Knockdown of RasGRP1 affects induced Ras-ERK activation more profoundly than knockdown of SOS1. (A) siRNA duplexes for mouse (mo. [control]) or human (Hu.) SOS1 were introduced via transient transfection into wild-type Jurkat or mutant JPRM441 cells. The resultant populations were stimulated 40 h after transfection by cross-linking the stably expressed CD25ζζ and analyzed as in Fig. 1. The relative percentages of SOS1 and RasGRP1 expression as well as ERK phosphorylation were determined by normalizing for Ras expression. Both Jurkat and JPRM441 cells express similar levels of RasGRP3 and SOS2 (data not shown). α-CD25, anti-CD25. (B) Wild-type Jurkat cells were transfected with the indicated siRNA duplexes. Forty hours later, cells were stimulated by cross-linking the TCR and analyzed by Western blot analysis as in panel A, normalizing for Grb2 expression. α-TCR, anti-TCR. (C) Mouse Rasgrp1 (control), RasGRP1, and SOS1 siRNA duplexes were introduced into purified human CD4⁺ T cells using an Amaxa electroporator. Forty hours later, cells were stimulated by cross-linking the TCR and CD4. Expression of the indicated proteins, as well as phosphorylation of ERK kinases, was determined by Western blot analysis (the same samples were run on two separate gels, analyzing SOS1 and Grb2 on one gel and RasGRP1, P-ERK, and Grb2 on the other). The percentage of ERK activation and RasGRP1 and SOS1 expression levels were determined by normalizing for Grb2 expression. α-CD4, anti-CD4. (D) Expression levels of SOS1 and RasGRP1 were compared between human CD4 T cells and the Jurkat T-cell clone by combining the primary antibodies into one mixture. Grb2 normalizes for the amount of protein loaded.

Suboptimal BCR stimulation reveals SOS function. (A and B) Wild-type (wt) or DT40 chicken B-cell lines with the indicated targeted gene deletions were stimulated for the indicated intervals with BCR-stimulating M4 monoclonal antibody at either a maximal dose (A [1:1,000]) or a suboptimal dose (B [1:20,000]). NP-40 lysates were analyzed by Western blotting. The same portion of the blot was blotted for P-ERK and stripped, blotted for P-MEK1/2 and stripped, blotted for MEK1/2 and stripped, and blotted for ERK. Blotting for ERK after P-ERK typically gives inverted patterns compared to P-ERK. Note that DT40 B cells express only one ERK protein (chicken ERK2). A separate portion of the same blot was blotted for α-tubulin. The relative percentage of ERK phosphorylation was determined by normalizing for α-tubulin. α-BCR, anti-BCR. (C) The indicated DT40 lines were pretreated with DMSO (control) or rottlerin, stimulated with a 1:1,000 dilution of M4 monoclonal antibody as indicated, and analyzed as in panels A and B. (D) Wild-type or Sos1⁻ Sos2⁻ DT40 chicken B cells were pretreated with DMSO (control) or rottlerin and stimulated with control phosphate-buffered saline (0) or decreasing dilutions of BCR-stimulating M4 monoclonal antibody (1:1,000, 1:5,000, and 1:20,000) and analyzed as before. Of note, rottlerin did not lead to general cell toxicity as BCR (or TCR- or CD25ζζ-)-induced phosphorylation of MAP kinases other than ERK was less affected (data not shown). (E) The indicated DT40 lines were pretreated with DMSO (control), rottlerin, Go6976, or GF109203X and maximally stimulated with a 1:100 dilution of M4 monoclonal antibody (different M4 batch from those used in panels A to D) or with a reduced amount of 1:500 as indicated, and analyzed as before. (F) Model of Ras-RAF-MEK-ERK activation following antigen receptor engagement on lymphocytes. The pathway downstream of DAG dominates lymphocyte Ras activation. Our collective data suggest an unknown dependence of SOS function on the DAG-PKC-RasGRP pathway.

