Two distinct Raf domains mediate interaction with Ras

A key event for Ras transformation involves the direct physical association between Ras and the Raf-1 kinase. This interaction promotes both Raf translocation to the plasma membrane and activation of Raf kinase activity. Although substantial experimental evidence has demonstrated that Raf residues 51-131 alone are sufficient for Ras binding, conflicting observations have suggested that the Raf cysteine-rich domain (residues 139-184) may also be important for interaction with Ras. To clarify the role of the Raf cysteine-rich domain in Ras-Raf binding, we have compared the ability of two distinct Raf fragments to interact with Ras using both in vitro Ras binding and in vivo Ras inhibition assays. First, we determined that both Raf sequences 2-140 and 139-186 (designated Raf-Cys) showed preferential binding to active, GTP-bound Ras in vitro. Second, we observed that Raf-Cys antagonized oncogenic Ras(Q61L)-mediated transactivation of Ras-responsive elements and focus-forming activity in NIH 3T3 cells and insulin-induced germinal vesicle breakdown in Xenopus laevis oocytes in vivo. This inhibitory activity suggests that Raf-Cys can interact with Ras in vivo. Taken together, these results suggest that Ras interaction with two distinct domains of Raf-1 may be important in Ras-mediated activation of Raf kinase activity.

Biological and structural characterization of a Ras transforming mutation at the phenylalanine-156 residue, which is conserved in all members of the Ras superfamily

Although Ras residue phenylalanine-156 (F156) is strictly conserved in all members of the Ras superfamily of proteins, it is located outside of the consensus GDP/GTP-binding pocket. Its location within the hydrophobic core of Ras suggests that its strict conservation reflects a crucial role in structural stability. However, mutation of the equivalent residue (F157L) in the Drosophila Ras-related protein Rap results in a gain-of-function phenotype, suggesting an alternative role for this residue. Therefore, we have introduced an F156L mutation into Ras to evaluate the role of this residue in Ras structure and function. Whereas introduction of this mutation activated the transforming potential of wild-type Ras, it did not impair that of oncogenic Ras. Further, Ras (156L) exhibited an extremely rapid off rate for bound GDP/GTP in vitro and showed increased levels of Ras.GTP in vivo. To determine the structural basis for these altered properties, we used high-resolution nuclear magnetic resonance spectroscopy. The F156L mutation caused loss of contact with residues 6, 23, 55, and 79, resulting in disruption of secondary structure in alpha-helix 1 and in beta-sheets 1-5. These major structural changes contrast with the isolated alterations induced by oncogenic mutation (residues 12 or 61) that perturb GTPase activity, and instead, weaken Ras contacts with Mg2+ and its guanine nucleotide substrate and result in increased rates of GDP/GTP dissociation. Altogether, these observations demonstrate the essential role of this conserved residue in Ras structure and its function as a regulated GDP/GTP switch.

Biomolecular applications of heteronuclear multidimensional NMR

Recent advances in stable-isotope enrichment and heteronuclear multidimensional NMR techniques have transformed NMR into a more powerful and versatile method for the structural and dynamic characterization of biomolecules. Current efforts still focus on improving the methodology to increase the number of systems amenable to NMR analysis, to generate better structures and to investigate biomolecular motions in solution.

Isolation and refolding of H-ras from inclusion bodies of Escherichia coli: refold procedure and comparison of refolded and soluble H-ras

A refolding and purification method for economically producing large quantities of H-ras isolated from Escherichia coli inclusion bodies is described. Experiments were performed to optimize the yield of refolded H-ras for structural analysis by NMR spectroscopy. Protein concentration, temperature, and the presence of 10% glycerol during refolding were varied. The yield of H-ras was highest when the protein was refolded at concentrations less than or equal to 0.1 mg/ml and was independent of the presence of 10% glycerol. The yield was slightly higher at 4 degrees C than at 25 degrees C. The refolded H-ras was purified by anion exchange chromatography to yield protein with a purity of > 90%, as determined by C4 reverse-phase HPLC, and a GDP-binding stoichiometry greater than 0.9. NMR structural analysis of refolded and soluble H-ras was conducted using [15N]glycine- and [15N]serine-enriched protein. The NMR data indicate that the refolded ras protein is structurally similar to ras isolated from the soluble fraction.

