Accumulation of 8,9-unsaturated sterols drives oligodendrocyte formation and remyelination

Abstract

Regeneration of myelin is mediated by oligodendrocyte progenitor cells-an abundant stem cell population in the central nervous system (CNS) and the principal source of new myelinating oligodendrocytes. Loss of myelin-producing oligodendrocytes in the CNS underlies a number of neurological diseases, including multiple sclerosis and diverse genetic diseases^1-3. High-throughput chemical screening approaches have been used to identify small molecules that stimulate the formation of oligodendrocytes from oligodendrocyte progenitor cells and functionally enhance remyelination in vivo^4-10. Here we show that a wide range of these pro-myelinating small molecules function not through their canonical targets but by directly inhibiting CYP51, TM7SF2, or EBP, a narrow range of enzymes within the cholesterol biosynthesis pathway. Subsequent accumulation of the 8,9-unsaturated sterol substrates of these enzymes is a key mechanistic node that promotes oligodendrocyte formation, as 8,9-unsaturated sterols are effective when supplied to oligodendrocyte progenitor cells in purified form whereas analogous sterols that lack this structural feature have no effect. Collectively, our results define a unifying sterol-based mechanism of action for most known small-molecule enhancers of oligodendrocyte formation and highlight specific targets to propel the development of optimal remyelinating therapeutics.

Extended Data Figure 1.. Expanded cholesterol synthesis pathway diagram.

The cascade cyclization of squalene epoxide, catalyzed by lanosterol synthase (LSS), provides the first sterol, lanosterol. Processing of lanosterol to cholesterol can proceed via the Kandutsch-Russell and/or Bloch pathways, which use the same enzymes and process substrates that vary only in the presence or absence of the C24 double bond. Intermediates in blue have been confirmed in our GC/MS-based sterol profiling assay using authentic standards. Sterol 14-reductase activity in mouse is shared by two genes, TM7SF2 and LBR. Consistent with past reports (ref. 21), inhibition of sterol 14-reductase activity can lead to accumulation of the expected upstream intermediate (FF-MAS) or 14-dehydrozymostenol, also known as cholesta-8,14-dien-3-β-ol. Green indicates enzyme targets and small molecules whose inhibition promotes oligodendrocyte formation.

Extended Data Figure 2.. CYP51 is the functional target by which imidazole antifungals enhance oligodendrocyte formation.

a) Azole molecules with varying degrees of potency for mammalian CYP51 inhibition. Throughout, green labels indicate molecules considered active, while red labels indicate inactive molecules. b) Percentage of MBP⁺ oligodendrocytes generated from a second, independent derivation of OPCs (OPC-1) at 72 h following treatment with the indicated concentrations of azoles. n = 4 wells per condition except DMSO (n = 24), with > 1,000 cells analyzed per well. c) GC/MS-based quantitation of lanosterol levels in a second derivation of OPCs (OPC-1) treated 24 h with the indicated azoles at 2.5 μM. n = 2 wells per condition. d, e) GC/MS-based quantitation of cholesterol levels in OPCs (OPC-5 and OPC-1) treated 24 h with the indicated azoles at 2.5 μM. n = 2 wells per condition. f, g) GC/MS-based quantitation of lanosterol levels in OPCs (OPC-5, OPC-1) treated 24 h with the indicated doses of ketoconazole. n = 2 wells per condition. Concentrations shown in panels f and g mirror those shown in Fig. 1c and panel b. h) Percentage of MBP⁺ oligodendrocytes generated from mouse primary OPCs at 72 h following treatment with the indicated imidazole antifungals at 3 μM. n = 4 wells per condition, with > 1,000 cells analyzed per well. i) GC/MS-based quantitation of lanosterol levels in mouse primary OPCs treated 24 h with the indicated imidazole antifungals at 3 μM. n = 2 wells per condition. j) Assessment of oligodendrocyte formation using an alternative image quantitation metric, fold increase in total neurite length. Panel is a re-analysis of data shown in Fig. 1c. n = 4 wells per condition except DMSO (n = 24), with > 1,000 cells analyzed per well. k) Percentage of oligodendrocytes generated from OPCs at 72 h following treatment with ketoconazole (2.5 μM) as measured by PLP1 immunostaining. Left, OPC-5; right, OPC-1. n = 8 wells per condition, with > 1,000 cells analyzed per well. l) LC/MS-based quantitation of lanosterol levels in OPC-5 treated 24 h with ketoconazole at 2.5 μM. n = 2 wells per condition. m) CYP51 mRNA levels measured by RT-qPCR following 96-h treatment with non-targeting or CYP51-targeting pools of cell-permeable siRNAs. n = 2 wells per condition. n) GC/MS-based quantitation of lanosterol levels in OPC-1 treated 96 h with the indicated pooled siRNA reagents. n = 2 wells per condition. o) Percentage of MBP⁺ oligodendrocytes generated from a second, independent batch of OPCs (OPC-1) at 72 h following treatment with the indicated reagents. n = 3 wells per condition, with > 1,000 cells analyzed per well. p) Percentage of MBP⁺ oligodendrocytes generated from an independent derivation of OPCs at 72 h following treatment with exogenous lanosterol. n = 4 wells per condition except DMSO and Ketoconazole (n = 8), with > 1,000 cells analyzed per well. q) Representative images of OPC-5 cells treated 72 h with the indicated siRNA reagents and lanosterol. Nuclei are labeled with DAPI (blue), and oligodendrocytes are indicated by immunostaining for myelin basic protein (green). Scale bar, 100 μm. All bar graphs indicate mean ± standard deviation, panels b, d, h, i, k, l, o and p are representative of two independent experiments, and all findings have been confirmed in a second independent derivation of OPCs (Fig. 1).

