1. Introduction

Hepatocellular carcinoma (HCC) is the main subtype of liver tumor and the third-leading cause of cancer death worldwide, whose incidence reflects the etiologies of liver diseases and their geographical distribution ^[1][2]. The global spreading of obesity and metabolic syndrome (MetS) have rapidly increased the incidence of nonalcoholic fatty liver disease (NAFLD) over the past two decades and, in parallel, those of NAFLD-related HCC in the industrialized society due also to the development of anti-viral therapies and effective viral hepatitis B (HBV) vaccination. Nowadays, NAFLD is the most common chronic liver disorder, affecting 25-30% of general population, and it is closely intertwined with insulin resistance (IR), overweight and type 2 diabetes mellitus (T2DM). NAFLD is defined as hepatic fat content >5% of liver weight (steatosis), a potential reversible condition which could evolve into nonalcoholic steatohepatitis (NASH), fibrosis, and cirrhosis. It has been reported that NASH patients with advanced fibrosis (F3-F4) had 7-fold higher rate to develop HCC and, in a small proportion, HCC arises in NAFLD individuals without fibrosis ^[3][4]. Additionally, the 30-40% of HCC cases occurred in subjects with cryptogenic cirrhosis, some of which were further affected by dyslipidemia, obesity, T2DM and possibly by NASH ^[5]. A study analyzing the Surveillance, Epidemiology and End Results (SEER) registries recorded a 9% annual increase of NAFLD-HCC cases from 2004 to 2009. The survey also examined the prevalence and mortality of 4.929 HCC individuals. Among them, the 14.1% of HCC was due to fatty liver, the 5% received NAFLD-related liver transplantation (LT) and the presence of NAFLD increased the risk of 1-year mortality, especially in older subjects with previous heart disease ^[6][7]. Therefore, NAFLD is currently representing not only a clinical and socio-economic burden for health, but it is predicted to overcome HCV, HBV and alcoholic hepatitis thus becoming the leading cause of HCC and LT ^{[1][2][3][4][6]}.

The outbreak of NAFLD-HCC results from continuous cycle of parenchymal disruption and tissue regeneration sustained by inflammation, oxidative stress, fibrogenesis and hypoxia. In this scenario, mitochondria, which are extremely adaptable in response to external cues, exert a key role as bioenergetic factories and for the regulation of liver metabolism. Compelling evidence has suggested that mitochondrial dysfunction may precede IR or arise before NASH development thereby reinforcing the concept that NAFLD may be considered a mitochondrial disorder. Loss of mitochondrial plasticity in terms of functions, morphology and dynamics may support hepatocellular injury and the onset of the Warburg effect, the mechanism by which hepatocytes exploit anaerobic glycolysis even in the presence of oxygen in order to sustain energy demand and cell proliferation ^{[8][9][10][11]}. Pathogenesis of NAFLD is multifactorial and closely related to nutrition and lifestyle, which may impact on mitochondrial functionality. However, it has emerged that both common and rare genetic variants predisposing to NAFLD spectrum may hasten mitochondrial dysfunction thus accelerating disease progression. Therefore, in the present entry we aimed to summarize how inherited factors may drive to HCC onset by modulating mitochondrial dynamics.

2. The Link among NAFLD, mitochondrial dysfunction and HCC: the Relevance of Genetics

Familial, twin, and epidemiological studies indicated that NAFLD has a strong heritable component, which contributes to the huge inter-individual phenotypic variability. Our group demonstrated that hepatic fat accumulation represents the main driver of the progression to the end-stage of liver damage in genetically predisposed individuals, and recently proposed a detailed review including all the candidate genes related to NAFLD susceptibility ^[12][13].

