Robert Fox - US grants

The mechanism by which protein molecules are localized in various compartments and membranes of the cell has stood as one of the great problems of cell biology. It is now clear that nascent polypeptide chains are carried into the lumen of that organelle where the leader (signal) sequence is removed. Completed polypeptides travel to the Golgi body where they are targeted to specific organelles. Plasma membrane receptors such as the low density lipoprotein (LDL) receptor are cycled from the plasma membrane through coated endocytic vesicles and endosomes, delivering LDL particles to the lysosomes while the receptors return to the cell surface. In humans, errors in the targeting, processing and recycling of LDL receptors lead to familial hypercholesterolemia resulting in premature atherosclerosis. The general goals of this research program are to probe the physical and chemical basis by which membrane proteins attain and maintain defined three-demensional structures in biological membranes, the mechanism by which such proteins carry out their biological function and the role of electric fields in these processes. The best starting point for such investigations is a detailed three-dimensional structure derived from diffraction studies at a resolution consistent with atomic modeling. The specific aims of this research program are directed at two aspects of these general problems. First, the closely related membrane active peptides: alamethicin and suzukacillin will be studied as examples of proteins which can spontaneously integrate into lipid bilayers from aqueous solution. We plan to characterize the structure of these peptides in crystals grown from aqueous solution to define the molecular surface which interacts with the lipid bilayer. EM image reconstruction techniques will be applied to two dimensional arrays of gated and ungated peptide channels to define the structural basis of the voltage-gating process in these systems. Secondly, we plan to prepare three-dmensional crystals of E. coli leader peptidase, and to characterize its molecular structure at high resolution in an effort to define the structural basis of leader peptide recognition and cleavage, and the role of this process in membrane protein biogenesis.

Immunolglobulin molecules have become valuable reagents in research and diagnostic laboratories and have been tested as clinical reagents in the treatment of disease. Monoclonal antibodies directed to a linear epitope in a globular protein molecule can sometimes be generated by immunizing with linear peptide sequences derived from the native protein. When these antibodies are identified they are generally of low titer and thus not of great utility in the research laboratory or clinic. This is particularly true when this approach has been applied to molecules of the immunoglobulin superfamily. The limited success in generating antipeptide antibodies with a high affinity for the folded protein from which the sequence was derived may arise because the linear peptide conformation does not resemble the structure found in the folded protein. Two methods are proposed to constrain the conformation of linear epitopes in model immunogens. Sequences which occur on the surface of a parent protein in a Beta-turn conformation will be introduced into a cyclic peptide structure which should promote a Beta-turn conformation. These sequences will also be substitited for a Beta-turn region of staphylococcal nuclease at the gene level to produce a hybrid protein immunogen. The affinity of polyclonal and monoclonal antibodies, elicited by the model immunogens, for the parent protein will be compared with antibodies raised against similar linear peptides. The conformation of the cyclic peptides and hybrid proteins will be characterized by NMR spectroscopy and x-ray crystallagrophy to allow a detailed molecular interpretation of the immunological data.

Globular protein molecules are largely composed of beta-sheet and alpha- helical secondary structures separated by sharp changes in chain trajectory at four residue beta-turns or larger loops on the surface of the molecule. These betaturn and loop structures can often be recognized in protein sequences, aiding in secondary structure prediction and in the identification of potential linear peptide epitopes for antibody production. While antipeptide antibodies which recognize a full protein chain have been prepared in a number of cases, the resulting monoclonal antibodies often have a low affinity for the folded protein from which the synthetic peptide immunogen sequence was derived. A protocol was proposed to use a second protein (staphylococcal nuclease) as a "host" to constrain the "guest" peptide immunogen into a native structure by incorporating that sequence into a hybrid protein at an appropriate site. Work will continue on two hybrid protein systems to develop this method and to understand its success or failure in structural and physical terms. Results during the last three years indicate that there is a strong relationship between the sequence and type of a beta-turn which can be dominant over globular protein context effects. Investigations will continue to define the relationship between amino acid sequence and beta-turn type. These experiments should provide an improvement in our ability to predict betaturn sites in globular proteins and should define protein engineering design principles for the hybrid proteins above and other new protein molecules in general. NMR and x-ray crystallography experiments will continue in an effort to define the physical basis by which a cis peptide bond is favored in a type VI beta-turn of staphylococcal nuclease. This system provides an opportunity to observe the influence of amino acid sequence on the equilibrium between two beta-turn types on the surface of a globular protein. A genetic and crystallographic analysis will continue to define sequences consistent with a type I' beta-turn in nuclease, and the physical and structural basis for these sequences preferences.

