Cdc42 and Rac1 activity is reduced in human pheochromocytoma and correlates with FARP1 and ARHGEF1 expression

in Endocrine-Related Cancer
View More View Less
  • 1 Institut des Neurosciences Cellulaires et Intégratives (INCI), CNRS UPR 3212, Strasbourg, France
  • 2 Caprion Proteome, Inc., Montréal, Québec, Canada
  • 3 Service de Chirurgie Digestive, Hépato-bilaire et Endocrinienne, CHRU Nancy, Hôpitaux de Brabois, Vandoeuvre les Nancy, France
  • 4 Centre de Ressources Biologiques (CRB), CHRU Nancy, Hôpitaux de Brabois, Vandoeuvres les Nancy, France

Abstract

Among small GTPases from the Rho family, Cdc42, Rac, and Rho are well known to mediate a large variety of cellular processes linked with cancer biology through their ability to cycle between an inactive (GDP-bound) and an active (GTP-bound) state. Guanine nucleotide exchange factors (GEFs) stimulate the exchange of GDP for GTP to generate the activated form, whereas the GTPase-activating proteins (GAPs) catalyze GTP hydrolysis, leading to the inactivated form. Modulation of Rho GTPase activity following altered expression of Rho-GEFs and/or Rho-GAPs has already been reported in various human tumors. However, nothing is known about the Rho GTPase activity or the expression of their regulators in human pheochromocytomas, a neuroendocrine tumor (NET) arising from chromaffin cells of the adrenal medulla. In this study, we demonstrate, through an ELISA-based activity assay, that Rac1 and Cdc42 activities decrease in human pheochromocytomas (PCCs) compared with the matched adjacent non-tumor tissue. Furthermore, through quantitative mass spectrometry (MS) approaches, we show that the expression of two Rho-GEF proteins, namely ARHGEF1 and FARP1, is significantly reduced in tumors compared with matched non-tumor tissue, whereas ARHGAP36 expression is increased. Moreover, siRNA-based knockdown of ARHGEF1 and FARP1 in PC12 cells leads to a significant inhibition of Rac1 and Cdc42 activities, respectively. Finally, a principal component analysis (PCA) of our dataset was able to discriminate PCC from non-tumor tissue and indicates a close correlation between Cdc42/Rac1 activity and FARP1/ARHGEF1 expression. Altogether, our findings reveal for the first time the importance of modulation of Rho GTPase activities and expression of their regulators in human PCCs.

Abstract

Among small GTPases from the Rho family, Cdc42, Rac, and Rho are well known to mediate a large variety of cellular processes linked with cancer biology through their ability to cycle between an inactive (GDP-bound) and an active (GTP-bound) state. Guanine nucleotide exchange factors (GEFs) stimulate the exchange of GDP for GTP to generate the activated form, whereas the GTPase-activating proteins (GAPs) catalyze GTP hydrolysis, leading to the inactivated form. Modulation of Rho GTPase activity following altered expression of Rho-GEFs and/or Rho-GAPs has already been reported in various human tumors. However, nothing is known about the Rho GTPase activity or the expression of their regulators in human pheochromocytomas, a neuroendocrine tumor (NET) arising from chromaffin cells of the adrenal medulla. In this study, we demonstrate, through an ELISA-based activity assay, that Rac1 and Cdc42 activities decrease in human pheochromocytomas (PCCs) compared with the matched adjacent non-tumor tissue. Furthermore, through quantitative mass spectrometry (MS) approaches, we show that the expression of two Rho-GEF proteins, namely ARHGEF1 and FARP1, is significantly reduced in tumors compared with matched non-tumor tissue, whereas ARHGAP36 expression is increased. Moreover, siRNA-based knockdown of ARHGEF1 and FARP1 in PC12 cells leads to a significant inhibition of Rac1 and Cdc42 activities, respectively. Finally, a principal component analysis (PCA) of our dataset was able to discriminate PCC from non-tumor tissue and indicates a close correlation between Cdc42/Rac1 activity and FARP1/ARHGEF1 expression. Altogether, our findings reveal for the first time the importance of modulation of Rho GTPase activities and expression of their regulators in human PCCs.

Introduction

The monomeric Rho (Ras homologous) GTPase family, a subclass of the Ras GTPase superfamily, contains 20 highly conserved members including the well-investigated RhoA, Rac1, and Cdc42. Most of the Rho GTPases undergo a tightly regulated activation/inactivation cycle by switching from an inactive GDP-bound to an active GTP-bound conformation upon signaling cues. Their activation is under the control of guanine nucleotide exchange factors (GEFs) that stimulate the exchange of GDP for GTP, and their inactivation requires GTPase-activating proteins (GAPs) that catalyze GTP hydrolysis by the Rho GTPases (Cherfils & Zeghouf 2013).

Rho GTPases regulate many cellular processes related to cancer biology, including cell cycle regulation, cell polarity establishment, cell migration, apoptosis and angiogenesis (Ellenbroek & Collard 2007, Vega & Ridley 2008). Consequently, Rho GTPases have been largely implicated in many types of cancer, and numerous studies have shown that aberrant expression of Rho GTPases is associated with tumor growth, invasion, and metastasis. The evidence for the role of Rho GTPases in various stages of tumorigenesis has been compiled in comprehensive reviews (Ellenbroek & Collard 2007, Grise et al. 2009, Karlsson et al. 2009, Parri & Chiarugi 2010, Alan & Lundquist 2013, Orgaz et al. 2014). However, the expression level of Rho GTPase genes or proteins does not necessarily reflect their activation state. Even so, the activation state of Rho GTPases in tumors has been very rarely investigated. Modulation of Rho-GTPase activity can occur through altered expression of Rho-GEFs or Rho-GAPs, a phenomenon that has also been observed in various tumors (Ellenbroek & Collard 2007, Vigil et al. 2010).

Expression of Rho GTPases and their regulators has been explored in many types of tumors with the exception of human NETs, a heterogeneous group arising from hormone/peptide-producing cells dispersed in the neuroendocrine system, including mainly the gastrointestinal and bronchopulmonary systems, and the adrenal, thyroid, and thymus glands (Taal & Visser 2004, Yao et al. 2008). Among NETs, PCCs arise from chromaffin cells of the adrenal medulla and are characterized by an excess of catecholamine release, leading mainly to hypertension, cardiomyopathy and severe stroke risk (Tsirlin et al. 2014). During the last decade, PCCs have been widely studied and have emerged as a relevant model for investigation of NETs. So far, the potential importance of Rho GTPases has not been investigated in human PCC.

The aim of this study was to analyze the activity of RhoA, Rac1, and Cdc42 as well as the expression of their putative regulators (Rho-GEFs and -GAPs) in human resections of PCC. Using an ELISA-based activity assay, we found that Rac1 and Cdc42 activities decrease in tumors compared with the adjacent non-tumor adrenal tissue from the same patient, whereas RhoA activity remains unchanged. Moreover, through a quantitative screening proteomic approach we have measured the expression of several Rho-GEFs and -GAPs in human PCC, and found that the expression of two GEFs, ARHGEF1, and FARP1, is significantly reduced in tumors compared with matched non-tumor tissue. Finally, using PCC cell models, we demonstrate that knocking down ARHGEF1 and FARP1 by siRNA significantly reduces Rac1 and Cdc42 activities.

Materials and methods

Subjects and samples

The study involved an analysis of data from 41 patients who underwent surgery for benign PCC originating from adrenal medulla between 2000 and 2009. All included patients had no clinical suspicion for malignant PCC (metastasis, locoregional involvement) or familial disease (negative germ-line mutation when performed, no personal or family history of familial disease) at operation. Furthermore, postoperative follow-up has not shown any tumor recurrence, metastasis, or late syndromic features in all included patients. Surgery was performed by Pr Laurent Brunaud at Nancy-Brabois University Hospital (Vandoeuvre les Nancy, France). After resection, the adrenal gland was cut longitudinally into two parts, and a piece of the tumor zone was dissected. Regarding the non-tumor adjacent control tissue, a piece of adrenal medulla was carefully dissected outside the tumor zone under a binocular magnifier (×4) and removed from the gland. This biological collection is governed by French legislation and regulations (registration numbers to the Ministry of Research DC-2014-2114 and AC-2014-2112, and CNIL data protection legislation 1209171). Informed consent was obtained from all participants, and the study was approved by the local ethic committees (October 10, 2008 by CPP Est-III).

