FOXO1a acts as a selective tumor suppressor in alveolar rhabdomyosarcoma

Abstract

Rhabdomyosarcoma (RMS), the most common pediatric soft-tissue sarcoma, has two major histological subtypes: embryonal RMS (ERMS), which has a favorable prognosis, and alveolar RMS (ARMS), which has a poor outcome. Although both forms of RMS express muscle cell-specific markers, only ARMS cells express PAX3-FOXO1a or PAX7-FOXO1a chimeric proteins. In mice, Pax3 and Pax7 play key roles in muscle cell development and differentiation, and FoxO1a regulates myoblast differentiation and fusion; thus, the aberrant regulation of these proteins may contribute to the development of ARMS. In this paper, we report that FOXO1a is not expressed in primary ARMS tumors or ARMS-derived tumor cell lines and that restoration of FOXO1a expression in ARMS cells is sufficient to induce cell cycle arrest and apoptosis. Strikingly, the effects of FOXO1a are selective, as enforced expression of FOXO1a in ERMS-derived tumor cell lines had no effect. Furthermore, FOXO1a induced apoptosis in ARMS by directly activating the transcription of caspase-3. We conclude that FOXO1a is a potent and specific tumor suppressor in ARMS, suggesting that agents that restore or augment FOXO1a activity may be effective as ARMS therapeutics.

Figure 1.

FOXO proteins are not expressed in ARMS. (A) Immunoblot analyses of FOXO1a, FOXO3a, and FOXO4 proteins were determined in samples of primary ARMS (t[2;13] only) and ERMS tumors. The expression of these proteins was also evaluated in cell lines derived from ERMS (Rh2, Rh6, JR1, and RD) and ARMS (Rh3, Rh4, Rh30, and Rh41; t[2;13] only) tumors. The expression of the muscle cell–specific markers myogenin and MyoHC is also shown. FOXO proteins were not expressed in ARMS tumors and in all but one ARMS-derived cell line; however, all FOXO proteins were expressed in ERMS tumors and cell lines. Fast Green dye was used to confirm equal loading and homogeneous protein transfer. (B) A Western blot analysis of the indicated proteins in differentiating C2C12 cells, over a 3-d time period, is shown. Uniform protein loading and homogenous transfer of SDS-PAGE gels were confirmed by staining transferred membranes with Fast Green.

Figure 2.

FOXO1a localization in ARMS and ERMS. Endogenous levels of FOXO1a were not detectable in RMS cells by immunofluorescence with the FOXO1a-specific antibody (αFOXO1a), but enforced expression of FOXO1a-WT or FOXO1a-TM demonstrated nuclear localization of FOXO1a in ARMS and ERMS. Note the nuclear shrinkage, a hallmark of apoptosis, in Rh3, Rh4, and Rh41 ARMS cells overexpressing FOXO1a or FOXO1a-TM proteins, as well as elongation of Rh30 cells expressing FOXO1a. In contrast, the two ERMS cell lines show little change in their morphology after overexpression of FOXO1a. DAPI staining was used to stain DNA. Bar, 50 μm.

Figure 3.

FOXO1a transcriptional activity is impaired in ARMS. (A) Transcriptional activity of FOXO1a was analyzed in JR1 and RD ERMS cells. The pGL3 basic luciferase reporter showed little activity with (light gray bars) or without (white bars) exogenous FOXO1a-WT. However, a reporter construct containing six FOXO DNA binding sites (dark gray bars) was robustly induced by FOXO1a-WT (black bars). (B) In contrast, ARMS cell lines show little induction in absolute luciferase activity by FOXO1a-WT. Note the much lower absolute luciferase activity scale in ARMS than in ERMS. Values are the means of at least triplicate samples, and error bars depict the standard deviation. All values were normalized using β-actin promoter–driven SEAP as a transfection efficiency control. Fold increases between the reporter construct containing six FOXO DNA binding sites expressing exogenous FOXO1a-WT versus empty vector are indicated.

Figure 4.

Suppression of FOXO1a expression in ARMS. (A) Real-time RT-PCR analysis of FOXO1a transcripts in ERMS and ARMS cell lines demonstrated that FOXO1a is suppressed in ARMS, and at levels below what one would expect from a haploid gene dose of FOXO1a. Relative levels of FOXO1a transcripts were normalized to those of glyceraldehyde-3-phosphate dehydrogenase transcripts using an arbitrary unit of measure. Values are the means of triplicate samples, and error bars depict variations between samples. (B) Immunoblot analyses revealed that high levels of AKT are expressed in ARMS and ERMS cells. However, AKT is inactive when compared with a PTEN knock-out cell line control, as indicated by the lack of phosphorylated AKT (P-AKT^Thr308 and P-AKT^Ser473). Equal loading and homogeneous protein transfer were confirmed by staining blots with Fast Green dye. (C) FOXO1a protein is suppressed in ARMS through a proteasome-dependent pathway. The indicated ARMS cells were treated for 8 h with the proteasome inhibitor MG132, and endogenous levels of FOXO1a were determined by immunoblot analyses.

