The PI3K/AKT/mTOR signaling pathway- also known as the “mTOR pathway”- contains two important components: mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2) (Figure 1). They are multiprotein complexes, comprised of both shared and unique components.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Mammalian target of rapamycin complex 1, Mammalian target of rapamycin complex 2 and their associated proteins. mTOR: Mammalian target of rapamycin; DEPTOR: DEP domain-containing mTOR-interacting protein; mLST8: Mammalian lethal with SEC13 protein 9; RAPTOR: Regulatory-associated protein of mTOR; PRAS40: 40 kDa Pro-rich AKT substrate; RICTOR: Rapamycin-insensitive companion of mTOR; mSIN1: Mammalian stress-activated map kinase-interacting protein 1; PROTOR: Protein observed with RICTOR.

The mTOR kinase- also known as “mTOR”- is one of three components which is present in both mTORC1 and mTOR2; mammalian lethal with SEC13 protein 9 (mLST8) and DEP domain-containing mTOR-interacting protein (DEPTOR) are two other proteins that are common to both mTORC1 and mTORC2. mLST8 interacts directly with mTOR to enhance its kinase activity, particularly within mTORC2 (its effect within mTORC1 is not clearly understood)[5]. DEPTOR prevents substrate binding to mTORC1 and mTORC2, which leads to inhibition of mTORC1 and mTORC2 activity[6,7].

mTORC1, which is sensitive to the effects of rapamycin, has two unique proteins: regulatory-associated protein of mTOR (RAPTOR) and 40 kDa Pro-rich AKT substrate (PRAS40; also known as AKT1S1). RAPTOR serves as a binding platform where substrates are presented to mTOR for subsequent activation of mTORC1[8]. Conversely, PRAS40, like DEPTOR, is a direct inhibitor of mTORC1 substrate binding which hinders mTORC1 activity[9].

Specific to mTORC2 are rapamycin-insensitive companion of mTOR (RICTOR), mammalian stress-activated map kinase-interacting protein 1 (mSIN1; also known as MAPKAP1) and protein observed with RICTOR (PROTOR) (Figure 1)[10]. There is some evidence that RICTOR contributes to the structural foundation of mTORC2; in the absence of RICTOR, mTORC2 becomes inactive[7]. The functions of mSIN1 and PROTOR remain unclear.

mTORC1: mTORC1 expression is driven by stimulants such as energy status, physiologic stress, and growth factors. Specifically, in the presence of growth factors, insulin receptor substrate 1 (IRS1) activates PI3K. PI3K phosphorylates the second messenger called phosphatidylinositol (4,5)-biphosphate (PIP-2), which becomes phosphatidylinositol (3,4,5)-triphosphate (PIP-3) upon phosphorylation. PIP-3 then promotes the phosphorylation of serine/threonine protein kinase (PKB/AKT) at protein residue Thr308 by 3-phosphoinositide-dependent protein kinase-1 (PDK1). Further downstream signaling through the effector tuberous sclerosis 1-tuberous sclerosis 2 complex (TSC1-TSC2) ultimately leads to the activation of mTORC1.

Activated mTORC1 phosphorylates its two downstream targets, 70S ribosomal protein S6 kinase (S6K1) and the eukaryotic initiation factor 4E binding protein 1 (4E-BP1). S6K1 and 4E-BP1 are major regulators of protein translation; they also drive cell proliferation, angiogenesis, and autophagy[11].

Under normal physiologic conditions, 4E-BP1 binds to the eukaryotic initiation factor 4E (eIF4E) to arrest protein translation. However, when 4E-BP1 is phosphorylated by mTORC1, its binding to eIF4E is interrupted, and this allows protein translation to occur. Concomitantly, phosphorylation of S6K1 by mTORC1 also results in protein translation; however, it also creates a negative feedback loop whereby the phosphorylated S6K1 attenuates PI3K signaling by suppressing IRS1 activity, leading to mTORC1 inhibition. Interestingly, mTORC1 inhibition by rapamycin and its analogues disrupts S6K1-mediated feedback inhibition of PI3K signaling, which allows for increased PKB/AKT phosphorylation (Figure 2). Furthermore, rapamycin-induced inhibition of mTORC1 leads to an accumulation of phosphorylated AKT which can then activate downstream effectors of alternative pathways to inhibit apoptosis and promote cell proliferation[12].

Open in New Tab   Full Size Figure   Download Figure

Figure 2 The PI3K/AKT/mTOR pathway. IRS1: Insulin receptor subsrate 1; PI3K: Phophoinositide-3-kinase; PIP-2: Phosphatidylinositol (4,5)-biphosphate; PIP-3: Phosphatidylinositol (3,4,5)-triphosphate; PTEN: Phosphatase and tensin homologue on chromosome 10 gene; PDK1: Phosphoinositide-dependent protein kinase-1; PKB/AKT: Serine/threonine protein kinase; TSC1-TSC2: Tuberous sclerosis 1-tuberous sclerosis 2 complex; 4EBP1: Eukaryotic initiation factor 4E binding protein 1; eIF4E: Eukaryotic initiation factor 4E; S6K1: 70S ribosomal protein S6 kinase; SFKs: SRC family kinases.

mTORC2: Similar to mTORC1, mTORC2 activity is also promoted by growth factors. The upstream regulatory mechanisms specific to mTORC2 are poorly understood; however, its role in the phosphorylation of PKB/AKT has been well-characterized. The full activation of PKB/AKT requires two steps of phosphorylation: first, at protein residue Thr308 by PDK1, and second, at residue Ser473 by mTORC2[13]. Therefore, mTORC2 indirectly promotes mTORC1 activity through activation of PKB/AKT (Figure 2). Another less understood function of mTORC2 involves the regulation of actin cytoskeleton organization. Unlike the inhibitory effects on mTORC1, the effects of rapamycin and its analogues on mTORC2 are minimal[14].