Non-PCC tissues used in this study were drawn from patients who underwent surgery for various types of epithelial cancers (lung, colon, bladder, prostate, breast, and kidney) at two different institutions (McGill University Health Center and Centre Hospitalier de l'Université de Montréal; Montreal, Quebec, Canada). Informed consent was obtained from all patients as part of an institutionally approved protocol and after approval from the institutional review boards of the two participating institutions.

Once dissected, PCC and epithelial cancer tumors as well as their respective matching adjacent non-tumor tissues were placed in physiologic media immediately after excision in the operating room. Samples were then directly frozen and stored at –80°.

Cell culture and transfection

Rattus norvegicus PC12 cell line (ATCC CRL-1721) is a noradrenergic single clonal line established from a transplantable rat adrenal PCC (for more details see Greene & Tischler 1976). PC12 cells were cultured as described previously (Gasman et al. 2004). Cells were seeded at 5 × 104/cm2 24 h before siRNA transfection. Lipofectamine RNAiMAX (Invitrogen) and 80 nM pool of four nontargeting siRNAs as a control or siRNA directed against FARP1 or ARHEF1 (ON-TARGETplus SMARTpool siRNA; Dharmacon) were used to transfect cells according to manufacturer’s instructions. Cells were cultured for 48 h before experiments, and FARP1 or ARHGEF1 silencing was estimated by real-time quantitative PCR (RT-qPCR).

Rho GTPase activity assays

GTP-bound Rac1, Cdc42, and RhoA were measured using the respective G-LISA Activation Assay Kit (Cytoskeleton Denver, CO, USA). For human samples, the protocol was adapted to compare Rho GTPase activities between samples. Small pieces of frozen human tumor or matched adjacent non-tumor tissue were weighed and lysed for 15 min at 4°C into Cytoskeleton G-LISA lysis buffer using 40 µL lysis buffer for 1 mg tissue. Lysates were centrifuged at 20,000 g for 1 min at 4°C, and the supernatant aliquoted, snap-frozen into liquid nitrogen and stored at –80°C. Protein amounts were determined using the Precision Red reagent (Cytoskeleton), and immediately before the assay protein concentration of the samples was adjusted to 1 mg/mL and split into four aliquots to estimate the total amounts of Rho proteins (Rhotot), the maximal activity expected (Rhomax), the Rho GTPase activity in the sample (Rhosample), and the background (Rhobkg). Rhotot, Rhomax, and Rhobkg were estimated by loading Rho GTPases in vitro with different guanine nucleotides, that is, nonhydrolysable GTP (GTPγS, Rhotot), GTP (Rhomax) or GDP (Rhobkg). Loadings were performed by incubating cleared lysates with 100 µM guanine nucleotides and 10 mM EDTA for 15 min at 30°C. GTPases were locked into the expected state by adding MgCl2 at a final concentration of 60 mM and stored at 4°C before processing for G-LISA. Activities of each condition were measured in triplicates and compared with 1 ng of purified and activated Rho GTPases according to the manufacturer’s instruction. ELISA plates were loaded with 25 µg (Rhosample and Rhobkg) or 12.5 µg proteins (Rhotot and Rhomax). Percentage of Rho GTPase activity for each sample was calculated as ((Rhosample – Rhobkg)/(Rhomax–Rhobkg)) × 100. The GTP hydrolysis rate (Rhomax/Rhotot) between normal and tumor samples was unchanged.

For Rho GTPase activity measurement in PC12 cells, 48 h after siRNA transfection, cells were washed twice in Locke’s solution and lysed in Cytoskeleton G-LISA lysis buffer for 15 min at 4°C and lysates centrifuged at 10,000g for 1 min at 4°C. Supernatants were aliquoted, snap-frozen in liquid nitrogen and stored at −80°C. Protein concentration was determined and adjusted to 1 mg/mL in lysis buffer before the assay was realized according to the manufacturer’s instructions.

Tissue fractionation

Frozen tumor and matched adjacent non-tumor tissues were cut into small pieces (∼10 mm3), and 3 mL homogenization buffer (0.25 M sucrose/10 mM Tris pH 7.4/100 units/mL DNase 1/5 mM MgCl2, complete protease inhibitor EDTA-free cocktail) was added per tissue sample and homogenized twice for 10 s and once for 20 s using a Polytron set at speed 4.0 (non-cancer) or thrice for 20 s using a Polytron set at speed 5.5 (other tissues) (PT10-35 Polytron; Kinematica, Lucerne, Switzerland). Homogenates were filtered through a 180 μm nylon and brought to 3 mL with the homogenization buffer, if necessary. Light membranes were obtained by isopycnic centrifugation using discontinuous sucrose gradients in which samples brought to 1.4 M sucrose were layered by 1.2 and 0.8 M sucrose. After centrifugation at 155,000g for 2 h at 4°C, the light membrane fraction located at the 0.8–1.2 M sucrose interface was collected, snap-frozen in liquid nitrogen and stored at −80°C. The light membrane fraction is enriched with plasma membrane, Golgi apparatus, endosomes and secretory pathway-associated membranes. The cytosol fractions were obtained by centrifuging 200 µL crude homogenates at 150,000g for 1 h at 4°C. The supernatant was collected, snap-frozen and stored at −80°C.

The amounts of protein were determined using the bicinchoninic acid assay according to the manufacturer’s instructions (Thermo Fisher Scientific).

Mass spectrometry analysis

About 30 µg samples (homogenates, cytosol, and light membranes) were incubated in a denaturing buffer at a final concentration of 7 M urea/175mM NH4HCO3/8.75% v/v acetonitrile and incubated for 30 min at room temperature. Samples were then diluted to 1 M urea with water and digested with trypsin (Promega) overnight at 37°C at a ratio of 1 µg trypsin per 10 µg protein for homogenate and cytosol samples, while for light membranes the ratio was set at 1 µg trypsin per 25 µg proteins. Samples were reduced with 10 mM tris(2-carboxyethyl)phosphine (final concentration), incubated for 30 min at room temperature and then acidified to 0.5M HCl. Finally, the samples were desalted using C18 96-well plates (3M). The C18 eluates from homogenate samples were evaporated and stored at 4°C before MS analysis. The C18 eluates from light membrane and cytosol samples were collected in injection plates for strong cation exchange (SCX), dried by vacuum evaporation and stored at −20°C.

To fractionate peptides by SCX chromatography, samples were solubilized with reconstitution buffer (0.2% v/v formic acid, 10% v/v acetonitrile for light membrane samples; 20 mM K2HPO4, 25% v/v acetonitrile for cytosol samples) and loaded on an SCX column. Three fractions were collected following elution using a salt gradient. At the end of each SCX fractionation batch, the collected fractions were stored at −80°C. Once the SCX fractionation was completed, the fractions were freeze-dried and then desalted. The eluates were divided equally into two 96-well plates: one plate for liquid chromatography (LC)-MS/MS analysis and the other plate as a back-up. All plates were vacuum evaporated and stored at −20°C until analysis by LC-MS/MS. Samples were resuspended in 92.5/7.5 water/ACN + 0.2% formic acid and analyzed by LC-MS/MS on a nanoACQUITY UPLC (Waters, Milford, MA, USA) coupled to a Q Exactive mass spectrometer (Thermo Fisher Scientific). Survey (LC-MS) and tandem mass spectrometry scans (MS/MS) were acquired in the same run. The resolutions for the MS and MS/MS scans were 70,000 and 17,500 respectively. Peptide separation was achieved using a Waters nanoACQUITY UPLC Symmetry Trap column (180 µm × 20 mm, 5 µm particle size) and a Waters nanoACQUITY UPLC BEH300 analytical column (150 µm × 100 mm, 1.7 µm particle size). The mobile phases were (A) 0.2% formic acid in water and (B) 0.2% formic acid in acetonitrile. For each sample approximately 3.6 µg was loaded onto the trap column for 3 min at a flow rate of 10 µL/min. Peptides were separated using a linear gradient (92.5% A to 84% A) for 26 min, followed by (84% A to 75% A) for 14 min and a wash at 60% B for 2 min. The flow rate was 1.8 µL/min. Protein identification was accomplished using data acquired by LC-MS/MS. The MS/MS spectra were matched to the corresponding peptide sequences found in the UniProt human protein database using Mascot (version 2.2.06; Matrix Science, Boston, MA, USA) software.

