Enhancers of Human and Rodent Oligodendrocyte Formation Predominantly Induce Cholesterol Precursor Accumulation

Abstract

Regeneration of myelin in the central nervous system is being pursued as a potential therapeutic approach for multiple sclerosis. Several labs have reported small molecules that promote oligodendrocyte formation and remyelination in vivo. Recently, we reported that many such molecules function by inhibiting a narrow window of enzymes in the cholesterol biosynthesis pathway. Here we describe a new high-throughput screen of 1,836 bioactive molecules and a thorough re-analysis of more than 60 molecules previously identified as promoting oligodendrocyte formation from human, rat, or mouse oligodendrocyte progenitor cells. These studies highlight that an overwhelming fraction of validated screening hits, including several molecules being evaluated clinically for remyelination, inhibit cholesterol pathway enzymes like emopamil-binding protein (EBP). To rationalize these findings, we suggest a model that relies on the high druggability of sterol-metabolizing enzymes and the ability of cationic amphiphiles to mimic the transition state of EBP. These studies further establish cholesterol pathway inhibition as a dominant mechanism among screening hits that enhance human, rat, or mouse oligodendrocyte formation.

Figure 1.. High-throughput screen reveals novel and known EBP inhibitors and thyroid hormone analogs as potent enhancers of oligodendrocyte formation.

(A) Percentage of MBP+ oligodendrocytes generated from OPCs following treatment with a collection of over 1,800 bioactive small molecules at a uniform screening dose of 100 nM for 72 h. Red dots represent previously characterized enhancers of oligodendrocyte formation. The green dot represents clomifene. (B) Chemical structure of clomifene. (C) Percentage of MBP+ oligodendrocytes generated from OPCs following treatment with doses of clomifene for 72 h. n ≥ 4 wells per condition. (D) GC-MS based quantification of zymostenol and zymosterol levels in OPCs treated with clomifene at 300 nM for 24 h. n = 2 wells per condition. For (C), (D), DMSO and 1 μM TASIN-1, a known EBP inhibitor, served as the negative and positive control, respectively. (E) Representative images of OPCs treated with DMSO, clomifene at 300 nM, and TASIN-1 at 1 μM. Nuclei are labeled with DAPI (blue), and oligodendrocytes are shown by immunostaining for MBP (green). Scale bar, 100 μm. (C) and (D) represent two independent experiments.

Figure 2.. EBP and sterol 14-reductase are common enzyme targets in screening hits.

(A), (B) Percentage of MBP+ oligodendrocytes generated from OPCs following treatment with bioactive small molecules at 2 μM (A) and 6 μM (B). n=4 wells per condition, except DMSO and amorolfine, a known sterol 14-reductase inhibitor, n=8 wells, with >1,000 cells analyzed per well. (C), (D) GC-MS based quantification of 14-dehydrozymostenol, zymostenol, and zymosterol levels in OPCs treated for 24 h with TASIN-1 (1 μM), a known EBP inhibitor, and the indicated screening hits at the concentrations shown in panel (A) and (B). n=2 wells per condition. € Representative images of OPCs treated with amorolfine at 600 nM, toremifene citrate at 2 μM, and fenspiride HCl at 6 μM. Nuclei are labeled with DAPI (blue), and oligodendrocytes are shown by immunostaining for MBP (green). Scale bar, 100 μm. For percentage of MBP+ oligodendrocytes for all bioactive small molecules that were screened, see Fig. S2. S14R, sterol 14-reductase.

Figure 3.. Remyelinating agents inhibit CYP51, EBP, and sterol 14-reductase.

(A) Validated screening hits from Najm et al, 2015, with annotations of subsequently-established inhibitory activity for EBP, CYP51, or sterol 14-reductase. (B) GC-MS based quantification of lanosterol levels in OPCs treated with ketoconazole (2.5 μM), a known CYP51 inhibitor, and megestrol (13.4 μM) for 24 h. n2 wells per condition. (C) GC-MS based quantification of 14-dehydrozymostenol, zymostenol, and zymosterol levels in OPCs treated for 24 h with dicyclomine, propafenone, and haloperidol at 2 μM. S14R, sterol 14-reductase.

Figure 4.. Many enhancers of oligodendrocyte formation inhibit cholesterol pathway enzymes.

(A) Validated screening hits from Elitt et al., 2018 with annotations of subsequently-established inhibitory activity for EBP, CYP51, or sterol 14-reductase. (B) Percentage of MBP+ oligodendrocytes generated from OPCs following treatment with bioactive small molecules at a uniform dose of 10 μM. n=4 wells per condition, except DMSO and amorolfine, n=16 wells, with >1,000 cells analyzed per well. (C) GC-MS based quantification of 14-dehydrozymostenol, zymostenol, and zymosterol levels in OPCs treated for 24 h with the indicated screening hits at 10 μM. n=2 wells per condition. (D) Percentage of MBP+ oligodendrocytes in OPCs treated with climbazole in dose response as shown. (E) GC-MS based quantification of lanosterol levels in OPCs treated with climbazole in dose response as shown. n=2 wells per condition. For percentage of MBP+ oligodendrocytes for all bioactive small molecules that were screened, see Fig. S3. S14R, sterol 14-reductase.

