Rac1 and Cdc42 are regulators of HRasV12-transformation and angiogenic factors in human fibroblasts

Abstract

Background: The activities of Rac1 and Cdc42 are essential for HRas-induced transformation of rodent fibroblasts. What is more, expression of constitutively activated mutants of Rac1 and/or Cdc42 is sufficient for their malignant transformation. The role for these two Rho GTPases in HRas-mediated transformation of human fibroblasts has not been studied. Here we evaluated the contribution of Rac1 and Cdc42 to maintaining HRas-induced transformation of human fibroblasts, and determined the ability of constitutively activated mutants of Rac1 or Cdc42 to induce malignant transformation of a human fibroblast cell strain.

Methods: Under the control of a tetracycline regulatable promoter, dominant negative mutants of Rac1 and Cdc42 were expressed in a human HRas-transformed, tumor derived fibroblast cell line. These cells were used to determine the roles of Rac1 and/or Cdc42 proteins in maintaining HRas-induced transformed phenotypes. Similarly, constitutively active mutants were expressed in a non-transformed human fibroblast cell strain to evaluate their potential to induce malignant transformation. Affymetrix GeneChip arrays were used for transcriptome analyses, and observed expression differences were subsequently validated using protein assays.

Results: Expression of dominant negative Rac1 and/or Cdc42 significantly altered transformed phenotypes of HRas malignantly transformed human fibroblasts. In contrast, expression of constitutively active mutants of Rac1 or Cdc42 was not sufficient to induce malignant transformation. Microarray analysis revealed that the expression of 29 genes was dependent on Rac1 and Cdc42, many of which are known to play a role in cancer. The dependence of two such genes, uPA and VEGF was further validated in both normoxic and hypoxic conditions.

Conclusion(s): The results presented here indicate that expression of both Rac1 and Cdc42 is necessary for maintaining several transformed phenotypes in oncogenic HRas transformed human cells, including their ability to form tumors in athymic mice. Our data also indicate that expression of either activated Rac1 or Cdc42 alone is not sufficient for malignant transformation of human fibroblasts, although each is required for specific transformed phenotypes. Furthermore, our study elucidates that the expression of several highly significant cancer related genes require the activities of Rac1 and/or Cdc42 which may also play a critical role in cellular transformation.

Figure 1

A role for Rac1 and Cdc42 in focus reconstruction. (a) Cells were grown in the presence or absence of tetracycline (1 μg/ml). PH3MT-tTak-C1 is the parental cell strain. PH3MT-VC-C2 is transfected with empty vector. Whole cell extracts were made and the Western blot was probed with both myc and FLAG antibodies. β-actin levels were used to indicate loading. (b) Cells expressing dominant negative mutants were plated on a lawn of non-transformed MSU-1.1 cells, and focus reconstruction was analyzed. Pictures are representative of replicate experiments.

Figure 2

Evidence that Rac1 activity is essential to maintain HRas^V12-induced tumor formation. Solid lines represent mice that were not administered tetracycline (Tet). Dotted lines represent mice injected with the same cell strain, but mice were given Tet in order to suppress dominant-negative expression. Tumors were measured weekly. When tumors reached a volume of approximately 0.5 cm³, mice were sacrificed and the tumors were removed for further analysis. The data are plotted using Kaplan-Meier analyses. (a) PH3MT-tTak-C1 (parent); N = 4 (- Tet), N = 4 (+ Tet) p > 0.1. (b) PH3MT-VC-C2 (vector control); N = 5 (- Tet), N = 6 (+ Tet) p > 0.1. (c) PH3MT-Rac1^N17-C1; N = 7 (- Tet), N = 9 (+ Tet), p < 0.001. (d) PH3MT-Rac1^N17-C2; N = 4 (- Tet), N = 6 (+ Tet), p > 0.1. (e & f) Western blots probed with myc (9E10) antibody to detect dominant-negative protein expression in two tumor-derived cell lines (Tumor 1 and Tumor 2) Both blots were probed with β-actin to verify loading.

