PI3K (phosphoinositide 3-kinase family)
PI3Ks are a conserved family of lipid kinases that phosphorylate the 3′-OH of the inositol ring of phosphoinositides, generating lipid second messengers that recruit pleckstrin-homology (PH) domain proteins — principally akt and pdk1 — to the plasma membrane. Class I PI3Ks are the primary transducers downstream of receptor tyrosine kinases (RTKs), producing PIP3 (PI-3,4,5-P₃) from PIP2 (PI-4,5-P₂) and thereby activating the PI3K–AKT signaling axis 1. Class III PI3Ks (VPS34) generate PI3P (PI-3-P) to initiate autophagy and endosomal membrane trafficking. PI3Ks are among the most mutated enzyme families in human cancer and are central nodes in the IGF-1 signaling axis implicated in aging across multiple model organisms.
Naming note: This page covers the PI3K catalytic subunit family (all three classes). The pi3k-akt-pathway page covers the full signaling network downstream of Class I PI3K. The primary canonical anchor here is the Class IA p110α subunit (UniProt P42336/PIK3CA); other isoforms are listed in the isoform table below.
Identity (primary anchor: PI3Kα/PIK3CA)
- UniProt: P42336 (PI3KCA_HUMAN); Swiss-Prot reviewed
- Gene: PIK3CA (HGNC:8975)
- NCBI Gene: 5290
- Length: 1,068 amino acids (~120 kDa)
- Mouse ortholog: Pik3ca (one-to-one; class I PI3K signaling highly conserved)
PI3K class taxonomy
Three structurally and functionally distinct classes, all catalyzing 3′-phosphorylation of inositol phospholipids but differing in substrate, product, and cellular role 2:
| Class | Substrate | Product | Subcellular role | Isoforms (catalytic) |
|---|---|---|---|---|
| Class I | PIP2 (PI-4,5-P₂) | PIP3 (PI-3,4,5-P₃) | RTK/GPCR signaling → AKT/mTOR activation | p110α, p110β, p110γ, p110δ |
| Class II | PI, PI4P | PI3P, PI(3,4)P₂ | Membrane trafficking, clathrin-mediated endocytosis, angiogenesis | PI3K-C2α, PI3K-C2β, PI3K-C2γ |
| Class III | PI | PI3P (PI-3-P) | Autophagy initiation, endosomal sorting, lysosome biogenesis | VPS34 (single isoform) |
Class I and Class III PI3Ks are the most aging-relevant: Class I drives the IIS/mTOR anabolic signaling axis that determines lifespan in multiple organisms; Class III (VPS34) nucleates the autophagy initiation complex in partnership with beclin-1 3.
Isoform paralog table
| Isoform | Gene | UniProt | Class | Regulatory partner | Tissue/cell enrichment |
|---|---|---|---|---|---|
| p110α | PIK3CA | P42336 | IA | p85α/β/γ (PIK3R1/R2/R3) | Ubiquitous |
| p110β | PIK3CB | P42338 | IA | p85α/β/γ | Ubiquitous; dominant under PTEN loss |
| p110γ | PIK3CG | P48736 | IB | p101 (PIK3R5), p84 (PIK3R6) | Immune cells; GPCRs |
| p110δ | PIK3CD | O00329 | IA | p85α/β/γ | Immune cells, preadipocytes |
| PI3K-C2α | PIK3C2A | O00443 | II | None (monomeric) | Ubiquitous |
| PI3K-C2β | PIK3C2B | O00750 | II | None (monomeric) | Ubiquitous |
| PI3K-C2γ | PIK3C2G | Q8WYR1 | II | None (monomeric) | Liver, heart |
| VPS34 | PIK3C3 | Q8NEB9 | III | VPS15 (PIK3R4) + Beclin-1 | Ubiquitous; ER/endosomes |
needs-canonical-id — UniProt accessions for Class II and Class III isoforms listed from primary literature; not independently re-verified against UniProt in this pass.
Class I PI3K: structure and activation
Architecture
Class IA heterodimers consist of a p110 catalytic subunit + a p85 regulatory subunit (encoded by PIK3R1/R2/R3 for Class IA; p101/PIK3R5 for Class IB). In the resting state, the regulatory subunit’s inter-SH2 (iSH2) domain contacts the p110 catalytic subunit and suppresses its basal lipid kinase activity, preventing spurious PIP3 generation 1.
