AEP is a cysteine protease of the C13 peptidase family that becomes active under acidic conditions and is distributed widely in tissues, with particularly high expression in the kidneys and testes. In healthy cells it resides mainly in the endo-lysosomal system, where it cleaves its substrates after asparagine residues. Under pathophysiological conditions, however, the enzyme can also appear in the cytoplasm and the nucleus, where its proteolytic activity has been linked to a striking range of age-associated processes.

Previous work from the same group showed that AEP cleaves the Tau protein at position N368, generating neurotoxic fragments that aggregate and contribute to the neurofibrillary tangles characteristic of Alzheimer's disease. The enzyme also processes alpha-synuclein into aggregation-prone truncated forms implicated in Parkinson's disease, and it cuts SET, an inhibitor of protein phosphatase 2A, thereby impairing DNA base excision repair and promoting the accumulation of DNA damage in neurons. Because aberrant AEP activation has been tied to neurodegeneration, cancer, atherosclerosis, and chronic inflammation, the team asked whether pharmacological inhibition could ameliorate aging itself.

Mice received a standard chow diet or feed supplemented with #11a at 0.056 grams per kilogram, equivalent to roughly 7.5 milligrams per kilogram per day, starting at approximately 18 months of age. Lifespan and health span analyses ran concurrently. The results were striking: mean survival reached 24 months in the control group and 28 months in the treated group, and the risk of death was significantly higher in controls, with a Mantel-Haenszel hazard ratio of 2.08 (95 percent confidence interval 1.02 to 4.27, p = 0.0454).

Body weight was unaffected, although food intake was measured only at the cage level and was numerically lower in treated cages at several time points, a caveat the authors acknowledge openly. Health span was assessed with a blinded frailty index of 29 phenotypes, scored for severity on a scale of 0, 0.5, or 1, with repeated assessments every four weeks from 18 months onward. Because the same mice were followed over time, the researchers applied a linear mixed-effects model treating each animal as a random effect, a statistical approach they report was far more efficient than a standard ANOVA.

The analysis showed that AEP blockade reduced aging phenotypes and delayed morbidity in both sexes, but the specific benefits differed. In females, #11a significantly decreased the severity of mouse grimace, body condition decline, forelimb grip strength loss, hearing loss, gait disorder, and vestibular disturbance. In males, severity decreased for mouse grimace, cataract, loss of fur color, tail stiffening, menace reflex deficits, rectal prolapse, and diarrhea.

Age-related vision loss and coat condition showed the most notable overall improvement, and histological analysis revealed improved lens structure, a lower incidence of cataracts, and increased cell densities in the inner and outer nuclear layers of the retina. Tolerability data from an intermediate time point were reassuring within their limits. Plasma concentrations of #11a were quantified by liquid chromatography-tandem mass spectrometry, and after five months of treatment, histopathological assessment of the heart, liver, spleen, lungs, and kidneys revealed no overt abnormalities.

Complete blood counts and serum biochemical analyses likewise did not reveal major abnormalities in the parameters assessed, consistent with earlier reports that partial reduction of AEP activity can be tolerated without overt phenotypes. The authors stress, however, that these datasets do not establish long-term safety beyond the current study duration. Functional testing at 23 months of age supported the frailty findings.

After five months of treatment, mice on #11a showed significantly enhanced falling time and elevated falling speed on exhaustive treadmill tests, along with substantially longer hanging times in the wire hang assay. In the open-field test, treated animals moved greater total distances, explored the center more, and spent less time immobile, though the authors caution that the data do not fully distinguish anxiety-related behavior from general locomotor changes. Cognitive performance also improved: the novel object recognition discrimination index rose from 32 percent plus or minus 15.69 percent in controls to 48 percent plus or minus 11.03 percent in treated mice (p = 0.0071), and Y maze exploration times increased from 68 seconds to 102 seconds (p = 0.0005).

Micro-computed tomography of the femur showed significantly increased connectivity density, although most other trabecular and cortical bone parameters were unchanged; histology revealed reduced white adipocytes and increased trabecular bone. Immunoblotting confirmed that #11a blocked AEP fragmentation of amyloid precursor protein at N585 and Tau at N368, consistent with the observed cognitive gains. The mechanistic centerpiece of the study involves NAD+, the essential enzymatic cofactor whose decline accompanies aging.

