Very-low-carbohydrate diet enhances human T-cell immunity through immunometabolic reprogramming

Our data suggest KD as a feasible and effective clinical tool to augment human T-cell immunity. This could impact various clinical issues intimately correlated to T-cell immune disorders. In conclusion, our study changes the perspective on nutrition as a clinical tool and could help to redefine the role of dietary interventions in modern medicine.

Ketogenic diet markedly improved specific responses of human T lymphocytes in a balanced way—including T effector and T regulatory cell function—and increased the formation of memory T cells both in vitro and in vivo . This effect was based on a redirection of T-cell metabolism toward aerobic mitochondrial oxidation, resulting in enhanced cellular energy supply and respiratory reserve. These functional changes were in line with transcriptomic alterations, linking the KD to a fundamental immunometabolic reprogramming of human T cells.

The ketogenic diet (KD) is characterized by very limited uptake of carbohydrates resulting in endogenous production of ketone bodies as alternative energy substrates that can be utilized via mitochondrial aerobic oxidative phosphorylation There are as yet unproven assumptions that KD positively affects human immunity. We investigated this topic in an in vitro model using primary human immune cells and in a nutritional intervention study enrolling healthy volunteers.

Here, we present the first study investigating the influence of KD on human immune responses in vitro and in a cohort of healthy subjects. Our results reveal profound beneficial effects of ketone bodies on human T-cell immunity. BHB improves effector and regulatory T-cell function and primes human T memory cell differentiation both in vitro and in vivo . These functional changes are based on a fundamental immunometabolic reprogramming, resulting in enhanced mitochondrial oxidative metabolism, thus conferring an increased immunometabolic capacity to human T cells. We provide molecular evidence that ketone bodies promptly improve human T-cell metabolism and immunity. By complementing classical approaches of modern medicine, nutritional interventions offer new perspectives for prevention and therapy of numerous diseases.

In this regard, the high-fat low-carbohydrate ketogenic diet (KD) is one highly discussed approach (Bolla et al , 2019 ; Ruiz Herrero et al , 2020 ). Restriction of carbohydrate intake leads to the endogenous production of ketone bodies such as beta-hydroxybutyrate (BHB) as evolutionary conserved alternative metabolic substrates, which can be utilized via mitochondrial oxidative phosphorylation (Puchalska & Crawford, 2017 ). In animal models, BHB has been shown to dampen inappropriate innate immune responses via suppression of the NLRP3 inflammasome, thus ameliorating chronic low-grade inflammation and associated diseases (Youm et al , 2015 ; Goldberg et al , 2017 ; Newman et al , 2017 ). Human adaptive immunity, however, has not yet been addressed (Stubbs et al , 2020 ).

Western diet is increasingly seen as being at the root of many diseases, such as metabolic syndrome, autoimmune disorders, and cancer, and thus is suspected to limit life expectancy during the 21 st century (Olshansky et al , 2005 ; Christ & Latz, 2019 ). It impairs cellular immunity and evokes systemic low-grade inflammation not only by causing obesity but also by direct reprogramming of immune cells toward a proinflammatory phenotype (Hotamisligil et al , 1993 ; Visser et al , 1999 ; Weisberg et al , 2003 ; Mathis, 2013 ; Guo et al , 2015 ; de Torre-Minguela et al , 2017 ; Dror et al , 2017 ; Christ et al , 2018 ). Nutritional interventions may hold promise as a tool to prevent and even to treat disease. Unfortunately, most recommendations on food intake and dietary guidelines yet lack substantiated scientific background (Archer et al , 2018 ; Ioannidis, 2018 ; Archer & Lavie, 2019 ; Ludwig et al , 2019 ). Novel nutritional concepts promote a restriction of carbohydrates in favor of fat to ameliorate detrimental low-grade inflammation (Paoli et al , 2015 ; Bosco et al , 2018 ; Myette-Côté et al , 2018 ). However, large observational studies investigating this approach are highly controversial (Dehghan et al , 2018 ; Seidelmann et al , 2018 ), and molecular data in humans are scarce.

