Two cooperative binding sites sensitize PI(4,5)P ₂ recognition by the tubby domain

Abstract

Phosphoinositides (PIs) are lipid signaling molecules that operate by recruiting proteins to cellular membranes via PI recognition domains. The dominant PI of the plasma membrane is phosphatidylinositol 4,5-bisphosphate [PI(4,5)P 2 ]. One of only two PI(4,5)P 2 recognition domains characterized in detail is the tubby domain. It is essential for targeting proteins into cilia involving reversible membrane association. However, the PI(4,5)P 2 binding properties of tubby domains have remained enigmatic. Here, we used coarse-grained molecular dynamics simulations to explore PI(4,5)P 2 binding by the prototypic tubby domain. The comparatively low PI(4,5)P 2 affinity of the previously described canonical binding site is underpinned in a cooperative manner by a previously unknown, adjacent second binding site. Mutations in the previously unknown site impaired PI(4,5)P 2 -dependent plasma membrane localization in living cells and PI(4,5)P 2 interaction in silico, emphasizing its importance for PI(4,5)P 2 affinity. The two-ligand binding mode may serve to sharpen the membrane association-dissociation cycle of tubby-like proteins that underlies delivery of ciliary cargo.

GōMartini 3: From large conformational changes in proteins to environmental bias corrections

“…Figure 3A shows the PCLδ1 PH domain with a PI(4,5)P2 lipid in its crystal structure binding pocket embedded in a POPC bilayer. Unbiased MD simulations starting from the membrane-bound structure confirm the high affinity of the PLCδ1 PH domain to a single PI(4,5)P2 lipid (Figure S2) which has also been shown previously based on atomistic as well as CG simulations 40,49,65,67,68 . We used three different structural bias models.…”

“…Moreover, the fully reparameterized Martini 3 model for proteins is presented, pointing to which improvements in the model may enable more accurate predictions of protein packing and protein interactions 38,39 . The GōMartini implementation together with Martini 3 has already shown that it can capture subtle changes in protein dynamics caused by interactions with membranes 40 , single point mutations 41 and mechanostability. 42 We also show how the virtual-site implementation can be used to implement an environmental bias to correct recently described inaccuracies of the model, such as underestimated dimensions of intrinsically disordered proteins (IDPs) 43,44 and low hydrophobicity of certain amphiphilic small peptides.…”

PI(4,5)P2role in Transverse-tubule membrane formation and muscle function

“…We surveyed the myofiber localization of eight selective biosensors available for five of the seven different PIPs, including PI(3)P, PI(4)P, PI(3,4)P 2 , PI(4,5)P 2 and PI(3,4,5)P 3 . Only the PI(4)P indicator, P4M (Brombacher et al, 2009, Hammond et al, 2014, Liu et al, 2018), and the two PI(4,5)P 2 indicators, GFP-tagged PH-PLC8 and Tubby-C (Santagata et al, 2001, Stauffer et al, 1998, Thallmair et al, 2022, Varnai and Balla, 1998), were found to localize at the T-tubule network in larval body wall muscles ( Supplemental Figure S2B ). In wildtype conditions, both PI(4)P and PI(4,5)P 2 co-localized with Dlg1, as well as with each other, along T-tubule membranes ( Figure 2A-2E ).…”

Pragmatic Coarse-Graining of Proteins: Models and Applications

“…Martini has been heavily used to study specific protein–lipid interactions, particularly the protein’s lipid “fingerprint”, i.e., the unique local lipid environment around a protein when inserted in a complex membrane . Some examples of this are the extensive work done on a variety of G protein-coupled receptors (GPCRs), − several pores and transmembrane channels, − and peripheral membrane proteins, ,− among many others. − …”