Currently, the rs738409 C >G single nucleotide polymorphism (SNP) in the Patatin-like phospholipase domain containing 3 gene (PNPLA3 or adiponutrin) is the major genetic variant associated with NAFLD onset and its progressive forms, including HCC. PNPLA3 is mainly localized on the endoplasmic reticulum (ER) and lipid droplets (LDs) surface in hepatocytes, adipocytes and hepatic stellate cells (HSCs), and it may be transcriptionally induced or post-translationally modified to provide triglyceride (TG) hydrolysis during the post-prandial or hyper-insulinemic state. Patients carrying the G allele lost PNPLA3 enzymatic activity, which impedes TG disposal and interferes with the activity of other lipases, such as PNPLA2 ^[14][15]. Beyond the triacylglycerol remodeling, PNPLA3 exerts widespread effects on human liver metabolome ^[16], influencing mitochondrial functions, glucose reprogramming and tumorigenesis. Huh-7 hepatoma cells overexpressing the PNPLA3 I148M variant showed high levels of lactate and γ-glutamyl-amino acids, thus mirroring the metabolic switching to aerobic glycolysis and mitochondrial failure, respectively ^[16]. Moreover, the rs738409 SNP impacts on retinol secretion in hepatic stellate cells (HSCs), leading to the myofibroblast-like phenotype and collagen deposition, and boosting fibrogenesis in subjects affected by NASH) ^[17] . In a small cohort of 54 NAFLD individuals, it has been demonstrated that carriers of the G risk allele had a severe profile of liver disease, characterized by enhanced steatosis, activation of pro-inflammatory pathways and an increased proliferative activity of hepatocytes ^[18]. Interestingly, Bruschi et al. demonstrated that the presence of I148M substitution in the PNPLA3 gene further affected metabolic reprogramming in TGFβ-activated HSCs, shifting towards aerobic glycolysis, lactate release and the activation of YAP/Hedgehog signaling ^[19].

The rs641738 C > T variant in the Membrane bound o-acyltransferase domain-containing 7/Transmembrane channel-like 4 (MBOAT7/TMC4) locus, encoding the MBOAT7 enzyme, was associated with the entire spectrum of NAFLD, including hepatocellular carcinoma (HCC) ^[20]. Recently, the role of the MBOAT7 variant in NAFLD progression has been evaluated in a large meta-analysis. Data were collected from 1047.265 subjects, of whom 8303 had liver biopsies, and displayed a correlation between the T minor allele and hepatic fat deposition, ALT levels, and advanced stages of NAFLD, such as fibrosis and HCC. In particular, carriers of the rs641738 variant show a 30% risk of developing HCC compared to non-carriers ^[21]. Physiologically, MBOAT7 localizes on mitochondrial-associated ER-membranes (MAMs) and mediates phosphatidylinositol (PI) acyl-chain remodeling in the Land’s cycle. Our group demonstrated that hepatic MBOAT7 expression is reduced during hyperinsulinemia and by the presence of the rs641738 C > T variant ^[22][23]. MBOAT7 downregulation induces an enrichment of saturated PIs, which are shunted towards the synthesis of TGs, thus contributing to fat accumulation. Though no evidence linking MBOAT7, mitochondrial lifecycle and metabolic reprogramming has been reported, it could be postulated that the wealth of saturated lipids induced by MBOAT7 downregulation may affect membrane composition and dynamics, possibly breaking ER–mitochondria communications.

The rs58542926 C > T variant in the Transmembrane 6 Superfamily member 2 (TM6SF2) gene induces TM6SF2 loss-of-function and hastens its hepatic protein turnover ^[24]. TM6SF2 dwells on ER-Golgi compartments where fat biosynthesis, LDs and lipoprotein formation occur. TM6SF2 inactivation induced by the presence of the polymorphisms impairs the assembly and trafficking of very low-density lipoprotein (VLDL), which remains trapped in hepatocytes ^[24]. In Huh-7 cells, TM6SF2 deficiency reduces the amount of polyunsaturated fatty acids (PUFAs), along with causing alterations in mitochondrial β-oxidation and higher numbers of lysosomal compartments ^[25]. In the small intestine of zebrafish, the TM6SF2 loss-of-function induces changes in ER architecture appearing with enlarged cisternae, supporting the notion that TM6SF2 may impact organelles’ morphology ^[26]. However, the rs58542926 polymorphism has been associated with the NAFLD/NASH spectrum, but its role in HCC development remains to be explored. A meta-analysis including 24,147 subjects affected by chronic liver disorders revealed that the presence of the T risk allele was correlated with a higher risk of developing NAFLD and its advanced stages, such as HCC ^[27]. Raksayot et al. performed a cross-sectional study in a cohort of 502 NAFLD patients and observed that carriers of the T allele are at a higher risk of HCC progression ^[28].

