Depletion of Plasma Membrane PI4P by ORP5 Requires Hydrolysis by SAC1 in Acceptor Membranes

This article has been Reviewed by the following groups

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Oxysterol binding protein (OSBP)-related proteins (ORPs) 5 and 8 have been shown to deplete the lipid phosphatidylinositol 4-phosphate (PI4P) at sites of membrane contact between the endoplasmic reticulum (ER) and plasma membrane (PM). This is believed to be caused by transport of PI4P from the PM to the ER, where PI4P is degraded by an ER-localized SAC1 phosphatase. This is proposed to power the anti-port of phosphatidylserine (PS) lipids from ER to PM, up their concentration gradient. Alternatively, ORPs have been proposed to sequester PI4P, dependent on the concentration of their alternative lipid ligand. Here, we aimed to distinguish these possibilities in living cells by orthogonal targeting of ORP5 to PM-mitochondrial contact sites. We demonstrate that ORP5 is unable to deplete PM PI4P or accumulate the lipid in mitochondrial outer membranes when acting alone, ruling out both lipid transport and sequestration models. However, when combined with orthogonal targeting of SAC1 to the mitochondrial outer membrane, ORP5 facilitates depletion of PM PI4P when targeted to PM-mitochondria contact sites. We conclude that PI4P depletion from the PM by ORP5 requires transport of PI4P to the tethered organelle membrane, which must be closely coupled to PI4P hydrolysis by SAC1. The data are most compatible with a lipid transfer reaction by ORP family members.

Article activity feed

  1. Review coordinated via ASAPbio’s crowd preprint review

    This review reflects comments and contributions by Amrita Anand, Richa Arya, Aurora Cianciarullo, Luciana Gallo, Dipika Mishra, Sanjeev Sharma, Ryman Shoko and Rajan Thakur. The comments were synthesized by Ehssan Moglad.


    The study conducted by Doyle et al. aimed to test the lipid phosphatidylinositol 4-phosphate (PI4P) transfer activity of the human ORP5 protein via orthogonal targeting of different sites of membrane contact, namely, between the plasma membrane (PM) and the mitochondrial outer membrane.

    Major comments

    Figure 1: The idea behind the experiment is great, however, there are some questions about the data presented:

    • Recommend using better representative images, the morphology of the cells in panels B and C appears distorted, are these cells undergoing some death? This would make interpretation of the data difficult.
    • In Figure 1B, the control experiment should be done with a known mitochondrial marker to test if the construct works as expected.
    • P14P measurement using the P4M probe is unclear from the images and quantification provided. Please provide a multi-plane image of the probe showing its distribution on the PM and at the ER-PM or ER-mito contact site.
    • Please provide an additional graph to show the relative change of PI4P at the PM compared to the rest of the cell or respective contact site.
    • For a better comparison, recommend showing the normal (control) distribution of PI4P in the images.

    Results: ‘As expected, we did not see the accumulation of PI4P at these contact sites (see graph in Fig. 1C), presumably due to SAC1 activity in the ER. Instead, the fluorescence of PM PI4P seemed to decline’: Please indicate whether this result is statistically significant.

    Figure 2: The experiment with the FKBP-PI4KC1001 construct is not discussed in the text. Also, further clarification would be helpful for the results presented in panel B. In +SAC1mito, it is showing accumulation after Rapa treatment, please discuss why PI4P is not showing accumulation.

    'The rationale was that without inhibition of PI4P synthesis, observing reductions in PM PI4P catalyzed by transport of PI4P out of the PM would require a rate that exceeded synthesis, which may not be possible through reduced flux at the much smaller surface area of induced PM-mitochondria contact sites, compared to ER-PM contact sites (compare Figs. 1C and D). We also imaged by Total Internal Reflection Fluorescence Microscopy (TIRFM) to more sensitively detect changes in PM PI4P with the high-affinity PI4P biosensor, P4Mx2.': Recommend revising the fragment for clarity.

    Figure 3: The shape of the cells across figure 3 varies substantially, can some text be added to discuss why this is the case?

    Figure 3: The authors have already shown in a previous paper that SAC1 predominantly acts only in the 'cis' configuration. However, induced coupling of overexpressed ORP between ER-PM and mito-PM using rapamycin might bring these membranes closer than usual or cause the formation of more membrane contact sites. Thus, there may be some possibility for SAC1 to act in 'trans'. Alternatively, there could be indirect changes in PM PI4P due to increased activity of endogenous ORP5 at these induced contact sites. To address this:

    • Would it be possible to confirm if there was an increase in the number/size of contact sites by checking for mapper expression and localization when ORP5 constructs are expressed and coupled with Rapamycin? Lipid binding mutants of ORP5 could also be used to show that those lipid binding mutants do not cause a depletion upon coupling with Rapamycin.
    • For experiments where SAC1 (Fig 3) was overexpressed along with FRB::FKBP-ORP5-ΔTMD, please show control conditions where the SAC1 alone was expressed without the ORP5. Also, in a control condition where lipid binding mutants of ORP5 are expressed along with SAC1, there should be minimal effects on PM PI4P depletion compared to WT ORP5. Adding these controls will further confirm prior observations and substantiate the effects of FRB:: FKBP-ORP5-ΔTMD expression, and rule out any potential artifacts from overexpression of just SAC1.

    A major confound across all the experiments is the activity of endogenous ORP5 that is not measured. Is it possible to perform experiments (such as in Figure 3) where the endogenous ORP5 is downregulated using siRNA or shRNA and a siRNA/shRNA-resistant version of ORP5 is overexpressed in this background? There could be potential compensatory effects from other ORP5, and this would require simultaneous knockdown of multiple ORPs.

    Minor comments

    Figure 1: It could be helpful to start Figure 1 using a scheme of the two hypotheses on how ORP5 regulates PI4P levels at the plasma membrane. This will help easily assess the data presented in the Figures for and against the hypotheses.

    Figure 1A: It is difficult to see the co-localization in the current color scheme. It would be helpful to use a different color combination, provide zoomed-in images, or use pointers to highlight.

    Figures 1C and D: Please specify in the legend the timepoint when rapamycin was added and the subcellular membrane the measurements were made from.

    In discussion: 'In principle, this observation does not demonstrate lipid transfer (though it is compatible with it). Although tethering at a site of membrane contact seems to facilitate access of PI4P to SAC1, ORP5 could simply be presenting the lipid to the phosphatase, as opposed to depositing PI4P into the membrane for subsequent hydrolysis by SAC1. If ORP5 works in such a presentation mode, it is not clear to which membrane the resulting PI lipid is released: either back into the PM, or into the tethered membrane. In other words, lipid transfer is not necessarily part of the reaction': Are there ways in which this can be tested? Suggest proposing some future experiments in the text.