Mandibular undifferentiated pleomorphic sarcoma: Molecular analysis of a primary cell population

Abstract Background Undifferentiated pleomorphic sarcomas are one of the most common subtypes of soft tissue sarcomas. These are aggressive mesenchymal tumors and are devoid of the major known biomarkers except vimentin. Our objective was to establish and characterize a primary cell population from a mandibular UPS specimen. Methods The tumor was surgically removed from the right mandible of a 24‐year‐old male with IRB approved signed consent. Tumor was dissected, cultured ex vivo, and a cell population, MUPS‐1, were isolated from outgrowths. Gene and protein expression profiles of both the primary tumor and the derived there from cells were obtained by quantitative RT‐PCR and immunohistochemistry and included markers of epithelial, endothelial, and mesenchymal differentiation. To better define potential biomarkers, MUPS‐1 cells were additionally characterized by RNA sequencing analysis. Results Pathological analysis of primary tumor tissue revealed a sarcoma demonstrating multiple pathways of differentiation simultaneously with myxoid, fibrous, and osseous tissue. The isolated cells had a spindle cell‐like morphology, were maintained in culture for greater than 20 passages, and formed colonies in soft agar indicating tumorigenicity. The cells, similar to the primary tumor, were strongly positive for vimentin and moderately expressed alkaline phosphatase. RNA‐seq analysis revealed the tumor over‐expressed several genes compared to normal tissue, including components of the Notch signaling pathway, NOTCH3 and JAG1. Conclusions We have successfully established an undifferentiated pleomorphic sarcoma cell population, which will provide a valuable resource for studying fundamental processes and potentially serving as a platform for exploring therapeutic strategies for sarcomas.


| INTRODUCTION
Soft tissue sarcomas (STS) encompass a diverse group of tumors arising from mesenchymal tissues such as cartilage, bone, fat, myxoid tissue, blood vessels, and muscle (Dodd, 2016;Matushansky et al., 2009;Tudor-Green, Gomez, & Brennan, 2017). STS are separated into 80-100 subtypes based on histological and molecular phenoytpes (Jo & Fletcher, 2014;Salawu et al., 2016;Tudor-Green et al., 2017). Undifferentiated pleomorphic sarcomas (UPS) are the most commonly diagnosed soft tissue sarcoma in adults; however, this should not be used as a diagnosis of exclusion. Many histologic mimics began as seemingly better differentiated tumors of fat (liposarcoma), muscle (leiomyosarcoma), or other mesenchymal cell types such as bone and cartilage which over time give rise (de-differentiate) into more aggressive, but morphologically undifferentiated tumors (Matushansky et al., 2009;Widemann & Italiano, 2018). Surgical resection is the primary treatment of UPS with some patients benefitting from the addition of chemotherapy. Even with surgical removal, the overall five-year survival is 54.7%, local recurrence rates are 13-42%, and many patients develop metastatic disease (Fletcher, 2014;Roland et al., 2016). Adjuvant chemotherapy or radiation therapy may be recommended for later stage tumors.
Histologically, UPS is composed of highly cellular pleomorphic spindle cells with markedly atypical nuclear features and with a high mitotic index. UPS commonly occurs in the extremities and retroperitoneum of adult Caucasian males and are rarer in women and other racial and ethnic populations (Jo & Fletcher, 2014;Roland et al., 2016). They are uncommon in children and young adults; and the peak incidence is in the sixth and seventh decades of life. UPS tumors, such as the one described here, were previously occasionally defined as malignant mesenchymomas when these pluripotential tumors simultaneously demonstrated two or more malignant mesenchymal elements. Over time this entity has disappeared from use and is now incorporated under the umbrella of UPS.
UPS of the jaw (gnathic) are rare, expansible lesions, occurring the maxilla or mandible with few cases reported. In each case, the tumor was detected as a bone-infiltrating mass. In this study, we describe a UPS extracted from the right mandible of a 24-year-old male. Pathological analysis of primary tumor tissue revealed a sarcoma demonstrating mixed morphologic patterns of the tissue with myxoid, fibrous, and osseous elements. This tissue was used to establish primary cell populations, which was characterized based on cell-type specific markers and RNA sequencing.

