AKT1 E17K and capivasertib: drugging below PI3K-alpha

The PI3K/AKT/PTEN pathway is one of the most frequently dysregulated signaling axes in cancer, and most of the drug attention has landed on the top of it. But there is a structural argument for hitting the node one step down. AKT integrates the output of the whole upstream cassette, so an AKT inhibitor catches pathway activation no matter which of several upstream genes is broken. That logic is what put capivasertib on the market and what is now drawing a new wave of AKT-selective candidates into the clinic.

Where AKT sits, and why that matters

PI3K-alpha (the catalytic subunit encoded by PIK3CA) phosphorylates the membrane lipid PIP2 to PIP3. PTEN is the phosphatase that reverses that reaction. AKT is recruited to the membrane by PIP3 through its pleckstrin homology (PH) domain, gets phosphorylated, and then drives growth and survival signaling downstream. Three different lesions all converge on the same outcome of too much active AKT: an activating PIK3CA mutation that makes more PIP3, loss of PTEN that fails to clear PIP3, or an activating mutation in AKT itself.

Because AKT is the convergence point, an AKT inhibitor is agnostic to which upstream gene caused the trouble. A PI3K-alpha-selective drug only addresses the PIK3CA branch; it does nothing for PTEN loss or a direct AKT mutation. That is the clinical pitch for going one node lower.

AKT1 E17K: the textbook activating mutation

The canonical activating mutation in AKT itself is E17K, a glutamate-to-lysine swap in the PH domain first reported by Carpten and colleagues in 2007. The mutation changes the electrostatics of the lipid-binding pocket so that AKT1 is pulled to the membrane in a PIP3-independent way. In other words, the mutant AKT no longer waits for PI3K to generate the signal; it constitutively localizes and activates. E17K shows up in breast, endometrial, ovarian, and colorectal cancers, and it is the cleanest example of pathway activation that originates at AKT rather than above it.

Capivasertib: the first approved AKT inhibitor

Capivasertib (Truqap) is an oral, ATP-competitive pan-AKT inhibitor and the first AKT inhibitor to win FDA approval. The pivotal CAPItello-291 trial paired capivasertib with fulvestrant in hormone-receptor-positive, HER2-negative advanced breast cancer that had progressed on an aromatase inhibitor. The biomarker design is the interesting part: enrollment captured tumors with alterations in any of PIK3CA, AKT1, or PTEN, precisely the three upstream lesions that funnel into AKT.

• Capivasertib + fulvestrant — in the AKT-pathway-altered population, median progression-free survival was 7.3 months versus 3.1 months with placebo plus fulvestrant (hazard ratio roughly 0.50). The overall-population benefit was smaller but still significant, which is the data that drove the biomarker-restricted label.
• The dominant adverse events were diarrhea and rash, the on-target consequences of inhibiting AKT in normal tissue, and an intermittent four-days-on, three-days-off schedule is used to keep them manageable.

Capivasertib is ATP-competitive and pan-AKT, meaning it hits all three AKT isoforms. That breadth is part of why the tolerability ceiling exists, and it is the gap the next generation is trying to close.

What is next: isoform- and mutant-selective AKT drugs

The frontier is selectivity. A drug that spares AKT2 (heavily involved in insulin signaling and glucose handling) or that preferentially engages the E17K-mutant PH domain could widen the therapeutic window that diarrhea and hyperglycemia currently cap. In early June 2026, Targeted Oncology reported that the FDA granted fast-track designation to an oral AKT1-selective candidate for HR-positive/HER2-negative advanced breast cancer carrying AKT/PI3K/PTEN-pathway alterations, signaling that the isoform-selective thesis is now being tested clinically rather than just argued on paper. Selectivity claims like these are exactly the kind of thing worth checking against structure: an isoform- or mutant-selective binder should show a visible binding-mode or score difference between the targeted pocket and its near-neighbors.

Try the docking yourself

AKT sits one step below PI3K-alpha, which is in the Liganx target catalog. Open Studio and pick PI3K-alpha to dock against the node directly upstream of AKT, including the PIK3CA H1047R and E545K hotspots that drive the pathway from the top. Comparing how an inhibitor scores against wild-type versus the activating mutant is the same ΔΔ exercise that separates a real selectivity story from a single headline number, and it is the conversion moment for any pathway-targeting program.

Liganx puts molecular docking online and free in the browser, so running molecular docking across a pathway hotspot and its wild-type counterpart is a couple of clicks.

Primary sources

• Carpten JD, et al. A transforming mutation in the pleckstrin homology domain of AKT1 in cancer. Nature 448, 439-444 (2007). doi:10.1038/nature05933
• Turner NC, et al. Capivasertib in hormone receptor-positive advanced breast cancer. N Engl J Med 388, 2058-2070 (2023). doi:10.1056/NEJMoa2214131
• U.S. Food and Drug Administration. FDA approves capivasertib with fulvestrant for breast cancer. (November 16, 2023). fda.gov