Protein Structure and Function: Applications of Bioinformatics Methods - John Rigden 2014

Bioinformatics Methods for Studying the Structure and Function of Disordered Proteins
Limitations of IDP Function Prediction Methods
Sequence-Function Independence and Ambiguity

In line with the aforementioned principles, several recent mutagenesis studies have focused on exploring the unconventional relationship between sequence and function in IDPs. In these experiments, the sequences of functional regions were scrambled and shuffled; remarkably, it was found that Protein Functions are insensitive to randomization. This phenomenon has been termed sequence autonomy (Ross et al. 2005; Tompa and Fuxreiter 2008). These findings highlight just how limited our understanding of the sequence–function paradigm in IDPs truly is.

A classic example is provided by METABOLISM/31.html">Transcription factors, where the acidic trans-activation domain (TAD) of Gen4p can be replaced by random acidic segments without any noticeable loss of biological activity (Hope et al. 1988). This pattern is likely universal, leading to the hypothesis that the assembly of the transcription preinitiation complex may not require the strict geometric complementarity typically needed for specific protein-protein recognition (Sigler 1988). A more recent and highly detailed study reported similar behavior for the chimeric transcription factor EWS fusion protein (EPF) (Ng et al. 2007). In the repeat-rich TAD of EFP, individual repeats can be freely permuted, arranged in random order, or even reversed while fully preserving EFP function.

Sequence autonomy has also been demonstrated for two other systems: linker Histones and Prions. In the case of linker histones, the binding region for the apoptotic nuclease DNA fragmentation factor 40 (DFF40) was investigated. It was established that any sufficiently long segment of the C-terminal domain (CTD) can bind and activate the enzyme, regardless of its primary sequence and position within the intact CTD (Hansen et al. 2006). For the Yeast prions Ure2p and Sup35p, amyloid formation serves as yet another general example of recognition sequence autonomy (Ross et al. 2005). These functional prions likely confer a selective advantage to their host Cells (Wickner et al. 1999). They contain Q/N-rich disordered prion domains that can be thoroughly randomized without losing the prion-like Properties of the protein (Ross et al. 2005).

THE CONCEPT OF fuzziness (Tompa and Fuxreiter 2008) is based, in part, on these findings. Fuzziness represents the inherent disorder of a bound IDP, which is characterized by two potentially interrelated states. In some cases, Molecular recognition and binding to a partner molecule do not entail folding or ordering of the IDP, as observed for the T-Cell receptor $\xi$-chains (Sigalov et al. 2004) and the umuD Gene product (Simon et al. 2008). In other cases, upon binding, the protein does not adopt a single dominant Structure, but instead exists as an ensemble of states, which can be viewed as polymorphism in the bound state. This behavior is exemplified by T-cell factor 4 (Tcf4) upon binding to $\beta$-catenin (Graham et al. 2001) and by the nuclear localization signal (NLS) signaling protein upon binding to $\alpha$-importin (Fontes et al. 2000). Undoubtedly, this phenomenon complicates the Prediction of IDP functions, as it challenges the dogma of a strict one-to-one correspondence between protein sequence and function.