Nonalcoholic fatty liver disease: cause or consequence of type 2 diabetes?
Valenti L1,2, Bugianesi E3, Pajvani U4, Targher G5.
1. Internal Medicine and Metabolic Diseases, Fondazione IRCCS Ca' Granda Ospedale
Maggiore Policlinico, Milano, Italy. [email protected].
2. Department of Pathophysiology and Transplantation, Università degli Studi di Milano,
Milano, Italy. [email protected].
3. Division of Gastroenterology, Department of Medical Sciences, A.O.U. Città della Salute e
della Scienza, Università di Torino, Torino, Italy.
4. Division of Endocrinology, Columbia University, New York, NY, USA.
5. Section of Endocrinology, Diabetes and Metabolism, Department of Medicine, University
and Azienda Ospedaliera Universitaria Integrata of Verona, Verona, Italy.
Abstract
Growing epidemiological evidence suggests that nonalcoholic fatty liver disease (NAFLD) is an
early predictor of and determinant for the development of type 2 diabetes and other features of the
metabolic syndrome. This finding may have important clinical implications for the diagnosis,
prevention and treatment of type 2 diabetes and its chronic complications. However, given the
complex and bi-directional relationships between NAFLD, insulin resistance and chronic
hyperglycaemia, it is extremely difficult to distinguish whether NAFLD is a cause or a consequence
of insulin resistance and type 2 diabetes. Indeed, at the molecular level, hepatic lipogenesis and
hepatic glucose production depend on differentially regulated branches of the insulin signalling
pathway. Furthermore, genetic studies suggest that excess hepatic fat is associated with progressive
liver disease, but does not always increase the risk of incident type 2 diabetes. Here, we will briefly
review the epidemiological, pathophysiological and molecular evidence linking NAFLD to the
development of type 2 diabetes. We will also discuss some recent genetic and therapeutic advances
that seem to challenge a causal role of NAFLD in the pathogenesis type 2 diabetes, and propose a
working hypothesis to explain this apparent conundrum. In conclusion, progressive liver disease
and type 2 diabetes are divergent though inter-related consequences of insulin resistance and the
metabolic syndrome.
Keywords: genetics – insulin resistance – nonalcoholic fatty liver disease – pathophysiology – type
2 diabetes
Nonalcoholic fatty liver disease (NAFLD) is defined by excessive hepatic fat content (HFC; >5% of
liver weight, mainly in the form of triglycerides) in the absence of excessive alcohol consumption
and other competing etiologies for hepatic steatosis. Paralleling the epidemic of obesity in both
developed and developing countries, NAFLD has become the leading cause of liver disease, and an
emerging important cause of liver failure and hepatocellular carcinoma (1–3). The prevalence of
NAFLD increases in subjects with obesity, the metabolic syndrome (MetS), or type 2 diabetes
mellitus (T2DM) (4–6). For many years, NAFLD has indeed been considered the hepatic
manifestation of systemic insulin resistance (IR) and the MetS (7, 8). Furthermore, IR and T2DM
are among the strongest predictors of the progression of NAFLD to advanced fibrosis and cirrhosis
(9–11), while insulin-sensitizing drugs have been used albeit with a variable success for NAFLD
treatment (12). However, NAFLD is now increasingly diagnosed also in individuals who do not
have T2DM or MetS. This supports the assertion that the conventional paradigm of NAFLD
representing the simple ‘hepatic manifestation’ of the MetS is outdated.
Several epidemiological studies have recently suggested that NAFLD may be an early predictor of
and a potentially important determinant for the development of both T2DM and other MetS traits
(13). This finding may have important clinical implications for the diagnosis, prevention and
treatment of T2DM. However, given the complex and bi-directional relationships between NAFLD,
IR and T2DM, it is extremely difficult to distinguish whether NAFLD plays a causal role in the
development of T2DM or whether it is merely a consequence of IR/T2DM.
Here, we will review the epidemiological, pathophysiological and molecular evidence linking
NAFLD to T2DM development. We will also discuss recent genetic and therapeutic advances that
seem to challenge a casual role of NAFLD in T2DM development, and propose a working
hypothesis to explain this apparent conundrum. Insulin resistance in NAFLD Insulin resistance is a
multifaceted syndrome that expresses itself in many ways, mainly represented by the features of the
MetS but not limited to them (13, 14). The physiological effect of insulin varies dramatically from
tissue to tissue, and different pathways within the same tissue vary in their degree of IR. Therefore,
the complex inter-relationship among NAFLD, IR and MetS cannot be easily detangled.
The most common cause of NAFLD in developed countries is an increased caloric intake exceeding
the rates of caloric expenditure, with consequent spillover of extra-energy in the form of lipid
precursors (i.e. nonesterified fatty acids, NEFA) from adipose tissue into ectopic depots: liver,
visceral fat (VF), skeletal muscle and pancreas (15). Insulin decreases circulating lipids through
inhibition of adipose tissue lipolysis and by directly suppressing hepatic production of very low
density lipoproteins (VLDL), but in IR states these actions are impaired (14). Specifically, NAFLD
develops when the rate of hepatic triglyceride (TG) synthesis, mainly due to increased hepatic
NEFA uptake and de novo lipogenesis (DNL) exceeds the rate of hepatic TG catabolism due to
NEFA oxidation and TG export as VLDL particles (13). In this sense, NAFLD is promoted by
adipose tissue and skeletal muscle IR, and represents a good marker of the severity of IR itself.
However, the liver is also a key site of insulin action: it is the main source of endogenous glucose
production, a major site for the synthesis and disposal of lipids and the primary site of insulin
extraction from plasma. In the liver, which normally extracts circulating NEFA with high efficiency
(25–40%), impaired suppression of lipolysis by insulin provides the energy and biochemical signals
required for the stimulation of gluconeogenesis (16). In non-diabetic individuals, increased
gluconeogenesis forms the basis of hepatic IR, i.e. rate of hepatic glucose production (HGP) that is
inappropriately elevated for the prevailing insulin levels (17, 18). Furthermore, increased HFC and
diacylglycerol (DAG) accumulation are associated with decreased insulin-stimulated tyrosine
phosphorylation of both Insulin Receptor Substrate (IRS)-1 and IRS-2 (15), but the causal link
between hepatic DAG and IR remains controversial (19). Fasting hyperglycaemia and
hyperinsulinaemia are closely associated with both HFC and IR, irrespective of body mass index
(20). Evaluation of the main sites of IR by the euglycaemic hyperinsulinaemic clamp studies,
coupled with tracers infusion in a group of non-diabetic nonobese patients with NAFLD, revealed
the presence of IR in the following target pathways: endogenous insulin release by the pancreas,
endogenous glucose production (hepatic IR); whole-body glucose disposal and its main
components (i.e. glucose oxidation and glycogen synthesis; muscle IR); lipolysis and lipid
oxidation (adipose tissue IR) (21). The mechanisms by which the majority of insulin-resistant
individuals are protected from developing T2DM remain a mystery, but it is the inability of
pancreatic b-cell to compensate for IR that renders individuals frankly diabetic. Fasting insulin
secretion and total stimulated insulin output are increased as a result of the prevailing
hyperglycaemia, and are correlated with IR in a positive fashion; in time, as hyperglycaemia
worsens total insulin output begins to decline (14, 22, 23). In NAFLD, a long-term increase in HGP
can exacerbate fasting and post-prandial hyperglycaemia and predispose to T2DM development
(13), although this does not explain the marked impairment in indices of dynamic b-cell
responsiveness, such as glucose sensitivity. An alternate hypothesis is that the main cause of the IR,
i.e. lipotoxicity, is also directly responsible for the b-cell failure (22). Many of the adverse effects of
IR are secondary to compensatory hyperinsulinaemia acting on tissues/pathways that remain insulin
sensitive, and much depends on the primary function of the organ/tissue involved (14). In the liver,
hyperinsulinaemia stimulates the transcription factor Sterol Regulatory Element-Binding Protein1c (SREBP-1c), which activates most genes involved in DNL. The contribution of DNL to HFC,
which is less than 5% in healthy subjects, increases to approximately 25–30% in NAFLD (24).
