Solution structure of the human Tax-interacting protein-1.

PDZ (PSD95/DLG/ZO-1) domains are protein–protein interaction modules composed of 90–100 amino acids. PDZ-containing proteins function in the organization of multiprotein complexes at specific cellular sites and control the spatial and temporal fidelity of intracellular signaling pathways. The PDZ fold comprises a compact six-stranded b-barrel (b1–b6) capped by two a-helices (a1–a2) (Doyle et al. 1996). Target protein recognition is accomplished via a long and deep groove formed between b2 and a2. The C terminus of target proteins inserts into this groove and augments the b-sheet structure of the domain. Certain PDZ domains interact with internal sequence motifs, whereas others bind lipids. Previous studies have established three PDZ classes based on their consensus recognition sequence. Class I prefer X-S/T-X-U-COOH, class II interact with X-U-X-U-COOH, and class III recognize X-D/E-X-U-COOH, where X is any amino acid and U is a hydrophobic residue. However, many PDZ domains bind promiscuously to targets from different classes, underscoring the need for a new classification scheme (Birrane et al. 2003). In general, PDZ proteins possess multiple domains facilitating their functions as protein–protein interaction modules in a variety of cellular contexts. The human Taxinteracting protein-1 (TIP-1) is an unusual PDZ protein predicted to be comprised entirely of a single PDZ domain. TIP-1 was initially identified as an interaction partner of the human T-lymphotropic virus type 1 (HTLV-1) Tax protein, a key viral regulatory protein involved in deregulating numerous cellular signal transduction pathways (Rousset et al. 1998). TIP-1 was also independently identified as a protein that associates with glutaminase L and was termed glutaminase-interacting protein (Olalla et al. 2001). TIP-1 participates in Rho signaling through its interaction with Rhotekin (Reynaud et al. 2000). In addition, TIP-1 binds with high affinity to the C terminus of b-catenin and inhibits its transcriptional activity (Kanamori et al. 2003). TIP-1 also interacts with the C-terminal tail of the Kir 2.3 potassium channel and functions as a negative regulator of its surface expression (Alewine et al. 2006). As a first step towards understanding the structural basis of TIP-1 interaction with cellular and viral proteins, we determined the solution structure of TIP-1 in the apo form and studied its interaction with a Kir 2.3 C-terminal peptide, using nuclear magnetic resonance (NMR) spectroscopy. We show that TIP-1 consists of a single PDZ domain flanked by unstructured N and C termini. Unique features of the TIP-1 PDZ domain include a short, two-stranded b-sheet within its b1–b2 loop, and a long b2–b3 loop that may participate in ligand binding. Extensive chemical shift perturbations are observed during titration of the protein with a Kir 2.3 peptide, indicating that TIP-1 undergoes conformational changes upon interaction with target proteins. M. A. Durney G. Birrane A. Soni J. A. A. Ladias (&) Molecular Medicine Laboratory and Macromolecular Crystallography Unit, Department of Medicine, Harvard Medical School, Boston, MA 02215, USA e-mail: johnladias@gmail.com