Ketone body metabolism and the NLRP3 inflammasome in Alzheimer's disease

Summary

Alzheimer's disease (AD) is a degenerative brain disorder and the most common form of dementia. AD pathology is characterized by senile plaques and neurofibrillary tangles (NFTs) composed of amyloid‐β (Aβ) and hyperphosphorylated tau, respectively. Neuroinflammation has been shown to drive Aβ and tau pathology, with evidence suggesting the nod‐like receptor family pyrin domain containing 3 (NLRP3) inflammasome as a key pathway in AD pathogenesis. NLRP3 inflammasome activation in microglia, the primary immune effector cells of the brain, results in caspase‐1 activation and secretion of IL‐1β and IL‐18. Recent studies have demonstrated a dramatic interplay between the metabolic state and effector functions of immune cells. Microglial metabolism in AD is of particular interest, as ketone bodies (acetone, acetoacetate (AcAc), and β‐hydroxybutyrate (BHB)) serve as an alternative energy source when glucose utilization is compromised in the brain of patients with AD. Furthermore, reduced cerebral glucose metabolism concomitant with increased BHB levels has been demonstrated to inhibit NLRP3 inflammasome activation. Here, we review the role of the NLRP3 inflammasome and microglial ketone body metabolism in AD pathogenesis. We also highlight NLRP3 inflammasome inhibition by several ketone body therapies as a promising new treatment strategy for AD.

1. INTRODUCTION

Alzheimer's disease (AD) is a progressive brain disorder marked by behavioral and cognitive impairment. AD is the most common cause of dementia, accounting for 60%–80% of cases. ¹ In the United States, around one in nine people age 65 and older has AD. ¹ , ² By 2060, a projected 13.8 million people age 65 and older in the United States will have AD ¹ , ² highlighting the need for continued exploration of innovative therapies to treat the symptoms and etiology of AD.

AD pathogenesis is a complex process, and pathological identification of AD remains the benchmark for diagnosis. Macroscopic features of a brain impacted by AD typically include medial temporal atrophy affecting the hippocampus and amygdala ³ along with dilation of the temporal horns of the lateral ventricles. ⁴ The microscopic features of an AD brain were first described by Alois Alzheimer more than 100 years ago. ⁵ Histopathological hallmarks of AD include extracellular aggregates of amyloid‐β (Aβ) and intracellular aggregations of neurofibrillary tangles (NFTs) composed of hyperphosphorylated tau. ⁶ , ⁷ , ⁸ , ⁹ A growing body of evidence suggests a synergistic relationship between Aβ and tau pathogenesis drives AD pathology progression as a result of the strong relationship between tau pathology and Aβ plaque burden. ¹⁰ , ¹¹ , ¹² Together, AD pathogenesis results in neuron damage/death, synaptic dysfunction, and behavior/memory impairment. ¹³ , ¹⁴

Neuroinflammation is a central component of AD pathogenesis and progression. ¹⁵ Several inflammatory pathways have been shown to drive AD pathogenesis, including nuclear factor kappa‐light‐chain‐enhancer of activated B cells (NF‐κB), type 1 interferons, and the nod‐like receptor family pyrin domain containing 3 (NLRP3) inflammasome. ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ Microglia, the primary immune effector cells of the brain, display a multifaceted role in AD pathogenesis. Microglial morphology is altered during the different stages of AD pathogenesis, implying microglia can rapidly adapt to changing environments. ²⁰ In regards to morphology, microglia are classified as ramified (resting), activated, and amoeboid (phagocytic). ²¹ In the early stages of AD development, activated microglia delay disease progression by Aβ clearance before plaque formation. ²² , ²³ Furthermore, clusters of microglia can act as a protective barrier to reduce neurotoxicity in plaque‐adjacent neurons. ²⁴ , ²⁵ , ²⁶ In contrast, microglia are a prominent contributor to neuroinflammation by the release of reactive oxygen species (ROS), pro‐inflammatory cytokines/chemokines, and other pro‐inflammatory mediators involved in AD pathogenesis. ²⁷ , ²⁸ Additionally, microglia have been implicated in AD progression, as phagocytized Aβ and tau are dispersed by microglia to unaffected regions of the brain, facilitating the dissemination of plaques. ²⁹ , ³⁰ , ³¹

Increasing evidence suggests AD pathogenesis is driven by metabolic dysfunction in the brain. ³² For example, metabolic syndromes such as diabetes, obesity, hypertension, and hyperlipidemia have been identified as major risk factors for AD. ³³ Recent studies have demonstrated a dramatic interplay between the metabolic state and effector functions of microglia in AD pathogenesis and progression. ³⁴ , ³⁵ , ³⁶ , ³⁷ A large‐scale proteomics analysis of AD cerebrospinal fluid (CSF) and brain samples demonstrated early metabolic changes associated with microglial activation. ³⁸ In this review, we discuss the role of the NLRP3 inflammasome in AD pathogenesis. We will also discuss ketone body metabolism by microglia and ketone body inactivation of the NLRP3 inflammasome to attenuate AD pathology. Finally, we will highlight several ketone body therapies that have the potential for use in the treatment of AD.

