Increasing age is the primary risk factor for Alzheimer’s disease (AD), even in individuals with high genetic risk. The mild stressor caloric restriction (CR)—or consuming ~2/3 normal daily calories—postpones and prevents diseases of aging in animal models and perhaps also in man. In contrast, diabetes mellitus and caloric excess (obesity, particularly during midlife) accelerate the onset of AD, suggesting a link between glucose/energy metabolism and amyloid precursor protein/β-amyloid (Aβ) metabolism. While the mechanism of CR benefits remains unclear, activation of sirtuins, notably SIRT1, may be a critical molecular pathway. SIRT1 deacetylase activity is regulated by NAD+/NADH—coupling cellular energy balance to epigenetic transcriptional regulation. Resveratrol, a potent SIRT1 activator and pharmacologic mimic of CR, is a polyphenol found naturally in red grapes, peanuts, and many other plant species. Similar to CR, treatment of transgenic mouse models of AD with resveratrol decreases behavioral deficits and central nervous system (CNS) Aβ deposition with aging [1].

We hypothesized that molecular mechanisms of aging, specifically SIRT1, may be exploited as a target for development of AD therapeutics. Given the proven safety of resveratrol and promising preclinical data, we enrolled 119 subjects in a phase 2 randomized, double-blind, placebo-controlled trial of resveratrol in subjects with mild-moderate AD (with dosage stepped up to 2 g pure, synthetic resveratrol by mouth daily, for 12 months) [2]. High-dose oral resveratrol treatment is safe and well-tolerated—the only significant adverse effect is weight loss. Low nanomolar native resveratrol is detectable in cerebrospinal fluid (CSF), suggesting CNS penetration and a high-affinity molecular target (or targets). Compared to placebo, resveratrol stabilizes the progressive decline in CSF Aβ40 and plasma Aβ40 levels as dementia advances. In individuals with biomarker-confirmed AD (CSF Aβ42 <600 ng/ml) at baseline, resveratrol also stabilizes CSF Aβ42 levels [2]. Despite the phase 2 trial being underpowered to detect clinical benefits, resveratrol attenuated decline in the Alzheimer’s Disease Cooperative Study-Activity of Daily Living (ADCS-ADL) score during the 12-month study. Aging is also a major risk factor for cancer, and fewer cancers were found in the resveratrol-treated group (one versus seven cancers in six participants in the placebo group). Collectively, these data support the notion that targeting molecular mechanisms of aging may point to therapeutic strategies that postpone or prevent diseases of aging—in parallel. With proven safety and suggestions of efficacy in the phase 2 trial, the putative benefits of resveratrol and other sirtuin activator compounds (STACs) should be further examined in clinical studies.

Paradoxically, resveratrol treatment increased brain volume loss in AD subjects, compared to the placebo-treated group. Since CSF tau and phospho-tau levels are unaffected (suggesting no treatment effect on neuronal loss), we hypothesize that resveratrol has potent anti-inflammatory effects in AD brain—with decreased CNS edema as the etiology of greater brain volume loss. Similar effects are found with anti-amyloid immunotherapies for AD [3] and effective drugs for multiple sclerosis (MS) are also known to be associated with “pseudoatrophy” [4]. To test the putative anti-inflammatory effects of resveratrol in AD brain, we measured pro- and anti-inflammatory cytokines and chemokines, and metalloproteinases, in banked samples of CSF and plasma from a subset of individuals with biomarker-confirmed AD (CSF Aβ42 <600 ng/ml) in the phase 2 trial. Consistent with our hypothesis, we found significant anti-inflammatory effects of resveratrol in the CSF of treated AD subjects. Our data also suggest that resveratrol treatment preserved the integrity of the blood-brain barrier (BBB) in AD. Collectively, these exploratory findings lend support to the notion that targeting molecular pathways of aging may lead to novel therapies to postpone or prevent diseases of aging, including AD.

One of the most striking results of this study is the significant decrease in the level of CSF MMP9 after resveratrol treatment. MMP9 has recently emerged as a major player in several brain pathologies, including neurodegeneration and neuro-inflammation [5]. MMP9 regulates BBB permeability via release of cytokines and free radicals as well as cleavage of vascular basal lamina and/or tight junctions in the neurovascular unit in both MS and AD [6–8]. The decrease in CSF MMP9 levels suggests that resveratrol treatment may reduce CNS permeability and limit the infiltration of leukocytes and other inflammatory agents into the brain. MMP9-mediated breakdown of the basal lamina and destruction of gap junctions in the neurovascular unit result in increased CNS permeability and inflammation in autoimmune encephalitis, hypoxic brain injury, and other diseases [9, 10]. MMP9 knockout reduces neuro-inflammation in experimental autoimmune encephalomyelitis (EAE) [11], while CSF MMP9 is elevated in patients with bacterial meningitis and BBB damage [12]. Moreover, inhibition of MMP9 alleviates the neurological damage associated with human immunodeficiency virus (HIV) infection [13], suggesting that MMP9 activation is a response to HIV infection. These data are also supported by enhanced expression and activity of MMP9 in serum, CSF, and demyelinating lesions in MS [14], and abundant evidence of increased MMP9 expression and activity in ischemic stroke [15, 16]. Animal studies have also revealed significant increases in MMP9 levels after traumatic brain injury [17], but damage to the BBB and behavioral deficits are significantly attenuated in MMP9 knockout animals [18, 19].

