The elucidation of novel SH2 binding sites on PLD2

Abstract

Our laboratory has recently reported that the enzyme phospholipase D2 (PLD2) exists as a ternary complex with PTP1b and the growth factor receptor bound protein 2 (Grb2). Here, we establish the mechanistic underpinnings of the PLD2/Grb2 association. We have identified residues Y(169) and Y(179) in the PLD2 protein as being essential for the Grb2 interaction. We present evidence indicating that Y(169) and Y(179) are located within two consensus sites in PLD2 that mediate an SH2 interaction with Grb2. This was demonstrated with an SH2-deficient GSTGrb2 R86K mutant that failed to pull-down PLD2 in vitro. In order to elucidate the functions of the two neighboring tyrosines, we created a new class of deletion and point mutants in PLD2. Phenylalanine replacement of Y(169) (PLD2 Y169F) or Y(179) (PLD2 Y179F) reduced Grb2 binding while simultaneous mutation completely abolished it. The role of the two binding sites on PLD2 was found to be functionally nonequivalent: Y(169) serves to modulate the activity of the enzyme, whereas Y(179) regulates total tyrosine phosphorylation of the protein. Interestingly, binding of Grb2 to PLD2 occurs irrespectively of lipase activity, since Grb2 binds to catalytically inactive PLD2 mutants. Finally, PLD2 residues Y(169) and Y(179) are necessary for the recruitment of Sos, but only overexpression of the PLD2 Y179F mutant resulted in increased Ras activity, p44/42(Erk) phosphorylation and enhanced DNA synthesis. Since Y(169) remains able to modulate enzyme activity and is capable of binding to Grb2 in the PLD2 Y179F mutant, we propose that Y(169) is kept under negative regulation by Y(179). When this is released, Y(169) mediates cellular proliferation through the Ras/MAPK pathway.

Figure 1. Schematic non-scale representation of mycPLD2 and GSTGrb2 mutants

(A) The amino acid residues involved in the creation of mutation and deletion series of mycPLD2 constructs are indicated in red together with their relative position (bar at the top). Also indicated are the PX domains and HKD domains. Boxed in the indicated PX domain of mycPLD2 are the sequences of interest used for these studies. (B) The typical SH3-SH2-SH3 domain structure of the fusion protein GSTGrb2 WT is depicted as boxes located C-terminally of the GST tag used for purification purposes. The impaired SH2 or SH3 functionality of GSTGrb2 R86K or P49/206L, respectively, are indicated as crossed boxes in their respective SH domains.

Figure 2. PLD2 interacts with Grb2 in vitro

The indicated amounts (in μg) of precleared COS-7 lysates expressing mycPLD2 were incubated with: (A) 1.5 mg purified Grb2, (B) 5 mg R86K or (C) 5 mg GST protein alone. After pull-down, the glutathione-Sepharose beads were subjected to SDS-PAGE and derived Western blots were probed with anti-myc and anti-GST antibodies. (D) In vitro binding competition of mycPLD2 to purified Grb2. The pulled-down complex Sepharose/GSTGrb2/mycPLD2 was incubated in the presence of increasing amounts of purified Grb2 protein (0, 1, 3 and 5 μg) and analyzed for the presence of mycPLD2.

Figure 3. The PLD2 residue Y¹⁷⁹ is involved in Grb2 interaction

PLD2/Grb2 interaction in vivo. COS-7 lysates from cells transiently expressing mycPLD2 Y179F point mutant or deletion mutant DY¹⁷⁹ (A), or DY¹⁷⁹ and DFYR178 (B) or Y179F/N181D (C) were immunoprecipitated with anti-human Grb2 or anti-myc antibodies. Immunoprecipitates were separated by SDS-PAGE and immunoblotted with anti-myc antibodies. (D) Densitometric analysis expressed as mean ± SEM (n=3) of the optical density of the immunoprecipitates normalized to myc signal (Grb2/myc IPs). (E) Shown is an anti-myc immunoblot of whole COS-7 lysates over expressing mycPLD2 WT and all Y¹⁷⁹ mutants.

