Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 May 5.
Published in final edited form as: Stem Cell Res. 2017 May 5;21:160–166. doi: 10.1016/j.scr.2017.05.002

MYB FUSIONS AND CD MARKERS AS TOOLS FOR AUTHENTICATION AND PURIFICATION OF CANCER STEM CELLS FROM SALIVARY ADENOID CYSTIC CARCINOMA

Alex Panaccione a, Yi Zhang a, Molly Ryan b, Christopher A Moskaluk c, Karen S Anderson b, Wendell G Yarbrough a,d,e,f,g, Sergey V Ivanov a,g
PMCID: PMC5518784  NIHMSID: NIHMS876947  PMID: 28500913

Abstract

Cancer stem cells (CSC) are considered the major cause of aggressive tumor behavior, recurrence, metastases, and resistance to radiation, making them an attractive therapeutic target. However, isolation of CSC from tumor tissue and their characterization are challenging due to uncertainty about their molecular markers and conditions for their propagation. Adenoid cystic carcinoma (ACC), which arises predominantly in the salivary glands, is a slow-growing but relentless tumor that frequently invades nerves and metastasizes. New effective treatment approaches for ACC have not emerged over the last 40 years. Previously, based on a highly conserved SOX10 gene signature that we identified in the majority of ACC tumors, we suggested the existence in ACC of SOX10+ cells with neural stem properties and corroborated this hypothesis via isolation from ACC tissue a novel population of CSC, termed ACC-CSC. These cells co-expressed SOX10 and other ACC-intrinsic neural crest stem cell markers with CD133, a CSC cell surface marker, and activated NOTCH1 signaling suggesting that ACC is driven by a previously uncharacterized population of SOX10+/CD133+ cells with neural stem cell properties. Here, we authenticated ACC identity of our primary cultures by demonstrating that most of them harbor MYB-NFIB fusions, which are found in 86% of ACC. We demonstrated using CyTOF, a novel mass cytometry technology, that these cells express high β-catenin and STAT3 levels and are marked by CD24 and CD44. Finally, to streamline development of ACC cell lines, we developed RT-PCR tests for distinguishing mouse and human cells and used immunomagnetic cell sorting to eliminate mouse cells from long-term cell cultures. Overall, this study describes a new population of CSC that activates signaling pathways associated with poor prognosis, validates their ACC identity, and optimizes approaches that can be used for purification of ACC-CSC and generation of cell lines.

1. INTRODUCTION

Adenoid cystic carcinoma (ACC) is a deadly cancer: with a prevalence rate of 1224 cases, 918 patients die from ACC in the U.S. every year (http://www.accoi.org/faq/acc-statistics/). ACC is treated by surgery with or without radiation, but only 40% of patients survive 15 years owing to intrinsic radiation resistance of ACC cells and their propensity to metastasize, relapse, and spread along nerves (1,2). The recurrence rate is high (53%) owing mostly to neural invasion, radio-resistance, and hematologic metastases (3).

Aggressive ACC behavior suggests that it may be driven by cancer stem cells (CSC). CSC possess properties of normal stem cells and are widely associated with invasion, recurrence, metastases, and resistance to cytotoxic therapies (4–6). Their identification in ACC will advance understanding of molecular etiology and cell of origin, improving diagnostics, predicting disease outcome, and developing effective therapies.

However, characterization of CSC is controversial when it is based solely on CD markers, whose expression is not stem cell-selective (7). In addition, CSC isolated from cell cultures are often not representative of tumor tissue and therefore lack clinical value (8–10). With the goal to identify clinically relevant CSC in ACC, we performed gene expression profiling of surgically resected tumor specimens to identify stem cell signaling and associated selective markers. This analysis demonstrated that most of ACC specimens selectively express SOX10, a marker of neural crest cells and oligonedraglial progenitors (11,12), providing a clue to how CSC can be identified and isolated from ACC tissue. Indeed, in line with a special role of SOX10 in this cancer, we identified in the majority of ACC the expression of a highly conserved SOX10 gene signature that contained a cluster of neural stem cell drivers and markers, such as NOTCH1, MAP2, GPM6B, and FABP7, as well as genes/proteins involved in WNT and NOTCH signaling (13,14). These findings suggested that SOX10 expression delineates activation of a neural stem cell program in ACC and marks a previously uncrecognized population of cells with neural stem cell properties.

The creation and maintenance of subcutaneous patient-derived xenografts (PDX) from fresh or cryopreserved ACC tissue (15) provided a renewable source of ACC cells for validation of our CSC hypothesis. As we previously demonstrated, these PDX models reproduced ACC morphology and maintained the SOX10 gene signature (13,15). To isolate SOX10+ CSC from grafted ACC tissue, we used a ROCK inhibitor-based cell culture protocol (16) and profiled our cell cultures for expression of SOX10 and other ACC-intrinsic stem cells markers as well as for expression of CD133, a CSC cell surface marker. In line with our hypothesis, we isolated from patient-derived xenografts (PDX) and clinical ACC specimens a new population of SOX10+/CD133+ cancer stem cells, which we named ACC-CSC. These cells co-expressed SOX10 with activated NOTCH1 and met basic CSC criteria by producing spheroids in vitro and generating aggressive ACC-like tumors in nude mice. In addition, ACC-CSC activated stem cell signaling with involvement of SOX10 and NOTCH1 (17). In this study, we authenticate cultured ACC-CSC using ACC-intrinsic MYB-NFIB fusions, investigate their signaling pathways, and test new tools, markers, and assays for their purification.