Extended Data Figure 3.. Effect of small-molecule inhibition of the cholesterol biosynthesis pathway on enhancing oligodendrocyte formation.

a) GC/MS-based quantitation of sterol levels in OPCs (OPC-5) treated 24 h with the indicated inhibitors of cholesterol biosynthesis. Left, cholesterol; right, desmosterol. n = 2 wells per condition. Inhibitors were used at the following doses unless otherwise noted: mevastatin, ketoconazole, MGI-39, 2.5 μM; YM53601, 2 μM; Ro 48–8071, amorolfine, TASIN-1, 100 nM; AY9944, 200 nM. b) GC/MS-based quantitation of sterol levels in a second derivation of OPCs (OPC-1). Left, cholesterol; right, desmosterol. n = 2 wells per condition. c) GC/MS-based quantitation of the sterol intermediates expected to accumulate following treatment of OPCs with the indicated inhibitors of cholesterol biosynthesis for 24 h. n = 2 wells per condition. d) GC/MS-based quantitation of the sterol intermediates expected to accumulate following treatment of a second derivation of OPCs (OPC-1) with the indicated inhibitors of cholesterol biosynthesis for 24 h. n = 2 wells per condition. In c and d, no accumulation of other sterol intermediates indicative of off-target effects within the cholesterol pathway were observed (see Source Data). e) Representative images of OPC-5 cells treated 72 h with the indicated small molecules. All treatments are at the highest concentration shown in Fig. 2b. Scale bar, 100 μm. f) Percentage of MBP⁺ oligodendrocytes generated from a second batch of OPCs (OPC-1) at 72 h following treatment with the indicated cholesterol pathway inhibitors. n = 4 wells per condition, except DMSO, n = 24, with > 1,000 cells analyzed per well. g) Percentage of MBP⁺ oligodendrocytes generated from mouse primary OPCs at 72 h following treatment with the indicated cholesterol pathway inhibitors at 300 nM. n = 4 wells per condition, except DMSO, n = 8, with > 1,000 cells analyzed per well. h) GC/MS-based quantitation of sterol intermediate levels in mouse primary OPCs treated 24 h with the indicated inhibitors of cholesterol biosynthesis at 300 nM. Left, 14-dehydrozymostenol levels following treatment with amorolfine; Right, zymostenol levels following treatment with TASIN-1. n = 2 wells per condition. i, j) GC/MS-based quantitation of sterol intermediate levels in OPC-5 (i) and OPC-1 (j) treated 24 h with the indicated doses of inhibitors of cholesterol biosynthesis. Left, 14-dehydrozymostenol levels following treatment with amorolfine; Right, zymostenol levels following treatment with TASIN-1. n = 2 wells per condition. Concentrations shown in panel i mirror those shown in panel f. All bar graphs indicate mean +/− standard deviation, and panels a, c, e-h are representative of two independent experiments.

