PI3Kβ prioritizes membrane interactions with RTK-derived pY peptides over Rac1(GTP) and GβGγ

(A) Cartoon schematic showing membrane tethered signaling inputs (i.e. pY, Rac1(GTP), and GβGγ) attached to a supported lipid bilayer and visualized by TIRF-M. Heterodimeric Dy647-PI3Kβ (p110β-p85α) in solution can dynamically associate with membrane bound proteins. (B) Cartoon schematic showing method for visualizing membrane tethered signaling inputs. (C) Kinetics of Rac1 nucleotide exchange measured in the presence of 20 nM Rac1(GTP) sensor (Cy3-p67/phox) and 50 nM P-Rex1. (D) Visualization of membrane conjugated RTK derived pY peptide (∼6,000/μm2), Rac1(GTP) (∼4,000/μm2), and GβGγ (∼4,800/μm2) by TIRF-M. Representative TIRF-M images showing the membrane localization of 20 nM nSH2-Cy3 in the absence and presence of membranes conjugated with a solution concentration of 10 μM pY peptide. Representative images showing the membrane localization of 20 nM Cy3-p67/phox Rac1(GTP) sensor before (GDP) and after (GTP) the addition of the guanine nucleotide exchange factor, P-Rex1. Equilibrium localization of 50 nM (low) or 200 nM (high) farnesyl GβGγ-SNAP-AF488. (E) Representative TIRF-M images showing the equilibrium membrane localization of 10 nM Dy647-PI3Kγ measured in the absence and presence of membranes equilibrated with 200 nM farnesyl GβGγ. (F) Representative TIRF-M images showing the equilibrium membrane localization of 5 pM and 10 nM Dy647-PI3Kβ measured in the presence of membranes containing either pY, Rac1(GTP), or GβGγ. The inset image (+GβGγ) shows low frequency single molecule binding events detected in the presence of 10 nM Dy647-PI3Kβ. (G) Bulk membrane absorption kinetics for 10 nM Dy647-PI3Kβ measured on membranes containing either pY, Rac1(GTP), or GβGγ. Membrane composition: 96% DOPC, 2% PI(4,5)P2, 2% MCC-PE.

Density dependent membrane binding behavior of Dy647-PI3Kβ measured in the presence of RTK-derived pY peptides

(A) Cartoon schematic showing conjugation of pY peptides (+/-Alexa488 label) using thiol reactive maleimide lipids (MCC-PE). (B) Representative image showing the single molecule localization of Alexa488-pY. Particle detection (purple circles) was used to quantify the number of pY peptides per μm2. (C) Relationship between the total pY solution concentration (x-axis) used for covalent conjugation, the bulk membrane intensity of covalently attached Alexa488-pY (left y-axis), and the final surface density of pY peptides per μm2 (right y-axis). (D) Relationship between the total pY solution conjugation concentration and bulk membrane intensity of measured in the presence of 50 nM nSH2-Cy3. (E-G) Membrane localization dynamics of Dy647-PI3Kβ measured on SLBs containing a range of pY surface densities (250–15,000 pY/μm2, based on Figure 1C). (E) Bulk membrane localization of 10 nM Dy647-PI3Kβ as a function of pY density. (F) Single molecule dwell time distributions measured in the presence of 5 pM Dy647-PI3Kβ. Data plotted as log10(1–CDF) (cumulative distribution frequency). (G) Step size distributions showing Dy647-PI3Kβ single molecule displacements from > 500 particles (>10,000 steps) per pY surface density. (H-J) Membrane localization dynamics of Dy647-PI3Kβ nSH2(R358A) and cSH2(R649A) mutants measured on SLBs containing ∼15,000 pY/μm2 (10μM conjugation concentration). (H) Bulk membrane absorption kinetics of 10 nM Dy647-PI3Kβ (WT, nSH2*, and cSH2*). (I) Single molecule dwell time distributions measured in the presence of 5 pM Dy647-PI3Kβ (WT, nSH2*, and cSH2*). Data plotted as log10(1–CDF) (cumulative distribution frequency). (J) Step size distributions showing single molecule displacements of > 500 particles (>10,000 steps) in the presence of 5 pM Dy647-PI3Kβ (WT, nSH2*, and cSH2*). Membrane composition: 96% DOPC, 2% PI(4,5)P2, 2% MCC-PE.

