Galectin-1 augments Ras activation and diverts Ras signals to Raf-1 at the expense of phosphoinositide 3-kinase

A, Raf-1 lysates. Mean values (±SD) of Raf-1 activity (32P incorporated into MBP) from quadruplicate experiments are shown. B, PI3-K activity in the cell lysates. The PIP spots of a representative experiment as detected on an X-ray film are shown. Similar results were obtained in 3 additional experiments. C, Galectin-1 co-immunoprecipitates with H-Ras(G12V) or with H-Ras(G12V/T35S) but not H-Ras(G12V/Y40C). The experiments were Fig 1A. Proteins were co-immunoprecipitated (IP) with anti-Ras Ab from lysates of galectin-1/ H-Ras (G12V), galectin-1/ H-Ras (G12V/T35S), or galectin-1/ H-Ras (G12V/Y40C) co-transfectants and immunoblotted (IB) with either anti-Ras Ab or anti galectin-1 (Gal-1) Ab. Immunoblots of a representative experiment are shown. Similar results were obtained in two additional experiments.

We hypothesized that interactions of active Ras with escort proteins could direct selectivity of Ras signals. In a recent study we found that galectin-1 is such an escort; it is associated with the constitutively active (GTP bound) H-Ras (G12V) mutant (13). Galectin-1 belongs to a family of β-galactoside-binding proteins which act both within cells via sugar-independent interactions with other proteins and outside of cells via sugar-dependent interactions with glycan chains of cellular glycoconjugates after their secretion which is independent of a classical signal sequence (14)(15)(16)(17)(18)(19). Our own findings suggested that active Ras is a binding partner of galectin-1 and that this partnership is required for the membrane anchorage of active H-Ras, but not its wild type counterpart (13). We proposed that active H-Ras is Cell culture and transfection procedures-We maintained cells in DMEM with 10% fetal calf serum (13). We transfected 1.5x10 6 Cos-7 cells with total amount of 1.5 µg DNA by dextran (Pharmacia) or 8 x 10 5 HEK 293 cells with total amount of 6 µg DNA by calcium phosphate (Sigma). The cells were harvested 48 h after transfections using lysis buffers containing protease and phosphatase inhibitors (13).
Lysates were normalized for protein and subjected to the specified assays and quantification as detailed bellow. Data were expressed as means± SD and differences between means were examined in Student's t Test. peroxidase-goat anti-mouse IgG (Jackson Laboratories) or 1:500 anti galectin-1 (Gal-1) Ab (23) and 1:5000 peroxidase-goat anti-rabbit IgG (Jackson Laboratories).

Co-immunoprecipitation and Western immunoblotting -We
We then visualized Ras proteins and the 14 kDa galectin-1 protein by ECL and quantified the bands by densitometry with Image Master VDS-CL (Amersham Pharmacia Biothech) using TINA 2.0 software (Ray Tests).  by Western immunoblotting with pan anti-Ras Ab, and lysates containing 500 µg protein for the determination of GFP-Ras-GTP by the glutathione S-transferase (GST)-RBD pull-down assay followed by Western immunoblotting with pan anti-Ras Ab as detailed previously (21). We determined ERK or phospho-ERK using 30 µg of lysates protein by Western immunoblotting, ECL and densitometry as detailed above. ERK immunoblots were incubated with 1: 2000 rabbit anti-ERK1/2 Ab (Santa-Cruz) then with 1:1000 peroxidase-goat anti rabbit IgG; phospho-ERK immunoblots were incubated with 1:10,000 mouse anti-phospho-ERK Ab (Sigma) then with 1:10,000 peroxidase-goat anti-mouse IgG.

