MECHANISMS, MODELS, AND THERAPEUTIC VULNERABILITIES OF CRTC1/MAML2-POSITIVE SALIVARY MUCOEPIDERMOID CARCINOMA
Adele Musicant
A dissertation submitted to the faculty at the University of North Carolina at Chapel Hill in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Curriculum in Genetics and Molecular Biology in the School of Medicine.
Chapel Hill 2021
Approved by:
Antonio L. Amelio
Albert S. Baldwin
Jimena Giudice
Gaorav P. Gupta
D. Neil Hayes
©2021 Adele Musicant ALL RIGHTS RESERVED
ii
ABSTRACT
Adele Musicant: Mechanisms, Models, and Therapeutic Vulnerabilities of CRTC1/MAML2-Positive Salivary Mucoepidermoid Carcinoma (Under the direction of Antonio L. Amelio)
Mucoepidermoid carcinoma (MEC) is a life-threatening salivary gland cancer for which treatment options and targeted therapies are lacking. This tumor type is frequently characterized by a recurrent t(11;19)(q21;p13) chromosomal translocation which generates the transcriptional coactivator fusion CRTC1/MAML2. While it is clear that CRTC1/MAML2 plays a key role in maintaining a transformed state in MEC, the mechanisms by which this fusion oncogene rewires gene expression programs that promote tumorigenesis remain poorly understood. Here, we show that CRTC1/MAML2 induces transcriptional activation of a non-canonical PGC-1α splice variant, PGC-1α4, which regulates PPARγ-dependent expression of the IGF-1 growth hormone. This dependence on autocrine regulation of IGF-1 transcription via PGC-1α4/PPARγ renders
MEC cells susceptible to IGF-1R inhibition with small molecules and to PPARγ inhibition with inverse agonists. These results yield insights into the aberrant co-regulatory functions of CRTC1/MAML2 and identify specific vulnerabilities that can be exploited for precision therapy.
The tumor “cell of origin” is a normal cell in which a cancer-causing mutation or genetic hit occurs. Identification and characterization of the cell of origin is an essential undertaking, as a better understanding of this cell type can aid in earlier diagnosis, hint
iii at future tumor behavior, and can even inform treatment strategies. MEC tumors exhibit extreme cellular heterogeneity, and so it has been challenging to determine what cell type is responsible for tumorigenesis in this cancer. We therefore also sought to characterize the MEC cell of origin by generating several genetically engineered mouse models in which the CRTC1/MAML2 fusion oncogene is inducibly expressed in specific salivary gland cell populations. Our results demonstrate that CRTC1/MAML2-positive
MEC originates from a salivary ductal precursor cell and that development of MEC tumors is aided by loss of Tp53 gene expression. While further studies are required to reduce off-target effects of oncogene induction within the cutaneous epithelia, the importance of determining the salivary MEC cell of origin cannot be understated.
iv
ACKNOWLEDGEMENTS
I want to first thank the entire Amelio lab. It may be a cliché, but our lab really is like a family. Tony, thank you for believing in me and taking me on as your first graduate student. I think the last 6+ years have been a learning experience for both of us and I am proud to be the first PhD out of your lab. I appreciate your ability to be encouraging and supportive while also challenging us to bring our very best to the bench every day. I can say with certainty that you’ve helped me become a better scientist. Miranda, thank you for your unwavering support and friendship. I had no idea when I first met you that this fast-talking, energetic, spontaneous redhead would become one of my closest friends in graduate school! I’m so thankful to have you in my life. Kay, thank you for being a wonderful scientist and friend. Your enthusiasm for science is something that I admire and strive to emulate. I’ll always be thankful for the stimulating conversations, the jokes, and the champagne! Jason, I’m so thankful that your six months in the States led to such a wonderful friendship. I miss the Greek lessons and drinking too much coffee with you! I know you’ll visit again soon but for now, σε ευχαριστώ φίλε μου.
Thank you also to my parents – all four of them! Dad, thank you for raising me to be curious and to always ask questions. That curiosity is what led me to science in the first place. Mom, thank you for your constant and unwavering support. I always knew you were there for me, even if I made you stop asking me when I would graduate! Beth, thank you for being the most positive person I know. I don’t think I could have gotten
v through all of these challenges without being reminded of my worth and potential when I most needed it. And Jim, thank you for reminding me to live a little. It’s so easy to be overwhelmed by graduate school and you have always been there to help me recognize when I need some balance.
And finally, thank you to some of my closest friends. You have all been such a source of support during my entire time in grad school. Robin, can you believe we met all the way back in first year group? You were one of my very first friends in graduate school and I have cherished every moment since. From running to wine nights to Zoom calls, I’m so thankful I got to share it all with you. Johanna, what would I do without you?
When grad school got hard, I always knew I could come to you for support and encouragement. Thank you for all the lunch chats and Randys nights! Michelle, my time in GMB wouldn’t have been the same without you. Thank you for being such a wonderful retreat roommate, presentation partner, and wine buddy! Joan, I feel so lucky to have you in my life. Despite being literally an ocean away, you’ve been a constant source of support through so much. Words cannot express how grateful I am to count you among my friends. And finally, Steph, my Northeastern twin. Can you believe we met almost nine years ago? I cannot imagine what my life would be like without you. So many of my favorite memories are with you: nights in Boston watching The Bachelor, our incredible trip to Europe for spring break, and hundreds of coffee dates in between! I love you chica; thank you for being the greatest best friend a girl could ever ask for.
vi
TABLE OF CONTENTS
LIST OF TABLES ...... ix
LIST OF FIGURES ...... x
LIST OF ABBREVIATIONS ...... xiii
CHAPTER 1: INTRODUCTION TO SALIVARY GLAND MALIGNANCIES ...... 1
1.1 Introduction ...... 1
1.2 Mucoepidermoid Carcinoma ...... 2
1.3 MEC Genetics and Molecular Characteristics ...... 4
CHAPTER 2: CRTC1/MAML2 ESTABLISHES A PGC1α-IGF1 CIRCUIT THAT CONFERS VULNERABILITY TO PPARγ INHIBITION ...... 7
2.1 Introduction ...... 7
2.2 Results ...... 10
2.2.1 Activation of IGF-1 Signaling is a Hallmark of CRTC1/MAML2-Positive Salivary Mucoepidermoid Carcinoma ...... 10
2.2.2 CRTC1/MAML2-Positive MEC Tumor Cell Lines Display Selective Sensitivity to Targeted IGF-1R Inhibition ...... 11
2.2.3 IGF-1R Signaling is Critical for CRTC1/MAML2-Positive MEC Tumor Cell Growth and Survival ...... 13
2.2.4 CRTC1/MAML2 Establishes a Synthetic PGC-1α4 Circuit that Regulates IGF-1 Expression in MEC Tumor Cells ...... 14
2.2.5 The PGC-1a4 Co-Regulator Activates PPARγ-dependent Transcription of IGF-1 in CRTC1/MAML2-Positive MEC Cells ...... 16
2.2.6 PPARγ Inverse Agonists Suppress CRTC1/MAML2-Positive MEC Tumor Growth ...... 17
vii 2.3 Discussion ...... 19
CHAPTER 3: DEVELOPING AN AUTOCHTHONOUS MOUSE MODEL OF MUCOEPIDERMOID CARCINOMA ...... 25
3.1 Introduction ...... 25
3.2 Results ...... 30
3.2.1 Targeted Transgene Expression is Achieved via Cell Type-Specific Promoters .... 30
3.2.2 CRTC1/MAML2-Positive Salivary MEC Originates in a Ductal Cell ...... 31
3.2.3 CRTC1/MAML2 Overexpression and TP53 Tumor Suppressor Loss Collectively Drive MEC Tumor Formation ...... 32
3.2.4 Murine Salivary MEC Tumors Recapitulate CRTC1/MAML2-Positive Human Salivary MEC Genetics ...... 33
3.3 Discussion ...... 35
CHAPTER 4: FUTURE DIRECTIONS IN MUCOEPIDERMOID CARCINOMA RESEARCH ...... 39
APPENDIX 1: MATERIALS AND METHODS ...... 43
APPENDIX 2: TABLES ...... 65
APPENDIX 3: FIGURES ...... 70
REFERENCES ...... 109
viii LIST OF TABLES
Table 1 Cancer pathway genes induced upon CRTC1/MAML2 expression……………………………………………………………………… 61
Table 2 Calculated IC50 values of IGF-1R inhibitors on the viability of CRTC1/MAML2-positive and -negative cancer cell lines……………… 62
Table 3 Calculated IC50 values of PPARγ inverse agonists on the viability of CRTC1/MAML2-positive cancer cell lines……………………… 63
Table 4 Primers used for qRT-PCR………………….………………………………. 64
Table 5 Primers used for detection of mycoplasma in cell cultures, MAML2 gene-specific cDNA synthesis, ChIP, and cloning and sequence verification of sgRNA constructs………………………….. 65
ix LIST OF FIGURES
Figure 1 Salivary gland structure………………………………………………………. 66
Figure 2 Salivary MEC is characterized by a t(11;19)(q14-21;p12-13) translocation………………………………….…... 67
Figure 3 CRTC1/MAML2 is expressed in >50% of human MEC samples………………………………………………………..………...…….. 68
Figure 4 Normal salivary glands, CRTC1/MAML2-positive MEC, and CRTC1/MAML2-negative MEC tumors are characterized by distinct gene signatures………………………………….. 69
Figure 5 An IGF-1 related gene signature is characteristic of fusion-positive MEC…………………………………………………………... 70
Figure 6 Curated Musicant_MEC_CRTC1-MAML2_IGF1 gene set distinguishes CRTC1/MAML2-positive MEC…………………….…….. 71
Figure 7 CRTC1/MAML2 and IGF-1 transcript levels are correlated in human MEC tumor samples………………………………...... 72
Figure 8 CRTC1/MAML2 and IGF-1 transcript levels are correlated in MEC cell lines………………………………………………….. 73
Figure 9 CRTC1/MAML2-positive xenografts display increased IGF-1 expression………………………………………………….. 74
Figure 10 A focused small molecule drug screen identifies sensitivity of CRTC1/MAML2-positive tumor cells to compounds targeting IGF-1R………………………………………………... 75
Figure 11 Sensitivity of three CRTC1/MAML2-positive cell lines to IGF-1R inhibition…………………………………………………………… 76
Figure 12 Validation of IGF-1R inhibition and shRNA-mediated knockdown……………………………………………………………………... 77
Figure 13 IGF-1R inhibition blocks MEC cell growth and survival…………………………………………………………………………. 78
Figure 14 CRTC1/MAML2 induces IGF-1 expression…………………………….…... 79
Figure 15 CRTC1/MAML2 induces expression of the PGC-1α4 splice variant…………………………………………………………………… 80
x Figure 16 The PGC-1α4 variant is highly expressed in CRTC1/MAML2-positive MEC.………………………………………………. 81
Figure 17 CRISPRi-based knockdowns confirm that IGF-1 is regulated by a specific PGC-1α isoform……………..……………………... 82
Figure 18 Overexpression of PGC-1α4 drives IGF-1 transcription…………….….… 83
Figure 19 Knockdown of key signaling components inhibits proliferation of MEC cells…………………………………………………….. 84
Figure 20 CRTC1/MAML2 overexpression activates IGF-1 and PPARγ response element luciferase reporters…………..………………… 85
Figure 21 IGF-1 activation is dependent on PPARγ-mediated signaling.……...…… 86
Figure 22 PPARγ knockdown inhibits IGF-1 expression and induces apoptosis in MEC cells……………….…………………………….. 87
Figure 23 PPARγ inverse agonists inhibit IGF-1 expression………...……...……….. 88
Figure 24 PPARγ inverse agonists suppress MEC cell growth, proliferation, and survival………………..…………………………………… 89
Figure 25 PPARγ inverse agonists suppress tumor growth in a xenograft model of salivary MEC……………...…………………………….. 90
Figure 26 Schematic of the CRTC1/MAML2 – PGC-1α4 – IGF-1 signaling pathway……………………….…………………………………….. 91
Figure 27 Cells targeted by Cre promoters used in transgenic lines………..………. 92
Figure 28 Bioluminescent imaging demonstrates transgene expression patterns in LumiFluor-positive animals……..…………………. 93
Figure 29 CRTC1/MAML2 is expressed in salivary gland tissue of DcppCre-C1/M2, MistCre-C1/M2, and K14Cre-C1/M2 transgenic mice………………………………………………...……………… 94
Figure 30 Ductal expression of CRTC1/MAML2 drives a dysplastic and hyperproliferative phenotype.………………………………. 95
Figure 31 CRTC1/MAML2 overexpression alone fails to induce MEC formation independent of Cre driver…..………….………………….. 96
C1/M2 LumiFluor Figure 32 Generation of Ductal-PTight PGK stable cells…………………... 97
Figure 33 CRTC1/MAML2-positive human MEC tumors display increased p53 dysregulation…………….…………………………………... 98
xi Figure 34 Salivary MEC tumors arise in K14Cre-C1/M2-p53f/f mice four months after tamoxifen administration……………...…………... 99
Figure 35 MSigDB pathways dysregulated in MEC GEMM tumors………...……… 100
Figure 36 Musicant_MEC_CRTC1-MAML2_IGF1 gene set distinguishes MEC GEMM tumors……………………….………………... 101
Figure 37 MEC GEMM tumors mimic all human MEC grades……………………… 102
Figure 38 Krt14-driven oncogene expression results in significant oral lesion formation…………………………………………….. 103
Figure 39 Direct injection of adenoviral Cre results in temporary LumiFluor signal…………………...………………………………………… 104
xii LIST OF ABBREVIATIONS
AFIP Armed Forces Institute of Pathology
ANOVA Analysis of variance
ATP Adenosine triphosphate
BLI Bioluminescent imaging
BRET Bioluminescence resonance energy transfer
BSA Bovine serum albumin cAMP Cyclic adenosine monophosphate cDNA Complementary DNA
ChIP Chromatin immunoprecipitation
CRE cAMP response element
CREB cAMP response element binding protein
CRISPRi Clustered regularly interspaced short palindromic repeats interference
CRTC1 CREB-regulated transcription coactivator 1
CSC Cancer stem cell
DEG Differentially expressed gene
DMEM Dulbecco’s Modified Eagle Medium
DMSO Dimethyl sulfoxide
DNA Deoxyribonucleic acid
DSB Double-strand break
EC50 Half-maximal effective concentration
ECL Enhanced chemiluminescence
EDTA Ethylenediaminetetraacetic acid
xiii EGF Epidermal growth factor
EGFR Epidermal growth factor receptor
FACS Fluorescence-activated cell sorting
FBS Fetal bovine serum
FFPE Formalin-fixed paraffin-embedded (tissue)
FGFR Fibroblast growth factor receptor
FOV Field of view
GC50 Half-maximal growth inhibitory concentration
GEMM Genetically engineered mouse model
GSEA Gene set enrichment analysis
H&E Hematoxylin and eosin
HDAC Histone deacetylase
HRP Horseradish peroxidase
HTSF High Throughput Sequencing Facility
IACUC Institutional Animal Care and Use Committee
IC50 Half-maximal inhibitory concentration
IGF-1 Insulin-like growth factor
IGF-1R Insulin-like growth factor 1 receptor
IHC Immunohistochemistry i.p. Intraperitoneal (injection)
IRB Institutional Review Board
MAML2 Mastermind-like transcriptional coactivator 2
MEC Mucoepidermoid carcinoma
xiv mRNA Messenger RNA
MSigDB Molecular Signatures Database
NEAA Non-essential amino acids
NHEJ Non-homologous end joining
Nu/Nu Athymic nude (mouse)
PBS Phosphate-buffered saline
PCR Polymerase chain reaction
PEI Polyethyleneimine
PGC-1α PPARγ coactivator 1α
PPARγ Peroxisome proliferator-activated receptor γ
PPP Picropodophyllin
PPRE PPARγ response element
PSG Penicillin streptomycin glutamine
PVDF Polyvinylidene difluoride qRT-PCR Quantitative real-time polymerase chain reaction
RNA Ribonucleic acid
ROI Region of interest
RTK Receptor tyrosine kinase
RTKi Receptor tyrosine kinase inhibitor
SCC Squamous cell carcinoma
SD Standard deviation
SDS-PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis
SEM Standard error of the mean
xv sgRNA Single guide RNA shRNA Short hairpin RNA
SV40 Simian virus 40
TBS-T Tris-buffered saline with Tween 20
TGL Translational Genomics Lab
UNC University of North Carolina
xvi CHAPTER 1: INTRODUCTION TO SALIVARY GLAND MALIGNANCIES
1.1 Introduction
The salivary glands are a group of exocrine organs comprised of three pairs of major glands (the parotid, submandibular, and sublingual glands) and hundreds of minor salivary glands scattered throughout the oral mucosa and the upper aerodigestive tract (Holsinger and Bui, 2007; Whelton, 2012). The salivary glands are responsible for producing serous and mucous secretions, collectively referred to as saliva, which function to enzymatically break down food and maintain oral health (Holmberg and
Hoffman, 2014).
The salivary glands begin to form early in development, with the human parotid gland developing as soon as the sixth embryonic week (Holsinger and Bui, 2007). Each gland develops through a process called branching morphogenesis, whereby successive rounds of epithelial budding, elongation, and lumen formation creates a mature gland with highly branched tree-like ductal structures (Patel et al., 2006). The distal ends of the ducts are capped by saliva-producing acinar cells, while the proximal end empties into the mouth (Figure 1). Branching morphogenesis during salivary gland development is a complex process orchestrated by the precise regulation of a variety of signaling pathways and molecular cascades, chief among which are receptor tyrosine kinases (RTKs). One such RTK, fibroblast growth factor receptor (FGFR), regulates cell proliferation and differentiation in developing tissues. Pharmacological inhibition of
FGFR signaling blocks epithelial cell proliferation and decreases branching
1 morphogenesis in the mouse submandibular gland (Hoffman et al., 2002). Deletion of
FGFR2b in mice results in salivary gland hypoplasia or, in some instances, complete lack of one or more salivary glands (Patel et al., 2006). Similarly, insulin-like growth factor 1 (IGF-1) signaling plays an important role in salivary gland growth and development (Amano and Iseki, 1993; Hoffman et al., 2002; Kerr et al., 1995; Ryan et al., 1992; Werner and Katz, 2004). Dysregulation of growth factor signaling, however, is associated with multiple types of cancer including salivary malignancies (Babina and
Turner, 2017; Pollak, 2008; Yu and Rohan, 2000).
Salivary cancers can arise in any of the major or the minor glands but are relatively rare, accounting for only 0.3-0.5% of all cancers (Eveson, 2011; Speight and
Barrett, 2002). The diagnosis, prognosis, and treatment of salivary gland malignancies are complicated by a variety of factors, not least of which involves appropriate grading and staging. As of 2011, the World Health Organization recognized as many as 13 benign and 24 malignant salivary gland tumor types (Eveson, 2011) Each of these tumors are marked by different pathologies and histologies, all of which can impact prognosis and treatment (Son et al., 2018). Finally, many salivary tumors are marked by highly variable intratumoral morphologies which can cause subjective differences in pathologist tumor grading, further complicating diagnosis (Betz, 2017).
1.2 Mucoepidermoid Carcinoma
The most common type of malignant salivary gland tumor is mucoepidermoid carcinoma (MEC), which accounts for roughly 34% of all salivary malignancies (Coca-
Pelaz et al., 2015; Lujan et al., 2010; Nance et al., 2008). These tumors arise approximately twice as often in females, with a mean age of 45 years at the time of diagnosis (Chenevert et al., 2011; Cipriani et al., 2019a). MEC most often occurs in the
2 parotid gland but can arise in any of the major or minor salivary glands (Goode et al.,
1998). Less frequently, MEC may develop in other glandular tissues, including the lung, breast, liver, sweat glands, thyroid, and cervix (Bean et al., 2019; Cortez et al., 2018;
Guo et al., 2014; Huo et al., 2015; Landman and Farmer, 1990; Lennerz et al., 2009;
Roydhouse et al., 2019; Watanabe et al., 2019; Xi et al., 2012).
MEC tumors are characterized by three main cell types: mucous, epidermoid, and intermediate cells. While other cell types, such as columnar and clear cells, may also be present in these tumors, mucous, epidermoid, and intermediate cell types are the hallmarks of MEC and their presence is often considered to be diagnostic (Coca-
Pelaz et al., 2015). Mucous cells are acinar-like in appearance, with a sizable cytoplasm and small nucleus. They are well-differentiated and characterized by expression of mucins. Epidermoid cells are also terminally differentiated and are more elongated and polygonal in appearance. Finally, the intermediate cells are an undifferentiated cell type that is thought to give rise to both epidermoid and mucous cells. Notably, the relative proportions of each cell type are correlated with histological grade (Son et al., 2018).
MEC is classified as low, intermediate, or high-grade depending on a variety of factors, including perineural invasion, cellular differentiation, necrosis, and anaplasia
(Coca-Pelaz et al., 2015). Low grade tumors display minimal infiltration, have a high proportion of mucous cells and have a better prognosis (76-95% 5-year survival)
(Nance et al., 2008). High grade tumors are highly infiltrative, tend to recur and metastasize, and have a much lower 5-year survival rate (30-50%) (Brandwein et al.,
2001; Son et al., 2018). Histological grading is a critical prognostic factor and heavily influences treatment decisions; thus, grading needs to be accurate and reproducible
3 (Pires et al., 2004). Two main classification systems have emerged for grading MEC tumors: the Brandwein and AFIP systems (Brandwein et al., 2001; Cipriani et al., 2019a;
Goode et al., 1998). In each system, points are assigned for each clinical variable and the sum of all points determines the tumor’s grade. Unfortunately, the two systems use different scoring criteria and therefore one tumor may be given two different grades depending on the system used. In general, the Brandwein system tends to assign higher tumor grades, while the AFIP system tends to assign lower tumor grades
(Brandwein et al., 2001).
Both low- and high-grade MEC tumors are typically treated by complete surgical resection, with high-grade tumors often receiving adjuvant radiotherapy (Coca-Pelaz et al., 2015; Pires et al., 2004). Neck dissection may be performed in cases with suspected regional metastasis (Eveson, 2011; Pires et al., 2004). A small number of clinical trials have been conducted using chemotherapy as a treatment option for MEC, however the results have been less than promising (Goode et al., 1998). Given the relative dearth of treatment options for MEC and the difficulty in reproducibly grading these tumors, it is critical that we gain a better understanding of the underlying molecular mechanisms contributing to MEC pathobiology.
1.3 MEC Genetics and Molecular Characteristics
The majority of MEC cases (50-85%) are characterized by a t(11;19)(q14-21; p12-13) chromosomal rearrangement (O’Neill, 2009; Seethala and Chiosea, 2016;
Seethala et al., 2010; Tonon et al., 2003). This translocation fuses the first exon of
CRTC1 (CREB-regulated transcriptional coactivator 1) to the transcriptional activation domain of MAML2 (Mastermind-like 2) (Figures 2A-E). Though MAML2 is involved in the Notch signaling pathway, CRTC1/MAML2 does not appear to dysregulate Notch
4 pathway genes, as the Notch-binding domain present in full-length MAML2 is missing in the fusion product. CRTC1/MAML2 does, however, bind to and constitutively coactivate the transcription factor cAMP response element binding protein (CREB) (Coxon et al.,
2005; Wu et al., 2005) and also exhibits a gain-of-function interaction with the master transcription factor MYC (Amelio et al., 2014). Moreover, CRTC1/MAML2 has been shown to interact with the AP-1 transcriptional complex (Canettieri et al., 2009) which controls cellular mitogen responses. Through CREB, AP-1, and MYC-mediated signaling, CRTC1/MAML2 can activate key tumor-supportive cascades that regulate cell survival, proliferation, and differentiation including the epidermal growth factor receptor
(EGFR) (Chen et al., 2014) and the insulin-like growth factor 1 receptor (IGF-1R) signaling pathways (Musicant et al., 2021). It is likely that each of the varied
CRTC1/MAML2 interactions, including those that are yet to be discovered, can play a role in promoting MEC tumorigenesis.
Fusion-positive tumors typically display a low overall mutational burden, often with CRTC1/MAML2 as the sole cytogenetic abnormality (Kang et al., 2017; Martins et al., 2004). Additionally, CRTC1/MAML2-positive cell lines require sustained expression of the fusion product for their continued proliferation and survival (Komiya et al., 2006).
These findings indicate that CRTC1/MAML2 itself functions as the primary oncogenic driver in fusion-positive MEC. While early literature posited that the CRTC1/MAML2 fusion was limited to low- and intermediate-grade tumors, recent work has provided evidence to the contrary. Mischaracterization of adenosquamous carcinomas or salivary duct carcinomas as high-grade fusion-negative MEC may have disguised the true rate
5 of CRTC1/MAML2 positivity in high-grade MEC (Chenevert et al., 2011; Seethala et al.,
2010; Tirado et al., 2007).
As the CRTC1/MAML2 gene fusion is considered to be the main driver of fusion- positive salivary MEC, recent studies have focused on elucidating the molecular mechanisms driven by this oncogene in order to more accurately identify and treat this disease. This dissertation will detail a unique mechanism of growth factor signaling regulation in CRTC1/MAML2-positive MEC as well as the development of a genetically engineered mouse model (GEMM) that will be critical for future in vivo MEC studies.
6
CHAPTER 2: CRTC1/MAML2 ESTABLISHES A PGC1α-IGF1 CIRCUIT THAT CONFERS VULNERABILITY TO PPARγ INHIBITION1
2.1 Introduction
Mucoepidermoid carcinoma (MEC) is the most common salivary gland malignancy and patients with advanced recurrent or metastatic tumors often suffer from unresectable, lethal disease marked by a 5-year survival rate of <40% (Bell and Hanna,
2012; McHugh et al., 2012). Salivary MECs can arise within the major or minor salivary glands and are characterized by significant intra-tumoral cellular heterogeneity. This heterogeneity is fueled by cancer stem cells that give rise to multiple cell types including epidermoid, mucous, and intermediate cells (Adams et al., 2015; Seethala et al., 2010;
Stewart et al., 1945). The intermediate cells are a poorly differentiated, proliferative cell type and are thought to give rise to the terminally differentiated epidermoid and mucous cell populations. Further, the relative proportion of intermediate cells is increased in high-grade tumors and correlates with poor prognosis (Batsakis, 1980). Therefore, transcriptional programs that control cellular identity and support tumor cell functions can directly influence tumor grade and therefore disease progression.
