The FEATURE framework for protein function annotation: modeling new functions, improving performance, and extending to novel applications.

Structural genomics efforts contribute new protein structures that often lack significant sequence and fold similarity to known proteins. Traditional sequence and structure-based methods may not be sufficient to annotate the molecular functions of these structures. Techniques that combine structural and functional modeling can be valuable for functional annotation. FEATURE is a flexible framework for modeling and recognition of functional sites in macromolecular structures. Here, we present an overview of the main components of the FEATURE framework, and describe the recent developments in its use. These include automating training sets selection to increase functional coverage, coupling FEATURE to structural diversity generating methods such as molecular dynamics simulations and loop modeling methods to improve performance, and using FEATURE in large-scale modeling and structure determination efforts.

Improving structure-based function prediction using molecular dynamics.

The number of molecules with solved three-dimensional structure but unknown function is increasing rapidly. Particularly problematic are novel folds with little detectable similarity to molecules of known function. Experimental assays can determine the functions of such molecules, but are time-consuming and expensive. Computational approaches can identify potential functional sites; however, these approaches generally rely on single static structures and do not use information about dynamics. In fact, structural dynamics can enhance function prediction: we coupled molecular dynamics simulations with structure-based function prediction algorithms that identify Ca(2+) binding sites. When applied to 11 challenging proteins, both methods showed substantial improvement in performance, revealing 22 more sites in one case and 12 more in the other, with a modest increase in apparent false positives. Thus, we show that treating molecules as dynamic entities improves the performance of structure-based function prediction methods.

Combining molecular dynamics and machine learning to improve protein function recognition.

As structural genomics efforts succeed in solving protein structures with novel folds, the number of proteins with known structures but unknown functions increases. Although experimental assays can determine the functions of some of these molecules, they can be expensive and time consuming. Computational approaches can assist in identifying potential functions of these molecules. Possible functions can be predicted based on sequence similarity, genomic context, expression patterns, structure similarity, and combinations of these. We investigated whether simulations of protein dynamics can expose functional sites that are not apparent to the structure-based function prediction methods in static crystal structures. Focusing on Ca2+ binding, we coupled a machine learning tool that recognizes functional sites, FEATURE, with Molecular Dynamics (MD) simulations. Treating molecules as dynamic entities can improve the ability of structure-based function prediction methods to annotate possible functional sites.

A two-hybrid screen identifies cathepsins B and L as uncoating factors for adeno-associated virus 2 and 8.

Vectors based on different serotypes of adeno-associated virus hold great promise for human gene therapy, based on their unique tissue tropisms and distinct immunological profiles. A particularly interesting candidate is AAV8, which can efficiently and rapidly transduce a wide range of tissues in vivo. To further unravel the mechanisms behind AAV8 transduction, we used yeast two-hybrid analyses to screen a mouse liver complementary DNA library for cellular proteins capable of interacting with the viral capsid proteins. In total, we recovered approximately 700 clones, comprising over 300 independent genes. Sequence analyses revealed multiple hits for over 100 genes, including two encoding the endosomal cysteine proteases cathepsins B and L. Notably, these two proteases also physically interacted with the corresponding portion of the AAV2 capsid in yeast, but not with AAV5. We demonstrate that cathepsins B and L are essential for efficient AAV2- and AAV8-mediated transduction of mammalian cells, and document the ability of purified cathepsin B and L proteins to bind and cleave intact AAV2 and AAV8 particles in vitro. These data suggest that cathepsin-mediated cleavage could prime AAV capsids for subsequent nuclear uncoating, and indicate that analysis of additional genes recovered in our screen may help to further elucidate the mechanisms behind transduction by AAV8 and related serotypes.