UM-HACC-2A: MYB-NFIB fusion-positive human adenoid cystic carcinoma cell line

Abstract

Objectives:

Limited availability of validated human adenoid cystic carcinoma (ACC) cell lines has hindered the mechanistic understanding of the pathobiology of this malignancy and the development of effective therapies. The purpose of this work was to generate and characterize a human ACC cell line.

Material and Methods:

Immediately after surgery, a tumor fragment from a minor salivary gland from the tongue of a female Caucasian was minced, dissociated, and a single cell suspension was plated in fibronectin-coated flasks. A culture medium containing bovine brain extract and rhEGF was optimized for these cells. Whole exome sequencing was used to evaluate the presence of MYB-NFIB translocation.

Results:

The University of Michigan-Human Adenoid Cystic Carcinoma (UM-HACC)-2A cells showed continuous growth in monolayers for at least 180 in vitro passages while maintaining epithelial morphology. Short-tandem repeat (STR) profiling confirmed a 100% match to patient DNA. Whole exome sequencing revealed the presence of the MYB-NFIB fusion in UM-HACC-2A cells, which was confirmed by PCR analysis. Western blots revealed high expression of epithelial markers (e.g. E-cadherin, EGFR, pan-cytokeratin) and proteins associated with ACC (e.g. c-Myb, p63). Developmental therapeutic studies showed that UM-HACC-2A cells were resistant to cisplatin (IC₅₀=44.7 µM) while more responsive to paclitaxel (IC₅₀=0.0006 µM). In a pilot study, we observed that UM-HACC-2A cells survived orthotopic transplantation into the submandibular gland. Notably, one of the mice injected with UM-HACC-2A cells exhibited lung metastasis after 6 months.

Conclusion:

UM-HACC-2A is a MYB-NFIB fusion-positive ACC cell line that is suitable for mechanistic and developmental therapeutics studies.

Introduction

Adenoid cystic carcinoma (ACC) is a rare, slow growing, malignant tumor of the salivary glands. ACC presents with 3 distinct histological subtypes (i.e. tubular, cribriform, solid), which represent increasing tumor aggressiveness and worse prognosis (solid type has worst prognosis). It is characterized by frequent perineural invasion and occasional metastatic spread to distant organs, which results in low long-term survival rates for patients diagnosed with advanced disease [1–5]. Surgery followed by radiation is the standard of care for patients with advanced disease [6]. Due to relentless (albeit slow) growth, the long-term survival rate (15-years) of patients with ACC is around 35–50% [1,7]. In general, ACC tumors do not respond well to therapies that rely on fast cell turnover for drug uptake and cytotoxic induction [3,6]. Limited availability of cell lines has hindered studies attempting to understand mechanisms underlying the pathobiology of ACC and slowed down the development of safe/effective therapies.

A translocation resulting in the fusion of MYB and NFIB in ACC was first identified by the Stenman laboratory [8]. This t(6;9)(q22–23;p23–24) translocation was originally found in ACC of the breast or head and neck region resulting in a fusion that usually involves exon 14 of MYB linked to the last coding exon(s) of NFIB [8]. The MYB-NFIB fusion is found in about 60–80% of salivary ACC patients [8–15]. More recently, the Moskaluk laboratory reported the presence of the fusion MYB-NFIB fusion in 83% of ACC patient derived xenografts (PDX) models [16].

Progress has been made in identifying novel therapeutic targets (e.g. MDM2) using low passage primary ACC cells and PDX models [17,18]. However, such studies are challenging due to the limited number of passages that primary cells typically grow in vitro. Ideally mechanistic and developmental therapeutic studies should be performed with established cell lines. However, the availability of human ACC cell lines is rather limited. Six ACC cell lines were found to be cross-contaminated with HeLa (ACC2, ACC3, ACCM), T24 bladder cancer cells (ACCS), established from mouse (ACCNS), or derived from rat (CAC2) [19]. A few validated ACC cell lines were generated through in vitro immortalization approaches using viral infection or transfection with hTERT [20,21]. Viral constructs containing the E6/E7 genes of HPV16 were used to infect and stably transform salivary gland tumor cell cultures [20]. This work resulted in the generation of two K14-positive ACC cell lines, i.e. UTSW-ACC52 and UTSW-ACC112. Alternatively, early passage ACC cells were immortalized with t(6;14) using h-TERT transfection, i.e. the MDA-ACC-01 (hTERT) cells [21] .

