Dosage-dependent regulation of VAV2 expression by steroidogenic factor-1 drives adrenocortical carcinoma cell invasion
Carmen Ruggiero,1,2,3,4 Mabrouka Doghman-Bouguerra,1,2,3,4 Silviu Sbiera,5 Iuliu Sbiera,5 Maddy Parsons,6 Bruno Ragazzon,7,8,9 Aurelie Morin,9,10 Estelle Robidel,9,10 Judith Favier,9,10 Jérôme Bertherat,7,8,9 Martin Fassnacht,11 Enzo Lalli1,2,3,4*
Adrenocortical carcinoma (ACC) is a rare endocrine malignancy with a dismal prognosis. Genomic studies have enabled progress in our understanding of the molecular bases of ACC, but factors that influence its prognosis are lacking. Amplification of the gene encoding the transcription factor steroidogenic factor-1 (SF-1; also known as NR5A1) is one of the genetic alterations common in ACC. We identified a transcriptional regulatory mechanism involving increased abundance of VAV2, a guanine nucleotide exchange factor for small GTPases that control the cytoskeleton, driven by increased expression of the gene encoding SF-1 in ACC. Manipulating SF-1 and VAV2 abundance in cultured ACC cells revealed that VAV2 was a critical factor for SF-1-induced cytoskeletal remodeling and invasion in culture (Matrigel) and in vivo (chicken chorioallantoic membrane) models. Analysis of ACC patient cohorts indicated that greater VAV2 abundance robustly correlated with poor prognosis in ACC patients. Because VAV2 is a druggable target, our findings suggest that blocking VAV2 may be a new therapeutic approach to inhibit metastatic progression in ACC patients.
INTRODUCTION
Adrenocortical carcinoma (ACC) is a rare endocrine malignancy with persistently poor prognosis (1, 2). Current polychemotherapeutic regi- mens or targeted therapies only have a limited efficacy in advanced- stage ACC (3-5). New hopes for more efficient treatments for ACC stem from studies that have identified important molecular actors implicated in its pathogenesis (6-11). Those include the transcription factor steroidogenic factor-1 (SF-1), encoded by NR5A1, a nuclear receptor that has a key role in adrenocortical development and reg- ulation of steroidogenesis (12). Our previous studies have revealed that increased SF-1 dosage (abundance or gene amplification) elicits increased proliferation in human H295R ACC cells and induces adrenocortical tumorigenesis in Sf-1 transgenic mice (8, 13). Further- more, increased SF-1 dosage is associated with poor clinical outcome in ACC and represents a stage-independent prognostic marker (14, 15). We have also provided the proof of principle that SF-1 inverse agonists can counteract the effects of SF-1 dosage on ACC cell proliferation (16).
Increased SF-1 abundance in ACC cells induces relevant modifi- cations of gene expression, affecting not only its classical steroidogenic target genes but also various other genes involved in a number of physiological processes and signaling pathways (17, 18). It is remarkable that some of those changes mirror similar changes in gene expression found in ACC compared to benign adrenocortical tumors and normal adrenal cortex tissue (8, 19, 20). Those data validated the use of our
cellular model as a discovery platform for new factors involved in the definition of the malignant phenotype in ACC, because currently, in vivo models for metastatic ACC are lacking.
Here, we focused on one of the SF-1 dosage-dependent target genes identified in our previous genomic studies, VAV2, encoding a guanine nucleotide exchange factor (GEF) known to have an impor- tant function in many aspects of tumor biology (21, 22). VAV2 was critical to SF-1-induced cytoskeleton remodeling and subsequent invasive capacity in ACC cells. Additionally, tumor VAV2 abundance was significantly and positively correlated with poor prognosis in three different ACC patient cohorts. These results identify VAV2 as a critical factor driving malignancy and potentially druggable target in ACC.
RESULTS
VAV2 abundance and activity of the small GTPases CDC42 and RAC1 are increased by augmented SF-1 dosage in ACC cells
We developed the H295R-TR SF-1 human ACC cell line that over- expresses SF-1 in a doxycycline (Dox)-dependent fashion to model the SF-1-dependent phenotypes induced by increased dosage of this transcription factor in ACC (8). Gene expression profiling showed that in this cell line, SF-1 regulates the expression of distinct sets of target genes according to its dosage (8, 17, 18). One of these genes is VAV2, encoding a multidomain Rho GEF (21, 22), whose abundance was significantly increased after induction of SF-1 (8, 17). Increased VAV2 expression correlated with increased SF-1 binding both to the upstream region and to intronic sites of the VAV2 gene (Fig. 1A), as detected by chromatin immunoprecipitation sequencing (ChIP-seq) (17). The abundance of VAV2 transcript (Fig. 1B) and protein (Fig. 1, C and D) was significantly increased in H295R-TR SF-1 cells after Dox stimula- tion, showing a delay after the increase of SF-1 abundance (Fig. 1, C and E), as expected for a transcriptionally regulated gene. Conversely, VAV2 abundance was not increased by Dox treatment in the parental cell line that does not overexpress SF-1 (H295R-TR) and in an H295R cell line
2017 @ The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science.
1Université Côte d’Azur, Sophia Antipolis, 06560 Valbonne, France. 2CNRS UMR7275, Sophia Antipolis, 06560 Valbonne, France. 3NEOGENEX CNRS Interna- tional Associated Laboratory, Sophia Antipolis, 06560 Valbonne, France. 4Institut de Pharmacologie Moléculaire et Cellulaire, Sophia Antipolis, 06560 Valbonne, France. 5Division of Endocrinology and Diabetes, Department of Internal Medicine I, University Hospital, University of Würzburg, 97080 Würzburg, Germany. 6Randall Division of Cell and Molecular Biophysics, King’s College London, London SE1 1UL, U.K. 7Inserm, U1016, Institut Cochin, 75014 Paris, France. 8CNRS UMR8104, 75014 Paris, France. 9Université Paris Descartes, Sorbonne Paris Cité, 74014 Paris, France. 1ºInserm, UMR970, Paris Cardiovascular Research Centre, 75015 Paris, France. 11Comprehensive Cancer Center Mainfranken, University of Würzburg, 97080 Würzburg, Germany. *Corresponding author. Email: ninino@ipmc.cnrs.fr
A
B
3.
Basal
VAV2 mRNA relative expresssion (fold)
**
2.
*
Dox
1.
SARDH
VAV2
BRD3
0
Time (hours)
9
12
24
48
C
F
kDa
kDa
SF-1
-50
p-VAV2
-118
VAV2
-118
VAV2
-118
B-Tubulin
-50
B-Tubulin
50
Dox
+
+
+
+
Dox
+
+
+
Time (hours)
9
12
24
48
Time (hours)
12
24
48
D
VAV2/B-tubulin (% of ctrl)
G
250
150
200
**
**
p-VAV2/VAV2 (%of ctrl)
150
100
100-
50
50
0
0
Time (hours)
9
12
24
48
Time (hours)
12
24
48
E
SF-1/B-tubulin (%of ctrl)
600-
**
400
*
*
200
0
Time (hours)
9
12
24
48
that overexpresses a transcriptionally inactive SF-1 under the control of Dox (fig. S1A) (8), confirming that transcriptional activity of SF-1 is re- quired to produce VAV2 mRNA increase. It is worth noting that induc- tion of SF-1 by Dox in the H295R-TR SF-1 clone was not homogeneous in all cells. About 30% of Dox-treated cells expressed low SF-1 levels, comparable to unstimulated cells (fig. S1B).
VAV2 activity is regulated by tyrosine phosphorylation by receptor tyrosine kinases (23, 24). The abundance of active, Tyr174-phosphorylated VAV2 increased consequently to Dox stimulation of H295R-TR SF- 1 cells, but the proportion of phospho-VAV2 compared to total VAV2 remained constant (Fig. 1, F and G). These results show that increased SF- 1 dosage increases the pool of active VAV2 in ACC cells.
Rho family guanosine triphosphatases (GTPases) are molecular switches regulating cytoskeletal architecture and cell motility, which cycle between a guanosine 5’-triphosphate (GTP)-bound active and a guanosine diphosphate-bound inactive conformation. Conversion to the active form is facilitated by GEFs, such as VAV2 (25, 26). We
A
0
Eff (%)
20
B
CDC42
20
Vehicle
Dox
FRET efficiency (%)
15
10
5
0
C
D
Vehicle
Dox
RAC1
Vehicle
Dox
FRET efficiency (%)
15
*
10
5
0
Vehicle
Dox
E
F
RHOA
FRET efficiency (%)
20
Vehicle
Dox
15
,
6
10
5
0
Vehicle
Dox
G
200-
Active GTPase (% of ctrl)
150
100-
50
0
CDC42
RAC1
RHOA
then determined whether the activity of the Rho family GTPases RHOA, RAC1, and CDC42 is modulated by SF-1 dosage in H295R cells. Both a biochemical and a fluorescence resonance energy trans- fer (FRET)-based assay in living cells showed that in Dox-treated H295R-TR SF-1 cells, the activity of CDC42 and RAC1, but not RHOA, was significantly increased compared to vehicle-treated cells (Fig. 2).