2. NLRP3 INFLAMMASOME

NLRP3 is a cytosolic pattern recognition receptor (PRR) protein expressed by a variety of cells including macrophages and microglia. ³⁹ , ⁴⁰ The NLRP3 protein consists of a C‐terminal leucine‐rich repeat (LRR) domain, a central NACHT domain that mediates adenosine triphosphate (ATP) hydrolysis, and an N‐terminal pyrin (PYD) domain which recruits proteins for inflammasome complex formation. ³⁹ , ⁴¹ , ⁴² Overall, the NLRP3 inflammasome consists of NLRP3 (sensor), apoptosis‐associated speck‐like protein containing a caspase recruitment domain (ASC) (adaptor), and caspase‐1 (effector). ³⁹ NLRP3 inflammasome formation occurs in two distinct stages, priming and activation. ³⁹ , ⁴³ Priming (signal 1) is initiated by pattern recognition receptors (PRRs) or tumor necrosis factor receptor (TNFR) activation, resulting in NF‐κB activation and increased expression of NLRP3, pro‐IL‐18, and pro‐IL‐1β. ⁴⁴ , ⁴⁵ , ⁴⁶ Activation (signal 2) is initiated by danger‐associated molecular patterns (DAMPs), protein aggregates, pathogens, and pore‐forming toxins, promoting NLRP3 inflammasome assembly and caspase‐1‐mediated release of IL‐18 and IL‐1β. ⁴⁴ , ⁴⁷ Numerous upstream cellular and molecular signaling pathways are induced by NLRP3 stimuli, including ion flux (calcium [Ca²⁺], sodium [Na⁺], chloride [Cl⁻], potassium [K⁺]), trans‐Golgi disassembly, mitochondrial dysfunction, metabolic alterations, and lysosomal disruption. ⁴⁸ , ⁴⁹ , ⁵⁰ , ⁵¹ , ⁵² , ⁵³ , ⁵⁴ , ⁵⁵ , ⁵⁶ , ⁵⁷ , ⁵⁸ , ⁵⁹ , ⁶⁰ , ⁶¹ NLRP3 inflammasome activation results in a form of lytic programmed cell death, termed pyroptosis. Gasdermin D (GSDMD) is a key regulator of pyroptosis, as activated caspase‐1 cleaves GSDMD into C‐ and N‐terminal domain fragments that subsequently bind to plasma membrane lipids (cardiolipin, phosphatidylserine, phosphatidylinositol). Once bound, the N‐terminal domain of GSDMD inserts itself into the plasma membrane to form pores, activate pyroptosis, and facilitate the release of pro‐inflammatory molecules. ⁶² , ⁶³ , ⁶⁴ , ⁶⁵ , ⁶⁶ , ⁶⁷

The NLRP3 inflammasome is recognized as a crucial element in host defense against fungal, viral, and bacterial pathogens. ⁶⁸ , ⁶⁹ , ⁷⁰ Dysregulation in NLRP3 inflammasome activation, however, is associated with the development of numerous disorders, including cancer, metabolic, neurodegenerative, and autoimmune diseases. ⁷¹ Gain of function mutations in NLRP3 are associated with cryopyrin‐associated periodic syndromes (CAPS) which is a hereditary inflammatory disorder consisting of a group of conditions (Muckle‐Wells syndrome [MWS], neonatal‐onset multisystem inflammatory disease [NOMID], familial cold autoinflammatory syndrome [FCAS]) that have overlapping symptoms and the same genetic cause. ⁷² , ⁷³ In regards to the central nervous system (CNS), the NLRP3 inflammasome has been suggested to play an important role in AD, traumatic brain injury (TBI), multiple sclerosis (MS), and Parkinson's disease (PD). ⁷¹ Neuroinflammation is a major factor in the numerous physiological changes in the brain following TBI, and many studies suggest activation of the NLRP3 inflammasome after a TBI event. ⁷⁴ , ⁷⁵ , ⁷⁶ , ⁷⁷ Furthermore, studies suggest pharmaceutical inhibition of the NLRP3 inflammasome reduces cell death and attenuates the neurological effects of TBI. ⁷⁸ , ⁷⁹ , ⁸⁰ In MS, a chronic disease of the CNS, Nlrp3 polymorphisms are correlated with susceptibility and severity of MS. ⁸¹ , ⁸² For example, Soares et al. ⁸² identified gain‐of‐function polymorphisms in IL1β (−511 C > T) and Nlrp3 (Q705K) correlated with progression and severity of MS, consistent with a previous report suggesting gain‐of‐function mutations in Nlrp3 as a comorbidity with MS. ⁸³ Furthermore, multiple NLRP3 inflammasome‐related proteins (NLRP3, ASC, caspase‐1, IL‐1β, IL‐18) are increased in PD patients, suggesting a strong correlation between NLRP3 inflammasome activation and PD pathology. ⁸⁴ , ⁸⁵ , ⁸⁶ , ⁸⁷ Additionally, several studies using Nlrp3 ^−/− , Asc ^−/− , and Caspase‐1 ^−/− mice demonstrated a critical role for NLRP3 inflammasome activation in experimental autoimmune encephalomyelitis (EAE), an animal model for MS, as all of these knockout mice displayed significantly attenuated EAE compared to their wild‐type littermates. ⁸⁸ , ⁸⁹ , ⁹⁰

