Determining the Function and Effects of a Novel FOXO1 Fusion

Tanner Wherley

A thesis presented to the Faculty at Cincinnati Children’s Medical Center in partial fulfillment of the requirements for the degree of Master of Science in the Graduate School of Arts and Sciences

Committee Chair: Rashmi Hedge, PhD

Biomedical Research Technologies Program College of Medicine University of Cincinnati

May 15th, 2020

Abstract: Rhabdomyosarcoma (RMS) is a heterogenous group of soft-tissue tumors with many underlying causes varying patient to patient. The two most encountered histological classifications, Alveolar and Embryonal, are driven by molecularly distinct mechanisms with an unknown fusion gene status in the later. Recent sequencing identified a novel fusion gene between FOXO1 and the Long Intergenic Non-coding RNA 00598 in an Embryonal RMS patient. Here we use C2C12 mouse myoblasts to investigate the resulting FOXO1-LINC fusion (FOXO1- LINC) and its role in the development of Embryonal RMS. We show that the FOXO1-LINC protein inhibits terminal differentiation of mouse myoblasts and disrupts the fusion of nuclei. The FOXO1-LINC protein acts as a dominant negative on FOXO1 function, blocking the activation of FOXO1 responsive luciferase production and target gene activation. Although expression of the FOXO1-LINC was not fully transformative, MyoG expression was inhibited during differentiation and the genotype/phenotype was mimicked using a constitutively active transcriptional co-activator with PDZ-binding motif (Taz) construct, suggesting potential roles for the fusion gene in the development of ERMS.

ii

iii

Table of Contents: Abstract……………………………………...... ….....….…..ii Table of Contents……………………………...... ……....…...iv Introduction………………………………...... …...... v Experimental Design and Protocols………...... vi Results…...... vii Conclusion...... x Statement of Work ...... xi Bibliography ...... xii Supplemental Figures ...... xv

iv

Introduction: RMS is the most common soft-tissue sarcoma of children and young adults (1). Histologically, the two most common subtypes, Alveolar and Embryonal (ARMS and ERMS), display a variable degree of skeletal muscle differentiation (2). Morphologically, ARMS tumors are characterized by ‘primitive’, compact, small round blue cancer cells adhered to the outsides of alveoli like structures. Harder to distinguish pathologically, ERMS often resembles the developing embryonic muscle with variable cell density and expanded eosinophilic cytoplasm (3,4). In low to intermediate risk patients, surgical removal and chemotherapy regimens, such as vincristine and actinomycin-D, have contributed to a slow climb in overall survival, but the 5-year survival for patients with metastatic or recurrent tumors remains grim (5,6). Determining the oncogenic drivers and aberrant biological cellular pathways hijacked in these tumors is vital to develop targeted therapies for patients present and future, reducing chemotherapy toxicity and providing hope for patients with recurrent or metastatic growth (7).

With the need of more accurate classifications reflecting patient prognosis and therapeutic strategies, classification parameters are shifting to also reflect the underlying tumor genomic landscape as Fusion Positive (FP) or Fusion Negative (FN) (2,8). Distinct to FP ARMS is a t(2;13) or t(1;13) PAX3/7-FOXO1 chimeric fusion protein, with few other PAX fusion artifacts noted (9,10). This chromosomal translocation fuses the DNA binding domain (DBD) of PAX with the potent transactivating domain (TAD) of FOXO1. Expression of a PAX3/7 fusion protein in combination with other genetic alterations drive oncogenicity in FP ARMS (11-13). A small subset of FN ARMS cases also exist, histologically diagnosed as ARMS, but lacking a PAX fusion gene (9,14). However, clinical analysis and microarray data reveal similar expression profiles between FN ARMS and FN ERMS (15,16). FN ERMS is known to be caused by many genetic alterations such as the loss of heterozygosity at 11.p15.5 and various mutations affecting receptor tyrosine kinase (RTK)/RAS/PIK3CA pathways (17,18). Despite these classification efforts, studying FP ERMS remains important as it is still poorly understood due to the small number of fusion identified in ERMS with different expressional profiles (19-22). For this report, ARMS and ERMS will refer to the histological patterning and diagnosis and FP/FN status will reflect the underlying genomic landscape affecting the tumor.

The expression of muscle specific markers indicate myogenic progenitors are the cells of origin for RMS, however other mesenchymal origins are reported (23-26). Myogenesis is the process of generating muscle during development or in response to injury and is often recapitulated in-vitro using an established primary mouse myoblast line, C2C12 (27,28). These satellite-cell derived myoblasts rapidly proliferate in a nutrient rich environment, however differentiate into mature myotubules when serum starved (28,29). Directing this process, MYOD1 acts as the master regulator of skeletal muscle differentiation, activating the expression of muscle specific genes such as MYOG to induce cell cycle arrest and confirm terminal differentiation (30). In the development of RMS, while MYOD1 is still able to bind DNA, complete activation of its downstream targets is disrupted, suggesting the inhibition of important MYOD1 cofactors (31-33). For this reason, MYOD1 and its target genes are useful markers to confirm the muscular cell lineage and gage the degree of myogenesis in RMS (34,35).

