The Pathogenesis of Nonalcoholic Fatty Liver Disease: Interplay between Diet, Gut Microbiota, and

14 Pages·2016·1.84 MB·English

Hindawi Publishing Corporation
Gastroenterology Research and Practice
Volume 2016, Article ID 2862173, 13 pages
http://dx.doi.org/10.1155/2016/2862173
Review Article
The Pathogenesis of Nonalcoholic Fatty Liver Disease: Interplay
between Diet, Gut Microbiota, and Genetic Background
JinshengYu,1SharonMarsh,2JunboHu,3WenkeFeng,4andChaodongWu5
1DepartmentofGenetics,WashingtonUniversitySchoolofMedicine,St.Louis,MO63110,USA
2FacultyofPharmacyandPharmaceuticalSciences,UniversityofAlberta,Edmonton,AB,CanadaT6G2H7
3DepartmentofGeneralSurgery,TongjiHospital,HuazhongScience&TechnologyUniversity,Wuhan,Hubei430030,China
4DepartmentofMedicine,UniversityofLouisville,Louisville,KY40208,USA
5DepartmentofNutritionandFoodScience,TexasA&MUniversity,Houston,TX77843,USA
CorrespondenceshouldbeaddressedtoJinshengYu;[email protected]
Received26January2016;Accepted14April2016
AcademicEditor:GregerLindberg
Copyright©2016JinshengYuetal.ThisisanopenaccessarticledistributedundertheCreativeCommonsAttributionLicense,
whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
Nonalcoholicfattyliverdisease(NAFLD)isthemostcommonchronicliverdiseaseintheworld,anditcomprisesaspectrumof
hepaticabnormalitiesfromsimplehepaticsteatosistosteatohepatitis,fibrosis,cirrhosis,andlivercancer.Whilethepathogenesis
of NAFLD remains incompletely understood, a multihit model has been proposed that accommodates causal factors from a
varietyofsources,includingintestinalandadiposeproinflammatorystimuliactingontheliversimultaneously.Priorcellularand
molecularstudiesofpatientandanimalmodelshavecharacterizedseveralcommonpathogenicmechanismsofNAFLD,including
proinflammationcytokines,lipotoxicity,oxidativestress,andendoplasmicreticulumstress.Inrecentyears,gutmicrobiotahas
gainedmuchattention,anddysbiosisisrecognizedasacrucialfactorinNAFLD.Moreover,severalgeneticvariantshavebeen
identifiedthroughgenome-wideassociationstudies,particularlyrs738409(Ile748Met)inPNPLA3andrs58542926(Glu167Lys)
inTM6SF2,whicharecriticalriskallelesofthedisease.Althoughahigh-fatdietandinactivelifestylesaretypicalriskfactors
for NAFLD, the interplay between diet, gut microbiota, and genetic background is believed to be more important in the
development and progression of NAFLD. This review summarizes the common pathogenic mechanisms, the gut microbiota
relevantmechanisms,andthemajorgeneticvariantsleadingtoNAFLDanditsprogression.
1.Introduction of patients with NAFLD are obese or even morbidly obese
andhaveaccompanyinginsulinresistancethatplaysacentral
Nonalcoholic fatty liver disease (NAFLD) is the most com- roleinthemetabolicsyndrome[6–9].Thus,NAFLDisalso
monchronicliverdiseaseintheworld.Itispresentin30% deemedtobehepaticmanifestationofmetabolicsyndrome
ofthegeneraladultpopulationandfoundpredominantlyin which is a cluster of complex conditions including central
obese people with high-fat diets and inactive lifestyles. In obesity,hypertension,hyperglycaemia,hypertriglyceridemia,
reality,NAFLDcomprisesaspectrumofhepaticabnormali- and low HDL (high density lipoprotein) that are predictive
tiesthatareobservableinliverhistologicalslides,fromasim- risk factors of cardiovascular disease, stroke, and diabetes
pleintrahepaticaccumulationoffat(steatosisornonalcoholic [10,11].
fattyliver,NAFL)tovariousdegreesofnecroticinflammation NAFLDhasbeenconsideredaconditionwitha“two-hit”
(nonalcoholicsteatohepatitis,NASH)[1–3].Simplesteatosis processofpathogenesissince1998whenDayandJamesfirst
(i.e., NAFL) rarely progresses to advanced disease whereas, proposedthishypothesis[12]withevidencefromtheBerson
in approximately 20% of patients with NASH, it progresses et al. study describing the role of lipid peroxidation in the
to fibrosis and cirrhosis and potentially to hepatocellular liver injury [13]. Essentially, the first hit is the development
carcinoma over a 15-year time period [4, 5]. The majority of hepatic steatosis via accumulation of triglycerides in
2 GastroenterologyResearchandPractice
hepatocytes, which increases the vulnerability of the liver 2.1. Adipose Tissue Inflammation. What exactly initiates
to various possible “second hits” that in turn lead to the adipose tissue inflammation in obesity is uncertain; but
inflammation, fibrosis, and cellular death characteristics of hypoxia and death of rapidly expanding adipocytes are
NASH. The second hit can be a variety of factors, such as believed to play a role [15]. Adipocytes under inflamma-
oxidativestress,endoplasmicreticulumstress,proinflamma- tion secrete cytokines and chemokines, particularly tumor
tory cytokines, and gut-derived bacterial endotoxin. As it necrosis factor-𝛼 (TNF-𝛼), interleukin-6 (IL-6), and CC-
hasevidentlyemergedthat(1)accumulationoftriglycerides chemokine ligand-2 (CCL2) [15, 16]. TNF-𝛼 was the first
in hepatocytes may be a protective mechanism from liver proinflammatory cytokine detected in adipose tissue and
damageand(2)hepaticinflammationcanprecedethesimple is involved in the regulation of insulin resistance. Studies
hepatic steatosis and can also be a cause of steatosis, it has indicated that neutralization of TNF-𝛼 activity by an anti-
beenbelievedthatmany“hit”factorsmayactsimultaneously TNF-𝛼monoclonalantibodyimprovesinsulinresistanceand
leading to the development of NAFLD, which supports the fattyliverdiseaseinanimals[17].IL-6isderivedfrommany
multihit model proposed in 2010 [14]. Indeed, among the cellsthroughoutthebodyincludingadipocytes.Serumlevels
proposed hit factors, many can interact with each other, ofthesecytokinescorrelateremarkablywellwiththepresence
forming a vicious circle. Recent advances in metagenomics ofinsulinresistance,andadiposetissue-derivedTNF-𝛼and
complicatetheunderstandingofthepathogenesisofNAFLD IL-6 have been shown to regulate hepatic insulin resistance
further in that dysbiosis and host-microbiota interactions viaupregulationofSOCS3,asuppressorofcytokinesignaling
arenowalsoimplicated.Moreover,genome-wideassociation [17].CCL2recruitsmacrophagestotheadiposetissue,result-
studies have discovered several promising candidate genes, inginevenmorelocalcytokineproductionandperpetuating
serving as the genetic background for the disease. These the inflammatory cycle; TNF-𝛼 and IL-6 induce a state of
genetic players appear to distinguish subgroups of NFLD insulinresistanceinadipocytes,whichstimulatestriglyceride
patients from obese and insulin resistance associated pop- lipolysis and fatty acid release into the circulation. At the
ulations. Although a high-fat diet and inactive lifestyles are sametime,extrahepaticadipocytesarecompromisedintheir
typical risk factors for NAFLD, the interplay between diet, natural ability to secrete adiponectin, an anti-inflammatory
gut microbiota, and genetic background can play a crucial adipokinethatfacilitatesthenormalpartitioningoflipidto
role in the development and progression of NAFLD. This adipocytesforstorage[18].Circulatingadiponectinregulates
review summarizes the common pathogenic mechanisms, hepatic fatty 𝛽-oxidation through AMP-activated protein
the gut microbiota relevant mechanisms, and the major kinase(AMPK)andacetyl-CoAcarboxylase(ACC)signaling
geneticvariantsleadingtoNAFLDanditsprogression(Fig- [19]. Together, these abnormalities accentuate fat loss from
ure1). adipocytesandpromoteectopicfataccumulation.