The p110 catalytic subunits share a conserved domain architecture:
- ABD (adaptor-binding domain) — binds the iSH2 domain of the regulatory subunit
- RBD (Ras-binding domain) — mediates interaction with GTP-loaded Ras family GTPases (particularly Ras for p110α; Rab for others)
- C2 domain — membrane anchoring; contributes to membrane recruitment
- Helical domain — site of activating hotspot mutations in PIK3CA (E542, E545 reside here)
- Kinase domain — catalytic ATP- and lipid-binding; H1047 hotspot resides here
RTK-mediated activation
Upon insulin, IGF-1, EGF, or other RTK ligand binding, autophosphorylated pYxxM motifs on the activated receptor (or on scaffold proteins such as IRS-2) recruit the SH2 domains of p85, relieving p85-mediated inhibition of p110 1. Activated p110 then catalyzes:
PIP2 (PI-4,5-P₂) → PIP3 (PI-3,4,5-P₃)
at the inner leaflet of the plasma membrane. PIP3 recruits akt and pdk1 simultaneously via their PH domains, co-localizing the two kinases and enabling PDK1 to phosphorylate AKT Thr308, partially activating AKT 4.
Class IB p110γ is activated via Gβγ subunits of heterotrimeric G-proteins (GPCRs), not RTKs — this is relevant to immune cell chemokine signaling but less directly to the IIS/aging axis.
Negative regulation by PTEN
PTEN (Phosphatase and TENsin homolog) is the principal antagonist of Class I PI3K: it dephosphorylates PIP3 → PIP2, directly reversing the PI3K reaction and terminating AKT recruitment. PTEN is among the most frequently lost tumor suppressors in human cancer; heterozygous PTEN loss increases PI3K–AKT tone and is associated with accelerated cellular aging phenotypes in some contexts 1. See pten for the full protein page.
Class III PI3K (VPS34) and autophagy
VPS34 (PIK3C3) is the sole Class III PI3K in mammals. Unlike Class I, it phosphorylates PI → PI3P and does not generate PIP3. VPS34 associates constitutively with the regulatory kinase VPS15 (PIK3R4) and functions in two distinct complexes scaffolded by beclin-1:
- PI3KC3-C1 (Beclin-1 + VPS34 + VPS15 + ATG14L): generates PI3P at ER-associated phagophore assembly sites → recruits WIPI2 and downstream ATG machinery → autophagosome nucleation
- PI3KC3-C2 (Beclin-1 + VPS34 + VPS15 + UVRAG): acts at late endosomes/lysosomes → endosomal membrane trafficking and autophagosome–lysosome fusion
Because VPS34-generated PI3P is essential for autophagy initiation, VPS34 activity connects mTORC1-mediated nutrient sensing to autophagic flux: mTORC1 suppresses autophagy in part by phosphorylating ATG14L (a PI3KC3-C1 subunit) and by inhibiting ULK1, which in turn activates Beclin-1 via Ser30 phosphorylation. Under nutrient deprivation (mTORC1 inhibited), this brake is released, VPS34 complex I becomes active, and autophagy is initiated 3. See beclin-1 and autophagy for detail.
Role in aging
PI3K as the molecular executor of IIS lifespan control
Reduced insulin/IGF-1 signaling (IIS) extends lifespan in every organism examined — from C. elegans (daf-2 LOF) to flies (dilp2/dilp3/dilp5 mutants) to mice (Igf1r+/−, GHRKO) 1. PI3K is the obligate executor of this effect: all known lifespan-extending IIS mutations reduce flux through Class I PI3K → reduced PIP3 → reduced AKT activity → FOXO transcription factors (foxo3 ortholog DAF-16 in worms, dFOXO in flies) released to the nucleus → pro-longevity gene expression (antioxidant defense, DNA repair, proteostasis, autophagy).