Here, we report the generation and establishment of a human adenoid cystic carcinoma cell line (UM-HACC-2A) that contains a MYB-NFIB translocation. We also describe cell culture conditions using fibronectin-coated plates and an optimized salivary gland medium (OSGM) that enabled us to generate this ACC cell line in absence of in vitro immortalization techniques.

Materials and Methods

Cell line generation and culture conditions

Patients with adenoid cystic carcinoma were recruited and consented in accordance with University of Michigan institutional policies. The patient who provided the specimen that enabled generation of the UM-HACC-2A cell line was a 53 year-old Caucasian female with a T3N1M0 ACC in a minor salivary gland at the base of the tongue. Tumor dissociation was performed as described [22]. Briefly, tumors were minced in small pieces, passed through a 25-ml pipette and centrifuged at 1,000 RPM, 4°C for 5 minutes. Tumor pieces were placed in a sterile Petri dish (Fisher Scientific, Pittsburgh, PA), digested in collagenase-hyaluronidase (Stem Cell Technologies, Vancouver, Canada) at 37°C for 45–60 minutes. Residual fragments were disrupted manually every 15 minutes using a 25-ml pipette followed by a 10-ml pipette. Single cell suspensions were prepared by passing the mixture through a 40-µm sieve (Fisher) placed in a Falcon tube containing 3–4 ml Fetal Bovine Serum (FBS; Invitrogen, Carlsbad, CA). Red blood cells were lysed for 1 minute using AKC lysis buffer (Invitrogen) and spun at 1,000 rpm, 4°C for 5 minutes. The resulting cells were cultured in Optimal Salivary Gland Medium (OSGM) and expanded in flasks coated with fibronectin (Sigma-Aldrich, St Louis, MO). OSGM consists of high glucose Dulbecco’s Modified Eagle’s Medium (DMEM; Invitrogen), supplemented with L-glutamine (Invitrogen), AAA antibiotic (Sigma-Aldrich), 10% Fetal Bovine Serum (FBS; Invitrogen), 2% Bovine Brain Extract (Lonza, Walkersville, MD), 20 ng/ml rhEGF (R&D Systems, Minneapolis, MN, USA), 0.4 µg/ml human hydrocortisone (StemCell Technologies), 5 µg/ml human insulin (Sigma-Aldrich) and 1% amphotericin B (Sigma-Aldrich). A cell line was successfully established from a patient sample and named University of Michigan-Human Adenoid Cystic Carcinoma (UM-HACC)-2A. Cells were passed (1:3) every week using 2 rounds of fresh 0.25% trypsin solution (Invitrogen) for 3 minutes. Cells were frozen at every passage number in freezing medium containing 8 parts of OSGM, 1 part of dimethylsulfoxide (DMSO; Fisher) and 1 part of FBS (Invitrogen). The average population doubling time (PDT) was calculated using the equation elapsed time/(log2) (end # of cells/start # of cells).

Detection of MYB-NFIB translocation: PCR, RT-PCR, Whole exome sequencing (WES)

UM-HACC-2A cells were plated at 2.0 ×10⁵ cells per 60 mm³ fibronectin-coated dishes and grown to 90% confluence in OSGM. RNA was extracted using TRIZOL (Invitrogen) according to manufacturer’s instructions. 1 µg RNA was transcribed to cDNA (Invitrogen) and used in two-step PCR. Primer sequences used to identify the MYB-NFIB translocations were previously described [8,9]. Briefly, exons MYB5 or MYB12 were amplified with exon NFIB9 in round one of the PCR reaction. Then, exons MYB6 or MYB14 were amplified in a second round of PCR in combination with step one amplicons (i.e. MYB5 or MYB12 + NFIB9), as described [8,9]. MYB-NFIB translocations were identified by running the PCR product on a 1.5% agarose gel. A low passage primary human ACC cells (UM-HACC-11) known to be MYB-NFIB fusion-positive was used as positive control. Two mucoepidermoid carcinoma cell lines (UM-HMC-1 and UM-HMC-3A) were used negative controls. In addition, GAPDH was simultaneously run for each sample as loading control. The presence of the MYB-NFIB gene fusion in the UM-HACC-2A cells was confirmed independently using whole exome sequencing performed at the University of Michigan Bioinformatics Core Research Facility. Alignment of sample UM-HACC-2A was searched for either a read that was partially mapped to both MYB and NFIB, or a read pair whose R1 and R2 was mapped to MYB and NFIB separately. In this pair, R1 was mapped to MYB and R2 was partially mapped to both MYB and NFIB. The fusion was confirmed by using BLAT [23] on UCSC Genome Browser [24], and PCR, as follows. Genomic DNA was extracted from the patient and from UM-HACC-2A using Wizard Genomic DNA Purification Kit (Promega, Madison, WI). PCR primers (Suppl. Fig. 1B) were designed to validate the MYB-NFIB fusion identified in the whole exome sequencing.

