Different metastasis promotive potency of small G-proteins RalA and RalB in in vivo hamster tumor model

Abstract

Background: Previously we have shown that oncogenic Ha-Ras stimulated in vivo metastasis through RalGEF-Ral signaling. RalA and RalB are highly homologous small G proteins belonging to Ras superfamily. They can be activated by Ras-RalGEF signaling pathway and influence cellular growth and survival, motility, vesicular transport and tumor progression in humans and in animal models. Here we first time compared the influence of RalA and RalB on tumorigenic, invasive and metastatic properties of RSV transformed hamster fibroblasts.

Methods: Retroviral vectors encoding activated forms or effector mutants of RalA or RalB proteins were introduced into the low metastatic HET-SR cell line. Tumor growth and spontaneous metastatic activity (SMA) were evaluated on immunocompetent hamsters after subcutaneous injection of cells. The biological properties of cells, including proliferation, clonogenicity, migration and invasion were determined using MTT, wound healing, colony formation and Boyden chamber assays respectively. Protein expression and phosphorylation was detected by Westen blot analysis. Extracellular proteinases activity was assessed by substrate-specific zymography.

Results: We have showed that although both Ral proteins stimulated SMA, RalB was more effective in metastasis stimulation in vivo as well as in potentiating of directed movement and invasion in vitro. Simultaneous expression of active RalA and RalB didn't give synergetic effect on metastasis formation. RalB activity decreased expression of Caveolin-1, while active RalA stimulated MMP-1 and uPA proteolytic activity, as well as CD24 expression. Both Ral proteins were capable of Cyclin D1 upregulation, JNK1 kinase activation, and stimulation of colony growth and motility. Among three main RalB effectors (RalBP1, exocyst complex and PLD1), PLD1 was essential for RalB-dependent metastasis stimulation.

Conclusions: Presented results are the first data on direct comparison of RalA and RalB impact as well as of RalA/RalB simultaneous expression influence on in vivo cell metastatic activity. We showed that RalB activation significantly more than RalA stimulates SMA. This property correlates with the ability of RalB to stimulate in vitro invasion and serum directed cell movement. We also found that RalB-PLD1 interaction is necessary for the acquisition of RalB-dependent high metastatic cell phenotype. These findings contribute to the identification of molecular mechanisms of metastasis and tumor progression.

Figure 1

Ral-dependent stimulation of spontaneous metastatic activity. (a) Expression of exogenous Ral proteins (RalA-G23V in pLXSN vector (HET-SR-RalA) and RalB-G23V in pBabe vector (HET-SR-RalB)) in HET-SR cell line was confirmed by Western blot analysis using anti RalA or anti-RalB antibodies followed by quantification. (b) Hematoxylin-eosin staining of control hamster lung and metastatic nodule formed by RalB-expressing cells (c.) RalB stimulates spontaneous metastatic activity (SMA) more than RalA. Indicated is number of lung metastases formed two months after subcutaneous injection of cells into adult immunocompetent hamsters; the average for 10 animals is shown with standard error (SE). (d, e) Simultaneous expression of RalA and RalB doesn't give additive effect on SMA. (d) Western blot analysis and quantification of RalA/B expression in HET-SR-RalA-RalB cells obtained by RalB transduction into HET-SR-RalA cells. (e) Graphs represent influence of RalA- and RalA-RalB combined expression on SMA of HET-SR cells.

Figure 2

Effect of active Ral proteins on HET-SR cells growth characteristics. (a) Ral proteins expression has no significant influence on HET-SR cells proliferation.; OD₅₉₅ absorption in MTT test is shown as average of three independent experiments ± SE. (b) RalB is more active than RalA in stimulation of colony formation in clonogenicity assay. Colonies were measured 6 days after seeding at low cell density (200 cells per 6-cm dish); colony size is shown in arbitrary units as average for three independent experiments ± SE. (c) Neither RalA nor RalB modulate tumorigenicity of HET-SR cells. Indicated is the size of tumors two months after injection as average for 10 animals ± SE.

Figure 3

Migrative and invasive properties of Ral-expressing HET-SR cells. (a,b) RalB is more active in stimulation of in vitro invasion. Invaded cells were stained with crystal violet 18 hours after seeding on Matrigel-coated Boyden chambers; (a) Representative pictures of three independent experiments are shown. (b) Graphs correspond to the number of invaded cells shown as average for three independent experiments ± SE. (c) Gelatin zymography was used to study the activity of matrix metalloproteases (MMPs) in conditioned media. RalA increases MMP-1 proenzyme level (proMMP-1) and drastically stimulates MMP-1 activity, while RalB has no effect on studied gelatinases. (d) RalA opposite to RalB increases activity of urokinase-like plasminogen activator (uPA). uPA activity in conditioned media was revealed by casein-plasminogen zymography. (e,f) Both Ral proteins stimulate cell motility in "wound healing" assay. (e) Representative pictures of wounds at 0 and 24 hours after scratching. (f) Graphs correspond to migration indexes shown as average for three independent experiments ± SE. (g,h) RalB is more potent in stimulation of growth factors-directed migration in transwell assay. (g) Migrated cells were stained with hematoxylin-eosin 18 hours after seeding in uncoated chambers. (h) Graphs correspond to the number of transwell migrated cells shown as average for three independent experiments ± SE.

Figure 4

Influence of RalA and RalB activation on CD24, Cyclin D1, Cav-1 and MAPKs. (a) RalA upregulates expression of CD24. (b) Both Ral proteins stimulate Cyclin D1 expression. (c) RalB downregulates Caveolin-1 expression. (d, e) Neither of Ral proteins modulate ERK and p38 kinases phosphorylation. (f) JNK1 phosphorylation is equally stimulated in RalA- and RalB-expressing HET-SR cells.

Figure 5

Among three main RalB effectors, PLD1 is essential for HET-SR-RalB high metastatic phenotype. (a) Western blot analysis was used to confirm expression of exogenous RalB G23V effector mutants: ΔN11, D49N and D49E (blocking interaction with PLD1, RalBP1 and exocyst complex respectively). (b) SMA test of HET-SR cells expressing active RalB mutants. Activated RalB as well as its effector domain mutants D49N and D49E give statistically significant increase of SMA compared to the control (empty vector expressing cells) (p < 0,05), while deletion of N-terminus leads to abrogation of activated RalB-dependent SMA stimulation.