The Lateral Ganglionic Eminence Does Not Generate Cortical Oligodendrocytes

Jialin Li; Feihong Yang; Yu Tian; Ziwu Wang; Dashi Qi; Zhengang Yang; Jiangang Song; Jing Ding; Xin Wang; Zhuangzhi Zhang

doi:10.7554/eLife.94317.1

eLife assessment

The authors provide solid evidence that any contribution of oligodendrocyte precursors to the developing cortex from the lateral ganglionic eminence is minimal in scope. The methods used support the conclusions, with some technical concerns that the authors can address with further experimentation. These are considered valuable additions to our understanding of the origins of oligodendrocytes in the forebrain during development.

https://doi.org/10.7554/eLife.94317.1.sa2

Significance of findings

valuable: Findings that have theoretical or practical implications for a subfield

landmark
fundamental
important
valuable
useful

Strength of evidence

solid: Methods, data and analyses broadly support the claims with only minor weaknesses

exceptional
compelling
convincing
solid
incomplete
inadequate

During the peer-review process the editor and reviewers write an eLife assessment that summarises the significance of the findings reported in the article (on a scale ranging from landmark to useful) and the strength of the evidence (on a scale ranging from exceptional to inadequate). Learn more about eLife assessments

Abstract

The emergence of myelinating oligodendrocytes represents a pivotal developmental milestone in vertebrates, given their capacity to ensheath axons and facilitate the swift conduction of action potentials. It is widely accepted that cortical oligodendrocyte progenitor cells (OPCs) arise from medial ganglionic eminence (MGE), lateral/caudal ganglionic eminence (LGE/CGE) and cortical radial glial cells (RGCs). Here, we used two different fate mapping strategies to challenge the established notion that the LGE generates cortical OPCs. Furthermore, we used a Cre-loxP-dependent exclusion strategy to reveal that the LGE/CGE does not give rise to cortical OPCs. Additionally, we showed that specifically eliminating MGE-derived OPCs leads to a significant reduction of cortical OPCs. Together, our findings indicate that the LGE does not generate cortical OPCs, contrary to previous beliefs. These findings provide a new view of the developmental origins of cortical OPCs and a valuable foundation for future research on both normal development and oligodendrocyte-related disease.

Introduction

Oligodendrocytes (OLs) are one of the most abundant cell types in the mammalian neocortex. The emergence of myelination OLs in vertebrates is a major event in the evolution of the central nervous system (CNS), as these cells play a crucial role in enhancing axonal insulation and the conduction of neural impulses. Previously, they have been described as support cells that myelinate axons and play essential roles in the function of neocortical circuits. Myelin loss or dysfunction is associated with many disorders, such as multiple sclerosis, depression and pediatric leukodystrophies^1,2.

During CNS development, OLs originate from oligodendrocyte precursor cells (OPCs), which arise from different regions of the embryonic neural tube. OPCs proliferate and migrate away from the ventricular germinal zones (VZ) into developing gray and white matter before differentiating into mature OLs^3,4. The developmental origin of cortical oligodendrocytes has been extensively studied in the CNS, especially in the neocortex^5-8. A landmark study⁹ demonstrated that cortical OPCs are generated in three waves, starting with a ventral wave, gradually transitioning to more dorsal origins. The first wave of cortical OPCs is derived from the medial ganglionic eminence (MGE) or the anterior entopeduncular area (AEP) at embryonic day E12.5 (Nkx2.1⁺ lineage). The second wave of cortical OLs is derived from the lateral/caudal ganglionic eminences (LGE/CGE) at E15.5 (Gsx2⁺/Nkx2.1^-lineage). The final wave occurs at P0, when OPCs originate from cortical glial progenitor cells (Emx1⁺ lineage). Our recent studies^10,11, however, challenge the notion that cortical OLs are generated in these three waves^9,12-14, as the cortical glial progenitor cells were also found to express Gsx2, which was known to be exclusively expressed in the developing subpallium^15-17, after E16.5¹¹. Therefore, Gsx2⁺/Nkx2.1^-lineage OLs may not represent LGE-derived cortical OLs. To date, whether the LGE generates cortical OPCs (OLs) during cortical development remains an unanswered question in the field of developmental neuroscience.

The present study was undertaken to answer this question. We carried out distinct fate mapping experiments and intersectional lineage analyses to reveal that cortical OPCs do not derive from the LGE. First, using in utero electroporation (IUE) to target the LGE in a Cre recombinase-dependent IS (Rosa-CAG-LSLFrt-tdTomato-Frt-EGFP) reporter mouse line¹⁸, we found that LGE-derived OPCs produced striatal OLs rather than cortical OLs. Second, by combining the PiggyBac transposon system with IUE, we found that no traced cells migrated into the cortex from the LGE. Third, by crossing the Nkx2.1-Cre and Emx1-Cre mouse lines with an H2B-GFP reporter mouse line, we found that nearly all cortical OPCs originated from Nkx2.1-Cre or Emx1-Cre lineage cells. Finally, we analyzed Olig2-conditional knockout mice (Nkx2.1-Cre; Olig2^F/F mice), and found that the number of cortical OPCs was significantly reduced from E14.5 to E16.5. These results indicated that the MGE was the sole ventral source of cortical OPCs. Taken together, our findings suggest that the LGE does not give rise to cortical OPCs (OLs) during neocortical development.

In summary, our findings collectively indicate that the MGE is the primary ventral source of cortical OPCs (OLs) and that the LGE does not generate cortical OPCs, challenging the conventional view that the LGE generates cortical OPC (OL) populations during cortical development. This research advances our understanding of CNS development and presents a new perspective on the origins of cortical OPCs (OLs).

Results

Fate mapping of LGE-derived OPCs by combining IUE with a Cre recombinase-dependent IS reporter

OPCs initially emerge in the ventral VZ of the MGE and can be labeled by platelet-derived growth factor receptor (a-subunit, Pdgfra) around E12.5. We hypothesized that if the LGE generated cortical OPCs, the Pdgfra⁺cell migration stream from the LGE to the cortex could be observed during cortical development. To confirm this speculation, we conducted a detailed analysis of Pdgfra mRNA expression using in situ hybridization (ISH). Our results showed that there was no observable Pdgfra⁺ cell migration stream from the LGE to the cortex from E13.5 to P0 (Figure 1A). Next, we used Cre recombinases in combination with an IS reporter mouse line¹⁸ to fate map LGE-derived OPCs (OLs). We delivered pCAG-Cre plasmids specifically into the LGE VZ of IS reporter mouse embryos at E13.5 by IUE (Figures 1B-C). In this experiment, cells generated from electroporated neural stem cells (NSCs) expressed tdTomato (tdT). Interestingly, we observed many tdT⁺ cells throughout the whole LGE (striatum) at E18.5. In contrast, there were very few tdT⁺ cells detected in the cortex (Figure 1D). OPCs express Olig2, Sox10, Pdgfra and NG2 (Cspg4) and mature neurons express NeuN ^19-25. To determine the identity of the tdT⁺ cells, we performed double staining of tdT with OLIG2, SOX10, PDGFRA, NG2 (CSPG4) and NeuN. Within the cortex, tdT⁺ cells did not express OPC marker genes but express NeuN; however, tdT⁺ cells expressed OPC marker genes and NeuN in the striatum (Figure 1E). We also delivered pCAG-Cre plasmids specifically into the LGE VZ of IS reporter mice at P0 by IUE and analyzed these mice at P49 (Figure 1F) to investigate whether the LGE generates cortical OLs at a later developmental stage. Our results showed that LGE progenitors generate LGE OLs and astrocytes, which expressed Sox10 and Sox9, respectively (Figure 1F).