Hence, in this condition the compensatory hyperinsulinaemia and the increased delivery of NEFA
turn the liver into a ‘fat producing factory’, starting a vicious cycle that worsens the atherogenic
lipid profile and increases the likelihood of cardiovascular complications. Increased amounts of
circulating and intracellular NEFA are also associated with an increase in nuclear factor kappa-b
(NF-jB), leading to increased transcription of interleukin-6 and other proinflammatory cytokines
(25). NAFLD, in particular nonalcoholic steatohepatitis (NASH), may also directly contribute to the
low-grade inflammatory state of IR through the hepatic production of several inflammatory
mediators (e.g., C-reactive protein, fibrinogen, monocyte chemotactic protein 1, and tumour
necrosis factora) (25), as well as other mediators potentially involved in the pathogenesis of
metabolic alterations and cardiovascular disease, such as fibrinogen, betatrophin (ANGPTL8),
Fibroblast growth factor (FGF)-21 and Fetuin A.
In this setting, the direct contribution of NAFLD to IR and MetS is tightly inter-related with that of
VF. HFC and VF are highly correlated with each other and to the degree of IR, but the relative role
of both fat depots on IR is controversial. The comparison of subjects with low VF and low HFC vs.
subjects with both high VF and HFC yielded contrasting results (26, 27), but a recent analysis
showed that having increased HFC with low VF is more deleterious than having high VF with low
HFC (28). The portal theory suggests that VF may be harmful because omental and mesenteric
adipocytes are more insulin-resistant and may release more NEFA into the portal vein than
subcutaneous adipocytes (29). The direct impact of VF on glucose and lipid metabolism is,
however, controversial. VF may also release more inflammatory cytokines than subcutaneous
fat and contribute to both systemic and hepatic inflammation (29). However, it is unlikely that FFAs
released by VF are responsible for generalized lipotoxicity, except through cytokines released by
the dysfunctional tissues (30). In fact, VF releases more pro-inflammatory cytokines than
subcutaneous fat and contributes to both systemic and hepatic inflammation (29). Adiponectin,
the main adipokine with insulin-sensitizing, anti-inflammatory and anti-fibrotic action (31), may be
also involved in the interplay between the liver and adipose tissue in NAFLD patients. Indeed,
circulating adiponectin decreases with NAFLD severity (32), IR and adipose tissue inflammation,
while experimental evidence suggests a protective role of this adipokine on HFC and liver disease
progression (33).
The pattern of IR in patients with NAFLD described so far is quite different from that found in
severe liver disease of any other etiology. In non-diabetic patients with cirrhosis, basal HGP is
normal and suppressed by insulin; lipolysis is increased in the post-absorptive state because of
‘accelerated starvation’ observed in cirrhotic patients, but it is suppressed by insulin in an almost
normal manner. Additionally, hyperinsulinaemia in cirrhotic patients is mainly caused by
hepatocellular dysfunction and intrahepatic shunting (34). The strong inter-relationship between
NAFLD and IR suggests that NAFLD may be both a marker and a promoter of
IR/hyperinsulinaemia. Beyond that, the presence of NAFLD may also differentiate those T2DM
patients who will develop specific metabolic and cardiovascular complications (35). However,
NAFLD developing in individuals carrying strong genetic risk factors, such as in the presence of
homozygosity for the rs738409 C>G variant in the Patatin-like phospholipase domain-containing
protein 3 gene (PNPLA3 I148M variant) or associated with rare loss-of-function mutations in
Apolipoprotein B (APOB), may be accompanied by less severe IR and fewer MetS features,
although, as discussed later, a combination of obesity, IR and genetic risk factors is most frequently
observed in clinical practice (13, 36). NAFLD as a cause of T2DM: the epidemiological evidence
Table 1 summarizes the principal data from observational cohort or case–control studies that have
assessed the association between NAFLD, as detected by ultrasonography or computed
tomography, and the risk of incident T2DM (37–53). The main characteristics of the study
populations, the length of follow-up period, the diagnostic criteria used for NAFLD and T2DM, the
relative risks of developing T2DM, and the list of covariates considered in multivariable regression
models have been specified. We did not include the large number of prospective, population-based
cohort studies that used serum liver enzyme levels (i.e. surrogate markers of NAFLD) to diagnose
NAFLD. These studies have consistently shown that mildly elevated serum liver enzyme levels are
associated with an increased risk of incident T2DM, independently of common risk factors in
various ethnic populations (54). The vast majority of the published studies have shown that NAFLD
diagnosed on ultrasonography substantially increased the risk of incident T2DM (Table 1). Only
one retrospective cohort study did not find any significant association between NAFLD and the risk
of T2DM after adjusting for confounding variables (37). The risk of incident T2DM varied
markedly from a 33% (42) to a 5.5-fold increase in patients with NAFLD (39). This wide interstudy variability might reflect differences in the demographic characteristics of the study
populations, in the length of follow-up period (ranging from a mean period of 3–12.8 years), the
varying degree of confounder adjustment, and differences in NAFLD severity (as detected either by
ultrasonography or non-invasive scoring systems of advanced hepatic fibrosis) in individual studies.
For example, a large prospective cohort study of over 25 000 Korean non-diabetic middle-aged men
demonstrated that NAFLD was an independent risk factor for incident T2DM, and that the absolute
risk increased sharply with the ultrasonographic severity of NAFLD at baseline (absent: 7.0%, mild:
9.8%, moderate-to-severe: 17.8%). The hazard ratios for T2DM development were higher in
subjects with mild steatosis (adjusted-HR 1.09, 95% CI 0.8–1.5) and in those with moderate-tosevere steatosis (adjusted-HR 1.73, 95% CI 1.0–3.0) compared to the normal group even after
adjusting for a wide range of T2DM risk factors (43). An independent retrospective study of an
occupational cohort of 38 291 Korean individuals showed that the HRs for incident T2DM
comparing NAFLD with low NAFLD fibrosis score (a reliable non-invasive marker of advanced
liver fibrosis) and NAFLD with intermediate or high fibrosis score vs. the non-NAFLD group were
2.0 (95% CI 1.8–2.2) and 4.7 (95% CI 3.7–6.1) respectively (46). Interestingly, Fukuda et al. (49)
examined the impact of NAFLD on the risk of incident T2DM in non-overweight individuals.
Using a population-based retrospective cohort study of 4629 non-diabetic Japanese individuals
enrolled in a health check-up program for more than 10 years, the authors found that the incidence
rates of T2DM were 3.2% in the non-overweight without NAFLD group, 14.4% in the nonoverweight with NAFLD group, 8.0% in the overweight without NAFLD group and 26.4% in the
overweight with NAFLD group respectively. The adjusted HR in the non-overweight with NAFLD
group (HR 3.6, 95% CI 2.1–5.7) was higher than that in the overweight without NAFLD group (HR
2.0, 95% CI 1.5–2.7), or that in the non-overweight without NAFLD group (49).
Similarly, Sung et al. (44) have demonstrated in another occupational cohort study of over 12 000
Korean individuals that the clustering of IR (by homeostasis model assessment, HOMA-IR),
overweight/obesity, and NAFLD (as detected by ultrasonography) was common and markedly
increased the risk of incident T2DM. In the fully adjusted regression model, the HRs for incident
T2DM were 3.9 (95% CI 2.8–5.4) for HOMA-estimated insulin resistance, 1.6 (95% CI 1.2–2.2) for
overweight/obesity and 2.4 (95% CI 1.7–3.4) for fatty liver. The HR for the presence of all three
risk factors in the fully adjusted model was 14.1 (95% CI 8.9–22.2). Notably, in this same
observational cohort, Sung et al. (47) have further assessed the risk of incident T2DM at 5 years of
follow-up in the following three subgroups of subjects: (1) in those in whom there was resolution of
NAFLD over 5 years; (2) in those in whom there was a development of new NAFLD between
baseline and the follow-up examinations; and (3) in subjects in whom there was an increase in
severity of NAFLD, from mild fatty liver noted at baseline to moderate-severe fatty liver, identified
at follow-up examination.
Interestingly, these data showed that changing NAFLD status over a 5-year period was associated
with markedly different risks of incident T2DM. In particular, the individuals in whom the severity
of fatty liver worsened over time showed a marked increase in risk of incident T2DM compared
with the risk in people with resolution of fatty liver. Conversely, there was a significant T2DM risk
reduction in those individuals in whom NAFLD on ultrasonography resolved over time (47).
