Fructose: a modulator of intestinal barrier function and hepatic health?

The idea that the intake of certain macronutrients like sugar may be linked to and even trigger the development of diseases is not a new concept. Already at the beginning of the last century, Emerson and Larimore reported a strong association of dietary intake of refined sugar with diabetes [1]. In line with these findings, in 1929, Banting also hypothesized that the intake of refined sugar may be a major cause of adult-onset diabetes [2]. However, similar to the ongoing discussion nowadays (also see [3, 4]), others suggested that the increase in obesity and the prevalence of diabetes type 2 may simply result from overnutrition [5, 6]. Following this hypothesis which is based on the first law of thermodynamics, weight gain would simply result from the imbalance of energy ingestion vs. energy expenditure. Supporting the latter, at the time food became more plentiful and easier to obtain. Along with the introduction of new inventions like elevators and cars but also machines in industrial production leading to an easier ‘avoidance’ of physical activity, lifestyle started to change dramatically. On the other hand, results from epidemiological [7‐10] and intervention studies [11‐14] as well as animal experiments [15, 16] provide evidence that the consumption of certain foods like sugar-sweetened beverages may impact the development of overweight and metabolic abnormalities like hypertension, insulin resistance, dyslipidemia, and non-alcoholic fatty liver disease (NAFLD) irrespective of overnutrition. Some more recent studies reported that the intake of fructose enriched diets may even in the absence of marked weight gain be causative in the development of metabolic abnormalities [14, 17]. Results of controlled intervention trials also suggest that glucose and fructose may not only differently affect glucose and insulin release, but may also affect gut microbiota composition and intestinal barrier function as well as the release of gastrointestinal hormones [18‐23]. Some of the so far accumulated data are rather contradictory; still, all in all, data so far fuel the assumption that different sugars may impact health differently. Starting from this background, the aim of the present narrative review is to provide some insights in dietary intake and metabolism of fructose and to summarize recent as well as older findings assessing the effects of fructose on intestinal microbiota, homeostasis, and barrier function. Furthermore, these findings are related to the development of metabolic diseases with a specific focus on the liver.

Fructose is naturally found in fruits, vegetables (e.g., in peppers or carrots), or honey. In addition, added sugars, like sucrose or high-fructose corn syrup (HFCS), also contribute markedly to the dietary intake of fructose [24] (for overview see Table 1). Nowadays, HFCS is being used in many countries as a replacement of sucrose in foods and beverages such as soft drinks, sweets, bakery goods, and dairy products. To produce HFCS, some of the glucose found in corn syrup is enzymatically converted to fructose [23, 25]. With implementing the latter technique in the USA in the 1970s, sucrose intake markedly decreased in the USA, while concurrently HFCS consumption increased resulting at first in a rather stable overall sugar intake [25]. Results of epidemiological studies suggest that from 2001 to 2021 the total daily intake of HFCS decreased by almost 40 % from 46 g/d to 28 g/d in the USA (data from US Department of Agriculture) [26]. Despite changes in the sugar market in 2017 [27], in most European countries, HFCS is not widely used and the intake of HFCS is, compared to the USA, rather low [28]. Indeed, sucrose is still the main added sugar found in foods and beverages in Europe [29]. And while there were also some decreases in average sugar intake in several European countries as well as in South- and Central American countries and in Australia, added sugar intake in most of these regions is still above the recommendations of the World Health Organization (WHO; [30]; energy derived from added sugar intake <10 % of total energy intake accounting to ~50 g of sugar in a 2,000 kcal diet [31]). For instance, surveys conducted in European and Latin American countries report that total sugar intake (= sucrose intake) in adults accounts for ~15–21 % of the total energy intake [32]. Results of studies also suggest that socio-economic status and the intake of added sugars negatively correlate with lower income households shown to have a markedly higher sugar intake [33].

Studies assessing fructose rather than sugar intake are rather limited. Results of our own studies and those of others suggest that the average fructose intake of healthy middle-aged individuals in Germany and Austria ranges from ~40 g to ~49 g fructose/d ([14, 34‐36] and unpublished data). Data from the Dutch National Food Consumption Survey 2007–2010 also revealed a similar dietary daily fructose intake in Dutch population (7–69 years, average fructose intake ~46 g/d with 67 % consumed as sucrose and 33 % as free fructose with the main food sources being soft drinks, juices, cakes and cookies) [37]. Studies from Brasilia suggest an average intake of 32 g (for women) to 35 g (for men) of total fructose per day mainly consumed as soft drinks or fruit juice. In Lebanon, daily intake of fructose seems to be on average 51 g/d. In this study, it was also estimated that the intake of added fructose stemming from HFCS was three times higher than the intake of naturally occurring fructose in fruits and vegetables [38, 39]. Taken together, these data suggest that free sugar intake in general and free fructose intake in particular are still markedly higher in wide parts of the general population than recommended in many countries world-wide.

As fructose and glucose uptake and metabolism are quite different and as these differences are discussed to be at least in part contributing to the different (patho-) physiological effects of these two monosaccharides, in the following, some of the key differences regarding the uptake and the metabolism of fructose are summarized.