| Tumor specimen and establishment of cell populations
Minced tumor pieces were collected following surgical resection of a right mandibular tumor following obtainment of informed consented under an institutionally approved IRB protocol. Part of the primary tumor was fixed in formalin and paraffin-embedded. Remaining pieces were dissected, minced and placed in culture. Primary cell populations (MUPS-1) were established as outgrowths in α-DMEM/10%FBS/antibiotics.

| Determination of cell growth rate
Cells were plated on a 96-well plate and cell viability was measured using an MTS assay (Cell Titer96, Promega, Madison, WI) and absorbance at 490 nm (BioTek, Winooski, VT).

| Soft agar colony formation assay
Soft agar colony formation was measured by seeding primary tumor cells, MUPS-1, (2,500 cells) and positive control human breast cancer cells, MDA-MB-231 (1,500 cells), in medium containing 10% FBS and 0.3% agarose (UltraPure low melting point agarose, Invitrogen, Carlsbad, CA). Cells were plated over a layer of solidified medium containing FBS and 2% agarose in six-well plates. Cultures were maintained in a humidified, 37 C incubator and media was changed every 3 days for 4 weeks before colonies were photographed and counted.

| Quantitative real-time PCR (qRT-PCR)
Total RNA was isolated from cells using the Qiagen RNeasy Mini Kit (Valencia, CA). Formalin-fixed paraffin-embedded (FFPE) tumor sections (10 μm thick) were deparaffinized with xylene and ethanol (Steg et al., 2006). Tissue pellets were washed with ethanol and dried at 55 C. RNA was isolated using the Roche High Pure RNA paraffin kit (Roche Diagnostics, Indianapolis, IN). All RNA was converted to cDNA using the Bio-Rad iScript cDNA Synthesis Kit (Herculus, CA). qRT-PCR reactions were performed using the RT 2 SYBRGreen/Rox qPCR master mix (SABiosciences, Frederick, MD) and relative gene transcriptional levels detected using the ABI Prism 7,500 Sequence  2.5 | Next-generation sequencing on Illumina platform RNA samples were submitted to the Genomics Core Laboratory in the Heflin Center for Genomic Sciences at our institution for sample preparation and sequencing. mRNA-sequencing was performed on the Illumina HiSeq2000. The quality of the total RNA was assessed using the Agilent 2,100 Bioanalyzer followed by two rounds of poly A+ selection and conversion to cDNA. TruSeq library generation kits were utilized as per the manufacturer's instructions (Illumina, San Diego, CA).
The cDNA libraries were quantitated using qPCR in a Roche Lig-htCycler 480 with the Kapa Biosystems kit for library quantitation (Kapa Biosystems, Woburn, MA) prior to cluster generation. Clusters were generated to yield approximately 725-825 K clusters/mm 2 .
Cluster density and quality were determined during the run after the first base addition parameters were assessed. Paired end 2X50bp sequencing runs were performed to align the cDNA sequences to the reference genome. 2.6 | RNA-sequencing data analysis TopHat version 2.0.9 (parameters: -r 150; −library-type fr-unstranded; −G; −transcriptome-index) was used to align the raw RNA-Seq fastq reads to the UCSC hg19 reference genome using the short read aligner Bowtie2 (1, 2, 3). TopHat also analyzes the mapping results to identify splice junctions between exons. Cufflinks version 2.2.0 (parameters: −g; −L; −b; −u) used the aligned reads from TopHat to assemble transcripts, estimate their abundances and test for differential expression and regulation (3, 4). In brief, correcting for sequence bias and using the upper-quartile normalization options were turned on to better improve the transcript assembly and abundance estimates. Cuffmerge (parameters: −o; −g; −s), which is part of Cufflinks merged the assembled transcripts to a reference annotation and is capable of tracking Cufflinks transcripts across multiple experiments.
Citrate antigen retrieval was performed for 20 min at 97 C. Samples were incubated in hydrogen peroxide block (Thermo Scientific, Fremont, CA) for 15 min and Ultra V block for 5 min. For specific staining, samples were incubated with primary antibodies for 30 min and HRP polymer-conjugated antibody for 15 min. Diaminobenzidine tetrachloride was used as the chromagen (Biocare Medical, Concord, CA) and hematoxylin for counterstaining. For immunocytochemistry, cells were grown in 4-well chamber slides until 70% confluent then fixed with 4% formaldehyde. Staining was conducted as previously described above for tissue sections.  appearance, a cell doubling time of 93.14 hr, organized into clusters when at confluency, and could be maintained in culture over 12 days  Figure 1. The cells also expressed ALP, but less than 1% expressed CK14 and no protein expression of CD34 or OPN was detectable by IHC using commercial antibodies specific for these proteins (Figure 3).
