Predicting protein function from structure: Unique structural features of proteases

Abstract

We have noted consistent structural similarities among unrelated proteases. In comparison with other proteins of similar size, proteases have smaller than average surface areas, smaller radii of gyration, and higher C ␣ densities. These findings imply that proteases are, as a group, more tightly packed than other proteins. There are also notable differences in secondary structure content between these two groups of proteins: proteases have fewer helices and more loops. We speculate that both high packing density and low ␣-helical content coevolved in proteases to avoid autolysis. By using the structural parameters that seem to show some separation between proteases and nonproteases, a neural network has been trained to predict protease function with over 86% accuracy. Moreover, it is possible to identify proteases whose folds were not represented during training. Similar structural analyses may be useful for identifying other classes of proteins and may be of great utility for categorizing the flood of structures soon to flow from structural genomics initiatives.T he genome sequencing projects currently underway have given birth to a new pursuit: determining the threedimensional structures of an organism's proteome. This new endeavor, dubbed ''structural genomics,'' has an initial goal of solving the structures of proteins that have little or no sequence identity to proteins of known structure so as to map out protein fold space most efficiently and to provide modeling scaffolds for proteins of biomedical interest (1-3). Structures solved to meet this goal will include proteins of unknown function as has recently been reported (4, 5). Deducing the functions of proteins from their structures would be beneficial, because it could suggest possible roles for a much larger group of homologous proteins from other organisms.Herein, we investigate whether a broad class of proteins of similar function, but not necessarily similar fold or catalytic mechanism, has distinguishing structural characteristics. We focus on the proteases, a very well studied class of proteins. Before the development of recombinant methods for protein expression, digestive enzymes were the subjects of many early structural and mechanistic studies, because they were easy to obtain in large quantities from natural sources (6). Today, the database of protease structures has grown to include a variety of molecules that play critical roles in many biological processes ranging from viral replication to the development and growth of an organism.As with nearly all biological processes, protease activity must be regulated tightly. Regulation is particularly important for proteases, because all proteins, at some level, are their natural substrates. Different mechanisms have arisen for protease regulation. These include inhibition by specific protease inhibitors as well as synthesis as zymogens with covalently attached, inhibitory prosegments (7). Proteases may also be restricted to certain parts of the cell (e.g., the proteasome) or function only under spe...

Cold-Active Winter Rye Glucanases with Ice-Binding Capacity

“…Indeed, by using Swiss PDB Viewer (Guex and Peitsch, 1997), the calculated loop (random coil) content percentage of the ScGLU-3 template was 48%, which suggests a high degree of fold flexibility. This value is significantly higher than the average loop content (37.6%) of 154 nonprotease proteins reported previously (Stawiski et al, 2000).…”

Sequence and Structural Features of Enzymes and their Active Sites by EC Class

“…From a study of structural features of the proteases, Stawiski et al found that they exhibit similar characteristics such as smaller than average surface areas and higher C α densities, regardless of whether or not they were evolutionarily related. 13 They also showed different secondary-structure content relative to the non-proteases. By using these features in a machine-learning approach, they were able to define a set of structural classifiers that could predict whether a protein is a protease or a nonprotease with an accuracy of over 86%.…”

Predicting Enzyme Class From Protein Structure Without Alignments

“…† Another surface-based attribute is the fractal dimension. This is calculated as by Stawiski et al, 29 by calculating the gradient of the log-log plot of probe radius against molecular surface (calculated by MSMS 45 ). Fractal dimension might be thought of as representing the "crinkliness" of the protein's surface.…”

“…Our method falls into this category as it characterises the simple structural properties of each class of enzymes at the top level of the EC scheme. Our initial work was based on the observations made by Stawiski and co-workers, 29 who noted broad structural differences between proteases and non-proteases. These included such properties as high C a density and lower than average surface areas in proteases when compared to non-proteases.…”