Excessive expression of progranulin leads to neurotoxicity rather than neuroprotection.

Frontotemporal dementia (FTD) is an early onset form of dementia characterized by frontotemporal lobar atrophy accompanied by behavioral, personality, language, and motor deficits. Heterozygous mutations in GRN gene encoding progranulin (PGRN) are the genetic causes of FTD. Since PGRN is a neurotrophic and anti-inflammatory factor, most FTD-related PGRN mutations are thought to cause FTD due to haploinsufficiency. Therefore, therapies that increase PGRN levels by the administration of recombinant PGRN or viral vectors are attracting attention as an approach to the treatment of FTD. However, the mechanisms underlying the neuroprotective effects of PGRN remain unclear. To investigate the neuroprotective mechanisms of PGRN in vivo, we generated human PGRN transgenic (Tg) mice using the CAG promoter. Unexpectedly, mice overexpressing wild-type human PGRN showed a shortened lifespan and cerebellar dysfunction, including the loss of Purkinje cells. Furthermore, PGRN Tg mice developed cognitive impairment, gliosis, and lysosomal abnormalities. FTD-causative R432C-PGRN mutant Tg mice also showed FTD-like phenotypes, such as neuronal loss, gliosis, and behavioral deficits. In cultured cells, overexpression of PGRN induced endoplasmic reticulum (ER) stress and apoptotic cell death, suggesting that continuous increases in PGRN expression through viral vectors or genetic manipulation are neurotoxic and that PGRN-replacement therapy may be required to maintain optimal PGRN levels for each neuron type and brain region.