2. MATERIALS AND METHODS

2.1. Cell culture

Detailed description of ROCK inhibitor-based cell culture is available from (16). Briefly, primary cultures were grown in F+Y Medium (3:1 (v/v) Dulbecco’s modified Eagle’s medium: Ham’s F12 nutrient mixture with 7.5% fetal bovine serum (all from Invitrogen), 0.4 μg/mL hydrocortisone (Sigma-Aldrich), 5 μg/mL insulin (Sigma-Aldrich), 8.6 ng/mL cholera toxin (Sigma-Aldrich), 10 ng/mL epidermal growth factor (Invitrogen), and 10 μM/L ROCK-inhibitor (Y-27632, SelleckChem) mixed in a 4:1 ratio with conditioned media from 3T3-J2 cells. To create conditioned media, irradiated (30 Gy) 3T3-J2 cells were incubated in a T-150 flask supplemented with 30 mL of DMEM media for 4 days. Media was filter-sterilized and then mixed in a 1:4 ratio with F+Y media.

2.2. Nude mouse patient-derived xenografts (PDX)

Generation and analysis of all PDX models but ACCX33 have been described in our previous studies (14,15). Briefly, fresh tissue cut into small (~1 mm3) pieces were implanted subcutaneously in the shoulder area of 4–6 weeks old female nude mice (NCr nu/nu) purchased from NCI-Frederick. Animals were operated and maintained in accordance with the Institutional Animal Care and Use Committee guidelines. ACCX33 was created from a clinical tumor specimen, ACC33, described in (17). Nude mice were purchased from Frederick National Laboratory of Cancer Research (Frederick, MD).

2.3

Human breast carcinoma MX-1 cells were purchased from Frederick National Laboratory of Cancer Research and grown with ROCK inhibitor as in (16). 293FT cells were purchased from Thermo Fisher Scientific (Witham, MA).

2.4

Orthotopic tumor models were generated via injection of 104 ACC-CSC cells mixed with Matrigel (Corning, Tewksbury, MA) at a 1:1 ratio into the salivary glands of nude mice.

2.5. Primers and RT-PCR reagents

For end-point RT-PCR (28 cycles), we used SuperScript III One Step RT-PCR kit (ThermoFisher, Waltham, MA) in combination with Bio-Rad Imaging system. We designed and used the following primers: mTusc2F: 5′-CCTCAGTGTTAACTGGTATTGGC GGA and mTusc2R: 5′-CAACTTGAGGGCAACTGGTTGAAATCA-3′ (for mouse RNA detection); hPPIAF: 5′-TGAGGGTAGGAGTCAAGATCAGC-3′ and hPPIAR: 5′-AGAGGCTCT ATATGCTACAAGCAGTACC-3′ (for human RNA detection); B-2-1: 5′-CGGGTGCTCTTACCC ACTGA-3′ (for single-primer mouse RNA detection), and 5′-ALU: 5′-ACATGGTGAAACCCC GTCTCTAC (for single-primer human RNA detection). To detect ACC-CSC, we used the following human-specific primers: hCD133F: 5′-GAAACTGCTTGAGCATCAGGATACTC-3′; hCD133R: 5′-CAGATCCAACATGAACTCCTGAAGC-3′; hCD24F: 5′-GTGGTGCGATCTCAGATCAGTGTA-3′, hCD24R: 5′-CTGTGGCCATATTAGATTACTGGAAC-3′; and hCD44F: 5′-TCTCCCACTTGGCA AGTCCTTTG-3′, hCD44R: 5′-TCCGTTGACAATGGCCAAGG-3′; hJAG1F: 5′-CCGACACGG TCTCGGATCA-3′, and hJAG1R: 5′-TTCACGGTCTCAATGGTGAACCAACAA-3′. To confirm MYB-NFIB fusions in PDX models cultured ACC cells, we used the following primers: ACCX11FUSF: 5′-GTAGAAGATCTGCAGGATGTGATCAAACAGG-3′; ACCX11FUSR: 5′-CCTCCTAGCCTT CGTTGGTG-3′; ACCX14FUSF: 5′-GAATATTCTTACAAGCTCCG-3′, ACCX14FUSR: 5′-CCT CCTAGCCTTCGTTGGTG-3′; and ACCX33FUSR: 5′-GAAGCTTTTGGGTTGAGACAATGGG-3′. To confirm human MYB expression in cultured ACC cells, we used the following primers: hMYBF: 5′-AAAGCGCCTCGCCAGCAAG-3′ and hMYBR: 5′-CCTCCTAGCCTTCGTTGGTG-3′. To validate activation of β-catenin and STAT3, we used the following primers: TCF7L2F: 5′-TGATTTCCTTCAAAGACGAGGG-3′ and TCF7L2R: 5′-CTTAAAGAGCCCTCCATCTTGC-3′; MYCF: 5′-AGCGACTCTGAGGAGGAACAA-3′ and MYCR: 5′-CTCAGCCAAGGTTGTGAGGT-3′.

2.6. Magnetic-activated cell sorting (MACS)

Trypsinized cells were sorted using CD24 and CD133 Micro-bead kits and depletion of mouse cells from ACC cultures was done with Mouse Cell Depletion kit (#130-104-694), all available from Miltenyi Biotec (Cambridge, MA). Cultured ACC cells were resuspended in MACS buffer (phosphate-buffered saline (PBS), pH 7.2, 0.5% bovine serum albumin (BSA), and 2 mM EDTA) and incubated for 15 minutes with magnetic bead- (CD133 & mouse cell depletion) or biotin-conjugated (CD24) antibodies. The CD24 kit subsequently washes and incubates the cells with a magnetic bead-conjugated anti-biotin antibody. Cells are then passed through a column encapsulated within a magnetic field using gravity flow, capturing bead-conjugated cells within the column. The column is washed three times with MACS buffer before removing the column from the magnet and cells are gently flushed from the column with MACS buffer.

2.7

Immunofluorescent staining was performed as described in (17).

2.8

CyTOF analysis was performed on CyTOF2 mass-spectrometer (Yale School of Medicine) with CD133-PE antibodies from Miltenyi Biotec (San Diego, CA) and heavy metal-tagged antibodies 165Ho-PE, 173Yb-STAT3, and 147Sm-β-catenin (Fluidigm, San-Francisco, CA). Cells were fixed, permeabilized, and stained for surface and intracellular targets using Fluidigm protocols.