Direct PI3K perturbations confirming this:
- Age-1 (PI3K catalytic subunit ortholog in C. elegans) LOF — the first well-characterized single-gene longevity mutation: age-1(hx546) extends mean lifespan ~40% and maximum lifespan ~60% at 20°C (and ~65%/~110% at 25°C) in C. elegans 5. The authors note the Age phenotype co-segregates with reduced hermaphrodite fertility (fer-15 locus) and caution that lifespan extension may be a pleiotropic effect of reduced fertility. The molecular identity of age-1 as the PI3K catalytic subunit was established by later work. needs-replication (single organism; molecular pathway clarified post-1988)
- p110α heterozygous knockout mice (Pik3ca+/−) — show improved insulin sensitivity in female mice; a ~18% lifespan extension in female mice (C57BL/6 background) has been reported. needs-human-replication; single study, mouse only; not ITP-validated. Note: the ~18% lifespan figure is NOT from Engelman 2006 (which does not report lifespan outcomes for Pik3ca+/− mice) — original source citation needed. unsourced
- Drosophila PI3K (dp110) overactivation — reduces lifespan; dominant negative dp110 extends lifespan; effect is dFOXO-dependent needs-human-replication
| Dimension | Status |
|---|---|
| Pathway conserved in humans? | yes — IIS/PI3K/AKT/FOXO signaling fully conserved |
| Phenotype (lifespan extension via PI3K reduction) conserved in humans? | partial — FOXO3 GWAS associations consistent; no human PI3K intervention data |
| Replicated in humans? | no — only observational GWAS and epidemiological evidence |
PI3Kδ and preadipocyte senescence
PI3Kδ (p110δ/PIK3CD) has a specific role in the survival anti-apoptotic pathway (SCAP) of senescent preadipocytes. Idelalisib (PI3Kδ-selective inhibitor) is part of the mechanism by which the dasatinib + quercetin (D+Q) senolytic combination selectively eliminates senescent preadipocytes while sparing non-senescent cells 6. This is an isoform-selective aging-relevant function distinct from the pan-PI3K IIS role. See senolytics for the D+Q preadipocyte SCAP detail.
PI3K hyperactivation and aging pathology
Chronically elevated PI3K activity (as in metabolic syndrome, obesity, aging adipose tissue) drives aging-associated pathology via:
- mTORC1 hyperactivation → suppressed autophagy → aggregate accumulation → disabled-macroautophagy and loss-of-proteostasis hallmarks
- FOXO nuclear exclusion → lost pro-longevity transcriptional program
- Senescent cell persistence → AKT→BAD phosphorylation protects senescent cells from apoptosis → chronic SASP → cellular-senescence hallmark (BAD phosphorylation by AKT confirmed in Engelman 2006 Box 3; Ser136 site designation requires separate citation from original BAD phosphorylation literature)
- Oncogenesis → PIK3CA gain-of-function mutations in ~30% of solid tumors 7
PIK3CA hotspot mutations (cancer context)
PIK3CA is the most-mutated kinase gene in human cancer. Three gain-of-function hotspot mutations account for the majority of cases 7:
| Mutation | Domain | Frequency | Mechanism |
|---|---|---|---|
| E542K | Helical domain | ~10% of PIK3CA-mutant cancers | Disrupts inhibitory contact with p85 regulatory subunit N-SH2; constitutive activation |
| E545K | Helical domain | ~25% of PIK3CA-mutant cancers | Same as E542K; most common helical hotspot |
| H1047R | Kinase domain (activation loop) | ~35% of PIK3CA-mutant cancers | Reorients activation loop toward membrane; increased membrane affinity + catalytic activity |
These mutations drive constitutive PIP3 generation independent of RTK input. The FDA-approved PI3Kα-selective inhibitor alpelisib (see alpelisib) specifically targets PIK3CA-mutant tumors 8.
Pharmacological targeting
| Drug | Target | Class specificity | Clinical status | Aging relevance |
|---|---|---|---|---|
| Alpelisib (BYL719) | p110α | PI3Kα-selective | FDA-approved: PIK3CA-mutant HR+/HER2− breast cancer (2019) | Cancer; hyperglycemia limits aging use |
| Idelalisib | p110δ | PI3Kδ-selective | FDA-approved: CLL, follicular lymphoma | Preadipocyte senolysis (D+Q) |
| Copanlisib | All class I | Pan-PI3K | FDA-approved (follicular lymphoma) | Research tool; too toxic for aging |
| Gedatolisib | PI3K + mTOR catalytic | Dual PI3K/mTOR | Phase 3 (breast cancer) | Research |
| LY294002 | All class I + class III (VPS34) | Pan | Research tool only (not clinical) | Gold-standard in vitro PI3K inhibitor |
| Wortmannin | All class I + class III | Covalent, pan | Research tool only (not clinical) | Irreversible inhibitor; fungal metabolite |
Metabolic safety caveat: At anti-cancer doses, PI3K inhibition disrupts insulin signaling in liver and muscle, causing grade 3 or 4 hyperglycemia in 36.6% of patients on alpelisib in the SOLAR-1 trial 8. This on-target toxicity makes full-dose PI3K inhibition inappropriate for aging geroprotection, motivating indirect approaches (metformin, AMPK activation) and isoform-selective strategies 6.