MMP9 is highly regulated both spatially and temporally with many target substrates including growth factors, cell surface receptors, and cell adhesion molecules. Low levels of MMP9 messenger RNA (mRNA) and protein expression are detected predominantly in neurons in the hippocampus, cerebellum and cerebral cortex of normal brain [20], but injury significantly increases the mRNA and protein levels and activity of MMP9 [5, 21, 22], which may be derived from brain cells or leukocyte invasion of the brain due to BBB compromise. Intercellular adhesion molecule-5 (ICAM-5), which mediates the regulation of dendritic spine elongation and maturation may be cleaved by MMP9 upon activation of N-methyl-d-aspartate (NMDA) receptors [23, 24], suggesting a role for MMP9 in synaptic function. Furthermore, MMP9 deletion increases the number of CA1 pyramidal neurons and decreases the length and complexity of dendritic spines [25]. Immune system dysfunction may develop with aging in parallel with upregulation of brain MMP9 [26–28]. However, a recent study showed that CSF MMP9 was significantly lower in AD subjects with decreased Aβ42 and Aβ40 and increased total tau and p-tau levels compared to healthy controls [29]. In the current study, the levels of CSF tau and p-tau were not altered by treatment but the levels of CSF Aβ42 and Aβ40 were altered in parallel with a reduction of MMP9. However, there was no difference in the level of CSF MMP9 between placebo and resveratrol-treated groups at baseline, and it is uncertain whether MMP9 in our study population with AD is different from healthy controls. MMP9 activation is likely driven by other MMPs [30], so we examined the level of MMPs in plasma and CSF. Leukocyte penetration into brain parenchyma in EAE models involves β-dystoglycan cleavage that is only abolished in double MMP2 and MMP9 knockout mice [31], suggesting the effects of other MMPs on MMP9 function. MMP10 and MMP3 were slightly increased in the plasma but not CSF of AD patients. MMP9 has overlapping substrates with other MMPs that share similar structures [5], so caution must be used in the interpretation of specific MMP9 targets. MMP9 also plays a role in post-natal brain development during a critical period of synaptic formation and maturation and axonal myelination [32]. In the adult brain MMP9 and MMP3 may be involved in neurogenesis and migratory response mechanisms [33]. MMP9 is upregulated in delayed and acute phases of post ischemic stroke models [34, 35].

MMP9 regulates the CNS immune response due to its ability to activate inflammatory markers and its involvement in BBB maintenance, leading to its regulation of entry of leukocytes into the brain parenchyma [5]. MDC/CCL22 is a small cytokine that belongs to the Cysteine-Cysteine (CC) family and is involved in transport of natural killer cells, chronically activated T lymphocytes (Th2) [36], monocytes, and dendritic cells into injury sites [37]. MDC is expressed in the CNS and is produced by CNS-infiltrating leukocytes and intra-parenchymal microglia in EAE models [38]. Activated microglia secrete MDC that induces chemotaxis of Th2, but not Th1, cells suggesting that MDC produced by microglia regulates neuro-inflammation via recruitment of Th2 cells into the injury site [38]. Leukocyte infiltration into CNS white matter lesions, which contain CD4⁺ and CD8⁺ T cells and activated macrophages/microglia, is a hallmark of MS [39]. Taken together, these findings support the hypothesis that the increase in CNS MDC with resveratrol may facilitate the intracerebral homing of specific leukocytes involved in brain injury in AD, providing a mechanism for responding to amyloid-associated inflammation [40]. MDC is involved in Th2-driven chronic inflammation [41], and this is consistent with the increase of CNS levels of IL-4, which mediates an adaptive immune response via Th2 cell induction [42, 43], leading to a long-term protective immune response. Our results are also consistent with the function MMPs that play an integral role in immune cell development, effector function, migration, and ligand-receptor interactions [44]. T helper cells (Th1 and Th2) secrete MMP9 [45], which plays a critical role in the migration of T cells from the blood stream to the brain and other tissues [46, 47]. Additionally, recent advances in neuro-inflammation implicate abnormal neurotrophic factor signaling, including fibroblast growth factors (FGFs) in HIV-associated neurocognitive decline (HAND) [48] and stroke [49]. The increase in CNS FGF levels after resveratrol treatment suggests an effect on growth factors, which may play a role in neuro-resilience in aging and AD.

Neuro-inflammation may contribute to cognitive impairment and play a significant role in AD progression. Activation of specific microglia/macrophage may be neuroprotective. Although resveratrol treatment did not affect CSF tau, resveratrol significantly attenuated the declines in CSF Aβ42 and Aβ40 levels (compared to placebo) and attenuated cognitive and functional decline (MMSE and ADCS-ADL) between the placebo and treated groups. Resveratrol also reduced the plasma levels of pro-inflammatory makers including IL-1R4, IL-12P40, IL-12P70, TNF-α, and RANTES, independent of CSF changes of the levels of these biomarkers.

Innate immune cells, including CNS resident microglia and peripheral bone marrow-derived macrophages can exhibit a dysfunctional or senescent profile characterized by impaired phagocytosis as AD progresses, indicating that modification of the microglia/macrophage activation state, instead of inhibiting their function, may hold therapeutic promise in AD [50–52]. Resveratrol may facilitate activation of microglia/macrophages therefore inducing a long-term adaptive immune response that may be clinically beneficial in AD subjects. Major impediments of current immunotherapy approaches to AD include limited evidence of significant clinical benefits, and the risk of excessive neuro-inflammation [53].

Resveratrol may maintain the integrity of the BBB via reduction of MMP9 and induce adaptive immune responses that may promote brain resilience to amyloid deposition. Resveratrol may slow cognitive decline in AD via a coordinated peripheral and central immune response that may also arrest neuronal death. In conclusion, the exploratory findings of the current study encourage further validation of the hypothesis that resveratrol may seal off a leaky BBB and contribute to cognitive and functional improvement in a larger follow-up study with AD patients.