2.9

Immunoblot analysis was performed with β-catenin, STAT3, p-STAT3, GAPDH, and anti-rabbit IgG, HRP-linked antibodies available from Cell Signaling (Beverly, MA).

3. RESULTS

3.1. Authentication of PDX models and ACC cell cultures via MYB-NFIB fusions

We recently generated 6 primary ACC cultures from patient-derived xenograft (PDX) models as well as from a surgically obtained specimen using a ROCK inhibitor-based approach to cell culture. Short tandem repeat (STR) analysis that we performed at that time confirmed origin of ACC cultures from corresponding tumors, and expression in these cells of ACC-intrinsic markers SOX10, FABP7, etc. strongly suggested their ACC identity (16). Here, we asked if ACC cell cultures harbor MYB-NFIB fusions, which are ACC-intrinsic genetic aberrations found in 86% of ACC patients and considered to be primary ACC-driving events (18,19). To identify these fusions and sequence them, we first performed targeted RNA sequencing of PDX models using the FusionPlex Solid Tumor panel and kit (ArcherDX, Boulder, CO). This technology uses anchored multiplex PCR to detect previously described and novel gene fusions, SNPs, and InDels in selected genes. Adapters with molecular barcodes are used for quantitative multiplex data analysis, read deduplication, and accurate mutation calling. This analysis revealed MYB fusions in PDX models ACCX11, 14, 33, and 5M1 and in one additional model, ACCX9. In 4 of 5 models, we identified canonical MYB-NFIB fusions between exons 12 or 14 of MYB (NM_005375) and exon 9 of NFIB (NM_005596) (FIG. 1). Similar MYB-NFIB fusions have been previously described by other research teams (20,21). In ACCX33, we found a previously uncharacterized fusion that joined exon 14 of MYB with a long non-coding RNA gene from chromosome 6q24.3, RP11-545I5.3 (GenBank: HG499759, pos. 135-605) (FIG. 1).

FIG. 1. Canonical ACC-intrinsic MYB-NFIB fusions and a novel MYB/lncRNA fusion identified in PDX models (ACC) and corresponding cell cultures (Acc).

FIG. 1

Open in a new tab

*-alternative MYB fusions with other NFIB exons were detected within the same tumor (data not shown).

We next designed PCR primers and confirmed MYB fusion breakpoints in PDX models and cell cultures using RT-PCR and Sanger sequencing. Overall, these analyses confirmed ACC identity of our PDX models and corresponding cell cultures validating them as authentic tools for studies aimed at the role of MYB fusions in ACC-CSC propagation.

3.2. Activation of β-catenin and STAT3 in ACC-CSC

We previously demonstrated that SOX10, NOTCH1, and their common target FABP7 play essential roles in propagation of CD133+ ACC-CSC and that NOTCH signaling is involved in their resistance to radiation (17). In addition to NOTCH, other signaling pathways, such as WNT/β-catenin and STAT3, are known to take part in CSC survival and aggressive behavior (22). To find out if WNT/β-catenin and STAT3 are engaged in ACC-CSC, we used CyTOF, a novel time-of-flight mass spectrometry-based technology for multiplex cytometry and single cell analysis to find out if these pathways are activated in ACC-CSC, which are marked by CD133 expression. CyTOF analysis of cultured Accx11 cells co-stained with CD133-PE antibodies, secondary antibody against PE, as well as β-catenin and human STAT3 metal-tagged antibodies demonstrated that the majority of CD133-positive Accx11 cells, or ACC-CSC, express high levels of β-catenin, and that most of β-catenin-positive cells are also enriched with STAT3. Specificity of CD133-PE antibodies used in CyTOF was validated on 293-FT cells that don’t express CD133 (FIG. 2A). Activation of STAT3 and β-catenin in ACC-CSC was confirmed by Western blot that showed higher levels of these proteins in CD133-expressing Accx11 and MX-1 than in CD133-negative 293FT cells (FIG. 2B). We also performed real-time PCR that demonstrated increased levels of a major β-catenin effector, TCF7L2 (23), and STAT3 target MYC (24) supporting our CyTOF data (FIG. 2C). These findings were in line with the recently reported cooperation between WNT/β-catenin and STAT3 signaling in CSC maitenance (25).

FIG. 2. CyTOF analysis of Accx11 for CD133, β-catenin, and STAT3.

FIG. 2

Open in a new tab

A, Top panels: Analysis of cells co-stained with CD133-PE, PE, and β-catenin antibodies show that ~ 70% of cells are positive for both CD133 and β-catenin. The majority of β-catenin-positive cells are also STAT3-positive. Bottom panels: CD133-negative cells 293FT produce no signal from CD133-PE. B, Immunoblot analysis is consistent with CyTOF data showing high β-catenin levels and STAT3 expression in Accx11 as compared to MX-1 and 293FT cell lines. C, Real-time RT-PCR data confirm transcriptional up-regulation of a β-catenin target/partner TCF7L2 and STAT3 partner MYC in CD133+ cells as compared to the CD133-negative fraction.

3.3. ACC-CSC purification via spheroids

To further purify ACC-CSC for their comprehensive characterization and cell line derivation, we used the most robustly growing ACC cell culture, Accx11, which spontaneously formed spheroids and developed into long-term cell culture (P>30). Based on previously published reports on spheroids isolated from other cancer cell cultures (26,27), we hypothesized that they are composed entirely of CD133+ cancer stem cells and can be used for ACC-CSC purification. In line with this expectation, we demonstrated via immunofluorescent staining on bulk Accx11 cells that spheroids express CD133 and NOTCH1, two ACC-intrinsic stemness markers (FIG. 3A) and confirmed that floating spheroids collected from culture media express higher levels of CD133 than bulk cells (FIG. 3B,C). In order to obtain pure ACC-CSC cultures, media containing unadhered spheroids was collected from plates containing bulk cells. Spheroids were allowed to settle at the bottom of 15 mL conical tubes for 15 minutes. After aspirating the media, the cells were washed with PBS and this was repeated twice more. The remaining spheroids were transfered into fresh plates with media. After several days, spheroids adhered to the surface and produced neural stem-like cells that developed processes reminiscent of axons and growth cones (FIG. 3D). Further passaging produced transparent semi-adherent cells with variable morphology, axon-like protrusions, and tendency to pile up and form spheroids (Suppl. FIG. 1). To find out if these cells are tumorigenic, we injected 104 cells into the salivary glands of nude mice. Confirming their tumor-initiating properties, spheroid cells produced ACC-like tumors in all four mice used in this experiment (FIG. 3E).