2.CommonPathogenicMechanisms 2.2. De Novo Lipogenesis (DNL). Presumably lipogenesis in
ofNAFLD livercouldbeincreasedduetothesteatoticnatureofNAFLD.
A number of prior studies have shown that diets enriched
Hepatic steatosis is a prerequisite to making a histological in both saturated fat and simple sugar carry a high risk of
diagnosis of NAFLD [2]. Several mechanisms may lead to hepaticsteatosis,atleastinpart,throughenhancedDNL[20–
steatosis, including (1) increased fat supply such as high-fat 23].TheroleofDNLinthedevelopmentofhepaticsteatosisis
diet and excess adipose lipolysis; (2) decreased fat export furthersupportedbyarecentstudyinsubjectswithmetabolic
in the form of very low density lipoprotein-triglyceride; (3) syndromeandahighcontentofliverfat[24].A3-foldhigher
decreased free fatty 𝛽-oxidation; and (4) increased de novo rateofdenovofattyacidsynthesisisseeninthesesubjects.
lipogenesis (DNL) [2]. Molecular mechanisms responsible In addition, specific dietary compositions may have differ-
for the accumulation of fat in the liver are not fully under- ent effects. Basically since carbohydrates are substrates for
stood;however,certaincytokinesderivedfrominflammation DNL,theamountofcarbohydrateinthedietwillpositively
sites, particularly from extrahepatic adipose tissues, can influencetheamountofDNLintheliver.Simplesugarsare
triggerthisprocess.Inaddition,theenhancementofhepatic convertedtofattyacidsmorereadilythancomplexcarbohy-
DNL is deemed to be a unique feature in steatosis. More drates[25,26],andfructoseisamorepotentinducerofDNL
importantly, insulin resistance appears to be at center stage thanglucose[27,28].Thisisalsosupportedbyepidemiologic
for the massive metabolic dysregulations of NAFLD that evidence linking dietary fructose to hepatic steatosis and
initiate and aggravate hepatic steatosis. At a certain point, NASH[20,29].Itisworthnotingthatdietaryfat,particularly
the simple steatosis transforms to steatohepatitis in about saturatedfat,stimulatesDNLbyupregulatingSREBP-1(sterol
20–30%ofNAFLDpatients.Thisbreakthrough-likeprocess responsive element binding protein), a key regulator of the
is mediated by the interplay of multiple hit factors. Patho- lipogenicgenesintheliver[30].However,notallindividuals
logical features of NASH include simple hepatic steatosis with hepatic steatosis had increased DNL nor upregulated
and, more characteristically, liver cell damage and accom- SREBP-1expression,asobservedintheMancinaetal.study
panying inflammation and/or fibrosis. Currently, a number showingaparadoxicaldissociationbetweenhepaticDNLand
of common pathogenic mechanisms have been proposed hepaticfatcontentduetothePNPLA3148Mallele[31].
andcharacterizedforthetransitionfromsimplesteatosisto
NASH, such as lipotoxicity, oxidative stress, mitochondrial 2.3.InsulinResistance. Studieshavehighlightedthefactthat
dysfunction,andendoplasmicreticulumstress. insulinresistanceisacharacteristicfeatureofNAFLD[7–9]
GastroenterologyResearchandPractice 3
High-fat diet, inactive Genetic variants in about10genes
lifestyle, obesity, and identified via genome-wide
other environmental association studies confer genetic
stimuli are risk factors predisposition to NAFLD, of which
of NAFLD PNPLA3 and TM6SF2 are at the top list
NAFLD
populations
PNPLA3
No genetic predisposition but insulin resistance TM6SF2 Genetic predisposition but no insulin resistance
and so on
No genetic predisposition and no insulin resistance Genetic predisposition and insulin resistance
Proinflammatory cytokines Inflammation Adipose hypoxia
(TNF-𝛼, IL-6, CCL2, etc.)↑ Adipocyte death
Adiponectin
MDF
Oxidative stress Obesity
Insulin
resistance
Insulin
ER stress
resistance
Steatosis Adipose
Steatohepatitis De novo Free fatty acid release Muscle
Gluconeogenesis
lipogenesis
Insulin
Glucose uptake resistance
TLRs
Lipotoxicity
Bacterial translocation Intestinal Tight junction damage
(endotoxemia, etc.) permeability↑
Blood stream
Gut-liver axis Enterocyte
Short-chain fatty acids↓
Choline absorption↓
Bile acid pool imbalance Intestinal lumen
Endogenous alcohol↑ Eubiosis Eubiosis
And so on Dysbiosis
Diet (saturated fatty acids, fructose, etc.) Probiotics
Figure1:Overviewatthepathogenesisofnonalcoholicfattyliverdisease(NAFLD).Theinterplaybetweendiet,microbiota,andhostgenetic
variantsplaysacrucialroleinthecomplexpathogenesisofNAFLDthroughavarietyofmechanisms.TheNAFLDpatientscannowbe
categorizedintodifferentpopulationsbasedontheirinsulinsensitivityandgeneticpredisposition.Insulinresistanceisatthecenterofthe
NAFLDpathogenicprocess,andanumberofkeyfactorsareinvolvedinthedevelopmentofNAFLD,suchasdiet,dysbiosis,gut-liveraxis,
geneticpredispositiongenes(PNPLA3andTM6SF2),oxidativestress,MDF(mitochondrialdysfunction),endoplasmicreticulum(ER)stress,
denovolipogenesis,lipotoxicity,andproinflammatorycytokines.
andiscausedbyavarietyoffactors,includingsolublemedia- hand, insulin resistant subjects with NAFLD show reduced
torsderivedfromimmunecellsand/oradiposetissue,such insulinsensitivity,notonlyatthelevelofthemuscle,butalso
as TNF-𝛼 and IL-6 [32]. Serine phosphorylation of insulin attheleveloftheliverandadiposetissue[7–9,33],whichcan
receptorsubstratesbyinflammatorysignaltransducerssuch leadtoafarmorecomplexmetabolicdisturbanceoflipidand
as c-jun N-terminal protein kinase 1 (JNK1) or inhibitor of glucose.However,notallpeoplewithNAFLDhaveincreased
nuclearfactor-𝜅Bkinase-𝛽(IKK-𝛽)isconsideredoneofthe insulinresistance,andNAFLD,perse,cannotbeconsidered
keyaspectsthatdisruptsinsulinsignaling[14].Ontheother a cause for insulin resistance but rather a consequence
4 GastroenterologyResearchandPractice
as shown by studies in subjects genetically predisposed to ROScandamagetheETC[51]andevencausemutationsin
NAFLD. Mutations in PNPLA3 (patatin-like phospholipase themitochondriaDNA[52].
domain containing 3) [34, 35], TM6SF2 (transmembrane 6
superfamily member 2) [6, 36], DGAT1 (diacylglycerol O- 2.6. Oxidative Stress. In the context of increased supply of
acyltransferase 1) [37], or hypobetalipoproteinemia [38, 39] fatty acids to hepatocytes, oxidative stress can occur and
genesarenotrelatedtoincreasedinsulinresistanceexceptfor be attributable to raised levels of reactive oxygen/nitrogen
severelyobeseindividualsinwhichitisassociated[40].Itis species(ROS/RNS)andlipidperoxidationthataregenerated
worthnotingthatinsulinresistanceischaracterizednotonly during free fatty acid metabolism in microsomes, peroxi-
by increased circulating insulin levels but also by increased somes, and mitochondria [53–55]. Peroxidation of plasma
hepaticgluconeogenesis,impairedglucoseuptakebymuscle, and intracellular membranes may cause direct cell necro-
and increased release of free fatty acids and inflammatory sis/apoptosis and megamitochondria, while ROS-induced
cytokines from peripheral adipose tissues [41], which are expressionofFas-ligandonhepatocytesmayinducefratrici-
the key factors promoting accumulation of liver fat and dalcelldeath.Recentstudiessupporttheideathatoxidative
progressionofhepaticsteatosis(Figure1). stress may be a primary cause of liver fat accumulation
and subsequent liver injury, and ROS may play a part even
2.4. Lipotoxicity. Studies have indicated that certain lipids in fibrosis development [56, 57]. Importantly, these species
canbeharmfultohepatocytesinNAFLD.Thisisparticularly can initiate lipid peroxidation by targeting polyunsaturated
true of the long-chain saturated fatty acids (SFAs) such as fatty acids (PUFAs), resulting in the formation of highly
palmitate(C16:0)andstearate(C18:0),whichareabundantin reactivealdehydeproducts,suchas4-hydroxy-2-nonenal(4-
animalfatanddairyproductsandproducedintheliverfrom HNE) and malondialdehyde (MDA). These reactive lipid
dietarysugar.Underphysiologicalconditions,SFAsaretrans- derivativeshavethepotentialtoamplifyintracellulardamage
portedtomitochondriafor𝛽-oxidationoresterifiedforeither bymediatingthediffusionofROS/RNSintotheextracellular
excretionintheformofVLDL(verylowdensitylipoproteins) space,thuscausingtissuedamage.