Initially discovered with the PAX3/7-FOXO1 fusion protein (11), the FOXO1 gene encodes a transcription factor acting downstream of many signaling pathways in response to starvation and stress (36). Mechanistically, this protein binds DNA as a monomer until phosphorylated at AKT phosphorylation sites (T24, S256, S319), where it then binds 14-3-3 for nuclear removal (37,38). Early in-vivo studies revealed that FOXO1 is involved in muscle fiber type specification during embryogenesis, however its exact role was found to be conflicting (39,40). Muscle specific triple knockout of three Forkhead box O proteins 1, 3 and 4 confirm that FOXO factors play essential roles in regulation of autophagy and the ubiquitin-proteasome system in mature muscle (41). Nonetheless, FOXO1 may play a role in myogenic growth and differentiation with unidentified roles.

Here we present preliminary and supportive in-vitro data for a novel fusion gene acting as a driver of ERMS. An 80kB deletion of 13 results in the fusion of FOXO1 Exon 1 (E1) with Exon 5 (E5) of the Long Intergenic Non-coding RNA 00598. Determining the function and effects of the FOXO1-LINC fusion gene will help establish correct therapeutic strategies as preliminary data indicates this fusion gene is conserved in other soft-tissue sarcomas with an unidentified role.

v

Experimental Design and Protocols: Cell Culture- Both Human Embryonic Kidney (HEK) 293 and mouse C2C12 myoblast cells were maintained in Delbecco’s Modified Eagles Medium (DMEM, Gilbco) supplemented with 10% Fetal Bovine Serum (FBS) and 1:100 antibiotic-antimycotic (10,000 units/mL of penicillin, 10,000 µg/mL of streptomycin, 5 µg/mL of Gibco Amphotericin B). Cells were passaged at 90-95% confluency in a 10-cm dish (~7.5x106 cells total, depending on the cell line used), and maintained under a passage number of 18. Differentiating C2C12 were cultured in DMEM with 2% Horse Serum (2%HS) and anti-anti for the indicated number of days. Cells were incubated at 37˚C in 5% CO2. Media was changed after the first day of 2%HS treatment and then every two days.

Plasmids- PiggyBac (PB) (System Biosciences) constructs were under a CAG promoter with a cytomegalovirus (CMV) enhancer, an internal ribosomal entry sight (IRES) and enhanced green fluorescent protein (eGFP). The PB-CAG-eGFP-IRES was used as the control (Ctl) and inserts were cloned into the PB-CAG-eGFP-IRES backbone using blunt end ligation. The PB-Flag-FOXO1-LINC-eGFP contains an N-terminal Flag Tag (DYKDDDDK) used for tracking the localization and immunoblotting (IB) for expression levels. The PB-Flag- caFOXO1 was created by subcloning the insert from a pcDNA3-Flag-FKHR-AAA mutant (Addgene plasmid #13508). The 3x Insulin Response Sequence (3xIRS) luciferase plasmid (3xIRS-luc) was in a pGL-2 promoter vector with a simian virus 40 (SV40) promoter (Addgene plasmid #13511). The PB-Flag-Taz-S4A-eGFP construct was given by Xiaohua Hu from Richard Lu’s laboratory.

Stable Cell Line Generation- Transfection of C2C12 was difficult, often resulting in less than 20% eGFP+, so we used the PB vector delivery system to create long term stable construct expressing C2C12 cell lines. Co- expression of the transposase allowed recognition of the 5’ and 3’ inverted terminal repeats (ITR) and incorporation of the insert into TTAA chromosomal sights at random. C2C12 cells were co-transfected using FuGENE HD (Promega) with a 1ug:2ug transposon to transposase DNA ratio in a 12-well dish at 65% confluent. FuGENE HD was used at a 1:3 DNA to FuGENE HD ratio. Media was changed 24 hours later and cells were passaged at a 1:5 ratio 48 hours after transfection to expand the eGFP+ population and ensure stable integration. After the first passage, cells were grown to confluent and suspended in Trypsin-EDTA (0.05%) (Gibco). The suspension was diluted accordingly and 30 cells/well were plated in column 1 and 7 of a 96-well plate. Serial dilutions were made from the first and seventh column with a multichannel pipette to obtain a single eGFP+ cell growing in a well. Plates were checked for single colony formation and eGFP expression 3 days post-plating. The single cell colony was then expanded to obtain stable lines from a single cell. The lipid nanoparticle delivery method Lipofectamine3000 (Thermo Fisher) was used, but yielded a lower transfection efficiency relative to FuGENE HD. A limitation to using the PiggyBac method was that genomic incorporation was random, therefore multiple lines were used after verification of construct expression.

Immunostaining- Cells were grown on coverslips, fixed with a 4% paraformaldehyde solution in Phosphate- buffered saline (PBS) for 15 minutes, and washed/permeabilized with PBS+0.1%Tween20 (PBST). Blocking was done for 30 minutes in a PBST+1%BSA solution and primary antibodies were diluted 1:200 in blocking buffer and incubated overnight at 4˚C. Coverslips were washed and incubated with a respective 1:200 Alexa Fluor (Thermo Fisher) secondary antibody in a 1:1000 DAPI solution for 1.5 hours. Coverslips were washed again and mounted to slides using Flouromount-G mounting media (SouthernBiotech). Images were analyzed and counted with ImageJ using threshold adjustment and binary pixel detection. Images were counted automatically and once again manually to verify accuracy.

Luciferase- HEK293 cells were co-transfected with 150ng pGL-2-promoter 3xIRS-luc construct, 200ng of the different PiggyBac constructs, and 5ng of the CMV-lacZ construct to drive -gal expression as a control. Luciferase was read 48 hours after transfection. Briefly, cells were lysed with a Glo Lysis Buffer (Promega) and the Bright-Glo assay reagent was added to detect luminescence with a Flextation 3 Multi-Mode Microplate Reader. The FluoReporter lacZ/Galactosidase kit (Molecular Probes) was used to detect the lacZ fluorescence production excited at 390 nm and read at 460 nm. Results were processed as luciferase relative to lacZ for each condition. A luciferase assay using C2C12 cells with differentiation media would have been optimal but due to the low transfection efficacy, HEK293 cells were used.