To further determine whether LGE-derived OPCs migrate into the cortex after E18.5, we used the same strategy as mentioned above (Figure S1A). We observed that at P10, the majority of the tdT⁺ cells were located in the striatum, whereas very few tdT⁺ cells were located in the neocortex (Figure S1B). All tdT⁺ cells in the neocortex expressed NeuN but did not express OLIG2 (Figure S1C). In fact, tdT⁺ cells in the neocortex did not express OPC marker genes; instead, they expressed NeuN in the neocortex. In contrast, tdT⁺ cells in the striatum expressed OPC marker genes (Olig2, Sox10 and Pdgfra) and NeuN (Figure S1D), consistent with our observation at E18.5 (Figure 1E). We also found that tdT⁺ cells expressed the astrocyte marker ALDH1L1 in the striatum at P10 (Figure S1D). Furthermore, through consecutive tissue sectioning and 3D reconstruction, we revealed that the traced cells tended to be localized in the striatum (Figure S1F). Collectively, our findings indicated that the LGE gives rise to striatal OLs, neurons and astrocytes, but not cortical OLs.

Lineage tracing of LGE-derived OPCs by combining IUE with the PiggyBac transposon system

To further investigate whether the LGE generates cortical OPCs during cortical development, we employed an alternative fate mapping strategy by combining IUE with the PiggyBac transposon system to map the fate of OPCs derived from LGE. This approach utilized a transposase enzyme (helper plasmid) to integrate the H2B-GFP sequence from a donor plasmid into the genome of the in utero electroporated LGE progenitors along the lateral ventricular zone. The H2B-GFP reporter vector was delivered specifically into the LGE VZ of wild-type embryos at E13.5 by IUE, and these embryos were examined at E18.5 or P10 (Figures 2A-B). In this system, most traced cells (GFP⁺) were located in the striatum, and very few traced cells were observed in the cortex (Figures 2C and E). Usually, a small population of RGCs located in the dorsal LGE or lateral/ventral pallium was labeled by IUE. These RGCs mainly migrated to the lateral/ventral pallium and amygdala (Figure 2C). We did not observe any GFP⁺ cells that were OPCs in the dorsal cortex, as these cells did not express OLIG2, SOX10 or PDGFRA (Figure 2D). Instead, they expressed NeuN (Figure 2D). In contrast, GFP⁺ cells expressed OLIG2, SOX10, PDGFRA and NeuN in the striatum at E18.5 (Figure 2D). The same results were also observed at P10 (Figure 2E). Again, our findings strongly suggested that the LGE does not generate cortical OPCs.

Fate mapping of LGE-derived OPCs by combining IUE with PiggyBac transposon system
(A) Schematic of the PiggyBac transposon reporter system used in this study.
(B) The experimental workflow.
(C) Representative coronal sections showing the distribution of GFP⁺ cells at E18.5.
(D) GFP⁺ cells expressed NeuN but not glial cell markers, such as OLIG2, SOX10 and PDGFRA in the cortex. However, GFP⁺ cells expressed NeuN, OLIG2, SOX10 and PDGFRA in the striatum.
(E) GFP⁺ cells expressed NeuN but not glial cell markers, such as OLIG2, SOX10 and PDGFRA in the cortex and GFP⁺ cells expressed NeuN, OLIG2, SOX10 and PDGFRA in the striatum at P10.

An exclusion strategy revealed that cortical OPCs do not originate from the LGE/CGE

While distinct fate mapping strategies provided evidence that the LGE did not generate cortical OPCs during cortical development, there remained a possibility that a small population of LGE RGCs or progenitors might generate a subpopulation of cortical OPCs at specific time points. To address this possibility, we crossed the Emx1-Cre^26-28, Nkx2.1-Cre²⁹ and H2B-GFP¹⁸ mouse lines to generate Emx1-Cre; Nkx2.1-Cre; H2B-GFP mice. Because Emx1-Cre was specifically expressed in cortical RGCs²⁶ and Nkx2.1-Cre was specifically expressed in MGE/AEP RGCs²⁹, cortical OPCs were traced (GFP⁺ cells) in using Emx1-Cre; Nkx2.1-Cre; H2B-GFP mice. GFP^- cell OPCs, on the other hand, were considered as LGE/CGE-derived cortical OPCs (OLs) (Figures 3A-B).

An exclusion strategy showed that cortical OPCs do not come from the LGE/CGE
(A) Scheme of the *H2B-GFP* reporter lines.
(B) Experimental design of the exclusion strategy to trace the lineage of LGE RGCs.
(C) Representative coronal sections showing the traced cells in the forebrain. The majority of SOX10- and PDGFRA-positive cells were GFP⁺ cells in the cortex at P10.
(D-E) The pie chart shows that the percentage of LGE/dMGE-derived cortical OPCs was approximately 3%. N = 6 mice per group.
(F) Many GFP⁺ cells expressed SOX10 and PDGFRA in the striatum at P10.
(G) Nearly all SOX10 and PDGFRA expressed GFP in the cortex at P0.
(H-I) The pie chart shows that the percentage of LGE/dMGE-derived cortical OPCs was less than 3%. N = 6 mice per group.

A previous report⁹ suggested that cortical OPCs originate from MGE, LGE/CGE and cortical RGCs. When MGE-derived cortical OPCs were eliminated, LGE/CGE-derived cells accounted for approximately 50% of cortical OPCs at P10. To confirm these findings, we performed double staining of GFP with OPC marker genes, such as Sox10 and Pdgfra in Emx1-Cre; Nkx2.1-Cre; H2B-GFP mice (Figure 3C). We observed that the majority of OPCs or OLs were colabeled with GFP in Emx1-Cre; Nkx2.1-Cre; H2B-GFP mice at P10 (Figure 3C). Statistical analysis showed that only approximately 2% of OPCs were GFP^- cells, and they were possibly derived from the LGE/CGE (Figures 3D-E), which was consistent with our observation at P0 (Figures 3G-I). Importantly, the expression of Nkx2.1-Cre is lower within the dorsalmost region of the MGE than in other Nkx2.1-expressing regions²⁹. Thus, these GFP^- OPCs (∼2%) may be derived from the dorsal MGE rather than from the LGE/CGE. Using this approach, we revealed that nearly all cortical OLs/OPCs were derived from MGE/AEP and cortical RGCs. Taken together, our findings indicated that the LGE/CGE did not generate cortical OLs, or that if they do, the number is negligible.