orstorageaslipiddroplets.InthepathophysiologyofNASH,
multiplemechanismsareconcurrentlyoperativetoproduce 2.7. Endoplasmic Reticulum (ER) Stress. The ER is a vast
liverinjuryinhepatocytesoverwhelmedbySFAandbyfree dynamic andtubularnetworkresponsibleforthesynthesis,
cholesterol(FC)fromdenovosynthesis[42,43].FCaccumu- folding/repair, and trafficking of a wide range of proteins
lationleadstoliverinjurythroughtheactivationofintracellu- [58].Underpathologicaland/orstressfulconditionssuchas
larsignalingpathwaysinKupffercells(KCs),hepaticstellate NASH,ithasbeenobservedthatERefficiencyintheprotein-
cells (HSCs), and hepatocytes. The activation of KCs and folding,repairing,and/ortraffickingmachineryisdecreased
HSCs promotes inflammation and fibrogenesis [44]. These whilethedemandofproteinsynthesisandfolding/repairis
lipids, including FC, SFA, and certain lipid intermediates increased [58, 59]. Such an imbalance between the load of
from excessive SFA, can activate a variety of intracellular neededprotein-foldingandtheresponse-relatedcapabilityof
responsessuchasJNK1andamitochondrialdeathpathway, the ER is termed ER stress, which can lead to the accumu-
resultinginlipotoxicstressintheendoplasmicreticulumand lationofunfoldedand/ormisfoldedproteinswithintheER
mitochondria,respectively[42,43,45].Inaddition,thetoll- lumen.Thistypeofcellularstressusuallytriggersanadaptive
likereceptor4(TLR4)isapatternrecognitionreceptorthat response,aimedatresolvingERstress,calledunfoldedpro-
activatesaproinflammatorysignalingpathwayinresponseto teinresponse(UPR)[60–62].TheUPRismediatedbyatleast
excessiveSFAs.Thispathwayisinitiatedbyrecruitingadap- threedifferentstress-sensingpathways:proteinkinaseRNA-
tor molecules such as toll/IL-1 receptor domain containing likeERkinase(PERK),inositol-requiringprotein1𝛼(IRE1𝛼),
adaptor protein (TIRAP) and myeloid differentiation factor and activating transcription factor 6 (ATF6) [62]. Coupled
88(MyD88)thatultimatelyleadtoactivationofnuclearfactor with inflammation, oxidative stress, insulin resistance, and
𝜅BwithproductionofTNF-𝛼[46]. apoptosis signaling, hepatic ER stress seems to play an
important role in regulating the composition and size of
2.5. Mitochondrial Dysfunction. Mitochondria are the most lipiddropletsaswellaslipidsynthesis,includingcholesterol
important energy suppliers of the cell and play a pivotal metabolism[58,59,63],throughSREBP.
roleinfattyacidmetabolism.Fattyacidoxidationisableto
beupregulatedtocompensateforsomedegreeofincreased 3.MicrobiotaAssociatedMechanisms
depositionoffat;however,multiplestudieshaveshownthat ofNAFLD
liver ATP levels are decreased in NAFLD [47–49]. This
discrepancy implicates mitochondrial dysfunction in the Gutmicrobiotawasfirstfoundtobealteredinpatientswith
state of liver fat overload that is characteristic of NAFLD. chronic liver disease more than 80 years ago. Derangement
Althoughthemechanismsresponsibleforthemitochondrial of the gut flora, in particular small intestinal bacterial
dysfunction remain poorly understood in NAFLD, reduced overgrowth (SIBO), occurs in a large percentage (20–75%)
enzymatic activities of mitochondrial electron transport of patients with chronic liver disease. In recent years, the
chain (ETC) complexes may be attributed to increased gut microbiome has gained much more attention due to
generation of reactive oxygen species (ROS) as a result of the advancement of the high-throughput next-generation
ETC leakage during mitochondrial 𝛽-oxidation in energy sequencing(NGS)technology.Priorstudiesofgutflorarelied
production(intheformofATP)[50].Studieshavefoundthat onculturedependenttechniques,whichwerelaborintensive
GastroenterologyResearchandPractice 5
and limited only to a countable number of species, as over carnitine palmitoyltransferase 1A expression [72], and a
80%ofthegutmicrobesarenotcultivatable[64].Incontrast, peroxisome proliferator-activated receptor-𝛾- (PPAR𝛾-)
NGS-based taxonomic assignments of the uncultured, dependent mechanism, shifting metabolism in adipose and
undefined microbes into operational taxonomic units livertissuefromlipogenesistofattyacidoxidation[73].
(OTUs) represent an effective and revolutionary approach
forstudiesonhighlycomplexgutmicrobiota,whichisbased 3.2.DietaryCholineMechanism. Dietarycholineisrequired
on clustering of the 16S rRNA sequences derived from the for very low density lipoprotein synthesis and hepatic lipid
NGSplatforms.Thisapproachallowsthecharacterizationof export;anddietarycholine-deficiencyhasbeenlinkedwitha
bothcompositionanddiversityoftheintestinalmicrobiota. varietyofconditionsincludinghepaticsteatosis.Buchmanet
Accordingtoevidencefromrelevantstudies,thegutmicro- al.[74]foundthat,inpatientswithparenteralnutrition,diets
biotamaycontributetothepathogenesisofNAFLDthrough deficient in choline can lead to increased hepatic steatosis,
severalmechanisms,including(1)increasedproductionand which can be reversed with choline supplementation. This
absorptionofgutshort-chainfattyacids;(2)altereddietary study suggests a role of choline in fat export out of the
choline metabolism by the microbiota; (3) altered bile acid hepatocytes. Recent studies indicate a role of the intestinal
poolsbythemicrobiota;(4)increaseddeliveryofmicrobiota- microbiota in the conversion of dietary choline to toxic
derivedethanoltoliver;(5)gutpermeabilityalterationsand methylamine, a substance that not only mimics a choline-
release of endotoxin; and (6) interaction between specific deficient diet by decreasing effective choline levels but also
dietandmicrobiota.Mostrecently,Mussoetal.broughtupa exposes the host to an inflammatory toxic metabolite [75].
newmechanismbythatchronickidneydiseasemaymutually Very recently, a metagenomic analysis of the microbial
aggravate NAFLD and associated metabolic disturbances communitieslivingintheintestinaltractsof15womenwith
through multiple paths including altered intestinal barrier a choline-depleted diet revealedthatincreasedGammapro-
functionand microbiotacomposition[65]. In fact, diet can teobacteria abundance and decreased Erysipelotrichi abun-
affectthecompositionanddiversityofgutmicrobiota;thus dancewereprotectiveagainstdevelopingsteatosis[76].
