Chronic liver disease is a social and economic burden ¹. Among the most common liver disorders is metabolic dysfunction–associated steatotic liver disease (MASLD), redefined in 2023 and previously described as non-alcoholic fatty liver disease (NAFLD)². Its prevalence is estimated at about 30% of the adult population worldwide ^3–5 while it can reach up to 90% in obese individuals and 60% in people with diabetes ¹⁶. Up to 20% of the normal-weight people are also concerned.

MASLD is characterized by the association of excessive fat accumulation in liver and at least one metabolic risk factor including overweight/obesity, type 2 diabetes, metabolic syndrome, hypertriglyceridemia andhypercholesterolemia. ². This steatosis results from the accumulation of lipids in liver parenchymal cells and triggers hepatic inflammation and fibrosis. Over time, it leads to cirrhosis and its associated complications ⁷. Thus, MASLD is correlated with higher risks of cancer and cardiovascular disease mortality ^8,9 and considered as the second most common reason for liver transplantation in the United States and Europe ^10,11.

In light of this, strategies promoting hepatic lipid clearance are of particular interest. Different pharmacological approaches are acknowledged including statins, SGLT-2 inhibitors, GLP-1 agonists, pioglitazone, vitamin E and resmetirom ^3,12,13. Resmetirom, a liver-directed thyroid hormone receptor beta-selective agonist, has recently been approved by FDA for MASLD treatment ^14–16. However, these treatments designed to cure rather than to prevent evidence side effects (e.g., muscle related issues for statins; nausea and diarrhea for resmetirom) ¹³ and may be quite expensive for patients and healthcare systems ¹⁷. This highlights the need for alternative, safer, and preventive options. In this context, lifestyle interventions including dietary changes and exercise represent the initial step in the treatment as well as nutritional or nutraceutical strategies which have become a major field for therapeutic innovation ¹.

Among natural bioactives of interest, polyphenols are plant compounds known for their antioxidant and anti-inflammatory properties. Regarding their role in MASLD, the last five years have accounted for more than half of the total publications related to this theme, demonstrating the growing interest in nutritional approaches in the field. Yet, only about 7% of these investigations have been conducted at the clinical level. Certain plant extracts are already known for their hepato-protective properties. For example, artichoke leaf extract (Cynara scolymus) is widely described in the literature for its choleretic, hepatoprotective ¹⁸ and lipid metabolism regulation properties ^19–22. Nevertheless, these strategies lack of relevant clinical data and optimization of the expected benefits, particularly through nutritional synergies, require further clinical validation.

TOTUM-448 is a patented polyphenol-rich blend of 5 different plant extracts (olive leaf (Olea europaea), bilberry (Vaccinium myrtillus), artichoke leaf (Cynara scolymus), chrysanthellum (Chrysanthellum indicum subsp. afroamericanum B.L. Turner), black pepper (Piper nigrum)) plus choline with potential synergistic properties on liver function. The aim of this work was to investigate whether and how this combination of different polyphenols can maintain, preserve or even improve hepatocyte metabolism in a lipotoxic context, supporting preventive strategies in the early stages of MASLD.

To address this working hypothesis, we collected human serum containing circulating bioactive metabolites resulting from TOTUM-448 ingestion and according to the Clinic’n’Cell patented methodology ^22–29, we evaluated their ex vivo influence on the behavior and function of human hepatocytes in a lipotoxic environment.

Hepatocytes were stressed by the incubation with 250µM of palmitate to mimic a MASLD context. Stress was performed in the presence of either the naïve or the enriched serum fractions to test whether metabolites were able to contribute to maintain hepatocyte function in such a lipotoxic environment. According to red oil staining, the presence of palmitate dramatically promoted lipid storage in hepatocytes (Fig. 3A and B). Neither the naïve nor the enriched human serum fraction had any impact on such lipid accumulation alone. In contrast, the impact of palmitate on the staining was significantly reduced when the cells were incubated with TOTUM-448 metabolites. To get more in-depth, we characterized the type of lipids involved in such a lipid accumulation. We found that red oil staining observation remarkably parallels with triglycerides and cholesterol contents (Fig. 3C and D). Palmitate significantly stimulated both triglycerides and cholesterol accumulation in human hepatocytes, while the presence of TOTUM-448 metabolites significantly prevented it (triglycerides (TG) -27%%, p ≤ 0.001; cholesterol − 52%, p ≤ 0.0001). We then, investigated hepatocytes transcriptomic activity of genes related to lipid metabolism. Both DGAT2 and SREBP1c were significantly up-regulated upon palmitate incubation (Fig. 3E and F). Such an up-regulation was blunted when hepatocytes were incubated with the enriched human serum fraction (-64%, p ≤ 0.001 and − 68%, p ≤ 0.0001 for DGAT2 and SREBP1c, respectively).