Similarly, in a cohort of 4604 Japanese non-diabetic individuals who were followed up for a mean
period of 11.3 years, Yamazaki et al. (48) reported that NAFLD at baseline was associated with an
approximately 2.5-fold increased risk of incident T2DM, and that NAFLD improvement was
associated with an approximately 70% risk reduction of developing T2DM, independently of
common risk factors. However, it is important to note that these two studies are not interventional
clinical trials (47, 48), so caution is needed in interpreting the data. Nevertheless, a recent
systematic review and meta-analysis of 20 observational studies, involving more than 115 000
individuals, found that NAFLD (as diagnosed either by abnormal serum liver enzymes or by
ultrasonography) was associated with an almost two-fold increased risk of incident T2DM over a
median period of 5 years (54). Although it would seem that there are subsets of NAFLD patients,
who have different levels of T2DM risk, there is to date only one small study also suggesting that
the presence of biopsy-confirmed NASH was a stronger risk factor of T2DM than simple steatosis.
In fact, in a retrospective cohort study of 129 Sweden adults with biopsy-proven NAFLD, Ekstedt
et al. (55) found that 80% of these patients developed either new onset T2DM or pre-diabetes, and
that patients with NASH had an approximately threefold higher risk of developing T2DM than
those with simple steatosis over a mean follow-up of 14 years. This finding is consistent with the
accepted notion that NASH patients tend to have a greater burden of metabolic diseases/T2DM, but
further studies in larger cohorts of patients with biopsy confirmed NAFLD are required to elucidate
whether the severity of NAFLD histology may differentially affect T2DM development.
Collectively, although there is now substantial evidence that NAFLD is strongly associated with
incident T2DM, risk and there is emerging evidence that this risk varies according to the severity of
NAFLD, further evidence is unquestionably needed, especially in non-Asian populations, as nearly
all of the studies reported in Table 1 – except for the recent study by Shah et al. –have been
conducted in various Asian populations. This finding might be important because the
anthropometric phenotypes and diets in Asian countries are very different from those in Western
countries. However, several population-based studies, involving American and European
individuals, have reported that mildly elevated serum liver enzymes are independent, long-term
predictors of new-onset T2DM (50). Some other caveats to keep in mind are that in none of the
published studies, with two exceptions (35, 47), T2DM diagnosis was based on 2-h post-load
plasma glucose levels, and that the adjustment for potential confounding variables was not always
complete. For example, data on waist circumference was often not reported, and family history of
T2DM, physical activity level or IR markers were not always included in multivariable
regression models. Prospective studies with a larger panel of known risk factors for T2DM
(including genetic variants) are required to firmly establish an independent contribution of NAFLD
to new-onset T2DM. Finally, controlled intervention trials will be necessary to determine whether
treating NAFLD/NASH leads to T2DM risk reduction.
Molecular features of NAFLD-related hepatic IR Hepatic IR is responsible for the increased
gluconeogenesis and HGP that predispose to T2DM development (see above), and is associated
with HFC. As alluded to earlier, HFC is determined by the balance of extrinsic (gastrointestinal
lipid absorption, adipose lipolysis), and intrinsic pathways – DNL, fatty acid oxidation and TG
secretion as VLDL (56). As the most parsimonious explanation for a shared genetic or
environmental determinant of NAFLD and IR is a cell-autonomous pathway, we focus this section
on hepatocyte-specific pathways. Of these, DNL accounts for as much as 30% of HFC in the
obese/insulin-resistant patient and thus becomes an attractive etiological possibility for NAFLDrelated IR. In fact, hepatic insulin signalling is intimately tied with the molecular machinery that
determines DNL (57), i.e. the production of long-chain fatty acids from two-carbon precursors
through sequential action of Acetyl CoA Carboxylase (ACC) and Fatty Acid Synthase (FAS),
followed by various fatty acid elongases and desaturases (58). Many of these enzymes are under
direct, albeit not exclusive, transcriptional control of SREBP-1c, which has the well-earned title of
master regulator of DNL and led to search for mechanistic determinants of SREBP-1c function.
SREBP-1c itself is regulated by fasting/refeeding, through hormonal action of insulin and glucagon
(59), although the specific pathways downstream of these hormones remain under active debate.
The most compelling evidence from this line of research implicates insulin-mediated activation
of the mechanistic target of rapamycin (mTOR), an evolutionarily conserved serine-threonine
kinase that links nutrient state to cellular growth. For instance, mice with liver-specific ablation of
the mTOR complex 1 (mTORC1)-defining subunit Raptor show reduced DNL, and are thus
protected from high-fat diet-induced fatty liver (60). Consistently, inhibition of mTOR/RAPTOR
interaction with rapamycin blocks insulin-dependent SREBP-1c activation in primary rat
hepatocytes and liver (61–63), leading to the widely accepted ‘bifurcation model’ whereby
mTORC1 is required for the lipogenic actions of insulin. Interestingly, mice lacking hepatocyte
Tuberous sclerosis complex-1 (TSC1), the endogenous inhibitor of mTORC1, also escape dietinduced fatty liver due to reduced AKT and SREBP-1c action (63, 64). AKT is a critical mediator
of insulin signal transduction (63–65), and mice lacking both hepatic isoforms (L-Akt1:Akt2) show
defects in feeding-induced SREBP-1c expression, similar to mice lacking hepatic insulin receptor
(INSR) (57, 66). L-Akt1:Akt2 mice also show marked glucose intolerance (57, 66), likely due to
loss of its inhibitory action on Forkhead O transcription factor-1 (FOXO1) (67). FOXO1, a member
of the FOXO family of transcription factors, is necessary for insulin suppression of HGP.
Physiological insulin levels prompt AKT mediated FOXO1 phosphorylation, leading to rapid
FOXO1 nuclear exclusion and degradation (68, 69). In the absence of insulin signalling, FOXOs are
transcriptionally active. In hepatocytes, the most consistently observed FOXO1-induced response
has been activation of Glucose-6-phosphatase (G6PC) expression (70, 71) in response to binding to
conserved insulin-regulatory sequences in the G6PC promoter sequence (72) coordinately with
NOTCH proteins (73). Although the bulk of FOXO studies in liver have thus justifiably focused on
its role in HGP, a parallel and consistent role in regulation of HFC has also emerged. Liver-specific
ablation of FOXO1 is associated with increased HFC (74, 75), an effect exacerbated by
simultaneous excision of all hepatic FOXOs and high-fat diet feeding (76). Expectedly, excess HFC
in FOXO null animals is secondary to a cell autonomous increase in DNL (76, 77), but the
molecular mechanism underlying this reproducible finding is still debate. It is likely that excess
DNL may be the endresult of multiple, simultaneous processes, as FOXO deficient hepatocytes
show disrupted autophagy (78) and altered NAD+ levels (76). However, an alternative hypothesis
has emerged that an altered substrate flux due to unbalanced action of both G6PC, mediating
glucose release from hepatocytes into the circulation (decreased), and glucokinase (GCK), which
facilitates glucose entry from the extracellular space and supports DNL (increased), may cause the
excess HFC in FOXO null animals (77). This finding would be consistent with the on-target hepatic
steatosis seen with administration of GCK activators (79, 80), while remaining are consistent with
the bifurcation model of hepatic insulin action on DNL and HGP (64).
Great progress has been made in identifying upstream AKT activators and downstream targets (81,
82), but mechanisms underlying the termination of the AKT signal have been understudied until the
discovery of the Pleckstrin homology (PH) domain leucine-rich repeat protein phosphatase
(PHLPP) gene family (83, 84). PHLPPs were first shown to slow AKT-dependent cancer
progression by dephosphorylating AKT at Ser473 (85–88), and have recently been proven to impact
obesity-related metabolic disease. Obese mice show a loss of PHLPP2 stability, leading to an
inability to ‘terminate’ AKT signalling – this causes a prolongation of insulin action – and
secondary induction of SREBP-1c, an effect that can be reversed by adenoviral PHLPP2 expression
(89). Consistently, knockdown of PHLPP2 in lean mice causes AKT hyperactivity, excess DNL and
fatty liver (89), similar to mice with constitutive active induction of AKT activity (90–92),
suggesting that endogenous PHLPP2 activity is required to terminate AKT signaling and avoid
excessive lipogenesis. In combination with data from liver-specific RICTOR knockout mice (92),
which are unable to maximally activate AKT due to the inability to phosphorylate ALT at Ser473,
and show consequent glucose intolerance but reduced HFC, a revised ‘bifurcation’ model has
emerged whereby AKT must be activated to inhibit FOXO-induced HGP, but also inactivated in a
timely fashion to avoid excess DNL (Fig. 1). Of note, in almost all of the above mouse models,
discordance was detected between DNL and HGP, and thus between HFC and circulating glucose
levels. Disruption of canonical insulin signaling by means of ablation of INSR (57), IRS (93) or
AKT (94), leads to increased HGP, but reduced DNL. Increased ‘late’ AKT activity by means of
altered PHLPP2, or reduced mTORC1 function by ablation of RAPTOR, reduces DNL, without
changes to HGP. Ablation of the key regulator of insulin-regulated HGP, FOXO, increased DNL.
