Progranulin Loss Impairs Cathepsin D Activity and Raises Insoluble TDP-43 in Mouse Brains

Loss of the progranulin protein disrupts the function of a key lysosomal enzyme and pushes the RNA-binding protein TDP-43 toward an insoluble state in the mouse brain, according to a new study published in Frontiers in Molecular Neuroscience. The work, carried out by researchers at KU Leuven and University Hospitals Leuven, examined mice engineered to lack one or both copies of the Grn gene, the mouse counterpart of GRN, a major genetic cause of frontotemporal dementia (FTD) in humans. By combining progranulin deficiency with partial loss of cathepsin D, a lysosomal protease, the team mapped how these two molecules interact to maintain lysosomal health and protein quality control in the brain.

Frontotemporal dementia is the second most prevalent form of early-onset dementia in people under 65 years of age, characterized by progressive degeneration of the frontal and anterior temporal lobes that produces behavioral and language deficits. Up to 40 percent of FTD cases are familial, and mutations in three genes—MAPT, GRN, and C9orf72—account for roughly 80 percent of those inherited cases. While MAPT mutations lead to tau accumulation, GRN and C9orf72 mutations are associated with aggregation of TDP-43, a nuclear DNA- and RNA-binding protein essential for RNA processing whose mislocalization or aggregation is harmful to neuronal function. Heterozygous loss-of-function mutations in GRN invariably produce FTD with ubiquitinated TDP-43-positive inclusions, and they cut circulating progranulin levels by about half in plasma and cerebrospinal fluid. How this reduction drives neurodegeneration, however, has remained unclear.

To probe the mechanism, the researchers generated six genotypes of mice: wild-type animals, heterozygous cathepsin D single knockouts, homozygous progranulin knockouts, double knockouts lacking both progranulin copies and one cathepsin D copy, heterozygous progranulin knockouts, and double heterozygotes. All molecular analyses were performed on brain tissue from 18-month-old mice, using western blotting, an in vitro enzymatic activity assay, biochemical fractionation, ELISA, and immunofluorescence microscopy. Homozygous cathepsin D-deficient offspring were excluded because complete cathepsin D deficiency causes early postnatal lethality.

The first major finding concerned cathepsin D itself. Two-way ANOVA revealed significant main effects of both the progranulin and cathepsin D genotypes on mature and immature cathepsin D levels. Post hoc comparisons showed that mature cathepsin D was significantly elevated in homozygous progranulin knockout mice compared with wild-type controls, and immature pro-cathepsin D was likewise significantly increased in both homozygous progranulin knockout conditions. The ratio of mature to immature cathepsin D remained unchanged across groups, suggesting that major dysregulation of cathepsin D processing was not evident under the conditions examined. Yet when the team measured enzymatic activity in cell lysates using a fluorescent substrate assay, activity was significantly reduced in homozygous progranulin-deficient mice despite the elevated protein abundance. Even after normalizing activity to mature cathepsin D levels, the effect of progranulin genotype remained significant, driven by the decrease in the homozygous knockout conditions. The authors interpret this uncoupling of abundance and activity as consistent with a proposed chaperone role for progranulin in supporting proper cathepsin D function.

The study next turned to lysosomal membrane markers. LAMP1 and LAMP2 are canonical structural markers of late endosomal and lysosomal compartments, with LAMP2 additionally involved in lysosomal trafficking and autophagy. Western blot analysis showed significant effects of both genotypes on LAMP1 and LAMP2 protein levels. Homozygous progranulin-deficient mice and heterozygous cathepsin D single knockouts showed significantly increased LAMP1 and LAMP2 compared with wild-type controls, whereas heterozygous progranulin-deficient mice did not differ significantly from wild type, indicating that complete loss of progranulin is required to elicit this effect. The LAMP2-to-LAMP1 ratio was unchanged between groups, suggesting the relative balance of these membrane markers was preserved despite the alterations in lysosomal biology. The authors note that the elevation could reflect either enhanced lysosomal biogenesis or impaired lysosomal turnover.

Immunofluorescence microscopy added a spatial dimension to these findings. In homozygous progranulin-deficient mice, LAMP1-positive lysosomes were not only increased in abundance and size but also exhibited pronounced perinuclear clustering, a pattern the researchers interpret as indicative of defects in lysosomal trafficking, positioning, or fusion dynamics. Because the clustering was so extensive, accurate segmentation and reliable quantification of lysosome number and size proved infeasible, and the authors present these imaging data as qualitative evidence of altered lysosomal organization. They caution that the data do not distinguish whether the phenotype reflects changes in trafficking, positioning, fusion, or compensatory lysosomal accumulation.

