University of Groningen
Computational modeling of cholesterol metabolism
Paalvast, Thijs
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Paalvast, T. (2019). Computational modeling of cholesterol metabolism. University of Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
271
Chapter 8
General Discussion
This thesis uses computational modeling to better understand how an imbalance between energy intake and expenditure leads to metabolic syndrome as well as investigates what perturbations are most likely to reverse the detrimental effects on cardiovascular risk. Pending progress in society on how individualism is best reconciled with the inability of individuals to resist forces conspiring to turn them obese, research like this may contribute to finding pharmacological adjuvants to achieve a healthy body weight.
Chapter 2
In Chapter 2, current models of cholesterol metabolism were reviewed [1].
By systematically evaluating key assumptions for their biological validity and considering how the assumptions may affect modeling results, we made sure to obtain a good understanding of the model. Moreover, by reproducing every model and applying a functional test to these models, we could also establish whether the model was sophisticated enough to show some general applicability. Passing such a functional test would make the model more attractive for use as a module in a larger computational model. The functional test consisted of mimicking the effect of statins, which for most of the models meant decreasing cholesterol synthesis and evaluating whether plasma LDL-C would decrease within the expected time-frame. It was then found that actually most models were not able to pass this functional test, making us conclude that the field as a whole is still in its infancy.
In hindsight however, there are some concerns over the ‘validity’ of our functional test, since while intensively studied, not all is known on the effect of statins. More specifically, we assumed that since statins inhibit cholesterol synthesis in vitro, this would also occur in vivo. At the time of our study, while some groups had reported decreased cholesterol synthesis on statins based on ratios of precursor metabolites, this was controversial since using either cholesterol balance or stable-isotope methods did not result in the same finding [2–4]. Interestingly, recent work in our group demonstrates that paradoxically, cholesterol synthesis may be enhanced in response to statins, at least in mice [5].
This emphasizes that there is a subtle interplay between experimental findings, the process of modeling and how the two are evaluated. Specifically, when a model is able to reproduce some phenomenon it does not necessarily do so because the model accurately emulates the biological processes culminating in the phenomenon under study, but just that the knowledge that the model is based on is free from conflicting
272
hypotheses. Conversely, when a model is not able to reproduce an experimental finding, it may demonstrate that there are inconsistencies within the current theoretical framework, and expose assumptions that may require revision. Thus, in a way, where experimental findings can be regarded as tests of individual hypotheses, modeling may be regarded as a test of the theoretical framework surrounding these hypotheses. In this regard, one may question the practice of demanding that computational models be validated by experimental findings, yet do not expect or demand theories accompanying experimental findings to be backed up by a computational model.
Chapter 3
In chapter 3 we presented a model of lipoprotein metabolism with the aim to better understand what explains the often observed inverse correlation between HDL and plasma TG levels. The agent-based model approach we take has the advantage that properties of individual lipoproteins can be taken into account when modeling lipid exchange between lipoproteins. Plasma TG induces lower HDL through increased clearance of HDL through CETP-mediated exchange of HDL-CE for VLDL-derived TG, subsequent hydrolysis of TG within HDL, in turn leading to faster clearance of the resulting smaller HDL through the kidney. Another interaction between plasma TG and HDL, is that during lipolysis of VLDL excess surface lipids are transferred to HDL, where they may be a substrate for LCAT, thus producing larger HDL, resulting in higher HDL levels. Incorporating these mechanisms in the model, we found that a high activity of the latter mechanism is incompatible with the common finding of tracer-kinetic studies, that high VLDL-production rates induce increased clearance and thus lower levels of HDL [6,7]. In fact, if it is assumed that the majority of excess surface lipids are a substrate for LCAT on HDL, then higher VLDL-production rates lead to increased levels of HDL. The implication of this finding is that the majority of excess surface lipids may not be cleared through HDL, but through some other as yet unknown route.
Though it is currently unclear what this route should be, it emphasizes that modeling can help uncover inconsistencies in a theoretical framework. Furthermore, we hope this work will inspire the search for this pathway.