Cross-pathway connections
- pi3k-akt-pathway — canonical downstream signaling network; this page covers the PI3K catalytic family only; full pathway on the dedicated page
- akt — immediate downstream effector of PIP3 (Class I); PH-domain recruitment → PDK1/mTORC2 dual phosphorylation
- pdk1 — co-recruited to plasma membrane by PIP3; phosphorylates AKT Thr308
- pten — antagonist; dephosphorylates PIP3 → PIP2; counteracts Class I PI3K
- mtor — downstream of AKT (via TSC2 inhibition); mTORC2 feeds back to phosphorylate AKT Ser473
- beclin-1 — Class III PI3K (VPS34) partner; scaffolds PI3KC3 complexes for autophagy initiation
- autophagy — Class III PI3K is essential for autophagosome nucleation; Class I suppresses autophagy via mTORC1
- insulin-igf1 — RTK/IRS-1 upstream; IIS reduction extends lifespan via reduced Class I PI3K flux
- ampk — antagonizes Class I PI3K net effect: AMPK activates TSC2 and inhibits mTORC1 even when AKT/PI3K is active
Limitations and gaps
- #gap/needs-human-replication — All lifespan-extension evidence from reduced Class I PI3K flux is from model organisms (C. elegans, Drosophila, mouse). Human observational data (FOXO3 GWAS) is consistent but not interventional.
- #gap/needs-replication — Direct PI3K reduction (e.g., Pik3ca+/−) lifespan data in mice is from limited studies in specific genetic backgrounds; not ITP-validated. The Pik3ca+/- lifespan claim (~18% female lifespan extension) requires its own primary citation; this result is NOT in Engelman 2006 and the original source must be identified.
- #gap/needs-replication — The original age-1 study (Friedman & Johnson 1988) acknowledges that the Age phenotype co-segregates with reduced hermaphrodite fertility (fer-15 locus) and that life extension may be a pleiotropic effect of reduced fertility, not a direct PI3K-aging mechanism. Subsequent work confirmed age-1 = PI3K, but the fertility confound interpretation was not fully resolved in the original paper.
- #gap/dose-response-unclear — Optimal degree of Class I PI3K attenuation for geroprotection in humans is entirely unknown; anti-cancer doses cause dose-limiting metabolic toxicity.
- #gap/long-term-unknown — Class II PI3K isoforms’ roles in aging are largely unstudied; PI3K-C2α has roles in insulin-stimulated glucose transport and cilia signaling but aging-specific functions are not established.
- #gap/needs-canonical-id — GenAge entry IDs for PIK3CA family members not confirmed in this pass; Class III VPS34 (PIK3C3) may have a GenAge entry.
- Class I vs Class III distinction in ageing interventions — Many pan-PI3K inhibitors (wortmannin, LY294002) block both Class I and VPS34 (Class III), simultaneously inhibiting AKT and autophagy — these have opposing effects on aging. Clean isoform-selective tools are critical for aging research interpretation. no-mechanism for the net effect of pan-PI3K inhibition in aging contexts.