Genomic characterization of salivary MEC tumors has revealed two distinct tumor subtypes: “fusion-negative” and “fusion-positive”. In fusion-positive tumors, a recurrent t(11;19) chromosomal translocation results in the oncogenic coactivator fusion
1 This chapter previously appeared in a slightly modified form as an article in the journal Cell Reports. The original citation is as follows: Musicant, A.M., et al. (2021). CRTC1-MAML2 establishes a PGC1α-IGF1 circuit that confers vulnerability to PPARγ inhibition. Cell Reports 34, 108768.
7 CRTC1/MAML2. Fusion-negative tumors, meanwhile, tend to harbor mutations in the tumor suppressor Tp53 (fusion-negative MEC) (El-Naggar et al., 1996; Kang et al.,
2017; Tonon et al., 2003; Wang et al., 2017). The majority of MEC cases (50-85%)
(O’Neill, 2009) are fusion-positive and display a strikingly low somatic mutational burden indicating that the CRTC1/MAML2 fusion is likely the primary oncogenic driver event.
This is not unheard of, as numerous other examples exist of cancers driven primarily by gene fusions in the absence of high mutational burden (Gao et al., 2018; Kadoch and
Crabtree, 2013; Missiaglia et al., 2012; Riggi et al., 2014). Although most patients with low-grade, fusion-positive tumors can often be successfully treated with surgical resection, some CRTC1/MAML2-positive patients develop high-grade tumors that display recurrent, chemoradiation-resistant disease (Chen et al., 2007; Seethala and
Chiosea, 2016; Warner et al., 2013). This underscores the critical need to develop targeted therapeutic strategies for this subset of salivary MEC patients.
Transcriptional co-regulators (coactivators and corepressors) primarily serve as bridges between DNA-bound transcription factors and the basic transcriptional machinery to regulate gene expression (Rosenfeld et al., 2006). Chromosomal translocations that create oncogenic fusion genes involving transcriptional co-regulators are predicted to cause profound changes to normal developmental, homeostatic, and/or cellular identity programs in cancer (Lee and Young, 2013; Mitelman et al., 2019;
Rabbitts, 1994; Tuna et al., 2019).
Molecular properties of the chimeric CRTC1/MAML2 oncoprotein have been extensively characterized and reveal that the t(11;19) chromosomal translocation fuses the coiled-coil domain of CRTC1, which promotes binding to the transcription factor
8 cAMP response element binding protein (CREB), with the strong transcriptional activation domain of MAML2 (Coxon et al., 2005; Wu et al., 2005). Consequently,
CRTC1/MAML2 functions as a rogue coactivator of the transcription factor CREB.
However, its regulatory properties unexpectedly also include gain-of-function interaction with and activation of the transcription factor and proto-oncogene MYC (Amelio et al.,
2014). Unfortunately, the absence of ligand binding sites on these transcription factors renders them impractical targets for developing selective inhibitors (Bishop et al., 2019;
Chen and Koehler, 2020). Thus, efforts have been directed towards defining the downstream pathways reprogrammed by CRTC1/MAML2. While candidate CREB and
MYC target genes have been identified (Amelio et al., 2014; Chen et al., 2015), the subordinate transcriptional pathways dysregulated by CRTC1/MAML2-mediated activation of these transcription factors remain poorly defined.
In this study, we performed transcriptomic profiling of fusion-positive salivary
MEC tumors and report that CRTC1/MAML2 initiates a transcriptional cascade that upregulates the potent growth hormone insulin-like growth factor 1 (IGF-1) via induction of a PGC-1α coactivator alternative splice variant, PGC-1α4. This has functional consequences for the growth, survival, and oncogenic transformation of salivary gland precursors. Elucidating this pathway demonstrates the unique capacity of the
CRTC1/MAML2 coactivator fusion to control complex and extensive transcriptional networks beyond CREB and MYC and provides the first evidence of a role for the PGC-
1α4 splice variant in human cancer pathogenesis. Integrating transcriptional profiling, drug screening and mechanistic data exposed multiple biologic nodes of drug
9 sensitivity, revealing that CRTC1/MAML2-positive MEC is therapeutically vulnerable to
PPARγ inhibition.
2.2 Results
2.2.1 Activation of IGF-1 Signaling is a Hallmark of CRTC1/MAML2-Positive Salivary Mucoepidermoid Carcinoma
IGF-1 receptor (IGF-1R) overexpression and/or ligand-induced activation of downstream signaling occurs in many cancers and is often required by oncogenes to promote transformation and malignancy (Pollak, 2008). Our previous work investigating the pathobiology of CRTC1/MAML2-positive MEC revealed that several genes involved in anabolic, pro-growth receptor tyrosine kinase (RTK) pathways, including IGF-1R signaling, are induced by the CRTC1/MAML2 fusion oncogene (Amelio et al., 2014).
Here, we confirmed that IGF-1 is upregulated >100 fold upon ectopic induction of
CRTC1/MAML2 expression (Table 1). Given the importance of IGF-1 in tumor cell growth and survival (Pollak, 2008; Yu and Rohan, 2000) and in salivary gland growth and development (Amano and Iseki, 1993; Kerr et al., 1995; Ryan et al., 1992; Werner and Katz, 2004), we hypothesized that IGF-1R signaling plays a key role in salivary
MEC tumorigenesis. To validate the clinical relevance of this finding, we first obtained a cohort of eighteen human primary salivary MEC samples from the surgical pathology department at UNC Hospitals (Chapel Hill, NC) and assessed CRTC1/MAML2 status
(Figure 3). In accordance with previous reports (Birkeland et al., 2017), we found that ten of these samples (56%) were CRTC1/MAML2-positive. Next, we performed RNA sequencing on all of these MEC samples (fusion-positive and -negative) to identify differentially expressed genes (DEGs) relative to six normal salivary gland controls.
Unbiased and unsupervised hierarchical clustering revealed that CRTC1/MAML2-
10 positive tumors constitute a gene expression subtype distinct from CRTC1/MAML2- negative tumors (Figure 4). This finding is in agreement with recent studies that suggested that CRTC1/MAML2-positive and -negative MEC are two distinct tumor types that are driven by distinct signaling mechanisms (Kang et al., 2017; Sheth et al., 2020;
Wang et al., 2017). Notably, we identified IGF-1 among 3,971 up-regulated genes that are significantly differentially expressed (fold change > 2 and padj < 0.05) in
CRTC1/MAML2-positive tumors relative to normal salivary glands. Gene set enrichment analysis (GSEA) of curated gene lists within the Molecular Signatures Database
(MSigDB) revealed that IGF-related signaling pathways rank among the top signatures associated with CRTC1/MAML2-positive tumors. Thus, we generated a refined gene list
(Musicant_MEC_CRTC1-MAML2_IGF1) by curating IGF-1 pathway-related genes with a fold change > 2 (padj < 0.05) that reflect a gene signature specific to CRTC1/MAML2- positive MEC tumors (Figures 5A and 5B). We then evaluated the ability of our curated
Musicant_MEC_CRTC1-MAML2_IGF1 gene list to distinguish CRTC1/MAML2-positive
MEC by comparison with the established MSigDB GNF2_IGF1 gene list (Figures 6A and 6B). Importantly, we examined gene expression in these MEC cases by qRT-PCR and found that CRTC1/MAML2 copy number is significantly correlated (r2 = 0.7251; p <
0.0001) with IGF-1 expression (Figure 7). Therefore, these results demonstrate that
CRTC1/MAML2-positive MEC is characterized by increased IGF-1 expression and suggest an important role for IGF-1R signaling in this MEC tumor subtype.
2.2.2 CRTC1/MAML2-Positive MEC Tumor Cell Lines Display Selective Sensitivity to Targeted IGF-1R Inhibition
To investigate the role of constitutive IGF-1 expression and downstream signaling in CRTC1/MAML2-positive MEC cells, we obtained a panel of five MEC
11 (H292, H3118, HMC1, HMC3A, and HMC3B) and three epidermoid (A253, A388, and
A431) cell lines. We first confirmed that all of the MEC cell lines are positive for expression of the CRTC1/MAML2 fusion transcript and that all of the epidermoid carcinoma cell lines are fusion-negative (Figure 8A). Further, all CRTC1/MAML2- positive MEC cell lines display robust IGF-1 expression relative to CRTC1/MAML2- negative cell lines, and this increased IGF-1 expression significantly correlates (r2 =
0.8664; p = 0.0008) with CRTC1/MAML2 expression in fusion-positive MEC cells
(Figures 8B and 8C). Finally, histologic analysis of tumor xenografts confirms that
CRTC1/MAML2-positive salivary MEC cell lines generate significant intra-tumoral cellular heterogeneity, which mimics the mucous, epidermoid, and intermediate cell types characteristic of human salivary MEC, demonstrating that these cell lines are capable of forming representative MEC tumors in vivo (Behboudi et al., 2006; Bell and
Hanna, 2012; Warner et al., 2013). In contrast, fusion-negative epidermoid carcinoma cells generate xenograft tumors with typical solid epidermoid morphology (Figure 9).
Importantly, immunohistochemical staining demonstrated that CRTC1/MAML2-positive tumors are associated with increased IGF-1 production in vivo (Figure 9).
To test whether CRTC1/MAML2-positive MEC cells are dependent on IGF-1 signaling, we performed a small molecule drug screen focused on RTK inhibitors
(RTKis) at concentrations ranging from 10 nM to 10 μM in all five CRTC1/MAML2- positive cell lines relative to a fusion-negative control cell line (Figure 10A). Calculation of the Z’-factor demonstrated robust assay quality such that the distribution between the positive (Bortezimib) and negative (DMSO) controls indicates low likelihood of false positive hits (Z’ > 0.75) (Figure 10B). Multiple independent inhibitors targeting IGF-1R
12 and EGFR emerged as significant hits (p ³ 3SD) in this screen (Figure 10A). Notably,
EGFR has been explored as a therapeutic target, though with limited clinical success
(Chen et al., 2014), and its emergence as a significant hit validated the robustness of our screen results. Compared to the non-MEC fusion-negative control cells, IGF-1R inhibitors induced selective and robust cell death in all CRTC1/MAML2-positive cell lines (Figure 10C), indicating the relative importance of IGF-1/IGF-1R signaling in
CRTC1/MAML2-positive MEC. Dose titrations of two separate IGF-1R inhibitors, BMS-
754807 and picropodophyllin (PPP), confirmed that the CRTC1/MAML2-positive cell lines are sensitive to IGF-1R inhibition, with IC50s for both compounds in the low nM range (Figure 11 and Table 2).
2.2.3 IGF-1R Signaling is Critical for CRTC1/MAML2-Positive MEC Tumor Cell Growth and Survival
We next wanted to examine the functional role of IGF-1R signaling on human
CRTC1/MAML2-positive MEC cell growth and survival via pharmacologic inhibition and/or genetic knockdown (validated in Figures 12A-E). Live-cell, kinetic proliferation assays revealed that wild-type HMC3A cells reach 50% confluency after 28 hr (GC50 =
28 hr), but pharmacologic inhibition of IGF-1R with PPP significantly increases this time to confluency (p < 0.05) relative to control (Figure 13A). To test if blocking IGF-1R impairs the tumorigenic potential of MEC cells to grow as clonogenic colonies ex vivo,
CRTC1/MAML2-positive cells were plated at low density and observed for colony formation. We found that both PPP and BMS-754807 markedly blunted the clonogenic potential of several independent MEC cell lines relative to vehicle-treated cells (Figures
13B and 13C). Several recent studies have documented the existence of cancer stem cell (CSC) subpopulations within various head and neck cancers, including salivary
13 MEC, and these CSCs are associated with malignant potential (Adams et al., 2015;
Curtarelli et al., 2018; Keysar et al., 2017). Since tumor spheroids are uniquely enriched for CSCs, we also quantified the effects of PPP on 3D tumor spheroid formation in
Matrigel and found that sustained IGF-1R inhibitor treatment (7 days) significantly blocks (p < 0.0001) MEC 3D sphere formation (Figure 13D). Finally, we investigated the mechanism underlying the observed decrease in proliferation and tumorigenic potential and found that PPP stimulates a significant dose-dependent increase (p < 0.001) in apoptosis compared to vehicle-treated cells (Figure 13E).
2.2.4 CRTC1/MAML2 Establishes a Synthetic PGC-1α4 Circuit that Regulates IGF-1 Expression in MEC Tumor Cells
To determine the mechanism by which CRTC1/MAML2 regulates IGF-1 expression in MEC cells, we utilized an engineered CRTC1/MAML2-inducible stable cell line (Amelio et al., 2014) (doxycycline-regulated HEK293-CMVTetRTetOC1/M2) (Figure
14A). Overexpression of CRTC1/MAML2 in this cell line significantly increases IGF-1 expression at both the transcript and protein levels (Figures 14B-E). CRTC1/MAML2 coactivates the transcription factor CREB, which in turn binds to specific DNA sequences called cAMP-response elements (CREs). Examination of IGF-1 promoter sequences revealed a non-canonical CRE site, indicating that CRTC1/MAML2 could possibly drive IGF-1 signaling through direct upregulation of IGF-1 transcription.
However, chromatin immunoprecipitation (ChIP) assays failed to identify
CRTC1/MAML2 enrichment at the IGF-1 promoter relative to the control canonical
CREB target gene NR4A2 (Figure 15A).
We previously showed that CREB regulates stress-inducible expression of a
PPARγ coactivator 1a splice variant (PGC-1a4) transcribed from an upstream (~13
14 Kbp) distal promoter of the PGC-1a locus (Bruno et al., 2014), which in turn selectively regulates IGF-1 expression (Ruas et al., 2012). Thus, we performed ChIP assays and identified CRTC1/MAML2 enrichment specifically at this distal PGC-1a promoter (Figure
15A). Moreover, luciferase reporter assays confirmed that CRTC1/MAML2 activates the distal CRE-containing PGC-1a promoter but not the proximal promoter (Figure 15B), suggesting that CRTC1/MAML2 may indirectly regulate IGF-1 expression by controlling
PGC-1a4 expression in MEC cells. We next performed shRNA-mediated
CRTC1/MAML2 knockdowns and found that PGC-1a4 levels are dramatically reduced similar to the control NR4A2, confirming that CRTC1/MAML2 upregulates PGC-1a4 transcription (Figures 15C and 15D). Expression of inducible dominant-negative CREB
(A-CREB) in a stable MEC cell line (HMC3A-PGKTetOn3GTRE3GSA-CREB) also blocked
PGC-1a4 expression, supporting the role of CREB-dependent regulation of the PGC-1a distal promoter by CRTC1/MAML2 (Figure 15E). To examine the kinetics of this expression circuit more closely, we performed a time course experiment in
CRTC1/MAML2-inducible stable cells and found that PGC-1a4 transcripts peak at 24 hrs (~12 hr after CRTC1/MAML2 induction) while IGF-1 transcripts begin to rise at ~24-
36 hrs, but peak 48 hrs after CRTC1/MAML2 induction and 24 hrs after PGC-1a4 upregulation (Figure 16A). Notably, compared to the other PGC-1a transcript isoforms,
CRTC1/MAML2-positive MEC cell lines are all characterized by robust expression of the
PGC-1a4 variant relative to CRTC1/MAML2-negative cell lines (Figures 16B and 16C).
To further validate that the PGC-1a4 variant is responsible for inducing IGF-1 expression, we engineered MEC cell lines with an inducible CRISPR interference
(dCas9-KRAB; CRISPRi) system (Gilbert et al., 2013). Unique sgRNAs that either
15 repress expression of all PGC-1a isoforms or that selectively repress only the PGC-1a1 variant revealed that the canonical PGC-1a1 variant has no effect on IGF-1 regulation
(Figures 17A and 17B). In contrast, repressing transcription of all variant isoforms causes a significant decrease (p < 0.0001) in IGF-1 expression (Figures 17A and 17B), further indicating that PGC-1a4 regulates IGF-1. To confirm that this non-canonical
PGC-1a4 isoform is responsible for promoting IGF-1 transcription, we next generated a
PGC-1a4-inducible stable cell line (doxycycline-regulated HEK293-
PGKTetOn3GTRE3GSPGC-1α4; validated in Figures 18A and 18B) and confirmed that selective overexpression of the PGC-1a4 splice variant is sufficient to upregulate IGF-1 expression, thus acting as the intermediate to CRTC1/MAML2 (Figure 18C). Finally, shRNA-mediated knockdown of PGC-1a4, as well as other components within this signaling circuit (IGF-1R and CRTC1/MAML2) significantly blunts proliferation of
CRTC1/MAML2-positive MEC cells (Figure 19), confirming that PGC-1a4 is both necessary and sufficient for activating the IGF-1 signaling circuit in fusion-positive MEC.
2.2.5 The PGC-1a4 Co-Regulator Activates PPARγ-dependent Transcription of IGF-1 in CRTC1/MAML2-Positive MEC Cells
PGC-1a is a co-activator that binds to the transcription factor peroxisome proliferator-activated receptor gamma (PPARγ) to exert its effects on transcription of metabolic target genes (Semple et al., 2006; Sonoda et al., 2008). Strikingly, analysis of the IGF-1 locus identified three canonical PPAR-response element (PPRE) binding motifs within the IGF-1 promoter region (Figure 20A). To test whether PGC-1a4 co- activates PPARγ to modulate IGF-1 transcription, we first confirmed that overexpression of CRTC1/MAML2 induces transcription of an IGF-1 promoter-driven luciferase reporter
16 comparable to levels of induction observed with a PPRE-driven luciferase reporter
(Figures 20B and 20C). Importantly, functional characterization revealed that both the
PPRE-driven and the IGF-1 promoter-driven luciferase reporter constructs are responsive to the PPARγ agonist GW1929, indicating that the cloned IGF-1 promoter fragment includes the PPRE motifs that recruit PPARγ (Figures 21A and 21B).
Therefore, we next tested whether the PPARγ inverse agonist SR10221, which functions to recruit transcriptional co-repressors to PPARγ that repress target gene transcription, is effective at inhibiting IGF-1 expression. Treatment with SR10221 effectively downregulated PPARγ-mediated transcriptional activity of the IGF1-
Luciferase reporter in a dose-dependent manner to below basal levels, comparable to that observed with the PPRE-Luciferase reporter (Figures 21C and 21D). Similarly, shRNA-mediated knockdown of PPARγ resulted in a robust decrease in IGF-1 expression (Figures 22A and 22B) and potent induction of apoptosis as evidenced by significant increases (p < 0.05) in caspase 3/7 activation (Figure 22C). This indicates that the PGC-1a4:PPARγ circuit controls IGF-1 expression and suggests that drugs targeting PPARγ may provide therapeutic benefit against CRTC1/MAML2-positive salivary MEC.
2.2.6 PPARγ Inverse Agonists Suppress CRTC1/MAML2-Positive MEC Tumor Growth
To investigate the functional role of PPARγ dependency on MEC cell growth and survival, we tested the effects of several PPARγ inverse agonists in our panel of
CRTC1/MAML2-positive MEC cell lines. First, to determine the efficacy of PPARγ inverse agonists as anti-MEC agents, we assessed the effect of three different inverse agonists (SR10221, SR2595, and T0070907) on cancer cell viability using the full panel
17 of CRTC1/MAML2-positive MEC cell lines (Figure 23A and Table 3). Dose titration of these compounds revealed that SR10221 potently reduced cancer cell viability at low micromolar concentrations (half-maximal inhibitory concentration [IC50] ~70 – 150 μM) in
HMC3A cells (Table 3). Further, treatment with SR10221 or SR2595 decreased expression of IGF-1 (Figures 23B and 23C) as well as the canonical PPARγ target genes PGK1 and PKM2 in CRTC1/MAML2-positive MEC cells (Figure 23B). Treatment of CRTC1/MAML2-positive cell lines with sub-lethal doses of SR10221 (½ IC50) significantly decreased proliferation, 2D colony formation, and 3D tumor spheroid growth (Figures 24A-C). Moreover, this decreased growth and tumorigenic potential coincided with potent induction of apoptosis as evidenced by significant increases in caspase 3/7 activation (Figure 24D).
To determine the efficacy of PPARγ inverse agonists in vivo, we first tested the potency of SR10221 in subcutaneous HMC3A tumor xenograft models labeled with our
LumiFluor bioluminescent reporter (Schaub et al., 2015). Tumor xenografts were allowed to grow to a palpable size (approximately 50 mm3) and animals were randomized into two cohorts. These cohorts then received intraperitoneal (i.p.) administration of 20 mg/kg SR10221 or vehicle control once daily for three weeks which resulted in the significant inhibition (p < 0.0001) of HMC3A tumor xenograft growth
(Figures 25A-C). Similarly, we also tested the in vivo potency of SR2595 since it displays superior pharmacokinetic properties, although we administered SR2595 at 60 mg/kg since SR10221 was shown to possess ~2- to 3-fold greater potency in vitro
(Marciano et al., 2015). Treatment with SR2595 at this concentration was generally well-tolerated without body weight loss (Figure 25D), but significantly blocked the
18 growth of CRTC1/MAML2-positive MEC tumor xenografts (Figures 25E-G). Collectively, these results suggest that inhibition of IGF-1 signaling using PPARγ inverse agonists alone or in combination with IGF-1R inhibitors may be a viable therapeutic strategy for the targeted treatment of CRTC1/MAML2-positive MEC.
2.3 Discussion
Recurrent chromosomal translocations that generate novel gene fusions have long been known to have the potential to function as cancer drivers (Rabbitts, 1994,
2009; Rowley, 2001). However, the advent of fusion detection algorithms applied to
“omics”-level data has not only enabled the discovery of additional gene fusions composed of splicing factors, signal transduction proteins, transcription factors, and/or transcriptional co-regulators (Fernandez-Cuesta et al., 2015; Jang et al., 2020; Kumar-
Sinha et al., 2015) but has also aided in elucidating the direct functional consequences of these fusions on cellular processes such as signaling and gene expression
(Latysheva et al., 2016; Lee and Young, 2013). Interestingly, these aberrant processes are frequently unique to distinct gene fusions which may in turn be characteristic of specific cancer subtypes (Fishbein et al., 2017; Gao et al., 2018). CRTC1/MAML2 represents one such recurrent gene fusion composed of two transcriptional co- activators that is pathognomonic for mucoepidermoid carcinoma (O’Neill, 2009; Tonon et al., 2003). Here, we report that the CRTC1/MAML2 fusion oncogene directs profound reprogramming of transcriptional networks and establishes a synthetic IGF-1 signal circuit through aberrant expression of an alternative PGC-1a splice variant in salivary
MEC (Figure 26).
19 Despite being the most common salivary gland malignancy, MECs are relatively rare compared to other head and neck cancers and consequently the pathobiology of
CRTC1/MAML2 remains poorly understood. While it is clear that patients with advanced stage and/or high-grade tumors display significantly worse overall survival (Chen et al.,
2013), the prognostic value of fusion status has recently been challenged (Seethala and
Chiosea, 2016). Most MEC patients are treated with surgical excision, which is sometimes accompanied by adjuvant radiotherapy (Nance et al., 2008). Unfortunately, resistance to chemoradiotherapies and a lack of targeted therapies for recurrent/metastatic disease poses significant challenges to treating patients with aggressive MEC tumors. For example, recent studies have indicated that paclitaxel, trastuzumab, or RTK inhibitors may be viable treatments for fusion-positive MEC, however preliminary results show only modest responses to these treatments (Coca-
Pelaz et al., 2015; McHugh et al., 2012) emphasizing the need for performing thorough molecular profiling studies to aid in the development of new targeted therapies.
As a transcriptional coactivator gene fusion, CRTC1/MAML2 interacts with and activates two master transcription factors: CREB and MYC (Amelio et al., 2014;
Tasoulas et al., 2019). Unfortunately, intrinsic disorder and the lack of available binding pockets has made designing small molecules that can effectively target transcription factors a major challenge (Bishop et al., 2019; Chen and Koehler, 2020; Wachtel and
Schäfer, 2018). In addition, two recent genomics studies revealed that CRTC1/MAML2 is the primary oncogenic driver in most salivary MECs and that these fusion-positive cases lack other cooperating mutations that occasionally serve as actionable targets in other cancers (Kang et al., 2017; Wang et al., 2017). In this study, we set out to identify
20 the most significantly altered and clinically relevant target genes and pathways downstream of CRTC1/MAML2 that can be exploited to develop novel therapeutic approaches for treating patients with salivary MEC. Using transcriptomic profiling, we show that CRTC1/MAML2 expression is correlated with increased IGF-1 expression and pathway activation in salivary MEC tumors. Autocrine and paracrine activation of the IGF-1 pathway is known to promote growth and survival of multiple tumor types
(Pollak, 2008; Yu and Rohan, 2000). Notably, endocrine IGF-1 signaling is broadly involved in normal tissue growth and development through its stimulation of IGF-1R and has been shown to play critical roles in regulating salivary gland growth and development and in stimulating regeneration following injury (Amano and Iseki, 1993;
Kerr et al., 1995; Ryan et al., 1992; Werner and Katz, 2004), thus supporting an important role for aberrant IGF-1 expression in salivary tumor development. Through the use of bioinformatics analyses and small molecule drug screening, we were successful in pinpointing IGF-1R as an actionable target in salivary MEC, suggesting that treatments directed against IGF-1R may provide therapeutic benefit. However, despite early excitement surrounding the potential utility of these treatments for many cancers (Gualberto and Pollak, 2009), targeting the IGF axis has yielded disappointing results in clinical trials (Chen and Sharon, 2013; Denduluri et al., 2015). Therefore, we also sought to explore the underlying mechanism governing IGF-1 activation by
CRTC1/MAML2 in salivary MEC and identified unexpected rewiring of this growth factor signal circuit coordinated by aberrant expression of an alternatively spliced target gene of CRTC1/MAML2, the PGC-1a4 variant.