Advanced age is the greatest risk factor for AD. ⁹¹ The term “inflamm‐aging” is often used to refer to the systemic, low‐grade level of inflammation produced by the innate immune system during normal aging. ⁹² , ⁹³ The study by Youm et al. determined a critical role of the NLRP3 inflammasome during “inflamm‐aging”. ⁹⁴ In their study, the NLRP3 inflammasome is demonstrated to be critical sensor of an age‐related aggregation of DAMPs. Their study also suggests reducing/eliminating NLRP3 inflammasome activation serves as a protective mechanism to reduce the risk of age‐related diseases, as ablation of the NLRP3 inflammasome protects mice from age‐related transcriptome changes, innate immune activation, neuroinflammation, and functional decline. ⁹⁴ Since neuroinflammation drives AD pathogenesis, and the NLRP3 inflammasome is involved in “inflamm‐aging,” there is great interest in investigating the role of the NLRP3 inflammasome in AD. Several studies report increased expression of inflammasome markers (IL‐1β, ASC, and caspase‐1) in human brain samples (hippocampus, cortex) of AD patients when compared to healthy controls. ¹⁶ , ⁹⁵ , ⁹⁶ The study by Halle et al. ⁵⁹ suggests the NLRP3 inflammasome is a sensor of Aβ, and once activated, promotes cathepsin B release and lysosomal damage. In their study, NLRP3 inflammasome activation occurred following Aβ phagocytosis by microglia and the NLRP3 inflammasome was required for Aβ‐induced activation of caspase‐1, the release of mature IL‐1β and secretion of other pro‐inflammatory cytokines/chemokines. ⁵⁹ The study by Heneka et al. also demonstrates the NLRP3 inflammasome is a major source of neuroinflammation‐associated AD pathology. ¹⁶ In their in vivo studies, human brain samples (cortical or hippocampal lysates) from AD, early onset AD (EOAD), and mild cognitive impairment (MCI) patients displayed a significant increase in cleaved caspase‐1 when compared to healthy controls. ¹⁶ Studies in APP/PS1 mice, a transgenic mouse model of AD where cerebral amyloidosis occurs at 6–8 weeks of age, ⁹⁷ demonstrated increased Aβ clearance, reduced brain caspase‐1 and IL‐1β levels, and protection from memory loss in APP/PS1/caspase‐1 ^−/− and APP/PS1/Nlrp3 ^−/− mice compared to APP/PS1 controls. ¹⁶ The NLRP3 inflammasome has also been shown to drive tau pathology in AD. ⁹⁸ , ⁹⁹ In the study by Ising et al., tau activated the NLRP3 inflammasome, and loss of NLRP3 inflammasome function in THY‐Tau22 transgenic mice, a mouse model of AD where tau pathology occurs without loss of motor function, ¹⁰⁰ attenuated tau aggregation and hyperphosphorylation by regulating tau phosphatases and kinases. ⁹⁸ In the study by Stancu et al., ¹⁰¹ NLRP3 deficiency in tauP301S (PS19) transgenic mice, a mouse model of AD where mice develop brain atrophy and neuronal loss with NFT and tau tangle pathology, displayed significantly decreased seeding and propagation of tau. ⁹⁹ Additionally, 8‐month‐old NLRP3 deficient PS19 tau mice displayed significantly less hippocampal atrophy than their wild‐type PS19 littermates. ⁹⁹ In line with other neurodegenerative diseases, it appears polymorphisms in Nlrp3 can dramatically affect AD development and progression. ¹⁰² In the study by Tan et al., Nlrp3 polymorphisms (rs35829419, rs2027432, and rs10754558) were genotyped in 1133 late‐onset AD (LOAD) patients and 1159 healthy controls. ¹⁰² The rs2027432 polymorphism is demonstrated to increase promoter activities of the Nlrp3 gene ¹⁰³ and was most associated with overall LOAD AD risk. ¹⁰² The rs35829419 polymorphism in Nlrp3 is a gain‐of‐function mutation resulting in increased IL‐1β and IL‐18 secretion. ¹⁰⁴ However, in the study by Tan et al., the minor A allele of rs35829419 was absent in AD patients, but present in healthy controls, and appeared to provide a protective benefit from acquiring LOAD AD. ¹⁰² Meanwhile, the Nlrp3 rs10754558 polymorphism is reported to affect Nlrp3 mRNA stability ¹⁰⁵ and only differed in apolipoprotein E (ApoE) ε4 carriers. ¹⁰² Together, these studies not only highlight the importance of the NLRP3 inflammasome in AD pathology, but also suggests polymorphisms in Nlrp3 as a risk factor for AD susceptibility.