DESCRIPTION: The overall goal of this proposal is the structure determination of recombinant human a1 glycine receptor chloride channel. The glycine receptor is part of a large family of ligand-gated channels that mediate signal transduction at the synapse. The structure of one member of this family, the acetylcholine receptor, has been studied in three dimensions by electron microscopy and a low resolution structure is available, however the sequence has not been mapped into this structure. Experiments are aimed at distinguishing plausible topological models for this family of channels by labeling cysteine residues engineered into the glycine receptor with chemical cleavage reagents. A major goal of the proposal is to crystallize purified holoreceptor as well as its extracellular, ligand binding domain. Functional glycine receptor composed of a single subunit type can be reconstituted from recombinant protein, and is produced in the co-P.I's lab in large enough yields for crystallization trials.

While may small proteins can refold spontaneously in vitro, the identification of the folding pathway, and the detection and structural characterization of folding intermediates have been difficult. Recently, attention has turned to the molten globule state and other nonnative equilibrium states of proteins which are thought to be models for kinetic intermediates in protein folding. Fragments of staphylococcal nuclease have been produced which also have properties similar to those of the molten globule, i.e. a somewhat compact structure with some secondary structure but without a defined tertiary structure. Pulsed hydrogen-deuterium exchange during refolding has been used to probe the protection of backbone amide hydrogens from solvent exchange during refolding of a number of proteins and most recently the staphylococcal nuclease Pro 117 yields Gly variant. The extent of exchange for 39 residues is determined by two-dimensional proton NMR after refolding for 5 ms to 10s. Three kinetic phases are inferred. Modest protection of amides in the early refolding intermediate composed to two beta-sheets formed by local sequence interactions is observed after a 5 ms refolding period. Native levels of protection throughout the molecule accrue more slowly in tow kinetic phases (k approximately 2s-1, k less than 0.01s-1). Protection factors were determined by varying the high pH labeling pulse after refolding for 100 ms. Little or no native or unfolded protein is present; instead, most molecules are in one or more partially folded states. The intermediate state has modest, yet significant, protection for residues in the beta-sheets (protection factors 10-60), and almost no protection in the alpha- helices (protection factors less than 10). The pattern of labeling is consistent with a role for beta turns and beta-hairpins in the formation of the early intermediate. Recently a chemical cleavage method has been developed where an EDT-Fe based reagent (EPD-Fe) can be attached to a protein via a cysteine side chain. The addition of ascorbate generates hydroxyl radicals at the iron center which diffuse and cleave the polypeptide backbone in a region close to the cysteine attachment site at residues accessible to solvent. The observed cleavage sites can be mapped by amino acid sequencing. The cleavage is dependent on protein conformation. We propose characterize the molten globule state of apomyoglobin and staphylococcal nuclease fragment structures using this newly developed chemical cleavage technique. We will prepare a number of cysteine variants of these proteins and characterize the cleavage patterns observed in the native and molten globule states.

DESCRIPTION (provided by applicant): Our previous studies have provided a proof-of-principle that a small disulfide-rich miniprotein can be developed that will block the infection of cells by Langat (LGT) virus, a naturally attenuated virus that is model for the pathogenic members of the tick-borne encephalitis (TBE) serogroup of the Flavivirus genus. A first generation miniprotein, termed MP-100, was selected by panning a conformationally restrained combinatorial miniprotein phage display library for binding with purified recombinant domain III (D3) of the LGT virus envelope (E) protein. The miniprotein MP-100 was shown to block infection of Vero and LLC-MK2 monkey kidney cell cultures by tick-borne LGT and Powassan viruses. Further studies indicated an antiviral effect in a mouse animal model. Our objective during this period of support is the development of a second-generation, more tightly binding miniprotein with improved antiviral activity against TBE serogroup flaviviruses compared to the current MP-100 sequence. Our goal is to develop an antiviral miniprotein that is effective against a broad range of potential flavivirus bioterrorist threat agents in the TBE serogroup, including Central European tick-borne encephalitis (strain Kumlinge), Kyasanur Forest Disease (KFD), Omsk Hemorrhagic Fever (OHF) and Russian Spring Summer encephalitis (RSSE) viruses. We will identify optimized tight-binding analogs of MP-100 to Kum-E-D3 and the related TBE serogroup E-D3s, and determine if they have enhanced antiviral activity in tissue culture cells and in animals. The development of anti-TBE virus miniproteins will serve as a model for the development of miniproteins against other flaviviruses that are also potential bioterrorist threat agents or emerging diseases, including dengue, Japanese encephalitis and West Nile.