any changes of gut microbiota that are observed in a diet-
stratified study should be interpreted with caution because 3.3.BileAcidPoolRelatedMechanisms. Withinhepatocytes,
thesechangescouldbeeitheradirecteffectofspecificdiets bileacidsaresynthesizedfromcholesterolthroughenzymatic
or an indirect effect of the gut-liver interactive axis which pathways and then conjugated with either glycine or tau-
hasbeenproposedandobservedrecently[66,67]. rine before secretion into bile and released into the small
intestine. In the small intestine, conjugated bile acids not
3.1. Short-Chain Fatty Acids (SCFAs) Relevant Mechanisms. only assist in lipid absorption and transport but have also
In the intestine, SCFAs are produced in the distal small beenincreasinglyrecognizedtofunctionasnuclearreceptor
intestine and colon where nondigestible carbohydrates bindersandtohaveaputativeroleinalteringthemicrobiome
like resistant starch, dietary fiber, and other low-digestible [77]. On the other hand, bacteria within the intestine can
polysaccharides are fermented by saccharolytic bacteria also chemically modify bile acids and thereby alter the
which include the phyla Bacteroidetes, Firmicutes, and compositionofthebileacidpool[78,79].Besidestheclassic
Actinobacteria.Acetateandpropionatearethemainproducts roleasdetergentstofacilitatefatabsorption,bileacidshave
of Bacteroidetes phylum and butyrate is mainly produced also been recognized as important cell signaling molecules
by Firmicutes phylum. As an energy precursor, SCFAs are regulating lipid metabolism, carbohydrate metabolism, and
implicated in the pathogenesis of NAFLD because of their inflammatory response [80]. These molecular functions are
possiblecontributiontoobesity.Thefirstevidenceregarding mediatedthroughtheirbindingandactivationofthenuclear
SCFAs was from Turnbaugh et al. study [68] showing that hormone receptor, farnesoid X receptor (FXR), and the G
thececumofob/obmicehasanincreasedconcentrationof protein coupled cell surface receptor TGR5 [81]. Intestinal
SCFAs and that transplantation of germ-free mice with the FXRactivityupregulatesendocrineFGF19expression,which
gut microbiome from ob/ob mice caused greater fat gain inhibitshepaticbileacidsynthesisviaCYP7A1signaling[82].
than transplants from lean animals. In humans, increased McMahanetal.showedthatactivationofbileacidreceptors
productionofSCFAsbythegutmicrobiotawasalsoobserved withareceptoragonistwasabletoimproveNAFLDhistology
in overweight and obese people, compared to lean subjects in an obese mouse model [83]. Due to the nature of the
[69]. In metagenomics analysis, the majority of studies complexinterplaybetweenthemicrobiomeandthehostbile
showedthatob/obmice[70]andobesepatients[71]exhibit acidpool,furtherstudiesarerequiredinthecontextofrisk
reduced abundance of Bacteroidetes and proportionally forNAFLDandNASH.
increased abundance of Firmicutes. However, how these
ratio changes affect energy imbalance leading to obesity 3.4.EndogenousAlcoholTheory. Thepossibleroleforendo-
and its complications including NAFLD needs further genousalcoholinNAFLDwasfirstimplicatedinob/obmice.
functional and species-level analyses. In fact, SCFAs have Copeetal.foundthatalcoholinthebreathofobeseanimals
more beneficial effects than their obesity-causing effects in is higher than that of lean animals [84], but they could
general[72].BeneficialeffectsofSCFAsarethroughseveral not find any difference in the breath alcohol concentration
ways,suchasimmunoregulation,enhancedintestinalbarrier betweenNASHpatientsandleancontrolsinahumanstudy
function,actingasahistonedeacetylase1(HDAC)inhibitor [85].Recently,Zhuetal.foundthatNASHpatientsexhibited
to decrease expression of lipogenic genes and to increase significantly elevated blood ethanol levels, while similar
6 GastroenterologyResearchandPractice
bloodethanolconcentrationswereobservedbetweenhealthy acids [96], studies have implicated the importance of LPS-
subjects and obese non-NASH patients [86]. Further, in TLR4/TLR9 signaling in the pathogenesis of NAFLD. Both
thismetagenomicsstudy,thecompositionofNASHmicro- TLR4-andTLR9-deficientmiceareprotectedfromhigh-fat
biomes was found to be distinct from those of healthy and diet-induced inflammation and insulin resistance [97, 98],
obese microbiomes, and Escherichia stood out as the only whilemicedeficientinTLR5developallfeaturesofmetabolic
abundant genus that differed between NASH and obese syndromeincludinghyperphagia,obesity,insulinresistance,
patients.BecauseEscherichiaareethanolproducers,thisfind- pancreaticinflammation,andhepaticsteatosis[99].Metage-
ing is in agreement with their previousreport that alcohol- nomic analysis indicated that TLR5 deficiency affected the
metabolizingenzymesareupregulatedinNASHlivers[87]. compositionofthegutmicrobiotaand,remarkably,transfer
However,Engstleretal.providedevidenceagainstthealcohol of the microbiota from TLR5−/− mice to healthy mice
theory [88]. In their study, ethanol levels were similar in resulted in transfer of disease [99]. Moreover, Wlodarska
portal vein and chyme obtained from different parts of the et al. found that NOD-like receptor family pyrin domain
GI tract between groups, while ethanol levels in vena cava containing6(NLRP6)inflammasomedeficiencyleadstoan
plasma were significantly higher in ob/ob mice, suggesting alteredtransmissible,colitogenicgutmicrobiota[100].When
that more ethanol was not metabolized in the liver due to fed with a methionine and choline-deficient diet (MCDD),
a significantly lower ADH activity observed in these ob/ob these inflammasome deficient mice developed NASH with
mice. They proposed that increased blood ethanol levels in significantlyhigherseveritythanwild-typeanimals[101].
patients with NAFLD may result from insulin-dependent
impairmentsofADHactivityinlivertissue,ratherthanfrom 3.6.SaturatedFattyAcids. Ithasbeenwellknownthatanimal
anincreasedendogenousethanolsynthesis.Thus,thealcohol meats are rich in saturated fatty acids (SFAs) which are
theorycurrentlyfacesconflictingresultsfromdifferentinves- highly correlated to an increased risk of obesity, diabetes,
tigators.Toclarifytheseconflictingresults,deMedeirosand andcardiovasculardiseases.Manystudieshaveindicatedthat
deLimahaveprovidedaninterestingmechanisticframework saturated fatty acids are more toxic than their unsaturated
explaining how NAFLD might be an endogenous alcohol counterparts [102, 103]. It is worth noting that SFAs are
fatty liver disease (EAFLD) [89]. However, this framework protective in alcohol induced fatty liver disease [104–106].
requiresexperimentalevidencetobevalidated. However, in liver and hepatocytes not exposed to alcohol,
SFAs appear to promote apoptosis and liver injury [107,
3.5.IntestinalPermeabilityandEndotoxemia. Thegutmicro- 108].IthasbeenshownthatSFAsincreasethesaturationof
biotaplaysapartinmaintainingtheintegrityoftheintestinal membrane phospholipids, thus initiating unfolded protein
barrier [90]; and changes in the composition of microbiota response (UPR) and leading to ER stress [108, 109]. SFAs
canleadtoincreasedintestinalpermeabilityandsubsequent also affect mitochondrial metabolism and promote ROS
overflowofharmfulbacterialby-productstotheliverthatin accumulation [23]. Furthermore, hepatocyte apoptosis has
turntriggershepaticinflammationandmetabolicdisorders. been shown to be dependent on the activation of JNK
Endotoxin,thatis,lipopolysaccharide(LPS),isderivedfrom stresssignalingpathwaysthatrespondtoprolongedERand
Gram-negative bacteria, and it has long been implicated in oxidative stress [109]. In addition, SFAs can interact with
chronicliverdiseases.Thefirstevidenceinsupportofarole gut microbiota to affect the progression of liver injury. For
for LPS in the pathogenesis of NASH was the observation instance,byanalyzingchangesintheintestinalmetagenome
that endotoxemia readily induces steatohepatitis in obese and metabolome of alcohol-fed mice, Chen et al. recently
rats and mice [91]. Further, murine NAFLD models of foundthatsynthesisofsaturatedlong-chainfattyacidissig-
bacterial overgrowth develop compositional changes of the nificantly reduced when compared with normal-chew mice
gutmicrobiotaandpresentincreasedintestinalpermeability, andthatsupplementationofsaturatedlong-chainfattyacids
with a concurrent reduction in the expression of tight recoversintestinaleubiosisandreducesethanol-inducedliver
junctionproteins[92].Inhumanstudies,Mieleetal.found injury in mice [110]. Moreover, de Wit et al. observed an
evidence of a disruption in the intestinal barrier of biopsy- overflow of SFAs to the distal intestine in mice on a high-
proven NAFLD patients, along with an increased rate of SFAdiet,which,ratherthanobesityitself,reducedmicrobial
smallbowelbacteriaovergrowth,suggestingthatalterations diversityandincreasedtheFirmicutes-to-Bacteroidetesratio
inthemicrobiomemayhavecontributedtodisruptionofgut in the intestine. Such a typical obesity microbiota profile
barrierintegrity[93].Inaddition,high-fatdietsmayfacilitate stimulated by SFAs favors the development of obesity and
LPS uptake through elevated chylomicron production in hepaticsteatosis[103].