Chapter 4
While it is known that metabolic syndrome develops slowly over time, most studies are not longitudinal in setup. While the reason for this is obvious, longitudinal studies in humans are incredibly expensive and a relevant time frame easily exceeds the time span of a PhD, important aspects of metabolic syndrome may be missed this way. In chapter 4 we presented an in-depth study of the response of APOE*3Leiden.CETP mice on a Western diet in order to find out whether these mice with a humanized lipid
273
profile, may act as substitute for humans in longitudinal studies, so that qualitatively similar information is gathered, however in a much shorter time frame. It was then found that there are indeed similarities in the response of these mice compared to humans in cross-sectional studies. Specifically, body weight and plasma TG decrease later in life in men, and a decrease in dyslipidemia was found in male APOE*3Leiden.CETP mice as well, whereas body weight plateaued. We however also observed a decrease in insulin resistance in these mice through time, while this is not observed in men, where insulin resistance generally increases throughout life [8]. We further observed a close relation between body weight and plasma TG, where plasma TG sharply increased when a certain body weight was reached. At the same time however, we observed that this relation altered through time, being less distinct at older age. Similarly, while at young age there was a clear relation between body weight and insulin resistance, this was no longer the case at older age. It is unclear why insulin resistance and plasma TG decreased through time in APOE*3Leiden.CETP mice while still on the Western diet. Current theories on the pathophysiology of metabolic syndrome hold that both insulin resistance and increased levels of plasma TG occur upon exceeding the storage capacity of fat and the leaking of fatty acids from adipose tissue [9]. The fact that this relation disappears in mice when on prolonged high-fat diet suggests that changes in energy expenditure and the handling of fatty acids may occur, and that either the time of onset or the magnitude of these changes vary between individuals. Apart from the changes of dyslipidemia through time, we also observed a marked interindividual heterogeneity in body weight and dyslipidemia between animals. While heterogeneity had also been reported in the background of the APOE*3L.CETP mice [10,11], there were little other clues towards why this heterogeneity occurs. Therefore, we decided to elucidate what factors may contribute to this heterogeneity using computational modeling, which is described in the next chapter.
Chapter 5
In chapter 5 the experimental data described in chapter 4 is used for computational modeling, so that we may learn about the factors behind the heterogeneity between animals, in the hope of finding clues to what makes humans either prone or non-prone to develop metabolic syndrome. We divided animals into a non-responder and responder group by using a cut-off value of 1.0 mM for plasma TG, so that animals with low plasma TG would be non-responders, and animals with higher plasma TG responders. Because plasma TG correlates with body weight and insulin resistance, this procedure may also be regarded as separating animals with and without the characteristics of metabolic syndrome. We then used Analysis of Dynamic Alterations in Parameter Trajectories (ADAPT), a methodology that fits the parameters of a model
274
of ordinary differential equations to the experimental data but allows for parameters of a model to gradually change over time. In this way, predictions are obtained for how the corresponding process may evolve through time.
ADAPT predicted that non-responders would have higher energy expenditure, decreased cholesterol absorption and increased fecal fat excretion. Interestingly, we could not find evidence for either decreased cholesterol absorption or increased energy expenditure in non-responders. It must be noted that the absence of differences between responders and non-responders in energy expenditure according to indirect calorimetry, may have been due to the detection limits inherent in the methodology. The prediction of changes in cholesterol absorption however were tested in two independent validation studies, suggesting that either the prediction was false or was only present in the particular cohort whose data was used for ADAPT. The final prediction of ADAPT however, that responders and non-responders had marked differences in fecal fat excretion, was not only validated in the original cohort, but also replicated in both the studies in which cholesterol absorption was measured. Interestingly, these changes in fecal lipid excretion were associated with a change in fecal bile acid profile, where animals that showed a higher hydrophobicity index of fecal bile acids showed less fat excretion. Specifically, fecal fatty acid excretion was highly negatively correlated with the fecal excretion of deoxycholic acid (DCA). Thus we found strong indications that perturbing bile acid homeostasis may be a useful strategy in the prevention of obesity.