21 PGC-1a has been shown to exert oncogenic (Frattini et al., 2018) or tumor suppressive (Torrano et al., 2016) activity depending on cell type and context- dependent metabolic cues (Mastropasqua et al., 2018; Sancho et al., 2015; Xing et al.,
2017). Notably, the PGC-1a family of transcriptional coactivators is composed of an expanding list of transcript isoforms generated by multiple promoters and alternative pre-mRNA splicing, and these resulting protein variants are known to display unique functional properties (Martínez-Redondo et al., 2015, 2016). The majority of the current literature focuses on the role of the major PGC-1a transcript, PGC-1a1, in various cancers including melanoma (Luo et al., 2016, 2020; Vazquez et al., 2013) and prostate cancer (Torrano et al., 2016; Valcarcel-Jimenez et al., 2019; Wallace and Metallo,
2016). However, rare, alternatively spliced transcripts such as PGC-1a4 are emerging as key players in multiple cancer types and thus represent an attractive target for pharmacological intervention (Kimes et al., 2014; Oltean and Bates, 2014; Wang and
Lee, 2018; Zhang and Manley, 2013). To date, though, the specific role of these PGC-
1a alternative splice variants in cancer has remained elusive. We demonstrate that the
CRTC1/MAML2 fusion selectively induces CREB-dependent expression of the PGC-
1a4 splice variant, which we and others have shown is associated with IGF-1 expression in anabolic skeletal muscle (Bruno et al., 2014; Ruas et al., 2012; Tasoulas et al., 2019). We confirm here that CRTC1/MAML2 directs similar anabolic pro-growth and pro-survival signaling by coordinating this autocrine PGC-1a4 – IGF-1 signal circuit in fusion-positive salivary MEC.
A model that may explain the transcriptional and pre-mRNA splicing effects of
CRTC1/MAML2 on PGC-1a4 is that this fusion acquires a strong transcriptional
22 activation domain from MAML2 and retains the CREB binding domain of CRTC1, while the splicing domain normally present in full-length CRTC1 is deleted (Amelio et al.,
2009; Tasoulas et al., 2019). Importantly, the PGC-1a4 variant retains the activation domain common to other PGC-1a transcript isoforms, which mediates binding to nuclear hormone receptor transcription factors such as PPARγ (Li et al., 2008;
Martínez-Redondo et al., 2015; Sonoda et al., 2008). Similar to PGC-1a, dichotomous roles for PPARγ have been described with evidence pointing to oncogenic activity in some cancers but tumor suppressor actions in other cancers. This is supported by data using PPARγ agonists or antagonists, respectively (Goldstein et al., 2017; Kardos et al.,
2016; Khandekar et al., 2018; Zou et al., 2019). We demonstrate here that PGC-1a4 upregulates IGF-1 in a PPARγ-dependent manner in CRTC1/MAML2-positive salivary
MEC, which sensitizes these tumor cells to treatment with PPARγ inverse agonists.
Given the limitations of anti-IGF1R monotherapy, the identification of PPARγ inverse agonists as potential therapeutics for the treatment of MEC is particularly exciting. A limitation of this study, however, is that therapeutic effects achieved with the PPARγ inverse agonists SR10221 and SR2595 were in the low micromolar range, despite the fact that functional ligand binding and induced PPARγ conformational changes are possible in the nanomolar range (Marciano et al., 2015; Shang et al., 2020). These differences in the inverse agonist concentrations required to achieve corepressor- dependent PPARγ repression may be due to higher binding affinity of the PGC-1a4 variant with PPARγ and/or increased levels of competing endogenous ligands (e.g. esterified lipids) in salivary MEC cells. Interestingly, our bioinformatic analyses revealed that several metabolic pathways with the potential to directly influence endogenous
23 ligand levels, including lipid biosynthesis and modification pathways, are significantly elevated in CRTC1/MAML2-positive salivary MEC.
In summary, our study identifies the CRTC1/MAML2 fusion oncogene as a master regulator of transcriptional networks that promote tumorigenic processes in part via a novel PPARγ-dependent PGC-1a4 – IGF-1 signaling circuit (Figure 26). Therefore, elucidation of the CRTC1/MAML2-regulated PGC-1a4-dependent transcriptional program may open new avenues for the identification of other pathways that can be exploited in order to stratify patients suitable for precision therapies. Our findings establish that targeting this ‘synthetic’ signaling circuit via IGF-1R inhibition or PPARγ inverse agonism, either individually or in combination, is a potential therapeutic option for patients with CRTC1/MAML2-positive salivary MEC.
24
CHAPTER 3: DEVELOPING AN AUTOCHTHONOUS MOUSE MODEL OF MUCOEPIDERMOID CARCINOMA
3.1 Introduction
In his 1971 State of the Union address, President Richard Nixon declared a “war on cancer”, proposing that $100 million be funneled into “an intensive campaign to find a cure for cancer” (Coleman, 2013). Though Nixon was misguided in his use of the word
“cure”– we now know that there is no single cure for all cancers – the aftermath of his declaration was marked by many successes including improved detection, diagnosis, and treatment options (DeVita Jr, 2004; Sporn, 1996). As a result, today’s cancer patients fare much better than those fifty years ago – with improved survival times and quality of life – though there is significant room for improvement. The 2016 launch of the
Obama administration’s cancer “moonshot” project (Kaiser and Couzin-Frankel, 2016;
Singer et al., 2016) is helping to promote significant advances in early disease detection and personalized medicine, however patients with advanced disease often do not survive beyond five years from their initial diagnosis and may respond poorly to novel therapeutics. Therefore, it is important that the field of cancer biology makes use of the best available tools to investigate the mechanisms underlying tumorigenesis and to test potential new therapies (Sharpless and DePinho, 2006).
Direct study of human tumor samples is important and can provide insights into certain tumor-associated molecular events, however these endeavors can only provide a snapshot of the tumorigenic process. In order to elucidate the complexities of disease
25 progression and test novel therapeutics, the development of preclinical model systems is critical (Walrath et al., 2010). In cancer biology, one of the most commonly used model systems is the immortalized cell line. Cell lines are typically derived from human tumor biopsies/resections that have been adapted to grow in highly controlled lab culture settings for use in in vitro assays. However, cells grown in a dish are incapable of recapitulating important aspects of tumor biology, including the stromal microenvironment and intratumoral heterogeneity (Bleijs et al., 2019), both of which can have significant effects on tumor behavior and patient outcomes. Therefore, many cancer studies make use of in vivo systems such as xenograft models and/or genetically engineered mouse models (GEMMs).
In a xenograft model, immortalized human cells or primary resected tumor samples are subcutaneously implanted into immunodeficient mice. The xenografted tumor can then grow in an in vivo environment that more accurately recapitulates aspects of the human tumor microenvironment, such as access to both growth factors and vasculature. Xenograft models are relatively simple and inexpensive, making them valuable tools for both academic and industrial research (Sharpless and DePinho,
2006). However, these models present their own set of problems. For example, xenograft studies are often performed on a relatively short time scale, which does not provide enough time for the development of a true cancer stromal microenvironment and may overestimate a tumor’s susceptibility to drug treatment (Walrath et al., 2010).
Moreover, xenografts are typically established using an immortalized cell line, which is monoclonal by nature. The process of establishing this monoculture may select for a cell line that, when xenografted, generates tumors that underrepresents the true
26 complexity and heterogeneity of the original human tumor (Bleijs et al., 2019). Finally, commonly used xenograft models rely on immunodeficient mice, which lack a functional immune system and thus are not amenable to the testing of immunotherapies and immunomodulatory agents, both of which have recently manifested as an exciting area of research (Frese and Tuveson, 2007; Leon et al., 2020; Sharpless and DePinho,
2006).
Thus, since the mid-1980s (Sharpless and DePinho, 2006), focus has shifted to mice that have been genetically modified to predispose them to tumors mimicking those seen in humans, both in terms of genetic makeup and site of origin. GEMMs are superior to xenograft models in that they allow for the interaction of host immune cells with the tumor throughout all stages of malignancy. They also allow for the co- development of the tumor stroma with the tumor itself (Walrath et al., 2010). GEMM technology has advanced significantly since its initial introduction. While early models often involved whole-body, constitutive knockout or overexpression of target genes, often resulting in issues such as embryonic and neonatal lethality (Chavez-Reyes et al.,
2003; Dang et al., 2003; de Oca Luna et al., 1995; Wakabayashi et al., 2003), current models are more refined, combining cell-specific genetic alterations with inducible systems such as Cre/LoxP and Flp-Frt technology (Lewandoski, 2002; Nagy, 2000;
Rodriguez et al., 2000; Sorrell and Kolb, 2005) to more accurately model the spontaneous, localized genetic changes that lead to human tumor development.
Despite these advancements in GEMM design, there is a dearth of available models for rare, understudied cancers. MEC, which is the most common type of malignant salivary gland tumor but affects only 1 in every 100,000 people (O’Neill,
27 2009), is one such disease where a validated pre-clinical GEMM has not been generated. Unfortunately, the extreme cellular heterogeneity observed in MEC tumors has hindered the development of a GEMM model by obscuring the cell of origin for this tumor type.
The tumor “cell of origin” is a normal cell in which a cancer-causing mutation or genetic hit first occurs (Visvader, 2011). This cell can then proliferate rapidly and form a full-blown tumor. Importantly, the concept of a tumor cell of origin should be distinguished from that of a cancer stem cell (CSC), which is a self-renewing cell that can give rise to the full complement of tumor cell subpopulations. In other words, the cell of origin is a tumor-initiating cell, while CSCs are tumor-propagating cells (Visvader,
2011). In many cancer types, the cell of origin and the CSC are phenotypically distinct.
Identification and characterization of the cell of origin is an essential undertaking in the field of cancer research, as a better understanding of this cell type can aid in earlier diagnosis, hint at future tumor behavior, and even inform treatment strategies (Hoadley et al., 2018). Further, establishing the appropriate cell of origin is crucial for developing accurate disease models, as overexpression of an oncogene might drive malignant processes in one cell type but not in another (Bailleul et al., 1990; Brown et al., 1998;
Perez-Losada and Balmain, 2003). While the histopathology of an individual tumor may sometimes provide insights into the cell of origin, MEC tumors exhibit significant cellular heterogeneity (Coca-Pelaz et al., 2015; Schwarz et al., 2011), and so it has been challenging to determine what cell type is responsible for tumorigenesis in this cancer.
The salivary glands are highly branched tree-like structures (Patel et al., 2006) composed of three main cell types: ductal, acinar, and myoepithelial cells (Figure 1).
28 Each of these differentiated cell populations is maintained by slow turnover and repopulation via lineage-restricted progenitor cells (Rocchi et al., 2021; Zajicek et al.,
1989). A putative stem cell population can also undergo accelerated cell division and differentiation in order to repopulate the salivary gland following extensive damage or injury. Work from other groups has suggested that these stem cells can be found within either the intercalated duct or acinar cell populations of the mature salivary glands.
These claims are based on the roles of the two cell populations in active cell division and repopulation of damaged salivary glands (Emmerson et al., 2017; Holmberg and
Hoffman, 2014; Knosp et al., 2012; Kwak et al., 2016; van Luijk et al., 2015; Maimets et al., 2016; May et al., 2018; Weng et al., 2018).
Stem cells are frequently favored as potential tumor cells of origin, as their plasticity, longevity, and capacity for self-renewal could conceivably contribute to the development of heterogeneous tumors (Visvader, 2011). These cells may reactivate dormant developmental signaling pathways in order to drive the rapid cellular proliferation seen in many malignancies. Thus, our hypothesis is that CRTC1/MAML2- positive MEC arises in an early salivary progenitor cell that is capable of differentiating into the multiple cell types seen in MEC tumors. In this study we utilized acinar and ductal cell-specific promoters (Maruyama et al., 2016) to drive inducible CRTC1/MAML2 overexpression and Tp53 loss in mice in order to (1) characterize the MEC cell of origin and (2) develop a tool for the future study of MEC tumorigenesis and treatment in vivo.
29 3.2 Results
3.2.1 Targeted Transgene Expression is Achieved via Cell Type-Specific Promoters
We developed multiple transgenic lines to study the contribution of various salivary gland cell types to MEC development. Each of these lines allowed for inducible expression of CRTC1/MAML2 (LoxP-STOP-LoxP-CRTC1/MAML2, abbreviated as
C1/M2) and/or our bioluminescent LumiFluor reporter (LoxP-STOP-LoxP-LumiFluor, abbreviated as LumiFluor) (Schaub et al., 2015). Expression of each transgene was controlled by a tamoxifen-inducible Cre recombinase (CreERT2) whose expression was driven by the promoter of a cell type-specific gene: Dcpp1, Mist1, or Krt14 (abbreviated as DcppCre, MistCre, and K14Cre, respectively) (Maruyama et al., 2016; Shi et al., 2009;
Vasioukhin et al., 1999) (Figure 27). Dcpp1 is expressed in the serous demilune cells of the sublingual gland and also in the intercalated ductal cells (Bekhor et al., 1994), which have been suggested as the location of a salivary gland stem cell population. Mist1 is expressed in salivary gland serous and seromucous acinar cells as well as in pancreatic acinar cells and secretory cells of the stomach, prostate and seminal vesicle (Pin et al.,
2000, 2001). Krt14 is expressed in basal cells of the salivary intercalated, striated, and excretory ducts and is also expressed in basal epithelial cells (Su et al., 1993).
We first assessed transgene activation by measuring bioluminescent signal in the head and neck region of MistCre-LumiFluor, K14Cre-LumiFluor, DcppCre-LumiFluor, and control mice at various time points (Figure 28) following tamoxifen administration.
K14Cre-LumiFluor animals displayed bioluminescent signal in the oral and salivary gland regions but also displayed significant transgene activation in epithelial tissues, including the paws and ears. MistCre-LumiFluor animals displayed activation specifically in the
30 submandibular and sublingual gland region, while DcppCre-LumiFluor animals did not exhibit any significant bioluminescent signal in the salivary glands at all.
We then analyzed CRTC1/MAML2 transcript levels in fresh-frozen salivary gland tissue from DcppCre-C1/M2, MistCre-C1/M2, K14Cre-C1/M2, and control animals to determine whether transgene activation was occurring in the submandibular, sublingual, or parotid glands. Within 4 months of tamoxifen administration, DcppCre-C1/M2, MistCre-
C1/M2, and K14Cre-C1/M2 animals all displayed robust submandibular gland expression of CRTC1/MAML2 and low levels of expression in the sublingual and parotid glands
(Figure 29).
3.2.2 CRTC1/MAML2-Positive Salivary MEC Originates in a Ductal Cell
Following confirmation of CRTC1/MAML2 expression in each transgenic mouse line, we wanted to investigate the histological and potential tumorigenic effects of oncogene overexpression in each of the three major salivary glands. Though we achieved robust CRTC1/MAML2 expression in the salivary glands of MistCre-C1/M2 mice, tissue samples from these animals did not display any histological differences compared to control animals (Figure 30). However, oncogene induction within the intercalated ductal cells and serous demilune cells (DcppCre-C1/M2 mice) resulted in slight ductal hyperplasia and dysplasia (Figure 30). Similarly, CRTC1/MAML2 induction in the basal duct cells (K14Cre-C1/M2 mice) did not result in tumor formation but did induce a dysplastic/hyperplastic morphology (Figure 30).
Notably, overexpression of CRTC1/MAML2 alone failed to induce MEC formation independent of the specific Cre driver allele (Figure 31). These results indicate that
31 oncogene induction specifically affects the ductal cell population but is not sufficient to drive MEC tumor formation.
Given these results, we wanted to generate a cell line that could be manipulated in vitro to investigate the ductal cell-specific effects of CRTC1/MAML2 expression. We therefore obtained an SV40-immortalized human salivary ductal cell line (Gao et al.,
2014) and engineered it for constitutive overexpression of our LumiFluor bioluminescent reporter (PGKLumiFluor) (Schaub et al., 2015) and doxycycline-inducible overexpression of
tdTO C1/M2 either a TdTomato reporter (PTight ) or the CRTC1/MAML2 oncogene (PTight )
tdTO LumiFluor (Figure 32A). Validation of the Ductal-PTight PGK cell line by flow cytometry revealed that administration of even 0.1 μg/mL doxycycline potently upregulated
TdTomato expression (Figures 32B and 32C). Similarly, validation of the Ductal-
tdTO C1/M2 PTight PGK cell line revealed that doxycycline administration induced
CRTC1/MAML2 overexpression at both the transcript and protein (Figure 32D) levels.
Finally, luciferase assays demonstrated that CRTC1/MAML2 induction in these cells activated the distal PGC-1a promoter (Figure 32E), suggesting that induced expression
tdTO C1/M2 of CRTC1/MAML2 in salivary ductal cells (Ductal-PTight PGK ) recapitulates the
PGC-1a4 – IGF-1 signaling circuit recently described by our lab (Musicant et al., 2021).
3.2.3 CRTC1/MAML2 Overexpression and TP53 Tumor Suppressor Loss Collectively Drive MEC Tumor Formation
Salivary MECs are generally characterized by a low mutational burden. In fact, the recurrent t(11;19) chromosomal translocation characteristic of the majority of salivary MECs is often the sole cytogenetic abnormality in these tumors. However, recent genomics studies revealed that fusion negative tumors frequently harbor mutations that functionally disrupt the tumor suppressor genes TP53 and/or CDKN2A
32 (Jee et al., 2013; Kang et al., 2017; Sheth et al., 2020; Wang et al., 2017). Our own
RNA sequencing data reveals that CRTC1/MAML2-positive human MEC tumors possess functionally wild-type TP53 but are characterized by decreased TP53 gene expression and display significantly increased p53 dysregulation (Figure 33), as measured using a breast cancer-derived p53 gene signature (Troester et al., 2006).
Thus, we explored the effects of heterozygous and homozygous TP53 (p53f/+ and p53f/f) or CDKN2A (p16f/+ and p16f/f) loss in our K14Cre-C1/M2 transgenic mouse line.
K14Cre-C1/M2-p16f/+, K14Cre-C1/M2-p16f/f, and K14Cre-C1/M2-p53f/+ animals all presented histologically with salivary gland ductal hyperplasia and dysplasia similar to that seen in K14Cre-C1/M2 mice but did not develop MEC-like tumors. However, a subset of animals with combined CRTC1/MAML2 overexpression and homozygous p53 loss (K14Cre-C1/M2-p53f/f) formed salivary tumors at approximately 4-5 months following tamoxifen administration (Figure 34A). These tumors appeared to originate from the submandibular gland (Figures 34B and 34C) and displayed all of the histological features characteristic of human salivary MEC (i.e. presence of mucous, epidermoid, and intermediate cell types) (Figures 34D and 34E). This indicates that CRTC1/MAML2- positive salivary MEC arises from a cell population within the salivary duct and that murine MEC tumor development depends on both CRTC1/MAML2 expression and p53 loss.
3.2.4 Murine Salivary MEC Tumors Recapitulate CRTC1/MAML2-Positive Human Salivary MEC Genetics
To determine the translational relevance of our GEMM model, we performed
RNA sequencing on fresh-frozen tissue from each of four murine salivary MEC tumors obtained from K14Cre-C1/M2-p53f/f mice. We also sequenced four matched normal
33 submandibular gland tissues from CRTC1/MAML2-negative mice. Gene set enrichment analysis (GSEA) revealed that several tumor-supportive pathways, including the
PI3K/AKT pathway and several metabolic pathways, were altered in the murine MEC tumor samples compared with normal tissues (Figure 35). We also wanted to investigate whether our previously identified PGC-1a4/IGF-1 signaling axis (Musicant et al., 2021) was active in these MEC tumors. Given the difficulties in detecting rare, alternatively spliced transcripts such as PGC-1a4 in RNA sequencing data, we instead evaluated the ability of our curated Musicant_MEC_CRTC1-MAML2_IGF1 gene list
(Musicant et al., 2021) to distinguish CRTC1/MAML2-positive murine tumor tissues
(Figure 36). Our data indicated a remarkable ability of this gene list to differentiate between the malignant and non-malignant sample populations.
Gene expression profiles vary between MEC tumors of different grades (Figure
37A). Thus, we asked whether the MEC tumors generated in our GEMM more closely resembled low, intermediate, or high-grade human MEC tumors. For this analysis, we compared RNA sequencing data from the three GEMM tumors with the highest sample quality to ten previously sequenced CRTC1/MAML2-positive human MEC tumors
(Musicant et al., 2021). Correlation scores were assigned based on similarities in gene expression profiles between each GEMM sample and the low, intermediate, or high- grade CRTC1/MAML2-positive human MEC samples, with a score of 1.0 indicating perfect correlation. These data indicated that two of the three GEMM tumors more closely resembled high-grade human MEC, while the remaining GEMM tumor was most similar to low-grade human MEC (Figure 37B). These results indicate that our K14Cre-
C1/M2-p53f/f mouse model faithfully recapitulates several histological and genetic
34 characteristics of human salivary MEC and that we may be able to use this model to investigate both low and high-grade MEC.
3.3 Discussion
The ability to model oncogenic processes in animals is valuable for a number of reasons. GEMMs in particular can provide insight into molecular processes that drive tumor initiation, progression, and even metastasis. However, the accurate modeling of a particular tumor type requires a working knowledge of the cell of origin of that tumor.
The cell of origin is a normal cell that acquires a cancer-promoting mutation which is then propagated to all daughter cells, forming a tumor (Visvader, 2011). Stem cells are often investigated as potential tumor cells of origin due to their innate plasticity and capacity for self-renewal. Though several studies have utilized injury models to investigate the putative stem and progenitor cell populations in normal salivary glands
(Emmerson et al., 2017; Holmberg and Hoffman, 2014; Knosp et al., 2012; Kwak et al.,
2016; van Luijk et al., 2015; Maimets et al., 2016; May et al., 2018; Weng et al., 2018), no study has yet explored the contributions of these cell types to the development of salivary MEC. Therefore, in this study we developed transgenic lines that could inducibly express the CRTC1/MAML2 oncogene in each of several different putative stem cell populations, including the acinar and intercalated ductal cell populations. Our results indicate that CRTC1/MAML2-positive salivary MEC arises in a ductal cell type.
Here, we have developed a mouse model of MEC that faithfully recapitulates both the genetic and the histologic features of disease seen in humans. This GEMM was engineered to inducibly express the CRTC1/MAML2 oncogene, which characterizes over 50% of human MEC tumors, irrespective of tumor grade (O’Neill,
35 2009; Seethala and Chiosea, 2016; Seethala et al., 2010; Tonon et al., 2003). As MEC is frequently characterized by p53 dysregulation, we also engineered our GEMM to feature inducible TP53 loss (Jee et al., 2013; Kang et al., 2017; Sheth et al., 2020;
Wang et al., 2017). Finally, transgene expression in our model was driven by a tamoxifen inducible Cre recombinase under the control of a Krt14 promoter. While Krt14 successfully targets a subset of progenitor cells in the salivary gland ducts, it is also expressed in a variety of other epithelial tissues including the oral mucosa, digestive tract, and skin (Chu and Weiss, 2002; Su et al., 1993). Thus, our current method of activating Krt14-CreERT2 via intraperitoneal injection of tamoxifen also leads to whole- body transgene activation. While salivary MEC tumors have developed in several
K14Cre-C1/M2-p53f/f mice, many additional mice of the same genotype reached endpoint far before tumor formation due to the appearance of severe off-target effects, including oral and skin lesions with significant keratinization (Figures 38A-C). We also stained tissues from these animals with mucicarmine to identify the presence of mucous cells that could indicate the presence of MEC-like disease. All oral and skin lesions from these animals were negative for mucicarmine staining, confirming that these keratinized lesions are unlikely to be MEC in origin (Figures 38D-E).
Future studies will attempt to reduce off-target effects of oncogene induction within the cutaneous epithelia through targeted delivery of tamoxifen. Targeted delivery may be achieved using several different methods, including trans-dermal injection or retroductal cannulation via the Wharton’s duct (Varghese et al., 2018). To test the first of these two methods, we performed a pilot study in which we delivered adenovirus expressing either Cre recombinase (Ad-Cre) or a GFP control (Ad-GFP) to the
36 submandibular gland via injection directly through the skin. This study was performed in animals expressing the LSL-LumiFluor transgene, thus injection of Ad-Cre but not Ad-
GFP should result in expression of the LumiFluor bioluminescent reporter. Our results revealed that trans-dermal injections can successfully target the submandibular gland, as we observed a bright bioluminescent signal in the salivary region as early as three days after Ad-Cre injection (Figure 39). This method is relatively simple to perform and would appear to be an ideal method of targeted transgene activation. However, the bioluminescent signal seen in mice injected with Ad-Cre faded within two weeks. This indicates that the injected Ad-Cre may have induced recombination in a population of non-stem/progenitor cells. Following the injury sustained upon transdermal injection, regeneration of the gland was therefore achieved by LumiFluor-negative cells and bioluminescent signal was lost. The Wharton’s duct cannulation procedure is more technically difficult but causes less damage to the salivary glands and thus may result in more robust, prolonged transgene activation. This method will be explored in future studies.
The ability to monitor salivary MEC disease progression from pre-neoplastic to metastatic disease in a GEMM will allow for the dissection of key genes and signaling pathways driving tumor cell proliferation and migration. During development, the salivary glands undergo a process called branching morphogenesis, which is mediated in part by increased activation of the FGFR and IGF-1R signaling cascades (Amano and
Iseki, 1993; Hoffman et al., 2002; Kerr et al., 1995; Ryan et al., 1992; Werner and Katz,
2004). Dysregulation of these growth factor signaling pathways, most notably the IGF-
1R signaling cascade, is associated with multiple cancer types, including salivary gland
37 cancers (Babina and Turner, 2017; Musicant et al., 2021; Pollak, 2008; Yu and Rohan,
2000). Several studies have indicated that reactivation of stem cell and fetal gene expression programs may drive tumorigenesis and may be correlated with increased tumor aggressiveness (Ben-Porath et al., 2008; Kim and Orkin, 2011; Spike et al.,
2012). We have recently obtained several RNA sequencing data sets from both human and mouse fetal salivary glands and are currently investigating these datasets along with our own GEMM and human MEC tumor datasets to determine if a similar fetal gene expression program is being reactivated in salivary MEC.
While further refinement of this newly developed GEMM is required to reduce mortality arising from off-target effects, this mouse model provides a much-needed tool to investigate many molecular questions in the field of salivary MEC biology.
38
CHAPTER 4: FUTURE DIRECTIONS IN MUCOEPIDERMOID CARCINOMA RESEARCH
The field of mucoepidermoid carcinoma research has made great strides in recent years. From transcriptomic sequencing studies (Jee et al., 2013; Kang et al.,
2017; Musicant et al., 2021; Sheth et al., 2020; Wang et al., 2017) to the identification of potential therapeutic vulnerabilities (Chen et al., 2014; Musicant et al., 2021; Wagner et al., 2018), the abundance of new data has provided hope to those currently battling this disease. However, there are a number of topics which remain to be addressed in future studies.
Over half of all MEC cases are characterized by a t(11;19)(q14-21; p12-13) chromosomal rearrangement that produces the CRTC1/MAML2 fusion oncogene
(O’Neill, 2009; Seethala and Chiosea, 2016; Seethala et al., 2010; Tonon et al., 2003).