Multiplexed multiple reaction monitoring assay

For each of the three proteins (ARHGAP36, ARHGEF1 and FARP1), five MRM-suitable peptides were selected by Caprion’s in-house MRM peptide selection software (Montreal, Quebec, Canada). If possible, peptides that were detected by MS were prioritized. The selected peptides were synthesized by JPT Peptide Technologies (Berlin, Germany). Synthesized peptides were resolubilized in 25%/75% water/DMSO (v/v), pooled and diluted with 0.2% formic acid in water to a concentration of 200 pmol/mL. This peptide mix was used to develop the MRM assay. The optimal two transitions (combination of peptide precursor and fragment ion mass-to-charge ratio that are monitored by the mass spectrometer) per peptide were determined using selected reaction monitoring (SRM)-triggered MS/MS on a QTRAP 5500 instrument (AB Sciex, Concord, Ontario, Canada) coupled to a nanoACQUITY UPLC (Waters). An SRM transition was predicted for each peptide. The detection of this transition triggered the acquisition of a full MS/MS spectrum of the target peptide. The two most intense fragment ions (b or y fragment ions only) in the MS/MS spectrum for each acquired peptide were recorded by in-house developed software. The mass spectrometer collision energy (CE) was optimized for each transition with five different CE values automatically generated by the in-house developed software. A solution containing all synthesized peptides at a concentration of 200 pmoL/mL was analyzed with the created MRM method. The two best peptides per proteins were selected to be monitored by the MRM assay.

The processed samples were resolubilized with 11 µL reconstitution solution containing five internal standard peptides each at 100 ng/mL. About 8 µL material (~10 µg) was analyzed by LC/MRM-MS. Peptide separation was achieved using a BioBasic C18 column (Thermo) (320 µm × 150mm, 5 µm particle size). The mobile phases were (A) 0.2% formic acid in water and (B) 0.2% formic acid in acetonitrile. Peptides were separated using a linear gradient (92.5% A to 60% A) for 21 min, followed by a wash at 60% B for 2 min. The flow rate was 10 µL/min. The transition peak areas were integrated using Elucidator software (Rosetta Biosoftware, Seattle, WA, USA) in combination with the software developed at Caprion for automated MRM peak integration.

Real-time quantitative PCR

Total RNA from human samples or PC12 cells were prepared using the GenElute Mammalian Total RNA Miniprep Kit (Sigma) and then treated with RNase-free DNaseI (Thermo Scientific). After checking RNA integrity and concentration by spectrophotometry and agarose gel electrophoresis, template RNA was transcribed into cDNA using the Maxima First Strand cDNA Synthesis Kit for RT-qPCR (Thermo Scientific), according to the manufacturer’s instructions (1 µg RNA/20 µL RT reaction). PCR was performed in 96-well plates using diluted cDNA samples, highly gene-specific primers (Supplementary Table 2, see section on supplementary data given at the end of this article) and SYBR Green PCR reagents (iQ SYBR Green Supermix; Bio-Rad).

Gene amplification and expression analyses were performed on a MyiQ real-time PCR machine (Bio-Rad) using a three-step procedure (20 s at 95°C, 20 s at 62°C, 20 s at 72°C) followed by a melting curve study to ensure specificity of the amplification process. PCR efficiency was evaluated by the standard curve analysis and the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as internal control. Gene expression in two different samples was compared using the comparative threshold cycle (Ct) method (Livak & Schmittgen 2001). Each reaction was performed in triplicate, and GAPDH was used as an internal control for gene expression. The mean ΔCt (Ct gene of interest – Ct GAPDH) was calculated for each condition, and expression levels were determined as 2–ΔCt. Gene expression levels from matched non-tumor samples were normalized to tumor tissue. Sequences of the primers are detailed in Supplementary Table 2.

Statistical analysis

For the differential expression analysis by MS, the intensity values for all detected components were log (base e) transformed with values <0 replaced by 0. Intensity data was normalized to account for small differences in protein concentration between samples. A subset of the samples was used to create a reference sample against which all samples were then normalized. The normalization factors were chosen so that the median of log ratios between each sample and the reference sample over all the peptides was adjusted to zero. Intensities below limit of detection (LOD = 100,000) after normalization were then linearly mapped to the range of (LOD/2, LOD) to avoid spurious large fold changes. Intensities above LOD were not changed. A two-way ANOVA model was used for the peptide-level analysis and is defined as follows: Iijk = M + Ci + Sj + εijk, where I is the peptide intensity, M the overall average intensity, C the ‘clinical group’ factor (matched non-tumor and tumor), S the ‘patient’ factor that takes into consideration the ‘pairing’ nature of the data, and ε is the random error. False detection rate and Q-value are calculated, based on the P values obtained from the ANOVA model, using Storey’s method to make multiple testing adjustments. Tukey’s HSD method is used to perform post hoc contrast among different groups.

One protein may have several identified and quantified peptides. The following ANOVA model, which is an extension of the two-way ANOVA used earlier in the peptide-level analysis, takes this into consideration by introducing a peptide factor in the model: Iijkl = M + Ci + Sj + εijkl, where I is the protein intensity, M an overall constant, C the clinical group, S the patient factor, and P the peptide factor. The number of levels for P is protein dependent, which is equal to the number of identified and quantified peptides for the protein.

For MRM analysis, differential intensity ratios were calculated in pairwise comparisons for each transition as the median of the ratio of the normalized intensities of each group. Paired Student’s t-test was applied for the expression analysis. Protein-level statistics were also computed by linearly combining the transitions of a given protein into a single variable and then applying a t-test.

For G-LISA assay, a Wilcoxon signed-rank nonparametric analysis was performed for matched samples. Eighteen pairs of human PCCs and matched non-tumor adjacent tissue were compared.

PCA was performed using R software implemented with the ‘Rcmdr’ package and ‘FactoMineR’ plug-in. Data were mean centered and normalized. Dimensions with eigenvalues >1 were considered for further analysis. The two dimensions (PC1 and PC2) accounted for 68% of the sample variance.

Results

Protein expression and activity of RhoA, Rac1 and Cdc42 in human pheochromocytomas

The expression level and the activity of Rho GTPases have never been explored in human PCC. To do so, we have used the ELISA-based assay (G-LISA; Cytoskeleton) to calculate the activity and the total amount of RhoA, Rac1, and Cdc42 in human PCC resections from 18 patients (see Materials and methods section). To avoid possible interindividual variations, each tumor tissue has been directly compared with its matched (from the same patient) adjacent non-tumor tissue. The global expression level of RhoA was slightly higher in PCC compared with adjacent non-tumor tissue (median values 2.40 vs 1.74), whereas the expression of Rac1 and Cdc42 remained statistically unchanged (Fig. 1A). However, we observed that, compared with non-tumor tissue, the relative activity of both Rac1 and Cdc42 in PCC is significantly decreased (median values 4.86 vs 10.63 for Rac1; 7.94 vs 16.34 for Cdc42), whereas the RhoA activity is unaffected (Fig. 1B). Fig. 1C and D illustrate the variation of Rho expression and the activity for each patient. Among the 18 patients, 15 exhibit a decrease of the Rac1 activity up to 5.3–fold and 16 exhibit a decrease of the Cdc42 activity up to 6.4-fold. Note that 13 patients (72%) exhibit a concomitant decrease of Rac1 and Cdc42 activities. Altogether, these results show a significant decrease of Rac1 and Cdc42 activities in human PCC, which is not correlated with a variation in their expression level.

Figure 1
Figure 1

The activities of Rac1 and Cdc42 are reduced in human PCCs compared with adjacent non-tumor tissue. (A) Colorimetric ELISA-based assay was used to estimate Rho GTPase activity in lysates from human PCCs and matched adjacent non-tumor tissue. Total amounts of Rho GTPases were estimated by loading proteins in vitro with nonhydrolysable GTP (GTPγS) and absorbance of the sample normalized to the absorbance of 1 ng of recombinant and activated Rho GTPases. Box-and-whisker plot represents the first quartile (bottom line), the median (line in the box) and the third quartile (upper line). Whiskers correspond to the 5th (bottom) and 95th (top) percentiles, and the black dots represent outlier observations. (B) Rho GTPase activity levels were determined by measuring the amounts of GTP-bound GTPases in the sample lysate, normalized to the maximal activity that can be detected. Maximal activity was obtained by loading proteins in vitro with GTP. Statistical significance for medians was determined by a Wilcoxon signed-rank nonparametric analysis for matched samples, n = 18 pairs of human PCCs, and matched adjacent non-tumor tissue. *P<0.05; **P<0.01; ***P <0.001; n.s., not significant. (C, D) Representation of total amounts and activity of Rho GTPases for each pair of tissue tested per patient. A full colour version of this figure is available at http://dx.doi.org/10.1530/ERC-15-0502.