To delve inside the mechanisms underlying NAFLD pathogenesis and to investigate possible synergisms among PNPLA3, MBOAT7 and TM6SF2 leading to hepatocytic metabolic rewiring, our group has generated in vitro models of genetic NAFLD. We stably silenced MBOAT7 (MBOAT7^−/−), TM6SF2 (TM6SF2^−/−), or both genes (MBOAT7^−/−TM6SF2^−/−), in HepG2 cells, homozygous for the I48M PNPLA3 variant, by exploiting clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9 (CRISPR/Cas9) technology ^[10][11][23]. MBOAT7^−/− spontaneously accumulated giant LDs associated with a dramatic increment in ROS and peroxides levels, while TM6SF2^−/− and MBOAT7^−/−TM6SF2^−/− models showed mitochondria with small and globular shapes, the loss of cisterns’ architecture, and ultrastructural electron density, suggestive of mitochondria degeneration. The numbers of mitochondria were progressively increased in all mutated cell lines, suggesting that either MBOAT7 and/or TM6SF2 deficiency impact mitochondrial biogenesis. Notably, the compound knockout re-routed its metabolism towards glucose-dependent ATP production, enhancing glycolytic enzymes, LDH and lactate release, thereby suggesting that the depletion of both MBOAT7/TM6SF2 combined with the genetic background of I148M PNPLA3 may affect mitochondria turnover, possibly accelerating metabolic reprogramming ^[10][13].

Recently, the opportunity has emerged to translate the genetics into clinics by aggregating these genetic variants into polygenic risk scores, which may better discriminate NAFLD patients who are at-risk of developing progressive liver damage and HCC ^[29]. In a large cohort of biopsied NAFLD patients, it has been observed that the cumulative number of risk alleles is associated with serum markers of disease severity and increased risk of developing HCC ^[10][30][31]. A cross-sectional study consisting of 2566 NAFLD participants evaluated the impact of these genetic polymorphisms on hepatocytic fat accumulation in HCC progression. For this purpose, the authors generated a polygenic risk score based on the presence of variants in PNPLA3, TM6SF2 MBOAT7 and GCKR genes, which was able to predict HCC occurrence much more effectively than the presence of a single genetic variant ^[29].

3. Mitochondrial Polymorphisms Are Correlated with NAFLD Pathogenesis

Several polymorphisms in mitochondrial genes have been associated with NAFLD development and progression. In genetically modified mice, the non-synonymous nt7778 G > T genetic variation in the mitochondrial ATP synthase protein 8 (mt-ATP8) increased susceptibility to diet-induced NASH ^[32]. The identification of mitochondrial haplotypes was even associated with NAFLD predisposition, opening new avenues for mito-genetic screening in patients, and new experimental applications ^[30].

Manganese-dependent superoxide dismutase (SOD), encoded by the nuclear SOD2 gene, mitigates oxidative damage by catalyzing the conversion of superoxide radicals to hydrogen peroxide. The rs4880 C47T variant in the SOD2 gene encodes the valine to alanine amino acid substitution at position 16 in the signal region targeting the protein to the mitochondrial matrix. The C47T mutation causes a reduction in MnSOD2 activity and the consequent failure to neutralize superoxide radicals. In case–control and familial studies, Al-Serri et al. demonstrated that the inherited T risk allele was an independent predictor of NASH severity and was strictly associated with fibrosis in both adults and children ^[31]. Conversely, the −866 G > A polymorphism localized in the promoter region of the Uncoupling protein 2 (UCP2) gene, involved in heat dissipation, has been associated with a reduced risk of obesity ^[33]. The A allele promotes UCP2 overexpression in the liver and has a protective role in progression from simple steatosis to NASH ^[34]. Likewise, the rs1800849 −55 C/T UCP3 variant ameliorates the circulating lipid profile and correlates with loss of body weight ^[35]. These findings were not confirmed by Aller et al. and Qian et al., who associated both the −866 G > A polymorphism and the rs1800849 variant with higher risk of IR, obesity, lower levels of adiponectin, severe steatosis, and inflammation in NAFLD subjects ^[36][37].