The CRTC1/MAML2 protein product has been shown to function as a rogue coactivator of a number of transcription factors including CREB, MYC, and AP-1 (Amelio et al.,
2014; Canettieri et al., 2009; Coxon et al., 2005; Wu et al., 2005). Through these interactions, CRTC1/MAML2 activates a myriad of tumor-supportive cascades, including those acting through EGFR (Chen et al., 2014) and IGF-1R (Musicant et al., 2021), and preclinical studies have indicated that therapeutics targeting these pathways may be beneficial for the treatment of salivary MEC. However, recent work from our lab has revealed that targeting the EGFR pathway specifically does not induce MEC tumor cell death (Parag-Sharma et al., 2021). Instead, EGFR inhibitors such as Erlotinib exert
39 cytostatic effects, causing MEC cells to enter a quiescent but viable state that can be reversed upon drug withdrawal. Therefore, EGFR inhibitors are not appropriate as single-agent MEC therapeutics, as they can halt tumor progression but cannot reduce tumor volume. Rather, our recent study demonstrates that combined treatment approaches using EGFR inhibitors with a sensitizing agent, such as a histone deacetylase (HDAC) inhibitor, is more clinically relevant (Parag-Sharma et al., 2021).
Similarly, though our lab has revealed that MEC cells are sensitive to IGF-1R inhibition
(Musicant et al., 2021), preclinical studies have revealed that tumors may develop resistance to RTK inhibitors and clinical trials have shown disappointing responses to these agents (Chen and Sharon, 2013; Coca-Pelaz et al., 2015; Denduluri et al., 2015;
McHugh et al., 2012). Therefore, our work suggests that IGF-1R inhibition should be replaced by or combined with other targeted therapies, such as PPARγ inverse agonists
(Musicant et al., 2021). Future molecular studies should focus on dissecting the relative importance of individual growth factor signaling cascades in order to identify those most promising for therapeutic targeting.
Early studies that identified the CRTC1/MAML2 fusion in human tumor samples correlated fusion status with tumor grade. Specifically, the CRTC1/MAML2 fusion was identified more frequently in low grade tumors and was correlated with improved prognosis (O’Neill, 2009). However, improved MEC grading systems have revealed that many tumors originally characterized as high grade, fusion-positive MEC are not MEC at all. Instead, these tumors are often reclassified as adenosquamous or salivary duct carcinomas (Chenevert et al., 2011; Chiosea et al., 2012; Seethala et al., 2010; Tirado et al., 2007). Subsequent reanalysis of CRTC1/MAML2 positivity rates in low,
40 intermediate, and high grade MEC indicate that the fusion event occurs at roughly equal rates in all tumor grades and is not correlated with disease outcome (Cipriani et al.,
2019b; Saade et al., 2016; Seethala and Chiosea, 2016). However, it is unclear whether the downstream pathways affected by CRTC1/MAML2 are differentially regulated in various MEC grades. For example, are MYC-regulated stemness or apoptosis pathways dysregulated more frequently in high grade tumors? Or are CREB-regulated proliferation- and differentiation-associated pathways dysregulated more frequently in low grade tumors? Future studies should explore differential gene expression programs in relation to tumor grade. The findings from such analyses could be informative when considering an individual’s prognosis and treatment options.
On a broader level, it is interesting to note that in all fusion-positive MEC patients, the breakpoints producing the CRTC1/MAML2 fusion oncogene are nearly identical at the genome level and indistinguishable at the protein level. While many studies have explored the impact of this translocation on cellular function and tumor progression, few studies have investigated the mechanisms that contribute to this specific genomic alteration. The limited investigations into susceptibility to chromosomal rearrangements have largely been performed in the context of hematological malignancies. These studies have suggested that most translocations occur due to erroneous repair of double-strand breaks (DSBs) via non-homologous end joining
(NHEJ) (Boehm et al., 1989; Lin and Scott, 2012; Novo and Vizmanos, 2006; Obe et al.,
2002; Povirk, 2006). Additionally, these studies have indicated that DSBs and aberrant
NHEJ are more likely to occur in regions of active transcription or increased chromatin accessibility (Mitelman et al., 2019). However, there are currently no studies in which
41 these hypotheses have been tested in the context of the t(11;19) translocation observed in MEC.
Finally, it will be important to focus future efforts on refining the GEMM described in Chapter 3. Whole-body oncogene activation in our K14Cre-C1/M2-p53f/f mouse model resulted in significant off-target effects within the cutaneous epithelia. To reduce these effects, we propose the targeted delivery of tamoxifen via either trans-dermal injection into the submandibular gland or ductal cannulation via the Wharton’s duct (Varghese et al., 2018). Our pilot study investigating the feasibility of the former method demonstrated successful targeting of the submandibular gland but short-lived recombination. The Wharton’s duct cannulation procedure is currently being explored in our lab but is quite technically difficult and will require a significant investment in terms of person hours. Despite these challenges, successful targeted tamoxifen delivery would deliver a substantial reward. A reliable mouse model that recapitulates both the genetic and microenvironmental features of human MEC would allow for the testing of novel therapeutics and would allow for exciting investigations into all stages of tumor development, from pre-neoplastic lesions through metastatic disease.
42 APPENDIX 1: MATERIALS AND METHODS
Clinical samples
All research involving human tumor tissues was reviewed and approved by The
University of North Carolina at Chapel Hill Institutional Review Board under IRB protocols 15-1604 and 17-2947. Tissues from these de-identified clinical subjects were identified from chart review and archived formalin-fixed paraffin-embedded (FFPE) salivary MEC (n = 18; 10 females and 8 males, average age of 49.6 yrs) or normal salivary gland (n = 6) tissue samples stored at room temperature less than ten years before blocks were sectioned and RNA isolation performed. For all cases, multiple H&E slides were reviewed by a pathologist and sections with tumor were selected for inclusion in the study. Adjacent serial unstained sections were then macrodissected and tumor material submitted for RNA extraction.
Animal models
All animal studies were reviewed and approved by The University of North
Carolina at Chapel Hill Institutional Animal Care and Use Committee under IACUC protocols 17-202 and 20-242. For xenograft studies, male 6-8 week old athymic nude mice (Nu/Nu) were obtained from the Animal Studies Core at the University of North
Carolina at Chapel Hill and housed in facilities run by the Division of Comparative
Medicine at the University of North Carolina (Chapel Hill, NC, USA).
Xenografts
For all xenograft studies, mice were subcutaneously injected with 1x106 UM-
HMC-3A cells resuspended in 50% HBSS and 50% Matrigel. When the average tumor size reached a palpable volume (~50 mm3), animals were randomized into two groups
43 (Vehicle and SR10221 or SR2595) so that the average tumor size in each group was approximately equal. The vehicle group was intraperitoneally (i.p.) injected daily with
250 µL of a 10% DMSO, 10% Tween-80, 80% PBS solution. Drug-treated animals were i.p. injected daily with a 10% DMSO, 10% Tween-80, 80% PBS solution containing either 20 mg/kg SR10221 or 60 mg/kg SR2595 (Sigma-Aldrich #SML2037) based on an average mouse weight of 25g. Caliper measurements were collected every 2 days and bioluminescent imaging (BLI) was performed every 5 days throughout the course of the study.
Genetically engineered mice
Dcpp1GCE (MGI:5661581; referred to here as DcppCre) (Maruyama et al., 2016) and MistCre mouse strains were generously provided by Dr. Catherine Ovitt (University of Rochester). K14CreERT (referred to here as K14Cre) mice were obtained from The
Jackson Laboratory (Stock #005107). LSL-CRTC1/MAML2 (referred to here as C1/M2) mice were generously provided by Dr. Lizi Wu (University of Florida). Rosa26-LSL-
GpNLuc mice were generated by knocking in the GpNLuc (Schaub et al., 2015)
LumiFluor optical reporter cDNA as described in (Carper et al., 2019). Trp53 floxed mice were obtained from the National Cancer Institute Mouse Repository (Strain
#01XC2). p16 floxed mice were a generous gift from Dr. Bernard Weissman (UNC
Chapel Hill). All animals were backcrossed to C57BL/6J or albino B6(Cg)-Tyrc-2J/J mice
(The Jackson Laboratory, Stock #: 000664 or Stock #: 000058) for 7-10 generations.
DcppCre and MistCre alleles were maintained at heterozygosity; all other alleles were bred to either heterozygosity or homozygosity.
44 Tamoxifen (Sigma-Aldrich #T5648) was dissolved at 10 mg/mL in corn oil (MP
Biomedical #901414). To conditionally induce Cre recombinase activation, 8-week old male and female animals were i.p. injected with 100 µL tamoxifen once a day for five days.
For transdermal adenovirus injections, 8-week old animals were injected with 107
IU adenovirus expressing either Cre recombinase or GFP. Adenovirus was brought to
50 µl total volume in PBS and injected transdermally into the submandibular gland using a 29.5-gauge needle once every other day for a total of three injections.
Cell lines
UM-HMC-1 (HMC1), UM-HMC-3A (HMC3A), and UM-HMC-3B (HMC3B) cells
(Warner et al., 2013) were kindly provided by Dr. Jacques Nör (University of Michigan,
Ann Arbor, MI, USA). NCI-H292 (H292) and NCI-H3118 (H3118) cells (Tonon et al.,
2003) were generously provided by Dr. Frederic Kaye (University of Florida, Gainesville,
FL, USA). A253, A388, and A431 cells were kindly provided by Dr. Bernard Weissman
(University of North Carolina, Chapel Hill, NC, USA). HMC1 (source: male, salivary gland MEC), HMC3A (source: female, hard palate MEC), and HMC3B (source: female, lymph node metastasis of hard palate MEC) parental and stably transduced lines were cultured in DMEM medium (Gibco #11965-118) supplemented with 10% FBS (Atlanta
Biologicals #S11550), 20 ng/mL EGF (Sigma-Aldrich #E9644), 400 ng/mL hydrocortisone (Sigma-Aldrich #H0888), 5 µg/mL insulin (Sigma-Aldrich #I6634), and
1X pen/strep/glutamine (PSG; Life Tech #10378016). H292 (source: female, lung MEC) cells were cultured in RPMI-1640 medium (Life Tech #11875119) supplemented with
10% FBS, 1X GlutaMAX (Life Tech #35050061), 1X NEAA (Gibco #11140050), 1X
45 NaPyr (Life Tech #11360070), and 1X PSG. HEK293A, HEK293-CMVTetRTetOC1/M2
(Amelio et al., 2014), HEK293-PGKTetOn3GTRE3GSPGC-1α4, Lenti X-293T (viral packaging cell line; Takara #632180), H3118 (source: female, parotid gland MEC), A388 (source: male, squamous cell carcinoma (SCC)), and A431 (source: female, vulvar SCC) cells were cultured in DMEM supplemented with 10% FBS, 1X GlutaMAX, and 1X PSG.
A253 (source: male, salivary gland SCC) cells were cultured in McCoy’s 5A medium
(Life Tech #16600108) supplemented with 10% FBS, 1X GlutaMAX, and 1X PSG. Cells were passaged using TrypLE (Gibco #12604013) every 2-3 days or when they reached
90% confluence. All cells were maintained in a 37°C, 5% CO2 atmosphere. All cell lines were confirmed mycoplasma-free by PCR as previously described (Young et al., 2010) and using mycoplasma detection primers (Table 5).
Clinical RNA isolation
FFPE tissue samples were sent to the UNC Lineberger Comprehensive Cancer
Center Translational Genomics Lab (TGL) for RNA isolation using the Maxwell 16 MDx
Instrument (Promega #AS3000) and the Maxwell 16 LEV RNA FFPE Kit (Promega
#AS1260) according to the manufacturer’s protocol (Promega #9FB167). Pathology review of a hematoxylin and eosin (H&E) stained slide was used to guide macro- dissection of unstained slides to enrich for tumor RNA. Total RNA quality was measured using a NanoDrop spectrophotometer (Thermo Scientific ND-2000C) and a TapeStation
4200 (Agilent G2991AA). Total RNA concentration was quantified using a Qubit 3.0 fluorometer (Life Technologies Q33216).
46 RNA sequencing
Total RNA sequencing libraries were prepared at TGL using a Bravo Automated
Liquid-Handling Platform (Agilent G5562A) and the TruSeq Stranded Total RNA Library
Prep Gold Kit (Illumina 20020599) according to the manufacturer’s protocol (Illumina
1000000040499). RNAseq library quality and quantity were measured using a
TapeStation 4200 (Agilent G2991AA), pooled at equal molar ratios, and denatured according to the manufacturer’s protocol (Illumina 15050107). Sequencing was performed at the High Throughput Sequencing Facility (HTSF) at UNC Chapel Hill. Two
RNA-seq libraries were sequenced per lane on a HiSeq2500 (Illumina SY–401–2501) with 2x50 bp paired-end configuration according to the manufacturer’s protocol (Illumina
15035786).
Bioinformatics
RNAseq data analyses were performed with FASTQ files aligned to the GRCh38 human genome (GRCh38.d1.vd1.fa) or the GRCm38 mouse genome using STAR v2.4.2 (Dobin et al., 2013) with the following parameters: --outSAMtype BAM Unsorted,
--quantMode TranscriptomeSAM. Transcript abundance for each sample was estimated using Salmon version 0.1.19 (Patro et al., 2017) to quantify the transcriptome defined by
Gencode v22. Gene level counts were summed across isoforms and genes with low expression (i.e. samples with fewer than 10 reads) were removed prior to downstream analyses. The R package DESeq2 (version 1.24.0) (Love et al., 2014) was used to test for differentially expressed genes between CRTC1/MAML2-positive MEC tumors and normal samples or between all MEC tumors and normal samples. The CRTC1/MAML2- regulated IGF-PI3K gene signature (Musicant_MEC_CRTC1-MAML2_IGF1 curated
47 gene set) includes genes that are differentially expressed between CRTC1/MAML2- positive MEC samples and normal samples with Benjamini–Hochberg false discovery rate (q-value) < 0.05 and a fold change > 2. The resulting list of genes was refined to include only genes shared with IGF1-related Molecular Signatures Database (MSigDB) curated gene sets. Gene set enrichment analysis was performed on a customized list of gene sets with GSEA software from the Broad Institute (Mootha et al., 2003;
Subramanian et al., 2005). The customized gene sets include a list of pathways from
MSigDB curated gene sets that are related to IGF1. Hierarchical clustering and heat- map plotting were performed using the ComplexHeatmap R package version 2.0.0 (Gu et al., 2016). An average linkage algorithm with a Euclidean distance function was applied to variance stabilizing transformed gene expression data for the gene signature.
RTKi drug screen
Compound screening was performed as previously described (Bevill et al., 2019;
Lipner et al., 2020). Briefly, drug screening was performed in a 384 well plate format, in duplicate. Cells were seeded in 45 µL of full culture media; A253 (1000 cells/well), H292
(600 cells/well) H3118 (7000 cells/well), HMC1 (500 cells/well), HMC3A (500 cells/well) and HMC3B (700 cells/well). Cells were seeded in 384-well plates using a BioTek microplate dispenser. Cells were allowed to adhere overnight, following which they were treated with 176 individual drugs at 6 doses (10 nM, 100 nM, 300 nM, 1 µM, 3 µM and
10 µM) using a Beckman Coulter Biomek FX Automated Liquid Handling instrument.
Each plate also included 1 µM Bortezomib and 1% DMSO as positive and negative controls for growth inhibition, respectively. Furthermore, a full 384 well plate with 1%
DMSO treatment was also seeded per cell line and was used to confirm minimal well-
48 location associated intra-plate variability. 72 hr post drug addition, cells were lysed, and cell viability was measured using Cell Titer Glo 2.0 (Promega #G9243) according to the manufacturer’s instructions. Luminescence was measured using a PHERAstar FS instrument and growth inhibition was calculated relative to DMSO-treated wells. Z’ scores were calculated as previously described (Zhang et al., 1999). The 1 µM
Bortezomib (positive control; 0% viability) and 1% DMSO (negative control; 100% viability)-treated cells were used to calculate the dynamic range and screen variability/reliability (Z’ score). A Z’ score of >0.75 was obtained for all cell lines tested.
The 3SD/5SD cutoffs were calculated based on the variability observed in the negative control wells.
Virus production and transduction
Lentiviral or retroviral expression plasmids were co-transfected with VSV-G envelope plasmid and either δ8.2 gag/pol (lentivirus) or pMD (retrovirus) helper plasmids into Lenti X-293T cells seeded in 10 cm tissue culture dishes. Transfections were performed using 1 mg/mL Polyethyleneimine (PEI) Transfection Reagent (VWR
#BT129700). Briefly, 1.5 µg VSV-G, 5 µg δ8.2 or pMD, and 6 µg lentiviral or retroviral plasmid were brought to 500 µL with OptiMEM (LifeTech #1158021) and vortexed briefly. In a separate tube, 25 µL PEI (2 µL PEI/µg DNA) was added to 475 µL OptiMEM and vortexed briefly. Both solutions were incubated at room temperature for 5 min, then combined and incubated for an additional 20 min at room temperature. This mixture was then added dropwise to the seeded Lenti X-293T cells. The next day, cell culture media was replaced with DMEM supplemented with 1x NaPyr, 10 mM HEPES, 1X GlutaMAX, and 1X PSG (no FBS). Two days later, media was collected and filtered through a 0.45
49 µm PVDF membrane and then viral particles were concentrated via ultracentrifugation
(100,000 g for 2 hr at 4°C) into a sucrose cushion. Concentrated virus was resuspended in cold PBS and either stored at -80°C or used immediately for transduction.
To generate stable cells, various cell lines were seeded at 50% confluency in 12- well plates. Concentrated virus and 4 µg/mL polybrene were added directly to cells and then plates were centrifuged at 1200g for 90 min at 30°C. To generate stable
LumiFluor-overexpressing cells, HMC3A cells were transduced with concentrated lentivirus expressing our GpNLuc reporter (Schaub et al., 2015) and selected with puromycin for at least 10 days prior to use in experiments. To generate stable Ductal cells with inducible TdTomato or CRTC1/MAML2 expression, SV-40 immortalized salivary ductal cells (Gao et al., 2014) were transduced with concentrated retrovirus
tdTO LumiFluor C1/M2 LumiFluor encoding either PTigh PGK or PTight PGK and stable cells were selected by fluorescence activated cell sorting (FACS) by gating for GpNLuc-positive cells. To generate shRNA-expressing cells, parental HMC3A or HMC3A-PGKGpNLuc cells were transduced with concentrated shRNA-expressing retro- or lentivirus expressing the shRNA of interest. Proliferation assays, quantitative real-time PCR, and Western blotting were performed on polyclonal cell populations within 14 days of transduction for all shRNA knockdown experiments. To generate cells that express the PGC-1α4 splice variant, HEK293A cells were first transduced with a concentrated lentivirus that expresses TetOn3G and stable cells selected with puromycin. Subsequently, these
HEK293-PGKTetOn3G stables were transduced with a concentrated lentivirus that expresses PGC-1α4 upon tetracycline/doxycycline administration and the resulting
HEK293-PGKTetOn3G-TRE3GSPGC-1α4 stable cells were selected with hygromycin for at
50 least 10 days prior to use in experiments. To generate MEC cells with inducible dominant-negative CREB (dnCREB; A-CREB), HMC3A cells were first transduced with a concentrated lentivirus that expresses TetOn3G and stable cells were selected with puromycin. These HMC3A-PGKTetOn3G stables were transduced with a concentrated retrovirus that expresses A-CREB (Ahn et al., 1998) upon tetracycline/doxycycline administration, and the resulting HMC3A-PGKTetOn3G-TetOdnCREB stable cells were selected by FACS by gating for tdTomato-positive cells. To generate MEC cells with inducible dCas9 expression, HMC3A cells were first transduced with a concentrated lentivirus that expresses TetOn3G and stable cells were selected with puromycin.
These HMC3A-PGKTetOn3G stables were transduced with a concentrated lentivirus encoding the doxycycline induced dCas9-KRAB-IRES-BFP cassette (Addgene #85449)
(Fulco et al., 2016). To select for positively transduced cells, 1 μg/mL doxycycline was applied to the cells for 72 hours and then BFP-positive cells were selected by FACS.
Finally, the HMC3A-PGKTetOn3G -TRE3GSdCas9-KRAB-IRES-BFP cells were transduced with a lentivirus constitutively expressing the respective PGC-1α sgRNA, a blasticidin resistance cassette, and a 2xNLS-mKate2 fluorescent reporter. Cells were selected with blasticidin for at least 7 days and successful selection was confirmed via nuclear mKate2 fluorescence signal.
Cell viability
HMC1, HMC3A, or HMC3B cells were seeded in 96-well clear-bottom, white- walled plates (Corning #3917) at 15,000 cells/well. The next day, media was replaced with 200 μL fresh media containing either vehicle (DMSO) or increasing concentrations of drug (BMS-754807 (Sigma-Aldrich #BM0003), PPP (Santa Cruz #SC-204008),
51 SR10221 (synthesis detailed in Musicant et al., 2021), SR2595 (Sigma-Aldrich
#SML2037), or T0070907 (Fisher #NC1015539); final concentration 1% DMSO). 72 hr later, the amount of ATP (a proxy for cell viability) in each well was measured using the
ATPlite Luminescence Assay System (PerkinElmer #6016949) according to the manufacturer’s instructions. Briefly, plates were removed from the cell culture incubator and equilibrated for 30 min at room temperature in the dark. Then, media was aspirated from each well and 100 μL reconstituted ATPlite 1-step reagent was added to each well.
Plates were shaken for 2 min (425 cpm, 3 mm orbit) in a Cytation 5 plate reader (BioTek
Instruments, Inc.) and then ATP levels were quantified by measuring total luminescence. IC50 values were calculated in GraphPad Prism (GraphPad, San Diego,
CA, USA) using a non-linear curve fit (log[agonist] vs response, four parameter-variable slope). All assays were performed in biologic triplicate with three technical replicates for each condition. Figure panels show one representative experiment (including three technical replicates each). IC50 values presented in Table 2 were calculated using the results from all three biologic replicates.
Cell proliferation assays (Cytation 5)
HMC3A-PGKGpNLuc cells (either non-transduced or transduced with shRNAs targeting PGC-1α, IGF-1R, or MAML2) were seeded in 12-well plates at 200,000 cells/well. The next day, fresh media was applied to each well and then individual wells were imaged at 0, 4, 8, 24, 28, 32, and 48 hr following the media change. Five separate fields of view were imaged for each well at 4X magnification using a fluorescent plate reader (BioTek Cytation 5, BioTek Instruments, Inc.). Between each imaging time point, plates were returned to the cell culture incubator. To determine cell confluence, a
52 primary mask was applied to each fluorescent image using the recommended parameters set out in the BioTek Technical Note “Measuring Confluence Using High
Contrast Brightfield” (https://www.biotek.com/resources/technical-notes/measuring- confluence-using-high-contrast-brightfield/). Confluence measurements from multiple fields of view in a single well were used as technical replicates. Each assay was performed in biologic triplicate.
Cell proliferation assays (IncuCyte Zoom)
HMC3A cells were seeded in 48-well plates at 15,000 cells/well. The next day, media was replaced with fresh media containing either vehicle (1% DMSO) or the indicated concentration of PPP (Santa Cruz #SC-204008) or SR10221. Then, plates were imaged at 10X magnification every 2-4 hours on an IncuCyte Zoom (Essen
BioScience), which maintained plates at a constant 37°C and 5% CO2 for the duration of the assay. Four separate fields of view were imaged for each well and an image mask was applied to each image to generate confluency measurements. Each assay was performed in biologic triplicate with at least three technical replicates per experiment.
Western blotting
Whole cell lysates were prepared in buffer containing 5% glycerol (Sigma-Aldrich
#G5516), 25 mM Tris (pH 7.4; Fisher #BMA51237), 150 mM NaCl (Fisher #BMA51202),
1 mM EDTA (Sigma-Aldrich #E7889), and 1% NP-40 (Fisher #50-147-289) supplemented with protease inhibitors (cOmplete, EDTA-free Protease Inhibitor
Cocktail, Roche #04693132001) and phosphatase inhibitors (PhosSTOP, Roche
#10917400). Lysates (30-50 mg) were loaded onto mini-10% tris-glycine polyacrylamide
53 gels and proteins were separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Proteins were transferred to a 0.45 μm nitrocellulose membrane (Fisher #0088018) using a Bio-Rad Trans-Blot Turbo system set at 1.3
Amps (constant) and 25 V for 30 min. Membranes were blocked for 1 hr at room temperature in TBS-T + 5% bovine serum albumin (BSA; Sigma #A2153) and then incubated overnight at 4°C with primary antibodies: IGF-1 (0.2 μg/mL; Abcam #ab9572),
IGF-1Rβ (1:2500; Cell Signaling #3018S), phosphorylated IGF-1Rβ (Tyr1135; 1:1000;
Cell Signaling #3918S), PPARγ (1:1000; Abcam #209350), actin (1:5000; Sigma
#A3854), tubulin (1:5000; Sigma #T6074), or Flag (1:5000; Sigma #A8592) diluted in
TBS-T + 5% BSA. Following primary antibody incubation, membranes were washed and probed with Horseradish peroxidase (HRP)-conjugated goat anti-mouse (1:5000;
Thermo Fisher #31432) or donkey anti-rabbit (1:5000; Thermo Fisher #31458) secondary antibodies diluted in TBS-T supplemented with 5% BSA for 1-2 hr at room temperature. Blots were imaged using Clarity ECL (Bio-Rad #170-5060) and
ImageQuant LHS4000 (GE).
2D colony formation
HMC1, HMC3A, and HMC3B cells were seeded in 24-well plates at 800, 600, and 1400 cells/well, respectively. The next day, media was replaced with 500 μL fresh media containing either vehicle (DMSO) or the indicated concentration of drug (BMS-
754807 (Sigma-Aldrich #BM0003), PPP (Santa Cruz #SC-204008), SR10221, SR2595
(Sigma-Aldrich #SML2037), or T0070907 (Fisher #NC1015539); final concentration 1%
DMSO). Seven days later, media was removed from all wells and cells were fixed in
10% buffered formalin (Fisher #SF994) for 5 min at room temperature. Wells were
54 washed once in ddH2O and then stained in 0.05% crystal violet (Sigma-Aldrich #C6158) for 30 min at room temperature. Wells were then washed an additional 3 times in ddH2O to remove any unbound stain and allowed to dry at room temperature overnight. Once dry, individual wells were imaged at 4X magnification on a BioTek Cytation 5 plate reader (BioTek Instruments, Inc.) and images were stitched using the Gen5 software
(v2.09; BioTek Instruments, Inc.) using the default parameters. Colony numbers were quantified manually in Image J (Schneider et al., 2012), where a colony is defined as a cluster of at least 50 individual cells.
