T H E   N I H   C A T A L Y S T     N O V E M B E R  –  D E C E M B E R  1999



Harris Bernstein

Harris Bernstein received his Ph.D. in biology from the Massachusetts Institute of Technology in Cambridge in 1987 and did postdoctoral work at the University of California, San Francisco, before joining the Genetics and Biochemistry Branch of NIDDK in 1992. He is now a senior investigator there.

My laboratory studies protein translocation across and insertion into cell membranes. Most of our work has focused on the transport of proteins across the mammalian endoplasmic reticulum (ER) and the bacterial inner membrane (IM), which are evolutionarily related processes. Although the cellular factors that decode ER and IM targeting signals and the components of the conserved protein-conducting channel ("translocon") have been identified, many steps in the transport pathway are still poorly understood. How membrane proteins are incorporated into the lipid bilayer with the correct topology is particularly enigmatic.

Work initiated in the 1980s showed that in mammalian cells a ribonucleoprotein complex called the signal recognition particle (SRP) recognizes both the signal sequences of presecretory proteins and the transmembrane segments of integral membrane proteins co-translationally and then targets nascent chain complexes to the ER. Interaction of SRP with an ER-bound receptor then catalyzes insertion of the nascent chain into the translocon. We have been particularly interested in understanding the mechanism by which the SRP 54-kD subunit (SRP54) recognizes signal sequences with a high degree of fidelity and releases them only after arriving at the ER. We have found that the three domains of SRP54—an NH2-terminal four-helix bundle ("N domain"), a central GTPase ("G domain"), and a COOH-terminal signal peptide binding domain ("M domain")—play distinct roles in the targeting cycle. Our studies show that the N domain promotes high-affinity signal peptide binding and that the GTPase acts as a switch that promotes signal peptide release at the ER. The data suggest that the N domain serves as a "lid" for the signal peptide binding pocket that is opened by a GTP-induced conformational change. We are currently continuing to explore how the three domains work together to promote accurate protein targeting.

While several laboratories were busy characterizing the mammalian SRP pathway, several other groups showed convincingly that presecretory proteins are targeted to the Escherichia coli IM post-translationally by molecular chaperones. Their results predicted that SRP would be unnecessary in prokaryotes and thus found only in eukaryotic cells. The surprising discovery of SRP in bacteria through sequence gazing in 1989, however, created an apparent paradox. Our most significant recent contribution has been to resolve this long-standing puzzle. Using a combination of genetic and biochemical methods, we showed that SRP targets integral membrane proteins to the IM in E. coli. In addition, we have obtained evidence that the SRP-targeting pathway has been widely conserved in prokaryotes not only because it increases the efficiency of membrane protein biogenesis, but also because it prevents the toxic accumulation of mislocalized membrane proteins in the cytoplasm.

Our studies on protein transport pathways in bacteria have also led us to some new insights into the function of the translocon. We have identified a mutant form of E. coli SecY, the most highly conserved translocon subunit, that has a specific defect in membrane protein insertion. Studies on this mutant demonstrate that the membrane protein insertion and protein translocation functions of the translocon are at least partially separable. We are now attempting to isolate additional translocon mutants in an effort to understand how the translocon performs two related but distinct functions. Finally, we have found that the bacterial SecA protein, which was previously thought to participate only in the export of proteins targeted post-translationally, also facilitates the insertion of membrane proteins targeted by SRP. Our most recent results suggest that SecA may play a wider role in membrane protein insertion than was previously expected.

In a departure from our work on ER/ IM translocation, we have recently begun to study an unusual phenomenon known as "nonclassical secretion." More than a dozen secreted cytokines (for example, interleukin 1 and basic fibroblast growth factor) and viral proteins (for example, HIV Tat) have been described that lack typical NH2-terminal signal sequences. The export of these proteins does not appear to involve passage through the normal ER-Golgi route. Currently, we are using genetic approaches to identify cellular factors that promote nonclassical secretion in Saccharomyces cerevisiae.

Charles Egwuagu

Charles Egwuagu received a masters degree in public health and a Ph.D. in epidemiology and microbiology from Yale University in New Haven, Conn., 1987. He joined the NEI Laboratory of Immunology in 1987 and is currently a senior investigator and head of the Section of Molecular Immunology.

The focus of my research is on molecular mechanisms that underlie the etiology and susceptibility to organ-specific autoimmune diseases. We emphasize a group of intraocular inflammatory diseases (uveitis) of presumed autoimmune etiology.

In the early part of my work, I demonstrated that T cells expressing T-cell antigen receptors (TCRs) of the Vb8 family are amplified in the retina of uveitic rats and might therefore be responsible for the induction of experimental autoimmune uveitis, the animal model of human uveitis. A significant number of the Vb8 TCRs contain a conserved Val-Gly motif in the third complementarity-determining region, suggesting that this TCR motif may provide an immunotherapeutic target. These results moved us to extend our studies to human diseases. We have found evidence of selective recruitment and amplification of Vg2+ T cells in tear ducts of patients with ocular sarcoidosis and are now examining other uveitic conditions.

An important and unresolved problem in autoimmunity is defining risk factors for development of an organ-specific autoimmune disease. Why are some individuals resistant while others are susceptible? Recent studies in my laboratory have shed some light on this. We discovered that ocular-specific antigens that are the targets for pathogenic autoimmune processes are expressed in the thymus of some animals. Animal species that possess the thymic antigens are resistant, while those that do not are susceptible to disease induction. Furthermore, the degree of susceptibility or resistance depends on the relative amounts of the autoantigens in the thymus. These data suggest a novel explanation for differences in susceptibility to autoimmune diseases: Resistance to an organ-specific autoimmune disease may be regulated at least in part by capacity to establish central tolerance to the relevant autoantigen. We have extended these studies to humans, and preliminary results indicate that the level of thymic expression of two putative ocular autoantigens (S-Antigen and IRBP) may serve as a useful indicator of susceptibility or resistance to uveitis. The general applicability of this concept to other autoimmune diseases remains to be established.