Microglia are a major source of neuroinflammation in many aging and neurodegenerative disorders, including AD. Several studies report NLRP3 inflammasome activation in microglia during Aβ and tau pathology. ⁹⁸ , ¹⁰⁶ , ¹⁰⁷ , ¹⁰⁸ NLRP3‐mediated microglial neuroinflammation is suggested to be an early event in AD pathogenesis, as studies demonstrate DAMPs recruit microglia to Aβ in the brain, and microglial expression of inflammasome markers (NLRP3, IL‐1β, IL‐18, caspase‐1, and ASC) have been observed in the early stages of AD pathology. ⁵⁹ , ¹⁰⁹ Furthermore, the NLRP3 inflammasome has been shown to be activated in microglia by low molecular weight Aβ protofibrils and oligomers, suggesting microglial activation may take place prior to Aβ deposition. ¹¹⁰ NLRP3‐mediated microglial neuroinflammation may also lead to tau and Aβ deposition in the brain, thereby promoting AD progression. ¹¹¹ , ¹¹² The study by Venegas et al. ¹¹¹ suggests accumulation of ASC into prion‐like ASC specks enhanced Aβ pathology by a cross‐seeding mechanism. Similarly, the study by Stancu et al. demonstrated tau seeds activate the NLRP3‐ASC‐dependent inflammasome in microglia to promote external tau‐seeding, highlighting a strong correlation between tau and Aβ activation of the NLRP3 inflammasome in microglia. ¹¹² In line with these studies, Friker et al. identified ASC‐Aβ complexes increased NLRP3 inflammasome activation more than ASC or Aβ alone. ¹¹³ Additionally, ASC secreted by pyroptotic cells can spread to nearby microglia to exacerbate the inflammatory response and inhibit the ability of microglia to phagocytize Aβ. ¹¹³ Other studies also suggest the NLRP3 inflammasome greatly influences microglial‐mediated clearance of Aβ. ¹⁶ , ¹¹⁴ Heneka et al. ¹⁶ demonstrated loss of NLRP3 or caspase‐1 resulted in enhanced Aβ clearance and shifted microglia to an anti‐inflammatory phenotype. Furthermore, Tejera et al. ¹¹⁴ showed systemic inflammation reduced Aβ clearance by microglia, and blocking NLRP3 inflammasome activation resulted in restored microglial Aβ clearance capacity. Overall, there is an abundance of evidence to suggest the NLRP3 inflammasome is a key mediator in microglial‐mediated neuroinflammation in AD.

3. KETONE BODY METABOLISM

Ketone bodies comprise a group of water‐soluble molecules (acetone, acetoacetate (AcAc), and β‐hydroxybutyrate (BHB)) produced primarily by the liver from fatty acid oxidation (FAO)‐derived acetyl‐CoA. ¹¹⁵ BHB and AcAc are the two main ketone bodies, with acetone originating by spontaneous enzymatic decarboxylation of AcAc. ¹¹⁶ Ketone bodies are associated with several important metabolic processes, including gluconeogenesis, FAO, lipogenesis, the tricarboxylic acid (TCA) cycle, and the biosynthesis of sterols. ¹¹⁷ Ketone body oxidation is critical during pregnancy, starvation, fasting, exercise, and low carbohydrate diets. ¹¹⁷ The human liver generates approximately 300 g of ketone bodies daily, accounting for 5%–20% of total energy dissipation depending on caloric state (fed, fasted, and starved). ¹¹⁸ , ¹¹⁹ , ¹²⁰