Despite intense research efforts throughout the last decade, mechanisms underlying the metabolic alterations associated with the intake of elevated amounts of fructose, and especially, the development of insulin resistance and associated diseases like NAFLD are not yet fully understood. As detailed above, due to its unregulated metabolism, fructose may be converted to various metabolites like diacylglycerol rather fast. Results of some animal studies suggest that hepatic insulin resistance and steatosis may result from an increase in hepatic diacylglycerol accumulation, and an associated protein kinase C activation leading to alterations of insulin-mediated Akt activation [64]. However, as discussed above, several studies assessing the uptake and distribution of fructose suggest that marked amounts of fructose may already be metabolized in the small intestine ([53] and Fig. 1). Accordingly, other (additional) mechanisms may be involved in the development of insulin resistance associated with an elevated fructose intake and NAFLD. In support of this assumption, results of animal studies of our own group have shown that a chronic intake of a fructose enriched drinking solution (30 % of fructose content) and the resulting development of fatty liver are associated with increased bacterial endotoxin levels in portal vein and an induction of its receptor, Toll-like receptor 4 (TLR4), in the liver as well as its subsequent signaling cascade [15]. In the same study, it was also shown that alterations alike are not present when animals are fed a 30 % glucose solution. Furthermore, the concomitant treatment of mice with non-resorbable antibiotics like polymyxin B and neomycin abolished the increase in bacterial endotoxin levels in portal vein being also associated with a diminishment of the development of liver steatosis and inflammatory alterations in mice [15]. Bacterial endotoxin and an activation of TLR4 signaling have also been shown to contribute to the development of insulin resistance [65, 66]. Supporting the hypothesis that an elevated fructose intake may impair intestinal barrier function, resulting in an increased translocation of bacterial endotoxin, mice lacking a functional TLR4 were found to be significantly protected from fructose-induced liver damage [67]. Also, targeting alterations of intestinal barrier function with drugs like metformin or bile acids and probiotics (also see below), respectively, have been reported by us and others not only to be associated with ‘normalized’ tight junction protein levels in small intestine but also with a lessening of the development of NAFLD and a normalization of markers of insulin resistance in liver tissue [68‐72]. Interestingly, in our studies only limited or no effects on tight junction proteins were found in colon [73]. These findings suggest that alterations associated with the intake of elevated amounts of fructose may not only result from changes of intestinal microbiota composition but also may result from direct effects of fructose, e.g., its metabolism, in enterocytes. In line with these findings, Guo et al. showed that the chronic intake of fructose in piglets decreased the expression of tight junction proteins and myosin light chain kinase (MLCK) in ileal tissue [74]. Wagnerberger et al. further reported that not only TLR4 is induced in livers of mice fed a fructose-rich diet. Rather, in this study it was shown that the expressions of other TLRs including TLR1, 2, 3, and 6-8 mRNA were all significantly higher in mice fed a fructose-rich diet than in controls. Furthermore, expression of TLRs in the liver was almost completely abolished when fructose-fed mice were concomitantly treated with the non-resorbable antibiotics polymyxin B and neomycin. Interestingly, while the disruption of TLR4 signaling or the treatment with non-resorbable antibiotics was associated with a reduction in markers of insulin resistance and inflammation, e.g., number of F4/80 positive cells, inducible nitric oxide synthase (iNOS), and tumor necrosis factor alpha (TNFα) expression, fat accumulation in liver tissue was only reduced by ~50–60 % [73]. These results further suggest that some of the fructose may reach the liver and may through its insulin-independent metabolism be quickly converted to triglycerides.

In line with the above summarized findings in animal studies, it has been shown in human studies that a three day long elevated intake of fructose (25 % of total energy as fructose) is associated with increased bacterial endotoxin levels and an induction of TLR2 and 4 mRNA expression in peripheral blood mononuclear cells in healthy volunteers [14]. In the same study, similar alterations were not found when the same subjects consumed comparable amounts of glucose for three days. In summary, these findings suggest that both intermediates of the hepatic fructose metabolism and pathogen-associated molecular patterns (PAMPs) and -dependent signaling cascades may contribute to the development of fructose-associated insulin resistance and the development of NAFLD (see Fig. 2). However, further studies are needed to determine doses and to further delineate molecular mechanisms. Some of the so far defined possible mechanisms underlying the increased translocation of PAMPs are highlighted in the following.

