Uncovering mechanisms of interorganelle lipid transport by enzymatic mass tagging

Lipid trafficking is critical for the biogenesis and expansion of organelle membranes. Lipid transport proteins (LTPs) have been proposed to facilitate lipid transport at contact sites between organelles. Despite the fundamental importance of LTPs in cell physiology, our knowledge on the mechanisms of interorganelle lipid distribution remains poor due to the scarcity of assays to monitor lipid flux in vivo. In this review, we highlight the recent development of a versatile method named METALIC (Mass tagging‐Enabled Tracking of Lipids in Cells), which uses a combination of enzymatic mass tagging and mass spectrometry to track lipid flux between organelles inside living cells. We discuss the methodology, its distinct advantages, limitations as well as its potential to unearth the pipelines of lipid transport and LTP function in vivo.

Lipid trafficking is critical for the biogenesis and expansion of organelle membranes.Lipid transport proteins (LTPs) have been proposed to facilitate lipid transport at contact sites between organelles.Despite the fundamental importance of LTPs in cell physiology, our knowledge on the mechanisms of interorganelle lipid distribution remains poor due to the scarcity of assays to monitor lipid flux in vivo.In this review, we highlight the recent development of a versatile method named METALIC (Mass tagging-Enabled Tracking of Lipids in Cells), which uses a combination of enzymatic mass tagging and mass spectrometry to track lipid flux between organelles inside living cells.We discuss the methodology, its distinct advantages, limitations as well as its potential to unearth the pipelines of lipid transport and LTP function in vivo.
Keywords: CFAse; cyclopropane fatty acid; lipid trafficking; lipid transport assay; lipid transport protein; lipidomics; mass tagging; membrane contact sites; METALIC; phospholipid Lipids are key biomolecules that underlie the biogenesis and the distinct architecture and function of organelles.Besides confining different biochemical reactions in organelles, lipids play a variety of roles in energy storage, signaling, protein recruitment, and modulation of membrane protein function among others.A key reason behind the versatile roles of lipids in cell physiology is their extraordinary chemical and structural diversity [1].Among the three major classes of lipids, which include sphingolipids and cholesterol, the most abundant class of lipids in membranes-glycerophospholipids-are made of a glycerol backbone with fatty acyl chains of varying number of carbon atoms and double bonds attached to the sn-1 or the sn-2 position, and a headgroup of different chemical moieties attached to the sn-3 position via a phosphodiester linkage (Fig. 1).Each organelle membrane is characterized by a signature lipid composition that can result from a combination of lipids with different headgroups and fatty acids.Strikingly, glycerophospholipid biosynthesis is mostly confined to the endoplasmic reticulum (ER), with some key lipids synthesized in mitochondria and endosomes.Therefore, lipid trafficking is an essential process for building and expanding organelle membranes.How cells orchestrate differential transport of lipids from their site of synthesis to establish organelle identity and biogenesis is a central question in lipid cell biology.
Research over more than a decade has revealed that a major means by which cells mediate lipid transport is via lipid transport proteins (LTPs) that solubilize lipids within a hydrophobic pocket and catalyze the transport between two membranes.This usually happens at regions of close proximity (10-30 nm) between two organelles, which are termed as 'membrane contact sites' (Fig. 1) [2].A number of putative LTPs implicated in phospholipid transport have been identified using tools of genetics, structural biology, and bioinformatics approaches [3][4][5].The evidence that LTPs actually transport lipids is often derived from in vitro assays monitoring lipid exchange between liposomes [6,7].While in vitro experiments have greatly aided to address the sufficiency of the candidate LTP to transport lipids as well as shed insights into the mechanism of lipid exchange [8,9], they fall short when it comes to addressing the organellar origins and destinations, regulation of lipid transport routes and substrate specificities of LTPs at different contact sites.Moreover, though lipid transport is an essential process, none of the known lipid transporters in yeast are essential, indicating redundancy in LTP function in vivo.Thus, our knowledge on interorganelle lipid trafficking has achieved little progress over decades due to the dearth of assays to monitor lipid flux inside living cells.
In 1990, Jean Vance developed one of the first assays to demonstrate phospholipid transport between ER and mitochondria [10].This assay took advantage of the sequential synthesis of phosphatidylserine (PS), phosphatidylethanolamine (PE), phosphatidylcholine (PC) that is spatially separated between ER and mitochondria.PS synthesized in the ER by PS synthase is transported and converted to PE in mitochondria by PS decarboxylase.PE is then transported back to the ER and converted to PC by the PE methyltransferases in the ER.Therefore, treatment of cell fractions copurifying with mitochondria with radiolabeled serine showed that PS, PE and PC can be detected by thin layer chromatography demonstrating to-and-fro lipid movement between ER and mitochondria.The principle behind this assay has been used in a number of other studies [3,11,12] to address mechanisms of phospholipid trafficking.Nevertheless, this seminal assay has limitations.First, the assay is limited to monitoring lipid transport between ER and mitochondria due to its reliance on endogenous phospholipid modifying enzymes.Secondly, the resolution is limited to the head group.Moreover, in yeast, the major PEsynthesizing enzyme that is usually localized to the inner mitochondrial membrane is additionally found in the ER in certain nutrient conditions [13], questioning the validity of the assay itself.Although the study of phospholipid trafficking has been mostly restricted to ER and mitochondria, radiolabeling approaches have also been used to follow PE trafficking to the plasma membrane [14] and sphingolipid trafficking between ER and late Golgi [15].Furthermore, several chemical biology approaches have been developed, which are based on treating cells with chemically functionalized headgroup precursors or fatty acids and subsequent click chemistry tagging for visualization.As some of these probes can be targeted to specific organelles, they can potentially be used to track interorganelle lipid transport.These and many other exciting technical advances to probe lipids have been discussed in excellent reviews elsewhere [16][17][18][19][20][21].

