Protein, disordered

Intrinsic disorder might not be encoded by the sequence, but rather might be the result of the absence of suitable tertiary interactions. If this were the general cause of intrinsic disorder, any subset of ordered sequences and any subset of disordered sequences would likely be the same within the statistical uncertainty of the sampling. On the other hand, if intrinsic disorder were encoded by the amino acid sequence, any subset of disordered sequences would likely differ significantly from samples of ordered protein sequences. Thus, to test the hypothesis that disorder is encoded by the sequence, we collected examples of intrinsically ordered and intrinsically disordered proteins, then determined whether and how their sequences were distinguishable. 

To obtain statistically significant comparisons of ordered and disordered sequences, much larger datasets were needed. To this end, disordered regions of proteins or wholly disordered proteins were identified by literature searches to find examples with structural characterizations that employed one or more of the following methods (1) X-ray crystallography, where absence of coordinates indicates a region of disorder (2) nuclear magnetic resonance (NMR), where several different features of the NMR spectra have been used to identify disorder and (3) circular dichroism (CD) spectroscopy, where whole-protein disorder is identified by a random coil-type CD spectrum. 

Number of Proteins and Residues in Databases of Intrinsically Disordered Protein Characterized by Various Methods... 

Three groups of disordered proteins have been assembled, with the groups defined by the experimental method used to characterize the lack of ordered structure. Because the focus has been on long regions of disorder, an identified disordered protein or region was not included in these groups if it failed to contain 40 or more consecutive residues. Disordered regions from otherwise ordered proteins as well as wholly disordered proteins were identified. Table I summarizes the collection of sequences in this database. 

Although the disordered proteins were characterized by three different methods, all three datasets of disordered amino acids showed semi-quantitatively similar changes for 16 of the 20 amino acids (Fig. 1, top). Because the three methods rely on completely different underlying biophysical principles for determining disorder, substantial compositional differences among the datasets were expected surprisingly, however, such differences were not observed. Thus, the compositions of the disordered proteins (Fig. 1, top) likely indicate inherent tendencies of this type of protein. 

Fig. 1. Comparisons of amino add compositions of ordered protein and disordered protein. (Top) Amino acid compositions of three disordered datasets. (Middle) Amino acid compositions of three ordered datasets. (Bottom) Compositions of disordered datasets relative to the Globular 3-D dataset (from Romero et al., Proteins Struct., Fund., Gen. 42, 38-48, copyright 2001. Reprinted by permission of Wiley-Liss, Inc., a subsidiary of John Wiley Sons, Inc.). The ordinates are (% amino acid in disordered dataset — % amino acid in Globular 3-D)/(% amino acid in Globular 3-D) = (D —0)/0. Negative values indicate that the disordered database has less than the ordered, positive indicates more than the ordered. Error bars are one standard deviation.

This procedure led to a substitution matrix for aligning disordered protein that was different from the commonly used substitution matrices, such as BLOSUM62 (Fig. 4). The matrix for disordered protein is generally better than order-based matrices for aligning disordered proteins whose sequence identities are between 20 and 50%. These results indicate that disordered and ordered protein can be distinguished by their patterns of evolutionary change. 

Fig. 4. Substitution matrix based on disordered protein families. Below the diagonal are the scores for each amino acid substitution. Above the diagonal are the differences between BLOSUM 62 and the disorder matrix. On the diagonal are the scores/differences. (From Radivojac et al., 2002, PSB 2002 7, 589-600, with permission of World Scientific Publishing Co. Pte Ltd.)...

The success of the initial PONDRs based on small databases of disordered protein motivated attempts to improve predictor accuracy. The main limitation for such attempts has been and continues to be the lack of low-noise structural data for both ordered and disordered protein, where noise means ordered regions misclassified as disordered and vice versa. 

The accuracies of the various PONDRs were estimated (Table IV) by applying them to the ordered sequences in 0 PDBS25 as summarized in Table II and to the merged set of disordered proteins described in Table I. Overall, the prediction accuracy of each PONDR was much better on the 222,116 ordered residues of 0 PDBS25 than on the 18,833 residues of the merged disorder set. Thus, prediction of order generalized much better than prediction of disorder. 

III. PONDR Estimations of the Commonness of Intrinsically Disordered Proteins... 

We are attempting to understand the biological significance of the large variations in frequency of putative LDRs, whether between different types of bacteria or archaea, or between pro- and eukaryota. We have carefully studied the literature of more than 90 example proteins selected from our disordered protein databases and found reports on the functions of most of the disordered regions (Dunker et al., 2002). The observed functions and the number of examples in each functional class are given in Table VI. As indicated, four major functional classes were found molecular recognition, molecular assembly or disassembly, protein modification, and entropic chains. 

Recently, there has been great interest in proteins that exhibit biological activity but lack a well-defined secondary or tertiary structure after purification (Dunker et al., 1998, 2001 Schweers et al., 1994 Uversky et al., 2000 Wright and Dyson, 1999). Such proteins are referred to as intrinsically disordered or unstructured. An analysis in 1998 of the Swiss Protein Database revealed that about 15,000 proteins in that database are likely to contain disordered segments at least 40 residues in length (Romero et al., 1998). Dyson and Wright (2002) review intrinsically disordered proteins in this volume. 

There are now numerous examples of proteins that are unstructured or only partially structured under physiological conditions yet are nevertheless functional (Dunker and Obradovic, 2001 Wright and Dyson, 1999). In many cases, such intrinsically disordered proteins adopt folded structures upon binding to their biological targets. As the proteins that constitute the transcriptional machinery have become... 

---