Results

In vitro characterization of UM-HACC-2A cells

We describe here the generation and characterization of a cell line derived from an adenoid cystic carcinoma (cribriform pattern) localized at a minor salivary gland at the base of the tongue of a 53 year-old Caucasian female. We have been able to expand the UM-HACC-2A cells for more than 180 passages so far (Fig. 1A). Photomicrographs show cells that continuously grow in monolayers maintaining an epithelial morphology, i.e. polygonal in shape with fairly regular dimensions (Fig. 1A). In the first passages, the cells proliferated very slowly taking several weeks to reach confluence (Fig. 1B). However, starting around passage 12, we began to observe fairly stable and exponential growth pattern (Fig. 1B). The average population doubling time calculated for cell passage numbers 0–4 was 3.47 days and 2.21 days for passage numbers 96–100. Supporting the epithelial appearance of UM-HACC-2A cells, western blot analyses revealed high and stable levels of several epithelial markers (e.g. EGFR, E-cadherin, pan-cytokeratin) (Fig. 1C). Head and neck squamous cell carcinoma cells (UM-SCC-22B) were used as a positive control for the epithelial markers. To confirm the identity of the UM-HACC-2A cells, STR profiling was performed at multiple passages (e.g. passage 14, 33, 60, 90, 115, 148) and compared to the patient DNA. The STR profiling confirmed a 100% match to the patient’s DNA throughout the expansion of the UM-HACC-2A cell line in culture (Fig. 1D). We observed a loss of heterozygosity (LOH) with deletion of one allele (i.e. 8) at the D7S820 locus between passage 14 and 33 (Fig. 1D).

Figure 1.

Establishment and authentication of UM-HACC-2A, a human adenoid cystic carcinoma cell line derived from a minor salivary gland of the base of tongue of a Caucasian female. (A) Photomicrographs of UM-HACC-2A cells cultured for 1–182 passages in vitro (40x magnification). (B) Graph showing the increasing growth rate of UM-HACC-2A cells over time. The average Population Doubling Time (PDT) = elapsed time (log/2(end # of cells/start # of cells). (C) Western analysis showing expression levels of epithelial markers EGFR, E-cadherin and Pan-cytokeratin. (D) Table reporting the results of the STR profiling of UM-HACC-2A cells at passages #14, 33, 60, 90, 115 and 148 when compared to patient DNA.

UM-HACC-2A cells exhibit a MYB-NFIB translocation

To begin to assess the presence of the MYB-NFIB fusion, we performed RT-PCR analysis, using a method that was previously described by the Stenmam research group [8,9]. The analysis revealed that a MYB-NFIB translocation was present in UM-HACC-2A cells (Fig. 2A). As a positive control for the RT-PCR analysis, we used low passage primary human ACC cells (UM-HACC-11) derived from a patient known to be MYB-NFIB fusion positive. As a negative control, we examined salivary mucoepidermoid carcinoma cells lines (UM-HMC-1, UM-HMC-3A) previously characterized in our laboratory [22] that do not carry the MYB-NFIB translocation.

Figure 2.

UM-HACC-2A is positive for the MYB-NFIB fusion. (A) RT-PCR analysis of the UM-HACC-2A cells, as previously described [8,9]. The low passage primary UM-HACC-11 cells were used as a positive control for the expression of the MYB-NFIB fusion. Salivary gland mucoepidermoid carcinoma cell lines (i.e. UM-HMC-1, UM-HMC-3A) were used as negative controls for the MYB-NFIB translocation, as these tumors typically do not express this fusion protein. (B) MYB-NFIB fusion sequence generated by whole exome sequencing of UM-HACC-2A cells. (C) PCR analysis of the WES fusion sequence in the patient’s tumor and in the UM-HACC-2A cells. (D) Western blot analysis of putative ACC markers (c-Myb, p63), proliferation markers (PCNA, p21), and pro-survival Bcl-2 over several passages.

