Our Science – Paterson Website

Bruce M. Paterson, Ph.D.

Laboratory of Biochemistry and Molecular Biology Head, Biochemistry of Gene Expression Section Senior Investigator Laboratory of Biochemistry, NCI/CCR 37 Convent Drive Building 37, Room 6118, MSC 4255 Bethesda, MD 20892-4255 Phone:   301-496-1746 Fax:   301-402-3095 E-Mail:   bruce.paterson@nih.gov

Biography

Dr. Bruce Paterson received his Ph.D. in zoology in 1971 from the University of California-Berkeley with Dr. R. Strohman, studying muscle-specific gene expression during myogenesis in vitro. He continued this work at the Weizmann Institute of Science in Rehovot, Israel, with Dr. D. Yaffe, from 1971 to 1974, followed by an additional year with Dr. J. Bishop at the University of Edinburgh, Scotland, studying mRNA complexity changes during myogenesis. He has been in the Laboratory of Biochemistry since 1975.

Research

Molecular Studies on Eukaryotic Gene Regulation

Tissue formation during development involves the determination, controlled proliferation and migration of cells that give rise to particular differentiated cell types in the embryo. Misregulation of this process can lead to the loss of proper embryonic development and uncontrolled cellular growth, a hall mark of the cancerous cell. Thus the study of gene regulation during development can provide insight into processes that play a role in the development of human diseases as well as potential approaches for therapeutic intervention.

Embryonic muscle formation in vertebrates and Drosophila (the fruit fly) provide excellent model systems in which to study the formation of one of the major tissues in higher organisms,muscle. The determination, proliferation, and differentiation of muscle cells during development in both vertebrates and invertebrates depend upon the function of the MyoD family of basic helix-loop-helix proteins, know as the muscle regulatory factors (MRFs). Determination of the first muscle precursor cells in vertebrates involves the activation of the MRFs in the early somitic mesoderm while gene expression characteristic of differentiated muscle remains repressed as the committed myoblasts migrate from the somite to the various areas of the developing body that will form the muscles of the trunk,limbs and thorax. Terminal differentiation is marked by the withdrawal of the myoblast from the cell cycle just prior to the activation of the muscle-specific gene set and both processes involve the MRFs. More than 90% of the genes expressed in the dividing muscle cell are transcriptionally silenced while the muscle-specific genes are activated. This switch in the pattern of gene expression is permanent and likely involves substantial changes in chromatin organization, mediated in part by the MRFs. Cell cycle control during terminal differentiation also involves the MRFs in a pathway that regulates the phosphorylation status of the retinoblastoma protein, Rb, in response to extra cellular cues that modulate myoblast differentition.

Import of cdk4 into the nucleus is dependent upon cyclin D1 which, in turn, is regulated by mitogen levels. In the absence of mitogens cyclin D1 is degraded and cdk4 is transported to the cytoplasm. We have shown that ectopically expressed MyoD binds directly to the G1 cyclin-dependent kinase, cdk4, to inhibit cell growth and the phosphorylation of Rb. The cdk4-MyoD interaction also blocks the trans activation functions of MyoD by disrupting DNA-binding by the MyoD/E-protein heterodimer. Therefore, normal levels of nuclear cdk4 inhibit MyoD function in dividing myoblasts while a decrease in the nuclear cdk4 concentration in response to reduced levels of growth factors and mitogens allows MyoD to function. We have identified a 15 amino acid domain on MyoD responsible for the interaction with cdk4. Expression of this domain either as a fusion protein with GST or GFP inhibits the kinase activity of cdk4 in vitro and in vivo, blocking its ability to phosphorylate the retinoblastoma protein, Rb. This results in the cessation of cell growth and induces myoblast differentiation in the presence of mitogens. We have a patent pending on the inhibitory activity of the 15 amino acid MyoD cdk4-binding domain domain. We would like to obtain structural information on the interaction of the MyoD-cdk4 binding domain inorder to use this information to design better inhibitors of the G1 cdks as potential reagents to reduce uncontrolled cell growth in human disease.