Citation: Endocrine-Related Cancer 23, 4; 10.1530/ERC-15-0502

Differential expression of Rho-GEFs and Rho-GAPs in human pheochromocytomas

A change in Rho GTPase activity could result from a differential expression of their regulators, Rho-GEFs and Rho-GAPs. To test this hypothesis, we performed quantitative mass spectrometry analysis of six additional pairs of human PCC and matched adjacent non-tumor tissue. Most Rho-GEFs and Rho-GAPs usually switch from the cytosol to membranous compartments to meet with and regulate their respective Rho GTPases. Therefore, to increase detection and sensitivity, we conducted the proteomic analysis on purified subcellular fractions in place of total homogenate. Two main subcellular fractions were isolated from each tissue sample, a membrane-enriched fraction and a fraction enriched in cytosolic proteins (see Materials and methods section).

In total, 13 Rho regulators have been detected, including 8 Rho-GEFs and 5 Rho-GAPs. Table 1 illustrates the median value of the relative expression changes (tumor vs non-tumor tissue) of the six PCC samples for all the GEFs and GAPs detected. Among these proteins, only two GEFs and one putative GAP (ARHGEF1, FARP1, and ARHGAP36, respectively) were significantly differentially expressed between the tumor and the matched non-tumor tissue. In line with the reduced activity of Rho GTPase observed in PCC, the expression level of the two Rho-GEFs is significantly decreased. Indeed, statistical analysis revealed that the relative expression level of FARP1 was significantly decreased in cytosol- and membrane-enriched fractions (median values: 0.69 and 0.51, respectively). ARHGEF1 was not detected in the membrane-enriched fraction, but its expression in the cytosol-enriched fraction was decreased in PCC compared with the non-tumor tissue (median value: 0.49). Conversely, the expression level of ARHGAP36 was increased in both fractions (4.45 and 1.55 as median values for the cytosol- and membrane-enriched fractions, respectively).

Table 1

Tumor-associated expression change of Rho-GAPs and Rho-GEFs in human PCCs.

Protein nameAccession numberRelative expression (Median)
CytosolMembranes
Rho-GEFARHGEF1199002.10.49**n.d.
ARHGEF6004840.20.681.07
ARFGEF15173728.3n.d.1.33
DOCK7033407.2n.d.1.54
DOCKk8203447.3n.d.0.81
DOCK11144658.3n.d.0.78
FARP1005766.20.69*0.51***
VAV2001134398.11.50n.d.
Rho-GAPARHGAP1004308.20.941.87
ARHGAP17001006634.10.80n.d.
ARHGAP18033515.20.5n.d.
ARHGAP36144967.34.45***1.55***
ARHGAP40001164431.10.54n.d.

List of Rho-GEFs and Rho-GAPs detected by MS in cytosol- and membrane-enriched fractions obtained from human PCCs. The median value corresponding to the change in relative protein expression levels in PCCs compared with adjacent non-tumor tissue is indicated. Statistical significance for medians was determined by an ANOVA analysis, n = 6 pairs of PCCs, and matched adjacent non-tumor tissue. Proteins significantly modulated are in bold.

*P<0.05 and Q< 0.05; **P < 0.01 and Q <0.01; ***P <0.001 and Q < 0.001; n.d., not detected.

The distribution of the relative expression changes for each of the six patients is illustrated in Fig. 2 and Table 2. Except for the cytosolic level of ARHGEF1 that is decreased in the PCC tissue versus control from each of the six patients, some variability can be observed. For example, cytosolic FARP1 from tumor samples is increased for two patients, whereas relative ARHGAP36 expression in the membrane-enriched fractions is decreased for two out of six patients. Therefore, the specific modulation of FARP1, ARHGEF1 and ARHGAP36 expression in PCC needs to be confirmed. To do so, we next performed a multiplexed MRM MS assay in total protein homogenates prepared from a cohort of 25 pairs of human PCC and their matched non-tumor tissue. Among these 25 pairs of samples, 13 were tested for the Rho-GTPase activity. As MRM is a targeted MS approach that uses synthetic peptide reference standards, it is used to confirm and quantify the presence of proteins of interest on smaller amounts of sample with high sensitivity, which eliminates the need for fractionation (Keshishian et al. 2007).

Figure 2
Figure 2

Variation of ARHGAP36, ARHGEF1, and FARP1 expression in cytosol- and membrane-enriched fractions from human PCCs. Six pairs of human PCCs and adjacent non-tumor tissue were subjected to subcellular fractionation. Cytosol- and membrane-enriched fractions were analyzed by MS to quantify the relative expression change of ARHGAP36, ARHGEF1, and FARP1. Data are normalized to the expression value of the matched non-tumor adjacent tissue. Box-and-whisker plot is as described in Fig. 1. Values of relative expression of ARHGAP36, ARHGEF1, and FARP1 for each patient (numbered 1 through 6) are shown in Table 2. Statistical significance for medians was determined by a two-way ANOVA analysis (see Materials and methods section): *P < 0.05 and Q < 0.05; **P < 0.01 and Q < 0.01; ***P < 0.001 and Q < 0.001; n.d., not detected.

Citation: Endocrine-Related Cancer 23, 4; 10.1530/ERC-15-0502

Table 2

Values of relative expression of ARHGAP36, ARHGEF1, and FARP1 for each patient (numbered 1 through 6).

123456P valueQ value
CytosolARHGAP3613.876.0910.182.822.120.927748.10−111294.10−10
ARHGEF10.620.380.600.230.260.640.0060.003
FARP11.980.192.70.600.770.050.0300.012
MembranesARHGAP3615.172.862.100.681.000.492340.10−95734.10−9
FARP11.570.120.630.390.650.251193.10−51737.10−5

As observed for subcellular fractions, we found that total expression levels of both FARP1 and ARHGEF1 were significantly reduced in PCC compared with the matched adjacent non-tumor adrenal tissue, whereas ARHGAP36 expression was increased (Fig. 3A). The relative expression for each patient is detailed in Supplementary Table 1. Statistical analysis of the 25 PCC tested revealed that the median values for the relative expression of FARP1 and ARHGEF1 are 0.22 (range 2.02–0.04) and 0.57 (range 1.10–0.04), respectively. Note that 92 and 84% display a lower expression of ARHGEF1 and FARP1, respectively (Supplementary Table 1). Regarding ARHGAP36, 72% of the PCC display higher expression but with a high variability (median value 1.42; range 0.37 101).

Figure 3
Figure 3

Global variation of ARHGAP36, ARHGEF1, and FARP1 proteins and mRNA expression in human PCCs. (A) Expression change of ARHGAP36, ARHGEF1, and FARP1 was quantified at the protein level by MRM-MS in 25 pairs of human PCCs normalized to their matched adjacent non-tumor tissue. Statistical significance for medians was determined by an ANOVA analysis: *P<0.05 and Q <0.05; **P<0.01 and Q< 0.01; ***P < 0.001 and Q <0.001. (B) Comparison of ARHGAP36, ARHGEF1, and FARP1 mRNA levels by qPCR in four pairs of human PCCs normalized to their matched adjacent non-tumor tissue. Box-and-whisker plot is as described in Fig. 1.

Citation: Endocrine-Related Cancer 23, 4; 10.1530/ERC-15-0502

To investigate whether modulation of ARHGAP36, ARHGEF1, and FARP1 expression was also detectable at the mRNA level, five samples of PCC and their non-tumor counterparts were analyzed by RT-qPCR (Fig. 3B). We observed that mRNA levels of ARHGEF1 and FARP1 were lower in PCC than in the corresponding non-tumor adrenal tissue. On average, ARHGEF1 and FARP1 mRNAs decrease by 1.4-fold and 3.5-fold, respectively. Conversely, ARHGAP36 mRNA level was largely higher as we detected a 852-fold increase on average, but as observed at the protein level with a high variability (range 8.3–1520).