To further understand the nature of the MYB-NIFB fusion, we performed whole exome sequencing, which identified R1 and R2 sequences from the sample UM-HACC-2A. BLAT searching determined R1’s first 142 bases mapped to the plus strand of chr6: 135519804–135519945 which was within an intron of MYB, while R2’s first 122 bases mapped to the plus strand of chr9:14109789–14109910 which was within an intron of NFIB, and R2’s last 30 bases mapped to the minus strand of chr6: 135519916–135519945 which overlapped with the end of where R1 mapped (Fig. 2B, Supplemental Fig. 1A). A closer look at R1 showed that it overlapped with R2 at position 113 to 151 in both. The last 9 bases in R1, though they couldn’t be mapped by either BLAT or BWA due to the small length, were actually identical with the end of where R2 mapped on chr9. The mapping positions of this read pair indicated a fusion event between MYB and NFIB, with the breakpoint covered within the overlapping region between R1 and R2 (Fig. 2B, Supplemental Fig. 1A). The breakpoint was mapped at chr6: 135519945 (or 135519944) and chr9: 14109909 (or 14109910). The ambiguity is because position 142 in R1 and position 122 in R2 can be mapped to either MYB or NFIB. To validate the WES data, we designed PCR primers that amplified the MYB-NIFB fusion (Suppl. Fig. 1B). As expected, a 250 bp product was detected in the patient’s tumor DNA and in the UM-HACC-2A cell line at passages 2, 60, 115, 148 (Fig. 2C).

UM-HACC-2A cells express high levels of c-Myb and p63

To better understand the phenotypic changes (e.g. increased proliferation) acquired by UM-HACC-2A cells over increasing passages, we performed western blot analysis for proteins involved in cell cycle (e.g. c-Myb, p63, PCNA, p21) and cell survival (e.g. Bcl-2). At low passage, UM-HACC-2A cells expressed low levels of c-Myb, p63, PCNA, and Bcl-2 (Fig. 2D). However, starting at passage 20, we observed a significant increase in expression of all of these key proteins (Fig. 2D). The expression levels of p21 remained fairly constant throughout the UM-HACC-2A passages evaluated in this experiment (Fig. 2D).

Orthotopic transplantation of UM-HACC-2A cells

The human tumor that was used to generate the UM-HACC-2A cells was diagnosed to be an adenoid cystic carcinoma with predominant cribriform histology, as depicted (Fig. 3). To evaluate the ability of UM-HACC-2A cells to survive transplantation, we injected these cells in the submandibular gland of immunodeficient mice in a small, pilot study with 4 mice. Immunohistochemical staining for a human-specific marker (HLA) detected human cells (i.e. UM-HACC-2A) in a lymphoid nodule located in the submandibular gland (brown cells, white arrows, 40 and 100x) 6.5 months after transplantation in a mouse (Fig. 3). Notably, UM-HACC-2A cells exhibited slow growth in vivo, which is also typically observed in patients with ACC. Lung metastasis (black arrows) was identified in 1 mouse (out of 4 mice evaluated) after orthotopic transplantation of UM-HACC-2A cells (Fig. 3). Surprisingly, we did not observe major local tumor growth or perineural invasion at the submandibular salivary gland (site of UM-HACC-2A injection) in the mouse that exhibited lung metastasis.

Figure 3.

Histology of the human tumor used to generate the UM-HACC-2A cell line, orthotopic transplantation of the cells, and metastatic lung. Photomicrograph of hematoxilin-eosin stained tissue section demonstrating a circumscribed collection of malignant myoepithelial cells with a distinctive cribriform pattern characteristic of ACC from patient that donated the tissue used to generate the UM-HACC-2A cells (left panels). Photomicrographs of HLA+ cells, i.e. UM-HACC-2A cells (white arrows), within a lymphoid nodule in the mouse submandibular gland (center panels). Hematoxilin-eosin stained section exhibiting a metastatic tumor nodule (black arrows) in the mouse lung (right panels). The top photomicrographs were taken at 40x, and the bottom ones at 100x.