These loss-of-function mouse models are obviously extreme examples, with complete absence of
critical signalling components, and moderate down-regulation of hepatic INSR expression, which is
associated with preserved regulation of glucose metabolism but increased cell-autonomous DNL,
may better recapitulate human NAFLD and T2DM (95). In sum, however, data from these
genetically manipulated mice do not support a role of hepatic DNL as a primary determinant of
hepatic IR. These data are echoed by mouse models of diet-induced NASH, methionine/cholinedeficient diet (MCDD)-feeding. MCDD-fed mice develop macrovesicular steatosis, lobular
inflammation and fibrosis (96), but tend to lose body weight and, as such, are more insulin-sensitive
than control animals (97), proving that excess HFC alone is insufficient to induce glucose
intolerance. Of course, this does not rule out the potential contribution of excess hepatic DAG/
ceramides to exacerbate systemic glucose homeostasis, or cell non-autonomous effects, as discussed
earlier. The key molecular features of unbalanced activation of hepatic insulin signalling pathways
have been confirmed in clinical samples of patients (12). Decreased expression of INSR protein has
indeed been observed in the liver of patients with NAFLD. Consistently, despite compensatory
hyperinsulinaemia, decreased INSR expression was accompanied by reduced phosphorylation status
of AKT1 and increased FOXO1 activity (98, 99).
In the liver of patients with NAFLD, reduced insulin signalling through AKT1 has been associated
with both enhanced oxidative stress (98) and increased DAG concentrations, leading to activation of
atypical protein kinase C isoforms (e.g., PKCe) able to contrast insulin signalling (100). However,
recent association studies and experimental data point to C:16-ceramides (101, 102) and possibly
lyso-phosphatidyl-cholines (103, 104) as important lipid mediators of hepatic IR in NAFLD.
While insulin ability to restrain HGP is impaired in NAFLD, it has been observed that FOXO1dependent IRS2 is up-regulated with preservation of the activation status of AKT2 and of induction
of SREBP-1c and lipogenic pathways (99). The substrate for DNL was linked to induction of GCK
by SREBP-1c (105), parallelingincreasing HFC (99, 106). In patients with NASH, this was
accompanied by activation of NOTCH signalling, which correlated with the severity of hepatic
steatosis, IR and gluconeogenesis (107).
NAFLD as a consequence of IR and T2DM: lesson from genetic studies During the last years,
genome wide association studies revealed the major genetic loci, whose common variants
predispose to NAFLD (108). We believe that this information will prove helpful in disentangling
the complex link between HFC and T2DM. According to a simple syllogism, if a genetic risk
variant causes NAFLD and NAFLD causes T2DM, then this genetic variant must also cause T2DM
with a comparable effect size. Indeed, in the case of validated germline genetic variants with a
proven causal role in NAFLD, reverse causation (because they are inherited and cannot be
modified) and confounding factors (because they are randomly inherited from the parents) cannot
bias the association with the risk of IR and T2DM. This is the principle underpinning Mendelian
randomization analysis, which is used to dissect possible causal relationships in association studies
(Fig. 2).
The major common genetic risk variant for NAFLD is undoubtedly represented by the I148M
variant in the PNPLA3 gene (109, 110), which is carried by one in three individuals in the general
population and by one in two patients with NAFLD. The PNPLA3 protein is a lipase that
hydrolyses triglycerides in hepatocytes and retinyl-esters in hepatic stellate cells, and is thus
involved in the regulation of lipid metabolism and several cell functions. The I148M loss-offunction variant predisposes to increased HFC and progressive liver disease in the presence of
environmental ‘stressors’, such asIR (111–113). The underlying mechanisms are still under debate,
but likely involve accumulation of the 148M protein variant on the surface of lipid droplets,
hindering access to other lipases, resulting in altered TG remodelling and retinol release (111, 114–
116), and influencing hepatic and circulating lipid profile (101, 117, 118). However, despite its
large effect on HFC, the I148M variant does not seem to markedly impact on systemic/hepatic IR,
NEFA levels, or the risk of T2DM (109, 110). Carriers of the I148M variant may have a small
increase in T2DM risk, mainly in severely obese patients (119). However, this marginal effect is
much smaller than the expected impact of increased HFC and is possibly mediated by
predisposition to more severe liver disease in these patients. This is not a unique example. The
Transmembrane 6 Superfamily Member 2 (TM6SF2) rs58542926 C>T E167K variant has more
recently been associated with NAFLD (10, 120). In this case, the mechanism is likely to be related
to hepatocellular retention of lipids due to altered lipidation of nascent VLDL (10, 120–123).
Similar to PNPLA3 I148M, the E167K variant does not alter T2DM risk (124). Similarly, the
Membrane-bound OAcyl Transferase protein 7 (MBOAT7) rs641738 C>T gene variant, which
regulates arachidonoyl-phosphatidyl-inositol metabolism, and circulating phosphatidylinositol
composition, has now been associated with NAFLD development and progression, without any
significant impact on T2DM prevalence (125). Another validated locus influencing NAFLD risk is
the Glucokinase regulator (GCKR) gene (126–128). The common P446L variant seems to
predispose to increased HFC by deregulation of glucose uptake and DNL in hepatocytes. In the
presence of the P446L variant, compartmentalization of energy as liver TG synthesized from
circulating glucose favours liver damage, but protects from T2DM development (126, 129, 130).
Since GCKR directly influences glucose metabolism, this cannot be generalized to conclude that
increased ability to store energy as TG in the liver protects from hyperglycaemia, and future studies
evaluating comprehensive genetic risk scores will be needed to better clarify this issue.
Furthermore, it cannot yet be ruled out that different forms of NAFLD characterized by
accumulation of specific lipids favouring IR, such as ceramides, may cause T2DM (101).
Rare loss-of-function variants in APOB responsible for familial hypobetalipoproteinaemia (FHBL)
provide the clearest example of the dissociation between HFC and IR, and underscore the
independent effect of severe HFC on liver disease progression (131). FHBL is indeed associated
with severe steatosis due to defective export of lipids from hepatocytes by VLDL (which requires
ApoB100 protein for correct secretion), progressive fibrosis and hepatocellular carcinoma (131,
132). However, FHBL is neither associated with systemic and hepatic IR, nor with increased NEFA
flux to the liver, nor with T2DM (36, 131–133). Therefore, although organ-specific deposition of fat
is a strong predictor of IR, HFC per se does not invariably result in IR and T2DM (as summarized
in Table 2), but in progressive NAFLD, which suggests that NAFLD may be a frequent
consequence rather than a cause of IR and T2DM, or that progressive NAFLD and T2DM may
represent two different outcomes stemming from IR (see below). In keeping with this view, genetic
variants that impair hepatic insulin signalling have been associated with hepatic fibrosis, as if IR
favors progressive NAFLD (134). Consistently, a new therapeutic approach with the new farnesoid
X receptor (FXR) agonist, obeticholic acid, reduced both HFC and liver damage in patients with
NASH, but it was associated with worsening of IR (135).
A unifying hypothesis Insulin resistance, compensatory hyperinsulinaemia and visceral obesity
contribute to the development of NAFLD first; in return, the insulin resistant fatty liver
overproduces glucose and VLDL thus boosting the mechanisms that lead to with progressive
exhaustion of pancreatic b cell reserve, leading to T2DM development at later stages of life. This
finding would explain the epidemiological observations of a strong association between the
presence and severity of NAFLD and the risk of developing incident T2DM, without implying that
HFC per se plays a causal role. In this context energy storage under as TG in lipid droplets in
hepatocytes may be considered as an attempt to protect the liver from lipotoxicity induced by
NEFA and from hyperinsulinaemia-driven DNL, and subjects who display a better ability to clear
fat from the liver would probably be those at lower risk of developing long-term hepatic
complications. At the same time, systemic IR and hypo-adiponectinaemia may contribute to the
development of both hepatic IR and T2DM, as adiponectin receptors have ceramidase activity
(136). A working model for the hypothesized mechanisms leading from IR/MetS to parallel
NAFLD and to T2DM pathogenesis is presented in Figure 3. The practical implications of this
model are the following: (a) HFC is strongly linked with metabolic derangements, independently of
many common and measurable risk factors, and thus represents an excellent marker of incident
T2DM risk (at least in patients without strong coexisting genetic risk factors for NAFLD); (b)
research attention should focus on specific lipid species favouring hepatic IR rather than on the
amount of HFC per se; and (c) novel therapeutic approaches for T2DM should be aimed to decrease
these specific lipid species. Indeed, DNL inhibition and steatosis reduction might benefit liver
disease progression, but may worsen glycaemic control. However, further prospective and
mechanistic studies are needed to better elucidate these fascinating issues.
