Oligodendroglia Generate Vascular Mural Cells and Neurons in the Adult Mouse Brain

Abstract

Oligodendroglia encompass oligodendrocyte precursor cells (OPCs) and oligodendrocytes (OLs). In the grey matter of the cortex, nearly all OPCs divide slowly, yet they do not differentiate solely into mature OLs, leaving the exact role of these OPCs in the grey matter enigmatic. Oligodendroglia were traced using the sex-determining region Y-related high mobility group-box 10 (Sox10) Cre-ER^T2 reporter mice. We compared the effect of tamoxifen dissolved in different solvents on the fate of Sox10 cells. We also compared the effect of tamoxifen dosage on the fate of Sox10 cells. The differentiation of labeled red fluorescent protein (RFP) cells was analyzed using immunofluorescence staining. Two groups of RFP cells, type A Sox10 (Sox10-A) cells and type B Sox10 (Sox10-B) cells, were identified in the cortex, striatum, hippocampus, thalamus, and hypothalamus. Sox10-A cells differentiate into platelet-derived growth factor receptor-β+, CD13+ pericytes, and smooth muscle myosin heavy chain 11+ smooth muscle cells when the mice received ethanol or high-dose tamoxifen. Sox10-B cells transform into glutamatergic neurons when the mice received high-dose tamoxifen. Sox10-B cells include perineuronal OPCs and OLs. This investigation provides evidence that a substantial proportion of oligodendroglia in the grey matter serve as mural cell precursors and neuronal precursors. These two phenomena may contribute to our understanding of the fate of oligodendroglia.

INTRODUCTION

Pericytes play a crucial role in maintaining the integrity of the blood–brain barrier (BBB) as a specialized group of perivascular cells in the brain. These cells closely associate with endothelial cells, covering the vasculature at a ratio ranging from 1 : 1 to 1 : 3.¹⁾ Studies have demonstrated that pericytes are vital for the preservation of neuronal health, as their elimination has been shown to result in neuronal loss.²⁾ Alzheimer’s disease (AD) has been speculated to have vascular origins, and researchers have observed significant alterations in capillaries near the pericyte soma in AD patients.³⁾ Additionally, individuals with the apolipoprotein E4 variant, known to increase the risk of AD, exhibit a lower abundance of pericytes.⁴⁾ Gene expression studies have revealed that the mural cells of blood vessels, including pericytes, undergo the most substantial changes in AD patients.^5,6) In addition to their role in maintaining the BBB, pericytes are responsible for regulating brain blood flow and have been found to be more susceptible to cell death following ischemia compared to other neural cell types.⁷⁾ Recent studies have also indicated that pericytes are the first responders during neuroinflammation, contrasting with the previously believed role of microglia.^8,9)

Oligodendrocyte precursor cells (OPCs) represent the 4th type of glial cell in the brain. These slowly proliferating cells make up approximately 5% of all the neural cells in the brain. As the name implies, OPCs constantly differentiate into mature oligodendrocytes (OLs).¹⁰⁾ In recent years, OPCs have been found to be very active, similar to microglia.¹¹⁾ Neuron-glia antigen-2 (NG2), platelet-derived growth factor receptor-α (PDGFRα), and OL lineage transcription factor 2 (Olig2) have been used for the identification of OPCs. They are also referred to as NG2 glia instead of NG2+ cells, which include pericytes and vascular smooth muscle cells (SMCs).^12,13) In white matter, OPC processes align parallel to axons, while in grey matter, they exhibit a radial morphology similar to microglia.¹⁴⁾ Interestingly, OPCs in the grey matter do not appear to differentiate solely into mature OLs.¹⁵⁾ These cells are consistently found in close proximity to vascular endothelial cells,¹⁶⁾ migrating along blood vessels,¹⁷⁾ with some OPCs even establishing contact with nearby blood vessels through their processes.¹⁸⁾ The debate surrounding whether OPCs differentiate into mature neurons has persisted for a considerable duration.^19–21) In a recent paper, it was demonstrated that OL lineage cells can transfer nuclear and ribosomal material to neurons,²²⁾ a phenomenon believed to take place in healthy brains.²³⁾ Despite the progressive strides made in comprehending the diversity among OPCs, the exact functions and offspring of these cells continue to remain enigmatic.

Sex-determining region Y-related high mobility group-box 10 (Sox10) has been identified as being exclusively expressed in OPCs, myelinating OLs, and newly formed OLs in the adult mouse brain.²⁴⁾ It serves as a more specific marker for the OL lineage compared to NG2, PDGFRα, and Olig2.¹⁴⁾ Given the unknown physiological functions of OL lineage cells, an inducible Sox10-Cre tracing system was utilized in our study. This approach led us to discover a unique cluster of oligodendroglia in the mouse brain that assumed the role of vascular mural cells under alcoholic conditions. Furthermore, our investigation demonstrated that the oligodendroglia-to-neuron conversion phenomenon observed in lineage tracing arises from tamoxifen toxicity, rather than being a healthy occurrence.

MATERIALS AND METHODS

Animals

Sox10 Cre-ER^T2 (#027651, Jackson Laboratory, Bar Harbor, ME, U.S.A.) and Ai9 (#007909, Jackson Laboratory) reporter mice were crossed, and male F1 generation (Sox10 Cre-ER^T2; Ai9) mice at 8 weeks of age were employed for induction of red fluorescent protein (RFP) using tamoxifen with various solvents, as specified in Results. Sox10 Cre-ER^T2; Ai9 mice without tamoxifen and Ai9 heterozygous mice served as controls. Aldh1l1 Cre-ER^T2 mice (#T052693) were kindly gifted by GemPharmatech Inc. (Nanjing, China) and crossed with Ai9 reporter mice to generate Aldh1l1 Cre-ER^T2; Ai9 mice. Male F1 generation mice at 8 weeks of age were used for RFP induction with tamoxifen.

