# Selective neuronal restoration of progranulin does not prevent the frontotemporal dementia like-phenotype of progranulin knockout mice

## TL;DR

Restoring progranulin in neurons of mice with frontotemporal dementia-like symptoms did not prevent brain inflammation or behavioral issues, suggesting microglial involvement is critical.

## Contribution

Demonstrates that neuronal progranulin restoration alone is insufficient to prevent FTD-like pathology in mice.

## Key findings

- Neuronal PGRN restoration did not reduce microgliosis or astrogliosis in progranulin knockout mice.
- Synaptic pruning and lipidomic changes were partially attenuated in rescued mice.
- Behavioral deficits like hyperactivity and memory impairment persisted despite neuronal PGRN restoration.

## Abstract

Progranulin (PGRN) is a neurotrophic and anti-inflammatory factor produced mainly by neurons and microglia in the central nervous system. Progranulin haploinsufficiency causes frontotemporal dementia (FTD). It is unclear to what extent neuronal versus microglial PGRN deficiency contributes to FTD pathology. In this study, we restored progranulin in neurons in progranulin knockout mice using Nestin-driven expression of mouse Grn transgene in a knockout background (NesGrn KOBG). They were compared with full PGRN KO mice and floxed control mice that carry a loxP flanked STOP codon in front of mGrn transgene (Grn-flfl). The expected neuron-only PGRN rescue was confirmed at RNA and protein level in brain tissue and primary cells, and single nucleus RNA sequencing. Despite neuronal PGRN-restoration, there was no difference in microgliosis, astrogliosis, and microglia phenotypes as assessed by histology, microglia morphometry and bulk RNAseq showing strong upregulation of microglia-associated genes equally in both KO lines. However, a microglial subpopulation with a phagocyte signature expressing Gpnmb, Lgals3, Atp6v0d2 and Apobec1 occurred only in PGRN KO brain, and accordingly, the loss of synapses and dendritic spines, which is caused by excessive synaptic pruning in PGRN KO mice, was partially attenuated in NesGrn KOBG mice. Lipidomic studies showed that phosphatidylserine eat-me-signals were increased in PGRN KO but not in NesGrn KOBG brain. Furthermore, some neuronal genes involved in axonal structure and dynamics were co-restored with progranulin in NesGrn KOBG mice. However, the modest improvement of neuronal health was not associated with an improvement of FTD-like behavior including hyperactivity, compulsive licking and impaired avoidance learning and memory. The results suggest that (still) viable neurons do not provide (sufficient) progranulin to prevent microgliosis but may shape the phenotype by presenting or hiding eat-me signals. Nonetheless, neuron-only-progranulin restoration may be insufficient to halt the progression of FTD.

The online version contains supplementary material available at 10.1186/s12974-025-03665-3.