Feasibility of UM-HACC-2A cells for developmental therapeutics studies

To determine if UM-HACC-2A cells are useful for in vitro experimental therapeutics studies, UM-HACC-2A cell viability assays were performed using cisplatin and paclitaxel as prototypic chemotherapeutic drugs (Fig. 4). We observed that UM-HACC-2A cells were more responsive to paclitaxel with an IC₅₀ dose of 0.0006 μM (72 hours respectively) as compared to cisplatin, that showed an IC₅₀ dose of 44.7 µM (72 hours respectively).

Figure 4.

Graph depicting the results of cell viability assays (SRB) of UM-HACC-2A cells exposed to increasing concentrations of cisplatin or paclitaxel for 48 to 72 hours. The data were normalized against the untreated controls for each time period and drug concentration. Data were derived from triplicate wells per condition.

Discussion

To develop mechanism-based therapies it is imperative that signaling pathways driving ACC tumorigenesis, perineural invasion and progression towards metastasis are understood. The limited availability of validated cell lines has been a major roadblock that has hindered the progress in understanding mechanisms underpinning the pathobiology of ACC. Here we describe the method and optimization of culture conditions that enabled us to generate a human ACC cell line.

One of the key developments that enabled us to generate this ACC cell line was to optimize the culture medium used for the growth of these cells. Using a salivary gland growth medium that was previously described [20,22] as our reference medium, we added Bovine Brain Extract (BBE). This was inspired by the recognition that many ACC patients present with perineural invasion, implying that neural cells secrete growth factors and/or cytokines that support the growth and survival of ACC cells [3, 26–29]. Indeed, the presence of BBE was critical for the establishment of the UM-HACC-2A cell line.

Salivary gland tumors are rare and typically consist of multiple cell types. There are 24 different types of benign or malignant salivary gland tumors that have been described to date [30]. Adenoid cystic carcinoma is considered a “biphasic tumor”, as it consists of malignant cells originating from both ductal/acinar and basal/myoepithelial cells [31,32]. We observed here that the UM-HACC-2A cells consistently express cytokeratins. P63 is a protein typically expressed in the in basal and myoepithelial cells of ACC and is often used for differential diagnosis with other salivary gland malignancies [31,32]. We observed strong expression of p63, particularly at passages higher than 20. As expected, UM-HACC-2A cells express numerous epithelial markers such as EGFR, Pan-cytokeratin and E-cadherin, which are well aligned with the epithelial appearance of the UM-HACC-2A cells.

Interestingly, around cell passage 20, we observed UM-HACC-2A cells began to proliferate faster. This finding correlated with an increase in the expression of proliferation markers (e.g. PCNA) and increased expression of the pro-survival Bcl-2 protein. Of note, Bcl-2 is a downstream target of MYB, which is highly expressed in UM-HACC-2A cells and typically overexpressed in ACC [9,33]. Further, our STR profiling showed genetic instability, as defined by loss of heterozygosity (LOH) with allele 8 drop out from the D7S820 locus at about the same time. Indeed, STR profiling has been used to show that cell lines may acquire additional genetic changes in culture [34] or when cancer specimens are compared to patient DNA [35]. Another possible explanation for the increase in growth rate and enhanced expression of these key proteins might be related to the senescence of stromal cells and progressively higher proportion of presence of actual tumor cells in our cultures. The average population doubling time decreased from 3.47 days for earlier cell passages to 2.21 days for higher cell passages. These results correlate well with previous studies that reported an increase in growth rates of patient-derived xenografts (PDX) models of ACC with increasing in vivo passages [36, 37].

Gene fusions contribute to the early steps of tumorigenesis and many are diagnostic and/or prognostic markers for disease [38]. MYB-NFIB is frequently found in ACC, but not in other salivary gland tumors [8–15]. Whole exome sequencing (WES) is considered an excellent approach to screen and identify fusion events [39]. For example, Chmielecki and colleagues reported a novel NAB2-STAT6 fusion in rare, solitary fibrous tumors (55%) compared to whole blood matched normal control samples using WES [39]. Yang and colleagues used WES to identify >9,000 fusion events near exons and intronic regions (exon-intron borders) in 15 cancer types [40]. WES has also been used to study other molecular events in ACC, revealing mutations in FGFR2, NOTCH1/2, PI3KCA and additional novel genes [41]. Here, WES demonstrated the presence of the MYB-NFIB fusion in the UM-HACC-2A cells. When the RT-PCR and the WES analyses are examined together, one may conclude that the fusion found through sequencing is consistent with the band observed upon the use of the MYB5/6-NFIB9 primer combination. However, the fusion reported by WES is not consistent with the bands observed when the MYB12/14-NFIB9 primer combination was used in the RT-PCR. We interpret these data as the likely presence of at least one alternative fusion event that is intronic in nature and that was not captured with WES performed here. We postulate that these cells constitute a useful model for study of fusion positive ACC tumors. However, it is important to point out that MYB-independent NFIB fusions have also been reported in ACC tumors [15]. The development of additional ACC cell lines would indeed be beneficial to the field, as a panel of cell lines could represent more broadly the full spectrum of complexities of this disease.