Ethics Statement

The animal study was reviewed and approved by the Animal Ethics Committee of Zhejiang Ocean University (SCXK ZHE2019-0031, #2021011, #2025059).

Immunofluorescence Staining

All mice were sacrificed using carbon dioxide and subsequently underwent transcardial perfusion with saline, followed by 4% paraformaldehyde. The brain tissues were then fixed in 4% paraformaldehyde at 4°C for 1 h. Vibratome sectioning (#ZQP-86, Zhisun Equipment Inc., Shanghai, China) was performed after embedding the brain tissues in low-melting-point agarose. Free-floating sections were permeabilized with 0.3% Triton X-100 and blocked with 1% bovine serum albumin in a 24-well cell culture plate. For immunofluorescence staining, the tissues were incubated overnight at 4°C with primary antibodies targeting specific proteins: cluster of differentiation 13 (CD13, #GTX75927, Genetex, Irvine, CA, U.S.A.), Sox10 (#AF2698, Beyotime, Shanghai, China), PDGFRβ (#AF1042, R&D Systems, Minneapolis, MN, U.S.A.; #14-1402-82, Thermo Fisher, Waltham, MA, U.S.A.), smooth muscle myosin heavy chain 11 (MYH11, #ab224804, Abcam, Cambridge, U.K.), NG2 (#ab5320, Millipore, Bedford, MA, U.S.A.; #ab129051, Abcam), aquaporin-4 (AQP4, #59678, CST, Danvers, MA, U.S.A.), PDGFRα (#558774, BD Biosciences, Franklin Lakes, NJ, U.S.A.), Ki67 (#12202, CST; #ab15580, Abcam), neuronal nuclei (NeuN, #266004, Synaptic Systems, Göttingen, Germany), microtubule-associated protein 2 (MAP2, #8707, CST), RFP (#600-401-379, Rockland, Limerick, PA, U.S.A.), gamma-aminobutyric acid (GABA, #A2052, Millipore), Reelin (#ab312310, Abcam), the proto-oncogene c-Fos (#2250, CST), doublecortin (DCX, #4604, CST), vesicular glutamate transporter 2 (VGLUT2, #MAB5504, Millipore), adenomatous polyposis coli (clone CC1, #GTX16794, Genetex), activating transcription factor 3 (ATF-3, #HPA001562, Millipore), Tau-1 (#MAB3420, Millipore), heat-shock protein 70 (Hsp70, #ab181606, Abcam), neurofilament (NF-M, #OB-MMS051-02, Oasis Biofarm, Zhejiang, China), γ-histone 2AX (γ-H2AX, #C2035S, Beyotime), cluster of differentiation 31 (#550274, BD Biosciences, Franklin Lakes, NJ, U.S.A.), glial fibrillary acidic protein (GFAP, #D262817-0025, Sangon Biotech, Shanghai, China), and Sox9 (#OB-PRB049-01, Oasis Biofarm). After 3 washes with phosphate-buffered saline, immunofluorescence secondary antibodies (#705-095-147, #112-095-003, #111-095-003, #115-095-003, and #706-095-148, Jackson ImmunoResearch, West Grove, PA, U.S.A.) were applied at room temperature for 1 h. Finally, the nuclei were counterstained with 4′,6-diamidino-2-phenylindole and the brain sections were imaged using a fluorescence microscope (Olympus, Tokyo, Japan, BX41).

Statistical Analysis

Statistical analysis was performed using GraphPad Prism 9.0. In all figures, data are represented as mean ± standard error of the mean. Significance was analyzed using one-way ANOVA followed by the Bonferroni post hoc test or an unpaired t-test. A p value <0.05 was considered statistically significant.

RESULTS

Sox10 Cells Differentiate into Pericytes and SMCs in the Brains of Adult Mice

To investigate the role of OL lineage cells in the mouse brain, we employed Sox10 Cre-ER^T2; Ai9 mice for fate mapping of these cells. After two days of tamoxifen administration (intraperitoneal injection of tamoxifen dissolved in sunflower seed oil at a daily dose of 40 mg/kg), followed by a 5-day interval, RFP+ oligodendroglia were found evenly distributed throughout the whole brain (Fig. 1A and Supplementary Fig. S5A). This group was designated as the 7D group. In a few brain sections of the 7D group, we observed the emergence of blood vessel-like RFP cells (1.85 ± 0.21%) in the cortex, which resembles vascular mural cells (n = 4 mice; Figs. 1B, 1C, 1L, Supplementary Figs. S5B and S5C). These cells were identified as either pericytes (11.82 ± 0.78% PDGFRβ+ and 11.71 ± 1.06% CD13+) or SMCs (25.54 ± 2.03% MYH11+) (Figs. 1D–1F, 1M–1O and Supplementary Figs. S5D–S5F). Additionally, AQP4 staining was employed to identify BBB-covered capillaries,²⁵⁾ revealing that 82.73 ± 1.88% of the converted cells were pericytes of the BBB (Fig. 1G and Supplementary Fig. S5G). As the Ai9 reporter mice have been reported by the Jackson Laboratory website to have very low levels of RFP expression prior to Cre introduction, we examined RFP expression in control mice. A few weak RFP signals were observed in the endothelial cells of both Ai9 heterozygous mice and Sox10 Cre-ER^T2; Ai9 mice without tamoxifen, but RFP expression in blood vessel-like cells was significantly higher in the 7D group (Supplementary Fig. S4). To confirm whether this small group of cells is continuously differentiating, mice were administered tamoxifen using the same strategy as in the 7D group, and the brain tissue was harvested 28 d later. This group was designated as the 1M group. We found that the 1M group had a percentage of blood vessel-like RFP cells of 1.56 ± 0.08%, along with 10.81 ± 1.63% PDGFRβ+, 11.69 ± 0.31% CD13+, and 21.67 ± 2.04% MYH11+ RFP cells, which were comparable to those in the 7D group (n = 4 mice, Figs. 1H, 1L–1O, Supplementary Fig. S5H). This indicated that the small amount of oligodendroglia-to-mural-cell conversion is not time course-dependent.