Figure 3

Evidence that Rac1 and/or Cdc42 activity is essential to maintain HRas^V12-induced tumor formation. Solid lines represent mice that were not administered tetracycline (Tet). Dotted lines represent mice injected with the same cell strain, but mice were given Tet in order to suppress dominant-negative expression. Tumors were measured weekly. When tumors reached a volume of approximately 0.5 cm³, mice were sacrificed and the tumors were removed for study. The data are plotted using Kaplan-Meier analyses. (a) PH3MT-Cdc42^N17-C1; N = 8 (- Tet), N = 8 (+ Tet), p > 0.1. (b) PH3MT-Cdc42^N17-C2; N = 6 (- Tet), N = 6 (+ Tet), p > 0.1. (c & d) Western blots probed with FLAG antibody to detect dominant-negative protein expression in two tumor derived cell lines (Tumor 1 and Tumor 2). (e) PH3MT-Rac1^N17/Cdc42^N17; N = 9 (- Tet), N = 9 (+ Tet) p < 0.001. (f) Western blots probed with either myc (9E10) or FLAG antibody to detect dominant-negative protein expression in two tumor derived cell lines (Tumor 1 and Tumor 2). All blots were probed with β-actin to verify loading.

Figure 4

Transformed phenotypes elicited by expression of Rac1^V12 or Cdc42^V12 in human fibroblasts. (a) MSU-1.1 fibroblasts expressing either GFP alone (MSU-1.1-GFP-VC), or GFP-tagged constitutively-activated proteins. (b) The indicated cell strains were grown in medium with reduced serum (0.5% SCS). Growth curves were plotted. Doubling time was calculated based on an equation derived from a best-fit exponential curve when cells were in log-phase growth. Error bars represent the SD of triplicate experiments. (c) The indicated cell strains were plated in 0.33% agarose and grown for three weeks. Each picture represents one representative field from each cell line. This experiment was completed in triplicate, each with similar results.

Figure 5

Activated Rac1 and Cdc42 independently regulate uPA expression. The indicated cell strains were serum starved for 24 hours then stimulated with medium containing 10% SCS. (a) Grown in the absence of tetracycline, PH3MT cell strains expressing a vector control (PH3MT-VC-C2), Rac1^N17(PH3MT- Rac1^N17-C1), Cdc42^N17 (PH3MT-Cdc42^N17-C2) or both Rac1^N17 and Cdc42^N17 (PH3MT-Rac1^N17/Cdc42^N17) were tested for uPA expression levels using ELISA. Data is presented as percent of control. Error bars indicate the SD from triplicate experiments. * indicates significant difference, p < 0.05. ** indicates a significant difference compared to PH3MT-Rac1^N17-C1 and PH3MT-Cdc42^N17-C2, p < 0.05. (b) conditioned medium was collected from MSU-1.1 cells expressing GFP alone (MSU-1.1-GFP-VC), GFP-tagged Rac1^V12 (MSU-1.1-GFP-Rac1^V12) or GFP-tagged Cdc42^V12 (MSU-1.1-GFP-Cdc42^V12), and uPA expression was analyzed. Data is presented as fold-induction of uPA expression. Error bars represent the SD from triplicate experiments. * indicates significant difference, p < 0.05.

Figure 6

Both Rac1 and Cdc42 regulate VEGF expression in non-hypoxic and hypoxic conditions. All cell strains were serum starved for 24 hours prior to exposure to the indicated agent. Conditioned media were collected after 24 hr and ELISA analyses were completed. The data is presented as either percent of control, or fold induction as indicated in the figure. Error bars represent the SD from triplicate experiments. (a) PH3MT cell strains expressing a vector control (PH3MT-VC-C2), Rac1^N17 (PH3MT- Rac1^N17-C1), Cdc42^N17 (PH3MT-Cdc42^N17-C2), or both (PH3MT-Rac1^N17/Cdc42^N17), were stimulated with either medium containing 10% SCS, 100 μM DFO, 100 μM CoCl₂ or Hypoxia (1% O₂). (b) MSU-1.1 cells expressing GFP alone (MSU-1.1-GFP-VC-C2), or GFP-tagged Rac1^V12 (MSU-1.1-GFP-Rac1^V12), or GFP-tagged Cdc42^V12 (MSU-1.1-GFP-Cdc42^V12). Cells were stimulated with medium containing 10% SCS. * denotes a significant difference (p < 0.01).