Figure 5.

Restoration of FOXO1a expression in ARMS cells induces apoptosis. (A) A FACS analysis of GFP control and FOXO1a-IRES-GFP–transduced ERMS and ARMS cell lines. The fine line shows nontransduced negative control cells, the dashed line displays GFP-only transduced cells, and the bold line indicates FOXO1a-IRES-GFP–transduced cells. The identity and percentages of GFP- and FOXO1a-IRES-GFP–positive cells is given for each cell line. Note the low percentage of FOXO1a-positive Rh3 cells, which is attributable to their rapid death. Only JR1 and RD ERMS lines could be FACS sorted for FOXO1a-IRES-GFP and expanded in culture, whereas all the FOXO1a-IRES-GFP–transduced ARMS cells either died (Rh3, Rh4, and Rh41) or failed to expand (Rh30). (B) The apoptotic index of ERMS and ARMS cells engineered to express FOXO1a (gray bars), FOXO1a-TM (black bars), or vector alone (white bars) was augmented in ARMS cells expressing FOXO1a, whereas the apoptotic index of ERMS cells expressing FOXO1a was unaffected. No significant differences were observed between ARMS or ERMS cells overexpressing FOXO1a-WT versus FOXO1a-TM. Values are the means of triplicate samples, and error bars depict SD.

Figure 6.

FOXO1a selectively induces G ₂ /M cell cycle arrest in ARMS cells. (A) FOXO1a (gray bars) and FOXO1a-TM (black bars) expression induced G₂/M phase cell cycle arrest in ARMS cells. Vector-only expressing ARMS cells are denoted by white bars. No cycle arrest was observed in ERMS lines. Values are the means of triplicate samples, and error bars depict variations between samples. (B) No morphological changes were observed in the JR1 and RD ERMS cell lines expressing FOXO1a-TM. GFP expression was overlaid to confirm that the cells had been transduced. Identical results were obtained with FoxO1a-WT (not depicted). (C) Typical morphological changes induced by FOXO1a in Rh4 and Rh30 ARMS cells are shown. Note the alignment of cells and the presence of multinucleated cells in the Rh30 cells expressing FOXO1a-TM. Rh3, Rh4, and Rh41 all undergo rampant apoptosis after FOXO1a-TM expression. GFP expression was overlaid to confirm that the cells had been transduced and expressed the MSCV-FOXO1a-TM-IRES-GFP retroviral vector. Identical pictograms were obtained with FoxO1a-WT (not depicted). Bars, 50 μm.

Figure 7.

FOXO1a activates caspase-3 transcription in RMS, but caspase-3 is only activated in ARMS. (A) A real-time RT-PCR analysis of caspase-3 transcripts in ERMS and ARMS cell lines in the absence or presence of FOXO1a-TM. Transcript levels were normalized to those of glyceraldehyde-3-phosphate dehydrogenase mRNA transcripts using an arbitrary unit of measure. Fold increases between GFP control–transduced (white bars) and FOXO1a-TM–transduced cells (black bars) are indicated. Values are the means of triplicate samples, and error bars depict variations between samples. (B) The indicated ARMS and ERMS cells were transduced with the control virus (MSCV-IRES-GFP) or a virus expressing FOXO1a-TM (MSCV-FOXO1a-TM-IRES-GFP). After 24–72 h, the levels of FOXO1a, total pro-Casp3 (Casp3), activated Casp3 (p20 and p17), and actin were assessed in cell extracts by immunoblot. Only low levels of FOXO1a-TM were detectable in Rh3 and Rh41 cells because of the rapid apoptosis of these cells, as indicated by high levels of activated Casp3. (C) Casp3 activity was assessed in the indicated cells 24 (Rh3 and Rh41 cells) or 72 h (Rh4, Rh30, RD, and JR1 cells) after transduction with the control virus (GFP only) or a virus expressing FOXO1a-TM. Caspase-3 activity was normalized to protein content, and fold increases in caspase-3 activity between GFP-only control (white bars) and FOXO1a-TM–transduced cells (black bars) are indicated using an arbitrary unit of measure. Values are the means of triplicate samples, and error bars depict variations between samples.

