Zainab Mustapha (Rutgers University, United States)
Abstract: Kindlin-2, K2, is a peripheral membrane protein and co-activator of integrin signaling in the cell, which is implicated in cell migration, adhesion, and cancer. K2 contains a pleckstrin homology (PH) domain, which, like many PH domains, binds to the phosphoinositide components of the cell membrane, specifically phosphatidylinositol-3,4,5-trisphosphate (PIP3), to enhance integrin activation. However, the mechanism of PIP3 recognition and binding is not fully understood, as no structural studies to date use full-length PIP3 in lipid bilayers, focusing instead on the soluble inositol headgroup.
In this study, we use a combination of solid-state and solution NMR to investigate the structure and dynamics of PIP3-bound K2-PH in a model membrane containing PIP3, phosphatidylcholine, phosphatidylserine, and cholesterol. We study changes in the bound state with respect to the model membrane and the unbound protein. Using proton detection and very fast MAS, ssNMR results show chemical shift perturbations in the backbone of the bound protein compared to the unbound form. These perturbations confirm interactions in areas predicted to interact with the membrane (in our MD simulations). We also describe how additional membrane components, such as cholesterol, can stabilize the binding of the protein.
Together, these results suggest that PIP3 binding induces structural changes in membrane-interacting regions of K2-PH, and that additional membrane components may help stabilize this interaction. Our findings help explain the mechanism of K2-PH binding to PIP3s in the context of a full-length lipid bilayer, which has broad implications for the PIP-based regulation of numerous important cellular processes.
-
Nice presentation!
I have a few questions:
1) Are the bilayers preserving their structural integrity upon spinning in MAS?
2) T398 seem to split into two. Would you have an explanation for this?
3) Can you quantify the fraction of PH bound to the bilayers?cheers
-
Thank you! Very good questions and observation
1) Usually, we spin the rotors containing liposomes only at 15 kHz, because they preserve their structural integrity and give good linewidth at this MAS rate. However, the 31P 1D data of the bound sample was spun at 40 kHz and the 31P static spectra taken before and after shows that the bilayer structure is preserved. So yes, the bilayers preserve their structural integrity2) Very nice observation, I am still trying to make sense of all the changes we see.
3) Yes, so we start out with solution NMR titration, which of course renders the membrane-associated protein invisible as the titration progresses. When we pellet the complex, we collect the supernatant and take 15N-HSQC to estimate how much of it is left in solution. We use this as a measure of what is bound to the bilayer.
I hope this answers your question. Please let me know if you have any suggestions or input. I’ll be happy to take them. Thank you again.
-
Dear Zainab,
Thank you for your presentation.1. I was wondering if you were able to determine the binding constants from your titration studies. If so, could you share the binding affinity you observed between the PH domain and PIP3?
2. Looking at your static 31P spectra, I noticed slight differences between the bound and unbound forms. Could you please elaborate on what might be causing these differences?Thank you!
-
Hi KSHAMA,
Very good questions.
1. We didn’t determine binding affinity from our solution NMR titrations. However, in a pioneering work with the soluble headgroup of PIP3, the Kd was measured to be about 2.12 uM. It’s worth thinking about if the presence of the tails would give a different measurement.
2. The presence of the protein may be inducing some sort of membrane curvature on the bound sample compared to the unbound. It can be that some of phosphates in the bilayer now have a different orientation as a result of protein binding. Overall, even though there are subtle differences between the two spectra, we think the bilayer may not be completely destroyed because at a different protein:lipid ratio (data not shown), the static spectra completely shows a different powder pattern.
Thanks for engaging with my research and I hope this answers your questions. Happy to take any suggestions or further questions.
Leave a Reply