Impact of ARHGEF1 and FARP1 expression change on Rho GTPase activity

We next decided to investigate the potential link between Rho-GEF expression change in PCC and the activity of RhoA, Rac1, and Cdc42. To do so, we measured the level of GTP-bound Rac1, Cdc42, and RhoA in rat PCC (PC12) cells, knocked down for endogenous FARP1 or ARHGEF1 by siRNA strategy. We had to choose PC12 cells for these experiments because chromaffin cells in primary culture are resistant to siRNA transfection, whereas more than 90% of the PC12 cell population is efficiently transfected with siRNAs. We could not detect endogenous FARP1 and ARHGEF1 by immunofluorescence or western blotting with the currently available antibodies (anti-FARP1 K-20 and anti-ARHGEF1 H-165; Santa Cruz Biotechnology). Therefore, we confirmed that PC12 cells transfected with FARP1 or ARHGEF1 siRNA exhibited a significant reduction (86.3%±3% and 92.3%±2.3%, respectively) in the level of endogenous FARP1 or ARHGEF1 mRNA expression using RT-qPCR analysis (not shown). Expression and silencing of ARHGAP36 in PC12 cells could not be efficiently detected. Rat ARHGAP36 sequence in the database is still predictive (access number XM_006257554) and despite trying different sets of primers the amplicon did not satisfy qPCR quality controls. Note that although ARHGAP36 is predicted to be a Rho-GAP family member based on sequence homologies, the lack of the ‘arginine finger’ motif that stimulates GTPase hydrolysis activity suggests that ARHGAP36 is catalytically inactive (Jelen et al. 2009). Therefore, we could not assess the consequence of ARHGAP36 silencing on Rho GTPases activity.

Interestingly, reduction of endogenous FARP1 decreased the level of GTP-loaded Cdc42 by approximately 50%, whereas it did not change the level of GTP-loaded Rac1 or GTP-loaded RhoA (Fig. 4). Moreover, reduction of endogenous ARHGEF1 decreased the level of GTP-loaded Rac1 by approximately 30% without affecting the level of GTP-loaded Cdc42 or GTP-loaded RhoA (Fig. 4). These results indicate that the activation/inactivation cycle of Cdc42 and Rac1 was significantly affected in cells exhibiting a reduced level of FARP1 and ARHGEF1, respectively. Therefore, the lower activity of Cdc42 and Rac1 observed in PCC may be the result of a decrease in FARP1 and ARHGEF1 expression.

Figure 4
Figure 4

ARHGEF1 and FARP1 silencing in PC12 cells inhibits the activity level of Rac1 and Cdc42, respectively. GTP-bound RhoA, Rac1, and Cdc42 levels in lysates from PC12 cells transfected with unrelated, ARHGEF1 or FARP1 siRNA were measured by colorimetric ELISA-based assay. Data are normalized as percentage of control (PC12 cells transfected with unrelated siRNA, considered as 100%) and represented as mean± s.e.m. n=4, *P< 0.05 (Kruskal–Wallis test).

Citation: Endocrine-Related Cancer 23, 4; 10.1530/ERC-15-0502

Correlation analysis of ARHGAP36, ARHGEF1, and FARP1 expression and Rho GTPase activity

The next questions we addressed were the correlation between inhibition of Cdc42 and Rac1 activity as well as the relationship between Rho-GTPases activity and the expression of the Rho-GEFs/GAPs. To do so, we performed a principal component analysis (PCA) using the data obtained from the 13 pairs of samples (human PCC and matched adjacent non-tumor tissue) that have been tested for both Rho-GTPase activity and Rho GEF/GAP expression by MRM. As illustrated in Fig. 5A, the distribution of the tissue samples using the two first principal components reveals a clear separation between tumor and non-tumor samples, suggesting that our set of data would be able to discriminate and predict the status of these two tissue types (Fig. 5A). Accordingly, analysis of the variables projection revealed that high activities of Cdc42 and Rac1 along with high expression of FARP1 and ARHGEF1 positively correlate with the non-tumor tissue, whereas RhoA activity and ARHGAP36 have a weaker association with non-tumor tissue (Fig. 5B). Moreover, the variables Cdc42 activity, Rac1 activity, FARP1 expression, and ARHGEF1 expression all display similar projections and cluster together, indicating that they are correlated with one another.

Figure 5
Figure 5

PCA of Rho GTPase activity and GEF/GAP expression in human PCC. RhoA, Rac1, and Cdc42 activities along with expression of ARHGAP36, ARHGEF1, and FARP1 determined in the same 13 patients (paired tissues, resulting in 26 samples) were submitted for PCA. (A) Two-dimensional representation of PC1 and PC2 scores of PCC (tumor) and matched non-tumor samples (matched NT). The first principal component explains 49.3% of the variations present in the data and the second explains 19.2%. Note that this two-dimensional data resulted in evident separation between non-tumor and tumor samples. (B) Variables factor map along the PC1 and PC2 axes. The variables Cdc42 activity, Rac1 activity, ARHGEF1 expression and FARP1 expression largely contribute to PC1, display similar projections and cluster together.

Citation: Endocrine-Related Cancer 23, 4; 10.1530/ERC-15-0502

Modulation of FARP1, ARHGEF1, and ARHGAP36 expression in various human tumors

Next, we asked whether FARP1, ARHGEF1, or ARHGAP36 expression change is a specific feature of PCC or could it be detected in other types of tumors. To address this question, we have quantified by an MRM-MS assay the expression of FARP1, ARHGEF1, and ARHGAP36 in samples from six different types of non-neuroendocrine human tumors (including tumors from colon, lung, bladder, breast, kidney, and prostate) compared with their expression in the corresponding non-tumor tissue samples from the same patient (Table 3). Significant decreases in FARP1 or ARHGEF1 expression have been exclusively detected in PCC and not in the other tumors. On the other hand, both FARP1 and ARHGEF1 expression significantly increased in breast and bladder tumors, respectively (Table 3). Regarding ARHGAP36 expression, no significant changes have been observed. Altogether, these results clearly demonstrate that tumor-associated change of FARP1, ARHGEF1, and ARHGAP36 expression varies from one type of tumor to the other. So far, the concomitant decrease of FARP1 and ARHGEF1 together with the increase of ARHGAP36 has appeared to be a specific feature of PCC.

Table 3

Tumor-associated expression change in FARP1, ARHGEF1 and ARHGAP36.

ColonLungBladderBreastKidneyProstate
ARHGAP360.941.351.361.051.220.84
ARHGEF11.541.492.01**2.451.460.98
FARP11.151.781.235.73**0.971.33

Samples from colon (n = 15), lung (n= 15), breast (n = 15), kidney (n =15), prostate (n = 15) and bladder (n = 11) were used for quantification of FARP1, ARHGEF1 and ARHGAP36 expression variations by MRM-MS. Data are normalized to non-tumor adjacent tissue from the same patient. Statistical significance for medians was determined by a paired t-test.

**P< 0.01 and Q <0.01. Significant variations are indicated in bold.

Discussion

Studies from many laboratories have revealed that Rho GTPases and their regulators are closely connected to cancer cell biology. Indeed, expression changes of Rho GTPases, GEFs and GAPs have been associated with various stages of tumorigenesis in different types of tumors (Ellenbroek & Collard 2007, Grise et al. 2009, Karlsson et al. 2009, Parri & Chiarugi 2010, Vigil et al. 2010, Alan & Lundquist 2013). However, to our knowledge, Rho GTPases and their regulators have never been investigated in resection of human NETs. In this study, we have directly compared the expression and the activity of RhoA, Rac1 and Cdc42 in human PCC resection and in the adjacent non-tumor adrenal tissue from the same patient. Thus, we report here that Rac1 and Cdc42 activities decrease in human PCC, whereas their expression remains unchanged. These results highlight the fact that the expression level of Rho GTPase proteins does not necessarily reflect their activation state. However, while most of the literature support that increased expression of Rho GTPases correlates with tumor formation and progression, the activity of Rho GTPases in human tumors remains poorly explored. In line with our results, few in vivo studies have reported a loss of function of Rho GTPase in favor of tumorigenesis. For example, mice lacking Cdc42 specifically in hepatocytes developed hepatocellular carcinoma and lung metastasis (van Hengel et al. 2008), whereas loss of function of Cdc42 in hematopoietic cells triggers myeloproliferative disorders (Yang et al. 2007). Moreover, Cdc42 expression and activity are downregulated in N-MYC amplified neuroblastoma cells (Valentijn et al. 2005). Decreased Rac1 activity is known to promote epithelial tumor cells invasion and malignant progression (Malliri et al. 2002). The inactivation of RhoA potentiates the development of KRAS-induced hepatocellular carcinomas (Chew et al. 2014) and favors T-cell lymphoma (Cleverley et al. 2000, Yoo et al. 2014). The consequence of the inhibition of Cdc42 and Rac1 activities in PCC is currently unknown. We have recently generated PC12 cell lines that stably express the dominant negative Cdc42T17N or Rac1T17N mutants. Preliminary data indicate that lower activities of Rac1 and Cdc42 do not impact PC12 proliferation in these cellular models (not shown). Identifying downstream pathways of Rac1 and Cdc42 that may impact PCC development will be the next challenge. Rac1 and Cdc42 have been previously associated with different stages of tumorigenesis, including tumor initiation (cell cycle progression, apoptosis, cell survival, differentiation, etc.), growth (polarity changes, proliferation, etc.), and invasion (adhesion property, migration, etc.). Since most evidence relies on gain-of-function experiments, the decrease in Rac1 and Cdc42 activities observed in PCC is likely to engage an unconventional signaling pathway, either directly or indirectly. This aspect will therefore require further investigation.