Along with lipotoxicity, the pro-inflammatory context drives altered hepatocyte function and subsequent MASLD onset. Using a DCFDA probe, we showed that palmitate promoted ROS production (Fig. 4A). The presence of TOTUM-448 metabolites tends to limit this production(p > 0.05). Regarding the inflammatory status, while only trends appeared when expressions of the targeted genes were considered separately (suppl. figure S1), using a 2-way Anova approach to globalize the analysis, we observed a significant group effect (2-way ANOVA, p = 0.0008) and significant differences in post hoc comparisons of overall inflammatory gene expression between groups NHS and NHS + P (250µM), p < 0.05, EHS and EHS + P(250µM), p < 0.05, and NHS + P (250µM) and EHS + P (250µM), p < 0.01, suggesting that palmitate consistently induced inflammatory gene expression while the presence of the metabolites significantly limited this induction (Fig. 4B). Still, the presence of metabolites did not achieve a return to baseline, as it remained significantly different from the control group (without palmitate).

Abnormal lipid accumulation in steatotic livers coincides with perturbed endoplasmic reticulum (ER) proteostasis in hepatocytes. Therefore, we checked for Unfolded Protein Response (UPR) markers in palmitate stressed hepatocytes. As expected, palmitate markedly and significantly promoted both CHOP and XBP-1 expression (Fig. 5A and B). Such an induction was abolished when cells were treated with TOTUM-448 metabolites (-36%, p ≤ 0.01 and − 66%, p ≤ 0.0001 for CHOP and XBP1, respectively. However, neither the naïve (NHS) nor the enriched human serum (EHS) fraction had any impact on those markers in the absence of a lipotoxic context. To support these data, we analyzed ATF-6 activity as the result of the presence of ATF-6 protein in the nucleus. Accordingly, ATF-6 was increased upon palmitate stress. However, limitation by TOTUM-448 metabolites remained a trend (-29%, p = 0.2).

In this manuscript, we first demonstrated that the ingestion of TOTUM-448 (1) led to bioavailable circulating metabolites and that these metabolites were able to drive the attenuation of palmitate-induced intracellular lipid storage including (2) TG (-27%, p ≤ 0.001) and (3) cholesterol (-52%, p = 0.0001), as well as the inhibition of (4) DGAT2 (-64%, p ≤ 0.001) and (5) SREBP1-c (-68%, p ≤ 0.0001) gene expression. TOTUM-448’s metabolites also inhibited (6) palmitate-induced inflammatory gene expression (p ≤ 0.01). A major effect was related to ER stress. While palmitate potently induced both (7) CHOP, (p ≤ 0.001) and (8) XBP1 mRNA expression (p ≤ 0.0001), (9) ATF-6 activity (p = 0.0053) and ultimately (10) Caspase-3 activity (p ≤ 0.0001), the presence of TOTUM-448’s metabolites normalized these markers (-36%, p ≤ 0.01; -66%, p ≤ 0.0001 and − 31%, p ≤ 0.01, for CHOP, XBP1 and Caspase-3, respectively; a trend was observed for ATF-6 (-29%, p = 0.2)).

First, the nutritional dimension of the dose used may be questioned. Each volunteer was exposed once per clinical phase to 4284 mg of TOTUM-448 representing an approximative global amount of 372 mg of polyphenols (see suppl. Table S1 for chemical characterization of TOTUM-448). The recommended dose for polyphenol daily intakes is about 1g, thus the nutritional dimension of the dose is quite relevant. Besides, the rational of the dose also echoes with paralleled clinical trials we have launched. Indeed, the dose chosen is also the one tested for long term supplementation and investigation in humans. The clinical trial is ongoing: NCT06704321.