The third strand of the investigation concerned TDP-43 solubility. Biochemical fractionation of brain lysates separated soluble and insoluble fractions, and western blotting revealed no significant changes in total or soluble TDP-43 across genotypes. The insoluble fraction told a different story: the progranulin genotype significantly influenced insoluble TDP-43 levels, with significant increases in homozygous progranulin knockout mice and, notably, also in heterozygous progranulin knockout mice compared with wild-type controls. This means that even a single functional copy of Grn was sufficient to shift TDP-43 toward the detergent-insoluble fraction, a biochemical change the authors describe as potentially reflecting impaired protein homeostasis. However, when the team immunostained whole-brain sections from end-stage mice for phosphorylated TDP-43, no cytoplasmic inclusions were detected in neurons or glial cells in any group at the analyzed timepoint. The authors therefore frame the insoluble TDP-43 accumulation as an early biochemical alteration rather than evidence of established TDP-43 proteinopathy, consistent with prior reports that similar homozygous progranulin knockout mice lack overt TDP-43 pathology.

Lifespan analysis revealed a striking synergistic effect. Kaplan–Meier survival analysis showed significant differences among the six groups, with double-knockout mice lacking both progranulin copies and one cathepsin D copy exhibiting a median survival of only 639 days, markedly shorter than all other genotypes. Wild-type, homozygous progranulin knockout, and heterozygous cathepsin D single knockout mice displayed normal or only modestly reduced lifespans, indicating that neither deficiency alone is sufficient to produce the pronounced survival phenotype. Functional testing at 12 and 18 months of age showed no genotype differences on rotarod, but hindlimb clasping scores at 18 months revealed significant impairment in the double-knockout mice compared with both single-mutant conditions, suggesting emerging neurological deficits in the combined knockout state. The authors note that the biochemical abnormalities were already evident at earlier ages, well before any detectable survival differences emerged.

The study also uncovered a bidirectional relationship between the two proteins. ELISA measurements of whole-brain lysates confirmed reduced progranulin in the knockout conditions, as expected, but showed that heterozygous cathepsin D loss significantly increased progranulin levels in progranulin-expressing animals. Because heterozygous cathepsin D single knockout mice also showed increased relative cathepsin D activity after normalization to mature protein levels, the authors suggest a feedback mechanism in which cathepsin D levels may influence progranulin abundance while progranulin modulates cathepsin D activity. Sex-stratified analysis of the ELISA data did not reveal major sex-dependent differences that altered the interpretation.

The authors are explicit about the limitations of their work. The cathepsin D activity assay was performed in vitro on cell lysates rather than within intact lysosomes, so the observed reduction reflects decreased catalytic capacity under assay conditions and may not fully represent physiological lysosomal proteolysis in living cells. The study did not evaluate other lysosomal proteases, such as cathepsin B and cathepsin L, which also process progranulin into granulin peptides, so broader alterations in lysosomal protease homeostasis cannot be excluded. All progranulin measurements came from whole-brain lysates, providing no information on subcellular or lysosome-specific localization. The mouse models used did not develop robust neurodegenerative phenotypes, marked behavioral impairment, or substantial survival deficits—a well-recognized limitation of progranulin-deficient models on C57BL/6-based backgrounds—and behavioral characterization beyond motor testing revealed no reproducible genotype-dependent abnormalities. Inflammatory signaling was not investigated. The authors conclude that the findings should be interpreted primarily as evidence that cathepsin D modulates molecular and cellular phenotypes associated with progranulin deficiency, and that whether these abnormalities directly contribute to neurodegeneration in GRN-associated FTD remains a hypothesis requiring further investigation, including studies of whether restoring cathepsin D function or enhancing lysosomal proteostasis can mitigate disease-relevant phenotypes.

Subject of Research: How progranulin deficiency disrupts cathepsin D activity, lysosomal homeostasis, and TDP-43 solubility in mouse models of frontotemporal dementia

Article Title: Progranulin haploinsufficiency disrupts cathepsin D function, resulting in lysosomal abnormalities and elevated TDP-43 levels in the mouse brain

Article References: Fazal, R., Tharkeshwar, A. K., Beel, S., Storm, A., Schevenels, B., Tomé, S. O., Thal, D. R., Van Den Bosch, L., Beckers, J., & Van Damme, P. (2026). Progranulin haploinsufficiency disrupts cathepsin D function, resulting in lysosomal abnormalities and elevated TDP-43 levels in the mouse brain. Frontiers in Molecular Neuroscience, 19, Article 1970154. https://doi.org/10.3389/fnmol.2026.1970154

Image Credits: AI Generated

DOI: 10.3389/fnmol.2026.1970154

Keywords: progranulin, cathepsin D, frontotemporal dementia, lysosomes, TDP-43, GRN mutations, mouse models, protein homeostasis, LAMP1, neurodegeneration, haploinsufficiency, KU Leuven