The MGE/AEP was the sole ventral source of the cortical OPCs

Olig2, a basic helix-loop-helix transcription factor, plays a critical role in the regulation of OPC fate determination. To investigate the role of Olig2 in the MGE, we crossed Nkx2.1-Cre mice with Olig2 ^F/+ mice to eliminate Olig2 expression in the SVZ/VZ of the MGE/AEP (Figures 4A-B). We referred to the conditional mutants as Olig2-NCKO mice. A previous study⁹ showed that MGE/AEP-derived OPCs migrate into the cortex around E14.5, and LGE/CGE-derived OPCs migrate into cortex around E15.5. We assumed that the MGE/AEP was the sole ventral source of cortical OPCs, and that inhibiting MGE/AEP-derived OPC generation would therefore significantly reduce the number of cortical OPCs at E16.5. To test this possibility, we analyzed Olig2-NCKO mice, in which the generation of MGE/AEP-derived OPCs was inhibited. Our results showed that the expression of OPC marker genes, such as Sox10 and Pdgfra was significantly reduced (> 10-fold) in the cortex of Olig2-NCKO mice, compared to the control mice at E14.5 and E16.5 (Figures 4C-E). Considering the recombination efficiency of the CRE enzyme, the limited presence of residual OPCs in the cortex of Olig2-NCKO mice may be arise from the MGE/AEP. Indeed, we observed few PDGFRA-positive cells in the MGE/AEP of the Olig2-NCKO mice (Figure 4F). Altogether, our findings indicated that the MGE/AEP was the sole ventral source of cortical OPCs.

The MGE may be the sole ventral source of cortical OPCs
(A) The CRISPR/Cas9 technique was used to generate an *Olig2* conditional knockout allele.
(B) Experimental design for the generation of *Olig2-NCKO* mice.
(C-D) The expression of SOX10 and PDGFRA was significantly reduced in the cortex of *Olig2-NCKO* mice compared with that of control mice.
(E) The number of SOX10- and PDGFRA-positive cells was significantly reduced in *Olig2-NCKO* mice compared with control mice. Student’s t-test, **P <.01, ***P <.001, n ≥ 5 mice per group, mean ± SEM.
(F) Few PDGFRA-positive cells were detected in the MGE/AEP of the *Olig2-NCKO* mice.
(G) A new model of the developmental origins of cortical OPCs.

Discussion

It is widely accepted in the field of developmental neuroscience that cortical oligodendrocyte precursor cells (OPCs) originate from three regions: the MGE/AEP, LGE/CGE and cortex. However, in the current study, we re-examined the developmental origins of cortical OPCs using diverse strategies. Our findings show that cortical OPCs do not come from the LGE/CGE, contrary to previous reports^9,12, which suggested that the LGE/CGE generates cortical OPCs (Figure 4G). In fact, no cortical OPCs migrating from the LGE to the cortex were observed via our two fate mapping strategies. Furthermore, an exclusion strategy suggested that the LGE/CGE were not responsible for generating cortical OPCs. In addition, according to our study of Olig2-NCKO mice, MGE/AEP seems to be the sole ventral source of cortical OPCs during neocortical development. Overall, these findings provide evidence that the LGE does not generate cortical OPCs in the mouse brain.

Why the LGE cannot generate cortical OPCs during cortical development

OLs are specialized cells of the CNS that produce myelin, a multilayered membrane that spirally enwraps axons and facilitates rapid nerve conduction. Recent studies have shown that cortical OLs are heterogeneous, as there are transcriptionally distinct subpopulations of cortical OLs during development and disease^8,30-32. Indeed, OLs are derived from distinct OPC pools, which acquire different abilities to respond to demyelination^9,33. In the spinal cord, the developmental origin of OLs has been hotly debated for many years. Some groups believed that OLs originate only from the ventral neural tube, whereas others tend to support multiple origins^12,34-36. Currently, it is generally accepted that spinal cord OLs are derived from two origins: the ventral region and dorsal region. One pioneering study⁹ purported that, in the forebrain, cortical OLs derive from multiple origins, such as the MGE, LGE and cortex.

In general, the specification of OLs depends on SHH protein released from the notochord and floor plate, as Shh expression is both necessary and sufficient for OL generation in the spinal cord³⁷. As in the spinal cord, OL generation in the cortex is also dependent on the Shh signaling pathway³⁸. During MGE development, Nkx2.1 and Six3 promote SHH secretion. Furthermore, the early-born neurons of the MGE also secrete SHH protein before E14.5³⁹. Thus, the high levels of SHH in the MGE contribute to the generation of OPCs from RGCs. In contrast, Shh signaling pathway activity is very weak in the LGE at the early stage (before E14.5). As the level of SHH gradually increases in the LGE, this results in a shift from neurogenesis to gliogenesis after E15.5, and the large number of the striatal neurons may block the migration of OPCs into the neocortex^40,41. In summary, the process of cortical OL generation can be summarized as follows: First, the time window during which the MGE generates OLs is very brief, perhaps occurring before MGE neurogenesis. The high level of SHH in the MGE allows for the production of a small population of cortical OPCs around E12.5. Subsequently, multipotent intermediate progenitors begin to express DLX transcription factors resulting in ending the generation of OPCs in the MGE^42-44. Second, radial glial cells (RGCs) undergo a gradual transition from neurogenesis to gliogenesis around E15.5, driven by the progressive elevation of SHH levels in the LGE. However, the LGE had already generated a substantial number of neurons during this period, blocking the migration of OPCs into the cortex. Finally, with the gradual increase of SHH concentration, cortical RGCs begin to cease neurogenesis and shift toward gliogenesis after E16.5^10,40,45. Therefore, the generation of cortical OPCs (OLs) is regulated by both precise temporal and spatial patterning, which is strictly dependent on the local environment.

Cortical-derived OLs/OPCs make up the majority of the cortical OLs/OPCs in the adult brain

During the early development of the cortex, both pallium and subpallium can generate cortical OPCs^9,46. Interestingly, most of the ventral-derived OPCs, particularly MGE-derived OPCs/OLs, are eventually eliminated due to the continued expansion of dorsal-derived OPCs in the adult cortex^6,9. It appears that OPCs generated later in development may possess a competitive advantage over their earlier-generated counterparts. Indeed, human glial progenitor cells (hGPCs) engrafted neonatally or transplanted into the adult mouse brain can effectively disperse throughout the forebrain, differentiate into mature OLs and remyelinate the demyelinated adult brain^1,47. These findings indicate that OPCs generated later in development have the ability to take over for earlier-generated OPCs in the adult brain.