Extended Data Figure 4.. Effect of independent chemical-genetic and genetic modulators of CYP51, sterol 14 reductase and EBP on oligodendrocyte formation and cholesterol biosynthesis.

a, d, g) Percentage of MBP⁺ oligodendrocytes generated from two independent derivation of OPCs at 72 h following treatment with the indicated concentrations of medroxyprogesterone acetate (a), 2-methyl ketoconazole (d) or TASIN-449 (g). n = 4 wells per condition, except DMSO, n = 12 in a and d. In g, for OPC-5, n = 4 except DMSO, n = 7; for OPC-1, n = 3 except DMSO, n = 6. b, e, h) GC/MS-based quantitation of sterol levels in two independent derivations of OPCs treated 24 h with medroxyprogesterone acetate at 10 μM (a), 2-methyl ketoconazole at 2.5 μM (e) and TASIN-449 at the indicated concentrations (h). n = 2 wells per condition. c, f) Rat CYP51 enzymatic activity following treatment with varying concentrations of medroxyprogesterone acetate (c) and 2-methyl ketoconazole (f) as measured by LC/MS-based quantitation of the CYP51 product FF-MAS. n = 2 independent enzymatic assays. i) Percentage of MBP⁺ oligodendrocytes generated from OPCs (OPC-5) infected with lentivirus expressing Cas9 and an independent guide RNA targeting EBP (cf. Fig. 2c). 8 wells per condition, with > 1,000 cells analyzed per well. Two-tailed Student’s t-test, * P = 0.0009. j) Functional validation of CRISPR-based targeting of EBP with a second sgRNA using GC/MS-based quantitation of zymostenol levels. n = 2 wells per condition. k) EBP mRNA levels measured by RT-qPCR in OPCs (OPC-5) infected with lentivirus expressing Cas9 and either of two guide RNAs targeting EBP. One well per condition, with results validated in an independent experiment. l) Representative images of the oligodendrocyte formation assay shown in Main Figure 2c. Nuclei are labeled with DAPI (blue), and oligodendrocytes are indicated by immunostaining for myelin basic protein (green). Scale bar, 100 μm. All bar graphs indicate mean +/− standard deviation, and panels a, d, g, i and k are representative of two independent experiments.

Extended Data Figure 5.. Effect of 8,9-unsaturated sterols on oligodendrocyte formation.