Immune responses result from a complex interplay of innate and adaptive mechanisms. To explore the full potential of KD in vivo , we analyzed TruCulture whole blood protein levels, which allow the detection of overall immune responses of adaptive and innate immune cells in a physiological environment during pathogen encounter. KD led to a significant upregulation of IFNγ, IL4, IL6, IL12sub40, IL23, and TNFα (Fig 6H ) and thus confirmed the enhancement of human adaptive immune response by KD. Importantly, innate immune master cytokines IL1α/β displayed unaltered expression levels (Fig EV4L ). TruCulture analysis of unstimulated whole blood samples also ruled out immunostimulatory effects of KD per se (Fig EV4M ).

Forty-four healthy volunteers conducted a 3-week KD with a limited carbohydrate consumption of < 30 g/day. Blood was taken and analyzed prior to starting the diet (T 0 ) and again after 3 weeks of strict adherence to the diet (T 1 ). PBMCs were isolated. If applicable, T-cell stimulation was performed through CD3/CD28 Dynabeads at a bead:cell ratio of 1:8. Pan/CD4 + /CD8 + T cells were separated via magnetic cell labeling. Mitochondrial metabolism was analyzed for each subpopulation using a Seahorse XF96 Analyzer.

Forty-four healthy volunteers conducted a 3-week KD with a limited carbohydrate consumption of < 30 g/day. Blood was taken and analyzed prior to starting the diet (T 0 ) and again after 3 weeks of strict adherence to the diet (T 1 ). PBMCs were isolated. T-cell stimulation was performed through CD3/CD28 Dynabeads at a bead:cell ratio of 1:8. Pan/CD4 + /CD8 + T cells were separated via magnetic cell labeling. Mitochondrial metabolism was analyzed for each subpopulation using a Seahorse XF96 Analyzer.

We next analyzed mitochondrial respiratory chain activity in human KD T cells. After 3 weeks of KD, T 1 CD4 + T cells exhibited upregulated basal oxidative respiration (Fig 6A ). For T 1 CD8 + T cells, profound changes were detected: elevated basal and maximal respiratory rates as well as augmented mitochondrial ATP production and spare respiratory capacity (Figs 6B and EV4A ). Of note, this superior oxidative capacity was not at the expense of glycolytic capabilities as seen for extracellular acidification and glycolytic proton efflux (Fig EV4B–G ). These metabolic alterations—leading to emphasized oxidative phosphorylation—translated into elevated cellular and mitochondrial ROS in T 1 T cells (Fig 6C ). Also in vivo , no impairment of mitochondrial membrane integrity occurred (Fig EV4H ). Cellular increase in ROS was restricted to T-cell activation, whereas higher mROS could also be detected in unstimulated T 1 T cells (Fig EV4I and J ). Consequently, T 1 CD4 + and T 1 CD8 + T cells displayed higher mitochondrial mass and augmented expression of OXPHOS protein complexes (Figs 6D–F and EV4K ). We then evaluated the effects of KD on memory cell development in human T cells in vivo . Memory cell subset analysis revealed elevated levels of both CD4 + and CD8 + CCR7 − CD45RA − CD45RO + effector memory cells (T Em ; Fig 6G ). In summary, KD led to an increase in mitochondrial mass, ROS production, and aerobic oxidative metabolism through enhanced respiratory chain activity. These KD-induced changes in human T-cell immunometabolism promote T mem development.