Ketone body metabolism consists of ketogenesis and ketolysis. Ketogenesis is a metabolic pathway where β‐oxidation of fatty acids produces BHB and AcAc. Ketogenesis is primarily carried out in the mitochondria of hepatocytes. ¹¹⁷ The mitochondrial enzyme 3‐hydroxymethylglutaryl‐CoA synthase 2 (HMGCS2) provides the essential rate‐limiting enzymatic reaction in ketogenesis, as HMGCS2 catalyzes the breakdown of β‐oxidation‐derived acetyl‐CoA and acetoacetyl‐CoA to generate β‐hydroxy β‐methylglutaryl‐CoA (HMG‐CoA) and free Coenzyme A (CoA). ¹¹⁷ , ¹²¹ HMG‐CoA is then cleaved by HMG‐CoA lyase to produce AcAc, which is further reduced to BHB by the enzymatic activity of D‐3‐hydroxybutyrate dehydrogenase 1. ¹¹⁷ , ¹²¹ Impaired ketogenesis is associated with several metabolic diseases including diabetic ketoacidosis. ¹²² Ketogenesis is mainly regulated by insulin. In diabetics, lack of or resistance to insulin prevents sugars from entering the cells resulting in dangerously high blood sugar levels. ¹²² Since cells can't rely on glucose for energy, large amounts of ketone bodies are produced for use as an alternative energy source. As ketone bodies are acidic, increased high anion gap metabolic acidosis occurs, resulting in kidney and respiratory distress. ¹²²

In contrast, ketolysis is the catabolic metabolism of BHB and AcAc for energy production. While ketogenesis takes place in specialized cells, almost all cells are capable of ketolysis. ¹²³ Ketone bodies are readily absorbed from the bloodstream through monocarboxylate transporter 1 (MCT1) which is expressed by most cells in the human body. ¹²³ , ¹²⁴ The principal enzymatic reaction that facilitates ketone body utilization as energy substrates is when succinyl‐CoA:3‐ketoacid‐CoA transferase (SCOT) converts ketone bodies to acetoacetyl‐CoA. ¹²⁵ Next, acetoacetyl‐CoA is cleaved into two molecules of acetyl‐CoA by acetoacetyl‐CoA thiolase (ACAT1). Acetyl‐CoA molecules are provided for ATP synthesis through oxidation in the respiratory chain and TCA cycle. ¹²⁵ As seen with impaired ketogenesis, impaired ketolysis can lead to metabolic diseases such as diabetic ketoacidosis. Additionally, impaired ketolysis, can result in ketoacidosis caused by genetic diseases that typically manifest during childhood. β‐ketothiolase, succinyl‐CoA: 3‐oxoacid CoA transferase deficiency, and MCT1 deficiency are three inherited disorders that cause extensive ketoacidosis due to impaired isoleucine metabolism, ketone body utilization, and ketone body transport, respectively. ¹²⁶ , ¹²⁷ , ¹²⁸ An overview of the ketogenesis and ketolysis pathways is shown in Figure 1.

FIGURE 1.

Overview of ketogenesis and ketolysis pathways. Ketogenesis (left) is a metabolic pathway in hepatocyte mitochondria that produces ketone bodies from fatty acids by β‐oxidation. Ketolysis (right) is the catabolic pathway in the mitochondria of peripheral tissues that utilizes ketone bodies for energy production. Together, these metabolic pathways enable fat‐derived energy to be produced in the liver and used by peripheral organs, such as the heart, brain, and skeletal muscles when carbohydrates cannot be used effectively or are not available.

Human brain activity demands a considerable amount of energy. Although the brain only consists of approximately 2% of the total human body mass, it accounts for nearly 20% of the oxygen and 25% of the glucose consumed by the body. The high energy demands of the brain are attributed to the generation and renewal of ion gradients, action potential generation, and neurotransmitter uptake. ¹²⁹ , ¹³⁰ Unlike most organs in the human body, the human brain is unable to efficiently utilize fatty acids for energy when glucose is unavailable. Ketone bodies are a primary alternative energy source for the brain when glucose is inadequate. Following an overnight fast, plasma ketone body levels are typically quite low, thus ketone bodies under most conditions only supply approximately 5% of the energy required for brain metabolism. ¹³¹ During prolonged fasting periods (5–6 weeks), however, ketone body levels rise significantly to supply approximately 60% of the energy required for brain metabolism. ¹³² , ¹³³ Since brain energy demands are highly dependent on the interplay between glucose and ketone body metabolism, applications to diseases of the CNS have been investigated for over a century when the ketogenic diet was first used as a treatment for epilepsy in the 1920s. ¹³⁴