Intermezzo
While part of the heterogeneity observed in the APOE*3Leiden.CETP mouse may be related to bile acid metabolism, it remains unclear whether this effect is causal. Recently, Tarasco et al. reported that non-responders present with a distinct liver morphology, with livers that are generally smaller, have a nodular surface and frequently have pre-neoplastic formations [12]. While our experimental setup did not include a systematical evaluation of the macroscopic appearance of the liver, we encountered non-responders with a cirrhotic appearance of the liver as well. Furthermore, liver histology revealed (pre-)neoplastic formations in two out of nine non-responders of which histology was obtained.
Tarasco et al. further found that non-responders presented with both higher plasma bile acid levels and lower fecal bile acid concentrations [12]. This is in agreement with our findings, that fecal bile acid output of non-responders was low compared to responders. Tarasco et al. postulated that the differences observed in bile acid metabolism may be due to the observed liver abnormalities [12]. Indeed, liver cirrhosis is associated with decreased bile acid production, especially cholate, and increased bile acid re-uptake, thus supporting this hypothesis. Interestingly, liver cirrhosis is also marked by a
275
decrease in fecal excretion of DCA [13]. This is thought to be the result of decreased biliary secretion of cholic acid, resulting in a decrease in colonic microbiota specialized in converting CA into DCA (Clostridium group XIV), and thus decreased fecal DCA excretion [14,15]. Moreover, liver cirrhosis in humans is frequently accompanied by fat malabsorption [13]. Thus, our findings of decreased fecal DCA excretion together with increased fat excretion in non-responder mice are in line with similar observations of liver cirrhosis in humans.
However, does liver cirrhosis also explain the differences in body weight and dyslipidemia between responders and non-responders? It should be noted that although dyslipidemia is a risk factor for developing liver cirrhosis, the actual presence of liver cirrhosis is associated with a decrease in dyslipidemia [16], purportedly through increased liver dysfunction and impaired ability to synthesize lipoproteins. Furthermore, liver cirrhosis is associated with cachexia [17]. While the mechanisms behind cachexia remain poorly understood, a proposed contributor is increased fat oxidation in muscle, triggered through inflammatory signals, and resulting in muscle wasting [17]. Such a mechanism would be in agreement with the prediction of ADAPT that non-responders have increased fat oxidation. All in all, liver cirrhosis is indeed associated with a decrease in dyslipidemia and body weight loss, but the mechanisms through which this is achieved is not entirely clear.
In chapter 5 it was proposed that the difference in dyslipidemia and body weight between responders and non-responders is due to a combination of decreased fat absorption and increased energy expenditure, mediated through changes in bile acid metabolism that are marked by a state of FXR-agonism. The very same mechanisms may contribute to the low plasma TG and cachexia observed in liver cirrhosis.
Obviously, developing NASH or liver cirrhosis in order to decrease dyslipidemia and obesity is not a desirable path. Nonetheless, the mechanisms in liver cirrhosis that mediate the decrease in dyslipidemia and obesity may be emulated through other means (i.e. pharmacologically), thereby achieving the desired mitigation in symptoms of metabolic syndrome without adversely affecting liver health.
Chapter 6
Realizing that bile acid metabolism may be one of the key factors that make responders prone and non-responders non-prone to develop metabolic syndrome, we decided to mimick bile acid metabolism of non-responder mice in responder mice by treatment with the FXR-agonist PX20606, which is selective for the intestine (Chapter 6). Just like we observed in non-responder mice, FXR-agonist treated mice showed an altered bile acid profile and increased fecal fatty acid excretion. Importantly, we observed that upon treatment with FXR-agonist, animals showed an abrupt decrease in plasma TG, TC and body weight. Furthermore, we observed that the decreased plasma TG was not
276
due to increased VLDL-TG production, suggesting increased clearance. However, using radio-labeled VLDL-like emulsion particles, we could not detect an increase in clearance rate in treated animals. In contrast, proteomics of the plasma revealed a relative increase of apoC2 compared to a relative decrease in abundance of apoC1 and apoC3, suggesting that VLDL-TG hydrolysis activity was indeed higher and that the clearance of labeled particles did not sufficiently reflect the clearance of endogenous VLDL.