What might be the mechanisms leading to a lower activity of Rac1 and Cdc42 in human PCC? We have hypothesized that a decrease in these two GTPase activities could result from a differential expression of their respective regulators (i.e., reduction of Rho-GEFs expression and/or overexpression of Rho-GAPs). Through a quantitative screening proteomic approach, we have been able to detect the expression of 13 Rho-GEFs and -GAPs and to quantitatively measure their expression changes in human PCC versus the matched non-tumor tissue. Interestingly, we found that the expression of the two Rho-GEFs, ARHGEF1, and FARP1, is significantly decreased, whereas ARHGAP36, one putative Rho-GAP, is significantly increased. These findings are in line with microarray analysis demonstrating that both FARP1 and ARHGEF1 mRNA levels are downregulated in PCC compared with normal tissue (data accessible at NCBI GEO database: Shankavaran U, Fliender S, Pacak K; Dec 31, 2013; accession GSE50442). Interestingly, we have demonstrated through a PCA that, altogether, Cdc42 and Rac1 activities with FARP1 and ARHGEF1 expression constitute good indicators to discriminate PCC from non-tumor adrenal medulla.

Downregulation of Rho-GEF or upregulation of Rho-GAP expression has been previously reported in various tumors. Indeed, FARP1 expression is downregulated in renal cell carcinoma, and ARHGAP36 is overexpressed in human medulloblastomas, whereas Lsc-knockout mice, the murine homolog of human ARHGEF1, display impaired B/T-lymphocyte proliferation (Francis et al. 2006, Siu et al. 2009, Rack et al. 2014). Moreover, expression of the two other GAPs, ARHGAP8 and oligophrenin-1, are upregulated in invasive cervical cancer and glioblastoma, respectively (Ljubimova et al. 2001, Song et al. 2008). These data confirm that in some cases, pathways involving Rho-GTPases need to be inhibited in the course of tumor progression. Note that we cannot completely rule out that potential mutations in Rac1 and Cdc42 could also impact their activity in human PCC. Indeed, several recent studies reported point mutations in Rho-GTPases as driver mutations in various types of cancer, including melanoma, lymphoma, breast, lung, and head and neck cancers (for review see Alan & Lundquist 2013 and Orgaz et al. 2014). However, none of these studies reported mutations leading to inactivation of the Rac1 or Cdc42 GTPase. Only one recent study reported an inactivating mutation in RhoA, specifically in angioimmunoblastic T-cell lymphoma (Yoo et al. 2014), whereas similar mutation of RhoA seems to work in a gain-of-function manner in diffuse-type gastric cancer (Kakiuchi et al. 2014). Further work will be required to identify potential mutations affecting Rho-GTPases in PCC.

While investigating a set of non-NET samples, we were unable to detect a significant decrease in FARP1 or ARHGEF1 expression nor increase in ARHGAP36 expression, whereas a higher expression of FARP1 and ARHGEF1 was observed in breast and bladder tumors, respectively. These data support the idea that tumor-associated change of Rho-GEFs and -GAPs varies considerably from one type of tumor to the other.

FARP1 (also known as CDEP) is a Rho-GEF containing a FERM (4.1protein/Ezrin/Radixin/Moesin) domain involved in membrane–cytoskeleton interaction (Koyano et al. 1997). To date, very few studies have documented the function of FARP1 except in neuronal cells in which FARP1 regulates synapse formation and dendrite morphology (Zhuang et al. 2009, Cheadle & Biederer 2012). So far, the nucleotide exchange activity of FARP1 has been demonstrated in vitro only with Rac1 and RhoA (Koyano et al. 2001, Cheadle & Biederer 2012). However, the sequence of FARP1 DH–PH region predicts that an interaction and a nucleotide exchange are highly probable with Cdc42 (Jaiswal et al. 2013). Accordingly, we have demonstrated that siRNA-based knocking down of Farp1 in rat PCC cell significantly inhibits the Cdc42 activity without affecting the RhoA or Rac1 activity. These results suggest that the reduced activity of Cdc42 observed in human PCC could be a consequence of the reduction of FARP1 expression.

ARHGEF1 (also known as p115Rho-GEF) belongs to the family of GEFs for Rho GTPases that contain a regulator of G protein signaling (RGS) domain (Siehler 2009). RGS-Rho-GEFs are GAPs toward G12/13 α subunit, and binding to Gα12/13 stimulates their GEF activity toward the Rho GTPases (Dutt et al. 2004). To become active, ARHGEF1 needs to translocate from the cytoplasm to the plasma membrane. Since we have detected ARHGEF1 in cytoplasmic fraction but not in the fraction enriched in plasma membrane, we cannot rule out that for some reasons, ARHGEF1 remains sequestered within the cytoplasm in PCC. As a GAP for G12/13 α subunit, ARHGEF1 downregulation might increase Gα12/13 activation at the plasma membrane, and constitutively active Gα12/13 proteins are known to be oncogenes in various cell types (Siehler 2009). The link between ARHGEF1 expression and Rac1 activity in PCC remains puzzling. In vitro experiments suggested that the nucleotide exchange activity of ARHGEF1 is specific to Rho isoforms (Glaven et al. 1996, Jaiswal et al. 2011). In this study, siRNA-based knocking down of Arhgef1 in rat PCC cells slightly inhibits the Rac1 activity without affecting the RhoA or Cdc42 activity. We have currently no explanation for the basis of the mechanism leading to Rac1 inactivation, and further investigation will be required.

ARHGAP36 contains a Rho-GAP domain lacking the ‘arginine finger’ motif that stimulates GTPase hydrolysis, suggesting that ARHGAP36 might function through non-catalytic mechanisms. Accordingly, the only publication on ARHGAP36 function suggested a Rho-independent role of ARHGAP36 in the Hedgehog pathway activation (Rack et al. 2014). Interestingly, upregulation of the Hedgehog pathway stimulates proliferation of rat PCC cells (Carr et al. 2014). Whether Hedgehog pathway is perturbed in human PCC will require our future attention.

To conclude, we have reported here for the first time a reduced activity of both Rac1 and Cdc42 in human PCC resection as well as tumor-associated expression changes of FARP1, ARHGEF1, and ARHGAP36. The next challenge will be to understand why these changes are associated with this type of tumor and to unravel the underlying mechanisms.

Supplementary data

This is linked to the online version of the paper at http://dx.doi.org/10.1530/ERC-15-0502.

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

Part of this work has been supported by grants from the ‘Ligue contre le Cancer’ (CCIR-GE 2013/14) and ‘Fonds Européen de Développement Economique Régional’ (FEDER 32309-2011/13) to SG; by a grant from CQDM (2011/13) to JL, EP and DC; and by a PhD fellowship from ‘Région Alsace’ and ‘Ligue Contre le Cancer’ to SH.

Acknowledgments

The authors are grateful to Nicolas Théry (INCI, CNRS UPR3212, Strasbourg, France) for his technical assistance in setting up the Rho GTPase activity assay in human tumor and to Kevin Dorgans (INCI, CNRS UPR3212, Strasbourg, France) and Ameur Manceur (Caprion, Montreal, Canada) for their help in PCA analysis. We thank Dr Christophe Dubessy (IRIB, Inserm U982, Rouen, France) for critical reading of the manuscript and fruitful discussion.