Figure 8.

FOXO1a directly binds the Casp3 promoter region upon primary myoblast differentiation. (A) Direct binding of FLAG-FOXO1a to Casp3 promoter/enhancer regions in transduced mouse primary myoblasts as assessed by ChIP assay. As cells differentiated, FOXO1a showed rapid binding to the Casp3 promoter/enhancer regions at −3.8/−3.3 and 11.0/11.5 kb, which contain FOXO1a binding sites. Cells transduced with an empty MSCV-IRES-GFP vector were used as a negative control. (B) The induction of differentiation of C2C12 cells increased promoter–reporter luciferase activity of constructs containing the FoxO1a binding regions at −3.8/−3.3 kb upstream of the 5′ UTR of mouse Casp3. C2C12 cells were transiently transfected with the Casp3 −3.8/−3.3 kb luciferase construct alone (closed circles), with wild-type FOXO1a-WT (open circles), with the dominant-negative FOXO1aΔTA (closed squares), or with the −3.8/−3.3 kb promoter region deleted for the FOXO binding sites (open squares), and differentiation was then induced. Values are the means of triplicate samples, and error bars depict variations between samples. (C) Identical to B but with a reporter construct containing the FoxO1a binding regions at +1.0/+1.5 kb downstream of the 5′ UTR of mouse Casp3. Note the 10-fold higher induction of this reporter construct compared with that of the construct in B.

Figure 9.

FOXO1a induces cell cycle arrest in ARMS Rh30 cells. (A) Rh30 cells expressing vector alone (circles) or FOXO1a-TM-ER^TAM (triangles) were treated with tamoxifen (4OH; filled symbols) for the indicated intervals. The cell cycle profiles showed that 4OH treatment arrested ARMS cells expressing the FOXO1a-TM vector; the growth of the cells that expressed vector alone was not affected. Values are the means of triplicate samples, and error bars depict variations between samples. (B) Representative morphological changes observed in Rh30 cell lines after 4OH treatment induced FOXO1a-TM-ER^TAM activity. (C) Immunoblot analyses showed that long-term culture of FOXO1a-TM-ER^TAM–expressing Rh30 cells in a medium containing tamoxifen selected for cells that lost FOXO1a-TM-ER^TAM expression. Cells transduced with pBabe were used as the control, and equal loading of the blot was confirmed by staining with Fast Green dye. (D) FOXO1a-TM-ER^TAM expression in Rh30-derived tumors in NOD/SCID mice. Rh30-derived tumors expressing either pBabe vector alone or FOXO1a-TM-ER^TAM with or without 4OH were stained with an anti-FOXO1a antibody. DAPI staining was used to calculate the cell number and density. (E) Cell proliferation was substantially reduced after 4OH treatment in Rh30-derived tumors expressing FoxO1a-TM-ER^TAM _. NOD/SCID mice bearing vector alone or FOXO1a-TM-ER^TAM–expressing tumors, with or without 4OH treatment, were injected with BrdU, and tumor sections were stained by indirect immunofluorescence with an anti-BrdU antibody. DAPI staining was used to calculate the cell number and density. Bars, 100 μm.

Figure 10.

FOXO1a activation induces regression of Rh30-derived tumors in NOD/SCID mice. (A) After 5 wk of growth, NOD/SCID mice bearing Rh30-derived tumors that expressed vector alone (n = 2; closed squares) or FOXO1a-TM-ER^TAM (n = 5; open triangles) were treated with tamoxifen, and tumor sizes were determined at the indicated intervals after treatment. FOXO1a-TM-ER^TAM–expressing tumors substantially decreased in size after tamoxifen treatment, whereas vector-only–expressing tumors continued to grow. (B) Representative TUNEL assays performed on sections of untreated Rh30-derived tumors from NOD/SCID mice. DAPI staining was used to calculate cell number and density. (C) Representative TUNEL assay after 9 d of tamoxifen treatment. Note the increase in the number of TUNEL-positive cells in tumors expressing FOXO1a-TM-ER^TAM compared with that seen in B. (D) Active Casp3 in sections of untreated Rh30-derived tumors expressing either vector only or FoxO1a-TM-ER^TAM in NOD/SCID mice. (E) Active Casp3 in sections of Rh30-derived tumors after tamoxifen treatment. Note the increase in the number of cells containing active Casp3 in tumors expressing FOXO1a-TM-ER^TAM compared with that in D. Bars, 100 μm.