Another interesting observation in FXR-agonist treated animals was that instead of reversing hepatic steatosis, the location of the liver fat accumulation changed, with more periportal steatosis in treated animals versus more pericentral steatosis in untreated animals. Upon integrating the experimental data into ADAPT, ADAPT predicted that there may be increased flux of intralumenal triglycerides to the liver. We speculate that because there is a shift in the fat uptake from proximal to distal, fat may be packaged either in smaller chylomicrons or be released as free fatty acid, leading to a different distribution of fat [18]. Furthermore, such a difference in tissue distribution in the uptake of postprandial fat may also lead to an altered energy expenditure [19], explaining the decrease in body weight. It should be noted that treatment with the FXR-agonist also resulted in decreased food intake and increased fecal fat excretion, however that these latter two factors do not sufficiently explain the decrease in body weight. Selective FXR-agonists may thus be able to decrease body weight through inducing both decreased energy intake and increasing energy expenditure. If these results prove translatable to humans, there would be a drug that, in effect, bestows non-prone properties to non-prone individuals.
Chapter 7
In chapter 7 recent advances in intestinal lipid absorption are discussed [20]. A major recent development is that trans-intestinal cholesterol excretion has now finally shown to be present in humans and accounts for as much as 35% of total cholesterol excretion [21]. Importantly, it was shown that TICE could be enhanced four-fold by using ezetimibe, showing that the activity of this pathway can be enhanced and targeting this pathway may be a promising strategy in the treatment of patients with cardiovascular disease [21]. Since in mice it has been shown that FXR-agonist treatment increases TICE, and the combination treatment of FXR-agonist and ezetimibe results in a massive increase in TICE [22], it would be interesting to see if the same effect could be achieved in humans. A major bottleneck in the translation of findings in mice to the human situation is that mice have a different bile acid profile. Where the primary bile salts synthesized in humans are cholic acid and chenodeoxycholic acid, mice, in addition to cholic acid, produce muricholic acid [23]. FXR-agonist treatment in mice results in a shift to the production of muricholic acids [22]. It would therefore
277
worthwhile to see whether combinatorial treatment with ezetimibe and FXR-agonist still has such a massive effect on TICE if mice would have a bile acid profile that is more like that observed in humans. Interestingly, the enzyme responsible for the production of muricholic acids in mice was recently identified to be Cyp2c70, bringing this goal within reach [24].
Concluding remarks
The main driver behind symptoms of metabolic syndrome is obesity. Obesity should be cured by promoting healthy food intake and sufficient physical exercise. If these conditions are met, most of the risk-factors that make up metabolic syndrome will disappear. Whatever the reason, reality is that for most people, this is simply not achieved. Since currently no attractive pharmaceutical treatment is available for obesity, there are huge market opportunities for drugs that actually deliver on their promise to reduce body weight. While it is arguably unethical to promote widespread use of drugs that allows people to maintain an unhealthy lifestyle while still achieving some sort of aesthetic ideal, it may also be regarded as ‘leveling the playing field’ for those that are more prone to gain weight than others. While my personaI opinion on this matter is still unclear, pharmaceutical treatment of obesity would be a step forward from the status quo where the only effective therapy for losing weight is bariatric surgery [25–27].