References

  • Alan JK & Lundquist EA 2013 Mutationally activated Rho GTPases in cancer. Small GTPases 4 159163. (doi:10.4161/sgtp.26530)

  • Carr AL, Sun L, Lee E, Li P, Antonacci C, Gorbea E, Finlay C & Li L 2014 The human oncogene SCL/TAL1 interrupting locus is required for mammalian dopaminergic cell proliferation through the Sonic hedgehog pathway. Cellular Signalling 26 306312. (doi:10.1016/j.cellsig.2013.11.010)

    • Search Google Scholar
    • Export Citation
  • Cheadle L & Biederer T 2012 The novel synaptogenic protein Farp1 links postsynaptic cytoskeletal dynamics and transsynaptic organization. Journal of Cell Biology 199 9851001. (doi:10.1083/jcb.201205041)

    • Search Google Scholar
    • Export Citation
  • Cherfils J & Zeghouf M 2013 Regulation of small GTPases by GEFs, GAPs, and GDIs. Physiological Review 93 269309. (doi:10.1152/physrev.00003.2012)

    • Search Google Scholar
    • Export Citation
  • Chew TW, Liu XJ, Liu L, Spitsbergen JM, Gong Z & Low BC 2014 Crosstalk of Ras and Rho: Activation of RhoA abates Kras-induced liver tumorigenesis in transgenic zebrafish models. Oncogene 33 27172727. (doi:10.1038/onc.2013.240)

    • Search Google Scholar
    • Export Citation
  • Cleverley SC, Costello PS, Henning SW & Cantrell DA 2000 Loss of Rho function in the thymus is accompanied by the development of thymic lymphoma. Oncogene 19 1320. (doi:10.1038/sj.onc.1203259)

    • Search Google Scholar
    • Export Citation
  • Dutt P, Nguyen N & Toksoz D 2004 Role of Lbc RhoGEF in Galpha12/13-induced signals to Rho GTPase. Cellular Signalling 16 201209. (doi:10.1016/s0898-6568(03)00132-3)

    • Search Google Scholar
    • Export Citation
  • Ellenbroek SI & Collard JG 2007 Rho GTPases: Functions and association with cancer. Clinical & Experimental Metastasis 24 657672. (doi:10.1007/s10585-007-9119-1)

    • Search Google Scholar
    • Export Citation
  • Francis SA, Shen X, Young JB, Kaul P & Lerner DJ 2006 Rho GEF Lsc is required for normal polarization, migration, and adhesion of formyl-peptide-stimulated neutrophils. Blood 107 16271635. (doi:10.1182/blood-2005-03-1164)

    • Search Google Scholar
    • Export Citation
  • Gasman S, Chasserot-Golaz S, Malacombe M, Way M & Bader MF 2004 Regulated exocytosis in neuroendocrine cells: A role for subplasmalemmal Cdc42/N-WASP-induced actin filaments. Molecular Biology of the Cell 15 520531. (doi:10.1091/mbc.E03-06-0402)

    • Search Google Scholar
    • Export Citation
  • Glaven JA, Whitehead IP, Nomanbhoy T, Kay R & Cerione RA 1996 Lfc and Lsc oncoproteins represent two new guanine nucleotide exchange factors for the Rho GTP-binding protein. Journal of Biological Chemistry 271 2737427381. (doi:10.1074/jbc.271.44.27374)

    • Search Google Scholar
    • Export Citation
  • Greene LA & Tischler AS 1976 Establishment of a noradrenergic clonal line of rat adrenal pheochromocytoma cells which respond to nerve growth factor. PNAS 73 24242428. (doi:10.1073/pnas.73.7.2424)

    • Search Google Scholar
    • Export Citation
  • Grise F, Bidaud A & Moreau V 2009 Rho GTPases in hepatocellular carcinoma. Biochimica et Biophysica Acta 1795 137151. (doi:10.1016/j.bbcan.2008.12.003)

    • Search Google Scholar
    • Export Citation
  • Jaiswal M, Dvorsky R & Ahmadian MR 2013 Deciphering the molecular and functional basis of Dbl family proteins: A novel systematic approach toward classification of selective activation of the Rho family proteins. Journal of Biological Chemistry 288 44864500. (doi:10.1074/jbc.M112.429746)

    • Search Google Scholar
    • Export Citation
  • Jaiswal M, Gremer L, Dvorsky R, Haeusler LC, Cirstea IC, Uhlenbrock K & Ahmadian MR 2011 Mechanistic insights into specificity, activity, and regulatory elements of the regulator of G-protein signaling (RGS)-containing Rho-specific guanine nucleotide exchange factors (GEFs) p115, PDZ-RhoGEF (PRG), and leukemia-associated RhoGEF (LARG). Journal of Biological Chemical 286 1820218212. (doi:10.1074/jbc.M111.226431)

    • Search Google Scholar
    • Export Citation
  • Jelen F, Lachowicz P, Apostoluk W, Mateja A, Derewenda ZS & Otlewski J 2009 Dissecting the thermodynamics of GAP-RhoA interactions. Journal of Structural Biology 165 1018. (doi:10.1016/j.jsb.2008.09.007)

    • Search Google Scholar
    • Export Citation
  • Kakiuchi M, Nishizawa T, Ueda H, Gotoh K, Tanaka A, Hayashi A, Yamamoto S, Tatsuno K, Katoh H, Watanabe Y, et al.2014 Recurrent gain-of-function mutations of RHOA in diffuse-type gastric carcinoma. Nature Genetics 46 583587. (doi:10.1038/ng.2984)

    • Search Google Scholar
    • Export Citation
  • Karlsson R, Pedersen ED, Wang Z & Brakebusch C 2009 Rho GTPase function in tumorigenesis. Biochimica et Biophysica Acta 1796 9198. (doi:10.1016/j.bbcan.2009.03.00)

    • Search Google Scholar
    • Export Citation
  • Keshishian H, Addona T, Burgess M, Kuhn E & Carr SA 2007 Quantitative, multiplexed assays for low abundance proteins in plasma by targeted mass spectrometry and stable isotope dilution. Molecular & Cellular Proteomics 6 22122229. (doi:10.1074/mcp.M700354-MCP200)

    • Search Google Scholar
    • Export Citation
  • Koyano Y, Kawamoto T, Kikuchi A, Shen M, Kuruta Y, Tsutsumi S, Fujimoto K, Noshiro M, Fujii K & Kato Y 2001 Chondrocyte-derived ezrin-like domain containing protein (CDEP), a rho guanine nucleotide exchange factor, is inducible in chondrocytes by parathyroid hormone and cyclic AMP and has transforming activity in NIH3T3 cells. Osteoarthritis Cartilage 9 (Supplement A) S64S68. (doi:10.1053/joca.2001.0446)

    • Search Google Scholar
    • Export Citation
  • Koyano Y, Kawamoto T, Shen M, Yan W, Noshiro M, Fujii K & Kato Y 1997 Molecular cloning and characterization of CDEP, a novel human protein containing the ezrin-like domain of the band 4.1 superfamily and the Dbl homology domain of Rho guanine nucleotide exchange factors. Biochemical and Biophysical Research Communications 241 369375. (doi:10.1006/bbrc.1997.7826)

    • Search Google Scholar
    • Export Citation
  • Livak KJ & Schmittgen TD 2001 Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25 402408. (doi:10.1006/meth.2001.1262)

    • Search Google Scholar
    • Export Citation
  • Ljubimova JY, Khazenzon NM, Chen Z, Neyman YI, Turner L, Riedinger MS & Black KL 2001 Gene expression abnormalities in human glial tumors identified by gene array. International Journal of Oncology 18 287295. (doi:10.3892/ijo.18.2.287)

    • Search Google Scholar
    • Export Citation
  • Malliri A, van der Kammen RA, Clark K, van der Valk M, Michiels F & Collard JG 2002 Mice deficient in the Rac activator Tiam1 are resistant to Ras-induced skin tumours. Nature 417 867871. (doi:10.1038/nature00848)

    • Search Google Scholar
    • Export Citation
  • Orgaz JL, Herraiz C & Sanz-Moreno V 2014 Rho GTPases modulate malignant transformation of tumor cells. Small GTPases 5 e29019.