We subsequently investigated the immunometabolic phenotype of human T cells in response to KD (KD T cells), which revealed similar changes as found in vitro . In unstimulated CD4 + /CD8 + T cells, only minimal transcriptional activity and no relevant change of T-cell cytokine levels T 0 /T 1 was detected (Appendix Fig S2A and B ), whereas ex vivo activated T 1 CD4 + T cells displayed an upregulation of IL2, IL4, and CTLA4 mRNA (Fig 5A ). KD led to an elevation of Th 2 transcription factor GATA3 (Fig 5B ) resulting in a decline of Tbet/GATA3 ratio (Appendix Fig S2B ), an augmentation of Th 2 T helper cell subset and a decline in Th 1 /Th 2 cell ratio (Fig 5C ). T 1 T reg subpopulation was also enhanced as shown by a higher abundance of CD4 + CD25 + Foxp3 + T cells and an upregulation of IL10 mRNA (Fig 5D ). Moreover, T 1 CD8 + T cells exhibited elevated IFNγ, GZMB, and CTLA4 cytokine levels and significantly increased cell lysis activity (Fig 5E and F ). Thus, these data demonstrate an enhanced human T-cell immune capacity through KD in vivo .

These findings were substantiated by gene set enrichment analysis (GSEA; Fig 4E ). Several immune response pathways—involving both inflammatory and regulatory gene sets—were upregulated in response to KD. Major alterations could also be detected for metabolic gene sets. In summary, these results highlight substantial transcriptomic changes in favor of T-cell action, memory cell differentiation, and oxidative metabolism, thus pointing to immunometabolic reprogramming of human CD4 + and CD8 + T cells through KD.

In vitro results indicated substantial alterations to T-cell immunometabolism in response to BHB. To analyze the global impact of KD on gene expression in vivo , transcriptome profiles of human T-cell RNA samples from healthy volunteers conducting a 3-week KD were generated (Fig 4B ). Differential expression analysis T 1 /T 0 revealed a significant regulation of gene expression in both CD4 + and CD8 + T cells in response to KD (Fig 4C and D ). T 1 T cell transcriptome reflected enhanced immune competence with upregulation of both genes related to T-cell activation and effector function (TNF, TNFRSF9, CRTAM, DUSP4, TRAV30) but also inhibitory and regulatory capacity (EGR2, NFKBID). Importantly, enrichment of TCR-related genes TRAV30 and NR4A1 (Nur77) was also detected. Nur77 is known to indicate T-cell receptor signaling strength and is associated with memory cell development (Li et al , 2020 ; Shin et al , 2020 ). Notably, Nur77 has recently been identified as a central regulator of T-cell immunometabolism (Liebmann et al , 2018 ). Furthermore, genes involved in mitochondrial function were also differentially expressed on a KD (NR4A1, TOMM7, ACSL6).

The impact of KD on human immune responses in vivo has not yet been investigated. We conducted a prospective immuno-nutritional in vivo study enrolling 44 healthy volunteers, who performed a 3-week KD. Blood was collected prior to the start (T 0 ) and at the end (T 1 ) of KD. Blood ketone body levels were closely monitored. The study design is depicted in Fig 4A ; characteristics of study participants are given in Table 1 . During the course of KD, a significant increase in blood ketone bodies (in the range of 1.0–1.5 mM) was detected until the 2 nd week, which could be sustained until the end of the study (Fig EV3A ). Fasting serum glucose, however, remained unchanged throughout the course of a 3-week diet (Fig EV3B ). Apart from a brief phase of increased fatigue (duration 1–5 days), no relevant side effects were reported (Table 1 ). KD led to a significant decrease in participants' body mass index (BMI; Fig EV3C ). However, this effect was mainly obtained by overweight volunteers, who achieved a profound weight loss, while participants with T 0 BMI < 25 experienced only marginal BMI changes (Table 1 ).

Collectively, we have shown an increase in mitochondrial mass, ROS production, and aerobic oxidative metabolism through enhanced respiratory chain activity in BHB + human T cells. This might result in an augmentation of memory T cell (T mem ) formation as they primarily rely on OXPHOS, and increased spare respiratory capacity has been viewed as a hallmark of their immunometabolism (van der Windt et al , 2012 ; O'Sullivan et al , 2014 ). Supporting this hypothesis, BHB-cultivated cells expressed elevated levels of T mem differentiation factor IL15 (Fig 3H ). We applied a T mem differentiation protocol to further investigate the impact of ketone bodies on T mem development (Fig 3I ). As shown in Fig 3J , BHB + CD4 + and CD8 + T cells displayed higher expression of IL7R and IL15 after differentiation. CD4 + and CD8 + central and effector memory T-cell (T cm /T em ) subset analysis revealed a significant increase in CCR7 − CD45RA − CD45RO + CD8 + T em (Fig 3K ). In summary, an enhanced aerobic mitochondrial metabolism in response to BHB directs human T cells toward T mem differentiation.