Figure 4. Conserved PLD2 residues Y¹⁶⁹ and Y¹⁷⁹ are required for full Grb2 binding

(A) Eukaryotic PLD homologies and conservation of Y¹⁶⁵, Y¹⁶⁹ and Y¹⁷⁹. Protein sequence alignment of the C-terminal of PLD PX domains are shown in gray and the Y residues subject to this study are depicted in red. Boxed sequences represent the putative SH2-Grb2 consensus binding sites (B) PLD2/Grb2 interaction in vivo. COS-7 lysates from cells transiently expressing mycPLD2 WT, Y169F, Y179F or Y169/179F were subjected to immunoprecipitation with anti-Grb2 or anti-myc antibodies. Immunoprecipitates were separated by SDS-PAGE, immunoblotted, and the mycPLD2 levels were determined with anti-myc antibody probing. Also shown is an anti-myc immunoblot of whole COS-7 lysates over expressing mycPLD2 WT, Y169F, Y179F and Y169/179F mutants. Also shown are anti-myc and anti-Grb2 immunoblots of whole COS-7 lysates.

Figure 5. Functional uncoupling of Y¹⁶⁹ and Y¹⁷⁹

(A) PLD activity of Control (non-transfected and transfected with empty vector (EV), WT or mutant plasmids. COS-7 lysates from cells transiently expressing mycPLD2 WT or mutants, were immunoprecipitated with anti-myc antibodies, and the immunoprecipitated proteins were analyzed for PLD activity in vitro. Results are expressed as % PLD activity (cpm/μg protein) over Control, as mean ± SEM (n=5, *p<0.001). (B) PLD2 WT and mutant expression levels determined by myc-immunoblot of whole COS-7 lysates (labeled as “W.B.” [not “I.P”]). (C) Role of residue Y¹⁶⁹ in Grb2-mediated PLD2 activity. Myc immunoprecipitates of COS-7 transfected with empty vector (Control) or with fully active pcDNA-mycPLD2 Y179F or DY¹⁷⁹, were incubated in the presence of purified GSTGrb2 WT (10 mg) or P49/206L (20 mg) and analyzed for PLD2 activity in vitro. Results are expressed as % PLD activity relative to Control treatment ± SEM (n=3, *p<0.001). (D) Tyrosine phosphorylation of PLD2 WT, Y169F, DY¹⁷⁹ or Y179F probed with either anti-phosphotyrosine or with myc antibodies. (E) Shown is the densitometric analysis (optical density in arbitrary units normalized to myc signal) as mean ± SEM (n=3).

Figure 6. PLD2 activity is not required for Grb2 binding

(A) In vitro PLD activity of HKD mutants. COS-7 cells were transfected with empty vector (EV), mycPLD2 WT or lipase inactive mutants K444R, K758R and K444/758R. After 48 hs, cellular lysates were myc-immunoprecipitated and analyzed for PLD activity. (B) Interaction of PLD2 inactive mutants with Grb2 in vivo. COS-7 lysates from cells transiently expressing mycPLD2 K444R, K758R or K444/758R were immunoprecipitated with either anti-myc or anti-Grb2 antibodies. Immunoprecipitates were separated by SDS-PAGE and immunoblotted with anti-myc. Transfected COS-7 whole cell lysates were also probed with anti-myc antibody (bottom panel).

Figure 7. PLD2 residues Y¹⁶⁹ and Y¹⁷⁹ recruit Sos

(A) PLD2 interaction with endogenous Sos. Cellular extracts of COS-7 cells transiently expressing WT or mutant plasmids were immunoprecipitated with an anti-myc antibody and the immunological presence of Sos (~170 kDa) analyzed by Western blot. As controls, basal expression of myc, Sos and Grb2 were analyzed simultaneously. (B) Densitometric analysis showing the reduction of in vivo Sos binding to mutant PLD2s. Results are expressed as mean ± SEM (n=3) of the optical density in arbitrary units of the myc-immunoprecipitates normalized to Sos signal [(IP myc, WB Sos)/WB Sos].

Figure 8. Analysis of the Ras/MAPK pathway and DNA synthesis

(A) Ras activation induced by COS-7 transfection of WT and mutant plasmids. Cell lysates were incubated in the presence of Raf-1-RBD coupled to agarose beads and active Ras was purified and detected as indicated in Materials and Methods. Shown is a representative experiment done with 3 separate transfections [with pooled samples (n=6)]. Shown also are immunoblots of transfected COS-7 lysates probed with anti-Erk, -phospho-Erk, or -PCNA antibodies. For comparison purposes also shown here are positive controls 24 h-starved COS-7 cells stimulated with 100 ng/ml EGF and its effect on Ras and Erk. (B) De novo DNA synthesis as [³H]-thymidine incorporation. COS-7 cells Control [(non-transfected and transfected with empty vector (EV)] or transiently expressing WT or mutant plasmids, were incubated with 1 μCi/ml for 16 hours. TCA-precipitable [³H]-DNA was counted and referred as “% over Control” (mean ± SEM, n=9, *p<0.001).