Cytoskeleton remodeling and invasive capacity of ACC cells are increased in a VAV2-dependent fashion
CDC42 and RAC1 have partially overlapping but distinct roles in regulating cytoskeleton remodeling, cell migration, and invasive capac- ity (25, 27). Their increased activity in Dox-treated H295R-TR SF-1 cells suggested that increased SF-1 dosage modulates cytoskeleton dynamics. Dox treatment of H295R-TR SF-1 cells significantly increased the percentage of cells displaying filopodia (a CDC42-dependent pheno- type) in H295R-TR SF-1 cells but not in the parental H295R-TR cells (Fig. 3, A to C). Similar results were obtained for lamellipodia-ruffles (a RAC-dependent phenotype) (Fig. 3, D and E). We took advantage
of the inhomogeneous up-regulation of SF-1 by Dox in the H295R-TR SF-1 cell line (fig. S1B) to precisely quantify the percentage of filopodia- forming cells and the number of filopodia per cell in either low or high SF-1-expressing cells. Both the percentage of filopodia-forming cells (Fig. 3F) and the number of filopodia per cell (Fig. 3G) were signif- icantly greater only in cells expressing high amounts of SF-1 compared to unstimulated cells (as in Fig. 3A). Filopodia were only present in the cell expressing high SF-1 levels. These results indicate a direct relationship between high SF-1 abundance and cytoskeletal remodeling in ACC cells. Consistent with previous reports in COS7 cells (24), transfection of green fluorescent protein (GFP)-VAV2 was able to induce cyto-
A
Vehicle
Because cytoskeletal remodeling is directly linked to cell migra- tion and invasion, we used a Matrigel invasion assay to measure the ef- fect that an increased SF-1 dosage has on the invasive capacity of H295R cells. Dox treatment of H295R-TR SF-1 cells induced a significant
Dox
A
Phalloidin
B
GFP
Merge
siVAV2
siC
SİVAV2
GFP
siC
kDa
B
C
VAV2
118
Filopodia-forming cells (% of ctrl)
200-
200-
Filopodia-forming cells (% of ctrl)
150
150
GFP-VAV2
ß-Tubulin
-50
*
Dox
+
+
100
100
50
50
0
0
C
D
D
Dox
12 hours
24 hours
48 hours
72 hours
E
Dox
2 hours 24 hours 48 hours 72
hours
250
150-
siC
siC
Lamellipodia-ruffles-forming cells (% of ctrl)
200-
Lamellipodia-ruffles-forming cells (% of ctrl)
200
VAV2/B-tubulin ratio (% of ctrl)
200
siVAV2
Filopodia-forming cells (% of ctrl)
siVAV2
100
150
*
150
150
*
100-
100
100
50
50-
50
50
0
0
0
0
12 hours 24 hours 48 hours 72 hours
Vehicle
Dox
Vehicle
Dox
12 hours 24 hours 48 hours 72 hours
Dox
Dox
F
G
% of filopodia-forming cells
100
**
Number of filopodia per cell
E
F
15
**
10
**
*
*
80
Number of filopodia per cell
siC
Lamellipodia-ruffles-forming cells (% of ctrl)
250
siC
60
10
8
siVAV2
200
siVAV2
40
5
6
150
20
4.
100
0
0
SF-1
Low
High
SF-1
Low
High
2.
50
Vehicle
Dox
Vehicle
Dox
0
0
Vehicle
Dox
Vehicle
Dox
skeleton remodeling also in H295R-TR SF-1 cells (Fig. 4A). We used RNA interference to assess the impact of endogenous VAV2 on the increased number of filopodia and lamellipodia-ruffles after Dox treatment in this cell line. Nucleofection of VAV2-specific small inter- fering RNA (siRNA) reduced VAV2 protein abundance by about 50% in H295R-TR SF-1 cells either cultured in basal conditions or after Dox stimulation (Fig. 4, B and C). The Dox-stimulated increase in VAV2 protein was then virtually abolished by VAV2 knockdown. Whereas VAV2 knockdown had no effect on the percentage of lamellipodia- ruffles- and filopodia-forming cells or on the number of filopodia per cell under basal conditions, it decreased all those parameters to the baseline level in Dox-treated cells (Fig. 4, D to F). These results suggest that increased VAV2 triggered by increased SF-1 dosage in H295R cells is critical to produce cytoskeleton remodeling.
Fig. 4. VAV2 is critical for cytoskeleton remodeling triggered by an increased SF-1 dosage in ACC cells. (A) Cytoskeletal morphology revealed by phalloidin staining (red) in H295R-TR SF-1 cells transfected with vectors encoding either GFP or GFP-VAV2 (green). Scale bars, 5 um. (B) VAV2 protein knockdown by nucleofection of specific siRNA (siVAV2) compared to control siRNA (siC) in H295R-TR SF-1 cells treated with either vehicle or Dox. ß-Tubulin is shown as a control. (C) Quantification of VAV2 knockdown. White histograms, cells nucleo- fected with siC; gray histogram, cells nucleofected with siVAV2. (D to F) Effect of VAV2 knockdown on the percentage of filopodia-forming (D) and lamellipodia- ruffle-forming cells (F) and the number of filopodia per cell (E) in Dox-treated H295R-TR SF-1 cells. Data are means ± SEM of at least three independent experiments, each analyzing more than 250 cells per condition. * P < 0.05; ** P < 0.01; *** P < 0.001, ANOVA with Tukey’s posttest for multiple comparisons.
increase in the number of cells migrated through the Matrigel matrix compared to cells treated with vehicle only (Fig. 5, A and B). Phospho- rylation of cofilin at Ser3 is known to suppress its function as an actin depolymerizing factor (28). Consistent with the invasion assay data, we observed cofilin dephosphorylation in Dox-treated cells, which per- sisted up to the end of the invasion assay time (fig. S1C). Conversely, no increase in invasion capacity could be measured after Dox treatment for cells expressing AF-2 mutant SF-1 or for the parental H295R-TR clone that does not overexpress SF-1 (Fig. 5B). In vivo invasion and metastatic potential of H295R-TR SF-1 cells were assessed using a chicken chorioallantoic membrane (CAM) assay (Fig. 5C) (29, 30). Dox-treated cells were also more invasive in this assay compared with vehicle-treated cells (Fig. 5D). We confirmed that VAV2 knockdown remained efficient up to the end of the Matrigel invasion assay protocol (Fig. 5, E and F) and tested the effect that VAV2 knockdown has on the invasive capacity of H295R-TR SF-1 cells treated with vehicle only or with Dox. Whereas VAV2 knockdown had no impact on the invasive capacity of those cells under basal conditions, it did reduce the increased cellular invasion observed after Dox treatment back to baseline levels (Fig. 5, G and H). Cotransfection of an RNA interference-insensitive wild-type GFP-VAV2 fusion, but not of a GEF-inactive GFP-VAV2 mutant (Fig. 5G), was able to rescue the defect of invasive capacity observed when endogenous VAV2 was depleted (Fig. 5H). These data show that VAV2 is a critical mediator of the increased invasion capac- ity determined by increased SF-1 abundance in ACC cells.
High VAV2 abundance is a negative prognostic marker in ACC
These results in the cellular model point to a more aggressive phe- notype of ACC cells with high VAV2 abundance and prompted us to analyze VAV2 abundance in tumors.
NR5A1 and VAV2 transcripts were significantly correlated in two different ACC cohorts [Cochin and the Cancer Genome Atlas (TCGA)], which had no overlap (Fig. 6, A and B, and table S1). Patients with ACC were subdivided into two groups based on their relative VAV2 transcript abundance in the tumor (low and high; Fig. 6, C and D). In both cohorts, high VAV2 expression was associated with a significantly reduced overall survival, whether analyzed as a continuous variable [Cochin cohort: hazard ratio (HR), 2.52; 95% confidence interval (CI), 1.35 to 4.71; P = 0.003; TCGA cohort: HR, 1.48; 95% CI, 1.07 to 2.06; P = 0.018] or after dichotomization between ACC with low and high VAV2 (Cochin cohort: HR, 4.97; 95% CI, 1.80 to 13.7; P = 0.002; TCGA cohort: HR, 2.35; 95% CI, 1.07 to 5.14; P = 0.033) (Fig. 6, E and F). VAV2 transcript abundance was also significantly inversely correlated to survival in patients from both cohorts who had complete surgical resection of their tumors (R0) (table S1 and fig. S2).