It should be noted that while bariatric surgery may be the most effective treatment for obesity available today, it is not completely understood how this weight loss is achieved. The most important aspect determining weight loss however, is considered to be reduced food intake [28]. This is thought to be achieved through increased nutrient concentrations in the distal intestine, which triggers enhanced secretion of incretins like glucagon-like peptide (GLP)-1 and peptide YY (PYY), in turn enhancing feelings of satiety. One of the contributing mechanisms to enhanced GLP-1 secretion may be increased TGR5-activity in the enteroendocrine L-cells through decreased mixing time of bile acid with ingested food, possibly leading to higher local bile acid concentrations [28]. Furthermore, plasma bile acids and FGF19 are increased after Roux-en-Y gastric bypass (RYGB) surgery [29], suggesting increased FXR-activity. If increased GLP-1 secretion is the main reason for the observed effects, then similar weight loss is expected with GLP-1 mimetics and TGR5-agonists. However, to date GLP-1 mimetics have achieved only modest weight loss [30], whereas selective TGR5-agonists show modest weight loss as well in pre-clinical studies [31,32], however have not yet reached the stage of clinical testing. Modest weight loss with GLP-1 and TGR5-agonists compared to bariatric surgery suggests that additional mechanisms are important in mediating weight loss.
278
In this thesis, we obtained strong indications that FXR-activity is related to proneness in obesity, and that activating FXR in the intestine reduces body weight in mice. Furthermore, computational modeling suggested that a major part of the decrease in body weight induced through intestinal FXR-agonism may be a different handling of absorbed fat distal in the small intestine, leading to higher energy expenditure. Our findings resonate with the finding of increased energy expenditure in rodents after RYGB-surgery, in which fat absorption is obligately shifted to more distal parts of the small intestine as well [33]. Interestingly, especially meal-induced thermogenesis appears increased in rodents after RYGB-surgery [34,35]. However, there are indications that in RYGB-surgery in rodents, meal-induced thermogenesis is due to hypertrophy and increased energy expenditure of the alimentary limb, and not due to peripheral effects, since at least brown adipose tissue activity is not increased after RYGB-surgery in rodents [36]. Moreover, it should be emphasized that bariatric surgery in humans does not lead to a higher basal energy expenditure when adequately taking body composition into account [37], while the presence of increased meal-induced thermogenesis in humans after RYGB-surgery is still contested [38,39]. Thus, FXR-agonism, even when regarded as an important factor in RYGB-surgery mediated weight loss in rodents, is not likely to increase energy expenditure in humans.
In mice, FGF19 has been reported to induce weight loss through increased fat oxidation through beiging of adipose tissue [40]. Increased FGF19-secretion may have also contributed to weight loss and decreased dyslipidemia in FXR-agonist treated mice in our study. However, FGF19-mimetics in humans are not likely to increase energy expenditure [29], since the increased FGF19 levels after bariatric surgery are not associated with increased energy expenditure either. Reasoning the other way around, since weight loss associated with bariatric surgery has mostly been attributed to decreased food intake, and FGF19 has not been reported to lead to decreased food intake, increased FGF19 alone cannot explain the positive effects of bariatric surgery. Since GLP1 and TGR5-agonists alone have only limited effects on food intake, this suggests that other incretins or other mechanisms may be responsible for the decreased food intake after bariatric surgery. Indeed, DPP4-inhibitors inhibit PYY-secretion, and in part reverts reduced food intake after RYGB-surgery [38]. Similarly, somatostatin, inhibiting all incretin-secretions, completely reverts reduced food intake in these patients [41]. In our studies, we observed increased fecal fat excretion in FXR-agonist treated mice together with reduced food intake. It may be speculated that FXR-agonism reduces food intake indirectly through altering the exposure of L-cells to fatty acids, leading to increased incretin secretion through activation of the fatty acid receptors GPR40 and GPR120 [42,43]. In contrast to possibly increased energy expenditure through increased FGF19-secretion in response to FXR-agonism, indirectly increased incretin response due to FXR-agonism is more likely to be
279
translatable to humans, since studies on bariatric surgery have already shown the importance of incretins in regulating food intake to achieve weight loss.