Real Time qPCR- RNA was isolated from cells in 12-well plates using TRIzol (ambion) solubilization and extraction as described (42). cDNA was made from RNA using PrimeScript (Takara). 10ng of cDNA was then mixed with the forward/reverse primers (Supp. Table 1) at 500nM and PowerUp SYBR Green Master Mix

vi

(appliedbiosystems). Thermal cycling was done as follows: 2 mins at 55˚C, 2 mins at 95˚C, and then 40 cycles of 95˚C for 5 sec and 60˚C for 20 sec. 18s mRNA was used as an internal control and for normalization.

Western Blotting- Protein was extracted from cells in 6-well plates using a 2x Laemmli sample buffer (Bio-Rad) with 5% 2-mercaptoethanol. Samples were ran on 6-12% polyacrylamide gels with running buffer (25mM Tris, 190 mM glycine, 0.1% SDS) and transferred with transfer buffer (25mM Tris, 190mM glycine, 20% methanol) to a polyvinylidene difluoride membrane for 1.5-3 hrs on ice. Membranes were blocked at least 30 minutes with a 5% dry milk solution made with TBST (20mM Tris, pH 7.5, 150mM NaCl, 0.1% Tween20). Primary antibodies were incubated overnight at 4˚C or for 1hr at room temperature (Supp. Table 2). HRP linked secondary antibodies were then incubated for 1.5 hrs and detected at 425 nm using chemiluminescence with a ChemiDoc MP imaging System (Bio-Rad).

Statistical Measurements- An independent student t-test was used to determine the significance for all experimental results shown. * p<.05, ** p<.01, *** p<.001.

Results: The FOXO1-LINC fusion protein inhibits terminal differentiation of mouse myoblasts. In 2016, Next-generation and Whole-genome sequencing done by Foundation Medicine identified a novel interstitial chromosomal deletion on chromosome 13 in an ERMS tumor of an 8-year old male patient. The 80kb interstitial deletion resulted in a chromosomal fusion between E1 of FOXO1 and E5 of a Long Intergenic Non- coding RNA 00598 (LINC00598) (Fig. 1A). The resulting mRNA sequence contains an open reading frame starting at the natural start codon in FOXO1 E1 and ending at the first in-frame stop codon, found 8 amino acids (aa) into LINC00598 E5. Previous data indicates that translation starts and stops as shown, and the resulting product is a 32kDa fusion protein termed FOXO1-LINC (Fig. 1B).

Figure 1. FOXO1-LINC inhibits terminal differentiation of C2C12. (A) Ideograph (NCBI) and chromosomal landscape of Wild-Type (WT) and patient, showing ~80kb chromosomal deletion of Ch.13 and the fusion of FOXO1 E1 to LINC00598 E5. (B) Established mRNA strand, showing the open reading frame from start codon (green) in FOXO1 E1 to the first stop codon in LINC00598 (red), and the resulting 32kDa product. (C) Diagram of the FOXO1-LINC insert in the PB vector. (D) Western blot of Control (Ctl) line 1 and FOXO1-LINC lines 1 and 2 showing MyHC, Flag and Gapdh expression at D0 and D5 of differentiation. (E) MyHC immunostaining of Ctl and FOXO1-LINC cells after D5 of differentiation. (F) Quantification of the fusion index (fused nuclei/total nuclei) at D5 of differentiation. An independent student t-test was used to determine significance (N=3, three images of three coverslips differentiated). * p<.05, ** p<.01, *** p<.001.

vii

We hypothesized expression of FOXO1-LINC will affect the normal differentiation of C2C12 mouse myoblasts. We first developed single cell stable lines expressing FOXO1-LINC by chromosomal integration, enabling constant long-term expression (Fig. 1C). Expression was validated with eGFP expression and immunoblotting (IB) for the N-terminal Flag-Tag sequence on the FOXO1-LINC (Supp. Fig. 1A,B). FOXO1-LINC lines grew nearly identical to the control (Ctl) in a nutrient rich condition (data not shown). We then differentiated the FOXO1-LINC lines by switching the media to 2%HS and culturing for five days (D5). Protein was isolated and SDS-PAGE was used to detect the expression of Myosin Heavy Chain (MyHC) as a marker of terminal differentiation. By D5, MyHC expression in the FOXO1-LINC lines were markedly reduced compared with control lines and densitometry confirmed a significant fold reduction in relative MyHC expression. Lines that expressed higher levels of FOXO1-LINC showed lower expression of MyHC. FOXO1-LINC was expressed higher at D5 (Fig. 1D, Supp. Fig. 1C,D). We then differentiated representative Ctl and FOXO1-LINC stable lines to detect MyHC by immunofluorescence. Consistent with the western data, FOXO1-LINC lines displayed less MyHC expression, and cells that expressed MyHC were bundled and shortened (Fig. 1E, Supp. Fig. 1E). The Fusion Index, indicating the proportion of nuclei in fused myotubules is less in the FOXO1-LINC line (Fig. 1F). Taken together, these results indicate that stable expression of FOXO1-LINC inhibits terminal differentiation of mouse myoblasts.