intestinalepithelialcells[94].Ontheotherhand,Yuanetal.
didnotfindthecorrelationbetweenGram-negativebacteria 3.7. Fructose. Fructose has been utilized as artificial sweet-
abundance and the concentration of serum endotoxin and ener in many commercial soft drinks that are consumed
there was no endotoxemia in the majority of pediatric largely by adolescents and in a variety of social circum-
NASH patients [95], highlighting the multihit hypothesis stances.Anumberofstudieshavefoundthatexcessfructose
for the pathogenesis of NASH. Nonetheless, LPS and other consumption is involved in the pathogenesis of NAFLD
exogenous stimuli are responded to first by innate immu- andthatupregulateddenovolipogenesisandinhibitedfatty
nity through pattern recognition receptors such as toll-like acid 𝛽-oxidation are distinct metabolic processes for the
receptors(TLRs)andNOD-likereceptors(NLRs).Although developmentofhepaticsteatosisinindividualswithNAFLD
TLRs might respond to nutritional lipids such as free fatty [20, 24, 111–113]. Further, Abdelmalek et al. observed that
GastroenterologyResearchandPractice 7
increased fructose consumption is associated with a higher withincreasedliverfatandhepaticinflammationandfibrosis.
fibrosis stage in patients with NAFLD, independent of age, This finding has subsequently been reproduced with solid
sex, BMI, and total calorie intake [29]. Using a fructose- evidenceasshowninameta-analysiscomprising16studies
inducedNAFLDmousemodel,recentstudieswithmetage- [129]. Compared with noncarriers, homozygous carriers of
nomics analysis found that fructose significantly decreased the variant had a 73% higher liver fat content, a 3.2-fold
Bifidobacterium and Lactobacillus and tended to increase greaterriskofhighnecroinflammatoryscores,anda3.2-fold
endotoxemia[114,115].Severalprobioticbacterialstrainsof greater risk of developing fibrosis. The association between
Lactobacillus protect mice from the development of high- the PNPLA3 variantandsteatosisorseverityofhistological
fructose-induced NAFLD [116–118]. In addition, increased liverdiseasehasbeenwidelyobservedinthemajorityofsub-
expressionofTLRshasbeenimplicatedinthedevelopment sequent genome-wide association studies [130] and several
offructose-inducedhepaticsteatosis[119]. case-controlstudies,includingthoseinChinese,Korean,and
Japanesepopulations[131–133].Itisworthnotingthatthelink
4.GeneticBackgroundofNAFLD betweenPNPLA3I148MvariantandNAFLDisindependent
ofmetabolicsyndrome(MS)anditsfeatures;thatis,mostof
Genomic variations that have a causative effect on the patientscarryingthisvariantarenotassociatedwithobesity,
development of human diseases can be divided into two diabetes, and atherogenic dyslipidemia, as demonstrated in
groups:onesinrarediseasesandonesincommondiseases. the recent meta-analysis [129]. Furthermore, the PNPLA3
The former follow Mendelian inheritance patterns that are genotype seems to also influence steatosis development in
characterized by a single, highly penetrant but uncommon patientswithhepatitisBandhepatitisCandalcoholabuse,
mutation in a specific gene being necessary and sufficient andithasbeenindependentlyassociatedwiththeprogression
to cause the disease. The latter consist of causative muta- of hepatitis, including fibrosis, cirrhosis, and HCC occur-
tions that are not subject to negative selection pressures, rence[134–136].TheassociationbetweenthePNPLA3variant
and disease susceptibility is due to the combined effects of I148M andtheriskofHCCdevelopmenthas been robustly
multiplerelativelycommoncausativepolymorphisms(minor validated in patients with NAFLD [137, 138], and it has
allelefrequency1–5%)thatarecarriedbyaffectedindividuals. beenestimatedthatthehomozygouscarriersofthep.148M
Like most common diseases, NAFLD has been implicated mutation carry a 12-fold increased HCC risk as compared
in an inherited component to susceptibility, meaning that to p.I148 homozygotes [139]. Finally, as described earlier,
genetic variation does influence disease risk. As reviewed subpopulations of NAFLD patients with PNLA3 mutation
by Macaluso et al. in 2015 [120], dozens of genes with are not associated with insulin resistance, a hallmark of
multiple polymorphisms have been discovered in genome- metabolic syndrome. Collectively, it seems that a distinct
wideassociationstudies(GWAS)thatmayberesponsiblefor entity might exist in which the PNPLA3 risk allele appears
risk of NAFLD in certain populations. It is believed that as to be a major driver of disease progression in combination
more large scale GWAS are complete, more genes could be with viral infection, alcohol abuse, lifestyle (unhealthy diet
identified.Forinstance,inearly2016whilewewerepreparing andinactivity),and/ornonlifestyle(cryptogenic)causes,for
thisreview,anovelvariantMBOAT7rs641738wasreported example,PNPLA3-associatedsteatohepatitis(“PASH”)[140].
tobeassociatedwiththedevelopmentandseverityofNAFLD
inindividualsofEuropeandescent[121].Amongallreported 4.2. TM6SF2 (Transmembrane 6 Superfamily Member 2).
genes, onlytwo ofthem(PNPLA3 and TM6SF2)havebeen Another widely validated and intriguing genetic player in
identified as potential genetic modifiers in more than one NAFLD is the nonsynonymous variant rs58542926 (c.449
largescalestudy[120,122],whicharethefocusofourreview. C>T) within the TM6SF2 gene at the 19p13.11 locus, which
According to genotypes in those key genes and sensitivity encodesanE167Kaminoacidsubstitution.Theroleofvariant
toinsulin,NAFLDpatientscanbecategorizedintodifferent E167K in TM6SF2 was first described by Kozlitina et al.
subpopulations(Figure1). [36] in an exome-wide association study in a multiethnic,
population-basedcohort,highlightingtheassociationofthe
4.1.PNPLA3(Patatin-LikePhospholipaseDomainContaining TM6SF2 variant with higher serum alanine aminotrans-
3). The PNPLA3 gene (adiponutrin) encodes a transmem- ferase(ALT)andaspartateaminotransferase(AST)levels—
brane polypeptide chain exhibiting triglyceride hydrolase assurrogatesforNASH—andwithreducedplasmalevelsof
activity[123],whichishighlyexpressedontheendoplasmic triglyceridesandlowdensitylipoprotein-(LDL-)cholesterol.