VAV2 protein abundance was studied by immunohistochemistry (IHC) on tissue microarrays (TMAs) including cases from another distinct ACC German cohort of patients (table S2). VAV2 was weakly to moderately abundant in normal and benign adrenal tissues with no significant differences between the different pathological groups [mean H-score, 1.33 ± 0.51 (in normal adrenals), 1.50 ± 0.54 (in aldosterone- producing adenomas), 1.60 ± 0.54 (in cortisol-producing adenomas), and 1.50 ± 0.57 (in endocrine-inactive adenomas)]. The nonadreno- cortical tissues analyzed showed no to very low VAV2 presence [mean H-score, 1.00 ± 0.00 (for colon, pancreas, and prostate carcinomas) and 0.00 ± 0.00 (for ovarian cancer)] (Fig. 6G). VAV2 abundance was highly variable within the ACC group increasing slightly, but not significantly, from primary to local and distant recurrences (mean
A
B
**
Invading cells (% of ctrl)
250
200
Vehicle
150
100
-
50
0
Dox
Vehicle
Dox
H295R-TR SF-1
H295R-TR SF-1 AF-2 mut
H295R-TR
WT
-
C
D
Air bubble
Cancer cells
2.07
Basement
Upper CAM
membrane
Invasion index
1.5-
1.0-
Lower CAM
0.5
0
Vehicle
Dox
E
F
400
siC
siVAV2
SiC
SİVAV2
SiC
siVAV2
VAV2/ß-tubulin ratio (% of ctrl)
300
kDa
118
200
VAV2
*
100
B-Tubulin
-50
Dox
+
+
0
Vehicle
Dox
G
H
Invading cells (% of ctrl)
500-
*
**
kDa
400-
VAV2
118
300-
200-
B-Tubulin
-50
siC + GFP
siVAV2 + GFP
siVAV2 + WT VAV2
siVAV2 + GEF- inactive VAV2
100-
0
Vehicle
Dox
siC
SÍVAV2
siC + GFP
siVAV2 + GFP
siVAV2 + WT VAV2
SİVAV2 + GEF-
inactive VAV2
H-score, 1.42 ± 1.15, 1.62 ± 1.09, and 1.91 ± 1.08, respectively). SF-1 and VAV2 were significantly correlated also in the German ACC cohort (x2 = 78.11, P < 0.00001) (Fig. 6H). VAV2 abundance was inversely correlated with overall survival (median survival, 26.00 ± 10.17 months versus 62.00 ± 32.62 months for patients with low VAV2; HR, 1.64; 95% CI, 1.01 to 2.66; P = 0.042) and espe- cially with disease-free survival in patients after complete tumor resection (median survival, 9.00 ± 4.3 versus 49.00 ± 25.8, respectively; HR, 3.25; 95% CI, 1.44 to 7.30; P = 0.004) (Fig. 6I and table S2). A multivariate model adjusted for tumor stage, sex, age, and resection status of the tumors confirmed the high prognostic role of VAV2 on ACC patient relapse and overall survival (adjusted HR for relapse, 5.84; 95% CI, 2.19 to 15.59; P < 0.0001; adjusted HR for death, 2.24; 95% CI, 1.21 to 4.13; P = 0.010) (Fig. 6J and table S2). Hormone production did not negatively influence the statis- tics, as also reported in another recent study per- formed on a large cohort of ACC patients (31). VAV2 abundance correlated with Ki67 labeling index (LI), with the values of the latter being sig- nificantly higher in tumors with high VAV2 abun- dance (mean Ki67 LI, 19.41 + 14.97) compared to tumors with low VAV2 abundance (mean Ki67 LI, 7.29 ± 14.85) (P = 0.008) (fig. S3).
DISCUSSION
Deregulated abundance of GEFs has an important role in cancer pathogenesis and spreading (26). Here, we describe a mechanism by which an in- creased dosage of the transcription factor SF-1 in ACC cells induces increased abundance of the multidomain VAV2 GEF, which is a critical driver of cell invasion in adrenocortical cancer cells and is strongly correlated with negative prognosis in ACC patients. The SF-1-related nuclear receptor LRH-1 (also known as NR5A2) has also been linked to the promotion of invasion and metastasis of breast (32) and pancreatic (33) cancer cells, but it is at present unknown whether VAV2 is involved in those phenotypes. The most frequent mechanism of activation of VAV2 in cancer is by phosphoryl- ation of regulatory tyrosine residues by oncogenic kinases (23, 24). It has been shown that in head and neck squamous carcinomas, an autocrine loop involving epidermal growth factor (EGF) receptor is responsible for VAV2 activation (34). In turn, VAV2 slows down receptor internalization and degradation, further increasing EGF signaling (35) in a manner regulated by the Cbl ubiquitin ligase (36). In hematopoietic cells, VAV2 is a target for the homeobox proteins HOXA1 and HOXA9 (37). The role of VAV proteins is multifaceted in cancer. It has been shown that, in addition to being impli- cated in the modulation of small GTPase activity
A
Cochin cohort
C
Cochin cohort
E
Cochin cohort
2
100
Relative expression (log)
1-
VAV2 (log)
Overall survival (%)
VAV2 low (n = 24)
1
80
0
0-
60
40
-1
VAV2 high (n = 23)
-1-
20
-2
0
HR, 4.97; P = 0.002
-2
-1
0
1
2
NR5A1 (log)
-2
0
50
100
150
Time (months)
Low
High
VAV2
B
TCGA cohort
D
TCGA cohort
F
TCGA cohort
14
13-
100
12
12
VAV2 (log)
Relative expression (log)
Overall survival (%)
80
VAV2 low (n = 39)
10
11-
60
8
10
40
VAV2 high (n = 40)
6
9-
20
5
10
15
8-
0
HR, 2.35; P = 0.033
NR5A1 (log)
7-
0
50
100
150
Time (months)
Low
High
VAV2
G
H-score = 0
H-score = 1
Disease-free survival (%)
100
100
80
VAV2 low (n = 22)
Overall survival (%)
80
60
60
VAV2 low (n = 58)
ACC
40
40
H-score = 2
H-score = 3
VAV2 high (n = 18)
20
20
VAV2 high (n = 47)
HR, 3.25; P = 0.004
HR, 1.64; P = 0.042
0
0
20
40
60
0
0
20
40
60
80
100
Time (months)
Time (months)
100-
Colon ca.
Disease-free survival (%)
100
80
VAV2 low (n = 22)
CPA
Overall survival (%)
80
60
60
VAV2 low (n = 41)
40
40
H
20
VAV2 high (n = 18)
20
VAV2 high (n = 33)
SF-1
0123
0
HR, 5.84; P < 0.0001
0-
HR, 2.24; P = 0.01
VAV2
0
20
40
60
0
20 40 60 80 100
Time (months)
Time (months)
Downloaded from http://stke.sciencemag.org/ on March 7, 2017
and cytoskeleton remodeling, VAV2 and VAV3 control a transcrip- tional program in breast cancer cells that is linked to metastasis for- mation in the lung (38). VAV proteins also have an important role in the tumor microenvironment, as shown by experiments demon- strating reduced xenograft growth of lung cancer and melanoma cells in Vav2-Vav3 double null mice (39). Because of their important role in tumor spreading and metastasis, VAV proteins have been proposed as potential therapeutic targets in cancer (21, 22).
In ACC, increased SF-1 abundance is significantly correlated to shorter recurrence-free and overall survival (14, 15). We have shown here a strong correlation between SF-1 and VAV2 expression in tumors (Fig. 6). These patient data together with the results from our cell line model strongly suggest that SF-1 is a major element responsible for increased VAV2 abundance in ACC cases characterized by a more severe prognosis. However, other mechanisms reinforcing VAV2 abundance in ACC may exist. The VAV2 gene is localized in 9q34, a chromosomal region that is frequently amplified in both children and adult ACC (9-11, 40, 41). Furthermore, VAV2 mRNA is a target for miR-195 (42), whose abundance is inversely correlated with malignancy in ACC (43).