a) Percentage of MBP⁺ oligodendrocytes generated from OPCs (OPC-5) at 72 h following treatment with methyl β-cyclodextrin (1 mM) for 30 min at 37 °C. n = 8 wells per condition, with > 1,000 cells analyzed per well. b) GC/MS-based quantitation of cholesterol (left) and desmosterol (right) levels in OPCs (OPC-5) treated with methyl β-cyclodextrin (Me-β-CD) at 1 mM or ketoconazole at 2.5 μM. n = 2 wells per condition. c, d) Percentage of MBP⁺ oligodendrocytes generated from OPC-1 (c) and OPC-5 cells (d) at 72 h following treatment with the indicated purified sterol intermediates. n = 4 wells per condition, except n = 8 for DMSO and ketoconazole, with > 1,000 cells analyzed per well. Green text highlights metabolites that accumulate after treatments that enhance oligodendrocyte formation (Fig. 2e, Extended Data Fig. 3c). e) Percentage of MBP+ oligodendrocytes generated from OPC­1 at 72 h following treatment with MAS-412 and MAS-414. n = 4 wells per condition, with > 1,000 cells analyzed per well. f) Representative images of OPC­5 cells treated 72 h with DMSO, MAS-412, or MAS-414 (3 μM). Nuclei are labeled with DAPI (blue), and oligodendrocytes are indicated by immunostaining for myelin basic protein (green). Scale bar, 100 μm. g) Percentage of MBP⁺ oligodendrocytes generated from OPC-1 at 72 h following treatment with 2,2-dimethyl-zymosterol. n = 4 wells per condition except DMSO (n = 12), with > 1,000 cells analyzed per well. h) Representative images of OPC-5 cells treated 72 h with vehicle and 2,2-dimethyl-zymosterol (2.5 μM). Nuclei are labeled with DAPI (blue), and oligodendrocytes are indicated by immunostaining for myelin basic protein (green). Scale bar, 100 μm. i) Percentage of MBP+ oligodendrocytes generated from OPC-5 (left) and OPC-1 (right) at 72 h following treatment with FF-MAS or T-MAS. n = 4 wells per condition except DMSO and Ketoconazole (n = 8), with > 1,000 cells analyzed per well. j) Percentage of MBP⁺ oligodendrocytes generated from OPC-5 and OPC-1 OPCs at 72 h following treatment with the indicated concentrations of cholesterol. n = 8 wells per condition, with > 1,000 cells analyzed per well. k, l) Percentage of MBP⁺ oligodendrocytes generated from OPC-5 and OPC-1 at 72 h following treatment with the indicated concentrations of sterols that are structurally identical aside from the presence or absence of the 8,9 double bond (structures in panel o). n ≥ 3 wells per condition (see dot plots as replicate values vary by condition), with > 1,000 cells analyzed per well. m) Percentage of MBP⁺ oligodendrocytes generated from OPCs (OPC-5) at 72 h following treatment with the indicated small molecules or combinations of small molecules (ketoconazole, 2.5 μM; Ro 48–8071, 11 nM; liothyronine, 3 μM). n = 3 wells per condition, except DMSO n = 11, ketoconazole n = 13, liothyronine n = 8 & liothyronine + Ro 48–8071 n = 4, with > 1,000 cells analyzed per well. n) GC/MS-based quantitation of lanosterol levels in OPCs (OPC-5) treated 24 h with the indicated small molecules or combinations of small molecules at concentrations stated in m. n = 2 wells per condition. o) Structures of zymostenol, 8,9-dehydrocholesterol, 5α-cholestanol, and cholesterol. p) Total cell number as measured by counting of DAPI+ nuclei in the experiment presented in panel m. q, r) Percentage of MBP+ oligodendrocytes generated from OPCs (OPC­5 and OPC-1) at 72 h following treatment with the indicated small molecules or combinations of small molecules in two independent batches of OPCs (ketoconazole, 2.5 μM; MAS­412, 5 μM). In q, n = 16 for DMSO, 8 for Ketoconazole, and 4 for remaining bars. In r, n = 8 wells per condition. s) Luciferase reporter assays were used to assess if 2,2-Dimethylzymosterol (5 μM), Ketoconazole (2.5 μM), and TASIN-1 (250 nM) modulate human ERα, GR, LXRβ, NFkB, NRF2, PGR, PPARδ, PPARγ, RARα, RARγ, RXRα, RXRβ, TRα, TRβ and VDR transcriptional activity in agonist mode and ERRα, RORα and RORγ in inverse-agonist mode. n = 2 wells per condition and n = 3 wells per positive control conditions. t) Effects of sterols (2,2-dimethylzymosterol 5 μM, FF-MAS 10 μM) and small molecules (Ketoconazole 2.5 μM, TASIN-1 100 nM) on the NR2F1-mediated activation of a NGFI-A promoter driven luciferase reporter. n = 2 wells per condition. u) Effects of 2,2-dimethylzymosterol (5 μM) on NR2C2-mediated activation of a NGFI-A promoter driven luciferase reporter in comparison to cells transfected with reporter only, untreated, or treated with a previously reported positive control (all-trans retinoic acid, ATRA, 5 μM). n = 2 wells per condition. v) LSS, DHCR7, LDLR mRNA levels measured by RT-qPCR following 24 h treatment with DMSO, Mevastatin (2.5 μM), Ro 48–8071 (500 nM), Ketoconazole (2.5 μM), TASIN-1 (100 nM), or Amorolfine (100 nM). n = 2 wells. All bar graphs indicate mean +/− standard deviation, and panels a-n, and t-v are representative of two independent experiments.