Conclusions. It has become increasingly clear that there are complicated, bidirectional links
between NAFLD and T2DM. These two diseases share many aspects of their pathophysiology, and
although IR is at the centre of both, growing evidence suggests divergent pathogenic effects of IR
on NAFLD and T2DM. However, many aspects of the close interactions between NAFLD and
IR/T2DM are not yet fully elucidated yet, as genetic risk factors for increased HFC have revealed
that there is a disconnection between these two conditions. We believe that a closer focus on the
mechanisms that underlie these observations will identify new therapeutic targets not only for liver
disease, but also for T2DM and other metabolic diseases.
Financial support: LV was supported by my First AIRC grant no 16888, and Fondazione IRCCS
Ca’ Granda INGM molecular medicine grant. Conflict of interest: The authors do not have any
disclosures to report.
Figures legend
Fig. 1. Revised mechanism underlying the ‘bifurcation’ of hepatocyte insulin signalling
underpinning the pathogenesis of progressive NAFLD (increased hepatic lipogenesis) and type 2
diabetes (increased hepatic glucose production). FOXOs, forkhead box O transcription factors;
INSR, insulin receptor; IRS, insulin receptor substrate; mTORC, mammalian target of rapamycin;
PHLPP2, pleckstrin homology domain leucine-rich repeat protein phosphatase 2; PI3, inositol-3phosphate; SREBP-1c, sterol regulatory element binding protein-1c.
Fig. 2. Mendelian randomization for determining the risk of insulin resistance and type 2 diabetes
related to NAFLD: a qualitative approach. Mendelian randomization is an instrumental variable
approach to infer causality in association studies in the presence of potential confounding and
reverse causation. The modifiable exposure in this case is hepatic fat content. Genetic risk variants
for NAFLD are used as the instrument. Causal effects of hepatic fat content on insulin resistance
(IR) and type 2 diabetes (T2DM) are estimated by examining the genetic risk variants for
association with hepatic fat content, as well as with IR/T2DM using a triangulation approach. The
association between hepatic fat content and IR/T2DM can arise from both directions. However, as
genetic risk factors are strongly associated with hepatic fat content, but there is no predisposing
effect on IR/T2DM risk, we might conclude that hepatic fat content does not directly cause
IR/T2DM, but the opposite should held true. This does not exclude that NAFLD may be caused
exclusively by strong genetic or other non-genetic factors independently of coexisting IR/T2DM.
Fig. 3. A unifying hypothesis. Major triggers of insulin resistance, NAFLD and type 2 diabetes
mellitus (T2DM) are Western diet, physical inactivity favoring sarcopenia and reduced fitness, and
expanded visceral adipose tissue with development of subclinical inflammation (bottom part of the
cartoon). This, in turn, leads to increased non-esterified fatty acids (NEFA) influx to the liver and
hyperinsulinaemia, which trigger NAFLD in susceptible individuals. At the same time, increased de
novo lipogenesis (DNL) and reduced adiponectin levels can induce accumulation of lipid
intermediates (like diacylglycerol, ceramides and fatty acyl-CoAs) that are associated with hepatic
insulin resistance. The outcome of these processes is two-fold (upper part): on one hand,
lipotoxicity from NEFA spillover from lipid droplets and free cholesterol, together with
inflammatory stimuli from the gut, may lead to fibrosis in NAFLD and advanced liver disease; on
the other hand, increased hepatic gluconeogenesis drives hyperglycemia, which contributes to
T2DM development, when the capacity of pancreatic b-cells to compensate is exhausted. Upregulated pathways are shown by solid arrows, down-regulated pathways by dashed arrows.
Inhibitory pathways are indicated by dots. CE, cholesterol esters; DAG, diacyl-glycerol; DNL, de
novo lipogenesis; FOXO1, forkhead O box transcription factor 1; GCK, glucokinase; GNG,
gluconeogenesis; G6PC, glucose-6-phosphatase; INSR, insulin resceptor; NEFA, non-esterified
fatty acids; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; TGs,
triglycerides; TLR4, toll-like receptor-4; T2DM, type 2 diabetes mellitus.
References
1. Bugianesi E, Leone N, Vanni E, et al. Expanding the natural history of nonalcoholic
steatohepatitis: from cryptogenic cirrhosis to hepatocellular carcinoma. Gastroenterology
2002; 123: 134–40.
2. Dyson J, Jaques B, Chattopadyhay D, et al. Hepatocellular cancer: the impact of obesity, type 2
diabetes and a multidisciplinary team. J Hepatol 2014; 60: 110–7.
3. Charlton MR, Burns JM, Pedersen RA, et al. Frequency and outcomes of liver transplantation for
nonalcoholic steatohepatitis in the United States. Gastroenterology 2011; 141: 1249–53.
4. Ruhl CE, Everhart JE. Determinants of the association of overweight with elevated serum alanine
aminotransferase activity in the United States. Gastroenterology 2003; 124:71–9.
5. Browning JD, Szczepaniak LS, Dobbins R, et al. Prevalence of hepatic steatosis in an urban
population in the United States: impact of ethnicity. Hepatology 2004; 40:1387–95.
6. Non-Alcoholic Fatty Liver Disease Study G, Lonardo A, Bellentani S, Argo CK, Ballestri S,
Byrne CD, et al. Epidemiological modifiers of non-alcoholic fatty liver disease: focus on high-risk
groups. Dig Liver Dis 2015; 47:997–1006.
7. Marchesini G, Brizi M, Morselli-Labate AM, et al. Association of nonalcoholic fatty liver disease
with insulin resistance. Am J Med 1999; 107: 450–5.
8. Marchesini G, Brizi M, Bianchi G, et al. Nonalcoholic fatty liver disease: a feature of the
metabolic syndrome. Diabetes 2001; 50: 1844–50.
9. Marchesini G, Bugianesi E, Forlani G, et al. Nonalcoholic fatty liver, steatohepatitis, and the
metabolic syndrome. Hepatology 2003; 37: 917–23.
10. Dongiovanni P, Petta S, Maglio C, et al. Transmembrane 6 superfamily member 2 gene variant
disentangles nonalcoholic steatohepatitis from cardiovascular disease. Hepatology 2015; 61: 50614.
11. Dongiovanni P, Romeo S, Valenti L. Hepatocellular carcinoma in nonalcoholic fatty liver: role
of environmentaland genetic factors. World J Gastroenterol 2014; 20:12945–55.
12. Dongiovanni P, Rametta R, Meroni M, Valenti L. The role of insulin resistance in nonalcoholic
steatohepatitis and liver disease development – a potential therapeutic target? Expert Rev
Gastroenterol Hepatol 2015; 10: 1–14.
13. Yki-Jarvinen H. Non-alcoholic fatty liver disease as a cause and a consequence of metabolic
syndrome. Lancet Diabetes Endocrinol 2014; 2: 901–10.
14. Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes 1988;
37: 1595–607.
15. Shulman GI. Ectopic fat in insulin resistance, dyslipidemia, and cardiometabolic disease. N
Engl J Med 2014;371: 1131–41.
16. Roden M, Stingl H, Chandramouli V, et al. Effects of free fatty acid elevation on postabsorptive
endogenous glucose production and gluconeogenesis in humans. Diabetes 2000; 49: 701–7.
17. Gastaldelli A, Baldi S, Pettiti M, et al. Influence of obesity and type 2 diabetes on
gluconeogenesis and glucose output in humans: a quantitative study. Diabetes 2000; 49:1367–73.
18. Adkins A, Basu R, Persson M, et al. Higher insulin concentrations are required to suppress
gluconeogenesis than glycogenolysis in nondiabetic humans. Diabetes 2003; 52:2213–20.
19. Kantartzis K, Machicao F, Machann J, et al. The DGAT2 gene is a candidate for the
dissociation between fatty liver and insulin resistance in humans. Clin Sci 2009; 116:531–7.
20. Seppala-Lindroos A, Vehkavaara S, Hakkinen AM, et al.Fat accumulation in the liver is
associated with defects in insulin suppression of glucose production and serum free fatty acids
independent of obesity in normal men. J Clin Endocrinol Metab 2002; 87: 3023–8.