reticulumandlipidmembranesofhepatocytesand adipose Inaddition,theyperformedafunctionalanalysisforTM6SF2
tissue [124]. It is also reported that PNPLA3 is highly inmousemodelsbysilencingthegeneviaadeno-associated
expressed in human stellate cells. The encoded protein has viralvectors.Silencingofthegeneshoweda3-foldincrease
retinyl esterase activity and allows retinol secretion from inhepatictriglycerideslevelsandadecreaseinplasmalevels
hepatic stellate cells while the mutation causes intracellular oftriglycerides,LDL-andhighdensitylipoprotein-(HDL-)
retention of this compound [125–127]. As the first genome- cholesterols,andverylowdensitylipoprotein(VLDL).Over-
wide association study with strong evidence for NAFLD, a all, their results demonstrated that the TM6SF2 gene regu-
reportfromRomeoetal.in2008showedthatageneticvari- lated hepatic triglyceride secretion and that the functional
ant,analleleinPNPLA3(rs738409[G],encodingIle148Met), impairment of TM6SF2 promoted NAFLD. An association
conferssusceptibilitytothediseaseinindividualsofseveral betweentheTM6SF2rs58542926SNPandtheseverityofliver
westernpopulations[128].Thisgeneticvariantwasassociated disease has also been found in patients with biopsy-proven
8 GastroenterologyResearchandPractice
NAFLD in a recent study reported by Liu et al. [141]. phenotype. More intriguingly, environmental reprogram-
More intriguingly, the E167K variant in TM6SF2 seems mingofmicrobiotaresultedinobesity-prone129S6/SvEvTac
able to disconnect the risk of NAFLD/NASH progression mice becoming obesity resistant. This study suggests that
from cardiovascular risk, which is supported mainly by the developmentofobesity/metabolicsyndromeistheresultof
Dongiovanni et al. study [142] showing that 188 (13%) out interactionsbetweengutmicrobiota,hostgenetics,anddiet.
of 1201 subjects who underwent liver biopsy for suspected
NASH were carriers of the E167K variant and that these 6.Conclusions
carriershadlowerserumlipidlevelsthannoncarriers,more
severesteatosis,necroinflammation,ballooning,andfibrosis NAFLD is best considered a multietiology disease trait,
and were more likely to have NASH and advanced fibrosis meaning that it is not caused by a single gene mutation
afteradjustingformetabolicfactorsandthePNPLA3I148M genetically and is not associated with only a single factor
risk variant. In addition, E167K carriers had lower risk of environmentally; but it is the outcome of genetic variant-
developing carotid plaque; and in Swedish obese subjects environmental factor interplay determining disease pheno-
assessed for cardiovascular outcomes, E167K carriers had type and progression. The genetic variants in PNPLA3 and
higherALTandlowerlipidlevelsbutalsoalowerincidenceof TM6SF2 are only responsible for ∼50% of NAFLD patients
cardiovascularevents.Consequently,carriersoftheTM6SF2 [120], and majority of PNPLA3-associated NAFLD patients
E167K variant seem to be more at risk for progressive are not obese and have no insulin resistance and its related
NASH,butatthesametimetheycouldbeprotectedagainst diabetesandcardiovasculardiseases[140].Infact,likemany
cardiovasculardiseases[143]. commondiseases,NAFLDispolygenic,wheretheheritable
componenttosusceptibilityvariouslyaccountsforupto30–
5.InterplaybetweenDiet,Microbiota, 50%ofrelativerisk[130].Moreover,individualenvironmen-
andHostGenetics tal factors, particularly the specific diets, interact with gut
microbiota up front before a final beneficial or damaging
Oneofthebiggestlessonswelearnedfromthemetagenomic signal is sent. Whether environmental factors, including
studies so far is that constitutive profiles of gut microbiota lifestyle, are the cause of NAFLD will be steered by the
candetermineliverpathologyinresponsetoahigh-fatdiet interactionwiththehostgeneticsaswellastheconstitutional
(HFD)inmice,reflectingakindofinteractiveeffectbetween profile of gut microbiota. Thus, careful, multifaceted study
dietandgutmicrobiota,thatis,aneteffectaftertheinterplay. designs are warranted in future analysis in order to “catch”
For instance, in a transplantation experiment [144], Le Roy thetruecausestoNAFLD.
et al. selected donor mice at first, based on their responses
to a HFD. The “responders” developed hyperglycaemia CompetingInterests
and had a high plasma concentration of proinflammatory
cytokines, and the “nonresponders” were normoglycaemic Nocompetinginterestsaredeclaredbytheauthors.
and had a lower level of systemic inflammation, although
bothdevelopedcomparableobesityontheHFD.Germ-free Acknowledgments
mice were then colonized with intestinal microbiota from
eithertheresponderorthenonrespondermiceandthenfed AuthorsthankDr.RichardD.Head,theDirectorofGenome
the same HFD. The responder-receiver (RR) group devel- Technology Access Center at Department of Genetics in
opedfastinghyperglycaemiaandinsulinaemia,whereasthe Washington University School of Medicine in St. Louis,
nonresponder-receiver (NRR) group remained normogly- Missouri,UnitedStates,forhisguidanceandsupportofthis
caemic.IncontrasttoNRRmice,RRmicedevelopedhepatic work.
macrovesicular steatosis, which was confirmed by a higher
liverconcentrationoftriglyceridesandincreasedexpression References
ofgenesinvolvedindenovolipogenesis.Pyrosequencingof
the 16S ribosomal RNA genes revealed that RR and NRR [1] M.Demir,S.Lang,andH.Steffen,“Nonalcoholicfattyliverdis-
micehaddistinctgutmicrobiotaincludingdifferencesatthe ease:currentstatusandfuturedirections,”JournalofDigestive
Diseases,vol.16,no.10,pp.541–557,2015.
phylum,genera,andspecieslevels.Theseresultssuggestthat
the gut microbiota can contribute to the development of [2] D.G.Tiniakos,M.B.Vos,andE.M.Brunt,“Nonalcoholicfatty
NAFLD, independent of obesity but acting like a constitu- liver disease: pathology and pathogenesis,” Annual Review of
tionalbackgroundofahostorgansystem.Theinterrelation- Pathology:MechanismsofDisease,vol.5,pp.145–171,2010.
shipbetweendiet,gutmicrobiota,andhostgeneticshasbeen [3] A. Abu-Shanab and E. M. M. Quigley, “The role of the gut
unraveled further in a recent study reported by Ussar and microbiotainnonalcoholicfattyliverdisease,”NatureReviews
coworkers[145].Inthisstudy,theyutilizedthreecommonly GastroenterologyandHepatology,vol.7,no.12,pp.691–701,2010.
usedinbredstrainsofmice:obesity/diabetes-proneC57Bl/6J, [4] P.Angulo,“Long-termmortalityinnonalcoholicfattyliverdis-
obesity/diabetes-resistant129S1/SvImJ,andobesity-pronebut ease:isliverhistologyofanyprognosticsignificance?”Hepatol-
diabetes-resistant129S6/SvEvTacmice.Analysisofmetabolic ogy,vol.51,no.2,pp.373–375,2010.
parametersandgutmicrobiotainallstrainsandtheirenvi- [5] C. Soderberg, P. Sta˚l, J. Askling et al., “Decreased survival of
ronmentallynormalizedderivativesrevealedstronginterac- subjects with elevated liver function tests during a 28-year
tionsbetweenmicrobiota,diet,breedingsite,andmetabolic follow-up,”Hepatology,vol.51,no.2,pp.595–602,2010.
GastroenterologyResearchandPractice 9
[6] Y. Zhou, G. Llaurado´, M. Oreˇsicˇ, T. Hyo¨tyla¨inen, M. Orho- [23] M.V.Machado,P.Ravasco,L.Jesusetal.,“Bloodoxidativestress
Melander, and H. Yki-Ja¨rvinen, “Circulating triacylglycerol markersinnon-alcoholicsteatohepatitisandhowitcorrelates
signaturesandinsulinsensitivityinNAFLDassociatedwiththe withdiet,”ScandinavianJournalofGastroenterology,vol.43,no.
E167KvariantinTM6SF2,”JournalofHepatology,vol.62,no.3, 1,pp.95–102,2008.
pp.657–663,2015. [24] J.E.Lambert,M.A.Ramos-Roman,J.D.Browning,andE.J.
[7] R.Lomonaco,C.Ortiz-Lopez,B.Orsaketal.,“Effectofadipose Parks,“Increaseddenovolipogenesisisadistinctcharacteristic
tissue insulin resistance on metabolic parameters and liver ofindividualswithnonalcoholicfattyliverdisease,”Gastroen-
histologyinobesepatientswithnonalcoholicfattyliverdisease,” terology,vol.146,no.3,pp.726–735,2014.