3D tumor spheroid formation
Individual wells of a 48-well plate were coated with 100 μL of a 25% Matrigel
(Fisher #CB4023A) solution diluted in cold PBS. The plate was spun in a Thermo
Scientific Sorvall LYNX 4000 centrifuge at 4°C for 10 min at 2000g to ensure even distribution across the well surface. The plate was then placed in a 37°C incubator for 1 hr to allow the Matrigel ‘bed’ to completely polymerize. Next, excess PBS was gently aspirated from the wells and 2000 HMC3A cells resuspended in 100 μL media were added to each well. The plate was centrifuged at 4°C for 5 min at 2000g to embed the cells into the Matrigel. The plate was returned to the incubator for an additional hour to allow for cell adhesion to the Matrigel. After 1 hr, excess media was carefully aspirated and 50 μL undiluted Matrigel was gently added to each well on top of the cells. The plate was again returned to the incubator for 1 hr to allow for Matrigel polymerization.
Finally, 500 μL media (containing either drug or vehicle at a final concentration of 1%
DMSO) was added to each well. Every 24 hr, wells were imaged at 4X magnification on a BioTek Cytation 5 plate reader (BioTek Instruments, Inc.) and images were stitched
55 using the Gen5 software (v2.09; BioTek Instruments, Inc.) using the default parameters.
Tumorsphere area was manually measured in ImageJ using the ROI area measurement tool. A minimum of 50 individual tumorspheres were quantified for each condition in each experiment.
Caspase 3/7 apoptosis assay
For drug treatment assays, HMC3A cells were seeded in 6-well plates at 350,000 cells/well. The next day, media was replaced with 2 mL fresh media containing either vehicle (DMSO) or the indicated concentration of drug (PPP (Santa Cruz #SC-204008) or SR10221; final concentration 1% DMSO). For shRNA assays, HMC3A cells stably transduced with shNS (control) or shPPARγ were seeded in 6-well plates at 350,000 cells/well. At the indicated time point (either post-treatment (PPP/SR10221) or post- seeding (shRNAs)), cells were trypsinized, spun down, and resuspended in 1 mL PBS supplemented with 1X GlutaMAX (Life Tech #35050061) and 10% FBS (Atlanta
Biologicals #S11550). Apoptotic cells were visualized using the CellEvent Caspase 3/7
Green Flow Cytometry Assay Kit (ThermoFisher #C10427) according to the manufacturer’s instructions. Briefly, 1 μL CellEvent Caspase 3/7 Green Detection
Reagent was added per 1 mL of resuspended cells. This mixture was gently vortexed and incubated at 37°C for 25 min. Next, 1 μL SYTOX AADvanced Dead Cell Stain was added per 1 mL of resuspended cells. This mixture was gently vortexed and incubated at 37°C for an additional 5 min. Samples were analyzed via flow cytometry on a BD
Accuri C6 instrument. An aliquot of unstained cells from each condition was reserved as a control to generate gates.
56 Chromatin immunoprecipitation (ChIP) assay
HEK293-CMVTetRTetOC1/M2 cells (Amelio et al., 2014) were seeded in 15 cm tissue culture dishes in either plain growth medium or growth medium containing 1
μg/mL doxycycline (Sigma-Aldrich #D9891). After 48 hr, formaldehyde (Fisher #BP531) was added dropwise to each plate to 1% final concentration. Plates were incubated at room temperature for 10 min and then crosslinking was stopped by adding glycine
(Fisher #BP381-1) to a final concentration of 120 mM and incubating for an additional 5 min. Cells were then transferred to a tube, pelleted at 4°C, and washed twice with cold
PBS. Finally, washed cells were resuspended in PBS supplemented with protease inhibitors (cOmplete, EDTA-free Protease Inhibitor Cocktail, Roche #04693132001) to a concentration of 5x106 cells/mL and stored at -80°C in aliquots of 1 mL until lysis. For lysis, individual cell aliquots were pelleted, resuspended in 200 μL SDS lysis buffer (1%
SDS, 10 mM EDTA, 50 mM Tris-HCl, pH 8.1) supplemented with protease inhibitors, and rocked at 4°C for at least 20 min. Next, 800 μL ChIP dilution buffer was added and aliquots were immersed in a 100% ethanol ice bath (-11 to -7°C) and sheared (2 Amp for 5 seconds) using a sonicator (Qsonica #Q700A) equipped with a microtip sonic dismembrator (Model 505, Fisher #4418). A fraction of this sheared sample was reserved as a Pre-IP control. To the rest of the sample, 1 μg anti-Flag M2 antibody
(Sigma #F1804) was added and samples were rocked at 4°C overnight. Antibody- chromatin complexes were isolated by rocking with Protein G SureBeads (BioRad
#1614821) at 4°C for 4 hr. Beads were washed once each in low salt buffer (0.1% SDS,
1% TritonX-100, 2 mM EDTA, 20 mM Tris-HCl, 15 mM NaCl), high salt buffer (0.1%
SDS, 1% TritonX-100, 2 mM EDTA, 20 mM Tris-HCl, 500 mM NaCl), and LiCl buffer
57 (250 mM LiCl, 1% NP-40, 1% Deoxycholate, 1mM EDTA, 10 mM Tris-HCl, pH 8.1) and then twice in TE buffer (1.2 mM EDTA, 10 mM Tris-HCl, pH 8.1). Samples were eluted twice with pre-heated (55°C) elution buffer for 15 min at room temperature with shaking.
Cross-links were reversed by adding NaCl to 200 mM and incubating for at least 4 hr at
65°C. Samples were treated for 30 min at 37°C with RNase A at 8 μg/μL final concentration. Finally, samples were treated with Proteinase K solution (10 μL 0.5 M
EDTA, 10 μL 1M Tris, 1 μL 20 mg/mL Proteinase K (LifeTech #AM2546) for a 500 μL sample) for 1 hr at 50 °C and then DNA was purified using the NucleoSpin Gel and PCR
Cleanup Kit (Macherey-Nagel #740609.250). Occupancy of target proteins of interest within the promoter regions was assessed by qPCR (Table 5).
Luciferase reporter assay
HEK293-CMVTetRTetOC1/M2 cells were seeded in 24-well plates at 100,000 cells/well. The next day, transfected each well with 500 ng of either PGC-1a proximal promoter or PGC-1a distal promoter driven luciferase reporter using Lipofectamine
(Invitrogen #50470) according to the manufacturer’s protocol and using a DNA:lipid ratio of 1:2. The next day, media was replaced with 1 mL fresh media with or without 1 μg/mL doxycycline (Sigma-Aldrich #D9891). After 24 hr, media was removed from all wells and
100 μL 1% Triton X-100 was added to each well. Plates were rocked for 15 min at room temperature. 50 μL of the supernatant from each well was transferred to a white, opaque 96-well plate, and 50 μL Bright-Glo Luciferase Assay Buffer (Promega #E264B) was added to each well. Total luminescence was quantified immediately on a BioTek
Cytation 5 plate reader (BioTek Instruments, Inc.). For the dose-response luciferase reporter assays, HEK293 cells were reverse transfected with 3xPPRE promoter or
58 IGF1-P2 promoter driven luciferase reporters using Lipofectamine 2000 Transfection
Reagent (Invitrogen #11668019) in 96-well plates at a density of 2x104 cells/well. Forty- eight hours post-transfection, cells were treated with either vehicle (DMSO) or the
PPARγ ligands GW1929 (Sigma-Aldrich #5668), SR2595 or SR10221 at the indicated concentrations. 24 hr after treatment, cells were lysed, and luciferase activity was quantified using the Dual-Glo Luciferase Assay System (Promega #E2920).
Luminescence was measured using Synergy Neo microplate reader (BioTek). Values were normalized using Renilla expression.
Bioluminescence imaging
Bioluminescent-fluorescent BRET signal was measured non-invasively as previously described (Schaub et al., 2015) with minor modification. Briefly, animals were i.p. injected with 250 μM (1:20 dilution, ∼500 μg/kg) Nano-Glo Luciferase Assay
Substrate (Promega, #N1120) in sterile PBS. Isoflurane-anesthetized animals were then imaged using an AMI Optical Imaging System (Spectral Instruments Imaging, Inc.) 5 min after injection. Images were captured with open filter and acquisition times of 5 min or less at the indicated settings. Data were analyzed using Aura imaging software
(v2.2.0.0).
Histology and immunohistochemistry
All animals showing tumors > 2 cm3 or other signs of distress were euthanized and subjected to full necropsy. For histological analysis, xenograft tumors were fixed in
10% neutral buffered formalin for approximately 1 week at room temperature. Following fixation, tissues were processed on an ASP6025 automated tissue processor (Leica
Biosystems) and embedded in paraffin wax. Blocks were sectioned at 4-6 μm, mounted
59 on glass slides, and FFPE tissue sections were deparaffinized prior to hematoxylin and eosin (H&E) or mucicarmine staining. Immunohistochemistry was performed on the
Discovery Ultra (Ventana Medical Systems) using manufacturers reagents on 4 μm sections. For anti-IGF-1 immunohistochemistry, anti-rabbit IGF-1 (Abcam #ab9572) was prepared using Discovery PSS Diluent (cat. #: 760-212). Antigen retrieval was performed using Ventana’s CC1 (pH 8.5) for 64 min at 90 °C. The slides were given a hydrogen peroxide block for 8 min at room temperature and then incubated in the primary antibody diluent (1:100) for 1 hr at room temperature, followed by anti-Rabbit
HRP secondary antibody (Vector Labs #MP-7401) for 32 min at room temperature. The slides were then treated with DAB and counterstained with Hematoxylin II for 12 min and then Bluing Reagent for 4 min.
Plasmids
The PGC-1a promoter constructs were generated by cloning the -500 to +52 bp of either the proximal or distal human PGC-1a promoter sequences via 5’ NheI and 3’
HindIII into the pGL4.15 luciferase reporter (Promega #E6701). The human IGF-1 promoter constructs were generated by cloning either a 2079 bp region encompassing the P1 promoter or a 1547 bp region encompassing the P2 promoter via 5’ BglII and 3’
HindIII into either the pGL4.15 (Promega #E6701) or the pGL3-Enhancer luciferase reporter (Promega #E1771). The A-CREB (Ahn et al., 1998) dominant-negative CREB cDNA was directionally cloned via 5′ NotI and 3′ MluI restriction enzyme sites into pRetroX-Tight-tdTomato (Amelio et al., 2014). The pTRE3G-MCS_PGK-GpNLuc construct was generated by excising the pTight promoter from the pRetroX-Tight-
MCS_PGK-GpNLuc construct (Addgene plasmid #70185) containing our previously
60 described LumiFluor optical reporter (Schaub et al., 2015) and subcloning the TRE3G promoter in its place. Subsequently, a N-terminal FLAG-tagged CRTC1/MAML2 cDNA was directionally subcloned via 5’ NotI and 3’ MluI restriction enzyme sites into this pTRE3G-MCS-GpNLuc vector. The pLV-TRE3G_PGC1a4-mCherry:T2A:Hygro construct was custom synthesized (Cyagen Bioscience) to contain the human
PPARGC1A variant 4 splice isoform (PGC-1a4) (Ruas et al., 2012) downstream of the
TetO response element promoter for tet/dox inducible expression. The pFLAG-CMV-
2_CRTC1/MAML2 (Tonon et al., 2003) construct was a gift from Frederic Kaye, the pLKO.1 shNS and shMAML2 (Chen et al., 2014) shRNA constructs were gifts from Lizi
Wu, the pLMN UltramiR-E shIGF1R (ULTRA-3308043 and ULTRA-3443222) and shPPARGC1A (ULTRA-3197172 and ULTRA-3197171) constructs were from transOMICS Technologies, and the PPRE x3-TK-Luc (Kim et al., 1998) construct was a gift from Bruce Spiegelman. PPARγ knockdown constructs were purchased from
Sigma-Aldrich (#SHCLNG-TRCN0000355926, #SHCLNG-TRCN0000001673, and
#SHCLNG-TRCN0000001674). All plasmids were amplified and purified from either
DH5a (Zymo Research #T3007) or Stbl3 (ThermoFisher Scientific #C737303) E. coli strains. sgRNA Design and Cloning
sgRNAs targeting the PGC-1a proximal or distal promoters were designed using the CRISPR-ERA resource (http://crispr-era.stanford.edu/).
The PGC-1a proximal promoter sequence was used as the query for generating sgRNAs targeting the PGC-1a1 isoform (NM_013261.5):
AGCAAAAAAAAAAAAAAAAAAAAAAAAAGCCCCGTTTGCGCTTTCAAACACTCCCTC
61 AATGAGAAAATGTCTCATAAAAATGCATCATGTGATAAGCTCTTGCTTTAGTCCCAA
ACTGAGCTTGAGTCCACTTGGAGATCTTAGAATTAAAAAGATTGCAGGGGATTTTG
GTTATTATATGGCCAGGGCTCCGTTTAGAGTCTGTGGCATTCAAAGCTGGCTTTAA
TCACAGCATGATGCTTGAAGCCTCCAAAAGTCTAAGTCTAAGTGTTTCCTTTCTTTC
TTTCTTTTCTTTTTCTTTCTTTTTTTTTTTTTTTAAAGCGTTACTTCACTGAAGCAGAG
GGCTGCCTTTGAGTGACGTCACGAGTTAGAGCAGCAAGCTGCACAGGGGAAGGG
AGGCTGGGTGAGTGACAGCCCAGCCTACTTTTTAATAGCTTTGTCATGTGACTGGG
GACTGTAGTAAGACAGGTGCCTTCAGTTCACTCTCAGTAAGGGGCTGGTTGCCTG
CA.
The PGC-1a distal promoter sequence was used as the query for generating sgRNAs targeting the PGC-1a4 isoform (JQ772118.1):
TAATAACAGGGGACAAAGGGTGAAATAATCCAGTAAAGTTTTGAGCAGCCACTTGA
CAACGTATTCCAAATAAATGAGAGGAGGAAAACCCTAGCTCTACCAACTGGGGCCA
TAAAACAGAGTCTTCTACACTCTCTTTAATGTCAATATTAACCAGTTTGCAAATGTTC
AGTCAGTTATGGGCATGGATGGATGTCCTGAATGGGTTCCCGGGATAAAGTGTCA
TCATAGGACAGAAATCACAGGGAGAGTGCACCAAGGAAAAATTACAGTACTGCTAT
ATTTACTTAGTGCCTCTGAACTAGGGTTTTATTTTCCACGGGTTGGAAAGGGAACC
ACCTGTCTCAATTGCTGATGTCAGAGAGCTCCCTCGAGACACAGGGCTGCTGGAA
AGCACATGATACTGTACATATTTGCTCTTACGTTCGTATCTGGCTAAGATTGGGTTT
CAGATTTGTGCCCTATTGTGGAGTTCATTTAGTAGTGACTCTGAGATGCCC.
The pgRNA-CKB vector (Addgene #73501) (Mandegar et al., 2016) was used to constitutively express the target sgRNAs. sgRNAs were cloned into the pgRNA-CKB vector using previously published methods (Gilbert et al., 2013). Briefly, the pgRNA-
62 CKB backbone was digested with Esp3I (NEB #R0734) and dephosphorylated using recombinant Shrimp Alkaline Phosphatase. The required sgRNAs were synthesized
(Eton Biosciences, Inc) as oligonucleotides with Esp3I compatible overhangs (Table 5).
Oligonucleotides were phosphorylated using T4 Polynucleotide Kinase (NEB #M0201) and annealed by adding 50 μM sodium chloride to the phosphorylated oligo reaction, denaturing the oligo mix via boiling (95 °C for 10 min), and performing an extended, slow cooling (~0.1°C/sec). 100 ng of the digested/dephosphorylated pgRNA-CKB backbone was ligated overnight with 1 μL of the 0.1 μM phosphorylated/annealed guide
RNA using T4 DNA ligase. All sgRNA constructs were sequence verified (Eton
Biosciences, Inc) using the primer listed in Table 5. Validation of isoform-specific knockdown revealed that one sgRNA construct (pSico_U6-Pan-PGC1a sgPGC) repressed expression of all PGC-1a isoforms, while another sgRNA construct,
(pSico_U6-PGC1a1 sgPGC) selectively repressed expression of the major isoform
(PGC-1a1).
PCR and qRT-PCR
For cell lines and fresh-frozen tissues, gene expression was measured by extracting RNA using a Nucleospin RNA kit (Machery-Nagel #740955) according to the manufacturer’s instructions. cDNA was synthesized from 1 μg of RNA using iScript cDNA synthesis kit (Bio-Rad #170-8890). For human tissues, RNA was extracted using the Maxwell 16 MDx Instrument (Promega #AS3000) and the Maxwell 16 LEV RNA
FFPE Kit (Promega #AS1260) according to the manufacturer’s protocol (Promega
#9FB167). cDNA was made from 1-4 μg of RNA using SuperScript IV Reverse
Transcriptase (Invitrogen #18090050) with dNTPs (NEB #N0446S), RNase inhibitor
63 (Applied Biosystems #N808-0119), 25 μM oligo d(T)20 primer (Invitrogen #100023441) and 25 μM MAML2-specific reverse primer (Table 5). CRTC1/MAML2 copy number was determined by establishing standard curves with 100 to 1x106 copies of FLAG-
CRTC1/MAML2 plasmid. Relative gene expression of target genes was determined using the 2ΔΔCt method and normalized to human RPL23 expression. qRT-PCR was performed using FastStart Universal SYBR Green Master (Rox) Mix (Roche
#04913850001) with 1/50 (tissue) or 1/100 (cells) volume of the cDNA iScript reaction, and 0.25 μM of primers (Table 4).
Quantification and Statistical Analysis
All statistical tests were executed using GraphPad Prism software or the statistical software R (version 3.1.2). Differences between variables were assessed by
2-tailed Student’s t test, one-way ANOVA with Holm-Sidak post hoc test, or Wilcoxon rank sum tests, where appropriate. For proliferation assays, pairwise comparisons were performed by Benjamini-Hochberg analysis. Data are expressed as mean ± SEM, P values less than 0.05 were considered statistically significant (∗p < 0.05, ∗∗p < 0.01,
∗∗∗p < 0.001, ∗∗∗∗p < 0.0001).
64 APPENDIX 2: TABLES
Gene Fold Regulation Function/Process IGF1 112.2055 anabolism, cell proliferation, and differentiation FOS 49.8665 transcription, cell proliferation, and differentiation TIMP3 9.7136 cell adhesion MMP1 5.2416 cell adhesion TEK 4.9933 cell proliferation and differentiation CDKN1A 4.5948 cell cycle IL8 3.5554 cytokine THBS1 3.2043 cell adhesion SERPINE1 3.0105 cell adhesion PLAUR 2.1886 signal transduction
Table 1. Cancer pathway genes induced upon CRTC1/MAML2 expression. (From Qiagen Human Cancer PathwayFinderTM RT2 Profiler PCR Array)
65 Fusion Status Cancer Cell Line Tissue BMS-754807 PPP C1/M2 Negative Non-MEC A-253 Salivary Gland 58.9 0.719
UM-HMC-1 Salivary Gland 0.213 0.066 UM-HMC-3A Salivary Gland 0.148 0.211 C1/M2 Positive MEC UM-HMC-3B Salivary Gland 0.088 0.216 NCI-H3118 Salivary Gland ND 0.4890
Table 2. Calculated IC50 values of IGF-1R inhibitors on the viability of CRTC1/MAML2-positive and -negative cancer cell lines. Cell viability was assessed by ATPlite assay and IC50 values were calculated in GraphPad Prism using a non-linear curve fit (log[agonist] vs response, four parameter-variable slope). IC50 values are reported in μM and are reflective of the average of three biological replicates (mean ± SEM).
66 Cancer Cell Line SR-10221 SR-2595 T0070907 UM-HMC-1 114.5 40.1 328.1 UM-HMC-3A 72.7 50.2 222.7 UM-HMC-3B 148.4 78.6 329.7 NCI-H292 76.6 ND ND NCI-H3118 98.5 ND ND
Table 3. Calculated IC50 values of PPARγ inverse agonists on the viability of CRTC1/MAML2-positive cancer cell lines. Cell viability was assessed by ATPlite assay and IC50 values were calculated in GraphPad Prism using a non-linear curve fit (log[agonist] vs response, four parameter-variable slope). IC50 values are reported in μM and are reflective of the average of three biological replicates (mean ± SEM).
67 Application Target Sequence (5' - 3') Quantitative real-time PCR CRTC1-MAML2 ATG GCG ACT TCG AAC AA (forward) GGG TCG CTT GCT GTT GGC (reverse)
IGF-1 TGT GGA GAC AGG GGC TTT TA (forward) ATC CAC GAT GCC TGT CTG A (reverse)
IGF-1 Exon 1 GAT AGA GCC TGC GCA ATG GAA (forward) GAG ATG GGA GAT GTT GAG AGC A (reverse)
IGF-1R AAA AAC CTT CGC CTC ATC CT (forward) TGG TTG TCG AGG ACG TAG AA (reverse)
NR4A2 TGA AGA GAG ACG CGG AGA AC (forward) AAA GCA ATG GGG AGT CCA G (reverse)
PGC-1⍺1 ATG GAG TGA CAT CGA GTG TGC T (forward) GAG TCC ACC CAG AAA GCT GT (reverse)
PGC-1⍺2 AGT CCA CCC AGA AAG CTG TCT (forward) ATG AAT GAC ACA CAT GTT GGG (reverse)
PGC-1⍺3 CTG CAC CTA GGA GGC TTT ATG C (forward) CAA TCC ACC CAG AAA GCT GTC T (reverse)
PGC-1⍺4 TCA CAC CAA ACC CAC AGA GA (forward) CTG GAA GAT ATG GCA CAT (reverse)
PGK1 TTG ACC GAA TCA CCG ACC TC (forward) ACC CGC TTC CCT TTA ACG TC (reverse)
PKM2 CGT CTG AAC TTC TCT CAT GGA A (forward) ATG GGG TCA GAA GCA AAG C (reverse)
RPL23 TGA TGG CCA CAG TCA AGA AA (forward) ACA CGC CAT CTT TTC TAC GG (reverse)
Table 4. Primers used for qRT-PCR.
68
Application Target Sequence (5' - 3') Mycoplasma detection Mycoplasma TGC ACC ATC TGT CAC TCT GTT AAC CTC (forward) GGG AGC AAA CAG GAT TAG ATA CCC T (reverse)
MAML2 cDNA synthesis MAML2 CTC TGA GGG ACT GAA GGG AT (reverse)
Chromatin IP GAPDH CCT TCT TGC CTT GCT CTT GCT AC (forward) GCC TGC CTG GTG ATA ATC TTT G (reverse)
NR4A2 GTT GCA CCC TCC CCA CAT G (forward) GGC CGC CAA TGT GCC TTT GTT TAT (reverse)
Distal PGC-1⍺ CAG TCA GTT ATG GGC ATG (forward) CAC TCA CAA CAA GGG CTT CG (reverse)
IGF-1 CCA CTC CTG CCA GAA CGC (forward) GGG AGG AGA AAT CCT CTC AC (reverse)
sgRNA Cloning sgPGC-1⍺1 TTG GAC TGG GGA CTG TAG TAA GAC (forward) AAA CGT CTT ACT ACA GTC CCC AGT (reverse)
sgPGC-1⍺ (Total) TTG GCC GGG ATA AAG TGT CAT CAT (forward) AAA CAT GAT GAC ACT TTA TCC CGG (reverse)
Sequence verification GAG ATC CAG TTT GGT TAG TAC CGG G
Table 5. Primers used for detection of mycoplasma in cell cultures, MAML2 gene-specific cDNA synthesis, chromatin IP, and cloning and sequence verification of sgRNA constructs.
69 APPENDIX 3: FIGURES
Figure 1. Salivary gland structure. The salivary glands consist of a branched ductal structure formed through a process known as branching morphogenesis. Saliva is produced by the acinar cells, which cluster in groups called acini. The saliva then travels through the ducts, which empty into the mouth. Diagram created in BioRender.
70 Figure 2. Salivary MEC is characterized by a t(11;19)(q14-21;p12-13) translocation. (A) The t(11;19) translocation fuses the first exon of CRTC1 with exons 2-18 of MAML2, creating the CRTC1/MAML2 fusion product. This fusion retains the CREB-binding domain from CRTC1 and the transcriptional activation domain from MAML2. CBD: CREB-binding domain; NES: nuclear export signal; NBD: Notch-binding domain; TAD: transcriptional activation domain. Schematic created in BioRender. (B-C) The ZytoLight MAML2 Dual Color Break Apart Probe can be used to detect MAML2 rearrangements. The red and green fluorescent probes bind to the 3’ and 5’ ends of the MAML2 gene, respectively. In cells with no MAML2 rearrangement (B), the two probes colocalize, and in cells with a MAML2 rearrangement (C), individual green and red probes can be detected. (D-E) Fluorescent in-situ hybridization in CRTC1/MAML2-negative A253 cells (D; yellow arrowheads indicate probe colocalization) and CRTC1/MAML2- positive H292 cells (E; green and red arrowheads indicate spatially distinct green and red probes).
71 Figure 3. CRTC1/MAML2 is expressed in >50% of human MEC samples. qRT-PCR analysis of CRTC1/MAML2 (C1/M2) expression in human salivary MEC tumor samples. C1/M2 copy number per 10 ng input RNA was calculated based on a standard curve. Samples with <500 C1/M2 transcripts per 10 ng RNA were classified as fusion-negative (grey bars), while those with ≥500 C1/M2 transcripts per 10 ng RNA were classified as fusion-positive (blue bars). Data is presented as the mean ± SEM (n = 4).
72 Figure 4. Normal salivary glands, CRTC1/MAML2-positive MEC, and CRTC1/MAML2-negative MEC tumors are characterized by distinct gene signatures. Heat map showing unsupervised clustering of the top 1500 most variable genes between MEC and normal salivary gland samples. Normal (e.g. N #1, N #2) and tumor (e.g. Case #1, Case #7) samples are indicated at the top of the heat map in black and green, respectively. CRTC1/MAML2 fusion status is indicated in blue and grey.
