dissociated co-cultures. By categorizing GFP-positive cells (green) into different MBP expression- and morphological-related stages (D), it was shown that C21orf91 overexpression resulted in fewer MBP expressing (pos) cells and also diminished their capacity to ensheath axons (myelin). They rather exhibited strong but non-organized MBP expression (NOM). Double immunostaining for MBP and GFAP (gray bars, blue stain) exposed NOM cells also to aberrantly coexpress both markers (E) which were confirmed via the orthogonal projection of a confocal, z-stacked microscopy image (F). Black arrows show which channels were merged; white dashed lines depict the y-axis cut through the z-stack which is shown in the white box on the right side of the images. Scale bars: 50 μm (B,C); 25 μm (D) and 20 μm (F). Data are shown as mean values (±SEM) deriving from n = 5 (A) and n = 9 experiments (D,E). Statistical significance was calculated using Kruskal-Wallis test with Dunn’s post-test (A) and Student’s two-sided, unpaired t-test (D,E): *p ≤ 0.05, **p ≤ 0.01, ns = not significant; #p ≤ 0.05 for the statistical analysis of GFAP + /MBP+ cells (gray bars) in (E).