Menu

Close

Probing the Subcellular Distribution of Phosphatidylinositol Reveals a Surprising Lack at the Plasma Membrane

James P. Zewe, April Miller, Sahana Sangappa, Rachel C. Wills, Brady D. Goulden, Gerald R. V. Hammond

Preprint posted on June 19, 2019 https://www.biorxiv.org/content/10.1101/677039v1.article-info

and

Defining the Subcellular Distribution and Metabolic Channeling of Phosphatidylinositol

Joshua G. Pemberton, Yeun Ju Kim, Nivedita Sengupta, Andrea Eisenreichova, Daniel J. Toth, Evzen Boura, Tamas Balla

Preprint posted on June 20, 2019 https://www.biorxiv.org/content/10.1101/677229v1.article-info

It’s time to rethink how some lipids are distributed inside cells. Brand new molecular probes made by two labs show that phosphatidylinositol gets around more than we realise.

Selected by Sruthi Balakrishnan

Background

The lipid phosphatidylinositol (PI) performs both structural and functional roles in cells. Structurally, it anchors proteins in the plasma membrane (PM) and maintains membrane integrity. During infections, bacteria can break this link between PI and the proteins, creating avenues into the host cell. Apart from this role at outer leaflet of the PM, PI is a source of lipid intermediates required for cellular signalling and conferring identity to compartments like the Golgi apparatus and endosomes [1].

In cells, PI is thought to be present mainly at the endoplasmic reticulum (ER), with a minor presence on other compartments like the Golgi apparatus and PM [2]. Most of the available information regarding PI abundance and localisation comes from subcellular fractionation assays, which have limited resolution and potential of cross-contamination between organelles. While lipid-binding probes have been used to great effect, there have been none developed for PI till date [3]. Accurate data about PI subcellular distribution is required to understand processes like signalling, which often have high turnover rates of PI-derived lipid intermediates.

The two preprints discussed here have independently developed methods for live-tracking PI at specific membranes, offering new insights into its localisation and abundance.

Key Findings

The first study, from Gerald Hammond’s group at the University of Pittsburgh, used a bacterial PI- specific phospholipase C (PI-PLC) to develop an activatable probe for PI. PI-PLC hydrolyses PI to produce another lipid known as DAG, for which there exists a high-specificity lipid probe. The levels of DAG now provide an indirect readout of PI. To control PI-PLC activation, the group split PI-PLC from Listeria monocytogenes into its N-terminus, fused to a protein domain called FKBP, and C-terminus, fused to a domain called FRB. FKBP and FRB dimerise in the presence of the drug rapamycin, allowing temporal control over PI-PLC activity. The FKBP domain can be targeted to different subcellular compartments, allowing spatial control of the probe.

Using different versions of the split PI-PLC constructs, they found that PI was most abundant on the Golgi apparatus and peroxisomes, followed by the mitochondria and endosomes. Contrary to expectations, the probe did not show any signal from the PM or the ER.

They then made a second set of probes, this time using a recruitable variant of an enzyme (PI4K) that converts PI to a lipid called PI4P. Levels of PI4P now serve as proxy readouts of PI. While initially there was no PI detected on the ER using this probe, the lack of signal was attributed to the presence of a PI4P-degrading enzyme known as Sac1.When Sac1 activity was blocked, PI was again detected at the ER, confirming its presence at the compartment.

The second study, from Tamas Balla’s group at NIH, used PI-PLC probes derived from Bacillus cereus. A thorough study of protein structure enabled them to reduce the catalytic activity of PI-PLC since the native form was toxic to the cell. This altered construct was sent to specific compartments using  the FKBP-FRB system. After rapamycin-induced dimerisation, they measured the DAG levels as an indirect readout of PI, using a probe similar to the one used by the first group.

Using PI-PLC constructs targetted to different compartments, they found that PI is present at the ER, mitochondria, Golgi apparatus, peroxisomes and endosomes. There was no signal for PI seen at the PM, consistent with the observations made in the first study.

The relevance of PI at the ER was confirmed using a second set of experiments. PI4P, which is a signalling lipid derived from PI, has a constitutive presence at the PM. When PI was selectively removed from the ER, PI4P levels at the PM dropped significantly. Removing PI from the PM, however, had no effect on the PI4P levels.