Increased mROS may be the result of amplified oxidative phosphorylation. Indeed, we observed increased mitochondrial mass in BHB + CD4 + and CD8 + T cells (Figs 3C and D, and EV2F ), and Western blot analysis revealed enhanced expression of ETC complexes in BHB + primary human T cells and CD4 + /CD8 + T-cell subsets (Figs 3E and EV2G ). With respect to hormesis, elevated (mitochondrial) ROS have also been associated with mitochondrial and cellular damage (Zorov et al , 2014 ). However, an increase in GSH levels shows that protective antioxidative capacity also improved in BHB + CD4 + and CD8 + T cells (Fig 3F ). To further exclude structural impairment of mitochondria evoked by ketogenic conditions, we assessed the mitochondrial integrity of these cells. As depicted in Fig 3G , BHB did not alter mitochondrial membrane potential of primary human T cells, neither in CD4 + nor in CD8 + T-cell subsets. Hence, ketone metabolism does not compromise mitochondrial integrity.

Human peripheral blood mononuclear cells (PBMCs) were cultivated for 48 h in RPMI containing 80 mg/dl glucose (NC) and supplemented with 10 mM beta-hydroxybutyrate (BHB). T-cell stimulation was performed through CD3/CD28 Dynabeads at a bead:cell ratio of 1:8. Flow cytometry analyses were performed using the indicated dyes and antibodies identifying T cells and CD4 + /CD8 + T-cell subsets. For JC1 analysis, magnetic CD4 + /CD8 + T-cell separation was performed prior to staining.

The mitochondrial respiratory chain complexes I, II, and III are viewed as the major physiological sources of cellular and mitochondrial reactive oxygen species ([m]ROS) (Liemburg-Apers et al , 2015 ), and moderate mROS levels have been shown to be required for T-cell activation (Kamiński et al , 2012 ; Sena et al , 2013 ). Therefore, an increase in tricarboxylic acid cycle/electron transport chain (TCA/ETC) activity due to an elevated supply of BHB may result in augmented production of reactive oxygen species, serving as a T-cell “second messenger”, hence amplifying T-cell immune capacity. To test this hypothesis, we evaluated cellular and mitochondrial ROS in T cells treated with BHB. Flow cytometry analysis revealed higher cellular ROS in BHB + stimulated T cells, which could also be detected in CD4 + and CD8 + T-cell subsets (Figs 3A and EV2D ). Additionally, in all cell subsets, elevated levels of mitochondrial superoxide were found (Figs 3B and EV2E ). These changes were limited to activated T cells (Fig EV2A and B ).

Human PBMCs were cultivated for 48 h in RPMI containing 80 mg/dl glucose (NC) and supplemented with 10 mM beta-hydroxybutyrate (BHB). T-cell stimulation was performed through CD3/CD28 Dynabeads at a bead:cell ratio of 1:8. Pan T cells and CD4 + and CD8 + T cells were isolated through magnetic cell separation. Mitochondrial metabolism was analyzed for each subpopulation using a Seahorse XF96 Analyzer.

OCR, maximum respiration, spare respiratory capacity, and basal respiration were measured in (A) pan T cells, (B) CD4 + T cells, and (C) CD8 + T cells, n = 9 (pan T), n = 3/4 (CD4 unstimulated), n = 7/8/9 (CD4 stimulated), n = 5 (CD8 unstimulated), and n = 8/7/7 (CD8 stimulated) individual experiments. Data depicted as mean ± SEM (OCR) and box plots with median, 25 th and 75 th percentiles and range (all other). Dots indicating individual values. * P < 0.05, ** P < 0.01, paired t -test/Wilcoxon matched-pairs signed rank test, as appropriate.