Expression of FOXO1-LINC disrupts the function of FOXO1 in vitro The FOXO family of transcription factors share conserved domains important for their biological function. Comparing the amino acid sequences between WT FOXO1 and the FOXO1-LINC fusion protein, 75% of the original FOXO1 DNA binding domain is conserved (Supp. Fig. 2A,B), however it lacks the nuclear localization/enhancer sequences (NLS/NES) and the transactivating domain (TAD) (Fig. 2A).

Figure 2. FOXO1-LINC inhibits the function of FOXO1 in C2C12 and HEK293 cells. (A) Diagram of FOXO1 compared with FOXO1-LINC protein. AKT regulation sights T24, S256 and S319 are shown. (B) Immunostaining of FoxO1-C-terminal to detect endogenous FoxO1 localization at 16hr of differentiation. Arrows denote nuclear FoxO1 seen throughout staining. (C) Expression of FoxO1 target genes at 24 hours differentiation. Data shown is relative to 18s internal control. (D) 3xIRS-luc activity in HEK293. Cells were transfected with indicated constructs in 10%FBS and luciferase was measured 48 hours later. An independent student t-test was used to determine significance (N=3 independent samples for RT qPCR and luciferase). * p<.05, ** p<.01, *** p<.001.

Based off this observation, we then hypothesized the FOXO1-LINC would affect the function of FoxO1. FoxO1 is dynamically regulated in C2C12 during differentiation, shuttling from the cytoplasm to the nucleus under 2%HS treatment where it directly binds DNA. To determine if the FOXO1-LINC affects the localization of endogenous

viii

FoxO1, we used a FoxO1-C-terminal antibody to detect endogenous FoxO1 localization (Note: the immunogen sequence is not retained in FOXO1-LINC, thus specific for endogenous FoxO1). 2%HS treatment for 16 hours (16hr) induced the highest nuclear FoxO1 activity in both control C2C12 and FOXO1-LINC (Fig. 2B), indicating FOXO1-LINC does not affect the localization of FoxO1. We then checked the expression of FoxO1 target genes in the FOXO1-LINC lines with 24 hours of 2%HS treatment via RT qPCR. Expression of Map1lc3a (LC3) and MuRF1 were significantly lower in the FOXO1-LINC lines, however Gadd45 was significantly higher (Fig. 2C). Due to time limitations while developing stable lines, the control for Fig. 2B and C were normal C2C12 without expression of the empty PB vector. However, the vectoral load would not have had an effect on the localization of FoxO1, as localization mirrored the FOXO1-LINC lines. Upon phosphorylation by the serine/threonine kinase Protein Kinase B (AKT) at T24, S256, S319, FOXO1 is shuttled out of the nucleus. Mutating the three AKT phosphorylation sites to alanine deems FOXO1 constitutively active, losing the ability to locate to the cytoplasm. To determine if the FOXO1-LINC protein was capable of specifically inhibiting FOXO1 activity, we co-transfected the pGL-2 3xIRS-luciferase reporter and the FOXO1-LINC gene with plasmids expressing WT FOXO1 and constitutively active FOXO1 (caFOXO1) in HEK293 cells. The FOXO1-LINC protein significantly inhibited FOXO1 and caFOXO1 luciferase activity, indicating a strong specific inhibition (Fig. 2D). Taken together, this data suggests the FOXO1-LINC fusion protein could act as a dominant negative to FOXO1 function independent of the in-vitro context, and should be explored more for an exact mechanism of action and its function in-vivo.

The FOXO1-LINC fusion protein is a potential oncogenic driver of ERMS in-vitro To underlie the role of FOXO1-LINC protein in the pathological development of ERMS, we first hypothesized that expression would sustain the proliferation of myoblasts when cultured in 2%HS, whereas normal cell cycle arrest would occur. However, Ki-67 immunostaining data revealed this was not the case (data not shown). FOXO1-LINC lines grew nearly identical in 10%FBS, but did show a slight, but significant proliferative advantage when cultured past confluency (Fig. 3A,B). Other methods to detect proliferation, such as an anti-phosphohistone H3 antibody, showed no consistent overlap with the FOXO1-LINC fusion and lines could not establish colonies when grown in anchorage-independent growth conditions (data not shown).

ix

Figure 3. FOXO1-LINC expression can potentially drive oncogenesis in multiple ways. (A) Ki-67 immunostaining of stable cell lines at confluent (D2) or over confluent (D4). (B) Quantification of Ki-67+ cells at D2 and D4 of 10%FBS treatment. Results displayed as Ki-67+ nuclei/total nuclei (N=3 images for 3 slides). (C) Expression of MyoG D1 and D3 of differentiation compared to Ctl. Data shown is relative expression to the Ctl and normalized to 18s. (D) Stable lines were fixed at D3 of differentiation and stained for Taz localization (White arrow=nuclear, green=cytoplasmic, yellow=cytoplasmic/nuclear). (E) Diagram representing the strategy to mimic the FOXO1-LINC phenotype with Taz-S4A. (F) C2C12 cells were transfected with PB-CAG-eGFP-IRES or PB-Flag-Taz-S4A-eGFP and differentiated for one day. MyoG and Ctgf expression shown is relative to control. (G) C2C12 were transfected with PB-Flag-Taz-S4A-eGFP and immunoassayed for MyHC. Image is showing no overlay of MyHC and eGFP expression. An independent student t-test was used to determine the significance for all experimental results shown. * p<.05, ** p<.01, *** p<.001.