Identification of residues critical for Ras(17N) growth-inhibitory phenotype and for Ras interaction with guanine nucleotide exchange factors

The Ras(17N) dominant negative antagonizes endogenous Ras function by forming stable, inactive complexes with Ras guanine nucleotide exchange factors (GEFs; e.g., SOS1). We have used the growth-inhibitory phenotype of Ras(17N) to characterize two aspects of Ras interaction with GEFs. First, we used a nonprenylated version of Ras(17N), designated Ras(17N/186S), which no longer associates with the plasma membrane and lacks the growth-inhibitory phenotype, to address the importance of Ras subcellular location and posttranslational modification for its interaction with GEFs. We observed that addition of an N-terminal myristylation signal to Ras(17N/186S) restored the growth-inhibitory activity of nonprenylated Ras(17N). Thus, membrane association, rather than prenylation, is critical for Ras interaction with Ras GEFs. Second, we used a biological selection approach to identify Ras residues which are critical for Ras(17N) growth inhibition and hence for interaction with Ras GEFs. We identified mutations at residues 75, 76, and 78 that abolished the growth-inhibitory activity of Ras(17N). Since GEF interaction is dispensable for oncogenic but not normal Ras function, our demonstration that single-amino-acid substitutions at these three positions impaired the transforming activity of normal but not oncogenic Ras provides further support for the role of these residues in Ras-GEF interactions. Finally, Ras(WT) proteins with mutations at these residues were no longer activated by mammalian SOS1. Altogether, these results suggest that the Ras intracellular location and Ras residues 75 to 78 are critical for Ras-GEF interaction.

Secondary structure and signal assignments of human-immunodeficiency-virus-1 protease complexed to a novel, structure-based inhibitor

We report comprehensive NMR studies in solution of the human-immunodeficiency-virus (HIV)-1 protease. Stable solutions of the protease were obtained by complexing the protein to a designed cyclic urea inhibitor DMP 323. A variety of triple-resonance experiments provided essentially complete 1H, 13C and 15N NMR signal assignments of the protease. These assignments, together with short-range NOE constraints, coupling constants and hydrogen-exchange data, yielded the secondary structure of the protease in solution. The results reported herein open the way to the determination of the high-resolution three-dimensional solution structures of protease/inhibitor complexes, as well as to studies of protease dynamics and solvent interactions.

Structural and dynamic differences between normal and transforming N-ras gene products: a 31P and isotope-edited 1H NMR study

[15N]Glycine was biosynthetically incorporated into normal cellular N-ras p21 and a position 12 transforming mutant, in order to produce p21 proteins containing several site-specific NMR probes at or near activating positions in the guanine nucleotide binding domain. We have previously assigned all five glycine resonances located in loops directly involved in binding of guanosine diphosphate in the wild-type p21 protein [Campbell-Burk, S., Papastavros, M. Z., McCormick, F., & Redfield, A. G. (1989) Proc. Natl. Acad. Sci. U.S.A 86, 817-820]. In this report, the corresponding glycine resonances in the p21 mutant have been assigned, and spectral differences between normal and mutant p21-guanosine diphosphate (p21.GDP) complexes have been investigated. Our combined 1H[15N] and 31P NMR results show that substitution of aspartate for glycine-12 produces perturbations in the phosphoryl binding domain, near the point of the mutation. Although many of the remaining glycines were unaffected, spectral differences were also observed outside the GDP binding domain. Two of the five active-site glycines in wild-type p21.GDP have very slow amide proton exchange rates with water (kappa less than 2.8 x 10(-5) s-1). The active-site glycines are located in solvent-exposed loops, so their apparent solvent inaccessibility may result from strong hydrogen bond formation between glycine amide protons and bound guanine diphosphate and/or other nearby groups in p21.