73 Figure 5. An IGF-1 related gene signature is characteristic of fusion-positive MEC. (A) Heat map of IGF1- related DEGs between fusion-positive MEC and normal salivary gland samples. Normal (e.g. N #1, N #2) and tumor (e.g. Case #1, Case #7) samples are indicated at the top of the heat map in black and green, respectively. CRTC1/MAML2 status is indicated in grey (negative) and blue (positive). Hierarchical clustering was performed using ComplexHeatmap R package version 2.0.0. (B) Violin plot highlighting the significance of IGF-1 pathway-related genes within our curated Musicant_MEC_CRTC1-MAML2_IGF1 gene set between fusion-positive MEC and normal salivary gland samples (Wilcoxon rank sum test; *p = 0.0075).
74 Figure 6. Curated Musicant_MEC_CRTC1-MAML2_IGF1 gene set distinguishes CRTC1/MAML2-positive MEC. Gene set enrichment plots for genes in the GSEA Molecular Signatures Database (MSigDB) GNF2_IGF1 dataset (A) and our curated Musicant_MEC_CRTC1- MAML2_IGF1 gene set (B) in fusion-positive MEC and normal salivary gland samples.
75 Figure 7. CRTC1/MAML2 and IGF-1 transcript levels are correlated in human MEC tumor samples. Comparison of CRTC1/MAML2 (C1/M2) and IGF-1 CT values in qRT-PCR data from human CRTC1/MAML2-positive MEC tumor (blue) and normal salivary gland (grey) samples. The fitted regression line demonstrates the correlation between the expression levels of the two genes (r2 = 0.7251, p < 0.0001). 95% confidence intervals are indicated by curved lines on either side of the linear regression.
76 Figure 8. CRTC1/MAML2 and IGF-1 transcript levels are correlated in MEC cell lines. (A, B) qRT-PCR analysis of CRTC1/MAML2 (C1/M2) (A) and IGF-1 (B) expression in MEC cell lines and control epidermoid carcinoma cell lines. Cell lines are listed along the x-axis, with CRTC1/MAML2 status indicated below in grey (negative) or red (positive). C1/M2 copy number per 10 ng input RNA was calculated based on a standard curve. Relative IGF-1 fold expression is shown normalized to RPL23 mRNA levels. Data is presented as the mean ± SEM (n = 3). (C) Comparison of C1/M2 and IGF-1 CT values in qRT-PCR data from cell lines. CRTC1/MAML2-negative cell lines are indicated in grey and CRTC1/MAML2-positive cell lines are indicated in red. The fitted regression (solid) line demonstrates that the expression levels of the two genes are correlated (r2 = 0.8664, p = 0.0008). 95% confidence intervals are indicated by curved lines. Gene expression data is the average of at least 3 independent experiments.
77 Figure 9. CRTC1/MAML2-positive xenografts display increased IGF-1 expression. Histologic and immunohistochemical analysis of CRTC1/MAML2-positive (HMC1, HMC3A, and HMC3B) and CRTC1/MAML2- negative (A253) xenografts. Representative xenograft sections were formalin fixed, embedded, and H&E stained (top; magnification: 200x; scale bar: 100 μm), revealing epidermoid cells in CRTC1/MAML2-negative xenografts and epidermoid, mucus, and intermediate cells in the CRTC1/MAML2-positive xenografts. Mucicarmine staining (middle; magnification: 400x; scale bar: 100 μm) revealed mucus cells (indicated by yellow arrows) in CRTC1/MAML2-positive xenografts. IHC staining for IGF-1 (bottom; magnification: 200x; scale bar: 100 μm) demonstrated that IGF- 1 expression is increased in CRTC1/MAML2-positive xenografts relative to CRTC1/MAML2-negative xenografts.
78 Figure 10. A focused small molecule drug screen identifies sensitivity of CRTC1/MAML2-positive tumor cells to compounds targeting IGF-1R. (A) Drug screen performed in five CRTC1/MAML2-positive MEC cell lines and one CRTC1/MAML2-negative epidermoid carcinoma cell line. 176 inhibitors were tested, in duplicate, across all cell lines at six concentrations ranging from 10 nM to 10 μM. DMSO (1%) and Bortezomib (1 μM) were used as negative and positive controls, respectively, on each plate. Several IGF-1R and EGFR inhibitors emerged as top hits, inducing cell death in CRTC1/MAML2-positive cell lines at >5 standard deviations above baseline (DMSO). (B) Analysis of drug screen robustness based on a Z’-factor analysis. Data points in red indicate viability of cells treated with positive control (1 μM Bortezomib; 100% cell death) and data points in blue indicate viability of cells treated with negative control (DMSO; 0% cell death). Cell lines are listed along the x-axis, with CRTC1/MAML2 status indicated below in grey (negative) or red (positive). Z’-factor >0.75 confirms a robust dynamic range of drug effects indicating good drug screen reliability. (C) Relative cell death induced by drug screen inhibitors. IGF-1R and EGFR inhibitors more effectively induced cell death in CRTC1/MAML2-positive MEC cells compared with CRTC1/MAML2-negative epidermoid carcinoma cells.
79 Figure 11. Sensitivity of three CRTC1/MAML2-positive cell lines to IGF-1R inhibition. Representative dose response curves showing viability of three CRTC1/MAML2-positive cell lines (HMC1, HMC3A, and HMC3B) treated with increasing concentrations of IGF-1R inhibitors (BMS-754807 and PPP). IC50 values were calculated using a non- linear curve fit (log[agonist] versus response, four parameter-variable slope) in GraphPad Prism. Experiment was performed in biologic triplicate (n = 3 technical replicates for each experiment), with one representative experiment shown. See Table 2 for a summary of the IC50 values.
80 Figure 12. Validation of IGF-1R inhibition and shRNA-mediated knockdown. (A) Representative western blot showing expression of total and phosphorylated (Tyr1135) IGF-1R protein in HMC1 cells treated with BMS-754807 or PPP. (B) Quantification of phosphorylated (Tyr1135) IGF- 1R protein levels following IGF-1R inhibitor treatment. Band intensities were normalized to β-actin and to total IGF-1R and are shown as the fold change in phosphorylated band intensity relative to the untreated condition. Data are presented as the mean ± SEM (n = 3 biological replicates; one-way ANOVA with Holm-Sidak post hoc test; ns = not significant, *p ≤ 0.05). (C) qRT-PCR analysis of IGF-1R mRNA levels in cells transduced with lentiviruses expressing either a non-specific shRNA (shNS) or two different shRNAs targeting IGF-1R (shIGF-1R_#1 and shIGF-1R_#2). Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3; Student’s t-test; **p ≤ 0.01). (D) Representative western blot demonstrating relative knockdown of IGF-1Rβ protein levels with shIGF1R_#1 and shIGF1R_#2 relative to control shRNA (shNS). (E) Quantification of IGF-1R protein levels following IGF-1R knockdown. Band intensities were normalized to β-actin and are shown as the fold change in band intensity relative to the shNS condition. Data are presented as the mean ± SEM (n = 3 biological replicates; Student’s t-test; **p ≤ 0.01).
81 Figure 13. IGF-1R inhibition blocks MEC cell growth and survival. (A) Cell proliferation assay showing relative confluency of HMC3A cells treated with DMSO or PPP (see Table 2 for IC50 values). GC50 values for each condition are shown in hours and indicate the time required for cells to reach 50% confluency. ND: not determined. Experiment was performed in biologic triplicate (n = 4 technical replicates per experiment) with one representative experiment shown (mean ± SEM; Benjamini-Hochberg pairwise comparisons; **p ≤ 0.01, ***p ≤ 0.001). Data collected using InCucyte live cell imager. (B) Number of colonies formed by HMC3A cells treated with DMSO or PPP. A colony is defined as a cluster of ≥50 cells. Representative images of wells treated with DMSO and PPP (IC50 concentration; see Table 2 for IC50 values) are shown to the left. Data are presented as the mean ± SEM (n = 3 biological replicates; Student’s t-test; ****p ≤ 0.0001). (C) Number of colonies formed by HMC1 (left), HMC3A (center), and HMC3B (right) cells treated with DMSO or BMS-754807 (see Table 2 for IC50 values). A colony is defined as a cluster of ≥50 cells. Data are presented as the mean ± SEM (n = 3 biological replicates; Student’s t-test; *p ≤ 0.05, **p ≤ 0.01). (D) Left, Representative images of HMC3A tumorspheres on day one (top) and seven days post treatment with DMSO (vehicle control) or increasing concentrations of PPP. Right, Tumorsphere formation in HMC3A cells treated with DMSO or increasing concentrations of PPP. Percent change in tumorsphere area is calculated as [TumorAreaDayX]/[TumorAreaDay1]*100. >50 individual tumorspheres per condition were tracked for each individual experiment. Data are presented as the mean ± SEM (n = 3 biological replicates; Student’s t-test; ****p ≤ 0.0001). (E) Apoptosis levels (measured as % cleaved Caspase 3/7) in HMC3A cells treated with DMSO or PPP. Data was collected via flow cytometry and normalized to a non-stained control for each condition. Data are presented as the mean ± SEM (n = 3; Student’s t-test; **p ≤ 0.01).
82 Figure 14. CRTC1/MAML2 induces IGF-1 expression. (A, B) qRT-PCR analysis of CRTC1/MAML2 (C1/M2) (A) and IGF-1 (B) mRNA levels in HEK293-CMVTetRTetOC1/M2 cells treated with and without 1 μg/mL doxycycline. Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3; Student’s t-test; ***p < 0.001, ****p < 0.0001). (C) AlphaLISA-based analysis of secreted IGF-1 in media from HEK293-CMVTetRTetOC1/M2 cells with and without 1 μg/mL doxycycline treatment. Data are presented as the mean ± SEM (n = 3; Student’s t-test; **p = 0.003). (D) Western blotting analysis of IGF-1 and tubulin protein expression levels in HEK293-CMVTetRTetOC1/M2 cells treated with and without 1 μg/mL doxycycline. (E) Left, Schematic depicting the AlphaLISA assay. Donor and acceptor beads each carry an antibody that detects the analyte of interest. Donor beads can be excited by light at 680 nm and can then convert molecular oxygen to its excited state. Acceptor beads can utilize the excited oxygen molecule to emit light at 615 nm. The transfer of this excited molecular oxygen can only occur when donor and acceptor beads are brought into close proximity through mutual binding of the analyte. Thus, quantification of light at 615 nm allows for indirect quantification of analyte levels. Right, Standard curve showing the relationship between AlphaLISA signal and IGF-1 protein concentration.
83 Figure 15. CRTC1/MAML2 induces expression of the PGC-1α4 splice variant. (A) Chromatin immunoprecipitation analysis of Flag-CRTC1/MAML2 (Flag-C1/M2) promoter occupancy in HEK293- CMVTetRTetOC1/M2 cells. Data are expressed as promoter occupancy in doxycycline-treated cells normalized to promoter occupancy in vehicle-treated cells. GAPDH and NR4A2 promoters were used as negative and positive controls, respectively. Data are presented as the mean ± SEM (n = 3; Student’s t-test; ***p < 0.001, **p < 0.01). (B) Luciferase assay measuring activation of the distal and proximal PGC-1α promoters in HEK293-CMVTetRTetOC1/M2 cells with and without 1 μg/mL doxycycline treatment. Data are presented as the mean ± SEM (n = 3; Student’s t-test; ***p < 0.001). (C) qRT-PCR analysis of PGC-1α4 mRNA expression in HMC3A cells with and without shRNA- mediated CRTC1/MAML2 knockdown. Relative fold expression is shown normalized to RPL23 mRNA levels and to mock transduced condition. shNS: non-specific shRNA. shMAML2_#1 and shMAML2_#3: shRNAs targeting CRTC1/MAML2 and MAML2. Data are presented as the mean ± SEM (n = 3; Student’s t-test; **p < 0.01). (D) qRT- PCR analysis of CRTC1/MAML2 (C1/M2) (left) and NR4A2 (right) mRNA levels in HMC3A cells transduced with a non-specific shRNA (shNS) or two different shRNAs targeting CRTC1/MAML2 (shMAML2_#1 and shMAML2_#3). Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3; Student’s t-test; ns = not significant, *p ≤ 0.05, **p ≤ 0.01). (E) qRT- PCR analysis of PGC-1α4 mRNA levels in HMC3A-PGKTetOn3GTRE3GSdnCREB cells treated with and without 1 μg/mL doxycycline. Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3; Student’s t-test; ***p < 0.001).
84 Figure 16. The PGC-1α4 variant is highly expressed in CRTC1/MAML2-positive MEC. (A) Temporal analysis of CRTC1/MAML2 (C1/M2), PGC-1α4, and IGF-1 expression kinetics in HEK293-CMVTetRTetOC1/M2 cells treated with 1 μg/mL doxycycline for 0-48 hours. C1/M2 and PGC-1α4 expression is indicated on the left y-axis, while IGF-1 expression is indicated on the right y-axis. Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3; Student’s t-test; ***p < 0.001). (B) qRT-PCR analysis of PGC-1α4 expression in CRTC1/MAML2-positive and -negative cell lines. Cell lines are listed along the x-axis, with CRTC1/MAML2 status indicated below in grey (negative) or red (positive). Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3). (C) qRT-PCR analysis of PGC- 1α1, PGC-1α2, PGC-1α3, and PGC-1α4 mRNA levels in CRTC1/MAML2-positive and CRTC1/MAML2-negative cell lines. Cell lines are listed along the x-axis, with CRTC1/MAML2 status indicated below. Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3).
85 Figure 17. CRISPRi-based knockdowns confirm that IGF-1 is regulated by a specific PGC-1α isoform. (A) qRT-PCR analysis of IGF-1 mRNA levels in HMC3A cells transduced with unique gRNAs targeting the PGC-1a1 isoform only (PGC-1a1 sgRNA) or all PGC-1a isoforms (Pan-PGC-1a sgRNA). Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3; Student’s t-test; ***p < 0.001). (B) qRT-PCR analysis of PGC-1α1 and PGC-1α4 mRNA levels in HMC3A cells transduced with unique gRNAs targeting the canonical PGC-1α1 transcript only or all PGC-1α isoforms. Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3; Student’s t-test; *p < 0.05, **p < 0.01).
86 Figure 18. Overexpression of PGC-1α4 drives IGF-1 transcription. (A) qRT-PCR analysis of PGC-1α4 mRNA levels in HEK293- PGKTetOn3GTRE3GSPGC-1α4 cells treated with increasing concentrations of doxycycline for 24 hr. Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3 biological replicates with 3 technical replicates each; Student’s t-test; ****p < 0.0001). (B) Western blot showing Flag- PGC-1α4 and tubulin protein expression levels in HEK293-PGKTetOn3GTRE3GSPGC-1α4 cells treated with increasing concentrations of doxycycline for 24 hr. Blot is representative of 3 independent biological replicates. (C) qRT-PCR analysis of PGC-1α4 and IGF-1 mRNA expression in HEK293-PGKTetOn3GTRE3GSPGC-1α4 cells treated with 1 μg/mL doxycycline for 0-48 hours. Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3; Student’s t-test; **p < 0.01, ****p < 0.0001).
87 Figure 19. Knockdown of key signaling components inhibits proliferation of MEC cells. Cell proliferation assays showing relative confluency of HMC3A cells with and without shRNA-mediated knockdown of PGC-1α (left), IGF-1R (center), or CRTC1/MAML2 (C1/M2) (right). Time to 50% confluence is shown by dashed lines. shNS = non- specific shRNA; shPGC-1α_#1 and _#2 = shRNAs targeting PGC-1α; shIGF-1R_#1 and _#2 = shRNAs targeting IGF-1R; shMAML2_#3 = shRNA targeting CRTC1/MAML2 and MAML2. Experiment was performed in biologic triplicate (n = 3 technical replicates per experiment) with one representative experiment shown (mean ± SEM; Benjamini-Hochberg pairwise comparisons; *p < 0.05, ****p < 0.0001). Data collected using BioTek Cytation 5 plate reader.
88 Figure 20. CRTC1/MAML2 overexpression activates IGF-1 and PPARγ-response element luciferase reporters. (A) Schematic of the human IGF-1 gene locus, highlighting the P2 promoter (located upstream of the second exon) as well as several PPARγ (PPARG) binding motifs located within this promoter region. (B, C) Light output generated by an IGF-1 P2 promoter-driven luciferase reporter (B) or a PPARγ-response element (PPRE)-driven luciferase reporter (C) in HEK293-CMVTetRTetOC1/M2 cells treated with and without 1 μg/mL doxycycline treatment. Data are presented as the mean ± SEM (n = 3; Student’s t-test; *p ≤ 0.05, **p ≤ 0.01, ***p < 0.001).
89 Figure 21. IGF-1 activation is dependent on PPARγ-mediated signaling. PPARγ-response element (PPRE)- driven (A, C) and IGF-1 promoter-driven (B, D) luciferase reporter assay showing activation and repression of basal activity of endogenously expressed PPARγ in doxycycline-treated HEK293-CMVTetRTetOC1/M2 cells. Cells were treated with increasing concentrations of the PPARγ agonist GW1929 (A, B) or the PPARγ inverse agonist SR10221(C, D) for 24 hr prior to measuring luciferase activity. IC50 values were calculated using a non-linear curve fit (log[agonist] vs response, four parameter-variable slope) in GraphPad Prism. Experiment was performed in biologic triplicate (n = 3 technical replicates per experiment) with one representative experiment shown (mean ± SEM).
90 Figure 22. PPARγ knockdown inhibits IGF-1 expression and induces apoptosis in MEC cells. (A) Representative western blot demonstrating knockdown of PPARγ protein levels using three separate PPARγ-targeted shRNAs (#1, #2, and #3) relative to a control shRNA (shNS). IGF-1 levels were also detected and were observed to decrease upon PPARγ knockdown. (B) Quantification of PPARγ and IGF-1 protein levels in PPARγ knockdown samples. Band intensities were normalized to β-actin and are shown as the fold change in band intensity relative to the shNS-treated condition. Data are presented as the mean ± SEM (n = 3 biological replicates; Student’s t-test; **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001). (C) Apoptosis levels (measured as % cleaved Caspase 3/7) in HMC1 and HMC3A cells transduced with three separate PPARγ-targeted shRNAs (#1, #2, and #3) or a control shRNA (shNS). 48 hr after transduction, data was collected via flow cytometry and normalized to a non-stained control for each condition. Data are presented as the mean ± SEM (n = 3; Student’s t-test; *p ≤ 0.05).
91 Figure 23. PPARγ inverse agonists inhibit IGF-1 expression. (A) Dose response curves showing viability of three CRTC1/MAML2-positive cell lines (HMC1, HMC3A, and HMC3B) treated with increasing concentrations of PPARγ inverse agonists (SR10221, SR2595, and T0070907). IC50 values were calculated using a non-linear curve fit (log[agonist] vs response, four parameter-variable slope) in GraphPad Prism. Experiment was performed in biologic triplicate (n = 3 technical replicates for each experiment), with one representative experiment shown. See Table 3 for IC50 values. (B) qRT- PCR analysis of IGF-1, PGK1, and PKM2 expression in HMC3A cells treated with vehicle (DMSO) or SR2595. Relative fold expression is shown normalized to RPL23 mRNA levels. Data are presented as the mean ± SEM (n = 3; Student’s t-test; ***p ≤ 0.001). (C) Representative western blot showing expression of IGF-1 protein in HMC3A cells treated with increasing concentrations of SR2595. (D) Quantification of IGF-1 protein levels in SR2595-treated HMC3A cells. Band intensities were normalized to β-actin and are shown as the fold change in band intensity relative to the vehicle-treated condition. Data are presented as the mean ± SEM (n = 3 biological replicates; Student’s t-test; ns = not significant, *p ≤ 0.05).
92 Figure 24. PPARγ inverse agonists suppress MEC cell growth, proliferation, and survival. (A) Cell proliferation assay showing relative confluency of HMC3A cells treated with DMSO or SR10221. Experiment was performed in biologic triplicate (n = 4 technical replicates per experiment) with one representative experiment shown (mean ± SEM; Benjamini-Hochberg pairwise comparisons; ***p ≤ 0.001). Data collected using IncuCyte live cell imager. (B) Colony formation in HMC1 cells treated with DMSO or SR10221 (see Table 3 for IC50 values). A colony is defined as a cluster of ≥50 cells. Data are presented as the mean ± SEM (n = 3; Student’s t-test; ***p ≤ 0.001). (C) Tumorsphere formation in HMC3A cells treated with DMSO or SR10221. Percent change in tumorsphere area is calculated as [TumorAreaDayX]/[TumorAreaDay1]*100. >50 individual tumorspheres per condition were tracked for each individual experiment. Data are presented as the mean ± SEM (n = 3 biological replicates; Student’s t-test; **p ≤ 0.01, ***p ≤ 0.001). (D) Apoptosis levels (measured as % cleaved Caspase 3/7) in HMC3A cells treated with DMSO or SR10221. Data was collected via flow cytometry and normalized to a non-stained control for each condition. Data are presented as the mean ± SEM (n = 3; Student’s t-test; ***p ≤ 0.001, ****p ≤ 0.0001).
93 Figure 25. PPARγ inverse agonists suppress tumor growth in a xenograft model of salivary MEC. (A) Growth of HMC3A xenograft tumors over time during treatment with either SR10221 or a vehicle control. Tumor xenograft volume was quantified by caliper measurements obtained on the indicated days following initiation of drug administration. Decreased tumor volume in the SR10221 cohort correlates directly with suppression of tumor growth in vivo (n = 4, mean ± SEM; Student’s t-test; ****p ≤ 0.0001). (B) Weight of resected vehicle- and SR10221-treated HMC3A tumor xenografts at study endpoint (n = 4, mean ± SEM; Student’s t-test; **p ≤ 0.01). (C) Representative in vivo bioluminescent images obtained at endpoint of Nu/Nu mice bearing subcutaneous HMC3A tumor xenografts and treated with either SR2595 or a vehicle control. Images were captured with an open filter and an acquisition time of 5 min (binning = 2; FOV = 25; Fstop = 2; object height = 1.5). (D) Weights of Nu/Nu mice bearing subcutaneous HMC3A xenograft tumors and treated with SR2595 or a vehicle control for up to 24 days (n = 4, mean ± SEM). (E) Comparison of HMC3A xenograft tumor growth following treatment with either SR2595 or vehicle control. Bioluminescence imaging (BLI) signals from tumor xenografts were quantified by region of interest (ROI) analysis of images obtained at endpoint (day 24 following initiation of drug administration). Decreased BLI signal in the SR2595 cohort correlates directly with suppression of tumor growth in vivo (n = 4, mean ± SEM; Student’s t-test; **p ≤ 0.01). (F, G) Volume (F) and weight (G) of resected vehicle- and SR2595-treated HMC3A xenograft tumors at study endpoint (n = 4, mean ± SEM; Student’s t-test; *p ≤ 0.05).
94 Figure 26. Schematic of the CRTC1/MAML2 – PGC-1α4 – IGF-1 signaling pathway. CRTC1/MAML2 acts as a rogue coactivator of the transcription factor CREB. In MEC, this interaction drives the production of a minor PGC-1α splice variant, PGC-1α4. PGC-1α4 in turn can coactivate the transcription factor PPARγ to stimulate production of the IGF-1 growth hormone. Secreted IGF-1 ligand can bind to cell surface IGF-1R receptors and stimulate key signaling pathways driving survival, growth, and proliferation, all of which are key tumor-supportive functions. Our study demonstrated that this CRTC1/MAML2 mediated signaling cascade can be inhibited using either PPARγ inverse agonists (such as SR10221 or SR2595) or IGF-1R antagonists (such as PPP or BMS-754807).
95 Figure 27. Cells targeted by Cre promoters used in transgenic lines. Dcpp1 is expressed in the intercalated duct cells and serous demilune cells. Mist1 is expressed in acinar cells. Krt14 is expressed in excretory, striated, and intercalated duct cells. Adapted from Maruyama et al., 2016.
96 Figure 28. Bioluminescent imaging demonstrates transgene expression patterns in LumiFluor-positive animals. Representative in vivo bioluminescent images of control (LumiFluor-negative), K14Cre-LumiFluor, MistCre- LumiFluor, and DcppCre-LumiFluor animals obtained 3- and 9-months post tamoxifen administration. Images were captured with an open filter and an acquisition time of 5 min (binning = 2; FOV = 25; Fstop = 2; object height = 1.5).
97 Figure 29. CRTC1/MAML2 is expressed in salivary gland tissue of DcppCre-C1/M2, MistCre-C1/M2, and K14Cre- C1/M2 transgenic mice. qRT-PCR analysis of CRTC1/MAML2 expression in control (CRTC1/MAML2-negative), DcppCre-C1/M2, MistCre-C1/M2, and K14Cre-C1/M2 animals 4 months post tamoxifen administration. CRTC1/MAML2 copy number per 10 ng input RNA was calculated based on a standard curve. Data is presented as the mean ± SEM (n = 3). SMG: submandibular gland. SLG: sublingual gland. PG: parotid gland.
98 Figure 30. Ductal expression of CRTC1/MAML2 drives a dysplastic and hyperproliferative phenotype. Representative H&E images from control (CRTC1/MAML2-negative), MistCre-C1/M2, DcppCre-C1/M2, and K14Cre- C1/M2 salivary glands ~6 months post tamoxifen administration. SMG: submandibular gland. SLG: sublingual gland. PG: parotid gland. Scale bar: 50 μm.
99 Figure 31. CRTC1/MAML2 overexpression alone fails to induce MEC formation independent of Cre driver. Disease (MEC)-free survival (DSS) and overall survival (OS) of control (CRTC1/MAML2-negative), MistCre-C1/M2, DcppCre-C1/M2, and K14Cre-C1/M2 animals 18 months post-tamoxifen induction.