The nautilus gene is the single member of the bHLH family of myogenic regulatory factors in Drosophila. We have previously shown that nautilus is required to convert S2 cells to a myogenic phenotype, that nautilus in combination with the Drosophila E-protein, daughterless, will convert mammalian fibroblasts to a myogenic state, and that nautilus can rescue a lethal C. elegans HLH-1 (MyoD) null. Furthermore, ablation of nautilus-expressing cells in the Drosophila embryo completely abrogates muscle formation. All of these results pointed to the importance of the nautilus gene product in Drosophila myogenesis. In the absence of a null for the nautilus gene, we injected nautilus dsRNA into early embryos to phenocopy a loss-of- gene function by RNAi. Greater than 75% of the injected embryos did not develop normally compared to embryos injected with beta-galactosidase dsRNA and showed a severe loss-of-muscle phenotype in several instances. A subsequent genetic study by another group, using EMS-induced mutations, noted that the genetic loss-of-function phenotype in two EMS alleles differed from that determined using RNAi. Briefly, embryos lacking nautilus were only missing a subset of muscles. This loss is tolerated and does not result in lethality at any stage of development. However, the EMS alleles are sterile. In order to resolve the discrepancy in phenotypes between the RNAi and EMS methods, we used direct gene targeting to introduce an armidillo-GFP expression cassette in the middle of the nautilus gene to disrupt gene expression in collaboration with Dr. Yikang Rong (LMCB NCI). The targeted disruption of nautilus has a range of muscle pattern phenotypes, from the severe loss of muscle organization to minor muscle pattern disruptions. Each stage of development is affected. The majority of pupae are unable to exit the pupal case but the small precentage that do hatch successfully usually die in the food. The adult females that survive have highly distended abdomens and cannot lay eggs, suggestive of a muscle defect. The males can mate but are weak, and both sexes perish within a few days after hatching. Loss of nautilus gene function is similar to that for the loss of the C. elegans HLH-1 gene, CeMyoD: both genes appear to be essential for the integrity of the muscle fibers but not absolutely necessary for fiber formation and both nulls are essentially lethal. Rescue of nautilus function in the disruption by an HSP-nautilus transgene are underway. A complementary strategy using an inducible inverted repeat of a nautilus transgene that can be expressed in a developmental and tissue-specific pattern to induce RNAi complements the phenotype of the targeting results. We conclude the initial RNAi studies with injected nautilus dsRNA revealed the most severe dusruption phenotypes noted in the gene targeting studies. Furthermore the results demonstrate that the nautilus gene is essential for normal embryonic myogenesis and viability but is not required for myogenic commitment.

In order to utilize RNAi in vivo, we have been developing Drosophila transformation vectors that allow the induction of dsRNA in selective tissues at particular times during development. Our first attempts using an inducible T7 polymerase with opposable T7 promoters flanking the gene of interest worked to produce dsRNA in cultured S2 cells but turned out to be lethal in transgenic Drosophila, presumably due to cryptic T7 binding sites in the genome. A second approach, using the gal4 system and a flp-induced hairpin structure, demonstrated we could achieve targeting of the white eye color gene but screening for the flp induced hairpin made the method less convenient. A third approach took advantage of the the stability of inverted repeats flanking a large DNA spacer. As a spacer we used the 900bp first intron of the cytoplasmic actin 5C gene and placed the expression of the hairpin under the control of gal4. The intron is spliced out to give a perfect dsRNA. This vector, called pINT-1, has proven successful in cultured S2 cells in blocking the expression of GFP and nautilus and has been recently tested in transgenic flies to target nautilus expression giving the same phenotype as the targeted disruption.

In most lower eukaryotes RNA silencing depends upon the action of RNA-dependent RNA polymerase (RdRP). We have presented evidence that siRNAs produced in Drosophila embryo extract serve as primers to convert the target RNA into dsRNA which is then degraded by the RNase III-related enzymes, called the Dicers, to produce secondary siRNAs while degrading the target RNA in the process. This was the first evidence for RdRP activity in Drosophila associated with the RNA the targeting process. We have developed a primer extension assay for the purification of Drosophila RdRP and have shown by nearest neighbor analysis that UTP is internally incorporated to reflect the base composition of the template strand. We have demostrated template-dependent incorporation of bromo-uridine in S2 cells in the presence of actinomycin D and a-amanitin that co-localizes with the template in P-bodies, the cellular sites of RNA degradation. In silico analysis of potential RdRPs has identified a possible candidate in Drosophila that does not have the canonical RdRP domain but is related to known viral RdRPs. This protein has been cloned, expressed in the baculovirus system and is under study.