One of the cytokines that has been implicated in the immunopathogenic mechanism of a number of organ-specific autoimmune diseases is g-interferon (IFN-g). However, whether IFN-g plays a role in the induction or recovery from the disease is still a matter of debate. Recent studies in the mouse have shown that IFN-g confers protection against experimental allergic encephalomyelitis, a model of multiple sclerosis. To explore the potential benefits of IFN-g in the management of uveitis, we generated transgenic (TR) rats and mice with targeted ectopic expression of IFN-g in the eye. These models enabled us to study the consequences of prolonged exposure of ocular tissues to this cytokine. The IFN-g rat strain is the first TR rat generated at NIH.

Analysis of these rats revealed that an important consequence of prolonged exposure of ocular cells to IFN-g—as may occur during chronic or recurrent uveitis—is the induction of choroidal inflammation, formation of retinal folds, activation of pro-inflammatory genes, and enhanced susceptibility to anterior and posterior uveitis. Thus, in contrast to the protective effect of systemic IFN-g in the mouse, constitutive secretion of IFN-g in the rat eye predisposes the animal to severe uveitis. The TR rats also show progressive degeneration of the neuroretina and selective apoptosis of ganglion cells. These are early signs of glaucoma and nutritional amblyopia. TR rats are clearly a unique and important animal model for studying etiologic mechanisms of glaucoma and uveitis.

During the course of studies on our IFN-g TR mice, we discovered that several members of the interferon regulatory factors (IRFs) family of transcription factors are constitutively expressed in the lens. We have also shown that the expression of these IRFs is tightly regulated. Perturbation of the levels, spatial distribution, and subcellular localization of ICSBP, IRF-1, and IRF-2 in the developing mouse lens are strongly correlated with disruption of lens differentiation and development of lens cataracts. Constitutive expression of IRFs, including the lymphoid-specific IRFs, ICSBP, and LSIRF/Pip in the ocular lens, makes a compelling case for IRFs in transcriptional regulation of lens genes. Taken together with our previous finding that aberrant activation of the JAK/STAT signaling pathway can alter the developmental fate of ocular cells, we believe that the IFN-g TR model provides a useful biologic system for understanding competing signaling pathways that influence the development of the vertebrate lens.

Jeffrey Rubin

Jeffrey Rubin received his M.D. and Ph.D. in molecular biology from Washington University (St. Louis) in 1983. Following an internal medicine residency program at The Jewish Hospital of St. Louis, he joined the Laboratory of Cellular and Molecular Biology at the NCI in 1986 as a biotechnology fellow. He is now a senior investigator in the LCMB.

From 1986 to 1996, my research dealt primarily with the purification and biological activities of two heparin-binding mitogens—keratinocyte growth factor (KGF, also known as FGF-7) and hepatocyte growth factor/scatter factor (HGF/SF). These proteins are mediators of mesenchymal-epithelial communication that can stimulate cell migration, differentiation, proliferation, and tissue morphogenesis.

Through collaborative studies, I have explored the role of these factors in development, tissue repair, reproductive tract biology, and neoplasia. We and others have shown that KGF has remarkable cytoprotective effects, consistent with the hypothesis that it functions as a homeostatic factor to maintain epithelial barrier function. This has led to its use in clinical trials to reduce mucositis associated with chemoradiotherapy. My colleagues and I also identified two truncated HGF/SF isoforms, designated HGF/NK1 and HGF/NK2, which bind with high affinity to Met (the HGF/SF tyrosine kinase receptor). We demonstrated that these isoforms act as partial agonists or antagonists of HGF/SF activity. We determined that the amino-terminal domain of HGF/SF retains the heparin-binding properties of the full-length protein, and we established an important role for proteoglycan in HGF/SF isoform signaling.

My ongoing KGF research is collaborative and concerns its potential effects on development and function of the immune system. A collaborative study of HGF/SF will provide a detailed map of its heparin-binding site. This could yield more potent agonists or antagonists with potential clinical uses.

For the past three years, the major focus of my research has been a soluble protein we discovered that has a cysteine-rich domain homologous with the putative Wnt-binding site of Frizzleds, the cell surface Wnt receptors. We have shown that this secreted Frizzled-related protein (sFRP-1) can bind directly to Wnt and modulate its activity. Thus, we believe it regulates Wnt-dependent developmental processes. sFRP-1 might also have an effect on Wnt signaling in neoplasia.

We have generated an abundant source of recombinant sFRP-1 and are currently studying its structure and biological activity. Gene targeting and transfer projects are underway to assess sFRP-1 function in vivo. With the support of an NCI Intramural Research Award, we have begun to screen peptide phage display combinatorial libraries to identify motifs responsible for binding to sFRP-1. Such information could lead to the development of analogs that would modulate Wnt and sFRP-1 activities.

In another series of experiments, we have characterized the promoter region of the human sfrp-1 gene and identified several potential binding sites for transcription factors, including members of the GATA family.

This work adds a new dimension to my research program, which has centered on the discovery and analysis of soluble polypeptide factors involved in the regulation of growth and differentiation. The projects have been highly interactive, involving collaborations on and off the NIH campus, and have the potential to generate reagents of therapeutic relevance.


Return to Table of Contents