FIG. 3. ACC-CSC purified via spheroids produce aggressive ACC-like tumors when injected into the salivary glands of nude mice.

FIG. 3

Open in a new tab

A, Spheroids spontaneously produced by Accx11 cells express NOTCH1 (green) and CD133 (red). B, Spheroid purified via media collection, decantation, and re-plating. C, RT-PCR confirms expression of CD133/PROM1 in spheroids (sph) collected from bulk Accx11 cell culture (bulk). PPIA was used as a loading control. D, Purified spheroids adhere and produce neural precursor-like patches of cells that develop axon-like (AL) and growth cone-like (GCL) processes at the edge of cell colony. E, spheroid cells injected into salivary gland tissue of nude mice produce aggressively growing tumors (T) that develop salivary gland-like structures (GLS) and invade into normal salivary glands (NSG), nerves, and muscle.

3.4. Magnetic-activated CD133- and CD24-based cell sorting (MACS)- and spheroid-based purification of ACC-CSC

In our previous study, we demonstrated that ACC-CSC can be purified via Mitenyi magnetic beads bound to CD133 antibodies and that purified CD133+ cells as well as spheroids selectively express SOX10 with activated NOTCH1 and initiate tumor growth in nude mice (17). To more comprehensively characterize ACC-CSC CD markers, we asked if CD24 and CD44 are also expressed in these cells. In our end-point RT-PCR experiments with primers specific for human CD24 and CD44 transcripts, we demonstrated that cultured Accx11 cells expressed both cell surface markers. In this experiment, Accx11 spheroids and MACS-purified CD133+ cells showed higher CD24 and CD44 levels than the CD133-negative cell fraction suggesting that ACC-CSC selectively express CD24 and CD44 in addition to CD133 (FIG. 4A). To confirm CD24 and CD44 expression in ACC tumors, we analyzed our expression array data generated on surgically resected ACC specimens (13). In this dataset, we observed high levels of CD24 and CD44 in the majority of ACC tumors (FIG. 4B). We then used MACS to produce CD133- and CD24-sorted Accx11 cells and analyzed these fractions for CD24/CD133 expression. Remarkably, CD24+ Accx11 cells were mostly CD133-positive and vice versa suggesting that a large proportion of ACC-CSC have the double-positive CD24+/CD133+ phenotype (FIG. 5A). Based on all these data, we concluded that CD24, CD44, and CD133 are all clinically informative markers for tracking CSC in ACC tissue, PDX models, and primary cell cultures that we created. These cell surface markers may help to further purify ACC-CSC via MACS or FACS as well as investigate their heterogeneity, signaling status, and relation to other tumor cells using CyTOF analysis (28).

FIG. 4. Expression of CD24 and CD44 in cultured ACC cells and tumors.

FIG. 4

Open in a new tab

A, CD24 and CD44 levels in MACS-sorted Accx11 cells and two different spheroid preparations as assessed by end-point RT-PCR with human-specific primers. Human MX-1 cells and mouse 3T3-J2 cells were used as controls. For loading control, see PPIA data in Suppl. FIG. 3. B, Expression of CD24 and CD44 in PDX models (ACCX and MAD04) and surgically resected ACC tumors (ACC) as estimated via Affymetrix gene expression analysis performed in (13).

FIG. 5. Comparison between CD24 and CD44-based MACS, spheroid purification, monoclonal ACC-CSC purification, and MACS-based depletion of mouse cells.

FIG. 5

Open in a new tab

A, Analysis of MACS purified Accx11 cell fractions; B, Comparison between spheroids (SPH) and two ACC-CSC clones produced from Accx11. C, Analysis of Accx11 and Accx19 cell fractions before and after mouse cell elimination with Miltenyi Mouse Cell Depletion kit. Lack of signal from mouse RNA in the wash and flow-through fractions ensures compete purification of human cells. Endpoint RT-PCR analysis with human- and mouse-specific primers for CD marker-encoding transcripts and mTusc2 and hPPIA mRNA (28 cycles). MX-1 and 3T3-J2 cells were used as controls.

3.5. Elimination of mouse cell contamination from cultured ACC cells

We noticed that at late passages (P>10) mouse cells introduced in ACC cell cultures from feeder layer or PDX models may overgrow human ACC cells (our unpublished observations). To detect mouse cell contamination of ACC cultures, we designed a simple RT-PCR test based on primate- and mouse-specific Alu and B2 repeats, respectively. We also used mouse- and human-specific primers for ubiquitously expressed genes TUSC2 and PPIA, respectively. All these primers were validated on primary Accx11 culture as well as on mouse fibroblasts and human breast cancer cells MX-1. This analysis identified mouse RNA contamination in the original (bulk) Accx11 cell culture but not in the negative control, MX-1 (Suppl. FIG. 2).

The same test was then used to compare efficiencies of different approaches to mouse cell elimination. While MACS CD sorting significantly diminished the signals from mouse RNA, it failed to completely eliminate mouse cells most likely due to non-specific binding to magnetic beads or cell clumping (FIG. 5A). Seeking more efficient approaches, we analyzed cells after purification by: 1) collection of floating spheroids produced by bulk Accx11 cells, 2) development of clonal populations of CD133+ cells, and 3) depletion of mouse cells with a MACS Miltenyi kit. Following these purification techniques, mouse cells were below the limits of PCR detection (FIGS. 5B and C). The third approach, however, is the most optimal as it maintains tumor heterogeneity and therefore may be better for experimental use.