The expression of MyoG has been previously described as a reliable marker for ARMS when compared to ERMS, with ERMS tumor cells displaying less than 25% nuclear staining, compared to 75-100% of nuclear staining seen in ARMS (35). In C2C12, MyoG expression is upregulated during differentiation to promote cell cycle arrest and ensure terminal differentiation, so we hypothesized the expression of MyoG will be lower in the FOXO1-LINC lines. MyoG expression was significantly lower at D1 and D3 of differentiation when compared to the Ctl (Fig. 3C). The ability of the FOXO1-LINC stable lines to sustain proliferation past confluency indicated that cell cycle arrest due to cell-to-cell contact is aberrant. The transcriptional co-activator with PDZ-binding motif (Taz) has recently been identified as an effector of myogenesis and plays an oncogenic role in the development of ERMS, where its expression is associated with worse patient prognosis (43,44). To further reveal a potential target for this tumor genetic background, we examined the localization of Taz in the FOXO1-LINC lines. At D3 of differentiation, the FOXO1-LINC lines displayed a mix of nuclear (white arrow), cytoplasmic (green arrow) and both nuclear/cytoplasmic (yellow arrow) Taz localization, while it was largely cytoplasmic in the Ctl (Fig 3D). We then transfected C2C12 with a constitutively active form of Taz, mutating four serine phosphorylation sites to alanine (Taz-S4A), inducing the constant nuclear localization of Taz (Fig. 3E). We found that Taz-S4A inhibits MyoG expression at D1 of differentiation, consistent with what was observed in the FOXO1-LINC lines (Fig. 3F). The overexpression of Taz was verified by Ctgf expression, a Taz target gene. Finally, because the Taz-S4A construct is in frame with eGFP, we differentiated the Taz-S4A transfected cells and detected MyHC at D3 as a terminal differentiation marker. MyHC was not seen co-expressed in cells that expressed Taz-S4A, suggesting Taz-S4A is capable of inhibiting the terminal differentiation of C2C12 myoblasts (Fig. 3G). The overlapping expression of MyHC and Taz-S4A should be checked at D5 of differentiation to verify this inhibition through differentiation and MyoG protein expression should be checked next. Other methods such as EdU incorporation after differentiation should be tried next. These results imply the FOXO1-LINC fusion protein plays a potential oncogenic role in ERMS by partial nuclear retention of Taz in-vitro.

Conclusion: The results presented provide primary in-vitro evidence for a FOXO1 Fusion Protein involved in ERMS. Expression of the FOXO1-LINC inhibited terminal differentiation, a characteristic conserved in ERMS cells lines such as RD cells (45,46). Supportive to the lack of differentiation, MyoG expression was significantly lower in the FOXO1-LINC lines, consistent with the lack of MyoG expression seen histologically in ERMS (34,35). The FOXO1-LINC acts as a novel dominant negative to FOXO1 function without effecting endogenous FOXO1 protein levels (data not shown). The disruption of FOXO1 function has also been reported with expression of PAX3/7-FOXO1, indicating a potentially conserved feature between this ERMS and ARMS (47). Although not fully transformative in C2C12, these results could differ in another model, such as human myoblasts. Nonetheless, the FOXO1-LINC protein could act through multiple mechanisms. It has been shown in C2C12 that the inhibition of autophagy in C2C12 inhibits MyoG expression and terminal differentiation (48,49). Lower expression of LC3 and higher expression of Gadd45 with an increase in CC3 fluorescence expression (Supp. Fig. 2C) suggests autophagy is dysregulated. Further experiments to detect autophagic flux by LC3 protein and LC3-GFP puncta with hydroxychloroquine treatment should be done to investigate. Concurrently, the nuclear Taz expression patterns could indicate FOXO1-LINC induces nuclear Taz, further inhibiting MyoG expression and terminal differentiation. However, this could be an effect of inhibited terminal differentiation rather than a cause, as the nuclear expression was not seen in every cell. Regardless of the mechanism of action, we show that over-expression of the FOXO1-LINC protein inhibits the terminal differentiation of C2C12, delays MyoG expression and disrupts normal FOXO1 function.

In this report, we use an in-vitro model to study FOXO1-LINC, where over-expression alone is capable of driving a phenotype in C2C12. Although it is unknown if the Ch.13 deletion resulting in the FOXO1-LINC fusion is homozygous, this in-vitro model is unreflective of the in-vivo tumorigenesis in that endogenous levels of FoxO1 and LINC00598 are still expressed. Homozygous deletion of this Ch.13 region in pluripotent stem cells with

x

CRISPR-Cas9 or knock-down FoxO1 paired with expression of the FOXO1-LINC could better recapitulate the in-vivo tumorigenesis. Regardless, over-expression models are often used to determine the function of novel fusion genes (47,50). Further experiments will aim to determine how the FOXO1-LINC fusion protein drives transformation in a ‘second hit’ in-vitro model as well as the in-vivo confirmation of in-vitro results. An RNA-Seq will help to determine differentially expressed genes and pathways affected by the FOXO1-LINC. Pairing the FOXO1-LINC expression with a MYCN amplification or a large T antigen in-vitro will shed light on its transformative potential. Obtaining the patient sample in paraffin will allow detection of Taz localization and experiments determining the histological expression patterns conserved/unique with other ERMS genetic backgrounds. Establishing mouse xenograft models using stable cell line engraftment will allow us to study tumor formation and test therapeutic treatments in-vivo. Taken together, our results provide supporting evidence for the exploration of the novel FOXO1-LINC fusion protein in-vivo.