The isolated catalytic domain of SOS requires the allosteric pocket and RasGRP1 to efficiently trigger a Ras-ERK-CD69 pathway. (A) Ten micrograms of vector, SOS1-cat, SOS1-cat-F929A, or SOS1-cat-W729E expression constructs was transiently transfected into Jurkat or JPRM441 cells together with 10 μg of GFP plasmid. Sixteen hours later, cells were analyzed by FACS and Western blotting as described before. Numbers inside the plots indicate percentages of cell populations in each quadrant. (B) Wild-type Jurkat T cells were transiently cotransfected with 4 μg vector and 10 μg GFP plasmids or with 5 μg mouse Rasgrp1 or human RasGRP1 siRNA together with 4 μg SOS1-cat and 10 μg GFP plasmids. Transfected cells were analyzed for CD69 and GFP expression by FACS and for protein expression by Western blot analysis 40 h after transfection. The percentage of CD69-positive cells over all transfected cells was determined in three independent experiments and is plotted as mean values with error bars. Small amounts of SOS1-cat plasmid were introduced to minimize SOS1-cat expression before RasGRP1 knockdown occurs. (SOS1-cat starts to be expressed 4 h after transfection, but RasGRP1 knockdown is maximal only after 40 h.) (C) JPRM441 cells were transiently transfected with 4 μg vector and 10 μg GFP plasmids or with 5 μg mouse Rasgrp1 or human RasGRP1 siRNA together with 4 μg SOS1-cat and 10 μg GFP plasmids and analyzed as in panel B. As controls, wild-type Jurkat cells were transfected with 4 μg vector or SOS1-cat together with 5 μg vector and 10 μg GFP plasmids. Panel C is a representative example of two independent experiments. (D) Jurkat-CD25ζζ, JPRM441-CD25ζζ, or stably reconstituted JPRM441-CD25ζζ-wtRasGRP1 cells were transiently transfected with 30 μg of vector (vec) or SOS1-cat plasmids together with 15 μg of GFP expression construct and analyzed as in panel A at 4, 6, and 8 h posttransfection. The percentage of CD69-positive cells over all transfected cells was plotted for the various time points as a bar graph. Protein expression was determined as before using NP-40 lysates prepared at the indicated time points.

A RasGTP mimetic enhances in vivo GEF activity of SOS and can substitute for RasGRP1 deficiency. (A) The indicated H-RasV12 variants were tested in Jurkat and JPRM441 T cells. The effects of H-RasV12C40 on the Ras-ERK-CD69 pathway were tested by FACS analysis for CD69 expression 16 h after transient transfection (10 μg of RasV12 variants with 10 μg of GFP). (B) RasGRP1-deficient JPRM441 cells were transfected with the indicated combinations of H-RasV12C40 (10 μg) and/or SOS1-F (30 μg) together with GFP (10 μg) plasmids and analyzed for CD69 induction and protein expression 16 h later as described before. (C) Jurkat and JPRM441 T cells were transfected with 2 μg of SOS1-cat or SOS1-cat-W729E and/or 5 μg H-RasV12C40, together with 5 μg GFP plasmids and analyzed for induction of CD69 expression by FACS 16 h after transfection.

PKC-RasGRP1 signaling dominates induced Ras-ERK activation. (A) Jurkat and JPRM441 T-cell lines were stimulated by cross-linking CD25ζζ with anti-CD25 (α-CD25) monoclonal antibody for the indicated time intervals in the presence of rottlerin or the DMSO control (−). The relative percentages of Ras and ERK activation and RasGRP1 expression were determined by normalizing for total Ras expression. Jurkat T cells express K- and N-Ras, both migrating as doublets and recognized by the anti-Ras antibody. All Ras-GTP pull-down assays were run on a separate minigel, typically using 10-fold more input material compared to the material presented in the Western blot analysis of the NP-40 whole-cell lysates. The result presented in panel A is a representative example of three independent experiments. Almost all panels presented in this study are representative examples of at least three independent experiments: exceptions are experiments shown in Fig. 6C and 7B that yielded very similar results in two independent experiments (one of which is shown). (B) Wild-type Jurkat T cells were stimulated through the TCR in the presence of rottlerin, Go6976, GF109203X, Ro318220, or DMSO as a control (−). Phosphorylation of ERK kinases was measured as in panel A, normalizing for α-tubulin expression. α-TCR, anti-TCR.

