Hypothesis

Age‑related depletion of geranylgeranyl pyrophosphate (GGPP) disrupts the prenylation of Rho family GTPases, leading to aberrant actin phosphorylation and uncoupling of ZO‑1 from perijunctional actomyosin. This cytoskeletal lesion preferentially increases permeability to macromolecules and bacterial products while leaving claudin‑based ion pores relatively intact. Concurrently, loss of GGPP‑dependent Rac1/Cdc42 activity impairs dynamic tight‑junction remodeling, causing a compensatory upregulation of claudin‑2–mediated cation leaks that alter luminal osmolarity, favor opportunistic bacterial overgrowth, and stimulate zonulin release. The resulting bacterial DNA translocation fuels systemic inflammation (inflammaging). Restoring GGPP levels via geranylgeraniol supplementation will re‑prenylate Rho GTPases, restore ZO‑1–actomyosin coupling, normalize zonulin, reduce bacterial translocation, and attenuate inflammatory markers.

Mechanistic Rationale

1. GGPP and GTPase prenylation – GGPP is the lipid donor for geranylgeranylation of RhoA, Rac1, and Cdc42. Prior work shows age‑related decline in GGPP synthesis via the mevalonate pathway [5]. Loss of geranylgeranylation locks RhoA in a GDP‑bound state, promoting ROCK‑mediated LIMK/cofilin phosphorylation and actin hyperstabilization, which we propose uncouples ZO‑1 from the contractile actin ring [4].
2. ZO‑1‑selective barrier – ZO‑1 specifically restricts flux of solutes >4Å (e.g., bacterial lipopolysaccharide, peptidoglycan) without affecting small ion channels [4]. Thus, preserved claudin-2 expression can coexist with increased macromolecular leak.
3. Claudin‑2 compensatory leak – Impaired Rac1/Cdc42 signaling reduces junctional endocytosis‑recycling cycles, leading to accumulation of claudin-2 at the membrane and increased Na+ permeability. This alters luminal ionic strength, creating a niche for facultative anaerobes (e.g., Lactobacillus spp.) that trigger zonulin release [6].
4. Link to inflammaging – Bacterial DNA in plasma correlates with zonulin and systemic TNF‑α/IL‑6 [3], establishing a causal chain from cytoskeletal uncoupling to microbial translocation and inflammation.

Testable Predictions

• Older adults with low serum GGPP (or low geranylgeraniol levels) will exhibit higher zonulin, increased lactulose/mannitol ratio (macro‑molecule leak), and elevated plasma bacterial DNA compared with age‑matched controls.
• Geranylgeraniol supplementation (200 mg/day for 12 weeks) will raise intracellular GGPP, restore ZO‑1–actin co‑immunoprecipitation, decrease zonulin and lactulose/mannitol ratio, reduce plasma bacterial 16S rRNA copies, and lower circulating TNF‑α and IL‑6.
• In vitro Caco‑2 monolayers treated with mevastatin (to deplete GGPP) will show ZO‑1 actin uncoupling and increased FITC‑dextran (4 kDa) flux, which is rescued by geranylgeraniol but not by exogenous ZO‑1 overexpression.
• Claudin-2 knock‑down in the same model will prevent the zonulin‑like epithelial response to GGPP depletion, indicating that the cation leak is upstream of zonulin release.

Experimental Design

Human trial: Double‑blind, placebo‑controlled RCT in 120 participants aged ≥70 years. Stratify by baseline zonulin (high vs low). Intervention: geranylgeraniol 200 mg daily; control: matched placebo. Primary outcomes at 0, 6, and 12 weeks: serum zonulin (ELISA), lactulose/mannitol ratio, plasma bacterial DNA (qPCR for 16S rRNA), TNF‑α, IL‑6. Secondary outcomes: fecal microbiota composition (16S sequencing), GGPP levels in PBMCs (LC‑MS/MS).

Mechanistic sub‑study: Endoscopic biopsies from a subset (n = 30) for immunofluorescence of ZO‑1, actin, phosphorylated myosin light chain, and geranylgeranylated RhoA (using click‑chemistry probes).

Cellular model: Caco‑2 cultures treated with 1 µM mevastatin ± 10 µM geranylgeraniol for 48 h. Assess ZO‑1–actin interaction (co‑IP), TER, FITC‑dextran (4 kDa) flux, claudin-2 surface expression (biotinylation), and zonulin‑like peptide release (ELISA).

Potential Confounds and Mitigation

• Zonulin assay specificity: Use a zonulin ELISA validated against haptoglobin-2 precursor and confirm with parallel measurement of intestinal fatty acid binding protein (FABP‑2) as a complementary permeability marker.
• GGPP variability: Measure downstream metabolites (geranylgeraniol, farnesol) to ensure supplementation reaches target tissue.
• Microbiota influence: Run baseline antibiotic washout (if ethically permissible) or include pretreatment microbiome as covariate.

Significance

If validated, this hypothesis shifts the therapeutic focus from attempting to upregulate already‑normal tight‑junction genes to correcting the lipid‑mediated cytoskeletal lesion that underlies age‑related barrier leak. It offers a nutritionally tractable intervention (geranylgeraniol or statin‑sparing mevalonate pathway modulators) with clear biomarkers for efficacy, directly addressing the mechanistic link between intestinal permeability, bacterial translocation, and inflammaging.