The potential application of RNA silencing in disease therapy has lead us to investigate the RNA silencing pathway in detail using Drosophila as the model system since dsRNA and siRNAs can be used to silence gene expression. We are interested in the initiation and targeting steps and have cloned several of the Drosophila genes in the major gene families involved in RNAi identified in genetic screens in C. elegans. Candidate genes are evaluated in S2 cells expressing b-galactosidase and GFP. dsRNA for potential genes involved in RNAi are cotransfected with GFP dsRNA or a GFP synthetic siRNA and evaluated for inhibition of GFP silencing. We have cloned Drosophila Dicers 1 and 2, the RNAse III-related proteins involved in the formation of siRNAs, three dsRNA-binding proteins, Argonaute 2, helicases and members of the nonsense mediated decay proteins or smgs. We have established that both Dicer-1 and Dicer-2 participate in the initial processing step as well as siRNA-mediated targeting and have evidence to suggest Dicer-1 and Dicer-2 are in the RNA-induced silencing complex (RISC). We have shown that Dicer-1 can cleave dsRNA processively from the ends in the absence of ATP or any NTPs, whereas Dicer-2 requires ATP and is distributive in action. We have identified dsRNA-binding proteins that appear to be involved in presenting miRNA precursors to Dicer-1 and proteins that may carry nascent siRNAs to the targeting protein, Argonaute 2. Only the single strands of the siRNA bind to Agronaute 2 and we have cloned and expressed a potential helicase involved in unwinding the siRNAs prior to interaction with Argonaute 2. We have cloned and expressed Argonaute 2 in the baculovirus system and have determined that it is associated with both Dicers as well as two of the dsRNA binding proteins in a protein complex that is likely to represent the minimal RISC. It has been previously proposed that single-stranded siRNAs bound to Argonaute 2 act catalytically to cleave the target RNA and this is being evaluated.

Schneider cells can be induced to activate a partial myogenic program by expressing daughterless in these cells. This myogenic conversion is potentiated by the co-expression of DMEF2 and nautilus. The cells exit the cell cycle, become multinucleated and express muscle-specific myosin heavy chain. Myogenic conversion of Schneider cells by daughterless is dependent upon the endogenous expression of very low levels of nautilus and DMEF2. Inhibition of the endogenous nautilus and DMEF2 by RNA interference demonstrated the expression of both genes was essential for myogenic conversion of Schneider cells by daughterless. Quantitative RT-PCR has shown that Schneider cells express 100-1000 fold less daughterless than nautilus. Raising the levels of daughterless protein by ectopic expression allows sufficient levels of the nautilus/daughterless heterodimer to form to activate the myogenic program. This work defined conditions for the application of RNAi to cultured Drosophila cells and established a myogenic system in which to potentially identify nautilus-responsive genes by micro array analysis. Unfortunately, the metalothionine promoter is sufficiently leaky so that substantial differences between the induced and uninduced cells were not that well defined in the initial microarray studies. The nautilus null fly larvae are being used instead in a collaboration with the Canadian Drosophila Microarray Centre and Dr. Tim Westwood.

We have recently started to explore the role of the RNAi pathway in heterochromatin formation in Drosophila and have evidence that transient heterochromatic patches can be induced that depend upon target integration, that are TSA sensitive, that require the function of the Argonaute and Dicer proteins, similar to heterochromatin maintenance in S. pombe, and that contain increased amounts of K9Me3 histone. Although the induction is transient it may be possible to stabilize the modification leading to permanent gene silencing. This could lead to an entirely new approach to gene silencing and the treatment of human diseases.

This page was last updated on 6/12/2008.