The Protein Society

Thirteenth Symposium

Boston, MA

July 24-28, 1999

 

 

NMR as a tool for discovering high affinity ligands -- Steve Fesik

Extension of SAR by NMR technology:

Using this method small molecules that bind to proximal subsites of a protein are identified, optimized, and linked together based on NMR-derived information. Four different approaches were described in the talk --

1. Combination: Combine two molecules binding to adjacent sites by a linker.

2. Merging: Merge two molecules binding to sites that are overlapping to produce a molecule that binds tighter.

3. Directed libraries: Use a tight binder as a starting point to design and synthesize a family of molecules by combinatorial chemistry, that will bind tighter.

4. Fragment optimization: Similar to #3, but, does not involve combinatorial chemistry. Traditional SAR.

Cryoprobe allows use of lower concentrations. (good spectrum at ~20 micromole concentration of protein within 15 minutes) He envisions screening about 200,000 compunds per month.

 

 

New chemical approaches to tracing cellular signal transduction cascades -- K.M. Shokat, Princeton Univ.

2% of all human proteins are kinases. Gene knockouts of kinases often have no phenotype because of compensation by related kinases. It is difficult to identify role of specific kinases in particular signal transduction pathway. Two examples were described to do that:

1. ATP binding sites of kinases were mutated so that they can uniquely accept non-natural ATP molecules (e.g. N6-benzyl ATP). This makes the kinases traceable and allows identification of specific substrates for the kinases and their role in signal transduction pathways.

2. ATP binding site of v-Src was mutated and used in an assay to screen compunds. The molecules that bound to the mutated kinase were also found to bind other similarly mutated mutated kinases. This result suggests that many kinase families can be easily targeted for selective inhibition and can be used in understanding the signal transduction pathways as well as for designing inhibitors as drugs.

 

 

Structural genomics of a microbe: A test case -- S-H. Kim, UC-Berkeley

We are moving from sequence genomics to structural genomics to identify functions of proteins. Molecular functions (Physics & Chemistry) & Cellular function (i.e. network of molecular functions).

 

Out of 103-105 genes,

20-70% ORFs are either unknown or hypothetical genes.

25-35% ORFs are for membrane proteins.

Only ~30% of all possible protein folds are known.

~8-20% of functional annotations in the protein database need correction.

 

From M. janaschii genome,

25 genes were cloned.

20 proteins were soluble, 5 insoluble

13 were purified

11 crystallized

5 structures solved.

 

MJ0577 (18.7 kD):

Unknown function

Structure shows a dimer

ATP was bound

=> it was inferred that this protein is a molecular switch which hydrolyzes ATP.

 

MJ0226:

Unknown function

Forms a dimer

A novel fold

Looked like RNA/DNA/nucleotide binding protein

Experiments showed that it hydrolyzed non-standard nucleotide triphosphates.

A yeast analog with same function was identified.

Still needs more testing.

 

Small heat shock protein:

About 12-43 kD. Forms large oligomers (200-800 kD).

Homologous to eye a-crystalline. (Mutations cause cataract).

Structure shows a 24mer homopolymer. It forms a hollow sphere with 8 triangular and 6 square shaped windows.

It may coat certain proteins while the cell is under stress. Windows allow the substrates and products to pass through.

 

 

Structure and function of the nickel enzyme methyl-coenzyme M reductase -- R.K. Thauer, Max Planck Institute.

Methyl-coenzyme M reductase catalyzes the final reaction step in formation of methane in methanogenic archaea.

This is a 300 kD yellow protein that consists of 3 subunits in an a2b2g2 arrangement. Also contains two molecules of nickel pophinoid coenzyme F430 that are tightly bound.

1.45A structure shows that the a subunit contains 5 modified amino acids at or very near the active site. 1-N-methylhistidine, 5-(S)-methylarginine, 2-methylglutamine, S-methylcysteine, & thioglycine. Chemical analysis have confirmed these unusual amino acids, and they are biosynthetically derived from methyl group of methionine via S-adenosylmethionine.

 

 

New functions from old scffolds: Understanding enzyme chemistry from superfamily analysis -- P.C. Babbitt, UCSF

Analysis of several different protein superfamilies have shown that the common fold unique to each is associated with a specific functional step. E.g. members of enolase superfamily are found in nature and perform many different metabolic functions using widely diverse substrates and overall chemical reactions. But, all these enzymes carry out the same basic chemical step, i.e. abstraction of alpha protons of carboxylic acid followed by formation of an enolate intermediate.

 

This kind of information can be used to identify functions of new proteins generated from the genome projects.

 

 

Structural analysis of signaling by SH2 and SH3 domains -- J.Kuriyan, Rockefeller Univ.