We next hypothesized that an elevated energy supply may evoke these effects. To analyze the functional impact of BHB on T-cell metabolism, we performed extracellular flux analyses. BHB-cultivated primary human T cells showed an increased maximal respiration and spare respiratory capacity (Fig 2A ). BHB + CD4 + T cells exhibited elevated basal oxidative respiration (Fig 2B ). In CD8 + T cells, a marked shift toward oxidative metabolism—consisting of elevated basal and maximal respiratory rates, mitochondrial ATP production, and spare respiratory capacity—was displayed (Figs 2C and EV1A ). Notably, these alterations were not based on suppressed glycolysis, as BHB + T cells retained their glycolytic capacity (Fig EV1B and C ). Taken together, these data provide evidence for a substantial metabolic shift of T cells toward mitochondrial oxidative phosphorylation (OXPHOS) in response to BHB.

Notably, BHB had no significant impact on functions of the innate immune system. We conducted functional analyses of human whole blood supplemented with 10 mM BHB. Both phagocytic capacity and cellular respiratory burst displayed no change due to BHB (Appendix Fig S1F and G ). Furthermore, innate master cytokine IL1β was not significantly upregulated in BHB + PBMC (Appendix Fig S1H ).

To evaluate the impact of ketone bodies on human immunity, human peripheral blood mononuclear cells (PBMCs) were cultivated for 48 h with D/L-BHB (BHB + ) under stimulating conditions. In dose-finding experiments, only 10 mM BHB significantly enhanced human Pan T-cell immune response (Appendix Fig S1A and B ). 10 mM racemic BHB corresponds to approximately 5 mM metabolically active D-BHB, resembling a near-maximum level of endogenous ketone production (Puchalska & Crawford, 2017 ). Thus, 10 mM D/L-BHB was used to further decipher the impact of short-term BHB incubation on human T-cell immunity. BHB led to a significant transcriptional upregulation of CD4 + T-cell cytokines interleukin (IL)2, IL4, IL8, and IL22 (Fig 1A ). Protein analyses also showed an upregulation of IL2, IL4, IL6, and IL8 (Fig 1B ). CD4 + subset analysis revealed a downregulation of the Th1 transcription factor Tbet—resulting in a decrease in Tbet/GATA3 ratio—yet no significant changes of the Th 1 /Th 2 cell ratio (Fig 1C , Appendix Fig S1E ). CD8 + T-cell response displayed markedly increased levels of interferon γ (IFNγ), cytolytic proteins perforin 1 (PRF1) and granzyme B (GZMB), cytotoxic T lymphocyte-associated antigen 4 (CTLA4), and tumor necrosis factor alpha (TNFα; Fig 1D ). IFNγ and TNFα protein secretion was also elevated in response to BHB, and functional analysis unveiled enhanced cell lysis activity (Fig 1E ). Of note, in the absence of activating stimuli, unstimulated T cells did not exhibit relevant expression levels of immune cytokines, irrespective of their nutritional situation (Appendix Fig S1C and D ).

Western diet is increasingly recognized as a true endangerment to public health. It accounts for a rapid increase in obesity, diabetes, cardio- and neurovascular diseases, and even cancer (Mattson et al, 2014; Cohen et al, 2015; Ludwig, 2016; Hotamisligil, 2017). Chronic low-grade inflammation induced by both adipose tissue and forced dietary uptake of carbohydrates is considered a main driver of these conditions. The resulting unspecific activation of the innate immune system is not only harmful in itself, but also strongly impairs adaptive immune responses and hampers the ability to create immunological memory (Neidich et al, 2017; Napier et al, 2019; Ritter et al, 2020).