Fig. 1. Fate Mapping of Sox10+ Cells in the Mouse Brain

A small group of blood vessel-like RFP cells appeared in the 7D group (A–C), which was further confirmed by the immunofluorescent staining for PDGFRβ (D), CD13 (E), MYH11 (F), and AQP4 (G). RFP+ cells in the cortex of the 1M group (H). Blood vessel-like RFP cells in the cortex, striatum, and thalamus of the water group (I–K). Percentages of blood vessel-like RFP cells (L, n = 8 sections from 4 mice in each group), PDGFRβ+ RFP cells (M, n = 4 mice in each group), CD13+ RFP cells (N, n = 4 mice in each group), and MYH11+ RFP cells (O, n = 4 mice in each group) in the different groups. Significance analysis was performed using one-way ANOVA followed by the Bonferroni post hoc test. Scale bars: 80 μm for (A) and 20 μm for (B–K). Data are represented as mean ± S.E.M. ^*p < 0.05, ^**p < 0.01, and ^***p < 0.001. Ctx: cortex; Hip: hippcampus; Hth: hypothalamus; Str: striatum; Tha: thalamus.

We subsequently administered tamoxifen dissolved in ethanol via drinking water (50 mg of tamoxifen in 250 mL of a 1% ethanol solution)²⁶⁾ and collected brain tissue 14 d later, referred to as the water group. Strikingly, the presence of RFP-converted mural cells increased significantly (6.68 ± 0.36% blood vessel-like RFP cells, 25.32 ± 1.71% PDGFRβ+, 39.04 ± 1.63% CD13+, and 58.33 ± 4.81% MYH11+) in the cortex of the water group (n = 4 mice; Figs. 1I, 1L–1O, Supplementary Figs. S1A–S1C and S5I). Furthermore, as shown in Figs. 1J and 1K, in the water group, these converted cells were also observed in the striatum (5.62 ± 0.26% blood vessel-like RFP cells, 10.19 ± 1.10% PDGFRβ+, 19.34 ± 1.36% CD13+, and 27.92 ± 3.29% MYH11+ RFP cells; Figs. 1J, 1L–1O and Supplementary Fig. S5J) and the thalamus (1.45 ± 0.13% blood vessel-like RFP cells, 12.58 ± 1.05% PDGFRβ+, 4.18 ± 0.31% CD13+; Figs. 1K–1O and Supplementary Fig. S5K). We conducted RFP staining and observed that these converted RFP cells were not artifacts in either the 7D group or the water group (Supplementary Figs. S1D and S1E). These Sox10+ RFP cells, possessing the capacity for oligodendroglia-to-mural-cell conversion, were designated as type A Sox10 (Sox10-A) cells.

A Group of Sox10-A Cells Continue to Express NG2, but Not Sox10 or PDGFRα

As Sox10-A cells have the potential to differentiate into mural cells, we investigated whether these converted mural cells retain the expression of NG2, given that NG2 is frequently employed for OPC labeling and is sometimes used for labeling mural cells as well.²⁷⁾ We found that a part of converted Sox10-A cells maintained NG2 expression (32.42 ± 1.53% NG2+ RFP cells in the 7D group and 49.87 ± 2.16% in the water group; Figs. 2A, 2B, 2I, Supplementary Figs. S6A and S6B), while co-localization of Sox10 or PDGFRα expression with Sox10-A mural cells was not detected (Figs. 2C–2F and Supplementary Figs. S6C–S6F). In addition, 4.28 ± 0.14% Ki67+ proliferating RFP cells were found in the cortex of the water group, and 1.70 ± 0.11% Ki67+ proliferating RFP cells were found in the cortex of the 7D group (arrows, Figs. 2G, 2H, 2J, Supplementary Figs. S6G and S6H). These results partially elucidate the reason why the glycoprotein NG2 is present on both OPCs and mural cells. This occurrence can be attributed to the fact that at least some Sox10-A cells continue to express NG2 even after differentiating into mural cells.

Fig. 2. Representative Images of the Immunofluorescent Staining for OPC-Related Markers

Sox10-A cells co-localized with NG2 in the 7D group (A) and the water group (B), but did not co-localize with Sox10 (C, D) or PDGFRα (E, F). Ki67 immunostaining in the 7D group and the water group (G, H). Percentages of NG2+ Sox10-A cells and proliferating cells in the different groups (I, J). The arrows depict Ki67+ RFP cells. n = 4 mice in each group. Significance analysis was performed using an unpaired t-test. Scale bar = 20 μm. Data are represented as mean ± S.E.M. ^***p < 0.001 and ^****p < 0.0001.

Oligodendroglia-to-Mural-Cell Conversion: Influence of Ethanol and Tamoxifen

To determine whether oligodendroglia-to-mural-cell conversion is associated with ethanol exposure, we administered tamoxifen to mice using either pure sunflower oil or a mixture of sunflower oil and ethanol in a ratio of 9 : 1 via gavage (40 mg/kg per day) for three days. Subsequently, the mice were sacrificed 7 d after the initial tamoxifen dosage. The 2 respective groups were designated as the oil group and the oil/ethanol group.

In the oil group, Sox10-A cells constituted 1.46 ± 0.13% of all RFP cells (accounting for 10.62 ± 0.58% CD13+ pericytes and 23.06 ± 1.21% MYH11+ SMCs). This proportion was significantly higher in the oil/ethanol group, where Sox10-A cells represented 2.90 ± 0.16% of all RFP cells (accounting for 18.89 ± 0.49% CD13+ pericytes and 33.54 ± 1.54% MYH11+ SMCs), and this distribution was exclusively observed in the cortex (n = 4 mice; Figs. 3A, 3B, and 3D–3F and Supplementary Figs. S2A, 2B, 2D, 2E and S7A, 7B). These findings indicate that ethanol facilitates oligodendroglia-to-mural-cell conversion.