One particularly remarkable outcome of our study consists in the highly significant correlation discovered between high VAV2 abun- dance and reduced ACC patient survival. Those results appear par- ticularly robust because they were obtained from three different patient cohorts using different platforms to assay both VAV2 mRNA expression (microarrays and RNA sequencing) and VAV2 protein abundance (IHC). Ki67 LI is currently considered the most reliable immunohistochemical marker to predict clinical outcome in localized ACC after complete surgical resection (31). However, reliable assess- ment of Ki67 in ACC suffers from frequent heterogeneity in the tumor (44) and considerable intra- and interobserver variability (45). In our series of cases, we observed a significant correlation between VAV2 H-score and Ki67 LI. Because our results were obtained using TMA IHC, the VAV2 H-score may represent a more robust prognostic marker than Ki67 LI for use in the clinic. In conclusion, our results show that VAV2 is a critical factor driving the invasive phenotype of ACC cells, which is robustly correlated with prognosis in ACC patients. The recent development of inhibitors of RAC1-VAV2 association (46, 47) has provided the proof of principle that this interaction repre- sents a druggable target in cancer.
MATERIALS AND METHODS Cell culture
Parental H295R-TR cells were cultured in Dulbecco’s modified Eagle’s medium/F12 supplemented with penicillin-streptomycin, 2% NuSerum (BD Biosciences), 1% ITS+ (BD Biosciences), and blasticidin (5 µg/ml; Cayla InvivoGen). Wild-type and AF-2 mutant H295R-TR SF-1 cells (8) were cultured in the same medium with additional zeocin (100 µg/ml; Cayla InvivoGen).
Reverse transcription quantitative polymerase chain reaction
A total of 500 ng of total RNA was reverse-transcribed using Super- script II Reverse Transcriptase (Invitrogen). Reverse transcription quantitative polymerase chain reaction (qPCR) was performed using the SYBR Green I dye assay on a LightCycler 480 (Roche Applied Science) instrument using TATA-binding protein (TBP) as a reference transcript. Primer sequences used were as follows: TBP, 5’-GAACAT-
CATGGATCAGAACAACA-3’ (forward) and 5’-ATTGGTGTTCT- GAATAGGCTGTG-3’ (reverse); VAV2, 5’-CATCAAGGTG GAGGTGCAG-3’ (forward) and 5’-GTACTTGGCCTCGGTC TCCT-3’ (reverse). Results were calculated using the 44Ct threshold cycle method (48).
Immunoblots
Cells were treated for the indicated times with Dox (1 ug/ml; Sigma- Aldrich) or with vehicle (ethanol) as a control. Protein extracts were prepared by harvesting cells in Laemmli buffer [50 mM tris-HCl (pH 6.8), 10% glycerol, 2% sodium dodecyl sulfate (SDS), and 0.02% bro- mophenol blue] containing 5% ß-mercaptoethanol. Extracts for the analysis of phospho-VAV2 (Tyr174) protein levels were prepared by harvesting cells in lysis buffer [1% Triton X-100, 20 mM tris-HCl (pH 7.6), 150 mM NaCl, 5 mM Na3 VO4, 1 mM phenylmethylsulfo- nyl fluoride, 30 mM ß-glycerophosphate, 10 mM NaF, and protease inhibitor cocktail; Calbiochem] at 4℃. Cell lysates were centrifuged at 18,000g for 5 min at 4℃, and the supernatants were immediately processed by adding Laemmli buffer and incubating at 100℃ for 5 min in a dry bath incubator. Proteins were separated by SDS- polyacrylamide gel electrophoresis and transferred to a polyvinylidene fluoride membrane. Immunoblot was performed using a chemi- luminescence system for protein detection (ECL Plus; GE Healthcare). Primary antibodies used were rabbit polyclonal anti-SF-1 (#07-618, Millipore/Upstate Biotechnology), rabbit monoclonal anti-Vav2 (EP1067Y) (#ab52640, Abcam), rabbit monoclonal anti-phospho- VAV2 (Tyr174, EP510Y) (#GTX61721, GeneTex), rabbit monoclonal anti-cofilin (D3F9) (#5175, Cell Signaling Technology), rabbit mono- clonal anti-phospho-cofilin (Ser3, 77G2) (#3313, Cell Signaling Technol- ogy), and mouse monoclonal anti-ß-tubulin (#T4026, Sigma-Aldrich).
Densitometry was performed on scanned immunoblot images using the ImageJ (http://imagej.nih.gov/ij/) gel analysis tool. The gel analysis tool was used to obtain the absolute intensity (AI) for each experimen- tal SF-1 or VAV2 band and corresponding control ß-tubulin band or for each phospho-VAV2 and phospho-cofilin band and corresponding control VAV2 or cofilin band, respectively. Relative intensity for each experimental band was calculated by normalizing the experimental AI to the corresponding control AI.
Immunofluorescence and filopodia and lamellipodia-ruffles detection and quantification
Cells were treated with Dox (1 g/ml) or with ethanol as a control. At the indicated time points, they were fixed (15 min at 22℃) with 4% paraformaldehyde in phosphate-buffered saline (PBS) and per- meabilized by two treatments with 0.1% Triton X-100 in PBS (PT) for 10 min each. After blocking (30 min) in 2% bovine serum albumin in PBS, cells were incubated overnight at 4℃ with rabbit polyclonal anti-SF-1 antibody (#07-618, Millipore/Upstate Biotechnology). Cells were washed three times with PT and incubated for 1 hour at room temperature with Alexa 488-conjugated goat anti-rabbit secondary antibody (1:200; Invitrogen). To visualize F-actin for filopodia and lamellipodia-ruffles formation, cells were incubated for 1 hour with Alexa Fluor 594 phalloidin (1:400; Invitrogen). They were washed again three times with PT and mounted in SlowFade Gold antifade reagent with 4’,6-diamidino-2-phenylindole (DAPI) (Invitrogen). Images for quantification were acquired with a Zeiss Axioplan 2 flu- orescence microscope coupled to a digital charge-coupled device camera, processed, and analyzed using ImageJ. Sampling of cells was performed randomly. More than 250 cells (from two wells) were
scored per condition per experiment to assess the number of filopodia- and lamellipodia-ruffles-forming cells or the average number of filopodia per cell. Filopodia were defined as thin, tubular, fingerlike cell protrusions filled with straight bundled cross-linked actin filaments. Lamellipodia were defined as sheetlike protrusive structures extending from the cell edge and consisting mostly of dynamic, crisscrossed actin filaments, with ruffles being of similar morphology but not adhered and moving centripetally toward the main cell body. All samples were pro- cessed equally and evaluated blindly regarding sample identity. Repre- sentative confocal images shown were acquired using a Zeiss LSM780 inverted confocal microscope system (Carl Zeiss). Fixed cells were ana- lyzed using a 63x oil immersion objective, maintaining the pinhole of the objective at 1 Airy unit. Images were scanned using an Argon 488 laser, a HeNe 542 laser, and a HeNe 633 laser under nonsaturating conditions (pixel fluorescence below 255 arbitrary units) and acquired for all samples using the same settings. SF-1 and VAV2 induction by Dox treatment was, in parallel, assessed by immunoblotting.
CDC42, RAC1, and RhoA biochemical activation assays
H295R TR-SF-1 cells were plated at 1 x 106 per well in a six-well plate. After 24 hours, they were treated with Dox (1 ug/ml) or with ethanol as a control for 12 hours. Cell lysates were harvested on ice and snap- frozen in liquid nitrogen, and the total protein concentration of each lysate was adjusted to 0.25 mg/ml for Cdc42-GTP detection and 0.5 mg/ml for Rac1-GTP and RhoA-GTP detection. Cdc42 activity was analyzed by G-LISA Cdc42 Activation Assay Biochem kit, colorimetric- based (#BK127, Cytoskeleton); Rac1 activity was analyzed by G-LISA Rac1 Activation Assay Biochem kit, luminescence-based (#BK126, Cytoskeleton); RhoA activity was analyzed by G-LISA RhoA Activation Assay Biochem kit, luminescence-based (#BK121, Cytoskeleton) according to the manufacturer’s instructions. SF-1 and VAV2 induction by Dox treatment was, in parallel, assessed by immunoblotting.