Extended Data Figure 6. Inhibiting CYP51, TM7SF2 and EBP is a unifying mechanism for many small-molecule enhancers of oligodendrocyte formation identified by high-throughput screening.

a) Percentage of MBP⁺ oligodendrocytes (relative to DMSO control wells) generated from OPCs (OPC-1 derivation) at 72 h following treatment with a library of 3,000 bioactive small molecules, each at 2 μM. Each dot represents the result for one small molecule in the library. Red, imidazole antifungals; blue, clemastine; green, EPZ005687, the top novel hit molecule which is discussed in Extended Data Fig. 7 below. b, c) Percentage of MBP⁺ oligodendrocytes generated from OPCs (Left: OPC-5; Right: OPC-1) at 72 h following treatment with ketoconazole, nine top molecules identified by bioactives screening (green), and nine randomly chosen library members (red) at a uniform dose of 5 μM. n = 4 wells per condition except DMSO and Ketoconazole, n = 12 wells, with > 1,000 cells analyzed per well. d) GC/MS-based quantitation of zymosterol, zymostenol, and 14-dehydrozymostenol levels in a second batch of OPCs treated 24 h with the indicated screening hits and randomly chosen library members at 2 μM. n = 1; for validation in a second derivation of OPCs, see Fig. 3a. Molecules are clustered by enzyme targeted (top labels). e) Percentage of MBP⁺ oligodendrocytes generated from OPCs at 72 h following treatment with the indicated doses of fulvestrant, one of the top 10 HTS hits. n = 4 wells per condition except DMSO, n = 12), with > 1,000 cells analyzed per well. f) GC/MS-based quantitation of lanosterol levels in OPCs treated 24 h with fulvestrant at 2 μM. n = 2 wells per condition. g, h, i) GC/MS-based quantitation of metabolite levels in OPCs treated 24 h with the indicated previously-reported enhancers of oligodendrocyte formation at the following doses: Benztropine, 2 μM; Clemastine, 1 μM; Tamoxifen, 100 nM; U50488, 5 μM; bexarotene, 1 μM; liothyronine, 3 μM. n = 2 wells per condition. j, k) Percentage of MBP⁺ oligodendrocytes generated from OPCs (OPC-5 left, OPC-1 right) at 72 h following treatment with the indicated previously-reported enhancers of oligodendrocyte formation. n = 4 wells per condition, except DMSO n = 20 for OPC-5 and n = 12 for OPC-1, with > 1,000 cells analyzed per well. All doses are μM. l) Representative images of OPCs treated 72 h with the indicated small molecules. All treatments in l are at the highest concentration shown in panel j. Scale bar, 100 μm. m) Structures of muscarinic receptor antagonists used in this study. n, q) Percentage of MBP⁺ oligodendrocytes generated from OPCs (OPC-5: top, OPC-1: bottom) at 72 h following treatment with ketoconazole or the indicated muscarinic receptor modulators at 2 μM, the concentration used during screening. n = 4 wells per condition except DMSO and Ketoconazole, n = 8, with > 1,000 cells analyzed per well. o) GC/MS-based quantitation of three metabolite levels in OPC-5 OPCs treated 24 h with U50488 (5 μM) or the indicated muscarinic receptor modulators (2 μM). Left, zymostenol; center, cholesterol; right, desmosterol. n = 2 wells per condition. p) Heatmap indicating inhibition of muscarinic receptor isoforms M1, M3, and M5 by the indicated small molecules (2 μM) assayed using GeneBLAzer NFAT-bla CHO-K1 cells. n = 2 wells per condition. r) GC/MS-based quantitation of three metabolite levels in OPC-1 OPCs treated 24 h with clemastine (1 μM) or the indicated muscarinic receptor modulators at 2 μM. n = 2 wells per condition. Left, zymostenol; center, zymosterol; right, cholesterol. p-Fluorohexahydro-sila-difenidol is abbreviated as Sigma H127. All bar graphs indicate mean +/− standard deviation, and panels b, c, e, i, j, k, n and q are representative of two independent experiments.

