º£½Çֱ²¥

>

Novel structural insights reveal how human IP3R-2 receptors control calcium release

By Shelby A. Harris

Inositol 1,4,5-trisphosphate (IP3) receptors (IP3Rs) control calcium (Ca²⁺) signals by tightly regulating the spatial and temporal release of Ca²⁺ from stores within the endoplasmic reticulum into the cytosol of the cell while bound to IP3. Ca²⁺ is an important second messenger ion that has roles in a variety of physiological processes in the body, notably cell apoptosis, gene expression, and muscle contraction, and its dysregulation disrupts function, leading to disease.

There are three subtypes of IP3R (IP3R-1–3), which share around 60–70% amino acid conservation and assemble as homo- or heterotetramers, forming one of the largest intracellular ion channel complexes at ~1.2 MDa. While the sequence identity of the IP3R subtypes is high, they still possess distinct spatiotemporal expression patterns and physiological properties. IP3R-2 exhibits the highest affinity for IP3, is more sensitive to potentiation by ATP (which acts as a regulator of IP3R), and IP3R-2 channels tend to support cell-wide Ca²⁺ oscillations rather than the large, sustained Ca²⁺ transients seen in other subtypes such as IP3R-3.

Erkan Karakas, Associate Professor
Molecular Physiology and Biophysics

CSB researchers Caifeng Liu and Erkan Karakas report the cryo-EM structures of human IP3R-2 (hIP3R-2) in distinct conformations in the presence and absence of IP3, Ca²⁺, and ATP in a recently published Nature Communications paper. Here, they while defining compact (apo, resting, and preactivated), loose (inactive-like), and asymmetric intermediates.

 

In a nominally Ca²⁺-free environment, hIP3R-2 was shown to adopt a homotetrameric, mushroom-shaped assembly. 3D refinement of the compact class with C4 symmetry yielded a high-quality map at 2.95 Å overall resolution and resembled the low-Ca²⁺ structures of rat IP3R-1 and human IP3R-3.

Consensus cryo-EM maps of hIP3R-2 at ligand-bound resting (a), intermediate state 3 (b), intermediate state 4 (c), and preactivated state (d), as shown in the publication.

However, the loose conformation was remarkably flexible and retained weak cytosolic intersubunit contacts; refinement strategies that improved the compact class did not markedly improve this class. Each subunit shares a conserved domain architecture with other IP3R subtypes, comprising two β-trefoil domains (βTF1/2), three Armadillo repeat domains (ARM1–3), a central linker, a zinc-finger–containing juxtamembrane domain (JD), a transmembrane domain (TMD), and a C-terminal domain. The TMD forms a closed channel with a pore constriction radius of ~0.7 Å, flanked by lipid-like densities and conserved extra helices (S1′/S1″). Compared to hIP3R-3, apo hIP3R-2 displays subtle structural variations, most notably an ARM2 domain tilted further toward adjacent subunits and a flexible, disordered self-binding peptide region near the IP3-binding pocket.

Cryo-EM structures of hIP3R-2 bound to IP3 and ATP show a clear activation pathway driven by specific domain movements. When IP3 binds, the receptor undergoes a small local change, with the ARM1 domain rotating ~4° toward βTF2. The protein then moves into a preactivated state where the ARM2 domain pulls back toward the channel-like domain, and ARM1 rotates an additional 16° to fully close the cleft. Reconstructions of asymmetric intermediate states reveal that these domain movements can occur independently within individual subunits. However, ARM2 movement in one subunit directly correlates with cleft closure in its neighbor, proving that subunits communicate with each other during activation.

Although hIP3R-2 follows an activation path similar to other isoforms, its response to calcium is unique. In datasets containing IP3, ATP, and calcium, calcium density is visible at the primary activation site between ARM3 and the JD. Despite this binding, the channels remain in a loose, inactive-like shape rather than opening their pores. Unlike hIP3R-3, no extra calcium-binding sites were detected. Because local flexibility can hide weak calcium density, these structures conservatively represent either desensitized or calcium-inhibited states, showing that calcium binding at the activation site does not automatically open the pore. These structures also explain how receptors interact with one another when exposed to IP3, ATP, and calcium.

Caifeng Liu, Researcher
Karakas Lab

Neighboring channels form dimers and larger assemblies through a cytosolic interface made of ARM1 and βTF1 domains. This interface only opens after ligand-driven domain shifts occur, meaning it is absent in the apo or IP3/ATP states. The transmembrane domains of adjacent channels align along a flat plane, making this arrangement fully compatible with resting side-by-side in the ER membrane, unlike previous hIP3R-3 dimer models, which would distort the lipid bilayer. Ligand binding acts as a trigger to expose the ARM1-βTF1 interface, driving direct channel clustering that underlies localized calcium puffs. Mutating this contact site with the G706K mutation stops channels from clustering without altering their individual structure or gating activity. This mutation provides a precise tool to separate spatial clustering from channel function in future calcium signaling research.

This paper presents the cryo-EM structures of hIP3R-2 conformations in the presence and absence of IP3, Ca²⁺, and ATP and demonstrates that IP3R channels can self-assemble without needing accessory proteins or cytoskeletal anchors. By mapping the conformational landscape, ligand binding, and inter-channel assemblies of IP3R-2, their findings establish the structural basis of its gating and spatial organization.

The Karakas lab concludes with the statement that integrating these structural insights with further research in electrophysiology, helper factor identification, and cross-isoform comparisons will be essential to fully resolve IP3R-mediated Ca²⁺ dynamics.

 

Explore Story Topics

Recent News