Intact learning and memory in mice incapable of de novo myelination

Technical Summary: Intact learning and memory in mice incapable of de novo myelination

Problem and Background
Previous research has established that the adult oligodendrocyte (OL) lineage exhibits plasticity during learning experiences, such as motor skill acquisition and memory consolidation. Studies utilizing conditional knockout of Myelin regulatory factor (Myrf) in adult OL precursors (OLPs) demonstrated that blocking the differentiation of new myelinating OLs impairs motor skill learning (specifically on a complex running wheel) and the long-term consolidation of fear and spatial memories. These findings supported a hypothesis that the generation of new, compact myelin sheaths is essential for these cognitive processes, potentially by accelerating action potential conduction (saltatory conduction) or fine-tuning circuit timing. However, Myrf deletion causes OLPs to undergo an abortive differentiation program and die as pre-myelinating OLs (pmOLs) before forming myelin. Consequently, it remained unclear whether the observed behavioral deficits were caused by the loss of new myelin sheaths and their canonical function (speeding conduction) or by the loss of a non-canonical function of the newly differentiating OLs (or the pmOLs themselves) prior to myelination.

Methodology
To distinguish between the canonical role of compact myelin and other functions of new OLs, the authors generated a conditional knockout (cKO) mouse model targeting Myelin basic protein (Mbp).

• Genetic Strategy: The authors inserted loxP sites flanking the Mbp transcription start site (exon 5b) within the Golli-Mbp transcription unit. Crossing Mbp^fl/fl mice with Cre drivers (Ng2-CreERTM, Pdgfra-CreERT2, or Sox10-iCreERT2) allowed for tamoxifen-inducible deletion of Mbp specifically in the OL lineage of adult mice. This approach prevents the synthesis of MBP protein and the formation of compact myelin sheaths and normal nodes of Ranvier in newly differentiated OLs, while allowing the cells to survive and persist (unlike Myrf-cKO where cells die).
• Validation: The authors confirmed that Mbp-cKO mice (specifically Ng2-Mbp-cKO and Sox10-Mbp-cKO) successfully eliminated Mbp mRNA and protein in newly generated OLs. Electron microscopy and immunolabeling revealed that while these mice formed loose, "shiverer-like" membrane wraps, they failed to form compact myelin or normal nodes of Ranvier (characterized by abnormal CASPR1+ paranodes and reduced NaV clustering).
• Behavioral Assays:
  • Motor Learning: Mice were trained on a "complex wheel" with unequally spaced rungs to assess motor skill learning and long-term memory retention (tested after 30 days).
  • Fear Conditioning: Mice underwent contextual fear conditioning to assess short-term (24h) and long-term (30-day) memory consolidation.
• Metabolic Control: To test the hypothesis that metabolic coupling (lactate transfer via MCT1) is required for learning, the authors generated Mct1-cKO mice (Pdgfra-Mct1-cKO and Sox10-Mct1-cKO) and subjected them to the same behavioral assays.
• Positive Control: The study re-validated the Myrf-cKO phenotype in their facility to confirm that blocking OL differentiation (and cell survival) recapitulates the previously reported learning deficits.

Key Results

1. Intact Motor Learning: Despite the inability to form compact myelin or normal nodes of Ranvier in newly generated OLs, Mbp-cKO mice learned the complex wheel task with a trajectory indistinguishable from wild-type littermates. They mastered the skill and retained it over a 30-day period. This result held true regardless of whether the deletion was restricted to OLPs (Ng2/Pdgfra-Cre) or extended to all OL lineage cells (Sox10-Cre).
2. Intact Fear Memory: Mbp-cKO mice displayed normal freezing behavior 24 hours and 30 days after fear conditioning, indicating that the consolidation of long-term contextual fear memory does not depend on the formation of new compact myelin sheaths.
3. Metabolic Coupling Not Required: Mct1-cKO mice, which lacked MCT1 in newly formed OLs, also showed no deficits in motor learning or fear memory consolidation, suggesting that lactate transfer from OLs to axons is not critical for these specific learning tasks.
4. Confirmation of Myrf-cKO Deficits: In contrast to the Mbp-cKO results, Pdgfra-Myrf-cKO mice (where new OLs die) failed to learn the complex wheel task and showed impaired long-term fear memory consolidation, confirming that the presence of new, surviving OLs is necessary for these processes.
5. Myelin Stability: The authors noted that pre-existing myelin sheaths (formed before tamoxifen administration) persisted for months in the motor cortex, suggesting that the behavioral performance of Mbp-cKOs was supported by pre-existing myelin, while the de novo myelin was functionally absent.

Significance and Claims
The paper concludes that the generation of new, compact myelin sheaths and the associated acceleration of action potentials (the canonical function of myelin) are not essential for motor skill learning or the consolidation of long-term fear memories in adult mice.

Instead, the data suggest that the requirement for new OLs in learning and memory depends on a non-canonical property of pre-myelinating or myelinating OLs. Since Myrf-cKO (which eliminates new OLs) causes deficits but Mbp-cKO (which allows new OLs to survive but prevents myelination) does not, the critical factor is likely the presence of the new OLs themselves in a pre-myelinating or early-differentiating state, rather than the myelin sheath they eventually produce. The authors propose that these cells may contribute to learning through mechanisms such as:

• Modifying the extracellular matrix (e.g., laying down perineuronal nets or "dandelion clock-like" structures).
• Synaptic pruning or phagocytosis.
• Other forms of metabolic or trophic support distinct from MCT1-mediated lactate transfer.

The study reframes the role of adult neurogenesis in the OL lineage, shifting the focus from the structural output (myelin speed) to the cellular process of differentiation and the potential non-canonical functions of the differentiating cells themselves.