Fig. 3. Oligodendroglia-to-Mural-Cell Conversion in the 3 Groups

Representative images of Sox10-A cells in the cortex of the oil group (A), the oil/ethanol group (B), and the high-dose group (C). Percentages of Sox10-A cells (D, n = 8 sections from 4 mice in each group), CD13+ Sox10-A cells (E, n = 4 mice in each group), and MYH11+ Sox10-A cells (F, n = 4 mice in each group) in the 3 groups. Significance analysis was performed using one-way ANOVA followed by the Bonferroni post hoc test. Scale bar = 20 μm. Data are represented as mean ± S.E.M. ^**p < 0.01, ^***p < 0.001, and ^****p < 0.0001. Ctx: cortex; HD: high dose; Hip: hippocampus; Hth: hypothalamus; Str: striatum; Tha: thalamus.

In addition, the mice received high doses of tamoxifen via gavage (160 mg/kg daily, sunflower seed oil : ethanol = 9 : 1) for 3 d, and brain tissue was harvested four days later.²⁸⁾ This group was designated as the high-dose group. Oligodendroglia-to-mural-cell conversion was significantly increased compared with the oil group and the oil/ethanol group, suggesting that high doses of tamoxifen also contribute to the conversion (4.26 ± 0.18% Sox10-A cells, 40.42 ± 2.22% CD13+ pericytes, and 35.00 ± 1.67% MYH11+ SMCs) (n = 4 mice; Figs. 3C–3F, and Supplementary Figs. S2C, S2F and S7C). These converted cells were also observed in the striatum (3.53 ± 0.14% Sox10-A cells, 28.89 ± 0.91% CD13+ pericytes, and 24.58 ± 3.15% MYH11+ SMCs; Figs. 3D–3F).

Oligodendroglia-to-Neuron Conversion Is Associated with Tamoxifen Dosage

Unexpectedly, RFP cells displaying large cell bodies and long projections were observed in layer 2/3 of the cortex (7.93 ± 0.37% of RFP cells) in the high-dose group, as well as in the piriform cortex where the OPC-to-neuron phenomenon has previously been documented (Fig. 4A and Supplementary Fig. S3A) (Rivers et al., 2008). These cells were also found in the granule cells of the dentate gyrus (DG), striatum, thalamus, and hypothalamus of the high-dose group, accounting for 3.77 ± 0.20, 0.25 ± 0.01, 0.35 ± 0.05, and 0.28 ± 0.02% of RFP cells, respectively. Notably, these cells exhibited neuronal markers NeuN, MAP2, Tau-1, and NF-M. (n = 4 mice, arrows; Figs. 4A–4E and 4O). Conversely, RFP+ neurons were absent in the oil group (referred to as the low-dose group for convenience) and in the 1M group (Figs. 4F, 4G and Supplementary Fig. S3B). RFP expression was confirmed in the RFP cells of the low-dose group and in the RFP neurons of the high-dose group (Figs. 4H and 4I). These Sox10+ RFP cells, possessing the capacity for oligodendroglia-to-neuron conversion, were designated as type B Sox10 (Sox10-B) cells.

Fig. 4. Sox10-B Cells Transform into Neurons

Representative images of immunostaining for NeuN, MAP2, Tau-1, and NF-M in layer 2/3 of the cortex of the high-dose group (A–D). NeuN immunostaining results in the DG of the high-dose group (E). NeuN immunostaining results in the cortex of the low-dose group (F) and in the cortex of the 1M group (G). Expression of RFP (H, I) in the low-dose group and high-dose group. Immunostaining for DCX (J), c-Fos (L), and VGLUT2 (N) in the DG, and for c-Fos (K) and VGLUT2 (M) in the cortex of the high-dose group. The percentage of RFP+ neurons in the low-dose and high-dose groups (O, n = 4 mice in each group). The percentage of VGLUT2+ Sox10-B cells in the cortex and DG of the high-dose group (P, n = 4 mice in each group). The arrows indicate Sox10-B cells. Significance analysis was conducted using one-way ANOVA followed by the Bonferroni post hoc test or an unpaired t-test. Scale bar = 20 μm. Data are presented as mean ± S.E.M. ^***p < 0.001 and ^****p < 0.0001. Ctx: cortex; HD: high dose; Hth: hypothalamus; LD: low dose; Str: striatum; Tha: thalamus.

Additionally, to explore the possible connection of the phenomenon with neural stem cells, immunostaining for DCX was conducted. RFP neurons did not show co-localization with DCX in the DG or piriform cortex (Fig. 4J and Supplementary Fig. S3C), indicating that Sox10-B cells may not be associated with stem cell-related neurogenesis. The formation of Sox10-B neurons is not associated with neuronal excitation, as validated by c-Fos staining (Figs. 4K and 4L). These neurons were predominantly identified as glutamatergic, positive for VGLUT2 in the cortex (84.76 ± 1.49%) and DG (78.14 ± 3.44%), but negative for GABA or Reelin (arrows, Figs. 4M, 4N, and 4P and Supplementary Figs. S3D–3G). Therefore, Sox10-B neurons are pyramidal neurons in the cortex or granule neurons in the DG.

