m6A and YTHDF1 regulate motor learning and drug sensitization differently across striatal neuron types
An RNA mark that regulates rapid, local protein synthesis
A study published in PNAS reports that N6-methyladenosine (m6A), a modification of messenger RNA (mRNA), and the protein that recognizes it, YTHDF1, are involved in motor learning and drug-induced sensitization in the mouse striatum. Removing the same gene produced behavioral changes that differed in direction between dopamine D1 receptor-expressing and D2 receptor-expressing neurons. Understanding normal learning and maladaptive responses requires distinguishing the target cell types.
Long-term retention of learning requires new protein synthesis. Yet regulating gene transcription in the nucleus alone makes it difficult to accommodate rapid, local changes at numerous synapses. This has drawn attention to regulation of translation, the process by which mRNA is decoded into protein. The researchers deleted Mettl14, which contributes to m6A formation, or Ythdf1, which encodes an m6A reader protein, in specific cell types in the striatum, a region involved in motor learning and action selection.
Learning and sensitization diverge in D1 and D2 neurons
The researchers compared conditional knockout mice with littermate controls. They assessed motor learning using the rotarod task, which measures how long a mouse remains on a rotating rod before falling, and monitored neural activity during behavior by recording calcium signals through optical fibers.
Removing Ythdf1 from D1 neurons impaired rotarod motor learning, whereas removal from D2 neurons produced no significant difference on this task. The D1- and D2-neuron comparisons included seven and six mice per genotype, respectively. In the Mettl14 knockout experiments, deletion in D1 neurons also impaired motor learning and learning-related changes in neural activity.
Drug responses changed in opposite directions depending on cell type. With repeated cocaine administration, Ythdf1 deletion in D1 neurons reduced the acute locomotor response and sensitization, while deletion in D2 neurons increased both. Sensitization is the increase in a drug response after repeated exposure. D2-neuron deletion also reduced the acute cataleptic response to the antipsychotic haloperidol and sensitization of that response. These two drug experiments are animal models for studying addiction-related responses and antipsychotic-induced parkinsonism, respectively.
At the molecular level, stimulus-responsive protein synthesis was central. The researchers cultured striatal neurons from mice with a whole-body Ythdf1 knockout and measured protein synthesis using labeling with L-homopropargylglycine (HPG), a methionine analog incorporated into newly made proteins. Knockout neurons had higher baseline synthesis but showed no further increase after treatment with the D1 receptor agonist SKF-81297. Synthesis increased in wild-type neurons after the same stimulus, so YTHDF1’s role cannot be explained simply as uniformly increasing protein production.
Therapeutic potential and the challenge of cell selectivity
Crosslinking and immunoprecipitation followed by sequencing (CLIP-seq) showed an increase in the number of RNA targets bound by YTHDF1 after cocaine administration. It remains unclear whether this was driven by a change in binding capacity, increased m6A modification, or altered target RNA expression. The similar behavioral outcomes of Mettl14 and Ythdf1 deletion support the interpretation that YTHDF1 is a major mediator of m6A action in the adult mouse striatum.
The study does not demonstrate improved memory or therapeutic efficacy in humans. Cell type-specific gene deletion cannot be directly applied to human treatment, and the heterozygous controls used in the conditional knockout experiments are not equivalent to fully wild-type animals. Given the impairment of normal motor learning and the increased cocaine sensitization after D2-neuron deletion, the therapeutic benefits of broadly inhibiting YTHDF1 are difficult to generalize. Directly examining synapse-specific local translation during behavior also remains a task for future studies.
The methods and results are available in the full text provided by Europe PMC.
Proceedings of the National Academy of Sciences, Volume 123, Issue 16, April 2026. SignificanceNeural plasticity is essential for learning and adaptation to environmental changes. However, maladaptive plasticity leads to neurological and psychiatric disorders. It is known that long-term plasticity requires new protein synthesis. By ...
This study provides a basis for assessing both cell type-specific benefits and risks when exploring YTHDF1 as a therapeutic target. Research into strategies to reduce antipsychotic-related motor side effects should examine not only D2-neuron responses but also normal motor learning and responses to other drugs. Distinguishing D1 and D2 neurons is also central to research on addiction-related sensitization. Indeed, deletion in D2 neurons increased cocaine sensitization. Follow-up target validation should measure both baseline protein synthesis and the capacity to regulate synthesis after stimulation. Because these findings come from gene deletion in mice and cultured cells, whether drugs can safely reproduce the same effects requires separate testing.