An important question to ask, is whether our findings that FXR-activity in ‘humanized’ mice determines their proneness to obesity, can be extrapolated to reduce proneness to obesity in humans to FXR-hypoactivity. Interestingly, while intestinal FXR-agonism in mice was shown to decrease obesity, it should be noted that ileal biopsies of healthy and obese individuals showed a positive correlation between FXR-activity and body weight [44]. In fact, obesity as well as high-fat diet, increases production and pool sizes of cholic and deoxycholic acid [45]. Deoxycholic acid is a secondary bile acid, produced through reduction of the 7-alpha hydroxyl goup in cholic acid by clostridium cluster XI (and XIVa) bacteria [45]. Deoxycholic acid is not easily rehydroxylated to cholic acid in humans, and is a strong agonist [46]. However, while FXR-agonism is apparently increased in obese individuals, FGF19 levels are lower in obesity [29], suggesting that there may be a functional FXR-hypoactivity. Thus, increased FXR-agonism in the obese should be regarded as a response to obesity, and not as a cause for obesity.
All in all, these observations suggest that FXR-agonism in humans could potentially reduce body weight and the associated dyslipidemia through indirectly increasing the incretin response through increasing abundance of nutrients at the distal part of the small intestine. Therefore, pharmacological FXR-activation in the intestine may successfully emulate the decreases in body weight as observed in bariatric surgery. Finally, in the future people suffering from proneness to obesity may be able to blame FXR-hypoactivity for their disposition.
References
1. Paalvast Y, Kuivenhoven JA, Groen AK. Evaluating computational models of cholesterol metabolism. Biochim Biophys Acta - Mol Cell Biol Lipids. 2015;1851(10).
2. Vanhanen H, Kesäniemi YA, Miettinen TA. Pravastatin lowers serum cholesterol, cholesterol-precursor sterols, fecal steroids, and cholesterol absorption in man. Metabolism. 1992;41(6):588–95.
3. Kallien G, Lange K, Stange EF, Scheibner J. The pravastatin-induced decrease of biliary cholesterol secretion is not directly related to an inhibition of
cholesterol synthesis in humans. Hepatology. 1999;30(1):14–20.
4. Miettinen TA, Gylling H. Synthesis and absorption markers of cholesterol in serum and lipoproteins during a large dose of statin treatment. Eur J Clin Invest.
280 2003;33(11):976–82.
5. Schonewille M, de Boer JF, Mele L, Wolters H, Bloks VW, Wolters JC, et al. Statins increase hepatic cholesterol synthesis and stimulate fecal cholesterol elimination in mice. J Lipid Res. 2016;57:1455–64.
6. Verges B, Adiels M, Boren J, Barrett PH, Watts GF, Chan D, et al.
Interrelationships between the kinetics of VLDL subspecies and hdl catabolism in abdominal obesity: A multicenter tracer kinetic study. J Clin Endocrinol Metab. 2014;99(11):4281–90.
7. Adiels M, Olofsson S-O, Taskinen M-R, Borén J. Overproduction of very low-density lipoproteins is the hallmark of the dyslipidemia in the metabolic syndrome. Arterioscler Thromb Vasc Biol. 2008;28(7):1225–36.
8. Dubowitz N, Xue W, Long Q, Ownby JG, Olson DE, Barb D, et al. Aging is associated with increased HbA1c levels, independently of glucose levels and insulin resistance, and also with decreased HbA1c diagnostic specificity. Diabet Med. 2014;31(8):927–35.
9. Unger RH, Clark GO, Scherer PE, Orci L. Lipid homeostasis, lipotoxicity and the metabolic syndrome. Biochim Biophys Acta. 2010;1801(3):209–14. 10. Burcelin R, Crivelli V, Dacosta A, Roy-Tirelli A, Thorens B. Heterogeneous
metabolic adaptation of C57BL/6 J mice to high-fat diet. Am J Physiol Endocrinol Metab. 2002;282:834–42.