SH2 binds to phosphotyrosine

SH3 binds to polyproline helices (PRO-X-X-PRO motifs)

 

He showed beautiful structural evidence suggesting significant structural changes caused by autophosphorylation of Src protein and binding of Src SH2 domain binding to it. This acts as a controlling switch to turn on and turn off the kinase activity of Src.

 

 

Studies of protein free energies and folding using a combination of molecular dynamics -- P. Kollman, UCSF.

Molecular dynamics simulations using parallel computing techniques has enabled them to carry out simulations on a small protein in water for 1 microsecond. They studied folding of a small 36 residue protein called villin. The structure of this protein is known and it contains 3 helices. They could fold the protein from unfolded to a marginally stable state (150 ns lifetime) during the simulation. This structure has a favorable solvation free energy and resembles the actual structure of the protein (6-8 A RMSD).

 

They observe that burying of the hydrophobic surfaces dominated the early phase of the folding process and appeared to be the driving force behind the reduction of the radius of gyration.

 

They also studied folding of some mutants of the protein. They could replace a charged residue with Ala without much perturbation in the free energy of the system for small peptides.

 

 

Progress in Ab Initio protein structure prediction -- D. Baker, Univ. of Washington.

He talked about their results from CASP3. Also, about their effort in predicting structures of all the proteins with known structure under 120 amino acids. They could predict native like structure in about 80% of these proteins.

 

 

Protein interactions from genome sequence -- D. Eisenberg, UCLA

Function from sequence by non-homology method by inferring from the protein-protein interactions. New computational approaches for inferring protein interactions from genome sequences:

1. Phylogenetic profiling -- proteins that function together in a pathway are likely to evolve together in a correlated fashion. Phylogenetic profile is a string of letters representing presence or absence of a particular protein in all the known genomes. Proteins with similar profiles tend to be functionally linked.

2. Rosetta stone method -- it detects interactions by fusion of interacting proteins into a single polypeptide chain.

 

 

Thermodynamics and dynamics in RNA:Protein recognition -- K. Hall, Washington Univ.

RBD1 and RBD2 of U1A are similar in structure and sequence, but, RBD2 does not bind to RNA.

 

RBD1-U1A stemloop bind with kD of 2 x 10-11 M in 200 mM NaCl, 1mM MgCl2, pH 7.0 buffer at 22oC

 

DCp (binding) is -3 Kcal/mol.

 

 

Assembly of RNA and protein components into the spliceosomal snRNPs -- K. Nagai, MRC

Spliceosome consists of 4 large RNA-protein complexes: U1, U2, U4/U6, U5.

Spliceosomal proteins are divided into 2 categories: Sm or core proteins, that are common to all the snRNPs (B/B', D1, D2, D3, E,F, and G); and those specific only to one of the snRNPs.

 

Structures of D3B and D1D2 complexes have been solved. These proteins have a common fold comprising of a N-terminal helix and a strongly bent 5 stranded antiparallel b sheet. The structures suggest that the core proteins form a ring-like structure and the snRNAs may bind in the positively charged central hole.

 

 

Biophysical and cellular studies suggest a small molecule therapeutic strategy for intervention in amyloid diseases -- J.W. Kelly, Scripps.

Transthyretin is converted to an insoluble fibrilar structure, referred to as amyloid. There are 2 types of these human diseases: senile systemic amyloidosis, and familial amyloid polyneuropathy. In the senile type of the disease, the symptoms like heart dysfunction and deposition of the fibrils in joints occur in patients over 80 years of age. In the familial type these symptoms occur at a younger age because of the mutations in transthyretin.

 

Transthyretin is a 127 a.a. protein that forms a homotetramer. It is present in plasma as well as CSF. It functions as a thyroxine carrier. This protein is usually degraded in lysosomes like other proteins. It can undergo conformational change and form amyloid fibrils instead. Mutated transthyretin has a higher propensity towards forming the amyloid fibrils.

 

They have identified certain small molecules that bind to transthyretin and prevent it from undergoing the conformational change to form the fibrils. They have also identified other small molecules that bind to the fibrils and dissolve them so that they can be degraded properly.

 

 

Studies on the structure and function of ribosomes -- H. Noller, UCSC

They are working on the whole 70S ribosome particle bound with mRNA and tRNA. They have nice large crystals that diffract well and could get a structure with about 7-8 A resolution.

 

They have combined this with information obtained using biochemical probing methods such as the use of directed hydroxyl radical probing. This allowed them to get extensive information about the three dimensional folding of rRNA, positions of tRNA, translation factors and other ribosomal proteins.

 

 

Structural principles of ribosome architecture from a 4.5 A resolution map of the 50S subunit. -- Tom Steitz, Yale Univ.

Similar to the talk from last years NJ shore meeting. He talked about extending the phasing to allow a 3A structure. That work is still in progress.