Then, the bioavailability of the ingredient was investigated. In this study, we were able to detect 12 different circulating metabolites resulting from the ingestion of the blend including 2 oleuropein glucuronide isomers (olive leaf ³⁰), 3 luteolin glucuronide isomers (metabolites that may derive from luteolin-7-O-glucoside, luteolin-7-O-glucuronide and luteolin presents in chrysanthellum, in both artichoke and olive leaves ³¹), 1 ferulic acid sulfate (metabolite that may derive from either chlorogenic acid present in artichoke leaf and, chrysanthellum or caffeoylquinic acid also present in artichoke leaf ^32,33), 3 hydroxytyrosol sulfate isomers, 1 hydroxytyrosol glucuronide, 1 tyrosol glucuronide, 1 homovanilic acid sulfate (metabolites that may derive from both hydroxytyrosol and oleuropein present in olive leaf ³⁰). These metabolomic analyses evidenced a clear and diverse bioavailability of the components of the blend and supported the rational for further investigation of the bioactivity of the serum.

One may also question the ex vivo model. In our clinical ex vivo approach, volunteers should be seen as metabolites producers rather than patients expecting health benefits with measurable clinical scores. Thus, this protocol is designed to collect circulating metabolites in a standardized way and only male healthy volunteers with no treatment or pathology, aged from 18 to 35, BMI ranged from 20 to 28, with normal kidney and liver functions are recruited. The collected serum, either naïve (control baseline) or enriched (with circulating metabolites of interest) are further used/incubated on cell cultures according to the patent DIRV#18–0058 (written invention disclosure by the French National Institute for Agronomic, Food and Environment Research INRAE) to investigate a wide range of cell activities/markers that represent the final readout/score. In this study we targeted hepatocytes as a major protagonist in lipid metabolism and MASDL onset.

Hepatocytes are responsible for glucose, xenobiotic metabolism and for lipids metabolism as well. In our hands, palmitate consistently promoted lipids accumulation in hepatocytes including triglycerides and cholesterol mimicking a relevant MASLD onset. The presence of metabolites resulting from TOTUM-448 ingestion potently prevented these MASLD hallmarks. From a mechanistic point of view, these results match with the inhibition of both, DGAT2, an enzyme involved in triglycerides synthesis ³⁴ and SREBP1c, involved in the de novo lipogenesis in the liver ³⁵. Besides, these results are fully consistent with published observations regarding the effect of the different components of the blend when studied separately. In patients with diagnosed NAFLD, supplementation with artichoke leaf extracts (Cynara scolymus L.) (600 mg daily for 9 weeks) reduced liver size and decreased serum total cholesterol and triglyceride concentrations ³⁶. In a high-fat diet rat model, artichoke leaf extract supplementation limited hepatic steatosis ³⁷. A clinical study on a cohort of steatotic patients showed that supplementation with Chrysanthellum americanum extract decreased fibrosis and hepatic steatosis ³⁸. Decreased choline intake is significantly associated with increased liver fibrosis in post-menopausal women with NAFLD ³⁹.

In MASLD, the lipotoxic context fuels inflammation and oxidative stress by stimulating ROS production, pro-inflammatory cytokine release and by promoting the recruitment of immune cells. In light of this, our cellular model fully addresses these criteria as palmitate increases DCF-DA staining and inflammatory gene expressions. Consistent with this, we observed an increase in global inflammatory gene expression in response to palmitate exposure. This detrimental effect was blunted in hepatocytes incubated with Totum-448-enriched serum. This observation must be tempered by the fact that the induction of the expression of CXCL1, IL-1β, IL-6, MCP-1 and TNFα did not reach significance when analyzed separately, suggesting a global effect on inflammatory status rather than a specific effect on a target gene. Consistently, in a recent randomized, controlled clinical study, Guo et al. demonstrated that NLRP3 inflammasome activation is significantly increased in patients with NAFLD, but that this activation can be strongly repressed by anthocyanins provided here by blueberry extract (Vaccinium myrtillus L.) ⁴⁰. Several polyphenols present in the enriched serum have already demonstrated antioxidant properties through scavenging of free radicals or inhibition of oxidative enzymes ^{41 42} including oleuropein in a similar HepG2 model ⁴³. Preclinically, olive leaf extracts have shown anti-inflammatory, antioxidant, and antifibrotic effects ⁴⁴. Finally, a 2021 randomized double-blind clinical study shows that supplementation with olive leaf extract (Olea europaea L.), characterized by its oleuropein and hydroxytyrosol content, shows a reduction in oxidized LDL ⁴⁵.

