A peer-reviewed study in mice found no evidence that sustained NeuroD1 expression turned microglia into neurons. The result held in physiological brain conditions and after traumatic brain injury, using several ways to track the labeled cells and examine their identity.
The researchers used virus-free genetic lineage tracing, a method that marks a cell population and follows its descendants over time. They activated NeuroD1 in microglia with either CX3CR1-CreER or TMEM119-CreER genetic systems crossed with LSL-NeuroD1-GFP mice. Matching CX3CR1-Ai14 and TMEM119-Ai14 reporter mice served as controls.
The marked cells kept their microglial identity
At 4, 10, 18, 30 and 45 days after tamoxifen activation, GFP-positive cells remained positive for IBA1, a microglial marker. They did not show the neuronal markers NeuN, TUJ1 or MAP2. The same pattern appeared when the researchers used TMEM119-based tracing, another way of labeling resident microglia.
Two-photon imaging, which allows repeated views of living tissue, also showed no obvious shift toward a neuronal shape. The labeled cells stayed highly branched, with short moving processes typical of microglia. Some labeled microglia gradually disappeared during the imaging period instead.
The disappearance was accompanied by a rise in a marker of programmed cell death. Among reporter-positive cells, TUNEL positivity was approximately 0.01% in controls and 1.7% at day 18 in the NeuroD1-expressing group. The higher TUNEL signal accompanied the progressive disappearance seen in the imaging experiments.
The study also found molecular changes involving autophagy, the process cells use to break down and recycle their components. Atg5 was significantly upregulated, while LC3 and P62 levels in reporter-positive cells were about four times higher than in controls. These findings accompanied the observed cell loss, without establishing which process came first.
Brain injury did not change the outcome
The researchers then tested the same question after traumatic brain injury. In injured brains, GFP-labeled cells did not co-localize with NeuN or MAP2, and single-cell RNA sequencing found no neuronal cell cluster. Microglial markers remained expressed in the labeled cells.
The negative result was also seen when resident microglia were labeled before or after injury, as well as in the TMEM119-based experiments. These results do not cover every possible injury model or induction protocol.
At 19 days after tamoxifen induction in the injury experiments, the NeuroD1-expressing mice showed no improvement over controls in grip strength or rotarod performance.
What the evidence can and cannot say
For the sequencing analysis, the researchers integrated four experimental groups and counted genes as significantly changed when the absolute base-2 logarithm of their fold change was greater than 0.25 and the p-value was below 0.05. In plain terms, the analysis used set thresholds for both the size of a gene-expression difference and the statistical evidence behind it.
The number of mice varied by experiment. Lineage quantifications in the cortex and striatum used four or five mice per group. Injury behavior experiments used six to 10 mice per group, while histological injury measurements used five per group.
This is a mouse result from the physiological and injury conditions tested. It does not show that microglia conversion is impossible under every injury model or induction protocol, or establish the outcome in people. The findings argue against NeuroD1-driven conversion in the tested settings, rather than settling every possible version of the approach.
Histological measurements were averaged across multiple brain sections per mouse and assessed independently in a double-blind manner. The researchers used one-way analysis of variance with Tukey testing, two-tailed independent t-tests and two-way analysis of variance for behavior, and reported that no outliers were excluded. The study also reports institutional ethical approval, efforts to minimize animal suffering and animal numbers, and no significant sex differences.
The record behind the study
The paper reports grant support from national, Shanghai, postdoctoral and laboratory programs awarded to Yanxia Rao, Bo Peng, Xiaoyu Li and Pei Ouyang. Its funding statement says the funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
The sequencing data were deposited in the Gene Expression Omnibus under accession GSE277222, allowing the reported dataset to be examined separately.
Paper data and sources
Original title: In vivo virus-free lineage tracing and live imaging reveal that NeuroD1 does not reprogram microglia into neurons
Authors: Li X, Li Y, Cao Y et al.
Journal/Repository: eLife
Status: Peer-reviewed
First online: 2026-10-19
DOI: Not available
Original paper · Full text