Hepatology,vol.55,no.5,pp.1389–1397,2012. [25] L.C.Hudgins,T.S.Parker,D.M.Levine,andM.K.Hellerstein,
[8] G. Pagano, G. Pacini, G. Musso et al., “Nonalcoholic steato- “Adualsugarchallengetestforlipogenicsensitivitytodietary
hepatitis,insulinresistance,andmetabolicsyndrome:further fructose,”JournalofClinicalEndocrinologyandMetabolism,vol.
evidenceforanetiologicassociation,”Hepatology,vol.35,no.2, 96,no.3,pp.861–868,2011.
pp.367–372,2002. [26] V.Lecoultre,L.Egli,G.Carreletal.,“Effectsoffructoseandglu-
[9] A. J. Sanyal, C. Campbell-Sargent, F. Mirshahi et al., “Non- coseoverfeedingonhepaticinsulinsensitivityandintrahepa-
alcoholic steatohepatitis: association of insulin resistance and ticlipidsinhealthyhumans,”Obesity,vol.21,no.4,pp.782–785,
mitochondrialabnormalities,”Gastroenterology,vol.120,no.5, 2013.
pp.1183–1192,2001. [27] E. J. Parks, L. E. Skokan, M. T. Timlin, and C. S. Dingfelder,
[10] M.-E. Dumas, J. Kinross, and J. K. Nicholson, “Metabolic “Dietarysugarsstimulatefattyacidsynthesisinadults,”Journal
phenotypingandsystemsbiologyapproachestounderstanding ofNutrition,vol.138,no.6,pp.1039–1046,2008.
metabolicsyndromeandfattyliverdisease,”Gastroenterology, [28] K. L. Stanhope, J. M. Schwarz, N. L. Keim et al., “Consum-
vol.146,no.1,pp.46–62,2014. ing fructose-sweetened, not glucose-sweetened, beverages in-
[11] D.SchuppanandJ.M.Schattenberg,“Non-alcoholicsteatohep- creases visceral adiposity and lipids and decreases insulin
atitis:pathogenesisandnoveltherapeuticapproaches,”Journal sensitivityinoverweight/obesehumans,”TheJournalofClinical
ofGastroenterologyandHepatology,vol.28,supplement1,pp. Investigation,vol.119,no.5,pp.1322–1334,2009.
68–76,2013. [29] M.F.Abdelmalek,A.Suzuki,C.Guyetal.,“Increasedfructose
[12] C. P. Day and O. F. W. James, “Steatohepatitis: a tale of two consumptionisassociatedwithfibrosisseverityinpatientswith
‘hits’?”Gastroenterology,vol.114,no.4I,pp.842–845,1998. nonalcoholicfattyliverdisease,”Hepatology,vol.51,no.6,pp.
[13] A. Berson, V. De Beco, P. Letteron et al., “Steatohepatitis- 1961–1971,2010.
inducing drugs cause mitochondrial dysfunction and lipid [30] J. Lin, R. Yang, P. T. Tarr et al., “Hyperlipidemic effects of
peroxidationinrathepatocytes,”Gastroenterology,vol.114,no. dietarysaturatedfatsmediatedthroughPGC-1𝛽coactivationof
4,pp.764–774,1998. SREBP,”Cell,vol.120,no.2,pp.261–273,2005.
[14] H. Tilg and A. R. Moschen, “Evolution of inflammation in [31] R.M.Mancina,N.Matikainen,C.Maglioetal.,“Paradoxical
nonalcoholicfattyliverdisease:themultipleparallelhitshypo- dissociation between hepatic fat content and de novo lipo-
thesis,”Hepatology,vol.52,no.5,pp.1836–1846,2010. genesis due to PNPLA3 sequence variant,” Journal of Clinical
[15] A.R.Johnson,J.J.Milner,andL.Makowski,“Theinflammation EndocrinologyandMetabolism,vol.100,no.5,pp.E821–E825,
highway:metabolismacceleratesinflammatorytrafficinobe- 2015.
sity,”ImmunologicalReviews,vol.249,no.1,pp.218–238,2012. [32] H.TilgandA.R.Moschen,“Inflammatorymechanismsinthe
[16] S. C. Matherly and P. Puri, “Mechanisms of simple hepatic regulationofinsulinresistance,”MolecularMedicine,vol.14,no.
steatosis:notsosimpleafterall,”Clinicsin Liver Disease,vol. 3-4,pp.222–231,2008.
16,no.3,pp.505–524,2012. [33] A.Gastaldelli,K.Cusi,M.Pettitietal.,“Relationshipbetween
[17] R. Barbuio, M. Milanski, M. B. Bertolo, M. J. Saad, and L. hepatic/visceralfatandhepaticinsulinresistanceinnondiabetic
A.Velloso,“Infliximabreversessteatosisandimprovesinsulin andtype2diabeticsubjects,”Gastroenterology,vol.133,no.2,pp.
signaltransductioninliverofratsfedahigh-fatdiet,”Journalof 496–506,2007.
Endocrinology,vol.194,no.3,pp.539–550,2007. [34] K. Sevastianova, A. Kotronen, A. Gastaldelli et al., “Genetic
[18] H. Tilg, “The role of cytokines in non-alcoholic fatty liver variationinPNPLA3(adiponutrin)conferssensitivitytoweight
disease,”DigestiveDiseases,vol.28,no.1,pp.179–185,2010. loss—induceddecreaseinliverfatinhumans,”TheAmerican
[19] C.M.Hasenour,E.D.Berglund,andD.H.Wasserman,“Emerg- JournalofClinicalNutrition,vol.94,no.1,pp.104–111,2011.
ingroleofAMP-activatedproteinkinaseinendocrinecontrolof [35] K.Kantartzis,A.Peter,F.Machicaoetal.,“Dissociationbetween
metabolismintheliver,”MolecularandCellularEndocrinology, fattyliverandinsulinresistanceinhumanscarryingavariantof
vol.366,no.2,pp.152–162,2013. thepatatin-likephospholipase3gene,”Diabetes,vol.58,no.11,
[20] X.Ouyang,P.Cirillo,Y.Sautinetal.,“Fructoseconsumption pp.2616–2623,2009.
asariskfactorfornon-alcoholicfattyliverdisease,”Journalof [36] J.Kozlitina,E.Smagris,S.Stenderetal.,“Exome-wideassocia-
Hepatology,vol.48,no.6,pp.993–999,2008. tionstudyidentifiesaTM6SF2variantthatconferssusceptibil-
[21] N.Assy,G.Nasser,I.Kamayseetal.,“Softdrinkconsumption itytononalcoholicfattyliverdisease,”NatureGenetics,vol.46,
linkedwithfattyliverintheabsenceoftraditionalriskfactors,” no.4,pp.352–356,2014.
CanadianJournalofGastroenterology,vol.22,no.10,pp.811– [37] K.Kantartzis,F.Machicao,J.Machannetal.,“TheDGAT2gene
816,2008. isacandidateforthedissociationbetweenfattyliverandinsulin
[22] S. Zelber-Sagi, D. Nitzan-Kaluski, R. Goldsmith et al., “Long resistanceinhumans,”ClinicalScience,vol.116,no.6,pp.531–
term nutritional intake and the risk for non-alcoholic fatty 537,2009.
liver disease (NAFLD): a population based study,” Journal of [38] M.E.Visser,N.M.Lammers,A.J.Nederveenetal.,“Hepatic
Hepatology,vol.47,no.5,pp.711–717,2007. steatosis does not cause insulin resistance in people with
10 GastroenterologyResearchandPractice
familialhypobetalipoproteinaemia,”Diabetologia,vol.54,no.8, [54] G.H.Koek,P.R.Liedorp,andA.Bast,“Theroleofoxidative
pp.2113–2121,2011. stress in non-alcoholic steatohepatitis,” Clinica Chimica Acta,
[39] A. Amaro, E. Fabbrini, M. Kars et al., “Dissociation between vol.412,no.15-16,pp.1297–1305,2011.
intrahepatictriglyceridecontentandinsulinresistanceinfamil- [55] B.NowickaandJ.Kruk,“Occurrence,biosynthesisandfunction
ialhypobetalipoproteinemia,”Gastroenterology,vol.139,no.1, ofisoprenoidquinones,”BiochimicaetBiophysicaActa(BBA)—
pp.149–153,2010. Bioenergetics,vol.1797,no.9,pp.1587–1605,2010.