Mechanism controlling synergistic Dy647-PI3Kβ membrane binding by pY and GβGγ

(A) Kinetic trace showing the membrane absorption of 200 nM AF488-SNAP-GβGγ and AF555-SNAP-GβGγ (0.0025%) measured by TIRF-M. Single molecule densities of AF555-SNAP-GβGγ were calculated for each frame in a field of view of 3000 μm2. (B) Representative TIRF-M images showing the equilibrium membrane localization of 5 pM and 10 nM Dy647-PI3Kβ on membranes containing either pY, GβGγ, pY/GβGγ, or pY(solution)/GβGγ. The inset image (+GβGγ and +pY/GβGγ) shows low frequency single molecule binding events detected in the presence of 10 nM Dy647-PI3Kβ. Supported membranes were conjugated with 10 μM pY peptide (final surface density of ∼15,000 pY/μm2) and equilibrated with 200 nM farnesyl-GβGγ before adding Dy647-PI3Kβ. pY(solution) = 10 μM. (C) Bulk membrane recruitment dynamics of 10 nM Dy647-PI3Kβ measured in the presence of either pY alone, pY/GβGγ, or pY(solution)/GβGγ. pY(solution) = 10 μM. (D) Single molecule dwell time distributions measured in the presence of 5 pM Dy647-PI3Kβ on supported membranes containing pY alone (τ1=0.55±0.11s, τ2=1.44±0.56s, α=0.54, N=4698 particles, n=5 technical replicates) or pY/GβGγ (τ1=0.61±0.13s, τ2=3.09±0.27s, α=0.58, N=3421 particles, n=4 technical replicates). Alpha(α) represents the fraction of particles characterized by the time constant (τ1). (E) Step size distributions showing single molecule displacements measured in the presence of either pY alone (D1=0.34±0.04 μm2/sec, D2=1.02±0.07 μm2/sec, α=0.45) or pY/GβGγ (D1=0.23±0.03 μm2/sec, D2=0.88±0.08 μm2/sec, α=0.6); n=3-4 technical replicates from > 3000 tracked particles with 10,000-30,000 total displacements measured. Alpha(α) represents the fraction of particles characterized by the time constant (D1). (F) Combined model of the putative nSH2 and GβGγ binding sites on p110β. The p110β-GβGγ binding site is based on an Alphafold multimer model supported by previous HDX-MS and mutagenesis experiments. The orientation of the nSH2 is based on previous X-ray crystallographic data on PI3Kα (p110α-p85α, niSH2, PDB:3HHM) aligned to the structure of PI3Kβ (p110β-p85α, icSH2, PDB:2Y3A). (G) Bulk membrane recruitment dynamics of 10 nM Dy647-PI3Kβ, WT and nSH2(R358A), measured on membranes containing either pY or pY/GβGγ. (H) Bulk membrane recruitment dynamics of 10 nM Dy647-PI3Kβ, WT and GβGγ binding mutant, measured on membranes containing either pY or pY/GβGγ. (A-H) Membrane composition: 96% DOPC, 2% PI(4,5)P2, 2% MCC-PE.

Membrane anchored pY peptides synergistically enhance Dy647-PI3Kβ membrane binding in the presence of Rac1(GTP)

(A) Cartoon schematic showing membrane conjugation of Cy3-Rac1 and AF488-pY on membranes containing unlabeled Rac1 and pY. (B) Representative TIRF-M images showing localization of Cy3-Rac1 (1:10,000 dilution) and AF488-pY (1:30,000 dilution) after membrane conjugation in the presence of 30 μM Rac1 and 10 μM pY. Membrane surface density equals ∼4,000 Rac1/μm2 and ∼5,000 pY/μm2. (C) Representative TIRF-M images showing the equilibrium membrane localization of 5 pM and 10 nM Dy647-PI3Kβ measured in the presence of membranes containing either pY/Rac1(GDP) or pY/Rac1(GTP). (D) Bulk membrane recruitment dynamics of 10 nM Dy647-PI3Kβ measured in the presence of pY/Rac1(GDP) or pY/Rac1(GTP). (E) Single molecule dwell time distributions measured in the presence of 5 pM Dy647-PI3Kβ on supported membranes containing pY/Rac1(GDP) or pY/Rac1(GTP). (F) Step size distributions showing single molecule displacements from > 500 Dy647-PI3Kβ particles (>10,000 steps) in the presence of either pY/Rac1(GDP) or pY/Rac1(GTP). Membrane composition: 96% DOPC, 2% PI(4,5)P2, 2% MCC-PE.

GβGγ and Rac1(GTP) stimulate PI3Kβ activity beyond enhancing localization on pY membranes

(A) Representative TIRF-M images showing localization of 20 nM Btk-SNAP-AF488 on SLBs containing either 2% PI(4,5)P2 or 2% PI(3,4,5)P3, plus 98% DOPC. (B) Bulk membrane recruitment kinetics of 20 nM Btk-SNAP-AF488 on an SLB measured by TIRF-M. (C-D) Kinetics of PI(3,4,5)P3 production measured in the presence of 10 nM Dy647-PI3Kβ and 1 mM ATP on SLBs with membrane anchored pY, Rac1(GTP), or GβGγ alone. Reactions in (C) were performed in the absence of PS lipids, while membranes in (D) contained 20% DOPS. (E) Cartoon schematic illustrating method for measuring Dy647-PI3Kβ activity in the presence of either pY/Rac1(GDP) or pY/ Rac1(GTP). Phase 1 of the reconstitution involves membrane equilibration of Dy647-PI3Kβ in the absence of ATP. During phase 2, 1 mM ATP is added to stimulate lipid kinase activity of Dy647-PI3Kβ. (F) Dual color TIRF-M imaging showing 2 nM Dy647-PI3Kβ localization and catalysis measured in the presence of 20nM Btk-SNAP-AF488. Dashed line represents the addition of 1 mM ATP to the reaction chamber. (G) Cartoon schematic showing experimental design for measuring synergistic binding and activation of Dy647-PI3Kβ in the presence of pY and GβGγ. (H) Representative single molecule TIRF-M images showing the localization of 20 pM Dy647-PI3Kβ in (G). (I) Kinetics of PI(3,4,5)P3 production monitored in the presence of 20 nM Btk-SNAP-AF488 and 20 pM Dy647-PI3Kβ. Membrane contained either pY or pY/GβGγ. (B, C, F, H, I) Membrane composition: 96% DOPC, 2% PI(4,5)P2, 2% MCC-PE. (D) Membrane composition: 76% DOPC, 20% DOPS, 2% PI(4,5)P2, 2% MCC-PE. All kinetic measurements of PI(3,4,5)P3 production were performed in the presence of 20 nM Btk-SNAP-AF488.