Modern medicine has to establish new strategies to enable immunometabolic reprogramming to prevent and even treat these damaging conditions. At this point, nutritional interventions as a clinical tool enter the stage. One current and highly discussed approach is KD, a high-fat low-carbohydrate KD. Ketosis represents an evolutionary conserved physiological state characterized by reduced carbohydrate uptake leading to moderate hepatic production of ketone bodies (“ketosis”, 0.5–5 mM BHB) (Krebs, 1966). KD is considered a tool to control body weight and has been shown to potently ameliorate inflammatory processes in mouse models (Paoli et al, 2013; Youm et al, 2015; Goldberg et al, 2017). To date, however, the impact of KD on human immunity remains elusive.

We present the first study investigating the effects of KD on human immune cells in vitro and in a prospective series of healthy subjects. We did not use murine models due to their substantial inherent limitations with respect to accurately reflecting human immune responses (Pulendran & Davis, 2020). We report profound immunometabolic effects of KD on human lymphocytes both in vitro and in vivo. KD T cells displayed enhanced reactivity upon specific stimulation, strengthened cell lysis capacity, increased differentiation of regulatory T cells, and pronounced T memory cell formation. KD promoted mitochondrial mass and expression of ETC complexes, enabling augmented mitochondrial oxidative phosphorylation leading to moderately increased cellular and mitochondrial ROS production and superior respiratory reserve. These alterations were accompanied by global transcriptomic changes of human CD4+ and CD8+ T cells, linking KD to a fundamental immunometabolic reprogramming. In contrast to previous mouse studies, innate immune responses were not compromised by KD, which further highlights the need for human in vivo data.

For the in vivo part of our study, healthy participants performed an individual 3-week ad libitum KD, resulting in blood ketone levels in the range of 1.0–1.5 mM. No relevant adverse effects were reported. Of note, morning fasting glucose levels remained in normal range during the course of the 3-week diet, most likely due to hepatic gluconeogenesis, consistent with previous data (Cahill & Veech, 2003). It can be assumed that longer periods of KDs are needed to achieve a stable lowering of blood glucose levels. BMI of overweight study participants (BMI>25) was significantly reduced after the study period, while the body weight of non-overweight subjects remained almost unaffected. This could indicate a particular responsivity of adipose tissue to the reduction of blood glucose peaks and subsequent reduced insulin secretion. However, in our series, covering a time span of 3 weeks, no significant differences in the immunological outcome between overweight and non-overweight participants were detectable, which underlines our in vitro findings pointing out direct immunomodulatory effects of BHB. Longer KD periods in the specific setting of obesity are probably required to detect potential indirect effects resulting from the reduction of adipose tissue.

T cells have been shown to incorporate BHB-derived acetyl-CoA into the TCA cycle (Zhang et al, 2020). Compared to fatty acid oxidation (FAO), metabolization of BHB displays higher efficiency due to NADH-based delivery of all reducing equivalents to complex I of the ETC (Cotter et al, 2013). Elevated energy supply through metabolization of BHB may thus improve T-cell immunity. We, indeed, found that use of BHB as an alternative energy fuel enhanced mitochondrial metabolism and increased cellular power reserves. Activated T cells obtain a large proportion of their ATP production through aerobic glycolysis both in vitro and in vivo (Michalek et al, 2011; Chang et al, 2013; Macintyre et al, 2014; Buck et al, 2016). Importantly, the here reported upregulation of mitochondrial oxidative phosphorylation was not on the cost of glycolysis but on top, as we observed unchanged glycolytic capacity of human KD T cells. As a result, higher overall levels of biochemical energy are yielded. It is conceivable that KD could also increase glutaminolysis, another ATP-generating pathway used by T cells (Wang et al, 2011; Wang & Green, 2012), for example, by increasing activity of glutamate dehydrogenase via SIRT4 inhibition (Li et al, 2011; Min et al, 2018). As T cells—in contrast to tumors—are capable of glutamine oxidation, it can be assumed that ATP production as measured by Seahorse analysis yields reliable results. However, the contribution of glutamine fermentation to cellular ATP production escapes detection via Seahorse analysis and may thus further contribute to T-cell energy content. Detailed deciphering of ATP-delivering pathways affected by KD requires further investigation.