Statement of Work: The FOXO1-LINC constructs were made by Michael Thompson and the PB-Taz-S4A construct was a gift from Xiaohua Hu of the Richard Lu lab. The 3xIRS luciferase and pcDNA3 Flag FKHR AAA mutant were gifts from Kunliang Guan. Joseph Pressey M.D. is the clinician who identified the fusion gene and provided guidance throughout with helpful information on the patient. All experiments, primary data recording and statistical analysis were done by Tanner Wherley under the guidance of Dr. Masahide Sakabe and Dr. Mei Xin.

xi

Bibliography: 1. Skapek, S. X., Ferrari, A., Gupta, A. A., Lupo, P. J., Butler, E., Shipley, J., … Hawkins, D. S. (2019). Rhabdomyosarcoma. Nature Reviews Disease Primers, 5(1). 2. Parham, D. M., & Barr, F. G. (2013). Classification of Rhabdomyosarcoma and Its Molecular Basis. Advances In Anatomic Pathology, 20(6), 387–397. 3. Keller, C., & Guttridge, D. C. (2013). Mechanisms of impaired differentiation in rhabdomyosarcoma. FEBS Journal, 280(17), 4323–4334. 4. Parham, D.M. (2001). Pathologic Classification of Rhabdomyosarcomas and Correlations with Molecular Studies. Modern Pathology, 14, 506–514. 5. O Ognjanovic, S., Linabery, A. M., Charbonneau, B., & Ross, J. A. (2009). Trends in childhood rhabdomyosarcoma incidence and survival in the United States, 1975-2005. Cancer, 115(18), 4218– 4226. 6. O Ortega, J. A., Donaldson, S. S., Ivy, S. P., Pappo, A., & Maurer, H. M. (1997). Venoocclusive disease of the liver after chemotherapy with vincristine, actinomycin D, and cyclophosphamide for the treatment of rhabdomyosarcoma. Cancer, 79(12), 2435–2439. Retrieved from 7. C Chen, C., Dorado Garcia, H., Scheer, M., & Henssen, A. G. (2019a). Current and Future Treatment Strategies for Rhabdomyosarcoma. Frontiers in Oncology, 9, 1458. 8. Shern, J. F., Chen, L., Chmielecki, J., Wei, J. S., Patidar, R., Rosenberg, M., … Khan, J. (2014). Comprehensive Genomic Analysis of Rhabdomyosarcoma Reveals a Landscape of Alterations Affecting a Common Genetic Axis in Fusion-Positive and Fusion-Negative Tumors. Cancer Discovery, 4(2), 216–231. 9. B Barr, F. G., Qualman, S. J., Macris, M. H., Melnyk, N., Lawlor, E. R., Strzelecki, D. M., ... & Sorensen, P. H. (2002). Genetic heterogeneity in the alveolar rhabdomyosarcoma subset without typical gene fusions. Cancer Research, 62(16), 4704-4710. 10. Wachtel, M., Dettling, M., Koscielniak, E., Stegmaier, S., Treuner, J., Simon-Klingenstein, K., … Schäfer, B. W. (2004). Gene Expression Signatures Identify Rhabdomyosarcoma Subtypes and Detect a Novel t(2;2)(q35;p23) Translocation Fusing PAX3 to NCOA1. Cancer Research, 64(16), 5539–5545. 11. Galili, N., Davis, R. J., Fredericks, W. J., Mukhopadhyay, S., Rauscher, F. J., III, Emanuel, B. S., … Barr, F. G. (1993). Fusion of a fork head domain gene to PAX3 in the solid tumour alveolar rhabdomyosarcoma. Nature Genetics, 5(3), 230–235. 12. Davis, R. J. (1994). Fusion of PAX7 to FKHR by the Variant t(1;13)(p36;q14) Translocation in Alveolar Rhabdomyosarcoma. Cancer Res., 54, 2869. 13. Marshall, A. D., & Grosveld, G. C. (2012). Alveolar rhabdomyosarcoma – The molecular drivers of PAX3/7-FOXO1-induced tumorigenesis. Skeletal Muscle, 2(1), 25. 14. Barr, F. G., Smith, L. M., Lynch, J. C., Strzelecki, D., Parham, D. M., Qualman, S. J., & Breitfeld, P. P. (2006). Examination of Gene Fusion Status in Archival Samples of Alveolar Rhabdomyosarcoma Entered on the Intergroup Rhabdomyosarcoma Study-III Trial. The Journal of Molecular Diagnostics, 8(2), 202– 208. 15. Williamson, D., Missiaglia, E., de Reyniès, A., Pierron, G., Thuille, B., Palenzuela, G., … Delattre, O. (2010). Fusion Gene–Negative Alveolar Rhabdomyosarcoma Is Clinically and Molecularly Indistinguishable From Embryonal Rhabdomyosarcoma. Journal of Clinical Oncology, 28(13), 2151– 2158. 16. Laé, M., Ahn, E., Mercado, G., Chuai, S., Edgar, M., Pawel, B., … Ladanyi, M. (2007). Global gene expression profiling of PAX-FKHR fusion-positive alveolar and PAX-FKHR fusion-negative embryonal rhabdomyosarcomas. The Journal of Pathology, 212(2), 143–151. 17. Shukla, N., Ameur, N., Yilmaz, I., Nafa, K., Lau, C.-Y., Marchetti, A., … Ladanyi, M. (2011). Oncogene Mutation Profiling of Pediatric Solid Tumors Reveals Significant Subsets of Embryonal Rhabdomyosarcoma and Neuroblastoma with Mutated Genes in Growth Signaling Pathways. Clinical Cancer Research, 18(3), 748–757. 18. Renshaw, J., Taylor, K. R., Bishop, R., Valenti, M., De Haven Brandon, A., Gowan, S., … Shipley, J. (2013). Dual Blockade of the PI3K/AKT/mTOR (AZD8055) and RAS/MEK/ERK (AZD6244) Pathways Synergistically Inhibits Rhabdomyosarcoma Cell Growth In Vitro and In Vivo. Clinical Cancer Research, 19(21), 5940–5951. 19. Calabrese, G., Franchi, P. G., Stuppia, L., Rossi, C., Bianchi, C., Antonucci, A., & Palka, G. (1992). Translocation (8;11)(q12–13;q21) in embryonal rhabdomyosarcoma. Cancer Genetics and Cytogenetics, 58(2), 210–211.