Based on the present work and other studies^8,9,46, cortical-derived OPCs probably constitute the majority of cortical OPCs, including mature cortical OLs. Thus, OLs may function as a homogeneous population in the adult forebrain. In fact, the number of LGE/CGE-derived cortical OPCs may also gradually decrease in the adult brain, as reported by Lui et al⁶. However, Lui et al. did not analyze LGE-derived and MGE-derived OPCs separately. Therefore, they did not exclude the possibility that the LGE/CGE does not generate cortical OPCs. In summary, cortical OPCs/OLs comprise MGE-derived and cortical RGC-derived OPCs/OLs during the neonatal stage. However, only cortical-derived OPCs/OLs remain in the adult cortex.

Our study provides strong evidence that the LGE does not give rise to cortical OPCs/OLs during forebrain development. Thus, our findings provide new evidence about the developmental origins of cortical OLs, which may guide future research on both normal development and disease.

Methods

Animals

All experiments conducted in this study were in accordance with the guidelines of Fudan University. We used Emx1-Cre^26,27, and Nkx2.1-Cre²⁹ mice, in conjunction with either Rosa-H2B-GFP¹⁸ or Rosa-IS¹⁸ for genetic fate mapping. Olig2^F/+mice were generated via the CRISPR/Cas9 strategy. LoxP sites flanked the coding region of exons 1 and 2. Wild-type mice or littermates without the Cre allele were used as controls. All mice were maintained on a mixed genetic background (C57BL/6J and CD1). The mice were allowed access to water and food ad libitum and maintained on a 12-hour light/dark cycle. The day of vaginal plug detection was considered embryonic day 0.5 (E0.5), and the day of birth was defined as postnatal day 0 (P0). Both male and female mice were used in all experiments.

Tissue preparation

Postnatal mice were deeply anesthetized and perfused intracardially thoroughly with ice-cold PBS followed by 4% paraformaldehyde (PFA). Embryos were harvested from deeply anesthetized pregnant mice. All brains were fixed overnight in 4% PFA at 4°C, cryoprotected in 30% sucrose for at least 24 hours, embedded in optimal cutting temperature (O.C.T.) (Sakura Finetek) on dry ice and ethanol slush and preserved at -80°C. All frozen mouse tissues were sectioned into 20 µm thick sections unless specifically mentioned and stained on glass slides.

Plasmid Construction

The pCAG-Cre vector was constructed as reported elsewhere. The coding sequence of Cre in the pCAG-Cre vector was replaced by the PiggyBac transposase to construct the pCAG-PiggyBac transposase vector. The coding DNA sequence (CDS) of Cre in the pCAG-Cre vector was replaced by H2B-GFP from the Rosa-H2B-GFP mouse strain to construct the PB-pCAG-H2B-GFP vector. The inverted terminal repeat (ITR) sequence of the PiggyBac transposon was cloned upstream of the CAG promoter and downstream of the beta-globin polyA sequence for recognition by the PiggyBac transposase.

In utero electroporation (IUE)

In utero electroporation (IUE) of wild-type or IS embryos was performed at E13.5. The pCAG-Cre (Addgene #13775), pCAG-GFP (Addgene #11150), pCAG-PiggyBac transposase or PB- pCAG-H2B-GFP (final concentration: 1-2 μg/μL) plasmid was mixed with 0.05% Fast Green (Sigma) and injected into the lateral ventricle of embryos (0.5 μL per embryo) using a beveled glass micropipette. Five electrical pulses (duration: 50 ms) were applied at 42 V across the uterine wall with a 950 ms interval between pulses. Electroporation was performed using a pair of 5 mm platinum electrodes (BTX, Tweezertrode 45-0488, Harvard Apparatus) connected to an electroporator (BTX, ECM830). The embryos were analyzed at E18.5 and P10.

Immunofluorescence

The sections were first washed with 0.05 M TBS for 10 min, incubated in Triton-X-100 (0.5% in 0.05 M TBS) for 30 min at room temperature (RT), and then incubated with blocking solution (5% donkey serum +0.5% Triton-X-100 in 0.05 M TBS, pH= 7.2) for 2 h at RT. The primary antibodies were diluted in 5% donkey serum, incubated overnight at 4°C, and then rinsed 3 times with 0.05 M TBS.

Secondary antibodies (1:500, all from Jackson ImmunoResearch) matching the appropriate species were added and incubated for 2-4 h at RT. Fluorescently stained sections were then washed 3 times with 0.05 M TBS for 10 min. The sections were then stained with 4’,6-diamidino-2-phenylindole (DAPI) (1:3000, Sigma) diluted in TBS for 1 min and then finally rinsed three more times with TBS. The sections were then coverslipped with Gel/Mount (Biomeda, Foster City, CA). The primary antibodies used in this study are shown in the key resource table.

In situ hybridization (ISH)

ISH was performed on 20-μm cryosections using digoxigenin riboprobes as previously described ⁴⁸. The sections were first postfixed in 4% PFA for 20 min. mRNA in situ hybridization was performed as previously described⁴⁹. The probes were constructed from cDNA isolated from the wild-type mouse brain and amplified by PCR using the following primers:

3D Reconstruction

Images of serial coronal sections along the rostral–caudal axis were collected via an Olympus VS 120 microscope using a 10X objective and used by Amira as a formwork for reconstruction. The coordinates of each labeled cell (tdT+) in the brain were determined. A stereotaxic mouse brain atlas was used to locate the brain anatomical structures where the labeled cells were positioned.

Image acquisition and statistical analysis

Images for quantitative analyses were acquired with an Olympus VS 120 microscope using a 10X objective. Bright-field images were acquired with an Olympus VS 120 microscope using a 10X objective. Other fluorescence images were taken with the Olympus FV3000 confocal microscope system using 20× or 40× objectives. The images were merged, cropped and optimized in Adobe Photoshop and Adobe Illustrator without distorting the original visual information.

Analyses were performed using Microsoft Excel. Unpaired t test was used to determine statistical significance. All quantitative data are presented as the mean ± SEM. Differences with p values < 0.05 were considered significant. For quantification of SOX10⁺, PDGFRA⁺ and GFP⁺ cells in the cortex at E18.5 and P10, six anatomically matched 20-μm thick coronal sections were selected (n = 6 mice per group). We counted SOX10⁺, PDGFRA⁺ and GFP⁺ cells in the cortex under a 20X objective lens. The data are presented as the number of SOX10⁺, PDGFRA⁺ and GFP⁺ cells per section for each cortex.

For quantification of SOX10⁺ and PDGFRA⁺ cells in the dorsal cortex at E14.5 and E16.5, five or six anatomically matched 20-μm thick coronal sections were selected (n = 6, E14.5; n=5, E16.5). We counted SOX10⁺ and PDGFRA⁺ cells in the cortex under a 20X objective lens. The data are presented as the number of SOX10⁺ and PDGFRA⁺ cells per section for each cortex.