11. Peyot M, Pepin E, Lamontagne J, Latour MG, Zarrouki B, Lussier R, et al. Beta-Cell Failure in Diet-Induced Obese Mice Stratified According to Body Weight Gain : Secretory Dysfunction and Altered Islet Lipid Metabolism Without Steatosis or Reduced Beta-Cell Mass. Diabetes. 2010;59:2178–87. 12. Tarasco E, Pellegrini G, Whiting L, Lutz TA. Phenotypical heterogeneity in
responder (R) and non-responder (NR) male ApoE*3Leiden.CETP mice. Am J Physiol. 2018;
13. Schwartz CC, Almond HR, Vlahcevic ZR, Swell L. Bile acid metabolism in cirrhosis. V. Determination of biliary lipid secretion rates in patients with advanced cirrhosis. Gastroenterology. 1979;77(6):1177–82.
14. Ridlon JM, Kang DJ, Hylemon PB, Bajaj JS. Gut microbiota, cirrhosis, and alcohol regulate bile acid metabolism in the gut. Dig Dis. 2015;33(3):338–45. 15. Ridlon JM, Alves JM, Hylemon PB, Bajaj JS. Cirrhosis, bile acids and gut
microbiota: Unraveling a complex relationship. Gut Microbes. 2013;4(5). 16. Jiang J, Zhang X, Wu C, Qin X, Luo G, Deng H, et al. Increased plasma apoM
levels in the patients suffered from hepatocellular carcinoma and other chronic liver diseases. Lipids Heal Dis. 2008;7:25.
17. Fukawa T, Yan-Jiang BC, Min-Wen JC, Jun-Hao ET, Huang D, Qian CN, et al. Excessive fatty acid oxidation induces muscle atrophy in cancer cachexia. Nat Med. 2016;22(6):666–71.
281
Intestinal phospholipid remodeling is required for dietary-lipid uptake and survival on a high-fat diet. Cell Metab. 2016;23(3):492–504.
19. Buhman KK, Smith SJ, Stone SJ, Repa JJ, Wong JS, Knapp FF, et al. DGAT1 is not essential for intestinal triacylglycerol absorption or chylomicron
synthesis. J Biol Chem. 2002;277(28):25474–9.
20. Paalvast Y, de Boer JF, Groen AK. Developments in intestinal cholesterol transport and triglyceride absorption. Curr Opin Lipidol. 2017;1.
21. Jakulj L, van Dijk TH, de Boer JF, Kootte RSRS, Schonewille M, Paalvast Y, et al. Transintestinal Cholesterol Transport Is Active in Mice and Humans and Controls Ezetimibe-Induced Fecal Neutral Sterol Excretion. Cell Metab. 2016;24(6):1–12.
22. de Boer JF, Schonewille M, Boesjes M, Wolters H, Bloks VW, Bos T, et al. Intestinal Farnesoid X Receptor Controls Transintestinal Cholesterol Excretion in Mice. Gastroenterology. 2017;152(5):1126–1138.e6.
23. Kuipers F, Bloks VW, Groen AK. Beyond intestinal soap - Bile acids in metabolic control. Nat Rev Endocrinol. 2014;10(8):488–98.
24. Takahashi S, Fukami T, Masuo Y, Brocker CN, Xie C, Kristopher W, et al. Cyp2c70 is responsible for the species difference in bile acid metabolism between mice and humans. J Lipid Res. 2016;57:2130–7.
25. Kuipers F, Groen AK. FXR: The key to benefits in bariatric surgery? Nat Med. 2014;20(4):337–8.
26. Rankin W, Wittert G. Anti-obesity drugs. Curr Opin Lipidol. 2015;26(6):536– 43.
27. Srivastava G, Apovian C. Future Pharmacotherapy for Obesity: New Anti-obesity Drugs on the Horizon. Curr Obes Rep. 2018;147–61.
28. Abdeen G, le Roux CW. Mechanism Underlying the Weight Loss and Complications of Roux-en-Y Gastric Bypass. Review. Obes Surg. 2016;26(2):410–21.
29. Bozadjieva N, Heppner KM, Seeley RJ. Targeting FXR and FGF19 to Treat Metabolic Diseases—Lessons Learned From Bariatric Surgery. Diabetes. 2018;67(9):1720–8.