Hepatocytes are also responsible for de novo lipids biogenesis along with lipid metabolism as well. Thus, the role of both, the smooth and the rough endoplasmic reticulum (ER), is pivotal. In light of this, dysregulation of lipid metabolism, increased inflammation or oxidative stress may either be a starting point or a consequence of the unbalance of the ER metabolism. This drives the establishment of a vicious cycle triggering the UPR (Unfold Protein Response) and leading to the activation of IRE1, PKR-like ER kinase (PERK) and ATF-6 to reset ER homeostasis ⁴⁶. Of interest, PERK activation results in an antioxidant response ^47,48 while both PERK and ATF-6 participate to the recovery from liver’s condition as demonstrated by knockout studies ^49–51

In this study, the influence of the TOTUM-448 metabolites on ER-stress management seems to be the cornerstone. Palmitate enhanced both XBP-1 ^52,53 and CHOP ⁵⁴ expressions and significantly stimulates ATF-6 activity. XBP-1 and CHOP up-regulation result from the activation of IRE1 and PKR-like ER kinase (PERK), respectively. This up-regulation is fully consistent with literature data. Association between chronic ER and steatotic liver was previously established in both obese mouse models (genetic and high-fat diet models) and obese patients ⁵⁵. Palmitate also significantly enhanced caspase-3 activity. Such an enhancement was previously described in liver biopsies from patients suffering from NASH ⁵⁶. In rats fed a high fat diet, CHOP expression and active form of caspase-3 were found increased in liver ⁵⁷. The presence of the human metabolites from TOTUM-448 potently limited this palmitate-induced ER stress and normalized the studied markers (-36%, p ≤ 0.01; -66%, p ≤ 0.0001 and − 31%, p ≤ 0.01, for CHOP, XBP1 and Caspase-3, respectively; a trend was observed for ATF-6 (-29%, p = 0.2)) supporting the relevance of this nutritional strategy to manage MASLD. The protective effects of the serum enriched with TOTUM-448-derived metabolites against palmitate-induced endoplasmic reticulum (ER) stress may, at least in part, be attributed to the action of specific polyphenols known to be present in the plant extracts composing TOTUM-448. Several of these compounds have already been reported to mitigate ER stress. For example, hydroxytyrosol, a major compound of olive leaf extract, has been shown to protect HepG2 cells from tunicamycin-induced ER stress ⁵⁸. Similarly, luteolin, found in artichoke and Chrysanthellum extracts, was reported to prevent palmitate-induced ER stress, autophagy and apoptosis in AML12 hepatocytes ⁵⁹. In addition, chlorogenic acid, which is present in artichoke leaf extract, demonstrated protective effects against palmitate-induced ER stress in primary rat hepatocytes ⁶⁰.

Inflammation, oxidative stress and ER stress being widely imbricated ^61,62 we are facing a chicken-and-egg situation and, the question regarding the way the TOTUM-448’s metabolites support hepatoprotection remains. Do they improve hepatocyte behavior and lipid metabolism by preserving ER homeostasis or do they limit ER stress by improving lipid metabolism?

Indeed, it has been reported that caloric restriction improves steatosis and alleviates ER stress ⁶³ while normalization of ER stress markers was correlated with reduced hepatic lipid content ⁶⁴. ER is the major site of lipid synthesis in hepatocytes including triglyceride synthesis and storage. Both, the sterol regulatory element-binding proteins (SREBP1c for fatty acid synthesis and SREBP2 for sterol synthesis) and the diacylglycerol acyltransferase (DGAT) are ER-localized transcription factors and enzymes, respectively, involved in the regulation and the catalysis of de novo lipogenesis ⁴⁶. Accordingly, TOTUM-448’s metabolites were found to reduce SREBP1c and DGAT2 expression to blunt the vicious cycle of MASLD and promote hepatocyte homeostasis. Notably, the metabolites had no effect on their own but showed protective effects in the presence of palmitate, supporting the preventive role of this ingredient in maintaining hepatocyte function and homeostasis (Fig. 6).

The study was conducted in accordance with the Declaration of Helsinki of 1975 (https://www.wma.net/what-we-do/medical-ethics/declaration-of-helsinki (accessed on November, 12th 2021)) revised in 2013. The human study was approved by the French Ethical Committee (NCT06047847/ N° ID RCB: 2023-A00579-36/ Comité de Protection des Personnes CPP Est III; approved 25th July 2023). The volunteers were informed of the objectives and the potential risks of the present study and provided their written informed consent before they participated in the study.