Extended Data Figure 7.. Effect of selective estrogen receptor modulators and EZH2 inhibitors on cellular EBP function and oligodendrocyte formation.

a) Structures of selective estrogen receptor modulators used in this study. b) Effects of ospemifene and toremifene on the estrogen-dependent growth of T47D cells. n = 3 wells per condition. c, d) Percentage of MBP⁺ oligodendrocytes generated from two independent batches of OPCs at 72 h following treatment with ospemifene and toremifene. n = 4 wells per condition except DMSO and Ketoconazole, n = 8, with > 1,000 cells analyzed per well. e) Representative images of OPCs treated 72 h with the indicated small molecules. All molecules were treated at 300 nM. Scale bar, 100 μm. f, g) GC/MS-based quantitation of two metabolite levels in OPCs treated 24 h with ospemifene and toremifene at 300 nM. Left, zymostenol; right, cholesterol. n = 2 wells per condition. h) Percentage of MBP⁺ oligodendrocytes generated from two independent batches of OPCs at 72 h following treatment with tamoxifen and 4-hydroxytamoxifen. Left, OPC-5; right, OPC-1. n = 4 wells per condition, except DMSO, n = 6 for OPC-1 (right). i) Effects of tamoxifen and 4-hydroxytamoxifen on the estrogen-dependent growth of T47D cells. n = 3 wells per condition. j) GC/MS-based quantitation of zymostenol (left axis) and zymosterol levels (right axis) in OPC-5 and OPC-1 treated 24 h with tamoxifen and 4-hydroxytamoxifen at the indicated concentrations. n = 2 wells per condition. Keto = Ketoconazole. k) Percentage of MBP⁺ oligodendrocytes generated from OPCs at 72 h following treatment with the indicated structurally analogous EZH2 inhibitors. n = 4 wells per condition, except DMSO, n = 12, with > 1,000 cells analyzed per well. l) Percentage of MBP⁺ oligodendrocytes generated from a second batch of OPCs at 72 h following treatment with the indicated structurally analogous EZH2 inhibitors. n = 4 wells per condition, except DMSO, n = 12, with > 1,000 cells analyzed per well. m) Percentage of MBP⁺ oligodendrocytes generated from mouse primary OPCs at 72 h following treatment with EPZ005687. n = 4 wells per condition, except DMSO, n = 12, with > 1,000 cells analyzed per well. n) Structure of EPZ005687 and structurally analogous EZH2 inhibitors. o) Representative images of OPCs treated 72 h with the indicated EZH2 inhibitors. All treatments are at 2 μM. Scale bar, 100 μm. p) GC/MS-based quantitation of two sterol intermediates following treatment of OPCs with the indicated EZH2 inhibitors at 1 μM for 24 h. Left, zymostenol; right, zymosterol. n = 2 wells per condition. q) GC/MS-based quantitation of two sterol intermediates following treatment of a second derivation of OPCs with the indicated EZH2 inhibitors at 1 μM for 24 h. Left, zymostenol; right, zymosterol. n = 2 wells per condition. r) GC/MS-based quantitation of two sterol intermediates following treatment of mouse primary OPCs with EPZ005687 at 2 μM for 24 h. Left, zymostenol; right, zymosterol. n = 2 wells per condition. All bar graphs indicate mean +/− standard deviation, and panels c, d, h, k-o, and r are representative of two independent experiments.