4. MICROGLIAL KETONE BODY METABOLISM

Even in a resting state, microglia are highly active and continually surveil their ever‐changing microenvironments in the brain parenchyma. ¹³⁵ Such constant activity requires considerable amounts of energy, with glucose used as the primary energy source for microglia. ¹³⁶ Glucose is transferred across the blood–brain barrier (BBB) by glucose transporters where it is utilized by microglia to generate ATP through oxidative phosphorylation and glycolysis. ¹³⁷ When glucose is not available, microglia are able to adapt and use alternative energy sources, such as amino acids and ketone bodies. ¹³⁶ Accumulating evidence suggests metabolic flexibility contributes to the plasticity of microglial immune responses, as pro‐inflammatory microglia preferentially use glycolysis, whereas, anti‐inflammatory microglia primarily utilize oxidative phosphorylation. ¹³⁸ The reliance of microglial behavior on metabolic signals, however, renders them vulnerable to nutritional deficiencies and metabolic disorders. ¹³⁹

A large‐scale proteomics study of postmortem brains of AD patients indicated protein modules associated with increased microglial carbohydrate metabolism were highly correlated with AD progression. ³⁸ Indeed, several mechanisms promoting a metabolic switch to glycolysis that promote AD progression have been identified. ³⁷ , ¹⁴⁰ For example, in the study by Piers et al., ¹⁴⁰ microglia with mutations in TREM2 undergo a metabolic shift to glycolysis resulting in hindered phagocytic activity. Additionally, microglia exposure to Aβ induces a metabolic shift from oxidative phosphorylation to glycolysis which is dependent on the mammalian target of rapamycin (mTOR)‐ hypoxia‐inducible factor‐1, alpha subunit (HIF‐1α) pathway. ³⁷ These processes are exacerbated in AD, an age‐related disease, since normal aging also contributes to microglial metabolic dysfunction. Aging has been shown to promote pro‐inflammatory microglial phenotypes along with metabolic deficiencies leading to imbalances in metabolic substrates and reductions in brain glucose consumption. ¹⁴¹

Since the microglial glycolytic switch appears to be a major event in AD progression, investigation into microglial ketone body metabolism as an alternative energy source is of great interest. Reports indicated significantly lower levels of BHB in the blood of AD patients ¹⁴² with higher levels of BHB correlated with increased scores on cognitive tests. ¹⁴² , ¹⁴³ , ¹⁴⁴ Additionally, Shippy et al. ¹⁴⁵ demonstrated that BHB levels in brain tissue and red blood cell samples from AD patients were significantly lower than those from non‐AD controls. The study by Jin et al. ¹⁴⁶ explored the impact of Aβ oligomers and BHB in the context of human iPS‐derived microglia (hiMG), an applicable model to study AD pathogenesis. In their study, Aβ oligomers activated hiMG to release pro‐inflammatory cytokines (IL‐6, TNF‐α, IL‐1β) and mediators (ROS). Furthermore, Aβ oligomers stimulated increased expression of glycolytic transcripts and decreased expression of genes in the β‐oxidation pathway in hiMG. These metabolic deficiencies were mitigated by supplementation with BHB (0.1–2 mM) suggesting BHB can modulate microglial metabolism by pushing activated microglia away from pro‐inflammatory metabolic states. ¹⁴⁶ Another study has shown that BHB increases the accumulation of key metabolites (α‐ketoglutarate and fumarate) generated by the TCA cycle leading to an anti‐inflammatory phenotype in lipopolysaccharide (LPS)‐stimulated BV2 microglia. ¹⁴⁷ Additionally, other studies suggest BHB alters microglial morphology as BHB‐treated microglia display a ramified (resting) morphology suggestive of an anti‐inflammatory, non‐activated state. ¹⁴⁵ , ¹⁴⁸ Overall, perhaps the most significant contribution to the anti‐inflammatory properties of ketone bodies, particularly BHB, is the inhibition of the NLRP3 inflammasome (discussed below). ¹⁴⁵ , ¹⁴⁶ , ¹⁴⁹

