Alexander Pertsemlidis, Ph.D.
Rank: Associate Professor
Our research interests integrate computational biology, cancer biology and genetics. We study regulatory RNA molecules called non-coding RNAs (called that because they do not code for proteins), including microRNAs (miRNAs) and long non-coding RNAs (lncRNAs), and how they regulate cancer cell growth and response to anti-cancer drugs.
- nature has already developed elegant but non-obvious solutions to most problem facing modern medicine
- there may be many different solutions to the same problem
- things that look the same at the tissue level may look very different at the molecular level
- develop sensitive, non-invasive methods for early cancer detection
- identify new drugs targeting specific adult and pediatric cancer subtypes, either directly or in combination with traditional therapeutic agents
- ncRNA regulation of cell viability and drug response in neuroblastoma
- Trisomy 21 and protection against neuroblastoma
- Therapeutic miRNAs in combination with conventional chemotherapy
- Molecular sensors for detecting cancer at the single-cell level
- Separating tumor and host miRNAs through TU-tagging
- Therapeutic regulation of the PI3K and Wnt signaling pathways
- miRNA, lncRNA, and mRNA expression signatures of cancer cells and normal cells
- interaction networks between miRNAs, lncRNAs and mRNAs
- ncRNAs and their regulatory targets characterized in vitro, in vivo, and in silico
- functional relationships between ncRNAs and disease
- candidate biomarkers, therapeutic targets, and therapeutic agents.
Non-coding RNA Biology
The major focus of my lab is on investigating roles of non-coding RNA regulation in cancer pathogenesis, specifically: (1) non-coding RNA regulation of cell viability, and (2) non-coding RNA regulation of drug response. The long term goals of these projects are the identification of ncRNAs for which serum expression is a biomarker of either the presence or progression of tumors or of the likely response of a tumor to drug treatment, and of ncRNA mimics or inhibitors that can be delivered as therapeutic agents. Both projects integrate in silico, in vitro, and in vivo approaches.
Given the steadily increasing use of high-throughput methods in biomedical research, modern biologists need to understand both how biological data is collected and how to express biological problems in terms of algorithms and data structures. I believe that students should combine wet-lab and dry-lab work in both their courses and research projects. Such interdisciplinary training in computational and systems biology is appropriate to how modern biomedical research is evolving.
Neuroblastoma (NB) is a tumor that originates from neural crest precursor cells. It is the most common cancer in infants and the most common extracranial solid tumor in children, accounting for 15% of all childhood cancer deaths. The disease is highly heterogeneous and is stratified into low- and high-risk categories. The overall prognosis for those with high-risk or relapsed disease remains poor despite the standard therapies of surgery, radiation, and chemotherapy. Low-risk neuroblastoma, however, frequently shows spontaneous regression, mainly in tumors with a near triploid number of chromosomes. The dichotomy between low- and high-risk neuroblastoma also raises interesting questions: (1) What are the molecular differences between low-risk and high-risk NB that lead to spontaneous regression in the former? Is the presence of additional copies of chromosomes in low-risk NB a clue as to the molecular mechanisms underlying regression? (2) Can specific genes or ncRNAs be altered to selectively kill NB cells or improve response to the drug?
Lung cancer is divided into two major groups. 15% of bronchogenic carcinomas are small cell lung carcinomas (SCLC). Untreated SCLC has the most aggressive clinical course of any type of pulmonary tumor, with a median survival of only 2-4 months from diagnosis. SCLC is typically diagnosed only when the disease has already metastasized, beyond the point at which surgical or radio/chemotherapeutic intervention is likely to be of benefit. The other 85% of bronchogenic carcinomas are non-small cell lung carcinomas (NSCLC), which is made up of several histological subtypes, including adenocarcinoma, squamous cell carcinoma, and large cell carcinoma. NSCLC progresses relatively slowly and is characterized by significant heterogeneity in its response to treatment. This dichotomy between NSCLC and SCLC raises interesting questions: (1) Are specific genes or non-coding RNAs differentially expressed between states or between extremes of a phenotype within a state, like a drug response? (2) Can a phenotype like a drug response be changed by altering intracellular levels of a gene or non-coding RNA? (3) Can one subtype be turned into the other? Is SCLC a state that can be turned off? A major molecular difference between NSCLC and SCLC is the expression of a neuroendocrine program in the latter, including markers such as CHGA, SYP, and NCAM. Interestingly, some SCLC has lost its neuroendocrine program, and some NSCLC has gained a neuroendocrine program
Development and application of statistical methods to test associations between genotype and phenotype and testing the conventional wisdom that common disease is explained by common variants. Power and sample size calculations, kinship calculation, and statistical tests for association.
Yiqiang Zhang, Ph.D.