@article{d1ea9986723340cd8ec2d508ccb3b5ac,
title = "“Sweet death”: Fructose as a metabolic toxin that targets the gut-liver axis",
abstract = "Glucose and fructose are closely related simple sugars, but fructose has been associated more closely with metabolic disease. Until the 1960s, the major dietary source of fructose was fruit, but subsequently, high-fructose corn syrup (HFCS) became a dominant component of the Western diet. The exponential increase in HFCS consumption correlates with the increased incidence of obesity and type 2 diabetes mellitus, but the mechanistic link between these metabolic diseases and fructose remains tenuous. Although dietary fructose was thought to be metabolized exclusively in the liver, evidence has emerged that it is also metabolized in the small intestine and leads to intestinal epithelial barrier deterioration. Along with the clinical manifestations of hereditary fructose intolerance, these findings suggest that, along with the direct effect of fructose on liver metabolism, the gut-liver axis plays a key role in fructose metabolism and pathology. Here, we summarize recent studies on fructose biology and pathology and discuss new opportunities for prevention and treatment of diseases associated with high-fructose consumption.",
keywords = "cancer, fructos, gut inflammation, metabolic disease, NASH",
author = "Febbraio, {Mark A.} and Michael Karin",
note = "Funding Information: M.A.F. is a Senior Principal Research Fellow at the NHMRC ( APP1116936 ) and is also supported by an NHMRC Investigator Grant ( APP1194141 ). Research in his laboratory was supported by project grants from the NHMRC ( APP1042465 , APP1041760 , and APP1156511 to M.A.F. and APP1122227 to M.A.F. and M.K.). M.K. is an American Cancer Research Society Professor and holds the Ben and Wanda Hildyard Chair for Mitochondrial and Metabolic Diseases. His research was supported by grants from the NIH ( P42ES010337 , R01DK120714 , R01CA198103 , R37AI043477 , R01CA211794 , and R01CA234128 ). Funding Information: In a recent study conducted by Wellen, Rabinowitz, and coworkers (Zhao et al., 2020), ACLY was specifically ablated in mouse hepatocytes. Using in vivo isotopic tracing, the authors found that this genetic manipulation did not prevent fructose-induced hepatosteatosis, although it should be noted that, in this study, the high-fructose diet did not markedly increase hepatic triglyceride content even in wild-type control mice (Zhao et al., 2020). The same authors previously observed that fructose is converted to acetate by the microbiota (Jang et al., 2018) and that acetate can generate acetyl-CoA independent of ACLY (Zhao et al., 2016). Accordingly, they tested the hypothesis that fructose-induced lipogenesis in the liver is driven by microbiota-derived acetate. They observed that the depletion of microbiota markedly suppressed the conversion of fructose into acetyl-CoA in the liver and hepatosteatosis (Zhao et al., 2020), suggesting that fructose metabolism in the GI tract may control hepatosteatosis (Figure 3). However, as discussed below, any experiment based on bulk depletion of the microbiota needs to be carefully interpreted because of the major role of Gram-negative gut bacteria in the generation of endotoxin (LPS), which seems to be a key player in fructose-induced hepatosteatosis (Todoric et al., 2020). Moreover, all bacteria release bacterial nucleic acids that further enhance hepatic inflammation. In a recent study (Jang et al., 2020), the Rabinowitz group performed intestinal-specific KHK-C (the more active KHK isozyme) loss- and gain-of-function experiments. They demonstrated that, on the one hand, KHK-C deletion increased fructose delivery to the liver and that the microbiota promoted hepatosteatosis (Jang et al., 2020). On the other hand, KHK-C overexpression decreased fructose-induced lipogenesis (Jang et al., 2020). The authors concluded that metabolism of fructose in the gut shields the liver from fructose-induced fatty liver (Figure 3). Arriving at the same general conclusion, but via a different mechanism, our team found that fructose-induced hepatosteatosis is controlled by the intestinal epithelial barrier via the gut-liver axis (Todoric et al., 2020). Consistent with previous reports in animals (Cho et al., 2021; Kavanagh et al., 2013; Spruss et al., 2012) and humans (Jin et al., 2014), we found that excessive fructose consumption resulted in barrier deterioration, dysbiosis, low-grade intestinal inflammation, and endotoxemia (Todoric et al., 2020). Although we attributed barrier deterioration to KHK-dependent conversion of fructose to F1P in enterocytes, the protective effect of intestinal KHK-C ablation suggests that fructose-induced microbial dysbiosis may be the primary driver of barrier deterioration. Indeed, microbial depletion with antibiotics leads to a partial reversal of fructose-induced barrier deterioration (Todoric et al., 2020). Using RNA sequencing, we confirmed the presence of an endotoxin-induced transcriptional signature defined by the marked upregulation