Extended Data Figure 8.. Effect of combinations of small-molecule treatments on oligodendrocyte formation, and ability of oligodendrocytes to track along and wrap electrospun microfibers after single small molecule treatments.

a, b) Percentage of MBP⁺ oligodendrocytes generated from OPCs (OPC-1: Left, OPC-5: Right) at 72 h following treatment with the indicated combinations of liothyronine and enhancers of oligodendrocyte formation. Unless noted, the following concentrations were used: ketoconazole, 2.5 μM; benztropine, 2 μM; clemastine 2 μM; tamoxifen 200 nM; liothyronine, 3 μM. n = 4 wells per treatment condition, with > 1,000 cells analyzed per well. Lio = liothyronine. c, d) Percentage of MBP⁺ oligodendrocytes generated from OPCs at 72 h following treatment with the indicated combinations of ketoconazole and enhancers of oligodendrocyte formation. n = 4 wells per treatment condition, with > 1,000 cells analyzed per well. Keto = ketoconazole.e) Representative images of OPCs treated 72 h with the indicated small molecules. Small-molecule concentrations are as in panel a. Scale bar, 100 μm. f) Fold-increase of MBP+ oligodendrocytes following plating of OPCs (OPC-5) onto microfibers and treatment for 14 days with the indicated pathway modulators. n = 2 wells per condition, except DMSO, n = 4. g) In an independent experiment, OPCs (OPC-5) were plated onto microfibers, treated with small molecules for 4 days, and fixed and stained after 14 days. The extent to which MBP+ oligodendrocytes tracked along the microfiber substrate was measured. n = 2 wells per condition. h) Total DAPI+ cell number in the experiment described in g. i) Representative images highlighting tracking along the microfiber substrate. Each image is a montage of 4 separate images within the same well. Green, MBP. Scale bar, 100 μm. j) High-resolution images of MBP+ oligodendrocytes tracking along microfibers. Green, MBP; Blue, DAPI. Ketoconazole, 2.5 μM. Scale bar, 50 μm. k) Confocal imaging of OPCs seeded onto aligned microfibers and treated 14 days with ketoconazole (2.5 μM). The plane of the cross-section is highlighted in yellow and the cross-section, in which green fluorescence appears to encircle several microfibers, is shown in the bottom panel. Green, MBP; Blue, DAPI. All bar graphs indicate mean +/− standard deviation, and panels a-d are representative of two independent experiments.

Extended Data Figure 9.. Effect of oligodendrocyte-enhancing small molecules on sterol levels in human cells and human cortical spheroids.

a) Representative images of toluidine blue stained sections of LPC-lesioned dorsal spinal cord from mice treated 8 days with ifenprodil (10 mg per kg) or tamoxifen (2 mg per kg). Scale bar, 20 μm. b) GC/MS-based quantitation of three metabolite levels in human glioma cells (GBM528) treated 24 h with the indicated small molecules at the following concentrations: Tamoxifen, 100 nM; Clemastine, 2 μM; Ifenprodil, 2 μM; Ketoconazole, 2.5 μM; Amorolfine, 100 nM. Left, lanosterol; center, zymostenol; right, 14-dehydrozymostenol. n = 2 wells per condition. c) GC/MS-based quantitation of three metabolite levels in two independent batches of human cortical spheroids treated 24 h with the indicated small molecules at 2 μM. Left, lanosterol; center, zymostenol; right, zymosterol. n = 3 spheroids per condition; representative of two independent experiments.

Extended Data Figure 10.. 27 small molecules and nine purified 8,9-unsaturated sterols shown here to enhance the formation of oligodendrocytes.

a) Schematic showing the proposed mechanism of action for enhancing oligodendrocyte formation by diverse small molecules which enhance oligodendrocyte formation. b) Molecules that enhance oligodendrocyte formation are grouped by enzyme inhibited (GCMS analysis in OPCs): CYP51, top; sterol 14-reductase, center; EBP, bottom. c) Purified 8,9-unsaturated sterols that enhance oligodendrocyte formation.