4. DISCUSSION

Development of targeted therapies for ACC has been stalled by lack of reliable cell lines/cultures and approaches for CSC isolation. Unfortunately, most ACC cell lines generated 20–30 years ago and shared between laboratories turned out to be grossly contaminated or misidentified (29,30). These cell lines are now blacklisted (http://iclac.org/databases/cross-contaminations/). However, of more than 100 manuscripts that relied on these cell lines only two have been retracted (31), and new studies with same cell lines have been recently published (32–38). While the most plausible explanation for the ACC cell line identity crisis is lack of proper authentication (39), ACC cells are also notoriously difficult to culture as they rapidly undergo senescence or die in regular cell culture media (our unpublished observations).

In this study, we continue characterization of a novel population of CSC, ACC-CSC, that we recently purified from ACC using a ROCK inhibitor and mouse cell feeder layer-based approach (17). To our knowledge, this report is the first study to detect MYB fusions in ACC cultures, despite the fact that MYB-NFIB fusions have been described in 86% of ACC patients (18). In a fraction of ACC tumors, MYB fusions with other genes, such as TGFBR3 and RAD51B, as well as MYBL1 fusions with NFIB have been also reported (19,40). Here, we identified a novel MYB fusion with RP11-54515.3, an lncRNA gene from chromosome 6q24.3. Intriguingly, this lncRNA is a cis natural antisense transcript (cis-NAT) for SHPRH, an E3 ubiquitin ligase. While exact functions of cis-NAT are not known, they may inhibit expression of corresponding sense transcripts via transcription interference or post-transcriptionally (41). Interestingly, SHPRH functions as a tumor suppressor protein that induces degradation of β-catenin, a CSC driver, and has been recently recognized as a target of axitinib, a small molecule inhibitor of WNT/β-catenin signaling (42). It may be hypothesized therefore that the MYB-RP11-54515.3 fusion may inhibit SHPRH translation and stimulate ACC-CSC via WNT/β-catenin signaling. This hypothesis will be validated on Accx33 cells that harbor this fusion. Overall, generation of ACC cells with oncogenically activated MYB provided long-awaited tools for studies focused on the MYB role in ACC.

Our next goal was to perform a more detailed characterization of ACC-CSC to identify therapeutically amenable pathways of their propagation. To this end, we used CyTOF, an innovative flow cytometry technology, which is based on time-of-flight mass-spectrometry (43). CyTOF analysis demonstrated that CD133+ AC-CSC are highly enriched with β-catenin and STAT3, two major targets actively explored in novel cancer-targeting therapies (42,44). We and others previously implicated both STAT3 and Wnt/β-catenin signaling in ACC (13,45–47), and this study, for the first time, links both pathways to CSC suggesting their cooperation and providing targets for ACC-CSC eradication.

An important step towards further ACC-CSC characterization is identification of CD markers which, in combination with CD133, may be used for FACS sorting and multiplexed CyTOF analysis of ACC-heterogeneity and their signaling states. Such studies may significantly advance understanding of ACC-CSC roles in tumor tissue and their relation to tumor growth, neural invasion, metastases, resistance to radiation, and recurrence. Association of ACC-CSC with the CD24 cell surface marker in this study provides a new insight into ACC cell of origin. Indeed, our hypothesis that ACC and basal-like breast cancers have similar CSC driven by SOX10 (14) has been recently supported by a study that implicated SOX10 in stimulation of CD24hi fetal mammary stem cells (48). Thus, CD24 appears to be a biologically and clinically relevant cell surface marker that can be used in combination with CD133 and CD44 in CyTOF studies and for more efficient ACC-CSC purification. As it was recently demonstrated, expression of CD24, CD44, and CD133 may be linked with activation of distinct pro-survival signaling pathways and have different effects on radiation resistance and patient survival (49). It remains to see if similar associations take place in ACC and if differential expression of CD markers may predict prognosis or resistance to therapy.

Another important goal of this study was to purify ACC-CSC from mixed mouse/human cell cultures and grow them as long-term cell cultures. In this regard, we found that Accx11 spheroids are made entirely of human CD133+ ACC-CSC, which made them a straightforward tool for ACC-CSC purification (FIG. 3. and Suppl. FIG. 3). Similarly, generation of monoclonal cell populations completely purified ACC-CSC from mouse cells (FIG. 5B). We then used MACS with antibodies to CD24 and CD133 as an alternative approach to ACC-CSC purification. Even though this approach turned out to be not as efficient to rid cultures of mouse cell contamination, these markers are human-specific and can be used with cell cultures that don’t produce spheroids as well as in analytical FACS/CyTOF and preparative FACS studies on unpurified primary ACC cultures at early passages. Finally, we concluded that the Miltenyi kit for mouse cell elimination is an effective means for ACC-CSC purification since this approach produced cell populations without detectable mouse cells (FIG. 5C).

5. CONCLUSION

Overall, we report in this study on the assays, tools, and markers for authentication, purification, and research of a novel population of CSC that we recently identified in ACC. Marked by expression of SOX10, similar CSC populations appear to drive breast basal-like carcinoma, melanoma, and other cancers that appear to arise from neural crest (14,50). We confirmed that ACC cultures maintained MYB fusions found in xenografts or primary tumors and identified a novel MYB fusion to a non-coding RNA. We demonstrated that ACC-CSC that we characterized earlier as SOX10+/CD133+ (17,51) also express CD24 and CD44 as well as signaling molecules commonly found in CSCs, STAT3 and β-catenin. To develop a new platform for drug screening with the goal to develop effective and specific therapies for ACC, we optimized ACC-CSC purification from mixed mouse/human cell cultures and created a novel ACC-CSC-initiated orthotopic model for pre-clinical studies.

Supplementary Material

1. SUPPL. FIG. 1.

ACC-CSC purified via spheroid collection.