Optimal activity of membrane-targeted SOS1 depends on RasGRP1 function and DAG levels. (A) Jurkat or JPRM441 cells were transiently transfected with 30 μg of the indicated expression plasmids together with 15 μg GFP expression construct. Sixteen hours later, cells were analyzed by FACS for GFP and CD69 expression and NP-40 lysates were prepared. Numbers inside the plots indicate percentages of cell populations in each quadrant. Note the diagonal pattern with SOS1-F in Jurkat cells. The percentage of CD69-positive cells in the transfected population was determined and is depicted (e.g., 9/(9 + 38) = 19%). Protein expression levels were determined as described before. (B) Jurkat cells were transiently transfected with 30 μg of SOS1-F plasmid or vector control, together with 30 μg FLAG-tagged DGKζ expression or vector control plasmids. All transfections included 15 μg GFP expression construct. Sixteen hours later, cells were analyzed as in panel A. SOS1-F expression was higher than in panel A (due to carrier effect of the other introduced plasmid) and induced a higher level of CD69.

Complex regulation of SOS activity. (A) Graphic representation of the published Ras-SOScat-Ras GTP crystal structure and proposed positive feedback loop based on this structure (26, 33). RasGDP binding to the GEF pocket results in displacement of the GDP group, leaving empty Ras bound to SOS. After association with GTP, RasGTP can bind to the other side of the SOS molecule, to the allosteric pocket. Loading of the allosteric pocket increases the in vitro GEF activity of a wild-type molecule but not that of a W729E mutant. (B) Twenty-five micrograms (each) of SOS1, SOS1-F, 5′-SOS1, and 5′-SOS1-F, as well as 10 μg of vector, SOS1-cat, SOS1-cat-F, H-RasV12, or RasGRP1 expression constructs (all in the same pEF6 vector backbone), was introduced together with 10 μg GFP plasmid. Sixteen hours later, populations were analyzed for GFP and CD69 expression (using anti-CD69APC antibody detectable in Fl4). The right panel gives a representation of the introduced SOS1 constructs as well as the resultant percentage of CD69-positive cells of all transfected, GFP-positive cells (determined as in Fig. 3A).

Basal DAG-RasGRP1 signaling is required for optimal SOS1-cat activity. (A) Analysis of basal Ras activity in the indicated unstimulated cell lines grown in 5% FCS. Relative percentages of RasGTP levels were determined by normalizing for total Ras levels by Western blot analysis. Note the K- and N-RasGTP doublets detectable on this high-percentage gel. Input for the RasGTP pulldown was 30-fold the material presented in the total Ras blot analyzing the NP-40 whole-cell lysates. (B) Jurkat and JPRM441 cells were transiently transfected with 5 μg of SOS1-cat plasmid or vector control plasmids, together with 30 μg FLAG-tagged DGKζ expression or vector control plasmids. All transfections included 15 μg GFP expression construct. Sixteen hours later, cells were analyzed as before. Panel B is a representative example of two independent experiments.

Synergy between H-RasV12C40 and SOS function requires an intact allosteric pocket in SOS and is present in both T and B lymphocytes. (A) Wild-type and RasGRP1 and -3 dually deficient DT40 cells were transfected with 5 μg of SOS1-cat, SOS1-cat-W729E, and/or 5 μg H-RasV12C40 expression constructs and analyzed for protein expression as described before. The anti-Ras antibody only weakly reacts with the endogenously expressed chicken Ras. (B) Raji B cells were transfected and analyzed as in panel A. Two micrograms of SOS1-cat or SOS1-cat-W729E and/or 6 μg of H-RasV12C40 expression constructs was used. (C) The left panel shows a proposed model of Ras signaling in lymphocytes leading to activation of the Ras-RAF-MEK-ERK pathway following antigen receptor engagement. The right panel shows the same positive feedback loop via SOS' allosteric pocket, which was recently demonstrated to enhance Ras-ERK signaling in vivo in EGF-stimulated HeLa cells (4).