Figure 1.. Imidazoles inhibit CYP51 to enhance oligodendrocyte formation.

a) Rat CYP51 enzymatic activity following treatment with azoles. n = 2 independent enzymatic assays. b) GC/MS-based quantitation of lanosterol levels in OPCs treated with the indicated azoles at 2.5 μM. n = 2 wells per condition. c, f, g) Percentage of MBP⁺ oligodendrocytes generated from OPCs following treatment with azoles (c), cell permeable siRNA reagents (f), and lanosterol (g). n ≥ 4 wells per condition; for exact well counts in all figures, see Statistics and Reproducibility. In f, *, P = 0.0005, two-tailed Student’s t-test. d) Representative images of OPCs treated with the indicated azoles. Nuclei are labeled with DAPI (blue), and oligodendrocytes are indicated by immunostaining for myelin basic protein (green). Scale bar, 100 μm. e) GC/MS-based quantitation of lanosterol levels in OPCs treated with the indicated reagents. n = 2 wells per condition. h) Structure of lanosterol. All bar graphs indicate mean ± standard deviation. Experiments in c, d, and g are representative of three independent experiments, while b, e, and f are representative of two independent experiments using OPC-5 cells; for validation in an independent derivation of OPCs, see Extended Data Fig. 2.

Figure 2.. Small-molecule inhibition of CYP51, TM7SF2, or EBP enhances oligodendrocyte formation via accumulation of 8,9-unsaturated sterols.

a) Abbreviated cholesterol biosynthesis pathway. For greater detail, see Extended Data Figure 1. b) Percentage of MBP⁺ oligodendrocytes generated from OPCs treated with the indicated pathway inhibitors. n ≥ 4 wells per condition. c) Percentage of MBP⁺ oligodendrocytes generated from OPCs expressing Cas9 and guide RNA targeting EBP. n ≥ 3 wells per condition. d) Functional validation of Cas9-based targeting of EBP using GC/MS-based quantitation of zymostenol levels. n = 2 wells per condition. e, f, g) Percentage of MBP⁺ oligodendrocytes generated from OPCs with the indicated purified sterols. n ≥ 4 wells per condition. h) Structures of various sterols. All bar graphs indicate mean ± standard deviation. See Statistics and Reproducibility for exact well counts. Experiments in b-g are representative of two or more independent experiments using OPC-5 cells; for validation in an independent derivation of OPCs, see Extended Data Fig. 3–5.

Figure 3.. Inhibition of TM7SF2 and EBP is a unifying mechanism for many small-molecule enhancers of oligodendrocyte formation.

a) Quantitation of sterol levels in OPCs treated with the indicated molecules at 2 μM (one well per condition; for validation in a second derivation of OPCs, see Extended Data Fig. 6). (b) Quantitation of sterol levels in OPCs treated with the indicated previously-reported enhancers of oligodendrocyte formation (n = 2 wells per condition except DMSO, n = 6). Representative of two independent experiments. c) Quantitation of EBP enzymatic activity in a biochemical assay. All treatments 10 μM. n = 3 independent enzymatic assays, except DMSO, n = 5. Bars indicate mean; error bars indicate standard deviation. p-Fluorohexahydro-sila-difenidol abbreviated as Sigma H127.

Figure 4.. Accumulation of 8,9-unsaturated sterols enhances remyelination in vivo and in human brain spheroids.

a) Representative electron microscopy images of LPC-lesioned dorsal spinal cord from mice treated with ifenprodil or tamoxifen. Scale bar, 5 μm. b) Tukey plot showing quantitation of remyelinated axons in LPC-lesioned spinal cord from mice in panel a. n = 6 animals per group except vehicle, n = 4. **, P = 0.0004; *, P = 0.048, two-tailed Mann-Whitney test. c) Quantitation of brain sterol levels in mice treated with miconazole, ifenprodil, or tamoxifen. n = 4 animals per group. P = 0.0007 for miconazole, P = 0.0003 for ifenprodil, P = 0.0006 for tamoxifen, two-tailed Student’s t-test. d) Quantitation of MyRF+ oligodendrocytes within human myelinating cortical spheroids following treatment with miconazole (2 μM) or ifenprodil (2 μM). n = 4 spheroids per treatment condition. P = 0.0009 for miconazole and P = 0.0009 for ifenprodil, two-tailed Student’s t-test. e) Representative images of spheroids. DAPI+ nucelei (blue) and MyRF+ oligodendrocytes (red) are labeled. Scale bar, 100 μm. In c and d, bar graphs indicate mean and error bars indicate standard deviation.