To better characterize the tumor cells and identify possible therapeutic targets, RNA sequencing analysis was conducted on RNA isolated from tumor cells and compared to dental pulp cells, which are of mesenchymal dental origin, similar to the MUPS-1 cells. Thirty genes were upregulated in the MUPS-1 cells compared to dental pulp cells, 11 of which were not expressed or expressed below the level of detection in p cells (Table 1). These genes were also among some of the most expressed in the MUPS-1 cells. Fourteen genes were chosen based on previously reported relationships to human cancers and all 14 were confirmed to be upregulated in the MUPS-1 cells by qRT-PCR (Table 2). JPH2 and WISP2 proteins were both confirmed to be expressed in primary MUPS-1 tissue and tumor cells (Figure 4). JPH2 may serve as a marker for UPS, as it was previously reported to be overexpressed in leiomyosarcomas compared to an endometrial stromal sarcoma (Davidson et al., 2013). Of particular interest, two genes  transport gene) message compared to healthy tissue from the same patient. One study used UPS cells isolated from sporadic and radiationinduced UPS via explant culture from primary tumors and passaged xenografts (May et al., 2017). UPS cell lines exhibited high levels of p-Akt, the active form, which could be reduced with a dual PI3K/ mTOR inhibitor. However, the PI3K/mTOR inhibitor increased phosphorylated insulin-like growth factor receptor 1 (IGF1R), which is a known resistance mechanism to this inhibitor and was abrogated, in turn, with the addition of an anti-IGF1R kinase inhibitor.
In the current study, we establish an UPS primary cell population of VIM, as expected, and ALP, which has been shown to be expressed in mesenchymal stem cells (Gerlach et al., 2012).
Past studies to characterize molecular biomarkers of UPS demonstrated that UPS has a low burden of somatic mutations with 16% carrying RB1 and/or p53 mutations (Cancer Genome Atlas Research Network. Electronic address edsc, Cancer Genome Atlas Research N, 2017). One theory is that UPS arises from the transformation of mesenchymal stem cells (MSCs), which would account for the lack of differentiation seen in these tumors (Matushansky et al., 2009).
Genetic analysis demonstrated that UPS cells showed more similarities with MSCs than cells from other sarcoma subtypes. Other studies have shown Ras/MAPK activation in 80% of UPS cases, and expression of phospho-STAT3 associated with better prognosis (Bekki et al., 2017;Serrano et al., 2016). Another study showed that UPS and leiomyosarcomas had high expression of programmed cell death protein (PD-1) and programmed death-ligand1 (PD-L1), and thus may benefit from treatment with immune checkpoint inhibitors (Pollack et al., 2017).
To characterize our cell model and identify possible biomarkers we preformed RNA-sequencing analysis. Genes that were highly expressed in our UPS included WISP2, a regulator of Wnt signaling, which has been showed to play a role in the invasiveness of breast and colon cancer cells, and in epithelial-to-mesenchymal transition in pancreatic cancer (Davies, Watkins, Mansel, & Jiang, 2007;Dhar et al., 2007;Grunberg, Hammarstedt, Hedjazifar, & Smith, 2014). Our study is limited by the enrollment and inclusion of one patient; however, we have extensively characterized the cells, including RNA-sequencing analysis. A limitation of our analysis is the use of dental pulp cells as our control sample, which would share a dental mesenchymal origin, but the pulp cells would be differentiated. Our analysis identified novel potential biomarkers for UPS and STS, and with further investigation could be useful in pathologic studies. In addition, analysis linked a known tumor-associated pathway with STS.
The Notch pathway has been widely studied and may provide possible targeted therapies.
An advantage of our study is there are currently few commercially available sarcoma cell lines for preclinical studies, which are necessary for advancing therapies for UPS and STS. Miserocchi et al. (2017) highlight the advantages of ex vivo culture strategies such as the one