2. SUPPL. FIG. 2.

Validation of human- and mouse-specific RT-PCR primers on bulk Accx11 cells, human breast cancer cell line MX-1, and mouse 3T3-J2 cells.

3. SUPPL. FIG. 3.

RT-PCR analysis of CD133-sorted and spheroid-purified Accx11 cells for mouse contaminations.

HIGHLIGHTS.

  • ACC cancer stem cell cultures are authenticated using ACC-intrinsic MYB fusions.

  • A novel MYB fusion, with a long non-coding RNA from 6q23.4, is identified.

  • ACC-CSC are marked by expression of CD24 and CD44.

  • ACC-CSC purification is optimized to establish cell lines for drug screening.

  • CyTOF is used to explore STAT3 and β-catenin as novel targets for ACC-CSC eradication.

Acknowledgments

This study was supported by funds from the Adenoid Cystic Carcinoma Research Foundation and by grants 5R21DE023228 (WGY) and 5R21DE022641 (SVI) from the National Institute of Dental and Craniofacial Research. This work was also supported in part by funds from the Department of Surgery, the Division of Otolaryngology, Yale School of Medicine, by an endowment to the Barry Baker Laboratory for Head and Neck Oncology, by Laura and Isaac Perlmutter Cancer Center Support Grant NIH/NCI P30CA016087, and the National Institutes of Health S10 Grants NIH/ORIP S10OD01058 and S10OD018338.

