Skip to Search
Skip to Navigation
Skip to Content

UConn Health Department of Cell Biology

Paul Epstein

Paul Epstein

Associate Professor
Department of Cell Biology

Education and Training

A.B., Columbia College
Ph.D., Albert Einstein College of Medicine


Phone: 860-679-2810
Fax: 860-679-3693 
Office: B261012

UConn Health
263 Farmington Avenue 
Farmington, CT 06030

Research Interests

Current Research (PDF)

Research in my laboratory centers on second messengers and signal transduction. Particular focus has been on cyclic nucleotide metabolism and protein phosphorylation, with emphasis on analysis of cyclic nucleotide phosphodiesterases (PDE). It has become apparent, in recent years, that PDE is a complex enzyme system consisting of many different forms, some of which are tissue selective. Hence, by inhibiting or altering the expression of specific forms of PDE, we can alter cAMP levels in specific cell types, and change fundamental physiological processes in one cell type, without affecting others. Thus, a main research goal in this laboratory is to clone the genes for different forms of PDE and examine their tissue distribution and regulation of their expression during development and in different pathophysiological states.

One focus of our research is to use an antisense approach to inhibition of specific PDE forms as a novel treatment method for certain types of cancer. Recently we have cloned the cDNAs for different forms of PDE in normal and leukemic lymphocytes, as a means of probing the structure of these enzymes, and to facilitate analysis of their biological functions We have shown: 1) that pharmacological inhibitors of calmodulin-dependent PDE (PDE1) and cAMP specific PDE (PDE4) induce apoptosis in seven different leukemic cell lines that we have growing in our laboratory, 2) used quantitative reverse transcription-polymerase chain reaction (RT-PCR), to show that the gene for the 63 kDa CaM-PDE (PDE1B1) is expressed in the leukemic cell lines and in normal HPBL following mitogenic stimulation with phytohemagglutinin, but not in unstimulated, normal, quiescent HPBL, 3) cloned and sequenced the entire cDNA for the human form of PDE1B1, which had not been achieved before, 4) designed and synthesized phosphorothioate antisense oligodeoxynucleotides (AS ODNs) for disruption of the CaM-PDE gene, and the corresponding nonsense (NS) ODNs, and tested them for their effects on these leukemic cell lines. The AS ODNs selectively induced apoptosis in the leukemic cells, whereas the NS ODNs had no effect. That the AS ODNs work through disruption of the PDE1B1 gene was shown through selective depletion of PDE1B1 mRNA and enzymatic activity after treatment with the AS ODNs. This work has recently been published (see references 1-3 listed below). Work is continuing on this system with respect to: 1) looking for splice variants of PDE1B1 in human leukemic cells, 2) looking at changes in expression of isoforms of PDE4 between normal and leukemic cells, 3) developing new, additional AS ODNs targeted to PDE1B1 and making the AS ODNs more potent by experimenting with chemical derivatization and different uptake methods, and 4) planning in vivo studies of the effectiveness of the AS ODNs, by looking at their ability to prolong life of SCID mice implanted with human leukemic cells. We have recently begun to conduct similar studies with breast cancer cell lines as well.

Another system in which we are examining PDE is in rat brain. When cGMP is generated in the brain, upon release of nitric oxide, one of the main receptors for this cGMP is the cGMP stimulated form of PDE (PDE2), which has a high affinity, allosteric binding site for cGMP, and is believed to play a potential role in regulation of learning and memory. Using a rat brain cDNA library, we cloned a PDE2 cDNA, and examined its pattern of distribution in regions of the brain (4). Analysis of the brain cDNA we isolated showed it to code for an entirely novel form of PDE2. The predicted protein sequence showed very close homology to that of the only other reported PDE2 clone, from bovine adrenal gland, except in their 5'-ends. The N-terminal 37 amino acids of the brain protein are highly hydrophobic and show no homology with the adrenal protein. Further isolation and analysis of a genomic clone for PDE2 proved that the cDNA we isolated represents an alternatively spliced product designed to target the enzyme to brain membranes. We have, since then, found evidence for and cloned two additional splice variants of PDE2, from brain (5), which are also novel and have never before been reported, and we are examining the distribution and properties of these novel forms of PDE2.


Epstein Schematic

Selected Publications

PubMed Listings