Sox10-B Cells: Perineuronal OLs

Perineuronal OLs occupy specific positions and maintain close contact with neurons.^22,29,30) The proportion of perineuronal oligodendroglia was 24.14 ± 1.75 and 11.11 ± 1.24% in the cortex, 15.14 ± 1.15 and 5.26 ± 0.37% in the DG, 23.80 ± 1.25 and 16.69 ± 0.31% in the striatum, 15.96 ± 1.06 and 13.85 ± 0.65% in the thalamus, and 20.14 ± 1.62 and 14.28 ± 1.11% in the hypothalamus of the low-dose group and the high-dose group, respectively (arrows, Figs. 5A–5D and 5I). The number of PDGFRα+ perineuronal RFP cells was 28 pairs in the low-dose group and 8 pairs in the high-dose group (Fig. 5J). The number of CC1+ perineuronal RFP cells was 22 pairs in the low-dose group and 7 pairs in the high-dose group (Fig. 5K). This indicates that Sox10-B cells include OPCs and OLs.

Fig. 5. Perineuronal Oligodendroglia in the Mouse Brain

Representative images of perineuronal oligodendroglia in layer 2/3 of the cortex and the DG in the low-dose group and high-dose group (arrows, A–D). Immunostaining for ATF-3 in layer 2/3 of the cortex and in the DG in the high-dose group and the 2D group (E–H). Arrows indicate ATF-3+ neurons. Percentages of perineuronal oligodendroglia in the low-dose group and high-dose group (I). Percentages of PDGFRα+ OPCs and CC1+ OLs in the low-dose group and high-dose group (J, K). Significance analysis was performed using one-way ANOVA followed by the Bonferroni post hoc test or an unpaired t-test. Scale bar = 20 μm. Data are represented as mean ± S.E.M. ^**p < 0.01, ^***p < 0.001, and ^****p < 0.0001. Ctx: cortex; HD: high dose; Hth: hypothalamus: LD: low dose; Str: striatum; Tha: thalamus. n = 4 mice in each group.

To assess whether oligodendroglia-to-neuron conversion involves neuronal stress, we conducted ATF-3, γ-H2AX, and Hsp70 immunostaining. We found that γ-H2AX and Hsp70, but not ATF-3, were expressed in the neurons in the cortex or DG of the high-dose group (Figs. 5E and 5F and Supplementary Fig. S8). Subsequently, we administered a single dose of tamoxifen (160 mg/kg, sunflower oil : ethanol = 9 : 1), and mouse brains were harvested 48 h later, designated as the 2D group. High levels of ATF-3 expression were observed in the pyramidal neurons of the cortex and granule cells of the DG (indicated by arrows, Figs. 5G and 5H). Therefore, the initiation of neuronal injury and the absence of perineuronal oligodendroglia appear to be linked to oligodendroglia-to-neuron conversion, and these missing perineuronal oligodendroglia could be Sox10-B cells.

High-Dose Tamoxifen Does Not Induce Astrocyte-to-Neuron Conversion

To determine whether high-dose tamoxifen induces astrocyte-to-neuron conversion, we administered tamoxifen via gavage to Aldh1l1 Cre-ER^T2; Ai9 mice to label astrocytes. After three days of tamoxifen administration (160 mg/kg daily, sunflower seed oil : ethanol = 9 : 1), followed by a 4-d interval, mouse brains were harvested and sectioned for immunofluorescence analysis. No RFP+ astrocytes were found to be positive for NeuN in the cortex (Figs. 6A and 6E). However, 91.60 ± 2.85 and 90.80 ± 3.19% of RFP+ astrocytes were positive for Sox9 and AQP4, 2 markers for astrocytes and astrocytic endfeet (Figs. 6B, 6C, and 6E). Additionally, 79.81 ± 4.60% of RFP+ astrocytes were GFAP+ reactive astrocytes (Figs. 6D and 6E). Unlike oligodendroglia, astrocytes do not transdifferentiate into neurons under high-dose tamoxifen treatment.

Fig. 6. Fate Mapping of Astrocytes in Aldh1l1 Cre-ER^T2; Ai9 Mouse Brains

Representative images show immunostaining for NeuN, Sox9, AQP4, and GFAP in the cortex (A–D). Arrows indicate astrocytic endfeet. The percentages of RFP+ cells co-localized with NeuN, Sox9, AQP4, and GFAP are shown (E). n = 3 mice in each group. Scale bar = 20 μm. Data are represented as mean ± S.E.M.

DISCUSSION

In this study, we have identified 2 distinct populations of Sox10 cells: one comprising mural cell precursors called Sox10-A cells, which have the potential to transform into vascular mural cells, and another consisting of cells termed Sox10-B cells, which can differentiate into pyramidal neurons in the cortex and granule neurons in the DG. The transition of Sox10-A cells to mural cells is attributed to the use of ethanol as a solvent for tamoxifen and the dosage of tamoxifen. The transformation of Sox10-B cells into neurons is associated with the toxicity of tamoxifen, especially at high dosages. However, unlike its effects on oligodendroglia, high-dose tamoxifen does not promote astrocyte-to-neuron conversion.

The substitution of pericytes in this study may be attributed to pericyte injury.³¹⁾ In our pilot study, wild-type mice were intraperitoneally injected with alcohol for 30 min, and during this period, some pericytes in the cortex were labeled with intravenously injected propidium iodide (data not shown). This labeling indicated the cell membrane leakage of pericytes after alcohol treatment. In a previous study, we observed inflammation-related autofluorescence in nearly all major brain cells, including endothelial cells, but not in pericytes.³²⁾ Considering that autofluorescence tends to increase over time, this phenomenon could be attributed to the vulnerability of pericytes.