xii

20. Sirvent, N., Trassard, M., Ebran, N., Attias, R., & Pedeutour, F. (2009). Fusion of EWSR1 with the DUX4 facioscapulohumeral muscular dystrophy region resulting from t(4;22)(q35;q12) in a case of embryonal rhabdomyosarcoma. Cancer Genetics and Cytogenetics, 195(1), 12–18. 21. Karanian, M., Pissaloux, D., Gomez-Brouchet, A., Chevenet, C., Le Loarer, F., Fernandez, C., … Tirode, F. (2020). SRF-FOXO1 and SRF-NCOA1 Fusion Genes Delineate a Distinctive Subset of Well- differentiated Rhabdomyosarcoma. The American Journal of Surgical Pathology, 44(5), 607–616. 22. Mosquera, J. M., Sboner, A., Zhang, L., Kitabayashi, N., Chen, C.-L., Sung, Y. S., … Antonescu, C. R. (2013). RecurrentNCOA2gene rearrangements in congenital/infantile spindle cell rhabdomyosarcoma. Genes, and Cancer, 52(6), 538–550. 23. Sun, X., Guo, W., Shen, J. K., Mankin, H. J., Hornicek, F. J., & Duan, Z. (2015). Rhabdomyosarcoma: Advances in Molecular and Cellular Biology. Sarcoma, 2015, 1–14. 24. Boscolo Sesillo, F., Fox, D., & Sacco, A. (2019). Muscle Stem Cells Give Rise to Rhabdomyosarcomas in a Severe Mouse Model of Duchenne Muscular Dystrophy. Cell Reports, 26(3), 689-701.e6. 25. Morotti, R. A., Nicol, K. K., Parham, D. M., Teot, L. A., Moore, J., Hayes, J., … Qualman, S. J. (2006). An Immunohistochemical Algorithm to Facilitate Diagnosis and Subtyping of Rhabdomyosarcoma: The Childrenʼs Oncology Group Experience. The American Journal of Surgical Pathology, 30(8), 962–968. 26. Hatley, M. E., Tang, W., Garcia, M. R., Finkelstein, D., Millay, D. P., Liu, N., … Olson, E. N. (2012). A Mouse Model of Rhabdomyosarcoma Originating from the Adipocyte Lineage. Cancer Cell, 22(4), 536– 546. 27. Sabourin, L. A., & Rudnicki, M. A. (2000). The molecular regulation of myogenesis. Clinical genetics, 57(1), 16-25. 28. Andrés, V., & Walsh, K. (1996). Myogenin expression, cell cycle withdrawal, and phenotypic differentiation are temporally separable events that precede cell fusion upon myogenesis. The Journal of cell biology, 132(4), 657-666. 29. YAFFE, D., & SAXEL, O. (1977). Serial passaging and differentiation of myogenic cells isolated from dystrophic mouse muscle. Nature, 270(5639), 725–727. 30. Tapscott, S. J. (2005). The circuitry of a master switch: Myod and the regulation of skeletal muscle gene transcription. Development, 132(12), 2685–2695. 31. Cao, Y., Yao, Z., Sarkar, D., Lawrence, M., Sanchez, G. J., Parker, M. H., … Tapscott, S. J. (2010). Genome-wide MyoD Binding in Skeletal Muscle Cells: A Potential for Broad Cellular Reprogramming. Developmental Cell, 18(4), 662–674. 32. MacQuarrie, K. L., Yao, Z., Fong, A. P., Diede, S. J., Rudzinski, E. R., Hawkins, D. S., & Tapscott, S. J. (2012). Comparison of Genome-Wide Binding of MyoD in Normal Human Myogenic Cells and Rhabdomyosarcomas Identifies Regional and Local Suppression of Promyogenic Transcription Factors. Molecular and Cellular Biology, 33(4), 773–784. 33. Tapscott, S., Thayer, M., & Weintraub, H. (1993). Deficiency in rhabdomyosarcomas of a factor required for MyoD activity and myogenesis. Science, 259(5100), 1450–1453. 34. Kumar, S., Perlman, E., Harris, C. A., Raffeld, M., & Tsokos, M. (2000). Myogenin is a Specific Marker for Rhabdomyosarcoma: An Immunohistochemical Study in Paraffin-Embedded Tissues. Modern Pathology, 13(9), 988–993. 35. Sebire, N. J. (2003). Myogenin and MyoD1 expression in paediatric rhabdomyosarcomas. Journal of Clinical Pathology, 56(>6), 412–416. 36. Greer, E. L., & Brunet, A. (2005). FOXO transcription factors at the interface between longevity and tumor suppression. Oncogene, 24(50), 7410–7425. 37. ZHAO, X., GAN, L., PAN, H., KAN, D., MAJESKI, M., ADAM, S. A., & UNTERMAN, T. G. (2004). Multiple elements regulate nuclear/cytoplasmic shuttling of FOXO1: characterization of phosphorylation- and 14- 3-3-dependent and -independent mechanisms. Biochemical Journal, 378(3), 839–849. 38. Tzivion, G., Dobson, M., & Ramakrishnan, G. (2011). FoxO transcription factors; Regulation by AKT and 14-3-3 proteins. Biochimica et Biophysica Acta (BBA) - Molecular Cell Research, 1813(11), 1938–1945. 39. Kitamura, T., Kitamura, Y. I., Funahashi, Y., Shawber, C. J., Castrillon, D. H., Kollipara, R., … Accili, D. (2007). A Foxo/Notch pathway controls myogenic differentiation and fiber type specification. Journal of Clinical Investigation, 117(9), 2477–2485. 40. Kamei, Y., Miura, S., Suzuki, M., Kai, Y., Mizukami, J., Taniguchi, T., … Ezaki, O. (2004). Skeletal Muscle FOXO1 (FKHR) Transgenic Mice Have Less Skeletal Muscle Mass, Down-regulated Type I (Slow Twitch/Red Muscle) Fiber Genes, and Impaired Glycemic Control. Journal of Biological Chemistry, 279(39), 41114–41123.