30. Genovese S, Mannucci E, Ceriello A. A Review of the Long-Term Efficacy, Tolerability, and Safety of Exenatide Once Weekly for Type 2 Diabetes. Adv Ther. 2017;34(8):1791–814.
31. Thomas C, Gioiello A, Noriega L, Strehle A, Oury J, Rizzo G, et al. TGR5-Mediated Bile Acid Sensing Controls Glucose Homeostasis. Cell Metab. 2009;10(3):167–77.
32. van Nierop FS, Scheltema MJ, Eggink HM, Pols TW, Sonne DP, Knop FK, et al. Clinical relevance of the bile acid receptor TGR5 in metabolism. Lancet Diabetes Endocrinol. 2016;5(3):224–33.
282 Experiments. Front Nutr. 2016;3(August).
34. Saeidi N, Meoli L, Nestoridi E, Gupta NK, Kvas S, Kucharczyk J, et al. Reprogramming of Intestinal Glucose Metabolism and Glycemic Control in Rats After Gastric Bypass. Science. 2013;341(6144):406–10.
35. Stylopoulos N, Hoppin AG, Kaplan LM. Roux-en-Y gastric bypass enhances energy expenditure and extends lifespan in diet-induced obese rats. Obesity. 2009;17(10):1839–47.
36. Abegg K, Corteville C, Bueter M, Lutz TA. Alterations in energy expenditure in Roux-en-Y gastric bypass rats persist at thermoneutrality. Int J Obes. 2016;40(8):1215–21.
37. Schmidt JB, Pedersen SD, Gregersen NT, Vestergaard L, Nielsen MS, Ritz C, et al. Effects of RYGB on energy expenditure, appetite and glycaemic control: A randomized controlled clinical trial. Int J Obes. 2016;40(2):281–90.
38. Holst JJ, Madsbad S, Bojsen-Møller KN, Svane MS, Jørgensen NB, Dirksen C, et al. Mechanisms in bariatric surgery: Gut hormones, diabetes resolution, and weight loss. Surg Obes Relat Dis. 2018;14(5):708–14.
39. Wilms B, Ernst B, Schmid SM, Thurnheer M, Schultes B. Enhanced Thermic Effect of Food After Roux-en-Y Gastric Bypass Surgery. J Clin Endocrinol Metab. 2013;98(9):3776–84.
40. Velazquez-Villegas LA, Perino A, Lemos V, Zietak M, Nomura M, Pols TWH, et al. TGR5 signalling promotes mitochondrial fission and beige remodelling of white adipose tissue. Nat Commun. 2018;9(1).
41. Le Roux CW, Welbourn R, Werling M, Osborne A, Kokkinos A, Laurenius A, et al. Gut hormones as mediators of appetite and weight loss after Roux-en-Y gastric bypass. Ann Surg. 2007;246(5):780–5.
42. Sankoda A, Harada N, Iwasaki K, Yamane S, Murata Y, Shibue K, et al. Long-chain free fatty acid receptor GPR120 mediates oil-induced GIP secretion through CCK in male mice. Endocrinology. 2017;158(5):1172–80.
43. Miyauchi S, Hirasawa A, Ichimura A, Hara T, Tsujimoto G. New frontiers in gut nutrient sensor research: free fatty acid sensing in the gastrointestinal tract. J Pharmacol Sci. 2010;112(1):19–24.
44. Jiang C, Xie C, Lv Y, Li J, Krausz KW, Shi J, et al. Intestine-selective
farnesoid X receptor inhibition improves obesity-related metabolic dysfunction. Nat Commun. 2015;6:1–18.
45. Ruderman B, Schneider H, Berchtold P. The “metabolically-obese,” normal-weight individual. Am J Clin Nutr. 1981;34:1617–21.
46. Yoshimoto S, Loo TM, Atarashi K, Kanda H, Sato S, Oyadomari S, et al. Obesity-induced gut microbial metabolite promotes liver cancer through senescence secretome. Nature. 2013;499(7456):97–101.