Identifying pericytes is often accomplished by examining the expression of PDGFRβ, CD13, and α-smooth muscle actin (α-SMA). However, the debate regarding whether brain pericytes express the contractile α-SMA still persists due to issues related to tissue fixation.³³⁾ It is worth noting that MYH11, previously considered specific to SMCs, has also been discovered to be expressed in a specific subtype of pericytes.³⁴⁾ Currently, differentiating between SMCs and pericytes remains a challenge due to the lack of available methods.³⁵⁾

Our findings highlight the unique presence of OPCs as a characteristic component of mural cells during brain angiogenesis,³⁶⁾ as well as the common occurrence of bipolar OPCs in diseased conditions, a factor that has received relatively less attention in the context of mural cells.³⁷⁾ Importantly, in the brains of AD patients, while astrocytes and microglia do not display signs of aging, OPCs exhibit aging characteristics.³⁸⁾ Additionally, a recent study has indicated that damage to oligodendroglia is an early event in the development of AD.³⁹⁾

NG2 is commonly employed as a marker for identifying OPCs, but it is also occasionally used for identifying mural cells.⁴⁰⁾ In our observations, we noticed that not all converted Sox10-A cells were positive for NG2. This variability could be due to the challenges associated with NG2 immunofluorescent staining using different fixation methods.⁴¹⁾ Relying solely on Sox10 fate mapping is inadequate for distinguishing between myelinating OLs and OPCs. The theory of new mural cell formation from myelinating OLs appears implausible, as myelinating OLs exhibit remarkable stability over time.⁴²⁾ The notion of OLs transforming into mural cells seems improbable as mature OLs are postmitotic cells that do not express NG2, while some converted Sox10-A cells still maintain NG2 expression.⁴³⁾

Our observation aligns with the hypothesis that there exists a distinct population of OPCs in the grey matter that does not participate in OL development.¹⁵⁾ To emphasize the morphological differences from conventional OPCs, some researchers have introduced terms such as polydendrocytes,⁴⁴⁾ or synantocytes,⁴⁵⁾ which resemble Sox10-A cells.

The replacement of neurons may be ascribed to neuronal injury, as indicated by the presence of the cell stress-related protein ATF-3 in neurons following tamoxifen administration.⁴⁶⁾ OPCs establish synapses with both glutamatergic and GABAergic neurons.¹⁰⁾ One immediate question arises: Why does the transformation of Sox10 cells into neurons exclusively occur in glutamatergic neurons rather than GABAergic neurons? Interestingly, immunostaining for c-Fos showed that the converted neurons may not be associated with neural excitation.

The presence of RFP was further validated through immunofluorescence staining of RFP, which also eliminates the possibility of artifacts caused by autofluorescence. Additionally, cells may transfer both proteins and RNA through tunneling nanotubes.⁴⁷⁾ Further research is required to determine whether Sox10-A mural cells and Sox10-B neurons result from direct differentiation or merely from the transfer of RFP from oligodendroglia to neurons.

This study has several limitations. We cannot rule out the possibility that tamoxifen and its metabolites can directly modulate oligodendroglial lineage dynamics, altering the fate or stress-responsiveness of oligodendroglia,^48,49) although Mayrhofer et al. have demonstrated OPC-to-neuron transfer of fluorescent proteins using Sox10-Cre (non-ERT2) mice.²²⁾ Additionally, tamoxifen may induce neuronal stress and SMC stress by affecting cholesterol epoxide hydrolase.⁴⁶⁾ Finally, this study does not examine Sox10 cells and SMCs in the periphery, even though SMCs play an important role in atherosclerosis and wound healing.^50,51)

Taken together, we have discovered 2 groups of Sox10 cells that can transform into vascular mural cells and neurons when subjected to injury. In our future studies, we will delve into the connection between their inhibition and neurodegeneration.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (Grant # 82101382 to FY), the Fundamental Research Funds for the Provincial Universities of Zhejiang (Grant # 2022J007 to FY), and the College Student Science and Technology Innovation Project of Zhejiang Province (Grant # 2023R411012 to KY).

Author Contributions

Conceptualization: FY, XH. Data curation: QY, TX, FY, MH. Formal analysis: QY, TX, FY. Funding acquisition: FY. Investigation: QY, TX, XH, MH, FY, KY, ZY. Methodology: QY, FY. Writing—original draft: QY, KY, FY. Writing—review and editing: FY.

Conflict of Interest

The authors declare no conflict of interest.

Data Availability

The original contributions presented in this study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Supplementary Materials

This article contains supplementary materials.