FRET analysis of GTPase activation
H295R-TR SF-1 cells were transfected using Lipofectamine 2000 (Invitrogen) with plasmids to express Raichu probes specific for Cdc42, Rac1, and RhoA, respectively (49), treated with vehicle or Dox (1 µg/ml) for 12 hours, and then fixed with formaldehyde. Images for FRET analysis were acquired with a Nikon AIR laser scanning con- focal microscope and a 40x oil objective (numerical aperture, 1.3), as previously described (50). To determine FRET signal by acceptor photobleaching, cyan fluorescent protein (CFP) and yellow fluorescent protein (YFP) channels were excited using the 458-nm and the 514-nm argon line, respectively. Pinholes were opened to give a depth of focus of 2 um for each channel. Scanning was performed on a sequential frame- by-frame basis for each channel. One prebleach image for each channel was collected. Bleaching of the YFP acceptor was obtained with a minimum of 50 iterations of the 514-nm argon laser line at 100% power (bleaching was considered efficient if greater than 70%). A second post- bleach image was then collected for each channel. Twenty cells were analyzed for each condition. To calculate FRET efficiency values, pre- and postbleach CFP and YFP images were imported into Image] for processing (AccPbFRET v3.16 plug-in). Images were background-sub- tracted and fade-compensated (for postbleach images), and a threshold was set to remove saturated pixels before analysis.
Transwell invasion assay through Matrigel
The cell invasion assay was performed with a basement membrane- coated CytoSelect 24-well cell invasion assay kit according to the
manufacturer’s instructions (#CBA-100, Cell Biolabs). Briefly, H295R TR-SF-1 cells (3 x 105 cells per well in complete culture medium) were plated in the upper chamber of the invasion plate. Complete culture medium was added to the lower chamber. In parallel, cells were plated in 24-well plates to check both SF-1 and VAV2 protein induction and the cell number per condition (from triplicate wells). Cells were incu- bated for 72 hours at 37℃ in 5% CO2 atmosphere. The media in the upper and lower chambers were then replaced by fresh culture medi- um containing vehicle or Dox and fresh medium containing 50% fetal calf serum, respectively. Cells were incubated at 37℃ in 5% CO2 atmo- sphere for a further 72 hours. The noninvasive cells were then removed from the upper part of the basement membrane, and the inserts were transferred to a clean well containing 400 ul of Cell Stain Solution and incubated for 10 min at room temperature. The inserts were then washed several times in water, allowed to air dry, and transferred to an empty well, where 200 ul of Extraction Solution was added per well. After 10 min of incubation on an orbital shaker, 100 ul from each sample was transferred to a 96-well microtiter plate, and optical density of 560 nm was measured in a microplate reader. The values obtained were normalized for the corresponding mean number of cells (from triplicate wells).
CAM assay
Work using chick embryos was carried out under animal experimen- tation authorizations nos. 381029 and B3851610001. CAM assay was performed as described previously (29, 30). Briefly, fertilized white Leghorn eggs (SFPA) were incubated at 37.5℃ with 50% relative hu- midity for 9 days. At this time [embryonic day 9 (E9)], the CAM was dropped by drilling a small hole through the eggshell into the air sac, and a 1-cm2 window was cut in the eggshell above the CAM. H295R- TR SF-1 cells, pretreated with vehicle or Dox, were detached with trypsin, washed with complete medium, and resuspended in PBS. Cells (3 x 10°) were loaded onto the CAM of each egg before the eggs were returned to the incubator. At E10, tumors began to be detectable. They were then treated for 10 days, every 2 days, by dropping 100 ml of PBS containing vehicle or Dox (1 µg/ml). At the end of the protocol, tumors were excised and weighed. Tumor cell invasion into the lower CAM was measured by qPCR for human-specific Alu sequences. For each sample, qPCR results were normalized by the weight of the primary tumor.
VAV2 knockdown
For knockdown experiments, H295R-TR SF-1 cells were transfected with VAV2-specific (5’-AGUCCGGUCCAUAGUCAACdTdT-3’) (35) and control (medium GC) siRNA oligos (Invitrogen) using the Amaxa nucleofection technique (Lonza) (20). Cells were electroporated with 80 pmol siRNA per 106 cells using solution R and the T-020 program and then plated in 24-well plates. For the analysis of morpho- logical changes in the actin cytoskeleton, 24 hours after nucleofection, cells were treated with Dox (1 ug/ml) or vehicle alone and, 24 hours later, processed for immunoblotting to measure VAV2 protein abun- dance or for immunofluorescence microscopy to assess the number of filopodia- and lamellipodia-ruffles-forming cells or the average num- ber of filopodia per cell.
For the analysis of cell invasion ability through Matrigel, after elec- troporation, cells were directly plated on the polycarbonate Matrigel- coated inserts in a 24-well plate. In parallel, cells were plated in 24-well plates to check the efficiency of VAV2 mRNA and protein knockdown or induction and the number of cells per condition (from triplicate wells). After 24 hours, cells were treated with Dox (1 ug/ml) and
subjected to invasion assay for 6 days, as described previously. The spectrophotometric readings were normalized for the corresponding mean number of cells (from triplicate wells).
Site-directed mutagenesis and VAV2 knockdown-rescue experiment
The siRNA-insensitive wild-type GFP-VAV2-tagged complementary DNA (cDNA) was obtained introducing three silent mutations into the siRNA target site by site-directed mutagenesis using the QuikChange kit (Stratagene) according to the manufacturer’s instructions. For the L342R-L343S GEF-inactive GFP-VAV2-tagged mutant, mutations were introduced in the siRNA-insensitive wild-type GFP-VAV2 cDNA through site-directed mutagenesis. Mutagenic primers sequences are available upon request.
For VAV2 knockdown-rescue experiments, H295R-TR SF-1 cells were transfected simultaneously with specific siRNA oligos (80 pmol siRNA per 10° cells) directed toward VAV2 and siRNA-insensitive wild-type GFP-VAV2, L342R-L343S GFP-VAV2 mutant, or pEGFP alone (2.5 µg) using the Amaxa nucleofection technique, as previously described above. Control cells were electroporated with control siRNA oligos (80 pmol siRNA per 106 cells) and pEGFP alone (2.5 µg). After electroporation, cells were directly plated on the polycarbonate Matrigel- coated inserts in a 24-well plate. In parallel, cells were plated in 24-well plates to check for efficiency of VAV2 knockdown, wild-type and mutant VAV2 abundance, and the number of cells per condition (from triplicate wells). After 24 hours, cells were treated with Dox (1 µg/ml) and processed as described previously. Spectrophotometric readings were normalized for the corresponding mean number of cells (from triplicate wells).
Immunostaining on formalin-fixed, paraffin-embedded ACC samples
A total of 148 adrenocortical tumor tissues from patients with ACC (n = 133) or benign lesions (n = 15) [aldosterone-producing (n = 6), cortisol-producing (n = 5), and endocrine-inactive adenomas (n = 4)] and six normal adrenal glands (derived from tumor nephrectomies) that were previously assembled in three TMAs were analyzed (51). Among the ACC samples, 105 samples were derived from primary tumors, 16 from local recurrence, and 12 from distant metastasis. Eight nonadrenocortical tumor tissues served as controls. The diagnosis of ACC was made on established criteria based on clinical, biochemical, and morphological data. All clinical data were collected through the German ACC Registry (www.nebennierenkarzinom.ukw.de). Patient data are listed in table S2. All patients gave informed consent, and the study was approved by the ethical committee of the University of Würzburg. Immunohistochemical detection was performed in all samples using an indirect immunoperoxidase technique after high- temperature antigen retrieval in 0.01 M citrate buffer (pH 6.5) in a pressure cooker for 13 min. The primary antibody was a monoclonal rabbit antibody against the VAV2 protein (clone EP1067Y, Abcam) diluted 1:250 in 25% AB serum in PBS and incubated for 1 hour at room temperature. Signal detection was performed with Advance HRP detec- tion system (Dako) and DAB chromogen according to the manufacturer’s instructions. Nuclei were counterstained with Mayer’s hematoxylin for 3 min. As negative control, universal rabbit negative control (Dako) was used. Immunostaining results were analyzed using a light micro- scope at high magnification. VAV2 staining intensity was evaluated independently by two investigators blinded to the clinical data (S.S. and I.S.). The interobserver agreement was very high, with Spearman r = 0.93. In case of discrepancies, a joined score was agreed on by re-
evaluating the slides together. Cytoplasmic staining intensity was eval- uated with a grading score of 0, 1, 2, or 3, which corresponded to negative, weak, moderate, and strong intensity. The proportion of pos- itive tumor cells was calculated for each specimen and scored 0 if 0%, 0.1 if 1 to 9%, 0.5 if 10 to 49%, and 1 if >50% of tumor cells were pos- itive for VAV2. A semiquantitative H-score was then calculated by multiplying the staining intensity grade with the proportion score (14). The cutoff point for separating samples with high or low VAV2 abundance was between the H-scores of 0 to 1 and 2 to 3.