Key resource table

Experimental models

Recombinant DNA

All secondary antibodies were from Jackson ImmunoResearch Labs.

Acknowledgements

This work was supported by the Ministry of Science and Technology of China (STI2030-2021ZD0202300), National Natural Science Foundation of China (NSFC 31820103006, 82271197, 81974175 and 32200776), Shanghai Municipal Science and Technology Major Project (No. 2018SHZDZX01), ZJ Lab, and Shanghai Center for Brain Science and Brain-Inspired Technology.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Potential conflicts of interest

We declare that there is no conflict of interest.

Author contributions

J.L., F.Y. and Y.T. performed all experiments and analyzed data. Z.W., and D.Q. helped to conduct experiments and analyze the data. Z.Y. and Z.Z. designed the experiments and analyzed the results. Z.Z. wrote this manuscript. Z.Y., J.S., J.D., X.W. and Z.Y. help to review this manuscript.

Figure legends

Figure S1. Lineage tracing of LGE-derived OPCs by combining IUE with a Cre recombinase-dependent IS reporter

(A) Experimental schedule for fate mapping of LGE-derived OPCs.

(B) Representative coronal sections showing the distribution of the tdT⁺ cells at P10.

(C) Nearly all tdT⁺ cells expressed NeuN but not OLIG2 in the cortex. tdT⁺ cells expressed NeuN and OLIG2 in the striatum.

(D) tdT⁺ cells did not express OLIG2, SOX10, PDGFRA or ALDH1L1. Instead, they expressed NeuN in the cortex. In sharp contrast, tdT⁺ cells in the striatum expressed OLIG2, SOX10, PDGFRA, ALDH1L1 and NeuN at P10.

(E) 3D reconstruction of consecutive brain sections demonstrated that the traced cells were mainly located in the striatum.