of toll-like receptors (TLR) 2,3,4,6,7, and 8, and their adaptor protein MyD88, and the induction of inflammatory chemokines and cytokines, such as CCL2, CCL5, and TNF, in the livers of fructose-fed mice (Todoric et al., 2020). Similarly, Spruss and coworkers found that TLR4-deficient mice were protected from fructose-induced NAFLD (Spruss et al., 2009). Using several different approaches, including intestinal-specific expression of the antimicrobial protein Reg3b and myeloid-specific MyD88 ablation, we provided further support for the role of endotoxin and other microbial-generated inflammatory signals in the enhancement of fructose-induced DNL and hepatosteatosis (Todoric et al., 2020). This occurs through a multicomponent pathway consisting of recruited hepatic macrophages that produce TNF (Figure 3), which, by engaging its type 1 receptor (TNFR1) on hepatocytes, leads to the induction of the critical lipogenic enzymes ACLY, ACC1, and FASN, which convert fructose-derived acetyl-CoA to C16 and C18 fatty acids (Todoric et al., 2020). Consistent with our previous demonstration that TNFR1 signaling blockade prevents NASH (Febbraio et al., 2019; Kim et al., 2018), we found that incubation of human hepatocytes with TNF also results in the induction of lipogenic enzyme mRNAs and the conversion of either fructose or glucose to lipid droplets (Todoric et al., 2020). Critically, fructose, but not cornstarch (glucose), isocaloric feeding led to the downregulation of enterocyte tight-junction proteins and subsequent barrier deterioration, which is in agreement with previous rodents and human studies (Jin et al., 2014; Kavanagh et al., 2013; Lambertz et al., 2017; Spruss et al., 2012). In the past (Taniguchi et al., 2015), we found that enterocyte IL-6 signaling stimulates epithelial cell proliferation through the activation of Yes-associated protein (YAP), thereby conferring resistance to mucosal erosion. To test whether YAP activation can prevent fructose-induced inflammation, hepatosteatosis, and NASH, we expressed an activated form of IL-6 signal transducer (IL6ST), also known as gp130, exclusively in enterocytes, or injected fructose-fed mice with the YAP-induced matricellular protein cellular communication network factor 1 (CCN1). Both manipulations prevented the downregulation of tight-junction proteins, endotoxemia, and ameliorated fructose-induced hepatosteatosis, and NASH (Todoric et al., 2020). Although most of the studies mentioned earlier confirm the importance of the enterocyte in regulation of liver fructose metabolism, due to different experimental conditions, they arrive at seemingly different conclusions. Rabinowitz and colleagues (Jang et al., 2020), using comparatively low amounts of fructose, found that enterocyte-specific KHK-C deficiency was not protective, but hepatic steatosis was quite low in these experiments, as also observed previously (Zhao et al., 2020). Using higher amounts of fructose that cause barrier deterioration, we suggested that accumulation of toxic F1P within enterocytes may initiate the inflammatory cascade underlying hepatic steatosis and predicted that KHK inhibition should be protective (Todoric et al., 2020), which has been subsequently demonstrated (Gutierrez et al., 2021). However, as discussed earlier, the protective effect of KHK inhibition is mainly manifested in the liver and instead of F1P-induced toxicity, barrier deterioration is probably due to dysbiosis. Notably, fructose-induced inflammation is only observed after prolonged exposure, and its magnitude may depend on the animal facility, a variable that profoundly affects the microbiota (Ussar et al., 2015). Indeed, given the critical pathogenic role of barrier deterioration and endotoxemia, there is little doubt that the microbiota is a key contributor to fructose-induced liver disease (Jadhav and Cohen, 2020).M.A.F. is a Senior Principal Research Fellow at the NHMRC (APP1116936) and is also supported by an NHMRC Investigator Grant (APP1194141). Research in his laboratory was supported by project grants from the NHMRC (APP1042465, APP1041760, and APP1156511 to M.A.F. and APP1122227 to M.A.F. and M.K.). M.K. is an American Cancer Research Society Professor and holds the Ben and Wanda Hildyard Chair for Mitochondrial and Metabolic Diseases. His research was supported by grants from the NIH (P42ES010337, R01DK120714, R01CA198103, R37AI043477, R01CA211794, and R01CA234128). The two authors contributed equally to the writing and revision of the article. M.A.F. prepared the draft figures that were used for the final images. M.K. holds US Patent No. 10034462 B2 on the use of MUP-uPA mice for the study of NASH and NASH-driven HCC. M.A.F. is a co-inventor of IC7Fc and hold patents for this molecule (US 60/920,822; WO/2008/119110 A1). Publisher Copyright: {\textcopyright} 2021",
year = "2021",
month = dec,
day = "7",
doi = "10.1016/j.cmet.2021.09.004",
language = "English",
volume = "33",
pages = "2316--2328",
journal = "Cell Metabolism",
issn = "1550-4131",
publisher = "Elsevier",
number = "12",
}