Both of the studies highlighted two important aspects of PI distribution in cells. The first being that PI has a higher presence in non-ER compartments than previously considered. The second is that free PI is most likely never present at the PM, but rather transferred to the PM only upon requirement. Additionally, both studies have added a whole new repertoire of tools to track PI dynamics in live, intact cells – something that has been lacking in the field until now.

What I like about these preprints

Firstly, I was excited to see that two independent research groups had made similar findings. It is always a win for science when results are corroborated by different scientists. In this case, both the studies beautifully complement each other as the same findings were made using varied approaches.

Secondly, the studies have generated tools to study PI in live cells with greater resolution than previously available.

Finally, they challenged the important assumption that most of the PI in cells is concentrated at the ER. The presence of PI in high abundance at other compartments prompts a reassessment of the standard membrane composition models.

Future Directions

The derivatives of PI present at the PM are not only required for compartment identity, but also undergo fast turnover when signalling is turned on [4].The marked absence of PI from the PM, therefore, points to its supply as a potential checkpoint. This also begs the question of why PI is not allowed to remain on the PM for any length of time.

The observations of substantial PI presence at nearly all subcellular compartments also warrants a rethinking of lipid metabolism dynamics.The detection of PI at the mitochondria, in particular, is remarkable since there are very few reports of PI derivatives at this compartment.The mitochondrial membrane is known to form contact sites with other compartments like the endoplasmic reticulum. It is also a reported to be a source for autophagosomal membranes [5]. Both of these processes often involve lipid derivatives of PI and further investigations into this aspect could unearth new mechanistic details.

Questions for the Authors 

  1. The mitochondrial outer membrane interacts with other cellular compartments, often via PI- derived lipids. Does selectively depleting PI on the mitochondria change these interactions?
  2. Removing PI from the ER reduces the PI4P levels at the PM. Does it also affect the numbers/ clustering of the PI-anchored proteins?
  3. In both studies, PI was detected at the ER at lower levels than anticipated.The readout was either of low intensity (preprint 1) or transient (preprint 2). Does this solely reflect dynamics of the proxy lipid or actual fluctuations in the levels of PI at the ER?

Bibliography

  1. Sunshine, & Iruela-Arispe, M. L. Membrane lipids and cell signaling: Current Opinion in Lipidology 28, 408–413 (2017).
  2. Yang,Y., Lee, & Fairn, G. D. Phospholipid subcellular localization and dynamics. Journal of Biological Chemistry 293, 6230–6240 (2018).
  3. Wills, C., Goulden, B. D. & Hammond, G. R.V. Genetically encoded lipid biosensors. MBoC 29, 1526–1532 (2018).
  4. Rhee, G. Regulation of Phosphoinositide Specific PLC.pdf. Annu. Rev. Biochem. 70, 281– 312 (2001).
  5. Hailey, W. et al. Mitochondria Supply Membranes for Autophagosome Biogenesis during Starvation. Cell 141, 656–667 (2010).

Tags: imaging, lipids, probes, tools

Posted on: 25th July 2019

(2 votes)




  • Author's response

    Gerald Hammond shared about Probing the Subcellular Distribution of Phosphatidylinositol Reveals a Surprising Lack at the Plasma Membrane

    Regarding the first question, we haven’t looked yet at whether contact sites between mitochondria and other organelles are disrupted – although if PI is a major component of the outer membrane, this may occur due to non-specific structural damage. What’s curious about your question for me is that this gets at a larger question – does PI have specific roles beyond acting as a precursor for GPI-anchored proteins and phosphoinositides? We don’t know. Hopefully, we’re starting to have tools at answer these questions now.
    I’ll leave question number 2 to the Balla lab, since this was their observation, but regarding question 3: we just don’t know yet! We know that DAG and PI4P are both rapidly metabolized at the ER. Plus we know that PI can “flop” into the luminal leaflet of the ER for GPI synthesis, and this has been measured. So is there relatively little in the outer leaflet of the ER, or does it slip past our probes before they can detect it? We will have to come up with some craftier experiments to really nail this down (though reviewers: this is for the next paper!!).

    Have your say

    Your email address will not be published. Required fields are marked *

    This site uses Akismet to reduce spam. Learn how your comment data is processed.

    Sign up to customise the site to your preferences and to receive alerts

    Register here
    Close