The aging brain does not simply accumulate random damage; it actively removes metabolically expensive, weakly connected neurons through a microglia-mediated phagocytic program that resembles developmental pruning. This eviction is restrained by Klotho, which suppresses astrocytic expression of complement component C1q, a key “eat‑me” tag for synapses and somata. When Klotho declines with age, C1q rises, tagging low‑activity neurons for clearance by microglia via CR3 receptors. Consequently, interventions that boost Klotho in astrocytes should reduce C1q deposition, preserve weakly active neurons, and improve network efficiency without globally increasing excitation.

Testable prediction: In aged mice, astrocyte‑specific overexpression of Klotho (using AAV8, which transduces hippocampal astrocytes efficiently)[https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0076310] will lower cortical and hippocampal C1q levels, decrease microglial phagocytic cups (Iba1+/CD68+ engulfing NeuN+ material), and retain neurons that show low baseline c‑Fos activity but are otherwise healthy. Conversely, astrocyte‑specific Klotho knockdown will accelerate C1q‑dependent neuronal loss, especially in females where endogenous Klotho is lower and adverse effects have been observed with global Klotho overexpression[https://lifespan.io/news/a-klotho-gene-therapy-extends-life-in-male-mice/].

Experimental design:

• Use 20‑month-old male and female C57BL/6J mice. Groups receive intra‑hippocampal AAV8‑Klotho, AAV8‑shKlotho, or AAV8‑GFP control (n=10 per sex per group).
• Four weeks post‑injection, quantify: (1) astrocytic Klotho immunoreactivity, (2) C1q puncta co‑localized with NeuN, (3) microglial phagocytic index (fraction of Iba1+ cells containing NeuN+ material), (4) c‑Fos+ neuronal density as a proxy for activity, and (5) total NeuN+ cell counts in CA1 and layer II/III cortex.
• Perform contralateral unilateral injection of a function‑blocking anti‑C1q antibody in a subset to test sufficiency of complement.
• Assess behavior (Y‑maze spontaneous alternation, novel object recognition) to link cellular changes to cognition.

Falsifiable outcomes: If neuronal eviction is purely a passive consequence of oxidative or proteostatic damage, then manipulating astrocytic Klotho or complement will not alter the rate of NeuN+ loss, phagocytic index, or the survival of low‑c‑Fos neurons. A lack of sex‑specific differences in these manipulations would also contradict the hypothesis that Klotho’s protective effect is modulated by baseline sex‑dependent expression.

Mechanistic insight: Klotho’s known inhibition of IGF‑1/IIS and Wnt signaling may converge on the transcriptional repression of C1q in astrocytes, linking metabolic sensing to immune tagging. This positions the complement‑microglia axis as an effector of a hypothesized "energy‑budget" checkpoint: neurons that fail to meet activity‑dependent metabolic thresholds (e.g., low ATP/NAD+ ratios) become vulnerable to C1q tagging when Klotho‑mediated suppression wanes. Thus, aging‑related neuronal loss reflects an adaptive, albeit potentially maladaptive, re‑allocation of limited resources rather than indiscriminate decay.

By directly linking astrocyte‑derived Klotho, complement tagging, and microglial phagocytosis to the selective removal of inefficient neurons, this hypothesis provides a clear, falsifiable framework to distinguish active eviction from passive damage in the aging brain.