Footnotes
Footnotes
-
doi:10.1038/nrg1879 · engelman-luo-cantley-2006-pi3k-review · review · model: multi-organism (yeast, worm, fly, mouse, human) · Engelman JA, Luo J, Cantley LC · Nature Reviews Genetics 2006 · ~3,271 citations · evolutionary and functional review of PI3K family; Class I activation mechanism; IIS/PI3K conservation across organisms (Table 1: age-1/AGE-1 = C. elegans PI3K ortholog); mouse genetic studies of p110 knockouts (germline deletion lethal; kinase-dead p110α heterozygotes show glucose intolerance — lifespan outcomes NOT reported in this paper); PI3K in tumorigenesis (PIK3CA hotspot mutations) · archive: locally downloaded ↩ ↩2 ↩3 ↩4 ↩5
-
doi:10.1038/nrm2882 · review · model: mammalian signaling (human, mouse) · Vanhaesebroeck B, Guillermet-Guibert J, Graupera M, Bilanges B · Nature Reviews Molecular Cell Biology 2010;11(5):329–341 · 1,702 citations · comprehensive isoform-specific PI3K signaling review; Class I/II/III taxonomy (Fig. 2); Class IA p85 regulatory subunit function; PIK3CA hotspot mutation mechanisms (E545K helical, H1047R kinase domain); VPS34 complex organization (Fig. 4b: PI3KC3-C1 with ATG14L/barkor for autophagy, PI3KC3-C2 with UVRAG for endosomal trafficking); Class II biology · archive: locally downloaded ↩
-
doi:10.1038/s41580-019-0129-z · review · model: mammalian (human, mouse) · Bilanges B, Posor Y, Vanhaesebroeck B · Nature Reviews Molecular Cell Biology 2019 · 510 citations · PI3K isoforms in cell signalling and vesicle trafficking; Class III VPS34/autophagy section · archive: not_oa (not downloaded) ↩ ↩2
-
doi:10.1016/j.cell.2017.04.001 · manning-toker-2017-akt-signaling · review · model: human (cellular and biochemical) · Manning BD, Toker A · Cell 2017;169(3):381–405 (April 20, 2017) · 3,500+ citations · AKT dual phosphorylation: PDK1 phosphorylates Thr308 (activation loop) and mTORC2 phosphorylates Ser473 (hydrophobic motif); both required for full activation; PIP3/PI3,4P2 PH-domain recruitment co-localizes AKT with PDK1; downstream AKT substrates include TSC2 (mTORC1 activation), FOXO (nuclear exclusion), GSK3, BAD; note: Manning 2017 covers AKT signaling, not PI3K family classification — cited here only for AKT Thr308/Ser473 mechanism · archive: locally downloaded ↩
-
doi:10.1093/genetics/118.1.75 · in-vivo (C. elegans) · Friedman DB, Johnson TE · Genetics 1988;118(1):75–86 · age-1(hx546) recessive mutant: mean lifespan extended ~40% at 20°C and ~65% at 25°C; maximum lifespan extended ~60% at 20°C and ~110% at 25°C (Table 1); authors note the Age phenotype co-segregates with fer-15 (reduced hermaphrodite fertility) and caution that life extension may be a pleiotropic effect of reduced fertility rather than a primary aging gene function; paper identifies age-1 as “the only instance of a well-characterized genetic locus in which the mutant form results in lengthened life” · archive: locally downloaded ↩
-
doi:10.1038/s41571-022-00633-1 · review · model: human (clinical + preclinical) · Vasan N, Cantley LC · Nature Reviews Clinical Oncology 2022 · 164 citations · PI3K inhibitor toxicity (grade ≥3 hyperglycemia ~37% alpelisib arm from SOLAR-1), feedback mechanisms, and clinical trial lessons · claims partially verified via PMC11215755 (green OA); full PDF download failed ↩ ↩2
-
doi:10.1126/science.1096502 · in-vitro (Sanger sequencing of human tumor samples) · Samuels Y et al. · Science 2004 · 3,400 citations · identified PIK3CA somatic mutations in 32% of colorectal cancers and in multiple other tumor types; E542K, E545K (helical domain), H1047R (kinase domain) as dominant hotspots · archive: not_oa (not downloaded) ↩ ↩2
-
doi:10.1056/NEJMoa1813904 · rct (Phase III SOLAR-1 trial) · n=572 total randomized; 341 PIK3CA-mutant cohort (169 alpelisib + 172 placebo) · André F et al. · NEJM 2019;380(20):1929–40 · 2,360 citations · alpelisib 300 mg/day + fulvestrant vs placebo + fulvestrant in PIK3CA-mutant HR+/HER2− advanced breast cancer; primary endpoint PFS: 11.0 months (95% CI 7.5–14.5) vs 5.7 months (95% CI 3.7–7.4); HR 0.65 (95% CI 0.50–0.85); P<0.001; grade 3 or 4 hyperglycemia: 36.6% alpelisib arm vs 0.7% placebo (Table 3) · archive: locally downloaded (verified) ↩ ↩2