Genetic studies over the past two decades have revealed that TSC, GATORopathies, and HH share a common theme: dysregulation of intracellular signaling pathways that control cell growth and differentiation. Although different sets of genes are involved, the end result is the disruption of developmental programs that shape brain structure and function.

Altered mTOR signaling in TSC and GATORopathies

The mTOR pathway acts as a central hub integrating signals from growth factors, nutrients, and energy status to regulate protein synthesis, cell growth, and metabolism. It functions through two complexes: mTORC1, which promotes protein synthesis and cell size, and mTORC2, which influences the cytoskeleton and cell survival. In healthy cells, these complexes are tightly regulated by upstream inhibitors. The TSC1/TSC2 complex restrains mTORC1 by inhibiting the small GTPase Ras homolog enriched in brain (RHEB), while the GATOR1 complex suppresses mTORC1 under conditions of amino acid scarcity (Fig. 2).

Fig. 2The alternative text for this image may have been generated using AI.

Mitogen-activated protein (MAP) kinase and mTOR signaling cross-talk. Key cellular inputs, including metabolites (glucose, amino acids) and growth factors, modulate cell growth, proliferation, survival, and cytoskeletal organization via the mTORC1 and mTORC2 complexes

In TSC, pathogenic germline loss-of-function mutations in TSC1 or TSC2, encoding hamartin and tuberin, are the root cause. This loss prevents inhibition of RHEB, resulting in constitutive activation of the mTORC1 pathway. The disease traditionally follows a “two-hit” model, where a germline mutation (the first hit) is followed by a brain somatic mutation in the remaining healthy allele (the second hit), leading to biallelic inactivation and subsequent focal lesion formation [24]. The timing and location of these second hits dictate the varying number, size, and location of tubers seen even among patients with the same germline mutation. While nearly 100% of SEGAs show biallelic inactivation [5], second hits in tubers are more difficult to detect due to low variant allele frequencies (VAF). However, modern targeted deep sequencing has identified somatic second hits in approximately 35% of tuber samples [26]. In cases where no second hit is found, researchers suggest that haploinsufficiency or “progenitor-cell-specific” sensitivity to mTORC1 dysregulation may suffice for lesion formation [15]. Beyond germline cases, brain somatic variants in TSC1 and TSC2 cause isolated FCD IIB, sometimes discussed as a forme fruste of TSC [25]. Additionally, systemic mosaicism, where a mutation occurs early in embryogenesis, affects multiple but not all tissues and accounts for 10–15% of cases [23]. Most pathogenic variants are truncating mutations, whereas missense mutations are less common and tend to cluster in functional domains (https://simple-clinvar.broadinstitute.org/). Importantly, although loss of heterozygosity is commonly found in renal and skin lesions of TSC, it is detected less consistently in cortical tubers, raising questions about whether all tuber cells require biallelic inactivation [10].

In GATORopathies, loss-of-function mutations in DEP-domain containing 5 (DEPDC5), Nitrogen permease regulator 2-like protein (NPRL2), or NPRL3 disrupt the GATOR1 complex, leading to inappropriate mTORC1 activation even when amino acids are scarce. Like TSC, these disorders follow a two-hit model in many cases, with a germline mutation combined with a somatic hit in progenitor cells [3]. However, GATORopathies differ from TSC in that they are usually brain-limited and not syndromic. This challenges genetic testing, as brain somatic variants can only be identified from surgical brain tissue or explanted depth electrodes [7]. Clinical presentation is variable: some patients develop FCD or hemimegalencephaly, while others present with focal epilepsy without obvious structural lesions on MRI [31]. This incomplete penetrance likely reflects the role of mosaicism.

Together, TSC and GATORopathies illustrate how mutations in different molecular “brakes” of the same signaling pathway can converge on a shared phenotype: malformations of cortical development and epilepsy.

Aberrant Shh signaling in hypothalamic hamartomas

In contrast to TSC and GATORopathies, the genetic basis of HH centers on the Shh pathway, which regulates embryonic patterning, progenitor identity, and neurogenesis. In syndromic cases such as Pallister–Hall syndrome, germline mutations in GLI family zinc finger 3 (GLI3), a transcriptional effector of Shh signaling, have been identified. Sporadic cases have been linked to somatic mosaic mutations in GLI3 and other genes in the Shh pathway, including GLI2, Protein kinase cAMP-activated aatalytic aubunit Alpha (PRKACA), and CREB binding lysine acetylransferase (CREBBP) [9, 16, 34]. More recently, mutations in genes required for ciliary function, such as Dynein Cytoplasmic 2 Heavy Chain 1 (DYNC2H1), Dynein 2 Intermediate Chain 1 (DYNC2I1), and Intraflagellar Transport 140 (IFT140), have been reported, as well as alterations in SMO, a key Shh receptor [12, 14]. These findings emphasize that Shh signaling depends on intact primary cilia, which act as the signaling hub for this pathway.

There are also reports of both germline and somatic Protein tyrosine phosphatase non-receptor type 11 (PTPN11) variants in HH [12]. PTPN11 encodes a component of the RAS-MAPK cascade (Fig. 2), closely connected to mTOR signaling, and its dysregulation drives uncontrolled cell growth. Germline gain-of-function variants are associated with Noonan syndrome, a multisystem disorder characterized by distinctive facial features, short stature, congenital heart defects, and variable neurodevelopmental delay [33]. Brain somatic gain-of-function variants in PTPN11 together with other RAS-MAPK signaling genes have also been reported in glioneuronal low-grade epilepsy-associated tumors [17]. Notably, hamartomas can occur as part of the clinical spectrum of Noonan syndrome, which also carries a predisposition to glioneuronal brain tumors.

Thus, similar to TSC and GATORopathies, HH may in some cases follow a two-hit model, with a germline mutation conferring predisposition and a somatic mutation in ciliary or Shh pathway genes promoting lesion formation. However, the overall genetic architecture of HH is more heterogeneous, and many cases remain unexplained by known mutations. Consistent with the concept of degeneracy that multiple pathways can serve similar functions, hyperactivation of either Shh or MAPK/mTOR signaling may result in comparable brain malformations, depending on cell type, developmental stage, and selection pressure.

Cross-talk between mTOR and Shh signaling

Although mTOR and Shh pathways are distinct, experimental data highlight multiple points of interaction. Shh signaling activates GLI transcription factors through Patched 1 (PTCH1) and Smoothened, Frizzled class receptor (SMO), and in certain contexts, this also activates PI3K-AKT, which inhibits TSC2 and thereby stimulates mTORC1 [32]. Conversely, mTORC1 can regulate the translation and localization of GLI proteins, amplifying or modulating Shh output. In addition, some downstream effectors of mTOR, such as Ribosomal protein S6 kinase B1 (RPS6KB1), directly influence GLI protein stability [35].

This cross-talk has been well studied in cancer biology, where combined activation of Shh and mTOR promotes tumorigenesis, and combined inhibition shows therapeutic benefit [6, 37]. In developmental brain malformations, the role of such interactions has not yet been explored. The observation that both pathways are dysregulated in epilepsy-associated lesions raises the possibility that their convergence may be central to epileptogenesis.