21. Bugianesi E, Gastaldelli A, Vanni E, et al. Insulin resistance in non-diabetic patients with nonalcoholic fatty liver disease: sites and mechanisms. Diabetologia 2005;48: 634–42.
22. Defronzo RA. Banting Lecture. From the triumvirate to the ominous octet: a new paradigm for
the treatment of type 2 diabetes mellitus. Diabetes 2009; 58: 773–95.
23. Ferrannini E, Mari A. beta-Cell function in type 2 diabetes. Metabolism 2014; 63: 1217–27.
24. Donnelly KL, Smith CI, Schwarzenberg SJ, et al. Sources of fatty acids stored in liver and
secreted via lipoproteins in patients with nonalcoholic fatty liver disease. J Clin Invest 2005; 115:
1343–51.
25. Targher G, Day CP, Bonora E. Risk of cardiovascular disease in patients with nonalcoholic fatty
liver disease. N Engl J Med 2010; 363: 1341–50.
26. Fabbrini E, Magkos F, Mohammed BS, et al. Intrahepatic fat, not visceral fat, is linked with
metabolic complications of obesity. Proc Natl Acad Sci USA 2009; 106:15430–5.
27. Gastaldelli A, Cusi K, Pettiti M, et al. Relationship between hepatic/visceral fat and hepatic
insulin resistance in nondiabetic and type 2 diabetic subjects. Gastroenterology 2007; 133: 496–506.
28. Gaggini M, Morelli M, Buzzigoli E, et al. Non-alcoholic fatty liver disease (NAFLD) and its
connection with insulin resistance, dyslipidemia, atherosclerosis and coronary heart disease.
Nutrients 2013; 5: 1544–60.
29. Frayn KN. Visceral fat and insulin resistance–causative or correlative? Br J Nutr 2000;
83(Suppl. 1): S71–7.
30. Gastaldelli A. Role of beta-cell dysfunction, ectopic fat accumulation and insulin resistance in
the pathogenesis of type 2 diabetes mellitus. Diabetes Res Clin Pract 2011;93(Suppl. 1): S60–5.
31. Kadowaki T, Yamauchi T. Adiponectin and adiponectin receptors. Endocr Rev 2005; 26: 43951.
32. Polyzos SA, Toulis KA, Goulis DG, Zavos C, Kountouras J. Serum total adiponectin in
nonalcoholic fatty liver disease: a systematic review and meta-analysis. Metabolism 2011; 60: 313–
26.
33. Marra F, Bertolani C. Adipokines in liver diseases. Hepatology 2009; 50: 957–69.
34. Petrides AS, Groop LC, Riely CA, Defronzo RA. Effect of physiologic hyperinsulinemia on
glucose and lipid metabolism in cirrhosis. J Clin Invest 1991; 88: 561–70.
35. Targher G, Bertolini L, Rodella S, et al. Non-alcoholic fatty liver disease is independently
associated with an increased prevalence of chronic kidney disease and proliferative/laser-treated
retinopathy in type 2 diabetic patients. Diabetologia 2008; 51: 444–50.
36. Dongiovanni P, Romeo S, Valenti L. Genetic factors in the pathogenesis of nonalcoholic fatty
liver and steatohepatitis. Biomed Res Int 2015; 2015: 460190.
37. Okamoto M, Takeda Y, Yoda Y, et al. The association of fatty liver and diabetes risk. J
Epidemiol 2003; 13: 15–21.
38. Fan JG, Li F, Cai XB, et al. Effects of nonalcoholic fatty liver disease on the development of
metabolic disorders. J Gastroenterol Hepatol 2007; 22: 1086–91.
39. Shibata M, Kihara Y, Taguchi M, Tashiro M, Otsuki M. Nonalcoholic fatty liver disease is a
risk factor for type 2 diabetes in middle-aged Japanese men. Diabetes Care 2007; 30: 2940–4.
40. Kim CH, Park JY, Lee KU, Kim JH, Kim HK. Fatty liver is an independent risk factor for the
development of Type 2 diabetes in Korean adults. Diabet Med 2008; 25: 476–81.
41. Yamada T, Fukatsu M, Suzuki S, et al. Fatty liver predicts impaired fasting glucose and type 2
diabetes mellitus in Japanese undergoing a health checkup. J Gastroenterol Hepatol 2010; 25: 352–
6.
42. Bae JC, Rhee EJ, Lee WY, et al. Combined effect of nonalcoholic fatty liver disease and
impaired fasting glucose on the development of type 2 diabetes: a 4-year retrospective longitudinal
study. Diabetes Care 2011; 34: 727–9.
43. Park SK, Seo MH, Shin HC, Ryoo JH. Clinical availability of nonalcoholic fatty liver disease as
an early predictor of type 2 diabetes mellitus in Korean men: 5-year prospective cohort study.
Hepatology 2013; 57: 1378–83.
44. Sung KC, Jeong WS, Wild SH, Byrne CD. Combined influence of insulin resistance,
overweight/obesity, and fatty liver as risk factors for type 2 diabetes. Diabetes Care 2012; 35: 717–
22.
45. Kasturiratne A, Weerasinghe S, Dassanayake AS, et al. Influence of non-alcoholic fatty liver
disease on the development of diabetes mellitus. J Gastroenterol Hepatol 2013; 28: 142–7.
46. Chang Y, Jung HS, Yun KE, et al. Cohort study of nonalcoholic fatty liver disease, NAFLD
fibrosis score, and the risk of incident diabetes in a Korean population. Am J Gastroenterol 2013;
108: 1861–8.
47. Sung KC, Wild SH, Byrne CD. Resolution of fatty liver and risk of incident diabetes. J Clin
Endocrinol Metab 2013; 98: 3637–43.
48. Yamazaki H, Tsuboya T, Tsuji K, Dohke M, Maguchi H. Independent association between
improvement of nonalcoholic fatty liver disease and reduced incidence of Type 2 diabetes. Diabetes
Care 2015; 38: 1673–9.
49. Fukuda T, Hamaguchi M, Kojima T, et al. The impact of non-alcoholic fatty liver disease on
incident type 2 diabetes mellitus in non-overweight individuals. Liver Int 2016; 36: 275–83.
50. Shah RV, Allison MA, Lima JA, et al. Liver fat, statin use, and incident diabetes: the MultiEthnic Study of Atherosclerosis. Atherosclerosis 2015; 242: 211–7.
51. Ming J, Xu S, Gao B, et al. Non-alcoholic fatty liver disease predicts type 2 diabetes mellitus,
but not prediabetes, in Xi’an, China: a five-year cohort study. Liver Int 2015;35: 2401–7.
52. Li WD, Fu KF, Li GM, et al. Comparison of effects of obesity and non-alcoholic fatty liver
disease on incidence of type 2 diabetes mellitus. World J Gastroenterol 2015;21: 9607–13.
53. Choi JH, Rhee EJ, Bae JC, et al. Increased risk of type 2 diabetes in subjects with both elevated
liver enzymes and ultrasonographically diagnosed nonalcoholic fatty liver disease: a 4-year
longitudinal study. Arch Med Res 2013;44: 115–20.
54. Ballestri S, Zona S, Targher G, et al. Nonalcoholic fatty liver disease is associated with an
almost twofold increased risk of incident type 2 diabetes and metabolic syndrome. Evidence from a
systematic review and metaanalysis. J Gastroenterol Hepatol 2016; 31: 936–44.
55. Ekstedt M, Franzen LE, Mathiesen UL, et al. Long-term follow-up of patients with NAFLD and
elevated liver enzymes. Hepatology 2006; 44: 865–73.
56. Savage DB, Semple RK. Recent insights into fatty liver, metabolic dyslipidaemia and their links
to insulin resistance. Curr Opin Lipidol 2010; 21: 329–36.
57. Haas JT, Miao J, Chanda D, et al. Hepatic insulin signaling is required for obesity-dependent
expression of SREBP-1c mRNA but not for feeding-dependent expression. Cell Metab 2012; 15:
873–84.
58. Postic C, Girard J. The role of the lipogenic pathway in the development of hepatic steatosis.
Diabetes Metab 2008; 34(Pt 2): 643–8.
59. Horton JD, Goldstein JL, Brown MS. SREBPs: activators of the complete program of
cholesterol and fatty acid synthesis in the liver. J Clin Invest 2002; 109: 1125–31.