References

1
1. Goldman S. A.
2. Mariani J. N.
3. Madsen P. M
2021Glial progenitor cell-based repair of the dysmyelinated brain: Progression to the clinicSemin Cell Dev Biol 116:62–70https://doi.org/10.1016/j.semcdb.2020.12.004
2
1. Zhou B.
2. Zhu Z.
3. Ransom B. R.
4. Tong X.
2021Oligodendrocyte lineage cells and depressionMol Psychiatry 26:103–117https://doi.org/10.1038/s41380-020-00930-0
3
1. Miller R. H.
1996Oligodendrocyte originsTrends Neurosci 19:92–96https://doi.org/10.1016/s0166-2236(96)80036-1
4
1. Huang W.
2. et al.
2020Origins and Proliferative States of Human Oligodendrocyte Precursor CellsCell https://doi.org/10.1016/j.cell.2020.06.027
5
1. Allen D. E.
2. et al.
2022Fate mapping of neural stem cell niches reveals distinct origins of human cortical astrocytesScience 376:1441–1446https://doi.org/10.1126/science.abm5224
6
1. Liu R.
2. et al.
2021In Vivo Clonal Analysis Reveals Development Heterogeneity of Oligodendrocyte Precursor Cells Derived from Distinct Germinal ZonesAdvanced Science 8https://doi.org/10.1002/advs.202102274
7
1. Shen Z.
2. et al.
2021Distinct progenitor behavior underlying neocortical gliogenesis related to tumorigenesisCell Rep 34https://doi.org/10.1016/j.celrep.2021.108853
8
1. Marques S.
2. et al.
2018Transcriptional Convergence of Oligodendrocyte Lineage Progenitors during DevelopmentDev Cell 46:504–517https://doi.org/10.1016/j.devcel.2018.07.005
9
1. Kessaris N.
2. et al.
2006Competing waves of oligodendrocytes in the forebrain and postnatal elimination of an embryonic lineageNat Neurosci 9:173–179https://doi.org/10.1038/nn1620
10
1. Li X.
2. et al.
2021Decoding Cortical Glial Cell DevelopmentNeurosci Bull 37:440–460https://doi.org/10.1007/s12264-021-00640-9
11
1. Zhang Y.
2. et al.
2020Cortical Neural Stem Cell Lineage Progression Is Regulated by Extrinsic Signaling Molecule Sonic HedgehogCell Reports 30:4490–4504https://doi.org/10.1016/j.celrep.2020.03.027
12
1. Richardson W. D.
2. Kessaris N.
3. Pringle N.
2006Oligodendrocyte warsNat Rev Neurosci 7:11–18https://doi.org/10.1038/nrn1826
13
1. Rowitch D. H.
2. Kriegstein A. R
2010Developmental genetics of vertebrate glial-cell specificationNature 468:214–222https://doi.org/10.1038/nature09611
14
1. Bergles D. E.
2. Richardson W. D.
2015Oligodendrocyte Development and PlasticityCold Spring Harb Perspect Biol 8https://doi.org/10.1101/cshperspect.a020453
15
1. Chapman H.
2. et al.
2018Gsx transcription factors control neuronal versus glial specification in ventricular zone progenitors of the mouse lateral ganglionic eminenceDev Biol 442:115–126https://doi.org/10.1016/j.ydbio.2018.07.005
16
1. Chapman H.
2. Waclaw R. R.
3. Pei Z.
4. Nakafuku M.
5. Campbell K
2013The homeobox gene Gsx2 controls the timing of oligodendroglial fate specification in mouse lateral ganglionic eminence progenitorsDevelopment 140:2289–2298https://doi.org/10.1242/dev.091090
17
1. Wen Y.
2. et al.
2021Transcription Factor VAX1 Regulates the Regional Specification of the Subpallium Through Repressing Gsx2Molecular Neurobiology 58:3729–3744https://doi.org/10.1007/s12035-021-02378-x
18
1. He M.
2. et al.
2016Strategies and Tools for Combinatorial Targeting of GABAergic Neurons in Mouse Cerebral CortexNeuron 91:1228–1243https://doi.org/10.1016/j.neuron.2016.08.021
19
1. Sun T.
2. et al.
2001Olig bHLH proteins interact with homeodomain proteins to regulate cell fate acquisition in progenitors of the ventral neural tubeCurrent biology: CB 11:1413–1420https://doi.org/10.1016/s0960-9822(01)00441-9
20
1. Zhou Q.
2. Choi G.
3. Anderson D. J
2001The bHLH transcription factor Olig2 promotes oligodendrocyte differentiation in collaboration with Nkx2.2Neuron 31:791–807https://doi.org/10.1016/s0896-6273(01)00414-7
21
1. Lu Q. R.
2. et al.
2002Common developmental requirement for Olig function indicates a motor neuron/oligodendrocyte connectionCell 109:75–86https://doi.org/10.1016/s0092-8674(02)00678-5
22
1. Finzsch M.
2. Stolt C. C.
3. Lommes P.
4. Wegner M
2008Sox9 and Sox10 influence survival and migration of oligodendrocyte precursors in the spinal cord by regulating PDGF receptor alpha expressionDevelopment 135:637–646https://doi.org/10.1242/dev.010454
23
1. Reiprich S.
2. et al.
2017Transcription factor Sox10 regulates oligodendroglial Sox9 levels via microRNAsGlia 65:1089–1102https://doi.org/10.1002/glia.23146
24
1. Kucharova K.
2. Stallcup W. B
2010The NG2 proteoglycan promotes oligodendrocyte progenitor proliferation and developmental myelinationNeuroscience 166:185–194https://doi.org/10.1016/j.neuroscience.2009.12.014
25
1. Sanchez-Gonzalez R.
2. Bribian A.
3. Lopez-Mascaraque L
2020Cell Fate Potential of NG2 ProgenitorsSci Rep 10https://doi.org/10.1038/s41598-020-66753-9
26
1. Gorski J. A.
2. et al.
2002Cortical excitatory neurons and glia, but not GABAergic neurons, are produced in the Emx1-expressing lineageJ Neurosci 22:6309–6314https://doi.org/10.1523/JNEUROSCI.22-15-06309.2002
27
1. Du H.
2. et al.
2021Transcription Factors Bcl11a and Bcl11b Are Required for the Production and Differentiation of Cortical Projection NeuronsCerebral Cortex https://doi.org/10.1093/cercor/bhab437
28
1. Li Z.
2. et al.
2023BMP7 expression in mammalian cortical radial glial cells increases the length of the neurogenic periodProtein Cell https://doi.org/10.1093/procel/pwad036
29
1. Xu Q.
2. Tam M.
3. Anderson S. A
2008Fate mapping Nkx2.1-lineage cells in the mouse telencephalonJ Comp Neurol 506:16–29https://doi.org/10.1002/cne.21529
30
1. Floriddia E. M.
2. et al.
2020Distinct oligodendrocyte populations have spatial preference and different responses to spinal cord injuryNat Commun 11https://doi.org/10.1038/s41467-020-19453-x
31
1. Marques S.
2. et al.
2016Oligodendrocyte heterogeneity in the mouse juvenile and adult central nervous systemScience 352:1326–1329https://doi.org/10.1126/science.aaf6463
32
1. Falcao A. M.
2. et al.
2018Disease-specific oligodendrocyte lineage cells arise in multiple sclerosisNat Med 24:1837–1844https://doi.org/10.1038/s41591-018-0236-y
33
1. Crawford A. H.
2. Tripathi R. B.
3. Richardson W. D.
4. Franklin R. J. M
2016Developmental Origin of Oligodendrocyte Lineage Cells Determines Response to Demyelination and Susceptibility to Age-Associated Functional DeclineCell Rep 15:761–773https://doi.org/10.1016/j.celrep.2016.03.069
34
1. Yu W.-P.
2. Collarini E. J.
3. Pringle N. P.
4. Richardson W. D
1994Embryonic expression of myelin genes: Evidence for a focal source of oligodendrocyte precursors in the ventricular zone of the neural tubeNeuron 12:1353–1362https://doi.org/10.1016/0896-6273(94)90450-2
35
1. Spassky N.
2. et al.
1998Multiple restricted origin of oligodendrocytesJ Neurosci 18:8331–8343https://doi.org/10.1523/JNEUROSCI.18-20-08331.1998
36
1. Fogarty M.
2. Richardson W. D.
3. Kessaris N
2005A subset of oligodendrocytes generated from radial glia in the dorsal spinal cordDevelopment 132:1951–1959https://doi.org/10.1242/dev.01777
37
1. Orentas D. M.
2. Hayes J. E.
3. Dyer K. L.
4. Miller R. H
1999Sonic hedgehog signaling is required during the appearance of spinal cord oligodendrocyte precursorsDevelopment 126:2419–2429https://doi.org/10.1242/dev.126.11.2419
38
1. Lu Q. R.
2. et al.
2000Sonic hedgehog--regulated oligodendrocyte lineage genes encoding bHLH proteins in the mammalian central nervous systemNeuron 25:317–329https://doi.org/10.1016/s0896-6273(00)80897-1
39
1. Flandin P.
2. et al.
2011Lhx6 and Lhx8 coordinately induce neuronal expression of Shh that controls the generation of interneuron progenitorsNeuron 70:939–950https://doi.org/10.1016/j.neuron.2011.04.020
40
1. Tong C. K.
2. et al.
2015A Dorsal SHH-Dependent Domain in the V-SVZ Produces Large Numbers of Oligodendroglial Lineage Cells in the Postnatal BrainStem Cell Reports 5:461–470https://doi.org/10.1016/j.stemcr.2015.08.013
41
1. Su Z.
2. et al.
2022Dlx1/2-dependent expression of Meis2 promotes neuronal fate determination in the mammalian striatumDevelopment 149https://doi.org/10.1242/dev.200035
42
1. Dai J.
2. Bercury K. K.
3. Ahrendsen J. T.
4. Macklin W. B
2015Olig1 function is required for oligodendrocyte differentiation in the mouse brainJ Neurosci 35:4386–4402https://doi.org/10.1523/JNEUROSCI.4962-14.2015
43
1. Silbereis J. C.
2. et al.
2014Olig1 function is required to repress dlx1/2 and interneuron production in Mammalian brainNeuron 81:574–587https://doi.org/10.1016/j.neuron.2013.11.024
44
1. Price J. D.
2. et al.
2022DLX1 and the NuRD complex cooperate in enhancer decommissioning and transcriptional repressionDevelopment 149https://doi.org/10.1242/dev.199508
45
1. Winkler C. C.
2. et al.
2018The Dorsal Wave of Neocortical Oligodendrogenesis Begins Embryonically and Requires Multiple Sources of Sonic HedgehogJ Neurosci 38:5237–5250https://doi.org/10.1523/JNEUROSCI.3392-17.2018
46
1. van Tilborg E.
2. et al.
2018Origin and dynamics of oligodendrocytes in the developing brain: Implications for perinatal white matter injuryGlia 66:221–238https://doi.org/10.1002/glia.23256
47
1. Windrem M. S.
2. et al.
2020Human Glial Progenitor Cells Effectively Remyelinate the Demyelinated Adult BrainCell Rep 31https://doi.org/10.1016/j.celrep.2020.107658
48
1. Yang L.
2. et al.
2021Transcriptional profiling reveals the transcription factor networks regulating the survival of striatal neuronsCell death & disease 12:262–262https://doi.org/10.1038/s41419-021-03552-8

