Functions of intrinsic disorder in transmembrane proteins

Abstract

Intrinsic disorder is common in integral membrane proteins, particularly in the intracellular domains. Despite this observation, these domains are not always recognized as being disordered. In this review, we will discuss the biological functions of intrinsically disordered regions of membrane proteins, and address why the flexibility afforded by disorder is mechanistically important. Intrinsically disordered regions are present in many common classes of membrane proteins including ion channels and transporters; G-protein coupled receptors (GPCRs), receptor tyrosine kinases and cytokine receptors. The functions of the disordered regions are many and varied. We will discuss selected examples including: (1) Organization of receptors, kinases, phosphatases and second messenger sources into signaling complexes. (2) Modulation of the membrane-embedded domain function by ball-and-chain like mechanisms. (3) Trafficking of membrane proteins. (4) Transient membrane associations. (5) Post-translational modifications most notably phosphorylation and (6) disorder-linked isoform dependent function. We finish the review by discussing the future challenges facing the membrane protein community regarding protein disorder.

Keywords: Ball-and-chain inhibition; Intrinsically disordered protein; Lipid interaction domain; Membrane protein; Receptor associated signalling complex.

Fig. 1

Atomistic model of the full-length human prolactin receptor based on integrations of multiple structural techniques including computation. Despite containing only 50% more residues than the folded part of the receptor, the intrinsically disordered intracellular domain occupies a much larger volume, which allows it to facilitate interactions with multiple partners in its micro-environment. The figure represents 10 models of the human PRLR receptor aligned to the ECD and TMD domains Reproduced in a modified form from [43] available under a Creative Commons BY 4.0 license

Fig. 2

Intrinsic disorder in the β₁-adrenergic receptor—a prototypical GPCR. a Disorder predictions demonstrate that the intracellular loop 3 (IL3) between helix 5 and 6 and the C-terminal are likely to be disordered, potentially along with a short N-terminal segment. These regions were removed from the protein before determination of its structure by crystallography [38] b (PDB:2Y00). The pattern of disorder is conserved across GPCRs although the length and sequence of the disordered segments vary considerably. Disorder predictions were done on (NP_000675.1) using IUPRED [15] and transmembrane helices were predicted using TMHMM 2.0 [210]

Fig. 3

Intrinsic disorder in the intracellular domains of the NMDA receptor. a Crystal structure of the ligand binding and transmembrane domains of the NMDA receptor (PDB: 4PE5) [211] showing a dimer-of-dimer structure with two GluN1 subunits (dark) and two GluN2B subunits (light). b In addition to the extracellular domains, each subunit of the NMDA receptor contains an intracellular domain of variable length, that is c poorly conserved and d predicted to be intrinsically disordered (IUPRED) [15]

Fig. 4

Disordered signaling complexes coordinate subsequent steps in pathway into a single structure. a The interactions between CaMKII and the disordered intracellular domains of the NMDA receptor anchor the calcium-activated kinase near a source of calcium-influx. The disordered tail of the NMDA receptor are likely to contribute to its binding of up to six sites in dodecameric CaMKII. b Signaling complexes associated with the pathways following activation of β₁-AR. The GPCR activates AC via a G-protein. Active AC produces cAMP that activates PKA that is linked to both AC and the GPCR via scaffolding proteins. Both proteins are also substrates of PKA, and their phosphorylation states are determined by the competition between PKA and phosphatases bound to the same scaffolding protein

Fig. 5

Ball-and-chain inactivation of ion channels. a Voltage activated potassium channels like the Shaker channel from Drosophila can be inactivated by a short region tethered to a disordered N-terminal. The length of the linker between the inactivating region and the channel determine the rate of inactivation, with short linkers producing fast inactivation and longer linkers resulting in slow inactivation. b Some channels do not have an inhibitory region in the core subunit, but recruit an auxiliary subunit that inactivates the channel by equivalent mechanism to the Shaker channel or c carries the autoinhibitory element in a disordered inter-segmental loop

Fig. 6

Interactions between the disordered region of membrane proteins and the membrane. a Dynamic membrane interaction allowing for switch-like behavior either by phosphorylation, partner binding intracellularly (illustrated by CaM) or extracellularly (illustrated by ligand binding) or by folding upon membrane binding induced by changes in lipid composition. To the left are shown static interactions where the membrane participates as an active binding partner to concentrate interaction in the 2D plane of the membrane normal. b Specific components of the membrane, here illustrated by PI(4,5)P₂ can regulates oligomerization state of membrane proteins leading to either desequestering or sequestering. c Membrane interacting modules can act to bridge different cellular compartments or d induce membrane curvatures

Fig. 7

The SLC9A family isoforms. a Overall cartoon structure of NHE1. The disordered regions are indicated by an ensemble of structures. b Phylogram for the SLC9A family showing the relative phylogenic distance between members. c Order and disorder in three members of the family NHE1 (top), NHE5 (middle) and NHE7 (bottom). DISOPRED2 [212] in red and PONDR-FIT [17] in blue. See [179] for disorder prediction of all nine isoforms Figure modified from [179] with permission. Copyright (2014) Elsevier