Figure 1.
Ras, PI3KC2β, and ITSN1 co-localize on intracellular vesicles.
(A) VN-Ras and VC-PI3KC2β (green) were co-transfected with CFP-ITSN1 (red) into COS cells. a. The PI3KC2β-Ras BiFC complex co-localizes with ITSN, represented by yellow in the overlay panel; b. The VN-Ras and VC-PI3KC2β YFP signal (green) does not bleed into the CFP channel; c. The CFP-ITSN1 signal does not bleed into the YFP channel (size bars = 20 μm). Note: the differences in signal strength of the BiFC signal in (a) vs (b) are due to a lower power setting for the laser in (b) so that pixel intensities can be accurately quantified and are not saturated. In (a) a higher laser power was used to illustrate the punctate localization of the Ras-PI3KC2β complex throughout the cell. (B) Ras interaction with PI3KC2β is disrupted by deletion of the RBD (ΔRBD) but not by mutation of the Pro-rich, ITSN1 binding sites (PRD-PA). The graph represents the average fluorescence intensity per cell ± S.E.M. from at least three independent experiments (*p<0.05, WT vs ΔRBD, PRD-PA vs ΔRBD). Western blot analysis demonstrates equal expression of all constructs (size bars = 20 μm).
Figure 2.
Ras is necessary for ITSN1 activation of AKT.
(A) YFP-ITSN1 overexpression stimulates AKT activation as measured by levels of phospho-AKT as previously described [22]. Co-expression of Ras17N or Ras17N/69N inhibits this response. Results represent the average fold activation of AKT ± S.E.M. from at least three independent experiments. (B) Western blot analysis of AKT activation from a representative experiment. Top two panels represent Western blots of HA immunoprecipitates of cell lysates to assess AKT activation as described in the Materials and Methods section. The lower three panels indicate the level of expression of ITSN1, Ras, and actin (a loading control).
Figure 3.
Ras forms a complex with PI3KC2β.
(A) PI3KC2β preferentially interacts with Ras17N. VC-tagged PI3KC2β was co-transfected with one of the following VN-tagged Ras constructs: WT, 61L, 17N, or 17N/69N. BiFC signal (green) demonstrates that PI3KC2β interacted with Ras17N >17N/69N >WT >61L. CFP (red) was used as a transfection control. The graph represents the average fluorescence intensity per cell ± S.E.M. from at least three independent (*p<0.05). Western blot analysis demonstrates equivalent expression of all constructs. (B) PI3KC2β does not interact with Ras12V. VC-tagged PI3KC2β was co-transfected with one of the following VN-tagged Ras constructs: WT,17N, or 12V. BiFC signal (green) demonstrates that PI3KC2β interacted with Ras17N >WT >12V. CFP (red) was used as a transfection control. The graph represents the average fluorescence intensity per cell ± S.E.M. from at least three independent experiments (*p<0.05). Western blot analysis demonstrates expression of all constructs (size bar = 20 μm).
Figure 4.
ITSN1 and Ras form a BiFC complex.
(A) VC-tagged ITSN1 was co-transfected with one of the following VN-tagged Ras constructs: WT, 61L, 17N, or 17N/69N. ITSN1 formed a complex (green) with Ras17N >WT >17N69N >61L. CFP (red) was used as a transfection control. (B) The graph represents the average fluorescence intensity per cell ± S.E.M. from at least three independent experiments (*p<0.05, **p<0.01). (C) A Western blot was performed to demonstrate equal expression of all constructs. (size bar = 20 μm).
Figure 5.
Binding of PI3KC2β-RBD to Ras.
(A) Nucleotide loaded Ras does not directly interact with the RBD of PI3KC2β. Ras-GDP or Ras-GTPγS were incubated with GST-Raf-RBD, GST-PI3KC2β-RBD or GST alone as a negative control. Bound proteins were then analyzed by Western blot with a Ras antibody. The RBD of Raf specifically bound Ras-GTPγS. Neither GST-PI3KC2β-RBD nor GST alone interacted with Ras-GDP or Ras-GTPγS. Top panel, Ras bound to GST proteins. Bottom panels, input amounts of proteins. (B) Nucleotide-free Ras was generated in vitro as described and then tested for binding to the various GST proteins as in (A). GST-PI3KC2β-RBD directly binds nucleotide-free Ras while little association was seen with the GST-Raf-RBD or GST alone. Panels are same as in A. (C) Repeat of (B) except nucleotide (1 mM) was present during the binding reaction. Panels are same as in (A). (D) Addition of nucleotide (1 mM) does not disrupt pre-bound PI3KC2β-RBD- nucleotide-free Ras. GST-PI3KC2β-RBD was first bound to nucleotide-free Ras. Following binding, the complex was incubated with 1 mM GDP or GTPγS at RT for 30 min and then washed with buffer. Bound proteins were then analyzed as in (A).
