The investigation began with a human clue. The team analyzed 257 community-dwelling Han Chinese adults aged 50 and older who underwent comprehensive cognitive assessment using the Montreal Cognitive Assessment, a screening tool considered more sensitive than the Mini-Mental State Examination for detecting early deficits. Among a panel of energy metabolism indicators, including blood glucose, creatinine, Klotho, and the albumin/globulin ratio, ghrelin showed the strongest correlation with cognitive scores. Higher circulating ghrelin levels were associated with poorer cognitive performance, and the relationship held after accounting for age, sex, and education. Notably, the association was sex-dependent: women with mild cognitive impairment had significantly higher ghrelin levels than cognitively normal women and than men at the same stage. The authors interpret this elevation not as evidence that ghrelin damages cognition, but as a possible compensatory response to metabolic stress, a biomarker of a brain under siege rather than a culprit.

To test what ghrelin actually does, the researchers turned to two mouse models. The first compared young two-month-old animals with nine-month-old mice, capturing an early stage of natural aging. The second used chronic injection of D-galactose, a sugar that accelerates senescence across tissues and is a widely used chemical model of aging. In both models, behavioral testing revealed clear deficits. In the novel object location and novel object recognition tests, aged and D-galactose-treated mice spent less time exploring objects in new positions or objects they had never seen, indicating impaired spatial and recognition memory. In the Morris water maze, they took longer to find a hidden platform, a classic sign of compromised hippocampal learning.

Beneath these behavioral changes lay a striking cellular signature. Western blotting of hippocampal tissue showed upregulation of the senescence markers p16 and p21 and the DNA damage marker gamma-H2AX, alongside loss of Lamin B1, a nuclear envelope protein that declines as cells age, and reduced HMGB1, a chromatin-associated protein. Immunofluorescence co-localization pinpointed these changes within Iba1-positive microglia in the hippocampal CA1 and CA3 regions, confirming that the brain’s innate immune cells were themselves senescing. At the same time, enzyme-linked immunosorbent assays revealed a flood of inflammatory mediators in the hippocampus, including IL-1alpha, IL-1beta, IL-6, IL-8, CXCL10, MCP1, and M-CSF, an inflammaging profile consistent with the senescence-associated secretory phenotype that aged microglia are known to adopt. Both total and acylated ghrelin were elevated in the plasma of these animals, mirroring the human data.

The therapeutic experiment followed. D-galactose-treated mice received daily ghrelin injections at either 20 or 80 micrograms per kilogram during the final four weeks of aging induction, and the same regimen was applied to naturally aged mice. The results were consistent across both models: ghrelin-treated animals performed significantly better in all three behavioral tests, with the higher dose producing the greatest benefit. Molecular analysis showed that ghrelin suppressed p16, p21, and gamma-H2AX while restoring Lamin B1 and HMGB1, effectively reversing the senescence signature. Immunofluorescence confirmed fewer senescent microglia, and the inflammatory cytokine storm in the hippocampus subsided. Interestingly, the protection extended beyond immune cells, as gamma-H2AX signals in neurons also declined after treatment, though astrocytes showed minimal senescence marker co-localization throughout.

What made this outcome puzzling was the receptor. Ghrelin’s classical receptor, GHSR, was expected to mediate any benefit, yet high-dose ghrelin actually reduced GHSR membrane localization, a sign of receptor desensitization, and only the low dose boosted the p-CaMKII/CaMKII ratio that reflects GHSR-driven synaptic plasticity. Even more tellingly, when the researchers knocked down GHSR in D-galactose-treated primary microglia, ghrelin still suppressed p16, p21, and gamma-H2AX. The anti-senescent effect, in other words, was largely GHSR-independent, pointing to some other molecular intermediary.

To find it, the team deployed network pharmacology, cross-referencing predicted ghrelin targets with genes linked to cognitive impairment and clustering the resulting protein-protein interaction network. Five hub genes emerged: SHBG, SRC, C5AR1, KLF4, and BRCA1. Of these, sex hormone-binding globulin, or SHBG, responded most consistently to ghrelin in vivo. SHBG is classically known as a blood protein that binds sex hormones, but elevated levels have previously been linked to hippocampal atrophy and cognitive decline in clinical studies. Immunofluorescence showed SHBG localized in the cytoplasm of BV2 microglial cells, and molecular docking predicted a physical interaction between ghrelin and SHBG, which co-immunoprecipitation in primary microglia then validated. D-galactose exposure raised hippocampal SHBG, and ghrelin lowered it in a dose-dependent fashion.

The downstream circuit came into focus through transcriptomics and targeted manipulation. RNA sequencing of ghrelin-treated microglia revealed 373 upregulated and 431 downregulated genes, with gene set enrichment analysis positively scoring autophagy pathways. Ghrelin reduced the autophagy substrate p62 while increasing Beclin1 and the LC3-II/LC3-I ratio, hallmarks of enhanced autophagic flux, the cellular recycling process that keeps senescent cells from accumulating damaged components. When the researchers overexpressed SHBG in microglia, the opposite occurred: p62 rose, Beclin1 and LC3 fell, and senescence markers worsened. SHBG knockdown increased JNK1 phosphorylation while SHBG overexpression reduced it, placing JNK1 downstream of SHBG. Blocking JNK1 with a pharmacological inhibitor abolished ghrelin’s ability to raise Beclin1 and LC3, and critically, combining JNK1 inhibition with SHBG overexpression produced no additional suppression, confirming that the two act in a single linear pathway. Finally, forcing Beclin1 expression partially rescued the senescent phenotype caused by SHBG overexpression, completing the picture of a SHBG/JNK1/Beclin1 axis that ghrelin engages to restore autophagy and push microglia back toward a youthful, homeostatic state.

The authors are careful about the limits of their work. The D-galactose and nine-month-old mouse models cannot fully recapitulate human cognitive impairment, the animal experiments used only male mice even though the human data revealed sex-dependent ghrelin patterns, and the long-term safety of ghrelin administration remains untested in larger animals or clinical trials. Some autophagy-related genes, including ERK2, PRKAA1, and CTSD, were untouched by the hormone, suggesting the effect is pathway-specific rather than a blanket activation of the autophagy network. Even so, the study delivers a compelling conceptual advance: a gut-derived metabolic hormone can act on brain immune cells through a non-canonical, receptor-independent route, damping the senescence-associated secretory phenotype that drives neuroinflammation. If the SHBG/JNK1/Beclin1 axis behaves similarly in human microglia, it could open a new class of therapeutic targets for age-related cognitive decline, a condition for which disease-modifying treatments remain desperately lacking despite decades of effort focused almost exclusively on neurons and amyloid.

Subject of Research: Ghrelin regulation of microglial senescence and autophagy in aging-related cognitive impairment

Article Title: Ghrelin Alleviates Aging‐Related Cognitive Impairment by Regulating Autophagy‐Related Signaling via the SHBG/JNK1/Beclin1 Axis in Microglia

Article References: Ghrelin Alleviates Aging‐Related Cognitive Impairment by Regulating Autophagy‐Related Signaling via the SHBG/JNK1/Beclin1 Axis in Microglia. (n.d.). https://doi.org/10.1111/acel.70732

Image Credits: AI Generated

DOI: 10.1111/acel.70732

Keywords: ghrelin, microglia, cognitive impairment, autophagy, SHBG, JNK1, Beclin1, immunosenescence, neuroinflammation, brain aging, D-galactose, Aging Cell