[40] C.N.A.Palmer,C.Maglio,C.Pirazzietal.,“Paradoxicallower [56] D.Pessayre,“Roleofmitochondriainnon-alcoholicfattyliver
serum triglyceride levels and higher type 2 diabetes mellitus disease,” Journal of Gastroenterology and Hepatology, vol. 22,
susceptibility in obese individuals with the PNPLA3 148M supplement1,pp.S20–S27,2007.
variant,”PLoSONE,vol.7,no.6,ArticleIDe39362,2012. [57] E.Novo,C.Busletta,L.V.D.Bonzoetal.,“Intracellularreactive
oxygenspeciesarerequiredfordirectionalmigrationofresident
[41] M. Gaggini, M. Morelli, E. Buzzigoli, R. A. DeFronzo, E.
andbonemarrow-derivedhepaticpro-fibrogeniccells,”Journal
Bugianesi,andA.Gastaldelli,“Non-alcoholicfattyliverdisease
ofHepatology,vol.54,no.5,pp.964–974,2011.
(NAFLD) and its connection with insulin resistance, dyslipi-
demia,atherosclerosisandcoronaryheartdisease,”Nutrients, [58] G. S. Hotamisligil, “Endoplasmic reticulum stress and the
vol.5,no.5,pp.1544–1560,2013. inflammatorybasisofmetabolicdisease,”Cell,vol.140,no.6,
pp.900–917,2010.
[42] M.FuchsandA.J.Sanyal,“LipotoxicityinNASH,”Journalof
Hepatology,vol.56,no.1,pp.291–293,2012. [59] U. O¨zcan, Q. Cao, E. Yilmaz et al., “Endoplasmic reticulum
stresslinksobesity,insulinaction,andtype2diabetes,”Science,
[43] P. Simonen, A. Kotronen, M. Hallikainen et al., “Cholesterol
vol.306,no.5695,pp.457–461,2004.
synthesis is increased and absorption decreased in non-alco-
[60] D. Ron and P. Walter, “Signal integration in the endoplasmic
holic fatty liver disease independent of obesity,” Journal of
reticulumunfoldedproteinresponse,”NatureReviewsMolecu-
Hepatology,vol.54,no.1,pp.153–159,2011.
larCellBiology,vol.8,no.7,pp.519–529,2007.
[44] G. Arguello, E. Balboa, M. Arrese, and S. Zanlungo, “Recent
[61] J.LeeandU.Ozcan,“Unfoldedproteinresponsesignalingand
insights on the role of cholesterol in non-alcoholic fatty liver
metabolic diseases,” The Journal of Biological Chemistry, vol.
disease,”BiochimicaetBiophysicaActa(BBA)—MolecularBasis
289,no.3,pp.1203–1211,2014.
ofDisease,vol.1852,no.9,pp.1765–1778,2015.
[62] C.Hetz,“Theunfoldedproteinresponse:controllingcellfate
[45] H.Malhi,S.F.Bronk,N.W.Werneburg,andG.J.Gores,“Free
decisionsunderERstressandbeyond,”NatureReviewsMolec-
fatty acids induce JNK-dependent hepatocyte lipoapoptosis,”
ularCellBiology,vol.13,no.2,pp.89–102,2012.
TheJournalofBiologicalChemistry,vol.281,no.17,pp.12093–
[63] V.Zambo,L.Simon-Szabo,P.Szelenyietal.,“Lipotoxicityinthe
12101,2006.
liver,”WorldJournalofHepatology,vol.5,no.10,pp.550–557,
[46] T.Sharifnia,J.Antoun,T.G.C.Verriereetal.,“HepaticTLR4
2013.
signalinginobeseNAFLD,”AmericanJournalofPhysiology—
[64] P. S. Langendijk, F. Schut, G. J. Jansen et al., “Quantitative
GastrointestinalandLiverPhysiology,vol.309,no.4,pp.G270–
fluorescenceinsituhybridizationofBifidobacteriumspp.with
G278,2015.
genus-specific16SrRNA-targetedprobesanditsapplicationin
[47] G. Serviddio, F. Bellanti, R. Tamborra et al., “Alterations of fecalsamples,”AppliedandEnvironmentalMicrobiology,vol.61,
hepaticATPhomeostasisandrespiratorychainduringdevel- no.8,pp.3069–3075,1995.
opment of non-alcoholic steatohepatitis in a rodent model,”
[65] G. Musso, M. Cassader, S. Cohney et al., “Emerging liver-
EuropeanJournalofClinicalInvestigation,vol.38,no.4,pp.245–
kidneyinteractionsinnonalcoholicfattyliverdisease,”Trends
252,2008.
inMolecularMedicine,vol.21,no.10,pp.645–662,2015.
[48] Y. Jiang, M. Zhao, and W. An, “Increased hepatic apoptosis
[66] I. A. Kirpich, L. S. Marsano, and C. J. McClain, “Gut-liver
inhigh-fatdiet-inducedNASHinratsmaybeassociatedwith
axis, nutrition, and non-alcoholic fatty liver disease,” Clinical
downregulation of hepatic stimulator substance,” Journal of
Biochemistry,vol.48,no.13-14,pp.923–930,2015.
MolecularMedicine,vol.89,no.12,pp.1207–1217,2011.
[67] P.Lin,J.Lu,Y.Wangetal.,“NaturallyoccurringstilbenoidTSG
[49] X. Jin, Y.-D. Yang, K. Chen et al., “HDMCP uncouples yeast reverses non-alcoholic fatty liver diseases via gut-liver axis,”
mitochondrial respiration and alleviates steatosis in L02 and PLoSONE,vol.10,no.10,ArticleIDe0140346,2015.
hepG2 cells by decreasing ATP and H2O2 levels: a novel [68] P. J. Turnbaugh, R. E. Ley, M. A. Mahowald, V. Magrini,
mechanismforNAFLD,”JournalofHepatology,vol.50,no.5,
E. R. Mardis, and J. I. Gordon, “An obesity-associated gut
pp.1019–1028,2009.
microbiomewithincreasedcapacityforenergyharvest,”Nature,
[50] C. P. Day, “Pathogenesis of steatohepatitis,” Best Practice & vol.444,no.7122,pp.1027–1031,2006.
ResearchClinicalGastroenterology,vol.16,no.5,pp.663–678,
[69] A.Schwiertz,D.Taras,K.Scha¨feretal.,“MicrobiotaandSCFA
2002.
inleanandoverweighthealthysubjects,”Obesity,vol.18,no.1,
[51] H. A. Sadek, P. A. Szweda, and L. I. Szweda, “Modulation of pp.190–195,2010.
2+
mitochondrialcomplexIactivitybyreversibleCa andNADH [70] R. E. Ley, F. Ba¨ckhed, P. Turnbaugh, C. A. Lozupone, R. D.
mediated superoxide anion dependent inhibition,” Biochem- Knight,andJ.I.Gordon,“Obesityaltersgutmicrobialecology,”
istry,vol.43,no.26,pp.8494–8502,2004. Proceedings of the National Academy of Sciencesof the United
[52] C.C.Kujoth,A.Hiona,T.D.Pughetal.,“Medicine:mitochon- StatesofAmerica,vol.102,no.31,pp.11070–11075,2005.
drialDNAmutations,oxidativestress,andapoptosisinmam- [71] R.E.Ley,P.J.Turnbaugh,S.Klein,andJ.I.Gordon,“Microbial
malianaging,”Science,vol.309,no.5733,pp.481–484,2005. ecology:humangutmicrobesassociatedwithobesity,”Nature,
[53] C.P.Day,“Non-alcoholicfattyliverdisease:currentconcepts vol.444,no.7122,pp.1022–1023,2006.
andmanagementstrategies,”ClinicalMedicine,vol.6,no.1,pp. [72] G. den Besten, K. van Eunen, A. K. Groen, K. Venema, D.-J.
19–25,2006. Reijngoud,andB.M.Bakker,“Theroleofshort-chainfattyacids