Scientists Crack How mTORC2 Selectively Activates Akt in a Longevity Pathway

The mTOR kinase assembles into two distinct mega-complexes—mTORC1 and mTORC2—that govern fundamentally different cellular signaling programs. mTORC2 is a critical node in PI3K and Ras signaling, phosphorylating AGC-family kinases Akt, PKC, and SGK1 to regulate metabolism, survival, and growth. Despite decades of study, the molecular basis for how mTORC2 selectively recognizes these substrates over closely related kinases had remained unknown, partly because kinase–substrate interactions are inherently transient and difficult to capture structurally.

To overcome this, the team developed semisynthetic Akt proteins using expressed protein ligation, attaching synthetic C-terminal peptides containing either unmodified or phosphorylated Ser473. They then engineered 'bisubstrate inhibitors'—Akt molecules covalently tethered to ATP or the mTOR inhibitor Torin1 at Ser473—to stabilize the otherwise fleeting mTORC2–Akt complex. These trapped complexes were analyzed by cryo-EM and cross-linking mass spectrometry, yielding high-resolution structural snapshots of Akt docked within mTORC2.

The structural data revealed a striking finding: mTORC2 does not rely primarily on the local amino acid sequence flanking the phosphorylation site. Instead, it recognizes secondary and tertiary structural features of Akt—specifically surface elements ~75 Å from the mTOR active site—through the mSin1 subunit's conserved region in the middle (CRIM) and pleckstrin homology (PH) domains. These recognition surfaces are conserved across at least 18 related AGC-family substrates, explaining the shared selectivity pattern. Biochemical assays confirmed that purified mTORC2 directly phosphorylates Akt Ser473 (not via Akt autophosphorylation), and that mTORC1 and mTORC2 show ~140–150-fold preference for their respective canonical substrates in vitro.

The study also elucidated a multi-step mechanism in which membrane localization, mSin1-substrate contacts, and active-site engagement together drive selectivity—explaining why inhibiting the shared mTOR catalytic site alone cannot differentiate between mTORC1 and mTORC2. These structural insights suggest that allosteric inhibitors targeting the mSin1–substrate interface could selectively block mTORC2 without disrupting mTORC1, a long-sought pharmacological goal.

Because overactive mTORC2–Akt signaling drives cancer, metabolic syndrome, and aging-associated pathologies, and because current pan-mTOR inhibitors are too toxic for broad clinical use, this structural framework offers a molecular blueprint for next-generation selective therapeutics.