Phosphatase and tensin homologue on chromosome 10 gene (PTEN) is a multiphosphatase tumor suppressor located on human chromosome 10q23.3. It blocks the downstream activity of PI3K-AKT signaling by degrading PIP-3[11]. Inhibition of PIP-3 by PTEN prevents activation of PKB/AKT which leads to the down-regulation of mTORC1 activity (Figure 2). In the absence of PTEN, activation of mTORC1 is unbridled and hepatocellular carcinogenesis occurs. Watanabe et al[15] showed high incidence of HCC (66%) in PTEN-deficient mice at the end of an 80-wk period. Wang et al[16] demonstrated that decreased PTEN protein expression in HCC tissue samples compared to paired surrounding tissue samples was associated with higher tumor pathologic grade, TNM stage, and more frequent incidence of metastasis.

The TSC1 and TSC2 proteins (also known as hamartin and tuberin, respectively) are regulators of cell proliferation which have been implicated in HCC carcinogenesis. In its active form, the TSC1-TSC2 complex inhibits mTORC1 activation. Specifically, TSC2 acts as a GTPase-activating protein (GAP) which degrades guanosine triphosphate (GTP) and prevents its binding with Rheb, a GTP-binding protein. As a result, Rheb’s ability to inhibit FKBP38, a negative regular mTORC1, is disabled and mTORC1 is inhibited. However, Akt-mediated phosphorylation deactivates the TSC1-TSC2 complex by decreasing its GAP-activity towards Rheb, which permits Rheb-GTP binding[17]. In turn, Rheb carries out its usual function of inhibiting FKBP38, and mTORC1 activation occurs (Figure 2)[17,18]. In fact, loss of either TSC1 or TSC2 promotes autonomous activation of mTORC1. Using liver-specific TSC1 knockout mice, Menon et al[19] demonstrated that chronic activation of mTORC1 in the absence of TSC1 induced hepatocyte damage, independent of hepatic steatosis, which leads to the spontaneous development of HCC.

It is important to note that PTEN and TSC1-TSC2 complex also function as integrating hubs for the regulation of mTOR via alternative signaling pathways. For example, the Src family kinases (SFKs) and the Wnt protein of the Wnt/β-catenin pathway are direct upstream regulators of PTEN and TSC1-TSC2 complex, respectively. Studies of breast cancer cell lines have shown that SFKs phosphorylate PTEN to inhibit its function[20], which then promotes mTORC1 activation. Conversely, the stimulation of Wnt prevents TSC2 phosphorylation through inhibition of GSKβ3, a protein constituent of Wnt/β-catenin pathway, thus inhibiting mTORC1 activation[17].

Recently, single-arm phase I/II studies have shown that everolimus (a second-generation mTORC1 inhibitor), has single-agent activity in de novo or recurrent advanced HCC. In a cohort of 36 patients, everolimus hindered disease progression in patients with advanced HCC when used at maximum tolerated dose of 70 mg weekly[29]. In a subsequent phase I/II study by Zhu et al[30], 28 patients with advanced HCC tolerated everolimus at the dose of 10 mg daily. The median progression free survival was 3.8 mo, suggesting a modest antitumor effect of everolimus in advanced HCC[31].

This study led to the global phase III randomized EVOLVE-1 trial, where everolimus was compared to placebo in patients with advanced HCC who discontinued sorafenib due to disease progression or drug intolerance. This trial unfortunately showed no OS benefit for everolimus in the salvage setting of advanced HCC[32]. As discussed in section 2 (b) of this review, mTORC1 and mTORC2 are two complementary components of the mTOR pathway: when mTORC1 is inhibited, mTORC2 is upregulated. This increase in mTORC2 activity generates a surplus of phosphorylated PKB/AKT which, despite mTORC1 inhibition, inhibits apoptosis and promotes cell proliferation via alternative pathways (Figure 2)[33]. This phenomenon may partially explain the unsatisfactory efficacy of everolimus demonstrated in the EVOLVE-1 trial, and suggests a potential mechanism for drug resistance against mTORC1 inhibitors in HCC. Given this theory, dual mTORC1 and mTORC2 inhibition has become an attractive pharmacologic target with therapeutic potential in advanced HCC treatment.

The safety of everolimus in combination with sorafenib has also been evaluated for the treatment of advanced HCC, as it posed the opportunity to target two major pathways involved in HCC pathogenesis. However, phase I studies demonstrated intolerable toxicities with this combination, rendering it infeasible as a therapeutic option[34,35].