Ras-GTP and phospho-ERK assays-
Raf-1 kinase assay-We determined Raf-kinase activity with a Raf-1 immunoprecipitaion kinase cascade assay kit (Upstate Biotechnology). HEK 293 cells were co-transfected with plasmids coding for K-Ras, H-Ras, H-Ras (G12V) or its effector loop mutants along with a plasmid coding for galectin-1 or with empty vector as detailed above. All combinations included 1.5 µg of plasmid DNA coding for Raf-1(wt). Twenty-four h after transfection the cells were serum starved for 24h then, either left un-stimulated (activated Ras transfectants), or stimulated with 100 ng/ml EGF for 5 min. The cells were then lysed in 0.5 ml of Raf-1 kinase assay kit lysis buffer. Raf-1 was immunoprecipitated from 300 µl lysate and assayed in the coupled kinase assay using myelin basic protein (MBP) from Sigma as a substrate and [γ-32 P] ATP (3000 Ci/mmol, Dupont, NEN) as a phosphate donor according to the manufacturer's instructions. The amount of 32 P -labeled MBP thus formed was estimated in a scintillation counter.
PI3-K assay-We determined PI3-K activity using a published procedure (25) in HEK 293 cells that were transfected and treated with EGF as detailed for the Raf-1 kinase assay (omitting the Raf-1 plasmid). We lysed the cells with 0.5 ml of PI3-K lysis buffer (25), immunoprecipitated the enzyme from 500 µl lysate with rabbit anti PI3-K p85 Ab (Upstate Biotechnology) and assayed its activity with 0.5 mg/ml phosphatidylinositol and 125µM ATP and 5 µCi [γ-32 P] ATP . We extracted the lipids with chloroform/methanol then separated them by TLC (25). Phopholipid markers were use for the identification of the 32 P-labeled phosphatidylinositol-3phosphate (PIP) product (25). The 32 P-labeled lipid products were visualized by an overnight exposure on an X-ray film and the spots then quantified by TINA analysis.  or the EGF response (Fig. 1D). Moreover, there was no effect of 100 mM lactose, which interferes with galectin-1 binding to glycan ligands, on the response to EGF stimulation in the co-transfectants (not shown).

Galectin-1 augments basal and EGF-induced increase in GFP-H-Ras-GTP and GFP-K-Ras-GTP and prolongs the EGF response-Thus
The above results, in keeping with the preferential co-immunoprecipitation of galectin-1 with Ras-GTP over Ras-GDP, suggest that galectin-1 binds preferentially to Ras-GTP, stabilizing it in a conformation that may also render Ras less sensitive to Ras-GTPase activating proteins (GAPs) (26). Accordingly, galectin-1 might have been expected to raise basal Ras-GTP levels by stabilizing the small percentage of Ras-GTP that exists under normal conditions; this was indeed the case (Fig. 1, B and C). In the presence of EGF, the level of Ras-GTP is higher, and more is available to bind to galectin. Thus, the fold increase in Ras-GTP in response to EGF was similar in the absence and in the presence of galectin-1--the values recorded in GFP-Ras transfectants and in GFP-Ras/galectin-1 co-transfectants were 2.1 and 2.5 respectively. This is consistent with the idea that when Ras-GTP molecules are formed, they "search" for galectin-1 partners. To build up in the membrane, they have to find such partners, and the more galectin there is, the more Ras-GTP can accumulate. We gained additional support for the above assertion by knocking down Thus, the galectin-1 enhancement of the EGF-stimulated increase in Ras-GTP (Fig.   1, B and C) promoted elevated Raf-1 activity ( Fig. 2A).
We then used GFP-H-Ras or GFP-K-Ras transfectants and GFP-Ras/galectin-1 co-transfectants to examine whether the enhancement of EGFinduced Raf-1 activity propagated to ERK. We found that galectin-1 enhanced the EGF-stimulated increase in the active phospho-ERK protein (Fig. 2, B and C).
Similar to its effects on the EGF-induced increase in GFP-Ras-GTP (Fig. 1, B and C), galectin-1 promoted a stronger and more prolonged activation of ERK and galectin-1-AS inhibited the EGF response (Fig. 2, B and C). We conclude that galectin-1-Ras interactions enhance Ras' activation of the Raf/MEK/ERK pathway.
This prompted the question whether such an effect of galectin-1 is confined to this pathway or might be operative in a different pathway.
Galectin-1 inhibits EGF-induced signal to PI3-K-We therefore next examined the effects of galectin-1 expression on a second well-established Ras effector, PI3-K type I, the catalytic subunit of which interacts directly with active Ras-GTP (4,5). We studied H-Ras or K-Ras transfected cells or cells co-transfected with galectin-1 and each of these two Ras isoforms. The cells were stimulated for 5 min with EGF and PI3-K was then immunoprecipitated from the cell extracts by anti PI3-K p85 Ab. The precipitates were subjected to a PI3-K assay. We found, as expected, that EGF stimulated PI3-K activity inducing, respectively, 1.6 ± 0.1 and 1.8 ± 0.3 (means ± SD, n=3) fold increases in PI3-K activity in H-Ras and in K-Ras transfectants (Fig. 3). However, co-transfection with galectin-1 strongly inhibited the EGF-mediated stimulation of PI3-K activity (Fig. 3). The degrees of inhibition of EGF stimulation by galectin-1 were respectively 84±12% and 63± 8% (means ± SD, In conclusion, our results provide evidence for a novel mechanism controlling the function of Ras. We have shown that galectin-1 augments and prolongs Ras activation at the same time that it shifts the interaction of the activated molecule away from PI3-K and towards Raf-1. These, and the documented preferential activation of Raf-1 by K-Ras and of PI3-K by H-Ras (31,32), suggest that the strength (27)