Article and author information

Author information

Jialin Li
State Key Laboratory of Medical Neurobiology and MOE Frontiers Center for Brain Science, Institutes of Brain Science, and Department of Neurology, Zhongshan Hospital, Fudan University, Shanghai 200032, P.R. China
- These authors contributed equally to this work.
Feihong Yang
Department of Anesthesiology, Shuguang Hospital Affiliated with Shanghai University of Traditional Chinese Medicine, Shanghai 201203, P.R. China
- These authors contributed equally to this work.
Yu Tian
State Key Laboratory of Medical Neurobiology and MOE Frontiers Center for Brain Science, Institutes of Brain Science, and Department of Neurology, Zhongshan Hospital, Fudan University, Shanghai 200032, P.R. China
- These authors contributed equally to this work.
Ziwu Wang
State Key Laboratory of Medical Neurobiology and MOE Frontiers Center for Brain Science, Institutes of Brain Science, and Department of Neurology, Zhongshan Hospital, Fudan University, Shanghai 200032, P.R. China
Dashi Qi
Center for Clinical Research and Translational Medicine, Yangpu Hospital, School of Medicine, Tongji University, Shanghai 200090, P.R. China
Zhengang Yang
State Key Laboratory of Medical Neurobiology and MOE Frontiers Center for Brain Science, Institutes of Brain Science, and Department of Neurology, Zhongshan Hospital, Fudan University, Shanghai 200032, P.R. China
Jiangang Song
Department of Anesthesiology, Shuguang Hospital Affiliated with Shanghai University of Traditional Chinese Medicine, Shanghai 201203, P.R. China
Jing Ding
State Key Laboratory of Medical Neurobiology and MOE Frontiers Center for Brain Science, Institutes of Brain Science, and Department of Neurology, Zhongshan Hospital, Fudan University, Shanghai 200032, P.R. China
Xin Wang
State Key Laboratory of Medical Neurobiology and MOE Frontiers Center for Brain Science, Institutes of Brain Science, and Department of Neurology, Zhongshan Hospital, Fudan University, Shanghai 200032, P.R. China
- Correspondence Xin Wang, Email: wang.xin@zs-hospital.sh.cn, Zhuangzhi Zhang, Email: zz_zhang@fudan.edu.cn
Zhuangzhi Zhang
State Key Laboratory of Medical Neurobiology and MOE Frontiers Center for Brain Science, Institutes of Brain Science, and Department of Neurology, Zhongshan Hospital, Fudan University, Shanghai 200032, P.R. China
ORCID iD: 0000-0002-9860-6689
- Correspondence Xin Wang, Email: wang.xin@zs-hospital.sh.cn, Zhuangzhi Zhang, Email: zz_zhang@fudan.edu.cn

Version history

Sent for peer review: November 26, 2023
Preprint posted: December 3, 2023
Reviewed Preprint version 1: March 5, 2024
Reviewed Preprint version 2: August 28, 2024
Version of Record published: September 11, 2024

Copyright

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Not revised: This Reviewed Preprint includes the authors’ original preprint (without revision), an eLife assessment, and public reviews.

Reviewing Editor
Samuel Pleasure
University of California, San Francisco, San Francisco, United States of America
Senior Editor
Kathryn Cheah
University of Hong Kong, Hong Kong, Hong Kong

Reviewer #1 (Public Review):

Summary:
In this manuscript the authors re-examine the developmental origin of cortical oligodendrocyte (OL) lineage cells using a combination of strategies, focussing on the question of whether the LGE generates cortical OL cells. The paper is interesting to myelin biologists, the methods used are appropriate and, in general, the study is well-executed, thorough, and persuasive, but not 100% convincing.

Strengths, weaknesses, and recommendations:
The first evidence presented that the LGE does not generate OLs for the cortex is that there are no OL precursors 'streaming' from the LGE during embryogenesis, unlike the MGE (Figure 1A). This in itself is not strong evidence, as they might be more dispersed. In fact, in the images shown, there is no obvious 'streaming' from the MGE either. Note that in Figure 1 there is no reference to the star that is shown in the figure.

The authors then electroporate a reporter into the LGE at E13.5 and examine the fate of the electroporated cells (Figures 1C-E). They find that electroporated cells became neurons in the striatum and in the cortex but no OLs for the cortex. There are two issues with this: first, there is no quantification, which means there might indeed be a small contribution from the LGE that is not immediately obvious from snapshot images. Second, it is unexpected to find labelled neurons in the cortex at all since the LGE does not normally generate neurons for the cortex! Electroporations are quite crude experiments as targeting is imprecise and variable and not always discernible at later stages. For example, in Figure 1D, one can see tdTOM+ cells near the AEP, as well as the striatum. Hence, IUE cannot on its own be taken as proof that there is no contribution of the LGE to the cortical OL population.

The authors then use an alternative fate-mapping approach, again with E13.5 electroporations (Figure 2). They find only a few GFP+ cells in the cortex at E18 (Figures 2C-D) and P10 (Figure 2E) and these are mainly neurons, not OL lineage cells. Again, there is no quantification.

Figure 3 is more convincing, but the experiments are incomplete. Here the authors generate triple-transgenic mice expressing Cre in the cortex (Emx1-Cre) and the MGE (Nkx2.1-Cre) as well as a strong nuclear reporter (H2B-GFP). They find that at P0 and P10, 97-98% of OL-lineage cells (SOX10+ or PDGFRA+) in the cortex are labelled with GFP (Figure 3). This is a more convincing argument that the LGE/CGE might not contribute significant numbers of OL lineage cells to the cortex, in contrast to the Kessaris et at. (2006) paper, which showed that Gsh2-Cre mice label ~50% of SOX10+ve cells in the motor cortex at P10. The authors of the present paper suggest that the discrepancy between their study and that of Kessaris et al. (2006) is based on the authors' previous observation (Zhang et al 2020) (https://doi.org/10.1016/j.celrep.2020.03.027) that GSH2 is expressed in intermediate precursors of the cortex from E18 onwards. If correct, then Kessaris et al. might have mistakenly attributed Gsh2-Cre+ lineages to the LGE/CGE when they were in fact intrinsic to the cortex. However, the evidence from Zhang et al 2020 that GSH2 is expressed by cortical intermediate precursors seems to rest solely on their location within the developing cortex; a more convincing demonstration would be to show that the GSH2+ putative cortical precursors co-label for EMX1 (by immunohistochemistry or in situ hybridization), or that they co-label with a reporter in Emx1-driven reporter mice. This demonstration should be simple for the authors as they have all the necessary reagents to hand. Without these additional data, the assertion that GSX2+ve cells in the cortex are derived from the cortical VZ relies partly on an act of faith on the part of the reader.