100 C1/M2 LumiFluor Figure 32. Generation of Ductal-PTight PGK stable cells. (A) Schematic of engineered cell lines. In the tdTO LumiFluor Ductal-PTight PGK cells, the TdTomato gene is under the control of a doxycycline-inducible PTight promoter, while the GpNLuc (LumiFluor) reporter is driven by a constitutively active PGK promoter. In the Ductal- C1/M2 LumiFluor PTight PGK cells, CRTC1/MAML2 is under the control of a doxycycline-inducible PTight promoter, while the LumiFluor reporter is driven by a constitutively active PGK promoter. (B) Flow cytometric analysis of GpNLuc and tdTO LumiFluor TdTomato reporter expression in parental ductal cells (blue trace), uninduced Ductal-PTight PGK cells (green tdTO LumiFluor trace), and doxycycline induced Ductal-PTight PGK cells. (C) Quantification of GpNLuc (green line) and tdTO LumiFluor TdTomato (red line) expression in parental ductal cells (DC) and Ductal-PTight PGK cells treated with 0-2 μg/mL doxycycline. Data was collected via flow cytometry and was normalized to the background fluorescence from tdTO LumiFluor parental ductal cells (0% induction) and to the highest fluorescence achieved in induced Ductal-PTight PGK cells (100% induction). (D) qRT-PCR (bar graph) and western blot (inset) analysis of CRTC1/MAML2 (C1/M2) tdTO LumiFluor expression in Ductal-PTight PGK cells with and without 0.5 μg/mL doxycycline treatment. For qRT-PCR, C1/M2 copy number per 10 ng input RNA was calculated based on a standard curve. (E) Luciferase assay measuring tdTO LumiFluor activation of the distal PGC-1α promoter in Ductal-PTight PGK cells with and without 1 μg/mL doxycycline treatment
101 Figure 33. CRTC1/MAML2-positive human MEC tumors display increased p53 dysregulation. Box and whisker plots showing correlation of high-grade MEC, low-grade MEC, and normal salivary gland samples with a breast cancer-derived p53 gene signature (Troester et al., 2006). CRTC1/MAML2-positive and -negative samples are shown by red and black points, respectively.
102 Figure 34. Salivary MEC tumors arise in K14Cre-C1/M2-p53f/f mice four months after tamoxifen administration. (A) Disease (MEC)-specific survival in K14Cre-C1/M2-p53f/f animals compared with all other animals in this study (CRTC1/MAML2-positive and -negative, p53 wild-type, p53f/+, p16 wild-type, p16f/+, and p16f/f). (B-C) Representative tumor images from K14Cre-C1/M2-p53f/f mice. (D) Representative H&E-stained murine MEC tumor from a K14Cre- C1/M2-p53f/f mouse. Colored arrows indicate mucous (white), epidermoid (black), and intermediate (blue) cells. Scale bar: 50 μm. (E) Representative mucicarmine-stained MEC tumor from the same animal as in (D). Scale bar: 50 μm.
103 Figure 35. MSigDB pathways dysregulated in MEC GEMM tumors. Heat map showing significantly differentially expressed gene lists (p < 0.05) within the Molecular Signatures Database (MSigDB). Tumor and normal samples are indicated above in red and black, respectively. Green indicates pathway upregulation, while red indicates pathway downregulation.
104 Figure 36. Musicant_MEC_CRTC1-MAML2_IGF1 gene set distinguishes MEC GEMM tumors. Heat map showing differential expression of genes in the curated Musicant_MEC_CRTC1-MAML2_IGF1 gene set in MEC GEMM tumors. Tumor gene expression was compared to gene expression in matched normal submandibular gland tissues. Normal and tumor samples are indicated above in red and blue, respectively.
105 Figure 37. MEC GEMM tumors mimic both low- and high-grade human MEC. (A) Heat map of significant DEGs between low- and high-grade MEC and normal salivary gland samples. Low- and high-grade MEC are indicated at the top in red and blue, respectively. Normal samples are indicated in green. Hierarchical clustering was performed using ComplexHeatmap R package version 2.0.0. (B) Gene expression data from MEC GEMM tumors were compared to gene expression data from low, intermediate, and high grade CRTC1/MAML2-positive human MEC tumors. Correlation between each GEMM sample was calculated such that a score of 1 indicates perfect correlation. Two MEC GEMM tumor samples (GEMMSEQ005_C1 and GEMSEQ008_C4) displayed a strong correlation with gene expression profiles seen in high grade CRTC1/MAML2-positive human MEC tumors. The remaining MEC GEMM tumor sample (GEMMSEQ006_C2) correlated best with the gene expression profiles of low grade CRTC1/MAML2-positive human MEC tumors.
106 Figure 38. Krt14-driven oncogene expression results in significant oral lesion formation. (A) Significant oral lesions are formed by Krt14-driven gene expression. (B-C) Representative H&E images from the oral tissue shown in (A). Scale bar: 100 μm. (D-E) Representative mucicarmine staining from the oral tissue shown in (A). Scale bar: 100 μm.
107 Figure 39. Direct injection of adenoviral Cre results in temporary LumiFluor signal. (A-C) Representative in vivo bioluminescent images of LSL-LumiFluor animals 10 days (A), 27 days (B), and 60 days (C) after being injected trans-dermally with either Ad-GFP (control, left) or Ad-Cre (middle and right). Images were captured with an open filter and an acquisition time of 5 min (binning = 2; FOV = 25; Fstop = 2; object height = 1.5). (D) Bioluminescence imaging (BLI) signals were quantified by region of interest (ROI) analysis at the indicated time points following adenovirus injection (n = 3 in each group; mean ± SEM).
108 REFERENCES
Adams, A., Warner, K., Pearson, A.T., Zhang, Z., Kim, H.S., Mochizuki, D., Basura, G., Helman, J., Mantesso, A., Castilho, R.M., et al. (2015). ALDH/CD44 identifies uniquely tumorigenic cancer stem cells in salivary gland mucoepidermoid carcinomas. Oncotarget 6, 26633– 26650.
Ahn, S., Olive, M., Aggarwal, S., Krylov, D., Ginty, D.D., and Vinson, C. (1998). A Dominant- Negative Inhibitor of CREB Reveals that It Is a General Mediator of Stimulus-Dependent Transcription of c-fos.
Amano, O., and Iseki, S. (1993). Expression, localization and developmental regulation of insulin-like growth factor I mRNA in rat submandibular gland. Archives of Oral Biology 38, 671–677.
Amelio, A.L., Caputi, M., and Conkright, M.D. (2009). Bipartite functions of the CREB co- activators selectively direct alternative splicing or transcriptional activation. EMBO Journal 28, 2733–2747.
Amelio, A.L., Fallahi, M., Schaub, F.X., Zhang, M., Lawani, M.B., Alperstein, A.S., Southern, M.R., Young, B.M., Wu, L., Zajac-Kaye, M., et al. (2014). CRTC1/MAML2 gain-of-function interactions with MYC create a gene signature predictive of cancers with CREB-MYC involvement. Proceedings of the National Academy of Sciences of the United States of America 111.
Babina, I.S., and Turner, N.C. (2017). Advances and challenges in targeting FGFR signalling in cancer. Nature Reviews Cancer 17, 318–332.
Bailleul, B., Surani, M.A., White, S., Barton, S.C., Brown, K., Blessing, M., Jorcano, J., and Balmain, A. (1990). Skin Hyperkeratosis and Papilloma Formation in Transgenic Mice Expressing a ras Oncogene from a Suprabasal Keratin Promoter. Cell 62, 697–708.
Batsakis, J.G. (1980). Salivary gland neoplasia: An outcome of modified morphogenesis and cytodifferentiation. Oral Surgery 49, 229–232.
Bean, G.R., Krings, G., Otis, C.N., Solomon, D.A., García, J.J., van Zante, A., Camelo-Piragua, S., van Ziffle, J., and Chen, Y.Y. (2019). CRTC1–MAML2 fusion in mucoepidermoid carcinoma of the breast. Histopathology 74, 463–473.
Behboudi, A., Enlund, F., Winnes, M., Andrén, Y., Nordkvist, A., Leivo, I., Flaberg, E., Szekely, L., Mäkitie, A., Grenman, R., et al. (2006). Molecular classification of mucoepidermoid carcinomas—Prognostic significance of theMECT1–MAML2 fusion oncogene. Genes, Chromosomes and Cancer 45, 470–481.
109 Bekhor, I., Wen, Y., Shi, S., Hsieh, C.-H., Denny, P.A., and Denny, P.C. (1994). cDNA cloning, sequencing and in situ localization of a transcript specific to both sublingual demilune cells and parotid intercalated duct cells in mouse salivary glands. Archs Oral Biol 39, 1011–1022.
Bell, D., and Hanna, E.Y. (2012). Salivary gland cancers: Biology and molecular targets for therapy. Current Oncology Reports 14, 166–174.
Ben-Porath, I., Thomson, M.W., Carey, V.J., Ge, R., Bell, G.W., Regev, A., and Weinberg, R.A. (2008). An embryonic stem cell-like gene expression signature in poorly differentiated aggressive human tumors. Nature Genetics 40, 499–507.
Betz, S.J. (2017). Evaluating Histologic Grading Systems and the Expression of Human Cytomegalovirus in Salivary Gland Mucoepidermoid Carcinoma.
Bevill, S.M., Olivares-Quintero, J.F., Sciaky, N., Golitz, B.T., Singh, D., Beltran, A.S., Rashid, N.U., Stuhlmiller, T.J., Hale, A., Moorman, N.J., et al. (2019). GSK2801, a BAZ2/BRD9 bromodomain inhibitor, synergizes with BET inhibitors to induce apoptosis in triple- negative breast cancer. Molecular Cancer Research 17, 1503–1518.
Birkeland, A.C., Foltin, S.K., Michmerhuizen, N.L., Hoesli, R.C., Rosko, A.J., Byrd, S., Yanik, M., Nor, J.E., Bradford, C.R., Prince, M.E., et al. (2017). Correlation of Crtc1/3-Maml2 fusion status, grade and survival in mucoepidermoid carcinoma. Oral Oncology 68, 5–8.
Bishop, T.R., Zhang, Y., and Erb, M.A. (2019). Pharmacological Modulation of Transcriptional Coregulators in Cancer. Trends in Pharmacological Sciences 40, 388–402.
Bleijs, M., Wetering, M., Clevers, H., and Drost, J. (2019). Xenograft and organoid model systems in cancer research. The EMBO Journal 1–11.
Boehm, T., Mengle-Gaw, L., Kees, U.R., Spurr, N., Lavenir, I., Forster, A., and Rabbitts, T.H. (1989). Alternating purine-pyrimidine tracts may promote chromosomal translocations seen in a variety of human lymphoid tumours. EMBO Journal 8, 2621–2631.
Brandwein, M.S., Ivanov, K., Wallace, D.I., Hille, J.J., Wang, B., Fahmy, A., Bodian, C., Urken, M.L., Gnepp, D.R., Huvos, A., et al. (2001). Mucoepidermoid carcinoma: A clinicopathologic study of 80 patients with special reference to histological grading. American Journal of Surgical Pathology 25, 835–845.
Brown, K., Strathdee, D., Bryson, S., Lambie, W., and Balmain, A. (1998). The malignant capacity of skin tumours induced by expression of a mutant H-ras transgene depends on the cell type targeted. Current Biology 8, 516–524.
Bruno, N.E., Kelly, K.A., Hawkins, R., Bramah-Lawani, M., Amelio, A.L., Nwachukwu, J.C., Nettles, K.W., and Conkright, M.D. (2014). Creb coactivators direct anabolic responses and enhance performance of skeletal muscle. EMBO Journal 33, 1027–1043.
110 Canettieri, G., Coni, S., della Guardia, M., Nocerino, V., Antonucci, L., di Magno, L., Screaton, R., Screpanti, I., Giannini, G., and Gulino, A. (2009). The coactivator CRTC1 promotes cell proliferation and transformation via AP-1. PNAS 106, 1445–1450.
Carper, M.B., Troutman, S., Wagner, B.L., Byrd, K.M., Selitsky, S.R., Parag-Sharma, K., Henry, E.C., Li, W., Parker, J.S., Montgomery, S.A., et al. (2019). An Immunocompetent Mouse Model of HPV16(+) Head and Neck Squamous Cell Carcinoma. Cell Reports 29, 1660- 1674.e7.
Chavez-Reyes, A., Parant, J.M., Amelse, L.L., de Oca Luna, R.M., Luna, O., Korsmeyer, S.J., and Lozano, G. (2003). Switching Mechanisms of Cell Death in mdm2-and mdm4-null Mice by Deletion of p53 Downstream Targets. Cancer Research 63, 8664–8669.
Chen, A., and Koehler, A.N. (2020). Transcription Factor Inhibition: Lessons Learned and Emerging Targets. Trends in Molecular Medicine 26, 508–518.
Chen, H.X., and Sharon, E. (2013). IGF-1R as an anti-cancer target - trials and tribulations. Chinese Journal of Cancer 32, 242–252.
Chen, A.M., Granchi, P.J., Garcia, J., Bucci, M.K., Fu, K.K., and Eisele, D.W. (2007). Local-regional recurrence after surgery without postoperative irradiation for carcinomas of the major salivary glands: Implications for adjuvant therapy. International Journal of Radiation Oncology Biology Physics 67, 982–987.
Chen, A.M., Lau, V.H., Farwell, D.G., Luu, Q., and Donald, P.J. (2013). Mucoepidermoid carcinoma of the parotid gland treated by surgery and postoperative radiation therapy: Clinicopathologic correlates of outcome. Laryngoscope 123, 3049–3055.
Chen, J., Li, J.L., Chen, Z., Griffin, J.D., and Wu, L. (2015). Gene expression profiling analysis of CRTC1-MAML2 fusion oncogene-induced transcriptional program in human mucoepidermoid carcinoma cells. BMC Cancer 15, 1–13.
Chen, Z., Chen, J., Gu, Y., Hu, C., Li, J.L., Lin, S., Shen, H., Cao, C., Gao, R., Li, J., et al. (2014). Aberrantly activated AREG-EGFR signaling is required for the growth and survival of CRTC1-MAML2 fusion-positive mucoepidermoid carcinoma cells. Oncogene 33, 3869– 3877.
Chenevert, J., Barnes, L.E., and Chiosea, S.I. (2011). Mucoepidermoid carcinoma: A five-decade journey. Virchows Archiv 458, 133–140.
Chiosea, S.I., Dacic, S., Nikiforova, M.N., and Seethala, R.R. (2012). Prospective testing of mucoepidermoid carcinoma for the MAML2 translocation: Clinical Implications. Laryngoscope 122, 1690–1694.
Chu, P.G., and Weiss, L.M. (2002). Keratin expression in human tissues and neoplasms.
111 Cipriani, N.A., Lusardi, J.J., Mcelherne, J., Pearson, A.T., Olivas, A.D., Fitzpatrick, C., Lingen, M.W., and Blair, E.A. (2019a). Mucoepidermoid Carcinoma: A Comparison of Histologic Grading Systems and Relationship to MAML2 Rearrangement and Prognosis. American Journal of Surgical Pathology 43, 885–897.
Cipriani, N.A., Lusardi, J.J., Mcelherne, J., Pearson, A.T., Olivas, A.D., Fitzpatrick, C., Lingen, M.W., and Blair, E.A. (2019b). Mucoepidermoid Carcinoma: A Comparison of Histologic Grading Systems and Relationship to MAML2 Rearrangement and Prognosis. American Journal of Surgical Pathology 43, 885–897.
Coca-Pelaz, A., Rodrigo, J.P., Triantafyllou, A., Hunt, J.L., Rinaldo, A., Strojan, P., Haigentz, M., Mendenhall, W.M., Takes, R.P., vander Poorten, V., et al. (2015). Salivary mucoepidermoid carcinoma revisited. European Archives of Otorhinolaryngology 272, 799–819.
Coleman, M.P. (2013). War on cancer and the influence of the medical-industrial complex. Journal of Cancer Policy 1, e31–e34.
Cortez, A.P., Holanda, M.E., Mororó, E.S., Rego, P.H.C.L. do, and Sales, R.H. (2018). Mucoepidermoid thyroid carcinoma combined with palate squamous cell carcinoma. Archives of Head and Neck Surgery 47.
Coxon, A., Rozenblum, E., Park, Y.S., Joshi, N., Tsurutani, J., Dennis, P.A., Kirsch, I.R., and Kaye, F.J. (2005). Mect1-Maml2 fusion oncogene linked to the aberrant activation of cyclic AMP/CREB regulated genes. Cancer Research 65, 7137–7144.
Curtarelli, R.B., Gonçalves, J.M., dos Santos, L.G.P., Savi, M.G., Nör, J.E., Mezzomo, L.A.M., and Rodríguez Cordeiro, M.M. (2018). Expression of Cancer Stem Cell Biomarkers in Human Head and Neck Carcinomas: a Systematic Review. Stem Cell Reviews and Reports 14, 769–784.
Dang, T.P., Eichenberger, S., Gonzalez, A., Olson, S., and Carbone, D.P. (2003). Constitutive activation of Notch3 inhibits terminal epithelial differentiation in lungs of transgenic mice. Oncogene 22, 1988–1997.
Denduluri, S.K., Idowu, O., Wang, Z., Liao, Z., Yan, Z., Mohammed, M.K., Ye, J., Wei, Q., Wang, J., Zhao, L., et al. (2015). Insulin-like growth factor (IGF) signaling in tumorigenesis and the development of cancer drug resistance. Genes & Diseases 2, 13–25.
DeVita Jr, V.T. (2004). The “War on Cancer” and its impact. Nature Clinical Practice Oncology 1, 55–55.
Dobin, A., Davis, C.A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut, P., Chaisson, M., and Gingeras, T.R. (2013). STAR: Ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21.
112 El-Naggar, A.K., Lovell, M., Killary, A.M., Clayman, G.L., and Batsakis, J.G. (1996). A mucoepidermoid carcinoma of minor salivary gland with t(11;19)(q21;p13.1) as the only karyotypic abnormality. Cancer Genetics and Cytogenetics 87, 29–33.
Emmerson, E., May, A.J., Nathan, S., Cruz-Pacheco, N., Lizama, C.O., Maliskova, L., Zovein, A.C., Shen, Y., Muench, M.O., and Knox, S.M. (2017). SOX2 regulates acinar cell development in the salivary gland.
Eveson, J.W. (2011). Salivary tumours. Periodontology 2000 57, 150–159.
Fernandez-Cuesta, L., Sun, R., Menon, R., George, J., Lorenz, S., Meza-Zepeda, L.A., Peifer, M., Plenker, D., Heuckmann, J.M., Leenders, F., et al. (2015). Identification of novel fusion genes in lung cancer using breakpoint assembly of transcriptome sequencing data. Genome Biology 16.
Fishbein, L., Leshchiner, I., Walter, V., Danilova, L., Robertson, A.G., Johnson, A.R., Lichtenberg, T.M., Murray, B.A., Ghayee, H.K., Else, T., et al. (2017). Comprehensive Molecular Characterization of Pheochromocytoma and Paraganglioma. Cancer Cell 31, 181–193.
Frattini, V., Pagnotta, S.M., Tala, Fan, J.J., Russo, M. v., Lee, S.B., Garofano, L., Zhang, J., Shi, P., Lewis, G., et al. (2018). A metabolic function of FGFR3-TACC3 gene fusions in cancer. Nature 553, 222–227.
Frese, K.K., and Tuveson, D.A. (2007). Maximizing mouse cancer models. Nature Reviews Cancer 7, 645–658.
Fulco, C.P., Munschauer, M., Anyoha, R., Munson, G., Grossman, S.R., Perez, E.M., Kane, M., Cleary, B., Lander, E.S., and Engreitz, J.M. (2016). Systematic mapping of functional enhancer-promoter connections with CRISPR interference. Science 354, 769–773.
Gao, R., Cao, C., Zhang, M., Lopez, M.-C., Yan, Y., Chen, Z., Mitani, Y., Zhang, L., Zajac-Kaye, M., Liu, B., et al. (2014). A unifying gene signature for adenoid cystic cancer identifies parallel MYB-dependent and MYB-independent therapeutic targets. Oncotarget 5, 12528–12542.
Gao, X., Oei, M.S., Ovitt, C.E., Sincan, M., and Melvin, J.E. (2018). Transcriptional profiling reveals gland-specific differential expression in the three major salivary glands of the adult mouse. Physiological Genomics 50, 263–271.
Gilbert, L.A., Larson, M.H., Morsut, L., Liu, Z., Brar, G.A., Torres, S.E., Stern-Ginossar, N., Brandman, O., Whitehead, E.H., Doudna, J.A., et al. (2013). CRISPR-Mediated Modular RNA-Guided Regulation of Transcription in Eukaryotes. Cell 154, 442–451.
Goldstein, J.T., Berger, A.C., Shih, J., Duke, F.F., Furst, L., Kwiatkowski, D.J., Cherniack, A.D., Meyerson, M., and Strathdee, C.A. (2017). Genomic activation of PPARG reveals a candidate therapeutic axis in bladder cancer. Cancer Research 77, 6987–6998.
113 Goode, R.K., Auclair, P.L., and Ellis, G.L. (1998). Mucoepidermoid carcinoma of the major salivary glands. Cancer 82, 1217–1224.
Gu, Z., Eils, R., and Schlesner, M. (2016). Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 32, 2847–2849.
Gualberto, A., and Pollak, M. (2009). Emerging role of insulin-like growth factor receptor inhibitors in oncology: Early clinical trial results and future directions. Oncogene 28, 3009–3021.
Guo, X. qin, Li, B., Li, Y., Tian, X. ying, and Li, Z. (2014). Unusual mucoepidermoid carcinoma of the liver misdiagnosed as squamous cell carcinoma by intraoperative histological examination. Diagnostic Pathology 9.
Hoadley, K.A., Yau, C., Hinoue, T., Wolf, D.M., Lazar, A.J., Drill, E., Shen, R., Taylor, A.M., Cherniack, A.D., Thorsson, V., et al. (2018). Cell-of-Origin Patterns Dominate the Molecular Classification of 10,000 Tumors from 33 Types of Cancer. Cell 173, 291- 304.e6.
Hoffman, M.P., Kidder, B.L., Steinberg, Z.L., Lakhani, S., Ho, S., Kleinman, H.K., and Larsen, M. (2002). Gene expression profiles of mouse submandibular gland development: FGFR1 regulates branching morphogenesis in vitro through BMP- and FGF- dependent mechanisms. Development 129, 5767–5778.
Holmberg, K. v., and Hoffman, M.P. (2014). Anatomy, biogenesis and regeneration of salivary glands. Monographs in Oral Science 24, 1–13.
Holsinger, F.C., and Bui, D.T. (2007). Anatomy, function, and evaluation of the salivary glands. In Salivary Gland Disorders, pp. 1–16.
Huo, Z., Wu, H., Li, J., Li, S., Wu, S., Liu, Y., Luo, Y., Cao, J., Zeng, X., and Liang, Z. (2015). Primary pulmonary mucoepidermoid carcinoma: Histopathological and moleculargenetic studies of 26 cases. PLoS ONE 10, 1–18.
Jang, Y.E., Jang, I., Kim, S., Cho, S., Kim, D., Kim, K., Kim, J., Hwang, J., Kim, S., Kim, J., et al. (2020). ChimerDB 4.0: An updated and expanded database of fusion genes. Nucleic Acids Research 48, D817–D824.
Jee, K.J., Persson, M., Heikinheimo, K., Passador-Santos, F., Aro, K., Knuutila, S., Odell, E.W., Mäkitie, A., Sundelin, K., Stenman, G., et al. (2013). Genomic profiles and CRTC1-MAML2 fusion distinguish different subtypes of mucoepidermoid carcinoma. Modern Pathology 26, 213–222.
Kadoch, C., and Crabtree, G.R. (2013). Reversible disruption of mSWI/SNF (BAF) complexes by the SS18-SSX oncogenic fusion in synovial sarcoma. Cell 153, 71–85.
114 Kaiser, J., and Couzin-Frankel, J. (2016). Biden seeks clear course for his cancer moonshot. Science 351, 325–326.
Kang, H., Tan, M., Bishop, J.A., Jones, S., Sausen, M., Ha, P.K., and Agrawal, N. (2017). Whole- exome sequencing of salivary gland mucoepidermoid carcinoma. Clinical Cancer Research 23, 283–288.
Kardos, J., Chai, S., Mose, L.E., Selitsky, S.R., Krishnan, B., Saito, R., Iglesia, M.D., Milowsky, M.I., Parker, J.S., Kim, W.Y., et al. (2016). Claudin-low bladder tumors are immune infiltrated and actively immune suppressed. JCI Insight 1.
Kerr, M., Lee, A., Wang, P.L., Purushotham, K.R., Chegini, N., Yamamoto, H., and Humphreys- Beher, M.G. (1995). Detection of insulin and insulin-like growth factors I and II in saliva and potential synthesis in the salivary glands of mice. Effects of type 1 diabetes mellitus. Biochemical Pharmacology 49, 1521–1531.
Keysar, S.B., Le, P.N., Miller, B., Jackson, B.C., Eagles, J.R., Nieto, C., Kim, J., Tang, B., Glogowska, M.J., Morton, J.J., et al. (2017). Regulation of head and neck squamous cancer stem cells by PI3K and SOX2. Journal of the National Cancer Institute 109.
Khandekar, M.J., Banks, A.S., Laznik-Bogoslavski, D., White, J.P., Choi, J.H., Kazak, L., Lo, J.C., Cohen, P., Wong, K.K., Kamenecka, T.M., et al. (2018). Noncanonical agonist PPARγ ligands modulate the response to DNA damage and sensitize cancer cells to cytotoxic chemotherapy. Proceedings of the National Academy of Sciences of the United States of America 115, 561–566.
Kim, J., and Orkin, S.H. (2011). Embryonic stem cell-specific signatures in cancer: insights into genomic regulatory networks and implications for medicine. Genome Medicine 3.
Kim, J.B., Wright, H.M., Wright, M., and Spiegelman, B.M. (1998). ADD1/SREBP1 activates PPARgamma through the production of endogenous ligand. Proceedings of the National Academy of Sciences 95, 4333–4337.
Kimes, P.K., Cabanski, C.R., Wilkerson, M.D., Zhao, N., Johnson, A.R., Perou, C.M., Makowski, L., Maher, C.A., Liu, Y., Marron, J.S., et al. (2014). SigFuge: Single gene clustering of RNA- seq reveals differential isoform usage among cancer samples. Nucleic Acids Research 42.
Knosp, W.M., Knox, S.M., and Hoffman, M.P. (2012). Salivary gland organogenesis. WIREs Dev Biol 1, 69–82.
Komiya, T., Park, Y., Modi, S., Coxon, A.B., Oh, H., and Kaye, F.J. (2006). Sustained expression of Mect1-Maml2 is essential for tumor cell growth in salivary gland cancers carrying the t(11;19) translocation. Oncogene 25, 6128–6132.
Kumar-Sinha, C., Kalyana-Sundaram, S., and Chinnaiyan, A.M. (2015). Landscape of gene fusions in epithelial cancers: Seq and ye shall find. Genome Medicine 7.
115 Kwak, M., Alston, N., and Ghazizadeh, S. (2016). Identification of Stem Cells in the Secretory Complex of Salivary Glands. Journal of Dental Research 95, 776–783.