REFERENCES

• 1)  Shepro D, Morel NM. Pericyte physiology. FASEB J., 7, 1031–1038 (1993).
• 2)  Nikolakopoulou AM, Montagne A, Kisler K, Dai Z, Wang Y, Huuskonen MT, Sagare AP, Lazic D, Sweeney MD, Kong P, Wang M, Owens NC, Lawson EJ, Xie X, Zhao Z, Zlokovic BV. Pericyte loss leads to circulatory failure and pleiotrophin depletion causing neuron loss. Nat. Neurosci., 22, 1089–1098 (2019).
• 3)  Nortley R, Korte N, Izquierdo P, Hirunpattarasilp C, Mishra A, Jaunmuktane Z, Kyrargyri V, Pfeiffer T, Khennouf L, Madry C, Gong H, Richard-Loendt A, Huang W, Saito T, Saido TC, Brandner S, Sethi H, Attwell D. Amyloid β oligomers constrict human capillaries in Alzheimer’s disease via signaling to pericytes. Science, 365, eaav9518 (2019).
• 4)  Halliday MR, Rege SV, Ma Q, Zhao Z, Miller CA, Winkler EA, Zlokovic BV. Accelerated pericyte degeneration and blood-brain barrier breakdown in apolipoprotein E4 carriers with Alzheimer’s disease. J. Cereb. Blood Flow Metab., 36, 216–227 (2016).
• 5)  Montagne A, Nation DA, Sagare AP, et al. APOE4 leads to blood-brain barrier dysfunction predicting cognitive decline. Nature, 581, 71–76 (2020).
• 6)  Yang AC, Vest RT, Kern F, et al. A human brain vascular atlas reveals diverse mediators of Alzheimer’s risk. Nature, 603, 885–892 (2022).
• 7)  Hall CN, Reynell C, Gesslein B, Hamilton NB, Mishra A, Sutherland BA, O’Farrell FM, Buchan AM, Lauritzen M, Attwell D. Capillary pericytes regulate cerebral blood flow in health and disease. Nature, 508, 55–60 (2014).
• 8)  Rafalski VA, Merlini M, Akassoglou K. Pericytes: the brain’s very first responders? Neuron, 100, 11–13 (2018).
• 9)  Duan L, Zhang XD, Miao WY, Sun YJ, Xiong G, Wu Q, Li G, Yang P, Yu H, Li H, Wang Y, Zhang M, Hu LY, Tong X, Zhou WH, Yu X. PDGFRβ cells rapidly relay inflammatory signal from the circulatory system to neurons via chemokine CCL2. Neuron, 100, 183–200.e8 (2018).
• 10)  Bergles DE, Richardson WD. Oligodendrocyte development and plasticity. Cold Spring Harb. Perspect. Biol., 8, a020453 (2016).
• 11)  Hughes EG, Kang SH, Fukaya M, Bergles DE. Oligodendrocyte progenitors balance growth with self-repulsion to achieve homeostasis in the adult brain. Nat. Neurosci., 16, 668–676 (2013).
• 12)  Eugenín-von Bernhardi J, Dimou L. NG2-glia, more than progenitor cells. Adv. Exp. Med. Biol., 949, 27–45 (2016).
• 13)  Smyth LCD, Rustenhoven J, Scotter EL, Schweder P, Faull RLM, Park TIH, Dragunow M. Markers for human brain pericytes and smooth muscle cells. J. Chem. Neuroanat., 92, 48–60 (2018).
• 14)  Nishiyama A, Serwanski DR, Pfeiffer F. Many roles for oligodendrocyte precursor cells in physiology and pathology. Neuropathology, 41, 161–173 (2021).
• 15)  Dimou L, Simon C, Kirchhoff F, Takebayashi H, Götz M. Progeny of Olig2-expressing progenitors in the gray and white matter of the adult mouse cerebral cortex. J. Neurosci., 28, 10434–10442 (2008).
• 16)  Miyamoto N, Pham LD, Seo JH, Kim KW, Lo EH, Arai K. Crosstalk between cerebral endothelium and oligodendrocyte. Cell. Mol. Life Sci., 71, 1055–1066 (2014).
• 17)  Tsai HH, Niu J, Munji R, Davalos D, Chang J, Zhang H, Tien AC, Kuo CJ, Chan JR, Daneman R, Fancy SP. Oligodendrocyte precursors migrate along vasculature in the developing nervous system. Science, 351, 379–384 (2016).
• 18)  Su Y, Wang X, Yang Y, Chen L, Xia W, Hoi KK, Li H, Wang Q, Yu G, Chen X, Wang S, Wang Y, Xiao L, Verkhratsky A, Fancy SPJ, Yi C, Niu J. Astrocyte endfoot formation controls the termination of oligodendrocyte precursor cell perivascular migration during development. Neuron, 111, 190–201.e8 (2023).
• 19)  Guo F, Maeda Y, Ma J, Xu J, Horiuchi M, Miers L, Vaccarino F, Pleasure D. Pyramidal neurons are generated from oligodendroglial progenitor cells in adult piriform cortex. J. Neurosci., 30, 12036–12049 (2010).
• 20)  Kang SH, Fukaya M, Yang JK, Rothstein JD, Bergles DE. NG2⁺ CNS glial progenitors remain committed to the oligodendrocyte lineage in postnatal life and following neurodegeneration. Neuron, 68, 668–681 (2010).
• 21)  Rivers LE, Young KM, Rizzi M, Jamen F, Psachoulia K, Wade A, Kessaris N, Richardson WD. PDGFRA/NG2 glia generate myelinating oligodendrocytes and piriform projection neurons in adult mice. Nat. Neurosci., 11, 1392–1401 (2008).
• 22)  Mayrhofer F, Hanson AM, Navedo MF, Xiang YK, Soulika AM, Deng W, Chechneva OV. Transfer of nuclear and ribosomal material from Sox10-lineage cells to neurons in the mouse brain. J. Exp. Med., 220, e20221632 (2023).
• 23)  Pluchino S, Prasad P. Oligodendroglia-to-neuron material transfer lights up the mouse CNS. J. Exp. Med., 220, e20230489 (2023).
• 24)  Zhang Y, Chen K, Sloan SA, Bennett ML, Scholze AR, O’Keeffe S, Phatnani HP, Guarnieri P, Caneda C, Ruderisch N, Deng S, Liddelow SA, Zhang C, Daneman R, Maniatis T, Barres BA, Wu JQ. An RNA-sequencing transcriptome and splicing database of glia, neurons, and vascular cells of the cerebral cortex. J. Neurosci., 34, 11929–11947 (2014).
• 25)  Li L, Wang N, Jin Q, Wu Q, Liu Y, Wang Y. Protection of Tong-Qiao-Huo-Xue Decoction against cerebral ischemic injury through reduction blood-brain barrier permeability. Chem. Pharm. Bull., 65, 1004–1010 (2017).