A major driving force for the 8 years that we invested in the attempt to generate an ACC cell line was to develop an experimental model that would enable in vitro screening of anti-cancer drugs. Here, we used the UM-HACC-2A cells to test the effect of cisplatin and paclitaxel in cell viability assays. UM-HACC-2A cells were sensitive to paclitaxel but not as responsive to cisplatin. These results actually reflect the poor response that most ACC patients have when treated with cisplatin [4–7]. We believe that the UM-HACC-2A is a cell line that will be ideally suited for studies aiming at identification of novel targeted and chemotherapy drugs, and provide a new platform for studies of combination therapies for ACC.

When we seeded the UM-HACC-2A cells into biodegradable scaffold and transplanted them in the subcutaneous of immunodeficient mice [22], we did not observe the generation of tumors (data not shown). This negative finding led us to injected the UM-HACC-2A cells directly into the mouse submandibular gland. When injected orthotopically, we observed that these cells survived and proliferated slowly for a relatively long period of time (i.e. 6 months). We believe that the slow growth of the UM-HACC-2A cells observed in this pilot study represents well the slow growth typically observed in human patients with minor salivary gland ACC [42]. HLA staining showed that the human ACC cells accumulated in a lymphoid nodule of the submandibular gland. We are unsure if they were found preferentially in this location because of that was the site of injection or because this was a more “welcoming” microenvironment for the growth of these cells. Lymph node metastases might be associated with salivary gland cancers, but they are not observed in the majority of the patients [43–46]. Nevertheless, the patient who donated the tissue used to generate the UM-HACC-2A cell line presented with a lymph node metastasis at the time of primary tumor resection. Notably, the UM-HACC-2A cell line was derived from a T3N1M0 tumor. It is postulated that the ability of this tumor to metastasize at a relatively early time point in its development might have contributed to our ability to generate this cell line. However, since this is the only ACC cell line that we were able to generate so far, this hypothesis remains to be adequately tested.

Distant metastases to lung, bones, and liver are common in patients with salivary gland tumors, including ACC [1–7,46]. Ali and colleagues reported that 50% of all patients with salivary gland tumors (n=301) develop distant metastases, and the most common site was lung (49%), bone (40%) and liver (9%) [47]. Another study reported that 56.3% of ACC patients developed distant metastasis and that the most common site was the lungs (72.2%) [48]. In our pilot studies, we observed the formation of a metastatic node in the lung of one mouse that received orthotopic injection of UM-HACC-2A cells. We should be very cautious interpreting this finding as we only found lung metastasis in one mouse out of 4 mice injected with the cells. As such, we cannot say that this is a model of distant metastasis. Indeed, we are currently performing larger studies to verify this finding. If lung metastases are observed frequently, the UM-HACC-2A cells could be perhaps explored as a model of ACC tumor dissemination.

Conclusion

In summary, we describe the generation and characterization of a novel human adenoid cystic cell carcinoma (ACC) cell line that contains the pathognomonic MYB-NFIB translocation and expresses high levels of c-Myb. We performed small feasibility studies that suggested the potential value of the UM-HACC-2A cell line for drug screening and as a xenograft model of ACC. We propose that the availability of validated cell lines and xenograft mouse models is critical to enable mechanistic studies that aim at improving the survival and quality of life of patients with ACC.

Highlights.

• -We report a method for generation and culture of an adenoid cystic carcinoma cell line (UM-HACC-2A)
• -UM-HACC-2A exhibit the MYB-NFIB fusion, which is frequently observed in adenoid cystic carcinoma
• -UM-HACC-2A is useful for drug screening in vitro and for orthotopic transplantation in SCID mice