Effector T cells may additionally be supported by elevated mitochondrial reactive oxygen species, which are required to mount an antigen-specific immune response (Sena et al, 2013). The role of ROS in T-cell immunity is exceptional and markedly differs from tissues and particularly from cancer cells: ETC-derived ROS serve as a second messenger during T-cell activation (Murphy & Siegel, 2013; Franchina et al, 2018), thus being considered pivotal for T-cell immunity (Devadas et al, 2002; Jones et al, 2007; Sena et al, 2013). Of note, T cells diverge from NADPH-dependent GSH synthesis, instead redirecting NADPH for ROS production via NADPH oxidase to fulfill their demand on reactive oxygen species (Jackson et al, 2004; Kwon et al, 2010). These findings gave rise to the concept of mitohormesis, opposing the idea of ROS as solely detrimental byproducts of an imperfect oxidative system, but emphasizing the role of ROS as essential signaling molecules (Ristow, 2014). In non-phagocytic cells, ETC complexes are known to account for the majority of ROS production (Desdín-Micó et al, 2018; Raimondi et al, 2020). We show that KD leads to an amplified expression and activity of ETC complexes in human T cells, creating higher levels of mitochondrial and—in case of T-cell activation—cellular ROS. A mild increase in ROS production has already been associated with protective effects of KD via adaptational upregulation of antioxidative capabilities and amelioration of oxidative stress, which is in line with the observed augmentation of GSH levels and the preservation of mitochondrial membrane potential during KD (Milder & Patel, 2012). Consequently, both enhanced overall energy supply and (m)ROS-signaling create the bioenergetic basis for augmented T-cell immune responses in KD T cells.

Importantly, KD revealed also as a possible strategy to prime T cells toward T memory cell formation. We discovered an increase in mitochondrial mass, spare respiratory capacity (SRC)—both considered hallmarks of T mem cells—and, indeed, an enrichment of T memory cell formation in vivo and in vitro (van der Windt et al, 2012, 2013). It has previously been reported that T memory cell development depends on oxidative metabolism to ensure superior respiratory reserve, sustaining higher levels of ATP upon reactivation (Buck et al, 2016; Simula et al, 2017). Moreover, T cells are known to perform active ketogenesis to utilize BHB for epigenetic modifications to support memory cell development (Zhang et al, 2020).

Ketogenic diet could also prove to be a valuable measure in the threatening situation of the current COVID-19 pandemic, as dysregulation and exhaustion of both CD4+ and CD8+ T cells as well as impairment of classical CD8+ T effector memory cells have been reported as central elements of COVID-19 immunopathology (Chen & John Wherry, 2020; Habel et al, 2020). Thus, clinical evaluation could be promising.

Our study demonstrates that KD induces a fundamental immunometabolic reprogramming in human T cells associated with profound transcriptomic changes. This leads to a balanced global enhancement of T-cell immunity, comprising enhanced cytokine production and secretion, strengthened cell lysis capacity, amplified Treg differentiation, and pronounced Tmem cell formation. KD thus holds promise as a feasible and effective clinical tool for a large range of conditions intimately associated with immune disorders. This provides the basis to proceed into further medical translation. Consequently, a clinical phase II study investigating KD in sepsis patients is currently recruiting (Rahmel et al, 2020).

These new immunological aspects of KD might contribute to the modern concept of metabolic therapy of cancer (Seyfried et al, 2017, 2020). KD not only targets the Warburg effect, but could also strengthen anti-tumor immunity (Ferrere et al, 2021). Moreover, the different effects of BHB-induced ROS elevation on tumor cells and T cells might even lead to additive beneficial effects: While oxidative stress compromises cancer cell viability, mild increase in ROS enhances T-cell immune capacity which in turn further restrains tumor growth. However, these issues need to be addressed in future clinical studies.

In conclusion, our study changes the perspective on nutrition as a clinical tool and could help to redefine the role of dietary interventions in modern medicine.