The PI3K/AKT/mTOR Pathway in Cancer

PI3K Activation and PIP3 Generation

The PI3K family comprises three classes; in cancer, class IA PI3Ks — heterodimers of a p110 catalytic subunit (α, β, or δ) and a p85 regulatory subunit — are the most important. PI3Kα (p110α, encoded by PIK3CA) is the isoform most frequently mutated in cancer. In normal signalling, activated RTKs recruit PI3Kα to phosphotyrosine motifs via p85 SH2 domains, relieving p85-mediated inhibition of p110α. The released p110α catalyses phosphorylation of PIP2 to PIP3 at the plasma membrane inner leaflet.

PTEN is the counteracting lipid phosphatase, converting PIP3 back to PIP2. PIK3CA hotspot activation can increase PIP3 production, while complete PTEN loss can reduce its removal. The effects depend on variant, allele state, protein expression and feedback; a single finding should not automatically be described as irreversible constitutive activation.

AKT Activation and Its 100+ Substrates

PIP3 recruits AKT through its pleckstrin-homology domain. PDK1 phosphorylation at Thr308 and mTORC2-associated phosphorylation at Ser473 contribute to activation, with the quantitative effect depending on isoform, substrate and context. A fixed percentage of maximal activity is not portable across assays.

AKT phosphorylates many substrates involved in survival, metabolism and growth, including TSC2, FOXO-family factors, GSK3 and BAD. The downstream outcome depends on cell type, timing and parallel pathways rather than one universal cancer-hallmark programme.

mTOR Complexes and Translational Control

mTOR forms two functionally distinct complexes: mTORC1 (with RAPTOR, mLST8, PRAS40, DEPTOR) drives anabolic biosynthesis, and mTORC2 (with RICTOR, mSin1, mLST8) activates AKT and regulates the cytoskeleton. mTORC1 integrates signals from growth factors (via AKT-TSC2-RHEB), amino acids (RAG GTPase-Ragulator lysosomal pathway), and cellular energy (AMPK opposition). When all conditions are met, mTORC1 phosphorylates S6K1 (Thr389) and 4EBP1 at multiple sites, releasing eIF4E from inhibition and driving cap-dependent translation of growth-promoting mRNAs.

S6K-dependent feedback can reduce upstream insulin-receptor-substrate signalling. Inhibiting mTORC1 can relieve this feedback and increase selected upstream or AKT readouts in some models. This is one of several resistance mechanisms and does not prove that broader dual inhibition will be more effective or tolerable.

Inhibitors and Biomarker Selection

PI3K-, AKT- and mTOR-directed agents inhibit different nodes and have been studied in different biomarker-defined populations. SOLAR-1 and CAPItello-291 are landmark examples, but their eligibility definitions, endocrine partners and endpoints should not be collapsed into one pathway-wide claim. Present-day indications and companion-test requirements should be verified against current labels.

Key Takeaways

• ·PIK3CA produces PIP3 and PTEN removes it; activating and loss-of-function events alter opposite sides of this lipid-signalling balance.
• ·PDK1 and mTORC2-associated phosphorylation contribute to AKT activation, which then changes a broad, context-dependent substrate network.
• ·Relief of S6K–IRS feedback is one possible consequence of mTORC1 inhibition, not the sole explanation for limited response.
• ·Biomarker-stratified PI3K- and AKT-inhibitor trials use different alteration lists and disease contexts; one definition cannot be substituted for another.
• ·The pathway intersects with receptor and RAS signalling across many cancers, but alteration prevalence and dependency vary by tumour type.