5. TARGETING THE NLRP3 INFLAMMASOME WITH KETONE BODY THERAPIES FOR AD

Since the NLRP3 inflammasome is a major contributor to AD pathogenesis, ¹⁶ NLRP3 inflammasome inhibition is an attractive target for new therapeutic interventions for AD. A growing body of evidence supports the role of ketogenic regimens in the reduction of NLRP3 inflammasome activation. ¹⁵⁰ The ketogenic diet is a common low‐carbohydrate regimen where fat is used as the primary energy source. During ketosis, ketone bodies are produced which exert neuroprotective and anti‐inflammatory properties. ¹⁵¹ Several reports using mouse models of AD demonstrate reduced AD pathogenesis and improved cognitive functions in mice fed a ketogenic diet. ¹⁵² , ¹⁵³ , ¹⁵⁴ , ¹⁵⁵ , ¹⁵⁶ Yao et al. induced ketosis by 2‐deoxy‐D‐glucose (2‐DG) supplementation in 3xTgAD mice, a triple‐transgenic model AD mouse model displaying plaque and tangle pathology, ¹⁵⁷ to demonstrate a ketogenic diet improved bioenergetic capacity, reduced Aβ burden, and increased Aβ clearance. ¹⁵⁶ The study by Kashiwaya et al. ¹⁵² found mice fed a ketogenic diet had reduced Aβ deposition in the hippocampus along with reduced levels of hyperphosphorylated tau in the hippocampus, amygdala, and cortex. Other studies also reported decreased Aβ levels ¹⁵³ , ¹⁵⁵ and attenuated neuroinflammation ¹⁵⁴ , ¹⁵⁵ in mice fed a ketogenic diet. Additionally, most of these studies demonstrated a ketogenic diet improved cognitive performance in learning and memory tests. ¹⁵² , ¹⁵⁴ , ¹⁵⁵ In humans, numerous clinical trials and case reports suggest a beneficial outcome for AD patients on a ketogenic diet. ¹⁴³ , ¹⁵⁸ , ¹⁵⁹ , ¹⁶⁰ , ¹⁶¹ , ¹⁶² , ¹⁶³ Similar to mouse studies, AD patients on a ketogenic diet displayed increased ketone body levels and improved cognition/memory performance. ¹⁴³ , ¹⁵⁸ , ¹⁵⁹ , ¹⁶² , ¹⁶³ Other studies suggest a ketogenic diet not only enhances memory/cognitive performance but also improves metabolic capabilities and CSF biomarker profiles of AD. ¹⁶⁰ , ¹⁶¹

To some, the ketogenic diet is unpalatable and can be unhealthy due to the high fat content of the diet. To combat this, medium‐chain triglyceride (MCT) drinks were developed as an alternative. MCTs are rapidly absorbed by the small intestine and are substantially more ketogenic than dietary fats. ¹⁶⁴ , ¹⁶⁵ MCT supplementation has been shown to safely elevate serum ketone body levels regardless of carbohydrate consumption; therefore, MCTs are being investigated to treat AD. ¹⁶⁶ , ¹⁶⁷ , ¹⁶⁸ , ¹⁶⁹ Several clinical trials in humans with mild to moderate AD have demonstrated improved cognitive performance following an MCT regimen when compared to the placebo group. ¹⁶⁶ , ¹⁶⁷ , ¹⁶⁸ Furthermore, the study by Croteau et al. suggests MCT supplementation increases brain energy metabolism in AD patients. ¹⁶⁹ In their study, MCT supplementation doubled brain ketone body consumption without affecting brain glucose utilization, implying MCTs can compensate for brain glucose deficits often seen in AD patients. ¹⁶⁹ A type of MCT, caprylic acid (CA), is naturally found in coconut oil, palm oil, and in the milk of various animals, including humans. CA is metabolized into ketone bodies that can be used as an alternative energy source for brain metabolism. ¹⁷⁰ Axona®, composed of CA, is a Food and Drug Administration (FDA)–approved prescription dietary supplement for the treatment of AD. ¹⁷¹ Several trials in humans were conducted where study participants were subjected to screening procedures to determine baseline levels to include, but not limited to, cognition, BHB serum levels, and ApoE genotyping. ¹⁴⁴ , ¹⁷¹ , ¹⁷² , ¹⁷³ Axona® has shown the ability to improve cognitive deficits in ApoE ε4‐negative patients with mild to moderate AD. ¹⁴⁴ , ¹⁷¹ , ¹⁷² , ¹⁷³ Furthermore, Axona® appears to be well‐tolerated and safe, as the primary adverse effects are gastrointestinal intolerance, such as bloating, indigestion, and diarrhea. ¹⁷³ However, since MCTs mainly benefit ApoE ε4‐negative patients, and most patients with AD are likely to be ApoE ε4‐positive ¹⁷⁴ , ¹⁷⁵ MCTs in their current form are expected to have minimal impact in the treatment of AD. ¹⁷⁶