Nicotinamide phosphoribosyltransferase, or NAMPT, is the rate-limiting enzyme for NAD+ biosynthesis from nicotinamide, and the team previously showed that AEP cleaves NAMPT at position N136, depleting NAD+ and triggering neuronal senescence and cognitive deficits. In the aged mice, #11a pronouncedly reduced active AEP levels in the hippocampus, and NAMPT fragmentation—validated with an antibody specific for the truncated C137 form—was correspondingly reduced. Levels of SIRT1, an NAD+-dependent deacetylase central to longevity regulation, and of PGC-1α, a master regulator of mitochondrial function downstream of SIRT1, were elevated in both hippocampus and quadriceps muscle.

The NAD+/NADH ratio increased significantly in brain, muscle, liver, and kidney tissues. Notably, #11a did not significantly alter CD38 protein abundance, CD38 NADase activity, or global poly(ADP-ribose) levels, so the findings do not establish direct target engagement of these alternative NAD+-consuming pathways. Metabolic readouts shifted in parallel.

Both the glucose tolerance test and the insulin tolerance test showed that #11a treatment significantly increased insulin sensitivity and reduced blood glucose concentrations. Plasma levels of alanine aminotransferase, aspartate aminotransferase, low-density lipoprotein cholesterol, and lactate dehydrogenase were significantly decreased. Histological and Oil Red O analyses revealed substantially reduced adipocyte accumulation in brown adipose tissue and inguinal white adipose tissue, while tibialis anterior muscle fiber diameters were significantly augmented.

Quantitative PCR showed elevated expression of metabolic genes including Adipoq, Glut4, and Pparg in brown fat, and Pgc1a, Ucp1, and Dio2 in white fat. By contrast, fed-state insulin, glucose, glycated serum protein, and the plasma lipid profile remained unchanged, and circulating reproductive hormones showed no statistically significant differences, findings the authors interpret as exploratory. Markers of cellular senescence and inflammation also receded.

Senescence-associated beta-galactosidase staining suggested reduced senescence in liver, lung, and white adipose tissue, and Masson's trichrome staining suggested reduced fibrosis-associated changes in liver and kidney. Interleukin-1 beta, interleukin-6, and tumor necrosis factor alpha were substantially decreased in brain, muscle, liver, lung, and kidney. In cell culture, #11a suppressed hydrogen peroxide-induced senescence-associated staining in HT22 neuronal cells and D-galactose-induced senescence-like changes in C2C12 myoblasts, reducing AEP activation, NAMPT cleavage, and p53, p21, and p16 elevation—though the authors note the HT22 data reflect an early stress response rather than a fully established senescence program.

The authors frame their results within a broader model in which the transcription factor C/EBPβ, a principal regulator of aging and the senescence-associated secretory phenotype, drives AEP expression, which in turn cleaves NAMPT and depletes NAD+. Their recent work showed that genetic ablation of AEP or expression of an AEP-resistant NAMPT N136A mutant extends lifespan in Thy1-C/EBPβ transgenic mice, and that #11a outperformed nicotinamide mononucleotide in that accelerated-aging model. Yet the new study is candid about its limits: food intake was measured only at the cage level and a pair-feeding study will be needed to separate direct drug effects from caloric effects; the cohort was not powered to detect sex-specific effects for most endpoints; and the mechanistic evidence in naturally aged mice remains largely correlative, not proof that the AEP-NAMPT-NAD+/SIRT1 pathway is necessary for the survival benefits.

The authors also highlight that AEP's physiological roles are tissue- and disease-context dependent—tubule-specific AEP knockout mice develop spontaneous renal fibrosis, and AEP inhibition accelerated diabetic nephropathy in a streptozotocin model—underscoring the need for careful chronic tolerability assessment. Intriguingly, females lived shorter lives than males in both groups, contrary to the usual pattern in C57BL/6 colonies, possibly reflecting pathogen burden, housing conditions, or the reproductive history of the cohort acquired at 18 months. For now, the message is measured: in these mice, under these conditions, silencing an overactive protein-cutting enzyme was associated with more months of life and visibly better ones.

Subject of Research: Pharmacological inhibition of asparagine endopeptidase (AEP/legumain) during natural aging in mice

Article Title: Inhibition of asparagine endopeptidase extends the lifespan and improves health span during aging

Article References: Yang, M., Meng, X., Xie, Z., Wang, M., Qian, Z., & Ye, K. (2026). Inhibition of asparagine endopeptidase extends the lifespan and improves health span during aging. iScience, 29(11), Article 117990. https://doi.org/10.1016/j.isci.2026.117990

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117990

Keywords: aging, longevity, AEP, legumain, NAD+, NAMPT, SIRT1, cellular senescence, frailty, neurodegeneration, mouse study, drug research