Statistical analysis
Statistical analysis was performed with GraphPad Prism 5.0 software using the Mann-Whitney test, one-way ANOVA with Dunnett’s or Tukey’s correction for multiple testing, and Fisher’s exact test. P < 0.05 was considered to be statistically significant.
For mRNA abundance, two independent ACC cohorts were ana- lyzed: the Cochin cohort included 47 ACC (Gene Expression Omnibus data set GSE49280 and ArrayExpress data set E-TABM-311) (9), and the TCGA cohort included 79 ACC (https://gdc-portal.nci.nih.gov/ projects/TCGA-ACC) (11). Patient data are listed in table S1. For the Cochin cohort, all samples were normalized using the Robust Multi- array Average algorithm (Bioconductor affy package), and probe set intensities were then averaged per gene symbol. For the TCGA cohort, mRNA sequencing data were extracted from Broad Institute GDAC Firehose (TCGA data version 2015_08_21), and all calculations were performed on log2 values of RSEM-normalized read counts. Differential abundance was measured with moderated t test (limma R package). To avoid introducing bias by identification of the “best cutoff,” both cohorts were divided into two equal-size groups according to their VAV2 transcript abundance. Survival curves were obtained by the Kaplan-Meier method. Differences in survival were assessed with the log-rank test. All analyses were performed using R 3.0.3 with custom scripts. Survival analysis for ACC patients in the German cohort was calculated as described previously (14) using the Kaplan-Meier method, and differences between groups were assessed with log-rank and Cox proportional hazards statistics using the SPSS software package (version 15.0.0) after adjustment for sex, age, and tumor stage (20). Overall survival was defined as time elapsed from the primary resection of ACC to death or last follow-up visit. The comparison of clinical and histo- pathological characteristics was performed using the Mann-Whitney test for two groups and the Kruskal-Wallis test for more than two groups of nonparametric variables, as appropriate. Correlation analyses were performed using a x2 test for categorical variables.
SUPPLEMENTARY MATERIALS
www.sciencesignaling.org/cgi/content/full/10/469/eaal2464/DC1
Fig. S1. SF-1 dosage-dependent effects in H295R-TR SF-1 cells.
Fig. S2. High VAV2 abundance is a negative prognostic marker in ACC patients after complete tumor resection.
Fig. S3. Distribution of Ki67 LI in patient tumors grouped by VAV2 abundance.
Table S1. Statistical data of the Cochin and TCGA cohorts of ACC patients.
Table S2. Statistical data of the German cohort of ACC patients.
REFERENCES AND NOTES
1. C. L. Ronchi, M. Kroiss, S. Sbiera, T. Deutschbein, M. Fassnacht, EJE prize 2014: Current and evolving treatment options in adrenocortical carcinoma: Where do we stand and where do we want to go? Eur. J. Endocrinol. 171, R1-R11 (2014).
2. S. G. Creemers, L. J. Hofland, E. Korpershoek, G. J. H. Franssen, F. J. van Kemenade, W. W. Herder, R. A. Feelders, Future directions in the diagnosis and medical treatment of adrenocortical carcinoma. Endocr. Relat. Cancer 23, R43-R69 (2016).
3. M. Fassnacht, M. Terzolo, B. Allolio, E. Baudin, H. Haak, A. Berruti, S. Welin, C. Schade-Brittinger, A. Lacroix, B. Jarzab, H. Sorbye, D. J. Torpy, V. Stepan,
D. E. Schteingart, W. Arlt, M. Kroiss, S. Leboulleux, P. Sperone, A. Sundin, I. Hermsen, S. Hahner, H. S. Willenberg, A. Tabarin, M. Quinkler, C. de la Fouchardière, M. Schlumberger, F. Mantero, D. Weismann, F. Beuschlein, H. Gelderblom, H. Wilmink, M. Sender, M. Edgerly, W. Kenn, T. Fojo, H .- H. Müller, B. Skogseid; FIRM-ACT Study Group, Combination chemotherapy in advanced adrenocortical carcinoma. N. Engl. J. Med. 366, 2189-2197 (2012).
4. A. Berruti, E. Baudin, H. Gelderblom, H. R. Haak, F. Porpiglia, M. Fassnacht, G. Pentheroudakis; ESMO Guidelines Working Group, Adrenal cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann. Oncol. 23, vii131-vii138 (2012).
5. M. Fassnacht, A. Berruti, E. Baudin, M. J. Demeure, J. Gilbert, H. Haak, M. Kroiss, D. I. Quinn, E. Hesseltine, C. L. Ronchi, M. Terzolo, T. K. Choueiri, S. Poondru, T. Fleege, R. Rorig, J. Chen, A. W. Stephens, F. Worden, G. D. Hammer, Linsitinib (OSI-906) versus placebo for patients with locally advanced or metastatic adrenocortical carcinoma: A double-blind, randomised, phase 3 study. Lancet Oncol. 16, 426-435 (2015).
6. F. Tissier, C. Cavard, L. Groussin, K. Perlemoine, G. Fumey, A .- M. Hagneré, F. René-Corail, E. Jullian, G. Gicquel, X. Bertagna, M .- C. Vacher-Lavenu, C. Perret, J. Bertherat, Mutations of B-catenin in adrenocortical tumors: Activation of the Wnt signaling pathway is a frequent event in both benign and malignant adrenocortical tumors. Cancer Res. 65, 7622-7627 (2005).
7. M. A. D. Pianovski, L. R. Cavalli, B. C. Figueiredo, S. C. L. Santos, M. Doghman, R. C. Ribeiro, A. G. Oliveira, E. Michalkiewicz, G. A. Rodrigues, G. Zambetti, B. R. Haddad, E. Lalli, SF-1 overexpression in childhood adrenocortical tumours. Eur. J. Cancer 42, 1040-1043 (2006).
8. M. Doghman, T. Karpova, G. A. Rodrigues, M. Arhatte, J. De Moura, L. R. Cavalli, V. Virolle, P. Barbry, G. P. Zambetti, B. C. Figueiredo, L. L. Heckert, E. Lalli, Increased steroidogenic factor-1 dosage triggers adrenocortical cell proliferation and cancer. Mol. Endocrinol. 21, 2968-2987 (2007).
9. G. Assié, E. Letouzé, M. Fassnacht, A. Jouinot, W. Luscap, O. Barreau, H. Omeiri, S. Rodriguez, K. Perlemoine, F. René-Corail, N. Elarouci, S. Sbiera, M. Kroiss, B. Allolio, J. Waldmann, M. Quinkler, M. Mannelli, F. Mantero, T. Papathomas, R. De Krijger, A. Tabarin, V. Kerlan, E. Baudin, F. Tissier, B. Dousset, L. Groussin, L. Amar, E. Clauser, X. Bertagna, B. Ragazzon, F. Beuschlein, R. Libé, A. de Reyniès, J. Bertherat, Integrated genomic characterization of adrenocortical carcinoma. Nat. Genet. 46, 607-612 (2014).
10. E. M. Pinto, X. Chen, J. Easton, D. Finkelstein, Z. Liu, S. Pounds, C. Rodriguez-Galindo, T. C. Lund, E. R. Mardis, R. K. Wilson, K. Boggs, D. Yergeau, J. Cheng, H. L. Mulder, J. Manne, J. Jenkins, M. J. Mastellaro, B. C. Figueiredo, M. A. Dyer, A. Pappo, J. Zhang, J. R. Downing, R. C. Ribeiro, G. P. Zambetti, Genomic landscape of paediatric adrenocortical tumours. Nat. Commun. 6, 6302 (2015).
11. S. Zheng, A. D. Chemiack, N. Dewal, R. A. Moffitt, L. Danilova, B. A. Murray, A. M. Lerario, T. Else, T. A. Knijnenburg, G. Ciriello, S. Kim, G. Assie, O. Morozova, R. Akbani, J. Shih, K. A. Hoadley, T. K. Choueiri, J. Waldmann, O. Mete, A. G. Robertson, H .- T. Wu, B. J. Raphael, L. Shao, M. Meyerson, M. J. Demeure, F. Beuschlein, A. J. Gill, S. B. Sidhu, M. Q. Almeida, M. C. B. V. Fragoso, L. M. Cope, E. Kebebew, M. A. Habra, T. G. Whitsett, K. J. Bussey, W. E. Rainey, S. L. Asa, J. Bertherat, M. Fassnacht, D. A. Wheeler; Cancer Genome Atlas Research Network, G. D. Hammer, T. J. Giordano, R. G. W. Verhaak, Comprehensive pan-genomic characterization of adrenocortical carcinoma. Cancer Cell 29, 723-736 (2016).