Landman, G., and Farmer, E.R. (1990). Primary cutaneous mucoepidermoid carcinoma: report of a case. J Cutan Pathol 18, 56–59.
Latysheva, N.S., Oates, M.E., Maddox, L., Flock, T., Gough, J., Buljan, M., Weatheritt, R.J., and Babu, M.M. (2016). Molecular Principles of Gene Fusion Mediated Rewiring of Protein Interaction Networks in Cancer. Molecular Cell 63, 579–592.
Lee, T.I., and Young, R.A. (2013). Transcriptional regulation and its misregulation in disease. Cell 152, 1237–1251.
Lennerz, J.K.M., Perry, A., Mills, J.C., Huettner, P.C., and Pfeifer, J.D. (2009). Mucoepidermoid Carcinoma of the Cervix: Another Tumor With the t(11;19)-associated CRTC1-MAML2 Gene Fusion. Am J Surg Pathol 33, 835–843.
Leon, E., Ranganathan, R., and Savoldo, B. (2020). Adoptive T cell therapy: Boosting the immune system to fight cancer. Seminars in Immunology 49.
Lewandoski, M. (2002). Conditional Control of Gene Expression in the Mouse. Nature Reviews Genetics 2, 743–755.
Li, Y., Kovach, A., Suino-Powell, K., Martynowski, D., and Xu, H.E. (2008). Structural and biochemical basis for the binding selectivity of peroxisome proliferator-activated receptor γ to PGC-1α. Journal of Biological Chemistry 283, 19132–19139.
Lin, G.G., and Scott, J.G. (2012). Causes of oncogenic chromosomal mutations. 100, 130–134.
Lipner, M.B., Peng, X.L., Jin, C., Xu, Y., Gao, Y., East, M.P., Rashid, N.U., Moffitt, R.A., Herrera Loeza, S.G., Morrison, A.B., et al. (2020). Irreversible JNK1-JUN inhibition by JNK-IN-8 sensitizes pancreatic cancer to 5-FU/FOLFOX chemotherapy. JCI Insight 5.
Love, M.I., Huber, W., and Anders, S. (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology 15. van Luijk, P., Pringle, S., Deasy, J.O., Moiseenko, V. v, Faber, H., Hovan, A., Baanstra, M., van der Laan, H.P., Kierkels, R.G.J., van der Schaaf, A., et al. (2015). Sparing the region of the salivary gland containing stem cells preserves saliva production after radiotherapy for head and neck cancer. Science Translational Medicine 7.
Lujan, B., Hakim, S., Moyano, S., Nadal, A., Caballero, M., Diaz, A., Valera, A., Carrera, M., Cardesa, A., and Alos, L. (2010). Activation of the EGFR/ERK pathway in high-grade mucoepidermoid carcinomas of the salivary glands. British Journal of Cancer 103, 510– 516.
116 Luo, C., Lim, J.H., Lee, Y., Granter, S.R., Thomas, A., Vazquez, F., Widlund, H.R., and Puigserver, P. (2016). A PGC1α-mediated transcriptional axis suppresses melanoma metastasis. Nature 537, 422–426.
Luo, C., Balsa, E., Perry, E.A., Liang, J., Tavares, C.D., Vazquez, F., Widlund, H.R., and Puigserver, P. (2020). H3K27me3-mediated PGC1α gene silencing promotes melanoma invasion through WNT5A and YAP. Journal of Clinical Investigation 130, 853–862.
Maimets, M., Rocchi, C., Bron, R., Pringle, S., Kuipers, J., Giepmans, B.N.G., Vries, R.G.J., Clevers, H., de Haan, G., van Os, R., et al. (2016). Long-Term in Vitro Expansion of Salivary Gland Stem Cells Driven by Wnt Signals. Stem Cell Reports 6, 150–162.
Mandegar, M.A., Huebsch, N., Frolov, E.B., Shin, E., Truong, A., Olvera, M.P., Chan, A.H., Miyaoka, Y., Holmes, K., Spencer, C.I., et al. (2016). CRISPR Interference Efficiently Induces Specific and Reversible Gene Silencing in Human iPSCs. Cell Stem Cell 18, 541– 553.
Marciano, D.P., Kuruvilla, D.S., Boregowda, S. v., Asteian, A., Hughes, T.S., Garcia-Ordonez, R., Corzo, C.A., Khan, T.M., Novick, S.J., Park, H., et al. (2015). Pharmacological repression of PPARγ promotes osteogenesis. Nature Communications 6, 1–7.
Martínez-Redondo, V., Pettersson, A.T., and Ruas, J.L. (2015). The hitchhiker’s guide to PGC-1α isoform structure and biological functions. Diabetologia 58, 1969–1977.
Martínez-Redondo, V., Jannig, P.R., Correia, J.C., Ferreira, D.M.S., Cervenka, I., Lindvall, J.M., Sinha, I., Izadi, M., Pettersson-Klein, A.T., Agudelo, L.Z., et al. (2016). Peroxisome proliferator-activated receptor γ coactivator-1 α isoforms selectively regulate multiple splicing events on target genes. Journal of Biological Chemistry 291, 15169–15184.
Martins, C., Cavaco, B., Tonon, G., Kaye, F.J., Soares, J., and Fonseca, I. (2004). A study of MECT1-MAML2 in mucoepidermoid carcinoma and Warthin’s tumor of salivary glands. Journal of Molecular Diagnostics 6, 205–210.
Maruyama, E.O., Aure, M.H., Xie, X., Myal, Y., Gan, L., and Ovitt, C.E. (2016). Cell-specific Cre strains for genetic manipulation in salivary glands. PLoS ONE 11, 1–12.
Mastropasqua, F., Girolimetti, G., and Shoshan, M. (2018). PGC1α: Friend or foe in cancer? Genes 9, 1–19.
May, A.J., Cruz-Pacheco, N., Emmerson, E., Gaylord, E.A., Seidel, K., Nathan, S., Muench, M.O., Klein, O.D., and Knox, S.M. (2018). Diverse progenitor cells preserve salivary gland ductal architecture after radiation-induced damage. Development (Cambridge) 145.
McHugh, C.H., Roberts, D.B., El-Naggar, A.K., Hanna, E.Y., Garden, A.S., Kies, M.S., Weber, R.S., and Kupferman, M.E. (2012). Prognostic factors in mucoepidermoid carcinoma of the salivary glands. Cancer 118, 3928–3936.
117 Missiaglia, E., Williamson, D., Chisholm, J., Wirapati, P., Pierron, G., Petel, F., Concordet, J.P., Thway, K., Oberlin, O., Pritchard-Jones, K., et al. (2012). PAX3/FOXO1 fusion gene status is the key prognostic molecular marker in rhabdomyosarcoma and significantly improves current risk stratification. Journal of Clinical Oncology 30, 1670–1677.
Mitelman, F., Johansson, B., Mertens, F., Schyman, T., and Mandahl, N. (2019). Cancer chromosome breakpoints cluster in gene-rich genomic regions. Genes Chromosomes and Cancer 58, 149–154.
Mootha, V.K., Lindgren, C.M., Eriksson, K.-F., Subramanian, A., Sihag, S., Lehar, J., Puigserver, P., Carlsson, E., Ridderstråle, M., Laurila, E., et al. (2003). PGC-1α-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nature Genetics 34, 267–273.
Musicant, A.M., Parag-Sharma, K., Gong, W., Sengupta, M., Chatterjee, A., Henry, E.C., Tsai, Y.- H., Hayward, M.C., Sheth, S., Betancourt, R., et al. (2021). CRTC1-MAML2 establishes a PGC1α-IGF1 circuit that confers vulnerability to PPARγ inhibition. Cell Reports 34.
Nagy, A. (2000). Cre Recombinase: The Universal Reagent for Genome Tailoring. Genesis 26, 99–109.
Nance, M.A., Seethala, R.R., Wang, Y., Chiosea, S.I., Myers, E.N., Johnson, J.T., and Lai, S.Y. (2008). Treatment and survival outcomes based on histologic grading in patients with head and neck mucoepidermoid carcinoma. Cancer 113, 2082–2089.
Novo, F.J., and Vizmanos, J.L. (2006). Chromosome translocations in cancer: computational evidence for the random generation of double-strand breaks. Trends in Genetics 22, 193–196.
Obe, G., Pfeiffer, P., Savage, J.R.K., Johannes, C., Goedecke, W., Jeppesen, P., Natarajan, A.T., Martıneź -López, W., Folle, G.A., and Drets, M.E. (2002). Chromosomal aberrations: Formation, identification and distribution. Mutation Research 504, 17–36. de Oca Luna, R.M., Wagner, D.S., and Lozanot, G. (1995). Rescue of early embryonic lethality in mdm2-deficient mice by deletion of p53. Nature 378, 203–206.
Oltean, S., and Bates, D.O. (2014). Hallmarks of alternative splicing in cancer. Oncogene 33, 5311–5318.
O’Neill, I.D. (2009). t(11;19) translocation and CRTC1-MAML2 fusion oncogene in mucoepidermoid carcinoma. Oral Oncology 45, 2–9.
Parag-Sharma, K., Tasoulas, J., Musicant, A.M., Nascimento-Filho, C.H.V. do, Zhu, Z., Twomey, C., Liu, P., Castilho, R.M., and Amelio, A.L. (2021). Synergistic efficacy of combined EGFR and HDAC inhibitors overcomes tolerance to EGFR monotherapy in salivary mucoepidermoid carcinoma. Oral Oncology 115, 105166.
118 Patel, V.N., Rebustini, I.T., and Hoffman, M.P. (2006). Salivary gland branching morphogenesis. Differentiation 74, 349–364.
Patro, R., Duggal, G., Love, M.I., Irizarry, R.A., and Kingsford, C. (2017). Salmon provides fast and bias-aware quantification of transcript expression. Nature Methods 14, 417–419.
Perez-Losada, J., and Balmain, A. (2003). Stem-cell hierarchy in skin cancer. Nature Reviews Cancer 3, 434–443.
Pin, C.L., Bonvissuto, A.C., and Konieczny, S.F. (2000). Mist1 Expression Is a Common Link Among Serous Exocrine Cells Exhibiting Regulated Exocytosis. Anat Rec 259, 157–167.
Pin, C.L., Rukstalis, J.M., Johnson, C., and Konieczny, S.F. (2001). The bHLH transcription factor Mist1 is required to maintain exocrine pancreas cell organization and acinar cell identity. Journal of Cell Biology 155, 519–530.
Pires, F.R., de Almeida, O.P., de Araújo, V.C., and Kowalski, L.P. (2004). Prognostic Factors in Head and Neck Mucoepidermoid Carcinoma. Archives of Otolaryngology - Head and Neck Surgery 130, 174–180.
Pollak, M. (2008). Insulin and insulin-like growth factor signalling in neoplasia. Nature Reviews Cancer 8, 915–928.
Povirk, L.F. (2006). Biochemical mechanisms of chromosomal translocations resulting from DNA double-strand breaks. DNA Repair 5, 1199–1212.
Rabbitts, T.H. (1994). Chromosomal translocations in human cancer. Nature 372, 143–149.
Rabbitts, T.H. (2009). Commonality but Diversity in Cancer Gene Fusions. Cell 137, 391–395.
Riggi, N., Knoechel, B., Gillespie, S.M., Rheinbay, E., Boulay, G., Suvà, M.L., Rossetti, N.E., Boonseng, W.E., Oksuz, O., Cook, E.B., et al. (2014). EWS-FLI1Utilizes Divergent Chromatin Remodeling Mechanisms to Directly Activate or Repress Enhancer Elements in Ewing Sarcoma. Cancer Cell 26, 668–681.
Rocchi, C., Barazzuol, L., and Coppes, R.P. (2021). The evolving definition of salivary gland stem cells. Npj Regenerative Medicine 6, 4.
Rodriguez, C.I., Buchholz, F., Galloway, J., Sequerra, R., Kasper, J., Ayala, R., Stewart, A.F., and Dymecki, S.M. (2000). High-efficiency deleter mice show that FLPe is an alternative to Cre-loxP [correspondence]. Nature Genetics 25, 139–140.
Rosenfeld, M.G., Lunyak, V. v., and Glass, C.K. (2006). Sensors and signals: A coactivator/corepressor/epigenetic code for integrating signal-dependent programs of transcriptional response. Genes and Development 20, 1405–1428.
119 Rowley, J.D. (2001). Chromosome translocations: dangerous liaisons revisited. Nature Reviews Cancer 1, 245–250.
Roydhouse, C., Gannan, E., West, L., Brandt, C., and Spokes, R. (2019). A Rare Case of Primary Multifocal Mucoepidermoid Carcinoma in the Skin of the Breast Masquerading as a Breast Mass. Annals of Clinical Case Reports 4, 4–7.
Ruas, J.L., White, J.P., Rao, R.R., Kleiner, S., Brannan, K.T., Harrison, B.C., Greene, N.P., Wu, J., Estall, J.L., Irving, B.A., et al. (2012). A PGC-1α isoform induced by resistance training regulates skeletal muscle hypertrophy. Cell 151, 1319–1331.
Ryan, J., Mantle, T., McQuaid, S., and Costigan, D.C. (1992). Salivary insulin-like growth factor-I originates from local synthesis. Journal of Endocrinology 135, 85–90.
Saade, R.E., Bell, D., Garcia, J., Roberts, D., and Weber, R. (2016). Role of CRTC1/MAML2 translocation in the prognosis and clinical outcomes of mucoepidermoid carcinoma. JAMA Otolaryngology - Head and Neck Surgery 142, 234–240.
Sancho, P., Burgos-Ramos, E., Tavera, A., Bou Kheir, T., Jagust, P., Schoenhals, M., Barneda, D., Sellers, K., Campos-Olivas, R., Graña, O., et al. (2015). MYC/PGC-1α balance determines the metabolic phenotype and plasticity of pancreatic cancer stem cells. Cell Metabolism 22, 590–605.
Schaub, F.X., Reza, M.S., Flaveny, C.A., Li, W., Musicant, A.M., Hoxha, S., Guo, M., Cleveland, J.L., and Amelio, A.L. (2015). Fluorophore-NanoLuc BRET reporters enable sensitive In Vivo optical imaging and flow cytometry for monitoring tumorigenesis. Cancer Research 75, 5023–5033.
Schneider, C.A., Rasband, W.S., and Eliceiri, K.W. (2012). NIH Image to ImageJ: 25 years of Image Analysis. Nat Methods 9, 671–675.
Schwarz, S., Stiegler, C., Müller, M., Ettl, T., Brockhoff, G., Zenk, J., and Agaimy, A. (2011). Salivary gland mucoepidermoid carcinoma is a clinically, morphologically and genetically heterogeneous entity: A clinicopathological study of 40 cases with emphasis on grading, histological variants and presence of the t(11;19) translocation. Histopathology 58, 557– 570.
Seethala, R.R., and Chiosea, S.I. (2016). MAML2 Status in Mucoepidermoid Carcinoma Can No Longer Be Considered a Prognostic Marker [Letter to the Editor]. American Journal of Surgical Pathology 40, 1151–1153.
Seethala, R.R., Dacic, S., Cieply, K., Kelly, L.M., and Nikiforova, M.N. (2010). A reappraisal of the MECT1/MAML2 translocation in salivary mucoepidermoid carcinomas. American Journal of Surgical Pathology 34, 1106–1121.
120 Semple, R.K., Chatterjee, V.K.K., and O’Rahilly, S. (2006). PPARγ and human metabolic disease. Journal of Clinical Investigation 116, 581–589.
Shang, J., Mosure, S.A., Zheng, J., Brust, R., Bass, J., Nichols, A., Solt, L.A., Griffin, P.R., and Kojetin, D.J. (2020). A molecular switch regulating transcriptional repression and activation of PPARγ. Nature Communications 11, 1–14.
Sharpless, N.E., and DePinho, R.A. (2006). The mighty mouse: Genetically engineered mouse models in cancer drug development. Nature Reviews Drug Discovery 5, 741–754.
Sheth, S., Betancourt, R., Parker, J., Sharma, K., Musicant, A., Yarbrough, W., Hackman, T., and Amelio, A. (2020). Abstract PR04: RNA sequencing of salivary mucoepidermoid carcinoma reveals distinct molecular profiles based on histologic grade. Clinical Cancer Research 26, PR04.
Shi, G., Zhu, L., Sun, Y., Bettencourt, R., Damsz, B., Hruban, R.H., and Konieczny, S.F. (2009). Loss of the Acinar-Restricted Transcription Factor Mist1 Accelerates Kras-Induced Pancreatic Intraepithelial Neoplasia. Gastroenterology 136, 1368–1378.
Singer, D.S., Jacks, T., and Jaffee, E. (2016). A U.S. “Cancer Moonshot” to accelerate cancer research. Science 353, 1105–1106.
Son, E., Panwar, A., Mosher, C.H., and Lydiatt, D. (2018). Cancers of the Major Salivary Gland. Journal of Oncology Practice 14, 99–108.
Sonoda, J., Pei, L., and Evans, R.M. (2008). Nuclear receptors: Decoding metabolic disease. FEBS Letters 582, 2–9.
Sorrell, D.A., and Kolb, A.F. (2005). Targeted modification of mammalian genomes. Biotechnology Advances 23, 431–469.
Speight, P., and Barrett, A. (2002). Salivary gland tumours. Oral Diseases 8, 229–240.
Spike, B.T., Engle, D.D., Lin, J.C., Cheung, S.K., La, J., and Wahl, G.M. (2012). A mammary stem cell population identified and characterized in late embryogenesis reveals similarities to human breast cancer. Cell Stem Cell 10, 183–197.
Sporn, M.B. (1996). The war on cancer. The Lancet 347, 1377–1381.
Stewart, F.W., Foote, F.W., and Becker, W.F. (1945). Muco-Epidermoid Tumors of Salivary Glands. Annals of Surgery 122, 820–844.
Su, L., Morgan, P.R., Harrison, D.L., Waseem, A., and Lane, E.B. (1993). Expression of keratin mRNAS and proteins in normal salivary epithelia and pleomorphic adenomas. The Journal of Pathology 171, 173–181.
121 Subramanian, A., Tamayo, P., Mootha, V.K., Mukherjee, S., Ebert, B.L., Gillette, M.A., Paulovich, A., Pomeroy, S.L., Golub, T.R., Lander, E.S., et al. (2005). Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles.
Tasoulas, J., Rodon, L., Kaye, F.J., Montminy, M., and Amelio, A.L. (2019). Adaptive Transcriptional Responses by CRTC Coactivators in Cancer. Trends in Cancer 5, 111–127.
Tirado, Y., Williams, M.D., Hanna, E.Y., Kaye, F.J., Batsakis, J.G., and El-Naggar, A.K. (2007). CRTC1/MAML2 fusion transcript in high grade mucoepidermoid carcinomas of salivary and thyroid glands and Warthin’s tumors: Implications for histogenesis and biologic behavior. Genes Chromosomes and Cancer 46, 708–715.
Tonon, G., Modi, S., Wu, L., Kubo, A., Coxon, A.B., Komiya, T., O’Neil, K., Stover, K., El-Naggar, A., Griffin, J.D., et al. (2003). t(11;19)(q21;p13) translocation in mucoepidermoid carcinoma creates a novel fusion product that disrupts a Notch signaling pathway. Nature Genetics 33, 208–213.
Torrano, V., Valcarcel-Jimenez, L., Cortazar, A.R., Liu, X., Urosevic, J., Castillo-Martin, M., Fernández-Ruiz, S., Morciano, G., Caro-Maldonado, A., Guiu, M., et al. (2016). The metabolic co-regulator PGC1α suppresses prostate cancer metastasis. Nature Cell Biology 18, 645–656.
Troester, M.A., Herschkowitz, J.I., Oh, D.S., He, X., Hoadley, K.A., Barbier, C.S., and Perou, C.M. (2006). Gene expression patterns associated with p53 status in breast cancer. BMC Cancer 6.
Tuna, M., Amos, C.I., and Mills, G.B. (2019). Molecular mechanisms and pathobiology of oncogenic fusion transcripts in epithelial tumors. Oncotarget 10, 2095–2111.
Valcarcel-Jimenez, L., Macchia, A., Crosas-Molist, E., Schaub-Clerigué, A., Camacho, L., Martín- Martín, N., Cicogna, P., Viera-Bardón, C., Fernández-Ruiz, S., Rodriguez-Hernandez, I., et al. (2019). PGC1α suppresses prostate cancer cell invasion through ERRα transcriptional control. Cancer Research canres.1231.2019.
Varghese, J.J., Schmale, I.L., Wang, Y., Hansen, M.E., Newlands, S.D., Ovitt, C.E., and Benoit, D.S.W. (2018). Retroductal Nanoparticle Injection to the Murine Submandibular Gland. Journal of Visualized Experiments.
Vasioukhin, V., Degenstein, L., Wise, B., and Fuchs, E. (1999). The magical touch: Genome targeting in epidermal stem cells induced by tamoxifen application to mouse skin. Proceedings of the National Academy of Sciences of the United States of America 96, 8551–8556.
Vazquez, F., Lim, J.H., Chim, H., Bhalla, K., Girnun, G., Pierce, K., Clish, C.B., Granter, S.R., Widlund, H.R., Spiegelman, B.M., et al. (2013). PGC1α Expression Defines a Subset of
122 Human Melanoma Tumors with Increased Mitochondrial Capacity and Resistance to Oxidative Stress. Cancer Cell 23, 287–301.
Visvader, J.E. (2011). Cells of origin in cancer. Nature 469, 314–322.
Wachtel, M., and Schäfer, B.W. (2018). PAX3-FOXO1: Zooming in on an “undruggable” target. Seminars in Cancer Biology 50, 115–123.
Wagner, V.P., Martins, M.D., Martins, M.A.T., Almeida, L.O., Warner, K.A., Nör, J.E., Squarize, C.H., and Castilho, R.M. (2018). Targeting histone deacetylase and NFκB signaling as a novel therapy for Mucoepidermoid Carcinomas. Scientific Reports 8, 1–11.
Wakabayashi, N., Itoh, K., Wakabayashi, J., Motohashi, H., Noda, S., Takahashi, S., Imakado, S., Kotsuji, T., Otsuka, F., Roop, D.R., et al. (2003). Keap1-null mutation leads to postnatal lethality due to constitutive Nrf2 activation. Nature Genetics 35, 238–245.
Wallace, M., and Metallo, C.M. (2016). PGC1α drives a metabolic block on prostate cancer progression. Nature Cell Biology 18, 589–590.
Walrath, J.C., Hawes, J.J., van Dyke, T., and Reilly, K.M. (2010). Chapter 4: Genetically Engineered Mouse Models in Cancer Research. In Advances in Cancer Research, pp. 113–164.
Wang, B.D., and Lee, N.H. (2018). Aberrant RNA splicing in cancer and drug resistance. Cancers 10.
Wang, K., McDermott, J.D., Schrock, A.B., Elvin, J.A., Gay, L., Karam, S.D., Raben, D., Somerset, H., Ali, S.M., Ross, J.S., et al. (2017). Comprehensive genomic profiling of salivary mucoepidermoid carcinomas reveals frequent BAP1, PIK3CA, and other actionable genomic alterations. Annals of Oncology 28, 748–753.
Warner, K.A., Adams, A., Bernardi, L., Nor, C., Finkel, K.A., Zhang, Z., McLean, S.A., Helman, J., Wolf, G.T., Divi, V., et al. (2013). Characterization of tumorigenic cell lines from the recurrence and lymph node metastasis of a human salivary mucoepidermoid carcinoma. Oral Oncology 49, 1059–1066.
Watanabe, J., Kai, K., Tanikawa, K., Hiraki, M., Mizukami, N., Aishima, S., Nakano, T., and Yamamoto, H. (2019). Primary mucoepidermoid carcinoma of the liver with CRTC1- MAML2 fusion: A case report. Diagnostic Pathology 14, 4–8.
Weng, P.L., Aure, M.H., Maruyama, T., and Ovitt, C.E. (2018). Limited Regeneration of Adult Salivary Glands after Severe Injury Involves Cellular Plasticity. Cell Reports 24, 1464- 1470.e3.
Werner, H., and Katz, J. (2004). The emerging role of the insulin-like growth factors in oral biology. Journal of Dental Research 83, 832–836.
123 Whelton, H. (2012). Physiology of Salivary Glands. British Dental Association 1–36.
Wu, L., Liu, J., Gao, P., Nakamura, M., Cao, Y., Shen, H., and Griffin, J.D. (2005). Transforming activity of MECT1-MAML2 fusion oncoprotein is mediated by constitutive CREB activation. EMBO Journal 24, 2391–2402.
Xi, J. jie, Jiang, W., Lu, S. hua, Zhang, C. yan, Fan, H., and Wang, Q. (2012). Primary pulmonary mucoepidermoid carcinoma: An analysis of 21 cases. World Journal of Surgical Oncology 10, 1–7.
Xing, F., Luan, Y., Cai, J., Wu, S., Mai, J., Gu, J., Zhang, H., Li, K., Lin, Y., Xiao, X., et al. (2017). The Anti-Warburg Effect Elicited by the cAMP-PGC1α Pathway Drives Differentiation of Glioblastoma Cells into Astrocytes. Cell Reports 18, 468–481.
Young, L., Sung, J., and Masters, J.R. (2010). Detection of mycoplasma in cell cultures. Nature Protocols 5, 929–934.
Yu, H., and Rohan, T. (2000). Role of the Insulin-Like Growth Factor Family in Cancer Development and Progression. Journal of the National Cancer Institute 92, 1472–1489.
Zajicek, G., Schwartz-Arad, D., Arber, N., and Michaeli, Y. (1989). The streaming of the submandibular gland II: parenchyma and stroma advance at the same velocity. Cell Tissue Kinet 22, 343–348.
Zhang, J., and Manley, J.L. (2013). Misregulation of pre-mRNA alternative splicing in cancer. Cancer Discovery 3, 1228–1237.
Zhang, J.-H., Chung, T.D.Y., and Oldenburg, K.R. (1999). A Simple Statistical Parameter for Use in Evaluation and Validation of High Throughput Screening Assays. Journal of Biomolecular Screening 4, 67–73.
Zou, Y., Watters, A., Cheng, N., Perry, C.E., Xu, K., Alicea, G.M., Parris, J.L.D., Baraban, E., Ray, P., Nayak, A., et al. (2019). Polyunsaturated fatty acids from astrocytes activate PPARγ signaling in cancer cells to promote brain metastasis. Cancer Discovery 9, 1720–1735.
124