  • Parri M & Chiarugi P 2010 Rac and Rho GTPases in cancer cell motility control. Cell Commununication and Signalling 8 23. (doi:10.1186/1478-811X-8-23)

    • Search Google Scholar
    • Export Citation
  • Rack PG, Ni J, Payumo AY, Nguyen V, Crapster JA, Hovestadt V, Kool M, Jones DT, Mich JK, Firestone AJ, et al.2014 Arhgap36-dependent activation of Gli transcription factors. PNAS 111 1106111066. (doi:10.1073/pnas.1322362111)

    • Search Google Scholar
    • Export Citation
  • Siehler S 2009 Regulation of RhoGEF proteins by G12/13-coupled receptors. British Journal of Pharmacology 158 4149. (doi:10.1111/j.1476-5381.2009.00121.x)

    • Search Google Scholar
    • Export Citation
  • Siu KW, DeSouza LV, Scorilas A, Romaschin AD, Honey RJ, Stewart R, Pace K, Youssef Y, Chow TF & Yousef GM 2009 Differential protein expressions in renal cell carcinoma: New biomarker discovery by mass spectrometry. Journal of Proteome Research 8 37973807. (doi:10.1021/pr800389e)

    • Search Google Scholar
    • Export Citation
  • Song JY, Lee JK, Lee NW, Jung HH, Kim SH & Lee KW 2008 Microarray analysis of normal cervix, carcinoma in situ, and invasive cervical cancer: Identification of candidate genes in pathogenesis of invasion in cervical cancer. International Journal of Gynecological Cancer 18 10511059. (doi:10.1111/j.1525-1438.2007.01164.x)

    • Search Google Scholar
    • Export Citation
  • Taal BG & Visser O 2004 Epidemiology of neuroendocrine tumours. Neuroendocrinology 80 (Supplement 1) 37. (doi:10.1159/000080731)

  • Tsirlin A, Oo Y, Sharma R, Kansara A, Gliwa A & Banerji MA 2014 Pheochromocytoma: A review. Maturitas 77 229238. (doi:10.1016/j.maturitas.2013.12.009)

  • Valentijn LJ, Koppen A, van Asperen R, Root HA, Haneveld F & Versteeg R 2005 Inhibition of a new differentiation pathway in neuroblastoma by copy number defects of N-myc, Cdc42, and nm23 genes. Cancer Research 65 31363145. (doi:10.1158/0008-5472.CAN-04-2469)

    • Search Google Scholar
    • Export Citation
  • van Hengel J, D’Hooge P, Hooghe B, Wu X, Libbrecht L, De Vos R, Quondamatteo F, Klempt M, Brakebusch C & van Roy F 2008 Continuous cell injury promotes hepatic tumorigenesis in cdc42-deficient mouse liver. Gastroenterology 134 781792. (doi:10.1053/j.gastro.2008.01.002)

    • Search Google Scholar
    • Export Citation
  • Vega FM & Ridley AJ 2008 Rho GTPases in cancer cell biology. FEBS Letters 582 20932101. (doi:10.1016/j.febslet.2008.04.039)

  • Vigil D, Cherfils J, Rossman KL & Der CJ 2010 Ras superfamily GEFs and GAPs: Validated and tractable targets for cancer therapy? Natural Review of Cancer 10 842857. (doi:10.1038/nrc2960)

    • Search Google Scholar
    • Export Citation
  • Yang L, Wang L, Kalfa TA, Cancelas JA, Shang X, Pushkaran S, Mo J, Williams DA & Zheng Y 2007 Cdc42 critically regulates the balance between myelopoiesis and erythropoiesis. Blood 110 38533861. (doi:10.1182/blood-2007-03-079582)

    • Search Google Scholar
    • Export Citation
  • Yao JC, Hassan M, Phan A, Dagohoy C, Leary C, Mares JE, Abdalla EK, Fleming JB, Vauthey JN, Rashid A, et al.2008 One hundred years after ‘carcinoid’: Epidemiology of and prognostic factors for neuroendocrine tumors in 35,825 cases in the United States. Journal of Clinical Oncology 26 30633072. (doi:10.1200/JCO.2007.15.4377)

    • Search Google Scholar
    • Export Citation
  • Yoo HY, Sung MK, Lee SH, Kim S, Lee H, Park S, Kim SC, Lee B, Rho K, Lee JE, et al.2014 A recurrent inactivating mutation in RHOA GTPase in angioimmunoblastic T cell lymphoma. Natural Genetics 46 371375. (doi:10.1038/ng.2916)

    • Search Google Scholar
    • Export Citation
  • Zhuang B, Su YS & Sockanathan S 2009 FARP1 promotes the dendritic growth of spinal motor neuron subtypes through transmembrane Semaphorin6A and PlexinA4 signaling. Neuron 61 359372. (doi:10.1016/j.neuron.2008.12.022)

    • Search Google Scholar
    • Export Citation

 

Society for Endocrinology

Sept 2018 onwards Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 1456 405 33
PDF Downloads 687 393 45
  • View in gallery

    The activities of Rac1 and Cdc42 are reduced in human PCCs compared with adjacent non-tumor tissue. (A) Colorimetric ELISA-based assay was used to estimate Rho GTPase activity in lysates from human PCCs and matched adjacent non-tumor tissue. Total amounts of Rho GTPases were estimated by loading proteins in vitro with nonhydrolysable GTP (GTPγS) and absorbance of the sample normalized to the absorbance of 1 ng of recombinant and activated Rho GTPases. Box-and-whisker plot represents the first quartile (bottom line), the median (line in the box) and the third quartile (upper line). Whiskers correspond to the 5th (bottom) and 95th (top) percentiles, and the black dots represent outlier observations. (B) Rho GTPase activity levels were determined by measuring the amounts of GTP-bound GTPases in the sample lysate, normalized to the maximal activity that can be detected. Maximal activity was obtained by loading proteins in vitro with GTP. Statistical significance for medians was determined by a Wilcoxon signed-rank nonparametric analysis for matched samples, n = 18 pairs of human PCCs, and matched adjacent non-tumor tissue. *P<0.05; **P<0.01; ***P <0.001; n.s., not significant. (C, D) Representation of total amounts and activity of Rho GTPases for each pair of tissue tested per patient. A full colour version of this figure is available at http://dx.doi.org/10.1530/ERC-15-0502.

  • View in gallery

    Variation of ARHGAP36, ARHGEF1, and FARP1 expression in cytosol- and membrane-enriched fractions from human PCCs. Six pairs of human PCCs and adjacent non-tumor tissue were subjected to subcellular fractionation. Cytosol- and membrane-enriched fractions were analyzed by MS to quantify the relative expression change of ARHGAP36, ARHGEF1, and FARP1. Data are normalized to the expression value of the matched non-tumor adjacent tissue. Box-and-whisker plot is as described in Fig. 1. Values of relative expression of ARHGAP36, ARHGEF1, and FARP1 for each patient (numbered 1 through 6) are shown in Table 2. Statistical significance for medians was determined by a two-way ANOVA analysis (see Materials and methods section): *P < 0.05 and Q < 0.05; **P < 0.01 and Q < 0.01; ***P < 0.001 and Q < 0.001; n.d., not detected.

  • View in gallery

    Global variation of ARHGAP36, ARHGEF1, and FARP1 proteins and mRNA expression in human PCCs. (A) Expression change of ARHGAP36, ARHGEF1, and FARP1 was quantified at the protein level by MRM-MS in 25 pairs of human PCCs normalized to their matched adjacent non-tumor tissue. Statistical significance for medians was determined by an ANOVA analysis: *P<0.05 and Q <0.05; **P<0.01 and Q< 0.01; ***P < 0.001 and Q <0.001. (B) Comparison of ARHGAP36, ARHGEF1, and FARP1 mRNA levels by qPCR in four pairs of human PCCs normalized to their matched adjacent non-tumor tissue. Box-and-whisker plot is as described in Fig. 1.

  • View in gallery

    ARHGEF1 and FARP1 silencing in PC12 cells inhibits the activity level of Rac1 and Cdc42, respectively. GTP-bound RhoA, Rac1, and Cdc42 levels in lysates from PC12 cells transfected with unrelated, ARHGEF1 or FARP1 siRNA were measured by colorimetric ELISA-based assay. Data are normalized as percentage of control (PC12 cells transfected with unrelated siRNA, considered as 100%) and represented as mean± s.e.m. n=4, *P< 0.05 (Kruskal–Wallis test).

  • View in gallery

    PCA of Rho GTPase activity and GEF/GAP expression in human PCC. RhoA, Rac1, and Cdc42 activities along with expression of ARHGAP36, ARHGEF1, and FARP1 determined in the same 13 patients (paired tissues, resulting in 26 samples) were submitted for PCA. (A) Two-dimensional representation of PC1 and PC2 scores of PCC (tumor) and matched non-tumor samples (matched NT). The first principal component explains 49.3% of the variations present in the data and the second explains 19.2%. Note that this two-dimensional data resulted in evident separation between non-tumor and tumor samples. (B) Variables factor map along the PC1 and PC2 axes. The variables Cdc42 activity, Rac1 activity, ARHGEF1 expression and FARP1 expression largely contribute to PC1, display similar projections and cluster together.