12. E. Lalli, Adrenocortical development and cancer: Focus on SF-1. J. Mol. Endocrinol. 44, 301-307 (2010).
13. P. Latre de Late, A. El Wakil, M. Jarjat, R. R. de Krijger, L. L. Heckert, P. Naquet, E. Lalli, Vanin-1 inactivation antagonizes the development of adrenocortical neoplasia in Sf-1 transgenic mice. Endocrinology 155, 2349-2354 (2014).
14. S. Sbiera, S. Schmull, G. Assie, H .- U. Voelker, L. Kraus, M. Beyer, B. Ragazzon, F. Beuschlein, H. S. Willenberg, S. Hahner, W. Saeger, J. Bertherat, B. Allolio, M. Fassnacht, High diagnostic and prognostic value of steroidogenic factor-1 expression in adrenal tumors. J. Clin. Endocrinol. Metab. 95, E161-E171 (2010).
15. E. Duregon, M. Volante, J. Giorcelli, M. Terzolo, E. Lalli, M. Papotti, Diagnostic and prognostic role of steroidogenic factor 1 in adrenocortical carcinoma: A validation study focusing on clinical and pathological correlates. Hum. Pathol. 44, 822-828 (2013).
16. M. Doghman, J. Cazareth, D. Douguet, F. Madoux, P. Hodder, E. Lalli, Inhibition of adrenocortical carcinoma cell proliferation by steroidogenic factor-1 inverse agonists. J. Clin. Endocrinol. Metab. 94, 2178-2183 (2009).
17. M. Doghman, B. C. Figueiredo, M. Volante, M. Papotti, E. Lalli, Integrative analysis of SF-1 transcription factor dosage impact on genome-wide binding and gene expression regulation. Nucl. Acids Res. 41, 8896-8907 (2013).
18. E. Lalli, M. Doghman, P. Latre de Late, A. El Wakil, I. Mus-Veteau, Beyond steroidogenesis: Novel target genes for SF-1 discovered by genomics. Mol. Cell. Endocrinol. 371, 154-159 (2013).
19. M. Doghman, M. Arhatte, H. Thibout, G. Rodrigues, J. De Moura, S. Grosso, A. N. West, M. Laurent, J .- C. Mas, A. Bongain, G. P. Zambetti, B. C. Figueiredo, P. Auberger, C. Martinerie, E. Lalli, Nephroblastoma overexpressed/cysteine-rich protein 61/connective tissue growth factor/nephroblastoma overexpressed gene-3 (NOV-CCN3), a selective adrenocortical cell proapoptotic factor, is down-regulated in childhood adrenocortical tumors. J. Clin. Endocrinol. Metab. 92, 3253-3260 (2007).
20. M. Doghman-Bouguerra, V. Granatiero, S. Sbiera, I. Sbiera, S. Lacas-Gervais, F. Brau, M. Fassnacht, R. Rizzuto, E. Lalli, FATE1 antagonizes calcium- and drug-induced apoptosis by uncoupling ER and mitochondria. EMBO Rep. 17, 1264-1280 (2016).
21. I. Hornstein, A. Alcover, S. Katzav, Vav proteins, masters of the world of cytoskeleton organization. Cell. Signal. 16, 1-11 (2004).
22. X. R. Bustelo, Vav family exchange factors: An integrated regulatory and functional view. Small GTPases 5, 9 (2014).
23. K. E. Schuebel, N. Movilla, J. L. Rosa, X. R. Bustelo, Phosphorylation-dependent and constitutive activation of Rho proteins by wild-type and oncogenic Vav-2. EMBO J. 17, 6608-6621 (1998).
24. P. Tamás, Z. Solti, P. Bauer, A. Illés, S. Sipeki, A. Bauer, A. Faragó, J. Downward, L. Buday, Mechanism of epidermal growth factor regulation of Vav2, a guanine nucleotide exchange factor for Rac. J. Biol. Chem. 278, 5163-5171 (2003).
25. A. B. Jaffe, A. Hall, Rho GTPases: Biochemistry and biology. Annu. Rev. Cell Dev. Biol. 21, 247-269 (2005).
26. D. Vigil, J. Cherfils, K. L. Rossman, C. J. Der, Ras superfamily GEFs and GAPs: Validated and tractable targets for cancer therapy? Nat. Rev. Cancer 10, 842-857 (2010).
27. A. J. Ridley, Rho GTPase signalling in cell migration. Curr. Opin. Cell Biol. 36, 103-112 (2015).
28. T. Sumi, K. Matsumoto, Y. Takai, T. Nakamura, Cofilin phosphorylation and actin cytoskeletal dynamics regulated by Rho- and Cdc42-activated LIM-kinase 2. J. Cell Biol. 147, 1519-1532 (1999).
29. M. Liu, C. S. Scanlon, R. Banerjee, N. Russo, R. C. Inglehart, A. L. Willis, S. J. Weiss, N. J. D’Silva, The histone methyltransferase EZH2 mediates tumor progression on the chick chorioallantoic membrane assay, a novel model of head and neck squamous cell carcinoma. Transl. Oncol. 6, 273-281 (2013).
30. I. Mikaelian, M. Malek, R. Gadet, J. Viallet, A. Garcia, A. Girard-Gagnepain, C. Hesling, G. Gillet, P. Gonzalo, R. Rimokh, M. Billaud, Genetic and pharmacologic inhibition of mTORC1 promotes EMT by a TGF-B-independent mechanism. Cancer Res. 73, 6621-6631 (2013).
31. F. Beuschlein, J. Weigel, W. Saeger, M. Kroiss, V. Wild, F. Daffara, R. Libé, A. Ardito, A. Al Ghuzlan, M. Quinkler, A. Oßwald, C. L. Ronchi, R. de Krijger, R. A. Feelders, J. Waldmann, H. S. Willenberg, T. Deutschbein, A. Stell, M. Reincke, M. Papotti, E. Baudin, F. Tissier, H. R. Haak, P. Loli, M. Terzolo, B. Allolio, H .- H. Müller, M. Fassnacht, Major prognostic role of Ki67 in localized adrenocortical carcinoma after complete resection. J. Clin. Endocrinol. Metab. 100, 841-849 (2015).
32. A. L. Chand, K. A. Herridge, E. W. Thompson, C. D. Clyne, The orphan nuclear receptor LRH- 1 promotes breast cancer motility and invasion. Endocr. Relat. Cancer 17, 965-975 (2010).
33. Q. Lin, A. Aihara, W. Chung, Y. Li, X. Chen, Z. Huang, S. Weng, R. I. Carlson, C. Nadolny, J. R. Wands, X. Dong, LRH1 promotes pancreatic cancer metastasis. Cancer Lett. 350, 15-24 (2014).
34. V. Patel, H. M. Rosenfeldt, R. Lyons, J .- M. Servitja, X. R. Bustelo, M. Siroff, J. S. Gutkind, Persistent activation of Rac1 in squamous carcinomas of the head and neck: Evidence for an EGFR-Vav2 signaling axis involved in cell invasion. Carcinogenesis 28, 1145-1152 (2007).
35. S. Thalappilly, P. Soubeyran, J. L. Iovanna, N. J. Dusetti, VAV2 regulates epidermal growth factor receptor endocytosis and degradation. Oncogene 29, 2528-2539 (2010).
36. L. Duan, S. M. Raja, G. Chen, S. Virmani, S. H. Williams, R. J. Clubb, C. Mukhopadhyay, M. A. Rainey, G. Ying, M. Dimri, J. Chen, A. L. Reddi, M. Naramura, V. Band, H. Band, Negative regulation of EGFR-Vav2 signaling axis by Cbl ubiquitin ligase controls EGF receptor-mediated epithelial cell adherens junction dynamics and cell migration. J. Biol. Chem. 286, 620-633 (2011).
37. C. Breitinger, E. Maethner, M .- P. Garcia-Cuellar, A. Schambony, K. D. Fischer, K. Schilling, R. K. Slany, HOX genes regulate Rac1 activity in hematopoietic cells through control of Vav2 expression. Leukemia 27, 236-238 (2013).
38. C. Citterio, M. Menacho-Márquez, R. García-Escudero, R. M. Larive, O. Barreiro, F. Sánchez-Madrid, J. M. Paramio, X. R. Bustelo, The Rho exchange factors Vav2 and Vav3 control a lung metastasis-specific transcriptional program in breast cancer cells. Sci. Signal. 5, ra71 (2012).