Most evidence supporting a beneficial role for the ketogenic diet in AD suggests the main circulating ketone body, BHB is primarily responsible for the positive effects. BHB has been demonstrated to attenuate AD pathogenesis by blocking Aβ entry into neurons, suggesting a neuroprotective role for BHB during neurodegeneration. ¹⁷⁷ , ¹⁷⁸ BHB has also been shown to prevent microglia activation and attenuate the microglial inflammatory response through inhibition of pro‐inflammatory cytokines (IL‐1β, TNF‐α, and IL‐6). ¹⁴⁵ , ¹⁴⁶ , ¹⁷⁹ , ¹⁸⁰ , ¹⁸¹ , ¹⁸² The study by Youm et al. ¹⁸³ provided a key mechanism for the neuroprotective effect of BHB, as they demonstrated BHB blocked NLRP3 inflammasome activation in human monocytes. Given these findings, Shippy et al. investigated the effect of BHB administration on the NLRP3 inflammasome in AD. ¹⁴⁵ Their results indicated exogenous BHB administration in mouse models of AD decreased overall AD pathology through inhibition of the NLRP3 inflammasome. ¹⁴⁵ BHB was able to cross the BBB and BHB administration resulted in less pronounced microgliosis, significantly fewer cortical plaques, and decreased cortical volume occupied by plaques. ¹⁴⁵ Furthermore, BHB‐treated mice exhibited significantly less processed caspase‐1 in their cortices, along with a marked reduction in ASC specks, indicating a decrease in NLRP3 inflammasome activation. ¹⁴⁵ The mechanism by which BHB inhibits NLRP3 inflammasome activation is still unclear. Several mechanisms of NLRP3 inflammasome activation have been elucidated including, but not limited to, lysosomal disruption, potassium ion efflux, and ASC nucleation‐induced polymerization/oligomerization. ⁴⁹ , ⁵⁵ , ¹⁸⁴ , ¹⁸⁵ In the study by Youm et al., BHB blocked intracellular potassium ion efflux in response to incubation with the NLRP3 activators monosodium urate (MSU), ATP, and ceramides. ¹⁸³ Furthermore, BHB also blocked ATP‐induced ASC oligomerization and speck formation, suggesting BHB blocks NLRP3 inflammasome activation by modulating an unknown upstream event that subsequently blocks this mechanism of NLRP3 inflammasome assembly. ¹⁸³ An overview of BHB inhibition of the NLRP3 inflammasome in AD is shown in Figure 2.

FIGURE 2.

BHB inhibits NLRP3 inflammasome activation to attenuate AD pathology. BHB is hypothesized to inhibit NLRP3 inflammasome activation by modulating an unknown upstream event involved in potassium ion efflux and ASC oligomerization/speck formation during NLRP3 inflammasome assembly. NLRP3 inflammasome inhibition significantly attenuates AD pathology by reducing neuroinflammation, plaques, and microgliosis. Created with BioRender.com.

6. CONCLUSIONS

In this review, we highlighted the important role of the NLRP3 inflammasome in AD pathogenesis. We also discussed how the emerging field of microglial metabolism has advanced our understanding of how systemic and cellular metabolism influences immune responses. Furthermore, BHB inhibition of the NLRP3 inflammasome provides a protective benefit from AD pathology, offering an exciting new avenue of AD treatment research. However, as discussed above, microglia play a multifaceted role in AD pathology and have been shown to be beneficial and detrimental in AD pathogenesis, depending on numerous factors, including metabolic state. Therefore, the possible conflicting functions microglia may have in AD pathogenesis pose a major challenge in targeting microglial‐specific metabolism as a therapeutic strategy. Ultimately, a deeper understanding of how ketone body metabolism by microglia modulates NLRP3 inflammasome activation may yield new insights into alternative therapeutic treatments for AD.

FUNDING INFORMATION

This work was supported by the William F. Vilas Trust Estate Early‐Career Investigator Award (to TKU) and the National Institutes of Health R01AG070973 and R01AG083883 (both to TKU). This work was also supported by P30AG062715 from the National Institutes of Health (awarded to the Wisconsin Alzheimer's Disease Research Center).

CONFLICT OF INTEREST STATEMENT

The authors declare no competing financial interests.

Shippy DC, Evered AH, Ulland TK. Ketone body metabolism and the NLRP3 inflammasome in Alzheimer's disease. Immunol Rev. 2025;329:e13365. doi: 10.1111/imr.13365

DATA AVAILABILITY STATEMENT

No data are available.