Polymorphisms in sirtuins (SIRTs) further contribute to the regulation of mitochondrial functionality and dynamics, possibly contributing to NAFLD/NASH advancement and its cardiovascular comorbidities ^[38]. Patients carrying the rs11246020 variant (V208I) in the SIRT3 gene displayed a higher susceptibility to NAFLD. Consistently, Sirt3 knockout mice fed an HFD showed IR and a worsened adiposity and NASH ^[39]. The rs107251 in the SIRT6 gene affected SIRT6 activity, influencing its role in DNA repair and the maintenance of telomeric chromatin ^[40]. It has been described that the rs7895833 G > A in the SIRT1 gene represents a risk factor for body fat content and high diastolic blood pressure ^[40][41]. Interestingly, low SIRT1 levels were detected in 70 cirrhotic HCC patients carrying the rs7895833 variant, and SIRT1 reduction was inversely correlated with high AFP, Child–Pugh score and tumor stage ^[42].

Recently, rs2642438 A165T polymorphisms at the N-terminal domain of the Mitochondrial amidoxime reducing component 1 (MARC1) gene have been described, localizing on mitochondrial outer membranes (MOMs). The A165T variant has been associated with low fat content in the liver and a reduced risk of NALD progression toward cirrhosis. Such findings were independently validated by Lukkonen et al., showing that carriers of the rs2642438 variant had alleviated NASH severity accompanied by an improvement of the hepatic lipid profile, mainly consisting of polyunsaturated phosphatidylcholines, thus suggesting that MARC1 could represent a candidate therapeutic target ^[43][44].

4. Rare NAFLD Pathogenic Variants Are Involved in Switching towards HCC

Part of the missing hereditability in NAFLD may be attributed to rare genetic variants with a large effect size.

Rare mutations in the telomerase reverse transcriptase (TERT) promoter may arise in NAFLD-cirrhosis, in 10-20% of both low-grade and high-grade dysplastic nodules and in familial HCC, supporting that TERT germline genetic variants may be involved in tumor initiation ^[45]. In a cohort of 40 NAFLD-HCC, 45 cirrhotic patients and 64 healthy controls, telomere length decreased with the progression of NAFLD towards cirrhosis and mainly with HCC ^[46]. Four rare mutations have emerged in the hTERT gene among NAFLD-HCC subjects such as the Glu113Arg_fs*79 frameshift in the second exon and 3 missense mutations (Ala67Val, Pro193Leu, Glu668Asp) which correlated with shorter telomere length. In particular, the Ala67Val and Glu668Asp SNPs led to TERT loss-of-function and decreased its hepatic expression. On the contrary, the Pro193Leu substitution did not affect TERT catalytic activity but reduced its chromatin binding capacity ^[47]. Furthermore, in a cross-sectional study, it has been observed that NAFLD-HCC patients showed an enrichment of rare genetic variants in Regulator of telomere elongation helicase 1 (RTEL1) and Telomeric repeat binding factor 2 (TERF2) genes, that are involved in telomere preservation, and in RB1 which mediates the oxidative stress response. Mutations in STK11, TSC1, TSC2, NF2 and SMAD4 candidate genes, which regulate cell growth and proliferation, were also strongly correlated to HCC risk ^[48].

HCC surveillance may be addressed to NAFLD subjects with family history of hypobetalipoproteinaemia caused by ApoB mutations. ApoB gene is involved in hepatic lipid metabolism and its genetic variants lead to an impaired synthesis of ApoB100 with the consequent alteration of hepatic VLDL export. Uncommon variants in ApoB gene result in the impairment of VLDL export and development of severe hepatic steatosis. Some of the genetic variants causative of ineffective ApoB100 synthesis may even alter ApoB48 isoform expressed in the enterocytes provoking malabsorption of fat and insoluble vitamins, retention of chylomicrons and alterations of intestinal barrier ^[49].

Finally, a novel association between variants in Sequestosome 1 (SQSMT1) and HCC onset have been identified in NAFLD-HCC patients. SQSMT1 encodes the ubiquitin-binding protein p62, an autophagosome cargo protein that targets other proteins for selective autophagy. p62 takes part to Mallory-Denk bodies (MDBs), a cytoplasmatic protein aggregates found in several chronic liver diseases including NAFLD as well as in HCC, and it is involved in the hepatocyte’s transformation through the activation of mTOR pathway and regulation of telomere length machinery ^[50][51].

This entry is adapted from the peer-reviewed paper 10.3390/ijms22084173