39. D. M. Brantley-Sieders, G. Zhuang, D. Vaught, T. Freeman, Y. Hwang, D. Hicks, J. Chen, Host deficiency in Vav2/3 guanine nucleotide exchange factors impairs tumor growth, survival, and angiogenesis in vivo. Mol. Cancer Res. 7, 615-623 (2009).
40. E. Letouzé, R. Rosati, H. Komechen, M. Doghman, L. Marisa, C. Flück, R. R. de Krijger, M. M. van Noesel, J .- C. Mas, M. A. D. Pianovski, G. P. Zambetti, B. C. Figueiredo, E. Lalli, SNP array profiling of childhood adrenocortical tumors reveals distinct pathways of tumorigenesis and highlights candidate driver genes. J. Clin. Endocrinol. Metab. 97, E1284-E1293 (2012).
41. O. Barreau, A. de Reynies, H. Wilmot-Roussel, M. Guillaud-Bataille, C. Auzan, F. René-Corail, F. Tissier, B. Dousset, X. Bertagna, J. Bertherat, E. Clauser, G. Assié, Clinical and pathophysiological implications of chromosomal alterations in adrenocortical tumors: An integrated genomic approach. J. Clin. Endocrinol. Metab. 97, E301-E311 (2012).
42. R. Wang, N. Zhao, S. Li, J .- H. Fang, M .- X. Chen, J. Yang, W .- H. Jia, Y. Yuan, S .- M. Zhuang, MicroRNA-195 suppresses angiogenesis and metastasis of hepatocellular carcinoma by inhibiting the expression of VEGF, VAV2, and CDC42. Hepatology 58, 642-653 (2013).
43. P. Singh, P. S. H. Soon, J .- J. Feige, O. Chabre, J. T. Zhao, N. Cherradi, E. Lalli, S. B. Sidhu, Dysregulation of microRNAs in adrenocortical tumors. Mol. Cell. Endocrinol. 351, 118-128 (2012).
44. H. Lu, T. G. Papathomas, D. van Zessen, I. Palli, R. R. de Krijger, P. J. van der Spek, W. N. M. Dinjens, A. P. Stubbs, Automated Selection of Hotspots (ASH): Enhanced automated segmentation and adaptive step finding for Ki67 hotspot detection in adrenal cortical cancer. Diagn. Pathol. 9, 216 (2014).
45. T. G. Papathomas, E. Pucci, T. J. Giordano, H. Lu, E. Duregon, M. Volante, M. Papotti, R. V. Lloyd, A. S. Tischler, F. H. van Nederveen, V. Nose, L. Erickson, O. Mete, S. L. Asa, J. Turchini, A. J. Gill, X. Matias-Guiu, K. Skordilis, T. J. Stephenson, F. Tissier, R. A. Feelders, M. Smid, A. Nigg, E. Korpershoek, P. J. van der Spek, W. N. Dinjens, A. P. Stubbs, R. R. de Krijger, An international Ki67 reproducibility study in adrenal cortical carcinoma. Am. J. Surg. Pathol. 40, 569-576 (2016).
46. B. L. Montalvo-Ortiz, L. Castillo-Pichardo, E. Hernández, T. Humphries-Bickley, A. De La Mota-Peynado, L. A. Cubano, C. P. Vlaar, S. Dharmawardhane, Characterization of EHop-016, novel small molecule inhibitor of Rac GTPase. J. Biol. Chem. 287, 13228-13238 (2012).
47. L. Castillo-Pichardo, T. Humphries-Bickley, C. De La Parra, I. Forestier-Roman, M. Martinez-Ferrer, E. Hernandez, C. Vlaar, Y. Ferrer-Acosta, A. V. Washington, L. A. Cubano, J. Rodriguez-Orengo, S. Dharmawardhane, The Rac inhibitor EHop-016 inhibits mammary tumor growth and metastasis in a nude mouse model. Transl. Oncol. 7, 546-555 (2014).
48. K. J. Livak, T. D. Schmittgen, Analysis of relative gene expression data using real-time quantitative PCR and the 2-AACT method. Methods 25, 402-408 (2001).
49. R. E. Itoh, K. Kurokawa, Y. Ohba, H. Yoshizaki, N. Mochizuki, M. Matsuda, Activation of Rac and Cdc42 video imaged by fluorescent resonance energy transfer-based single-molecule probes in the membrane of living cells. Mol. Cell. Biol. 22, 6582-6591 (2002).
50. E. Pacary, J. Heng, R. Azzarelli, P. Riou, D. Castro, M. Lebel-Potter, C. Parras, D. M. Bell, A. J. Ridley, M. Parsons, F. Guillemot, Proneural transcription factors regulate different steps of cortical neuron migration through Rnd-mediated inhibition of RhoA signaling. Neuron 69, 1069-1084 (2011).
51. C. L. Ronchi, S. Sbiera, L. Kraus, S. Wortmann, S. Johanssen, P. Adam, H. S. Willenberg, S. Hahner, B. Allolio, M. Fassnacht, Expression of excision repair cross complementing group 1 and prognosis in adrenocortical carcinoma patients treated with platinum-based chemotherapy. Endocr. Relat. Cancer 16, 907-918 (2009).
Acknowledgments: We thank L. Buday for the gift of the GFP-VAV2 vector and M. Matsuda for the CDC42, RAC1, and RHOA Raichu probes. Funding: This work was supported by the European Union Seventh Framework Programme (FP7 2007-2013) under grant agreement no. 259735 (ENS@T-CANCER), Association pour la Recherche sur le Cancer grant SFI20111203563, and the French National Research Agency (ANR) through the BeyondTASKs (ANR-11-BSV1- 005-01), LOCALDO (ANR-15-CE14-0017-01), and “Investments for the Future” Labex SIGNALIFE (ANR-11-LABX-0028-01) grants. C.R. was a recipient of ESF-ENS@T and Ville de Nice postdoctoral fellowships. Author contributions: C.R., M.D .- B., M.P., A.M., E.R., and J.F. performed the cell biology and in vivo experiments and analyzed the data. S.S. and I.S. performed IHC and analyzed the data, with the supervision of M.F. B.R. analyzed the genomic data with the supervision of J.B. E.L. analyzed the data, supervised the whole project, and wrote the manuscript together with C.R., receiving inputs from all the other authors. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data and materials are available from the authors upon request.
Submitted 19 October 2016
Accepted 16 December 2016
Published 7 March 2017 10.1126/scisignal.aal2464
Citation: C. Ruggiero, M. Doghman-Bouguerra, S. Sbiera, I. Sbiera, M. Parsons, B. Ragazzon, A. Morin, E. Robidel, J. Favier, J. Bertherat, M. Fassnacht, E. Lalli, Dosage-dependent regulation of VAV2 expression by steroidogenic factor-1 drives adrenocortical carcinoma cell invasion. Sci. Signal. 10, eaal2464 (2017).
Science Signaling NAAAS
Dosage-dependent regulation of VAV2 expression by steroidogenic factor-1 drives adrenocortical carcinoma cell invasion Carmen Ruggiero, Mabrouka Doghman-Bouguerra, Silviu Sbiera, Iuliu Sbiera, Maddy Parsons, Bruno Ragazzon, Aurélie Morin, Estelle Robidel, Judith Favier, Jérôme Bertherat, Martin Fassnacht and Enzo Lalli (March 7, 2017) Science Signaling 10 (469), . [doi: 10.1126/scisignal.aal2464]
The following resources related to this article are available online at http://stke.sciencemag.org. This information is current as of March 7, 2017.
| Article Tools | Visit the online version of this article to access the personalization and article tools: http://stke.sciencemag.org/content/10/469/eaal2464 |
| Supplemental Materials | "Supplementary Materials" http://stke.sciencemag.org/content/suppl/2017/03/03/10.469.eaal2464.DC1 |
| Related Content | The editors suggest related resources on Science's sites: http://stke.sciencemag.org/content/sigtrans/5/244/ra71.full http://stke.sciencemag.org/content/sigtrans/5/244/pe43.full |
| References | This article cites 51 articles, 19 of which you can access for free at: http://stke.sciencemag.org/content/10/469/eaal2464#BIBL |
| Permissions | Obtain information about reproducing this article: http://www.sciencemag.org/about/permissions.dtl |
Science Signaling (ISSN 1937-9145) is published weekly, except the last December, by the American Association for the Advancement of Science, 1200 New York Avenue, NW, Washington, DC 20005. Copyright 2017 by the American Association for the Advancement of Science; all rights reserved.