60. Peterson TR, Sengupta SS, Harris TE, et al. mTOR complex 1 regulates lipin 1 localization to
control the SREBP pathway. Cell 2011; 146: 408–20.
61. Porstmann T, Santos CR, Griffiths B, et al. SREBP activity is regulated by mTORC1 and
contributes to Akt dependent cell growth. Cell Metab 2008; 8: 224–36.
62. Brown NF, Stefanovic-Racic M, Sipula IJ, Perdomo G. The mammalian target of rapamycin
regulates lipid metabolism in primary cultures of rat hepatocytes. Metabolism 2007; 56: 1500–7.
63. Yecies JL, Zhang HH, Menon S, et al. Akt stimulates hepatic SREBP1c and lipogenesis through
parallel mTORC1-dependent and independent pathways. Cell Metab 2011;14: 21–32.
64. Li S, Brown MS, Goldstein JL. Bifurcation of insulin signaling pathway in rat liver: mTORC1
required for stimulation of lipogenesis, but not inhibition of gluconeogenesis. Proc Natl Acad Sci
USA 2010; 107:3441–6.
65. Taniguchi CM, Emanuelli B, Kahn CR. Critical nodes in signalling pathways: insights into
insulin action. Nat Rev Mol Cell Biol 2006; 7: 85–96.
66. Lu M, Wan M, Leavens KF, et al. Insulin regulates liver metabolism in vivo in the absence of
hepatic Akt and Foxo1. Nat Med 2012; 18: 388–95.
67. Lin HV, Accili D. Hormonal regulation of hepatic glucose production in health and disease.
Cell Metab 2011; 14:9–19.
68. Nakae J, Kitamura T, Silver DL, Accili D. The forkhead transcription factor Foxo1 (Fkhr)
confers insulin sensitivity onto glucose-6-phosphatase expression. J Clin Invest 2001; 108: 1359–
67.
69. Frescas D, Valenti L, Accili D. Nuclear trapping of the forkhead transcription factor FoxO1 via
Sirt-dependent deacetylation promotes expression of glucogenetic genes. J Biol Chem 2005; 280:
20589–95.
70. Schmoll D, Walker KS, Alessi DR, et al. Regulation of glucose-6-phosphatase gene expression
by protein kinase Balpha and the forkhead transcription factor FKHR. Evidence for insulin response
unit-dependent and –independent effects of insulin on promoter activity. J Biol Chem 2000; 275:
36324–33.
71. Zhang W, Patil S, Chauhan B, et al. FoxO1 regulates multiple metabolic pathways in the liver:
effects on gluconeogenic, glycolytic, and lipogenic gene expression. J Biol Chem 2006; 281:
10105–17.
72. O’Brien RM, Granner DK. Regulation of gene expression by insulin. Physiol Rev 1996; 76:
1109–61.
73. Pajvani UB, Shawber CJ, Samuel VT, et al. Inhibition of Notch signaling ameliorates insulin
resistance in a FoxO1-dependent manner. Nat Med 2011; 17: 961–7.
74. Pajvani UB, Qiang L, Kangsamaksin T, et al. Inhibition of Notch uncouples Akt activation from
hepatic lipid accumulation by decreasing mTorc1 stability. Nat Med 2013; 19: 1054–60.
75. Zhang K, Li L, Qi Y, et al. Hepatic suppression of Foxo1 and Foxo3 causes hypoglycemia and
hyperlipidemia in mice. Endocrinology 2012; 153: 631–46.
76. Tao R, Wei D, Gao H, et al. Hepatic FoxOs regulate lipid metabolism via modulation of
expression of the nicotinamide phosphoribosyltransferase gene. J Biol Chem 2011; 286: 14681–90.
77. Haeusler RA, Hartil K, Vaitheesvaran B, et al. Integrated control of hepatic lipogenesis versus
glucose production requires FoxO transcription factors. Nat Commun 2014;5: 5190.
78. Liu HY, Han J, Cao SY, et al. Hepatic autophagy is suppressed in the presence of insulin
resistance and hyperinsulinemia:inhibition of FoxO1-dependent expression of key autophagy genes
by insulin. J Biol Chem 2009; 284:31484–92.
79. Ceuninck FD, Kargar C, Charton Y, et al. S 50131 and S 51434, two novel small molecule
glucokinase activators, lack chronic efficacy despite potent acute antihyperglycaemic activity in
diabetic mice: lipids as usual suspects. Br J Pharmacol 2013; 169: 999–1010.
80. De Ceuninck F, Kargar C, Ilic C, et al. Small molecule glucokinase activators disturb lipid
homeostasis and induce fatty liver in rodents: a warning for therapeuticapplications in humans. Br J
Pharmacol 2013; 168:339–53.
81. Sarbassov DD, Guertin DA, Ali SM, Sabatini DM. Phosphorylation and regulation of Akt/PKB
by the rictormTOR complex. Science 2005; 307: 1098–101.
82. Alessi DR, Pearce LR, Garcia-Martinez JM. New insights into mTOR signaling: mTORC2 and
beyond. Sci Signal 2009; 2: pe27.
83. Gao T, Furnari F, Newton AC. PHLPP: a phosphatase that directly dephosphorylates Akt,
promotes apoptosis, and suppresses tumor growth. Mol Cell 2005; 18: 13–24.
84. Brognard J, Sierecki E, Gao T, Newton AC. PHLPP and a second isoform, PHLPP2,
differentially attenuate the amplitude of Akt signaling by regulating distinct Akt isoforms. Mol Cell
2007; 25: 917–31.
85. Newton AC, Trotman LC. Turning off AKT: PHLPP as a drug target. Annu Rev Pharmacol
Toxicol 2014; 54:537–58.
86. Li X, Stevens PD, Liu J, et al. PHLPP is a negative regulator of RAF1, which reduces colorectal
cancer cell motility and prevents tumor progression in mice. Gastroenterology 2014; 146: 1301–
12.e10.
87. Ghalali A, Ye ZW, Hogberg J, Stenius U. Phosphatase and tensin homolog deleted on
chromosome 10 (PTEN) and PH domain and leucine-rich repeat phosphatase cross-talk (PHLPP) in
cancer cells and in transforming growth factor beta-activated stem cells. J Biol Chem 2014;289:
11601–15.
88. Wang P, Zhou Z, Hu A, et al. Both decreased and increased SRPK1 levels promote cancer by
interfering with PHLPP-mediated dephosphorylation of Akt. Mol Cell 2014; 54: 378–91.
89. Kim K, Qiang L, Hayden MS, et al. mTORC1-independent Raptor prevents hepatic steatosis by
stabilizing PHLPP2. Nat Commun 2016; 7: 10255.
90. Calvisi DF, Wang C, Ho C, et al. Increased lipogenesis, induced by AKT-mTORC1-RPS6
signaling, promotes development of human hepatocellular carcinoma. Gastroenterology 2011; 140:
1071–83.
91. Li L, Wang C, Calvisi DF, et al. SCD1 Expression is dispensable for hepatocarcinogenesis
induced by AKT and Ras oncogenes in mice. PLoS ONE 2013; 8: e75104.
92. Hagiwara A, Cornu M, Cybulski N, et al. Hepatic mTORC2 activates glycolysis and lipogenesis
through Akt, glucokinase, and SREBP1c. Cell Metab 2012; 15:725–38.
93. Guo S, Copps KD, Dong X, et al. The Irs1 branch of the insulin signaling cascade plays a
dominant role in hepatic nutrient homeostasis. Mol Cell Biol 2009; 29: 5070–83.
94. Leavens KF, Easton RM, Shulman GI, Previs SF, Birnbaum MJ. Akt2 is required for hepatic
lipid accumulation in models of insulin resistance. Cell Metab 2009; 10: 405–18.
95. Cook JR, Langlet F, Kido Y, Accili D. Pathogenesis of selective insulin resistance in isolated
hepatocytes. J Biol Chem 2015; 290: 13972–80.
96. Kohli R, Feldstein AE. NASH animal models: are we there yet? J Hepatol 2011; 55: 941–3.
97. Rinella ME, Elias MS, Smolak RR, et al. Mechanisms of hepatic steatosis in mice fed a
lipogenic methionine choline-deficient diet. J Lipid Res 2008; 49: 1068–76.
98. Valenti L, Rametta R, Dongiovanni P, et al. Increased expression and activity of the
transcription factor FOXO1 in nonalcoholic steatohepatitis. Diabetes 2008; 57:1355–62.