Figure 6.
Nucleotide free Ras inhibits PI3KC2β activity.
(A) The in vitro kinase activity of purified PI3KC2β was assayed in the presence of nucleotide-free (nf) Ras or Ras loaded with GTPγS (5 μM, 2.5 μM, and 1 μM). LY294002 (10 μM) was used as a negative control. The radioactive phospholipid products were extracted and eluted on a TLC plate. (B) Nucleotide-free Ras dose dependently inhibits PI3KC2β. LY294002 (10 μM; LY) inhibits PI3KC2β activity while Ras-GTP (5 μM, 2.5 μM, and 1 μM) does not. The graph represents relative kinase activity normalized to PI3KC2β alone ± S.E.M. from 4–6 independent experiments.
Figure 7.
Mutations in the effector region of Ras disrupt PI3KC2β binding.
(A) Point mutations in the effector region of Ras12V disrupt interactions with specific Ras targets. (B) VC-tagged PI3KC2β was co-transfected with either VN-tagged Ras17N, 17N/69N, or one the effector mutants in the background of Ras17N/69N. BiFC signal is pseudo-colored green. Effector mutations that disrupt Class I PI3K binding to Ras12V disrupt Class II PI3K binding to Ras17N/69N. CFP (red) was used as a transfection control. (C) Graph represents the average fluorescence intensity per cell ± S.E.M. from at least three independent experiments. (*p = 0.02). (D) Western blot analysis demonstrates equal expression of all constructs. (E) Mutation of Thr392 to Asp or Lys379 to Ala in full-length PI3KC2β disrupts interaction with Ras17N. The ΔRBD mutant was also included as a negative control. Graph represents the average of three independent experiments (*p<0.05).
Figure 8.
Biological activity of PI3KC2β versus Raf RBDs.
(A) Expression of the Raf-RBD but not the PI3KC2β-RBD or GST alone inhibited EGF-stimulation of a Gal-Elk reporter assay [23]. GST-RBDs were expressed equally. GST alone migrated at a faster rate and was not visible in this image. Results represent the average relative activation ± S.E.M. from at least three independent experiments. (*p<0.05 compared to unstimulated GST, **p<0.01 compared to EGF stimulated GST). (B). PI3KC2β-RBD dose-dependently inhibits the effect of Ras17N on Src-mediated transformation. NIH/3T3 cells were transfected with 100 ng of SrcY527F expression construct in the presence or absence of Ras17N. Co-expression of the PI3KC2β-RBD reverses the inhibitory effect of Ras17N on Src transformation whereas the Raf-RBD does not. In contrast, expression of the Raf-RBD alone, but not the PI3KC2β-RBD, significantly inhibited Src-mediated transformation. The results represent the average relative focus forming activity ± S.E.M. from three independent experiments performed in triplicate. Asterisks denote samples that were significantly different from Src alone (*p<0.05).
Figure 9.
Model for ITSN1-Ras-PI3KC2β pathway.
Growth factor stimulation of receptor tyrosine kinases leads to the recruitment of the Grb2-Sos complex resulting in dissociation of GDP from Ras. We propose that PI3KC2β competes for binding this transient nucleotide-free (nf) Ras trapping it in the nucleotide-free state and preventing GTP loading. In addition, binding of nucleotide-free Ras to PI3KC2β inhibits its lipid kinase activity. The PI3KC2β-Ras complex may then translocate to distal sites such as early endosomes (EE) where ITSN1 then binds to PI3KC2β leading to the release of nucleotide-free Ras and activation of the lipid kinase activity of PI3KC2β. In addition, once released from the ITSN1-PI3KC2β complex, Ras binds GTP resulting in Ras activation and recruitment of effectors, e.g., Raf, Class I PI3Ks, etc. Although the above model describes a potential role for nf-Ras in ITSN1 and PI3KC2β function, we propose that nf-Ras may be regulated by additional proteins besides these molecules. Our findings also raise the possibility that the nucleotide-free forms of other GTPases may play a similar role in cell signaling.