Figure 5.. Top enhancers of both rat and mouse OPC differentiation inhibit EBP and sterol 14-reductase.

(A) Validated screening hits from Deshmukh et al., 2013 and Lariosa-Willingham et al., 2016 with annotations of subsequently-established inhibitory activity for EBP, CYP51, or sterol 14-reductase. (B), (D) Percentage of MBP+ oligodendrocytes generated from OPCs following treatment with bioactive small molecules. OPCs in (B) were treated with compounds at a uniform dose of 1.5 μM while those in (D) were treated in dose response as shown. n=4 wells per condition, except DMSO, amorolfine and TASIN-1, n=8 wells, with >1,000 cells analyzed per well. (C), (E) GC-MS based quantification of 14-dehydrozymostenol, zymostenol, and zymosterol levels in OPCs treated for 24 h with the indicated screening hits at their effective concentrations. n=2 wells per condition. (F) Percentage of MBP+ oligodendrocytes in OPCs treated with vanoxerine in dose response as shown. (G), (H), (I) GC-MS based quantification of zymostenol, zymosterol, and 14-dehydrozymostenol in OPCs treated with vanoxerine and donepezil in dose response for 24 h. n=2 wells per condition. Results in (D) represent two independent experiments. For percentage of MBP+ oligodendrocytes for all bioactive small molecules that were screened, see Fig. S4. S14R, sterol 14-reductase.

Figure 6.. Top molecules reported to enhance human oligodendrocyte formation inhibit EBP, CYP51, and S14R in human GBM528 cells.

(A) Validated screening hits from Li et al. 2022 with annotations of subsequently-established inhibitory activity for EBP, CYP51, or sterol 14-reductase. (B), (E), (F) GC-MS based quantification of zymostenol and zymosterol levels in GBM528 cells treated for 24 h with the indicated screening hits at 10 μM. (C) GC-MS based quantification of lanosterol levels in GBM528 cells treated for 24 h with clotrimazole at 10μM. (D), (G) GC-MS based quantification of 14-dehydrozymostenol levels in GBM528 cells treated for 24 h with the indicated molecules at 10 μM. n=2 wells per condition. (H) Percentage of MBP+ oligodendrocytes generated from OPCs following treatment with bioactive small molecules in dose response as shown. n=4 wells per condition, except DMSO and amorolfine, n=8 wells with >1,000 cells analyzed per well.

Figure 7.. Compounds in clinical trials inhibit EBP.

(A), (C), (E) Percentage of MBP+oligodendrocytes generated from OPCs following treatment with bioactive small molecules. OPCs in (A) were treated with quetiapine at 3 μM while those in (C) and (E) were treated with bazedoxifene and GSK 239512 in dose response as shown. n=4 wells per condition, except DMSO, amorolfine, and TASIN-1, n=8 wells, with >1,000 cells analyzed per well. (B), (D), (F) GC-MS based quantification of zymostenol, and zymosterol levels in OPCs treated for 24 h with the indicated screening hits at their effective concentrations. n=2 wells per condition. (G) Quantification of EBP enzymatic activity in a biochemical assay. All treatments 10 μM. n=2 independent enzymatic assays. Bars indicate mean and error bars represent s.d. Results in (C) represent two independent experiments.

Figure 8.. High logP hit molecules likely mimic sterol cation intermediate of highly druggable enzymes.

(A) Drug Score analysis performed in DoGSiteScorer using crystal structures of EBP and tamoxifen, EBP and U18666A, and CYP51. (B) logP values for all library molecules tested. Dots represent the screening hits found in fig. 2 (A), (B). (C) Box plot comparing logP values for hit molecules and molecules that did not increase oligodendrocyte maturation. Included in the left box are our previously reported screening hits in addition to those found in fig. 2 (A), (B). (D) Total number of hit molecules that have been classified as either a CYP51, sterol 14-reductase, EBP, or HSD17B7 inhibitor in mouse OPCs. (E) Total number of molecules characterized as either a CYP51, sterol 14-reductase, or EBP inhibitor in human GBM528 cells. (F) Energy diagram comparing the structure of the sterol cation intermediate within EBP’s enzymatic process to known EBP inhibitors. Red circles indicate tertiary amines likely to be cationic at physiological pH. S14R, sterol 14-reductase.