Note that Tripathi et al. (2011, "Dorsally- and ventrally-derived oligodendrocytes have similar electrical properties but myelinate preferred tracts." J. Neurosci. 31, 6809-6819) found that the Gsh-Cre+ OL lineage contributed only ~20% of OLs to the mature cortex, not ~50% as reported by Kessaris et al. (2006). If it is correct that these Gsh2-derived OLs are from the cortical anlagen as the current paper claims, then it would raise the possibility that the ventricular precursors of GSH2+ intermediate progenitors are not uniformly distributed through the cortical VZ but are perhaps localized to some part of it. Then the contribution of Gsh2-derived OLs to the cortical population could depend on precisely where one looks relative to that localized source. It would be a nice addition to the current manuscript if the authors could explore the distribution of their GSH2+ intermediate precursors throughout the developing cortex. In any case, Tripathi et al. (2011) should be cited.

Finally, the authors deleted Olig2 in the MGE and found a dramatic reduction of PDGFRA+ and SOX10+ cells in the cortex at E14 and E16 (Figure 4A-F). This further supports their conclusion that, at least at E16, there is no significant contribution of OLs from ventral sources other than the MGE/AEP. This does not exclude the possibility that the LGE/CGE generates OLs for the cortex at later stages. Hence, on its own, this is not completely convincing evidence that the LGE generates no OL lineage cells for the cortex.

https://doi.org/10.7554/eLife.94317.1.sa1

Reviewer #2 (Public Review):

Traditional thinking has been that cortical oligodendrocyte progenitor cells (OPCs) arise in the development of the brain from the medial ganglionic eminence (MGE), lateral/caudal ganglionic eminence (LGE/CGE), and cortical radial glial cells (RGCs). Indeed a landmark study demonstrated some time ago that cortical OPCs are generated in three waves, starting with a ventral wave derived from the medial ganglionic eminence (MGE) or the anterior entopeduncular area (AEP) at embryonic day E12.5 (Nkx2.1+ lineage), followed by a second wave of cortical OLs derived from the lateral/caudal ganglionic eminences (LGE/CGE) at E15.5 (Gsx2+/Nkx2.1- lineage), and then a final wave occurring at P0, when OPCs originate from cortical glial progenitor cells (Emx1+ lineage). However, the authors challenge the idea in this paper that cortical progenitors are produced from the LGE. They have found previously that cortical glial progenitor cells were also found to express Gsx2, suggesting this may not have been the best marker for LGE-derived OPCs. They have used fate mapping experiments and lineage analyses to suggest that cortical OPCs do not derive from the LGE.

Strengths:
(1) The data is high quality and very well presented, and experiments are thoughtful and elegant to address the questions being raised.

(2) The authors use two elegant approaches to lineage trace LGE derived cells, namely fate mapping of LGE-derived OPCs by combining IUE (intrauterine electroporation) with a Cre recombinase-dependent IS reporter, and Lineage tracing of LGE-derived OPCs by combining IUE with the PiggyBac transposon system. Both approaches show convincingly that labelled LGE-derived cells that enter the cortex do not express OPC markers, but that those co-labelling with oligodendrocyte markers remain in the striatum.

(3) The authors then use further approaches to confirm their findings. Firstly they lineage trace Emx1-Cre; Nkx2.1-Cre; H2B-GFP mice. Emx1-Cre is expressed in cortical RGCs and Nkx2.1-Cre is specifically expressed in MGE/AEP RGCs. They find that close to 98% of OPCs in the cortex co-label with GFP at later times, suggesting the contribution of OPCs from LGE is minimal.

(4) They use one further approach to strengthen the findings yet further. They cross Nkx2.1-Cre mice with Olig2 F/+ mice to eliminate Olig2 expression in the SVZ/VZ of the MGE/AEP (Figures 4A-B). The generation of MGE/AEP-derived OPCs is inhibited in these Olig2-NCKO conditional mice. They find that the number of cortical progenitors at E16.5 is reduced 10-fold in these mice, suggesting that LGE contribution to cortical OPCs is minimal.

Weaknesses:
(1) The authors use IUE in experiments mentioned in point 2 of 'Strengths' above (Figures 1 and 2) and claim that the reporter was delivered specifically into LGE VZ at E13.5 using this IUE. It would be nice to see some sort of time course of delivery after IUE to show the reporter is limited to LGE VZ at early times post-IUE.

(2) In the experiments mentioned in point 3 of 'Strengths' (Figure 3), statistical analysis showed that only approximately 2% of OPCs were GFP-negative cells. This 2% could possibly be derived from the LGE/CGE so does not totally rule out that LGE contributes some cortical OPCs.

(3) In the experiments mentioned in point 4 of 'Strengths' (Figure 4), they do still find cortical OPCs at E16.5 in the Olig2-NCKO conditional mice. It is unclear whether this is due to the recombination efficiency of the CRE enzyme not being 100%, or whether there is some LGE contribution to the cortical OPCs.

Impact of Study:
The authors show elegantly and convincingly that the contribution of the LGE to the pool of cortical OPCs is minimal. The title should perhaps be that the LGE contribution is minimal rather than no contribution at all, as they are not able to rule out some small contribution from the LGE. These findings challenge the traditional belief that the LGE contributes to the pool of cortical OPCs. The authors do show that the LGE does produce OPCs, but that they tend to remain in the striatum rather than migrate into the cortex. It is interesting to wonder why their migration patterns may be different from the MGE-derived OPCs which migrate to the cortex. The functional significance of these different sources of OPCs for adult cortex in homeostatic or disease states remains unclear though.

https://doi.org/10.7554/eLife.94317.1.sa0

Significance of findings

Strength of evidence

Abstract

Introduction

Results

Fate mapping of LGE-derived OPCs by combining IUE with a Cre recombinase-dependent IS reporter

Fate mapping of LGE-derived OPCs by combining IUE with a Cre recombinase-dependent IS reporter

Lineage tracing of LGE-derived OPCs by combining IUE with the PiggyBac transposon system

Fate mapping of LGE-derived OPCs by combining IUE with PiggyBac transposon system

An exclusion strategy revealed that cortical OPCs do not originate from the LGE/CGE

An exclusion strategy showed that cortical OPCs do not come from the LGE/CGE

The MGE/AEP was the sole ventral source of the cortical OPCs

The MGE may be the sole ventral source of cortical OPCs

Discussion

Why the LGE cannot generate cortical OPCs during cortical development

Cortical-derived OLs/OPCs make up the majority of the cortical OLs/OPCs in the adult brain

Methods

Animals

Tissue preparation

Plasmid Construction

In utero electroporation (IUE)

Immunofluorescence

In situ hybridization (ISH)

3D Reconstruction

Image acquisition and statistical analysis

Key resource table

Acknowledgements

Data availability statement

Potential conflicts of interest

Author contributions

Figure legends

References

Article and author information

Author information

Jialin Li#

Feihong Yang#

Yu Tian#

Ziwu Wang

Dashi Qi

Zhengang Yang

Jiangang Song

Jing Ding

Xin Wang

Zhuangzhi Zhang

Version history

Copyright

Peer review process

Editors

Jialin Li

Feihong Yang

Yu Tian