Adenoid cystic carcinoma (ACC) is a rare cancer that frequently impacts the head and neck. The disease can be found in any of the major salivary glands, minor salivary glands, lacrimal or ceruminous glands. Current epidemiologic estimates of ACC suggest an incidence of 3-4.5 cases per million and that ACC accounts for approximately 1% of all head and neck cancers. In the United States, there are 1200 new salivary gland ACC cases each year.
One model that is used to study ACC is the patient-derived xenograft (PDX), where investigators transplant primary ACC patient tumors into immunodeficient mice. During the inception of the model, microarray experiments and immunohistochemical assays were principally employed to demonstrate the similarity of the transcriptome and proteome of PDX models and ACC tissue. Though several investigations have studied the transcriptome of the ACC PDX model, few have applied high throughput sequencing technology. In particular, virtually no investigations have utilized RNA sequencing to characterize the transcriptome of the ACC PDX model to reevaluate its utility as a research model.
Here, we examined the transcriptome of the ACC PDX model through paired end RNA sequencing data provided by the ACC Research Foundation. We analyzed sequencing reads from ten ACC PDX models and six normal salivary gland tissue samples. We utilized the pseudo aligner Kallisto to quantify transcripts in both the ACC PDX model and normal salivary gland tissue. Then, we utilized the DESEQ2 pipeline to complete a differential expression analysis of the ACC PDX and normal salivary gland tissue samples.
Through unsupervised hierarchical clustering, we observed that ACC PDX samples clustered together and away from normal salivary gland tissue. Through differential expression analysis, we observed that of 18084 genes considered, 6993 genes were found to be differentially expressed between ACC PDX and normal salivary gland tissue (FDR <.05). Overall, we observed that many ACC tumor defining genes were found to be differentially expressed in the ACC PDX model. However, we also observed that several housekeeping genes were found to be unexpectedly differentially expressed in the ACC PDX model.
Most research surrounding ACC has focused on the oncogene, MYB, because of its high prevalence among ACC tumors. MYB is a transcription factor that coordinates cell progression through mitosis. Previous studies have suggested that MYB plays an instrumental role in driving the malignant, aggressive nature of salivary gland ACC. Interestingly, a gene fusion known as MYB-NFIB has emerged as a defining molecular event of the disease. Through our differential expression analysis, we observed that MYB and its paralog MYBL1 were found to be differentially expressed between ACC PDX and normal salivary gland tissue. NFIB was not found to be differentially expressed. However, this finding is sensible given that the MYB-NFIB fusion is not present in all ACC PDX models.
Through our work, investigators will have a better understanding of the advantages and limitations of the utilization of the ACC PDX as a model for ACC research.