As detailed above, results of several studies suggest that an intake of large amounts of fructose can alter intestinal microbiota composition and barrier function. These results tempt the assumption that a manipulation of intestinal microbiota composition through the supplementation of probiotics may alter the fructose-induced alterations. Indeed, as detailed in the following, results of several studies suggest that a concomitant intake of probiotics may attenuate at least some of the effects of fructose. For instance, the concomitant treatment of fructose-fed mice with the probiotic Lactobacillus casei Shirota significantly diminished liver damage compared to mice only fed fructose [68]; interestingly, this was not associated with a protection from the loss of tight junction proteins or changes of intestinal microbiota composition in small intestine. Ritze et al. and Zhao et al. both reported that a concomitant intake of the probiotic Lactobacillus rhamnosus GG markedly attenuated the development of fructose-induced liver damage in mice [95, 96]. The protective effect of these probiotics was associated with a protection against the loss of tight junction proteins and impairments of intestinal barrier function. Also, Lactobacillus rhamnosus GG has been reported to ‘revert’ intestinal dysbiosis and decreased the relative abundance of inflammation-related bacteria such as Desulfovibrionaceae, Clostridia, and Proteobacteria in feces of fructose-fed animals [81]. The probiotic strains Lactobacillus plantarum ATG-K2 and ATG-K6 but also Lactobacillus plantarum strains such as NA136 have also been shown to ameliorate the induction of pro-inflammatory markers like TNFα and interleukin-6 as well as markers of lipogenesis, e.g., fatty acid synthase and sterol regulatory element-binding protein 1c in small intestinal tissue of rodents fed a fat- and fructose-rich diet [97, 98]. Furthermore, in the study of Wang et al. the concomitant treatment of mice fed a fat- and fructose-rich diet with the probiotics L. rhamnosus LS-8 and L. crustorum MN047 attenuated the development of NAFLD, insulin resistance, and decreased circulating lipopolysaccharides (LPS) levels [80]. It has also been shown that treating high fructose diet-fed mice with the bacterial strain called Lactobacillus brevis DM9218 resulted in an improved intestinal barrier function combined with a reduction in LPS in the liver [99]. It has further been reported that the administration of the probiotic Lactobacillus kefiri and Lactobacillus fermentum CECT5716 may diminish the effects of high fructose-induced dysbiosis [100, 101].

Taken together, these data further suggest that some probiotics may attenuate or at least diminish fructose-induced alterations of intestinal barrier function, and subsequently, the development of metabolic diseases like NAFLD. However, the above summarized data also suggest that not all beneficial effects on the development of fructose inflicted NAFLD found for probiotics may be related to an improved intestinal barrier function. Rather, results so far suggest that depending on the bacterial strain some of the probiotics might—probably through metabolites—alter liver metabolism directly while others seem, either through direct interaction or yet to be defined metabolites/mechanisms, to alter metabolism or signaling cascades within the enterocytes. Further studies are at need to determine mechanisms and metabolites involved.

High fructose consumption, be it through food and beverages sweetened with sucrose or HFCS, may not only lead to the development of overweight but may also contribute to the development of other metabolic diseases like diabetes type 2 and NAFLD. Despite intense research effort, mechanisms underlying these alterations associated with fructose and their contribution to the development of the latter disease are not yet fully understood. Studies suggest that fructose may affect intracellular signaling through direct measures due to insulin-independent metabolism [40]. However, in recent years, it has also been shown in various species that a high and chronic intake of fructose can be associated with changes in fecal microbiota but also a loss of tight junction proteins in small intestine. A causal link between the changes in fecal microbiota composition and the loss of tight junctions in small intestine could be established based on the findings employing certain probiotics like Lactobacillus rhamnosus GG and Lactobacillus brevis DM9218 that have been shown to attenuate the development of intestinal barrier dysfunction [96, 99]. However, results of recent other studies also suggest that fructose may add to intestinal barrier dysfunction and the subsequent translocation of PAMPs like bacterial endotoxin through more direct interaction with the enterocytes and their metabolism. Indeed, it has been shown that fructose metabolism directly disturbs intracellular NO homeostasis, e.g., leading to a loss of tight junction proteins ([84, 88] and see Fig. 2).

Further studies are needed to determine mechanisms and doses necessary to inflict these alterations. Also, studies are needed to elucidate the impact of intestinal microbiota and the interaction with other compounds found in foods and beverages like polyphenols herein. It remains yet to be determined if and how specific (probiotic) bacterial strains can alter or even attenuate these alterations. As most of the results showing an effect of fructose on intestinal microbiota composition and barrier function were obtained in in vitro and animal studies, studies in humans are needed in the future to determine if alterations alike are also found in humans and if so, whether these alterations contribute to the development of intestinal barrier dysfunction, and subsequently, the development of metabolic diseases like NAFLD and diabetes type 2. Moreover, whether sex- or age-specific differences have an impact on alterations associated with high fructose intake has not yet been determined to our knowledge. Despite these open questions, and especially when taken the high intake of sugar and fructose found in some populations into consideration, a restrictions of sugar intake, and herein, especially fructose intake may already now be a way to lessen intestinal barrier dysfunction and the development and progression of metabolic diseases associated with altered intestinal barrier function like NAFLD. Indeed, in recent years, guidelines have been issued and different political measures including restrictions of the promotion for sugar-rich products like sweets and sugar-sweetened beverages, the introduction of progressive taxes on sugary drinks and foods, and restrictions of the availability of specific foods have been established in many countries all aiming to reduce sugar intake [102]. Also, changing food choice settings, e.g., removing sugary snacks and beverages from near tills to reduce impulse buying and improving nutritional literacy especially in children, adolescents, and young adults through new didactic tools in school settings may also add to reduce overall sugar intake, and thereby, positively impacting on health outcomes in the general population.