99. Rametta R, Mozzi E, Dongiovanni P, et al. Increased insulin receptor substrate 2 expression is
associated with steatohepatitis and altered lipid metabolism in obese subjects. Int J Obes 2013; 37:
986–92.
100. Kumashiro N, Erion DM, Zhang D, et al. Cellular mechanism of insulin resistance in
nonalcoholic fatty liver disease. Proc Natl Acad Sci USA 2011; 108: 16381–5.
101. Luukkonen PK, Zhou Y, Sadevirta S, et al. Ceramides dissociate steatosis and insulin
resistance in the human liver in non-alcoholic fatty liver disease. J Hepatol 2016;64: 1167–75.
102. Raichur S, Wang ST, Chan PW, et al. CerS2 haploinsufficiency inhibits beta-oxidation and
confers susceptibility to diet-induced steatohepatitis and insulin resistance. Cell Metab 2014; 20:
687–95.
103. Oresic M, Hyotylainen T, Kotronen A, et al. Prediction of non-alcoholic fatty-liver disease and
liver fat content by serum molecular lipids. Diabetologia 2013; 56: 2266–74.
104. Lehmann R, Franken H, Dammeier S, et al. Circulating lysophosphatidylcholines are markers
of a metabolically benign nonalcoholic fatty liver. Diabetes Care 2013; 36:2331–8.
105. Kim SY, Kim HI, Kim TH, et al. SREBP-1c mediates the insulin-dependent hepatic
glucokinase expression. J Biol Chem 2004; 279: 30823–9.
106. Peter A, Stefan N, Cegan A, et al. Hepatic glucokinase expression is associated with
lipogenesis and fatty liver in humans. J Clin Endocrinol Metab 2011; 96:E1126–30.
107. Valenti L, Mendoza RM, Rametta R, et al. Hepatic notch signaling correlates with insulin
resistance and nonalcoholic Fatty liver disease. Diabetes 2013; 62: 4052–62.
108. Dongiovanni P, Valenti L. Genetics of nonalcoholic fatty liver disease. Metabolism 2016; 65:
1026–37.
109. Romeo S, Kozlitina J, Xing C, et al. Genetic variation in PNPLA3 confers susceptibility to
nonalcoholic fatty liver disease. Nat Genet 2008; 40: 1461–5.
110. Sookoian S, Pirola CJ. Meta-analysis of the influence of I148M variant of patatin-like
phospholipase domain containing 3 gene (PNPLA3) on the susceptibility and histological severity
of nonalcoholic fatty liver disease. Hepatology 2011; 53: 1883–94.
111. Dongiovanni P, Donati B, Fares R, et al. PNPLA3 I148M polymorphism and progressive liver
disease. World J Gastroenterol 2013; 19: 6969–78.
112. He S, Mcphaul C, Li JZ, et al. A sequence variation (I148M) in PNPLA3 associated with
nonalcoholic fatty liver disease disrupts triglyceride hydrolysis. J Biol Chem 2010; 285: 6706–15.
113. Pirazzi C, Valenti L, Motta BM, et al. PNPLA3 has retinyl-palmitate lipase activity in human
hepatic stellate cells. Hum Mol Genet 2014; 23: 4077–85.
114. Ruhanen H, Perttila J, Holtta-Vuori M, et al. PNPLA3 mediates hepatocyte triacylglycerol
remodeling. J Lipid Res 2014; 55: 739–46.
115. Smagris E, Basuray S, Li J, et al. Pnpla3I148M knockin mice accumulate PNPLA3 on lipid
droplets and develop hepatic steatosis. Hepatology 2015; 61: 108–18. 116. Donati B, Motta BM,
Pingitore P, et al. The rs2294918 E434K variant modulates PNPLA3 expression and liver damage.
Hepatology 2016; 63: 787–98.
117. Hyysalo J, Gopalacharyulu P, Bian H, et al. Circulating triacylglycerol signatures in
nonalcoholic fatty liver disease associated with the I148M variant in PNPLA3 and with obesity.
Diabetes 2013; 63: 312–22.
118. Peter A, Kovarova M, Nadalin S, et al. PNPLA3 variant I148M is associated with altered
hepatic lipid composition in humans. Diabetologia 2014; 57: 2103–7.
119. Palmer CN, Maglio C, Pirazzi C, et al. Paradoxical lower serum triglyceride levels and higher
type 2 diabetes mellitus susceptibility in obese individuals with the PNPLA3 148M variant. PLoS
ONE 2012; 7: e39362.
120. Kozlitina J, Smagris E, Stender S, et al. Exome-wide association study identifies a TM6SF2
variant that confers susceptibility to nonalcoholic fatty liver disease. Nat Genet 2014; 46: 352–6.
121. Mahdessian H, Taxiarchis A, Popov S, et al. TM6SF2 is a regulator of liver fat metabolism
influencing triglyceride secretion and hepatic lipid droplet content. Proc Natl Acad Sci USA 2014;
111: 8913–8.
122. Holmen OL, Zhang H, Fan Y, et al. Systematic evaluation of coding variation identifies a
candidate causal variant in TM6SF2 influencing total cholesterol and myocardial infarction risk.
Nat Genet 2014; 46: 345–51.
123. Smagris E, Gilyard S, Basuray S, Cohen JC, HOBBS HH. Inactivation of Tm6sf2, a gene
defective in fatty liver disease, impairs lipidation but not secretion of very low density lipoproteins.
J Biol Chem 2016; 291: 10659–76.
124. Pirola CJ, Sookoian S. The dual and opposite role of the TM6SF2-rs58542926 variant in
protecting against cardiovascular disease and conferring risk for nonalcoholic fatty liver: a metaanalysis. Hepatology 2015; 62: 1742–56.
125. Mancina RM, Dongiovanni P, Petta S, et al. The MBOAT7-TMC4 variant rs641738 increases
risk of nonalcoholic fatty liver disease in individuals of European descent. Gastroenterology 2016;
150: 1219–230.
126. Speliotes EK, Yerges-Armstrong LM, Wu J, et al. Genome-wide association analysis identifies
variants associated with nonalcoholic fatty liver disease that have distinct effects on metabolic
traits. PLoS Genet 2011; 7:e1001324.
127. Santoro N, Zhang CK, Zhao H, et al. Variant in the glucokinase regulatory protein (GCKR)
gene is associated with fatty liver in obese children and adolescents. Hepatology 2011; 55: 781–9.
128. Petta S, Miele L, Bugianesi E, et al. Glucokinase regulatory protein gene polymorphism affects
liver fibrosis in non-alcoholic Fatty liver disease. PLoS ONE 2014; 9:e87523.
129. Valenti L, Alisi A, Nobili V. Unraveling the genetics of fatty liver in obese children: additive
effect of P446L GCKR and I148M PNPLA3 polymorphisms. Hepatology 2012; 55: 661–3.
130. Beer NL, Tribble ND, McCulloch LJ, et al. The P446L variant in GCKR associated with
fasting plasma glucose and triglyceride levels exerts its effect through increased glucokinase
activity in liver. Hum Mol Genet 2009; 18:4081–8.
131. Di Filippo M, Moulin P, Roy P, et al. Homozygous MTTP and APOB mutations may lead to
hepatic steatosis and fibrosis despite metabolic differences in congenital hypocholesterolemia. J
Hepatol 2014; 61: 891–902.
132. Cefalu AB, Pirruccello JP, Noto D, et al. A novel APOB mutation identified by exome
sequencing cosegregates with steatosis, liver cancer, and hypocholesterolemia. Arterioscler Thromb
Vasc Biol 2013; 33: 2021–5.
133. Amaro A, Fabbrini E, Kars M, et al. Dissociation between intrahepatic triglyceride content and
insulin resistance in familial hypobetalipoproteinemia. Gastroenterology 2010;139: 149–53.
134. Dongiovanni P, Valenti L, Rametta R, et al. Genetic variants regulating insulin receptor
signalling are associated with the severity of liver damage in patients with nonalcoholic fatty liver
disease. Gut 2010; 59: 267–73.
135. Neuschwander-Tetri BA, Loomba R, Sanyal AJ, et al. Farnesoid X nuclear receptor ligand
obeticholic acid for non-cirrhotic, non-alcoholic steatohepatitis (FLINT): a multicentre, randomised,
placebo-controlled trial. Lancet 2015; 385: 956–65.
136. Holland WL, Miller RA, Wang ZV, et al. Receptormediated activation of ceramidase activity
initiates the pleiotropic actions of adiponectin. Nat Med 2011; 17:55–63.