xiii

41. Milan, G., Romanello, V., Pescatore, F., Armani, A., Paik, J.-H., Frasson, L., … Sandri, M. (2015). Regulation of autophagy and the ubiquitin–proteasome system by the FoxO transcriptional network during muscle atrophy. Nature Communications, 6(1), 6670. 42. Rio, D. C., Ares, M., Jr, Hannon, G. J., & Nilsen, T. W. (2010). Purification of RNA using TRIzol (TRI reagent). Cold Spring Harbor protocols, 2010(6), 5439. 43. Mohamed, A., Sun, C., De Mello, V., Selfe, J., Missiaglia, E., Shipley, J., … Wackerhage, H. (2016). The Hippo effector TAZ (WWTR1) transforms myoblasts and TAZ abundance is associated with reduced survival in embryonal rhabdomyosarcoma. The Journal of Pathology, 240(1), 3–14. 44. Sun, C., De Mello, V., Mohamed, A., Ortuste Quiroga, H. P., Garcia-Munoz, A., Al Bloshi, A., … Zammit, P. S. (2017). Common and Distinctive Functions of the Hippo Effectors Taz and Yap in Skeletal Muscle Stem Cell Function. STEM CELLS, 35(8), 1958–1972. 45. Xu, Q., & Wu, Z. (2000). The Insulin-like Growth Factor-Phosphatidylinositol 3-Kinase-Akt Signaling Pathway Regulates Myogenin Expression in Normal Myogenic Cells but Not in Rhabdomyosarcoma- derived RD Cells. Journal of Biological Chemistry, 275(47), 36750–36757. 46. Tiffin, N., Williams, R. D., Shipley, J., & Pritchard-Jones, K. (2003). PAX7 expression in embryonal rhabdomyosarcoma suggests an origin in muscle satellite cells. British Journal of Cancer, 89(2), 327– 332. 47. Schmitt-Ney, M., & Camussi, G. (2015). The PAX3-FOXO1 Fusion Protein Present in Rhabdomyosarcoma Interferes with Normal FOXO Activity and the TGF-β Pathway. PLOS ONE, 10(3), e0121474. 48. Fortini, P., Ferretti, C., Iorio, E., Cagnin, M., Garribba, L., Pietraforte, D., … Dogliotti, E. (2016). The fine tuning of metabolism, autophagy and differentiation during in vitro myogenesis. Cell Death & Disease, 7(3), e2168. 49. Sin, J., Andres, A. M., Taylor, D. J. R., Weston, T., Hiraumi, Y., Stotland, A., … Gottlieb, R. A. (2015). Mitophagy is required for mitochondrial biogenesis and myogenic differentiation of C2C12 myoblasts. Autophagy, 12(2), 369–380. 50. Tomlins, S. A., Laxman, B., Varambally, S., Cao, X., Yu, J., Helgeson, B. E., ... & Mehra, R. (2008). Role of the TMPRSS2-ERG gene fusion in prostate cancer. Neoplasia (New York, NY), 10(2), 177.

xiv

Supplementary:

Supplemental Figure 1. (A) eGFP expression of 5 representative stable lines generated taken at different time points during generation with Flag verification. (B) Full membrane image immunoblotting with anti-Flag antibody using stable lines isolated at full confluency cultured in 10%FBS for 2 days. (C) MyHC protein expression at D0 and D5 of differentiation for both control lines (1,2) and FOXO1-LINC lines (2,3). (D) Fold change in MyHC protein expression relative to Gapdh at D5 of differentiation (N=3 repeated experiments). (E) Representative MyHC images used for calculating the Fusion Index at D5 of differentiation.

xv

Supplemental Figure 2. (A) BLAST results using the FOXO1 DBD (aa 159-235) and the full aa sequence of the FOXO1-LINC. (B) Full aa sequence alignment showing 75% of the DNA binding domain of FOXO1 conserved. Query 12289 is FOXO1 WT DBD, query 12291 is FOXO1-LINC sequence. (C) Cleaved-Caspase 3 detection in C2C12 stable lines with D1 of differentiation.

Table 1 and 2. Primers and antibodies used.

xvi