Footnotes

Conflict of interests

Authors declare no conflict of interests.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Singh FM, Mak SY, Bonington SC. Patterns of spread of head and neck adenoid cystic carcinoma. Clin Radiol. 2015;70(6):644–53. doi: 10.1016/j.crad.2015.01.013. [DOI] [PubMed] [Google Scholar]
  • 2.Oplatek A, Ozer E, Agrawal A, Bapna S, Schuller DE. Patterns of recurrence and survival of head and neck adenoid cystic carcinoma after definitive resection. Laryngoscope. 2010;120(1):65–70. doi: 10.1002/lary.20684. [DOI] [PubMed] [Google Scholar]
  • 3.Shingaki S, Saito R, Kawasaki T, Nakajima T. Adenoid cystic carcinoma of the major and minor salivary glands. A clinicopathological study of 17 cases. J Maxillofac Surg. 1986;14(1):53–6. doi: 10.1016/s0301-0503(86)80259-4. [DOI] [PubMed] [Google Scholar]
  • 4.Beck B, Blanpain C. Unravelling cancer stem cell potential. Nature reviews Cancer. 2013;13(10):727–38. doi: 10.1038/nrc3597. [DOI] [PubMed] [Google Scholar]
  • 5.Tysnes BB. Tumor-initiating and -propagating cells: cells that we would like to identify and control. Neoplasia. 2010;12(7):506–15. doi: 10.1593/neo.10290. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Pattabiraman DR, Weinberg RA. Tackling the cancer stem cells - what challenges do they pose? Nature reviews Drug discovery. 2014;13(7):497–512. doi: 10.1038/nrd4253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Jaggupilli A, Elkord E. Significance of CD44 and CD24 as cancer stem cell markers: an enduring ambiguity. Clin Dev Immunol. 2012;2012:708036. doi: 10.1155/2012/708036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Liu Y, Nenutil R, Appleyard MV, Murray K, Boylan M, Thompson AM, et al. Lack of correlation of stem cell markers in breast cancer stem cells. Br J Cancer. 2014;110(8):2063–71. doi: 10.1038/bjc.2014.105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Madjd Z, Erfani E, Gheytanchi E, Moradi-Lakeh M, Shariftabrizi A, Asadi-Lari M. Expression of CD133 cancer stem cell marker in melanoma: a systematic review and meta-analysis. Int J Biol Markers. 2016;31(2):e118–25. doi: 10.5301/jbm.5000209. [DOI] [PubMed] [Google Scholar]
  • 10.Leon G, MacDonagh L, Finn SP, Cuffe S, Barr MP. Cancer stem cells in drug resistant lung cancer: Targeting cell surface markers and signaling pathways. Pharmacol Ther. 2016;158:71–90. doi: 10.1016/j.pharmthera.2015.12.001. [DOI] [PubMed] [Google Scholar]
  • 11.Nonaka D, Chiriboga L, Rubin BP. Sox10: a pan-schwannian and melanocytic marker. Am J Surg Pathol. 2008;32(9):1291–8. doi: 10.1097/PAS.0b013e3181658c14. [DOI] [PubMed] [Google Scholar]
  • 12.Gorris R, Fischer J, Erwes KL, Kesavan J, Peterson DA, Alexander M, et al. Pluripotent stem cell-derived radial glia-like cells as stable intermediate for efficient generation of human oligodendrocytes. Glia. 2015;63(12):2152–67. doi: 10.1002/glia.22882. [DOI] [PubMed] [Google Scholar]
  • 13.Ivanov SV, Panaccione A, Brown B, Guo Y, Moskaluk CA, Wick MJ, et al. TrkC signaling is activated in adenoid cystic carcinoma and requires NT-3 to stimulate invasive behavior. Oncogene. 2013;32(32):3698–710. doi: 10.1038/onc.2012.377. [DOI] [PubMed] [Google Scholar]
  • 14.Ivanov SV, Panaccione A, Nonaka D, Prasad ML, Boyd KL, Brown B, et al. Diagnostic SOX10 gene signatures in salivary adenoid cystic and breast basal-like carcinomas. Br J Cancer. 2013;109(2):444–51. doi: 10.1038/bjc.2013.326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Moskaluk CA, Baras AS, Mancuso SA, Fan H, Davidson RJ, Dirks DC, et al. Development and characterization of xenograft model systems for adenoid cystic carcinoma. Lab Invest. 2011;91(10):1480–90. doi: 10.1038/labinvest.2011.105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Liu X, Ory V, Chapman S, Yuan H, Albanese C, Kallakury B, et al. ROCK inhibitor and feeder cells induce the conditional reprogramming of epithelial cells. Am J Pathol. 2012;180(2):599–607. doi: 10.1016/j.ajpath.2011.10.036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Panaccione A, Chang MT, Carbone BE, Guo Y, Moskaluk CA, Virk RK, et al. NOTCH1 and SOX10 are Essential for Proliferation and Radiation Resistance of Cancer Stem-Like Cells in Adenoid Cystic Carcinoma. Clin Cancer Res. 2016;22(8):2083–95. doi: 10.1158/1078-0432.CCR-15-2208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Brill LB, 2nd, Kanner WA, Fehr A, Andren Y, Moskaluk CA, Loning T, et al. Analysis of MYB expression and MYB-NFIB gene fusions in adenoid cystic carcinoma and other salivary neoplasms. Mod Pathol. 2011;24(9):1169–76. doi: 10.1038/modpathol.2011.86. [DOI] [PubMed] [Google Scholar]
  • 19.Drier Y, Cotton MJ, Williamson KE, Gillespie SM, Ryan RJ, Kluk MJ, et al. An oncogenic MYB feedback loop drives alternate cell fates in adenoid cystic carcinoma. Nat Genet. 2016;48(3):265–72. doi: 10.1038/ng.3502. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Persson M, Andren Y, Mark J, Horlings HM, Persson F, Stenman G. Recurrent fusion of MYB and NFIB transcription factor genes in carcinomas of the breast and head and neck. Proc Natl Acad Sci U S A. 2009;106(44):18740–4. doi: 10.1073/pnas.0909114106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Mitani Y, Li J, Rao PH, Zhao YJ, Bell D, Lippman SM, et al. Comprehensive analysis of the MYB-NFIB gene fusion in salivary adenoid cystic carcinoma: Incidence, variability, and clinicopathologic significance. Clin Cancer Res. 2010;16(19):4722–31. doi: 10.1158/1078-0432.CCR-10-0463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Matsui WH. Cancer stem cell signaling pathways. Medicine (Baltimore) 2016;95(1 Suppl 1):S8–S19. doi: 10.1097/MD.0000000000004765. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Cadigan KM, Waterman ML. TCF/LEFs and Wnt signaling in the nucleus. Cold Spring Harb Perspect Biol. 2012;4(11) doi: 10.1101/cshperspect.a007906. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Banerjee K, Resat H. Constitutive activation of STAT3 in breast cancer cells: A review. Int J Cancer. 2016;138(11):2570–8. doi: 10.1002/ijc.29923. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Chen MW, Yang ST, Chien MH, Hua KT, Wu CJ, Hsiao SM, et al. The STAT3-miRNA-92-Wnt signaling pathway regulates spheroid formation and malignant progression in ovarian cancer. Cancer Res. 2017 doi: 10.1158/0008-5472.CAN-16-1115. [DOI] [PubMed] [Google Scholar]
  • 26.Lee SH, Hong JH, Park HK, Park JS, Kim BK, Lee JY, et al. Colorectal cancer-derived tumor spheroids retain the characteristics of original tumors. Cancer Lett. 2015;367(1):34–42. doi: 10.1016/j.canlet.2015.06.024. [DOI] [PubMed] [Google Scholar]
  • 27.Christensen K, Aaberg-Jessen C, Andersen C, Goplen D, Bjerkvig R, Kristensen BW. Immunohistochemical expression of stem cell, endothelial cell, and chemosensitivity markers in primary glioma spheroids cultured in serum-containing and serum-free medium. Neurosurgery. 2010;66(5):933–47. doi: 10.1227/01.NEU.0000368393.45935.46. [DOI] [PubMed] [Google Scholar]