• 26)  Whitfield J, Littlewood T, Soucek L. Tamoxifen administration to mice. Cold Spring Harb. Protoc., 2015, 269–271 (2015).
• 27)  Wang D, Wu F, Yuan H, Wang A, Kang GJ, Truong T, Chen L, McCallion AS, Gong X, Li S. Sox10⁺ cells contribute to vascular development in multiple organs-brief report. Arterioscler. Thromb. Vasc. Biol., 37, 1727–1731 (2017).
• 28)  Donocoff RS, Teteloshvili N, Chung H, Shoulson R, Creusot RJ. Optimization of tamoxifen-induced Cre activity and its effect on immune cell populations. Sci. Rep., 10, 15244 (2020).
• 29)  Bernstein HG, Keilhoff G, Dobrowolny H, Guest PC, Steiner J. Perineuronal oligodendrocytes in health and disease. the journey so far. Rev. Neurosci., 31, 89–99 (2019).
• 30)  Szuchet S, Nielsen JA, Lovas G, Domowicz MS, de Velasco JM, Maric D, Hudson LD. The genetic signature of perineuronal oligodendrocytes reveals their unique phenotype. Eur. J. Neurosci., 34, 1906–1922 (2011).
• 31)  Brash JT, Bolton RL, Rashbrook VS, Denti L, Kubota Y, Ruhrberg C. Tamoxifen-activated CreERT impairs retinal angiogenesis independently of gene deletion. Circ. Res., 127, 849–850 (2020).
• 32)  Yang Y, Yu Q, Li B, Li S, Yang Z, Yuan F, Liu Z. A single dose of lipopolysaccharide elicits autofluorescence in the mouse brain. Front. Aging Neurosci., 15, 1126273 (2023).
• 33)  Erdener Ş.E, Küreli G, Dalkara T. Contractile apparatus in CNS capillary pericytes. Neurophotonics, 9, 021904 (2022).
• 34)  Deaton RA, Bulut G, Serbulea V, Salamon A, Shankman LS, Nguyen AT, Owens GK. A new autosomal Myh11-CreER(T2) smooth muscle cell lineage tracing and gene knockout mouse model—brief report. Arterioscler. Thromb. Vasc. Biol., 43, 203–211 (2023).
• 35)  Armulik A, Genové G, Betsholtz C. Pericytes: developmental, physiological, and pathological perspectives, problems, and promises. Dev. Cell, 21, 193–215 (2011).
• 36)  Chavali M, Ulloa-Navas MJ, Pérez-Borredá P, Garcia-Verdugo JM, McQuillen PS, Huang EJ, Rowitch DH. Wnt-dependent oligodendroglial-endothelial interactions regulate white matter vascularization and attenuate injury. Neuron, 108, 1130–1145.e5 (2020).
• 37)  Yi C, Verkhratsky A, Niu J. Pathological potential of oligodendrocyte precursor cells: terra incognita. Trends Neurosci., 46, 581–596 (2023).
• 38)  Zhang P, Kishimoto Y, Grammatikakis I, Gottimukkala K, Cutler RG, Zhang S, Abdelmohsen K, Bohr VA, Misra Sen J, Gorospe M, Mattson MP. Senolytic therapy alleviates Aβ-associated oligodendrocyte progenitor cell senescence and cognitive deficits in an Alzheimer’s disease model. Nat. Neurosci., 22, 719–728 (2019).
• 39)  Depp C, Sun T, Sasmita AO, et al. Myelin dysfunction drives amyloid-β deposition in models of Alzheimer’s disease. Nature, 618, 349–357 (2023).
• 40)  Zeng H, He X, Tuo QH, Liao DF, Zhang GQ, Chen JX. LPS causes pericyte loss and microvascular dysfunction via disruption of Sirt3/angiopoietins/Tie-2 and HIF-2α/Notch3 pathways. Sci. Rep., 6, 20931 (2016).
• 41)  Pfeiffer F, Sherafat A, Nishiyama A. The impact of fixation on the detection of oligodendrocyte precursor cell morphology and vascular associations. Cells, 10, 1302 (2021).
• 42)  Tripathi RB, Jackiewicz M, McKenzie IA, Kougioumtzidou E, Grist M, Richardson WD. Remarkable stability of myelinating oligodendrocytes in mice. Cell Rep., 21, 316–323 (2017).
• 43)  Hermann A, Brandt MD, Loewenbrück KF, Storch A. “Silenced” polydendrocytes: a new cell type within the oligodendrocyte progenitor cell population? Cell Tissue Res., 340, 45–50 (2010).
• 44)  Nishiyama A, Watanabe M, Yang Z, Bu J. Identity, distribution, and development of polydendrocytes: NG2-expressing glial cells. J. Neurocytol., 31, 437–455 (2002).
• 45)  Butt AM, Kiff J, Hubbard P, Berry M. Synantocytes: new functions for novel NG2 expressing glia. J. Neurocytol., 31, 551–565 (2002).
• 46)  Denk F, Ramer LM, Erskine EL, Nassar MA, Bogdanov Y, Signore M, Wood JN, McMahon SB, Ramer MS. Tamoxifen induces cellular stress in the nervous system by inhibiting cholesterol synthesis. Acta Neuropathol. Commun., 3, 74 (2015).
• 47)  Zhang C, Schekman R. Syncytin-mediated open-ended membrane tubular connections facilitate the intercellular transfer of cargos including Cas9 protein. eLife, 12, e84391 (2023).
• 48)  Barratt HE, Budnick HC, Parra R, Lolley RJ, Perry CN, Nesic O. Tamoxifen promotes differentiation of oligodendrocyte progenitors in vitro. Neuroscience, 319, 146–154 (2016).
• 49)  Gonzalez GA, Hofer MP, Syed YA, Amaral AI, Rundle J, Rahman S, Zhao C, Kotter MRN. Tamoxifen accelerates the repair of demyelinated lesions in the central nervous system. Sci. Rep., 6, 31599 (2016).
• 50)  Ogita H, Isobe Y, Takaku H, Sekine R, Goto Y, Misawa S, Hayashi H. Synthesis of potent and selective inhibitors against the proliferation of human coronary artery smooth muscle cells. Chem. Pharm. Bull., 51, 117–121 (2003).
• 51)  Parfejevs V, Debbache J, Shakhova O, Schaefer SM, Glausch M, Wegner M, Suter U, Riekstina U, Werner S, Sommer L. Injury-activated glial cells promote wound healing of the adult skin in mice. Nat. Commun., 9, 236 (2018).