  • 28.Han L, Qiu P, Zeng Z, Jorgensen JL, Mak DH, Burks JK, et al. Single-cell mass cytometry reveals intracellular survival/proliferative signaling in FLT3-ITD-mutated AML stem/progenitor cells. Cytometry A. 2015;87(4):346–56. doi: 10.1002/cyto.a.22628. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Phuchareon J, Ohta Y, Woo JM, Eisele DW, Tetsu O. Genetic profiling reveals cross-contamination and misidentification of 6 adenoid cystic carcinoma cell lines: ACC2, ACC3, ACCM, ACCNS, ACCS and CAC2. PLoS One. 2009;4(6):e6040. doi: 10.1371/journal.pone.0006040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Retraction: Epithelial mesenchymal transition is required for acquisition of anoikis resistance and metastatic potential in adenoid cystic carcinoma. PLoS One. 2014;9(9):e108117. doi: 10.1371/journal.pone.0108117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Retraction notice to “Nef from SIVmac239 decreases proliferation and migration of adenoid-cystic carcinoma cells and inhibits angiogenesis” [OO 47 (2011) 847–854] Oral oncology. 2012;48(1):95. doi: 10.1016/j.oraloncology.2011.11.011. [DOI] [PubMed] [Google Scholar]
  • 32.Sumida T, Ishikawa A, Kamata YU, Nakano H, Yamada T, Mori Y. Estrogen Enhances Malignant Phenotypes in Human Salivary Adenoid Cystic Carcinoma Cells. Anticancer Res. 2016;36(6):2793–8. [PubMed] [Google Scholar]
  • 33.Maruyama S, Cheng J, Yamazaki M, Zhou XJ, Zhang ZY, He RG, et al. Metastasis-associated genes in oral squamous cell carcinoma and salivary adenoid cystic carcinoma: a differential DNA chip analysis between metastatic and nonmetastatic cell systems. Cancer Genet Cytogenet. 2010;196(1):14–22. doi: 10.1016/j.cancergencyto.2009.08.002. [DOI] [PubMed] [Google Scholar]
  • 34.Wu H, Huang S, Chen Z, Liu W, Zhou X, Zhang D. Hypoxia-induced autophagy contributes to the invasion of salivary adenoid cystic carcinoma through the HIF-1alpha/BNIP3 signaling pathway. Mol Med Rep. 2015;12(5):6467–74. doi: 10.3892/mmr.2015.4255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Fang J, Bao YY, Zhou SH, Fan J. Apigenin inhibits the proliferation of adenoid cystic carcinoma via suppression of glucose transporter-1. Mol Med Rep. 2015;12(5):6461–6. doi: 10.3892/mmr.2015.4233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Sun L, Liu B, Lin Z, Yao Y, Chen Y, Li Y, et al. MiR-320a acts as a prognostic factor and Inhibits metastasis of salivary adenoid cystic carcinoma by targeting ITGB3. Mol Cancer. 2015;14:96. doi: 10.1186/s12943-015-0344-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Sumida T, Murase R, Onishi-Ishikawa A, McAllister SD, Hamakawa H, Desprez PY. Targeting Id1 reduces proliferation and invasion in aggressive human salivary gland cancer cells. BMC Cancer. 2013;13:141. doi: 10.1186/1471-2407-13-141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Wang SS, Gao XL, Liu X, Gao SY, Fan YL, Jiang YP, et al. CD133+ cancer stem-like cells promote migration and invasion of salivary adenoid cystic carcinoma by inducing vasculogenic mimicry formation. Oncotarget. 2016;7(20):29051–62. doi: 10.18632/oncotarget.8665. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Yu M, Selvaraj SK, Liang-Chu MM, Aghajani S, Busse M, Yuan J, et al. A resource for cell line authentication, annotation and quality control. Nature. 2015;520(7547):307–11. doi: 10.1038/nature14397. [DOI] [PubMed] [Google Scholar]
  • 40.Mitani Y, Liu B, Rao PH, Borra VJ, Zafereo M, Weber RS, et al. Novel MYBL1 Gene Rearrangements with Recurrent MYBL1-NFIB Fusions in Salivary Adenoid Cystic Carcinomas Lacking t(6;9) Translocations. Clin Cancer Res. 2016;22(3):725–33. doi: 10.1158/1078-0432.CCR-15-2867-T. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Rosikiewicz W, Makalowska I. Biological functions of natural antisense transcripts. Acta Biochim Pol. 2016;63(4):665–73. doi: 10.18388/abp.2016_1350. [DOI] [PubMed] [Google Scholar]
  • 42.Qu Y, Gharbi N, Yuan X, Olsen JR, Blicher P, Dalhus B, et al. Axitinib blocks Wnt/beta-catenin signaling and directs asymmetric cell division in cancer. Proc Natl Acad Sci U S A. 2016;113(33):9339–44. doi: 10.1073/pnas.1604520113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Giesen C, Wang HA, Schapiro D, Zivanovic N, Jacobs A, Hattendorf B, et al. Highly multiplexed imaging of tumor tissues with subcellular resolution by mass cytometry. Nat Methods. 2014;11(4):417–22. doi: 10.1038/nmeth.2869. [DOI] [PubMed] [Google Scholar]
  • 44.Xiang M, Kim H, Ho VT, Walker SR, Bar-Natan M, Anahtar M, et al. Gene expression-based discovery of atovaquone as a STAT3 inhibitor and anti-cancer agent. Blood. 2016 doi: 10.1182/blood-2015-07-660506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.de Araujo VC, Furuse C, Cury PR, Altemani A, de Araujo NS. STAT3 expression in salivary gland tumours. Oral Oncol. 2008;44(5):439–45. doi: 10.1016/j.oraloncology.2007.05.011. [DOI] [PubMed] [Google Scholar]
  • 46.Daa T, Kashima K, Kaku N, Suzuki M, Yokoyama S. Mutations in components of the Wnt signaling pathway in adenoid cystic carcinoma. Mod Pathol. 2004;17(12):1475–82. doi: 10.1038/modpathol.3800209. [DOI] [PubMed] [Google Scholar]
  • 47.Frierson HF, Jr, El-Naggar AK, Welsh JB, Sapinoso LM, Su AI, Cheng J, et al. Large scale molecular analysis identifies genes with altered expression in salivary adenoid cystic carcinoma. Am J Pathol. 2002;161(4):1315–23. doi: 10.1016/S0002-9440(10)64408-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Dravis C, Spike BT, Harrell JC, Johns C, Trejo CL, Southard-Smith EM, et al. Sox10 Regulates Stem/Progenitor and Mesenchymal Cell States in Mammary Epithelial Cells. Cell Rep. 2015;12(12):2035–48. doi: 10.1016/j.celrep.2015.08.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Sahlberg SH, Spiegelberg D, Glimelius B, Stenerlow B, Nestor M. Evaluation of cancer stem cell markers CD133, CD44, CD24: association with AKT isoforms and radiation resistance in colon cancer cells. PLoS One. 2014;9(4):e94621. doi: 10.1371/journal.pone.0094621. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Panaccione A, Guo Y, Yarbrough WG, Ivanov SV. Expression Profiling of Clinical Specimens Supports the Existence of Neural Progenitor-Like Stem Cells in Basal Breast Cancers. Clin Breast Cancer. 2017 doi: 10.1016/j.clbc.2017.01.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Yarbrough WG, Pannaccione A, Chang MT, Ivanov SV. Clinical and Molecular Insights Into Adenoid Cystic Carcinoma: Neural Crest-Like Stemness as a Target. Laryngoscope Investigative Otolaryngology. 2016 doi: 10.1002/lio2.22. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1. SUPPL. FIG. 1.

ACC-CSC purified via spheroid collection.

NIHMS876947-supplement-1.jpg (269.9KB, jpg)
2. SUPPL. FIG. 2.

Validation of human- and mouse-specific RT-PCR primers on bulk Accx11 cells, human breast cancer cell line MX-1, and mouse 3T3-J2 cells.

NIHMS876947-supplement-2.jpg (55KB, jpg)
3. SUPPL. FIG. 3.

RT-PCR analysis of CD133-sorted and spheroid-purified Accx11 cells for mouse contaminations.

NIHMS876947-supplement-3.jpg (183.9KB, jpg)

RESOURCES