The advent of METALIC
We developed a new approach that is based on genetically encoded enzymatic modification of lipids and mass spectrometry to study lipid trafficking in vivo in near-native conditions.This assay has been named METALIC, which stands for Mass tagging-Enabled TrAcking of Lipids In Cells [22].The central idea is that two different endo-or exogenous lipid modifying enzymes artificially targeted to two organelles introduce a distinct chemical group on lipids in the respective organelle, which is exploited as a 'mass tag'.Upon transport to the second organelle in the chosen pair, mass-tagged lipids acquire a second mass tag introduced by the second enzyme.The mere detection of doubly mass-tagged lipids by mass spectrometry (MS) from whole cell lipid extracts thus serves as a proxy to assess interorganelle lipid exchange in a quantitative manner (Fig. 2A).
For the first mass tagging enzyme, METALIC takes advantage of the exclusive localization of the native phosphatidylethanolamine (PE) methyltransferase (PEMT), Cho2/Opi3 in yeast, to the endoplasmic reticulum, ER.The triple methylation of PE headgroup to produce phosphatidylcholine (PC) serves as one of the mass tags.The second enzyme is a bacterial cyclopropane fatty acyl phospholipid synthase (CFAse) that introduces a methylene (-CH2) group to the double bonds in fatty acid chains in phospholipids, resulting a cyclopropane ring that serves as the second mass tag.As both PEMT and CFAse require S-adenosylmethionine (SAM) as the cofactor for their activity, metabolic labeling with an isotope-labeled precursor [23] allows to precisely time the mass-labeling of lipids in the respective organelles and monitor the kinetics of lipid transport by quantifying the doubly mass tagged lipid species using mass spectrometry (Fig. 2B).

CFAse and the promise of METALIC
A stand-out feature of the METALIC toolbox is the use of bacterial CFAse as an agent to mass tag lipids.CFAse is a soluble enzyme that introduces a cyclopropane moiety in the double bonds of fatty acids in phospholipids, irrespective of their head group (Fig. 2).Because of its soluble nature, it allows for targeting to different organelles via specific targeting sequences.The use of CFAse together with the ERlocalized PEMT, thus, opens unprecedented avenues to monitor lipid flux between ER and any other organelle in the cell.While the PEMT reaction is an endogenous activity of eukaryotic cells, CFAse activity is alien to yeast and human cells although cyclopropane lipids are found in plants [24].Remarkably, we found that CFAse is active both in yeast and human cells, where it modifies a variety of phospholipids irrespective of the head group upon targeting to different organelles.Importantly, we found that cyclopropane lipids are metabolized at a similar rate as oleic acid, the 18-carbon fatty acid with one double bond commonly found in phospholipids.In addition, the expression of CFAse in any of the tested organelles did not impact growth of both yeast and human cells suggesting that the production of cyclopropane lipids in membranes does not lead to any obvious functional impairment of the tested organelles.These results perhaps align with previous observations that cyclopropane lipids have similar biophysical properties as unsaturated lipids, for instance, in enhancing membrane fluidity due to the similar kinks they introduce in the fatty acyl chain [25,26].Interestingly, in vivo studies in bacteria have found that preferred substrate of CFAse is the fatty acid linked to the 2nd carbon of the glycerol backbone (sn-2 position) in phospholipids, with a further preference for double bonds at the 9th, 10th or 11th carbon in the 16-carbon palmitate or 18carbon oleate with other positions not acted upon [27].In this context, whether phospholipids bearing polyunsaturated fatty acid with a highly branched conformation serve as substrates of CFAse remains to be determined.
METALIC is emerging as an effective tool to unravel the pipelines of lipid distribution between organelles.It gives a high-resolution read-out as the complete identity of the transported lipid species is revealed by mass spectrometry.Using METALIC, we have achieved one of the first demonstrations of LTP activity inside living cells.Specifically, we show that ER-mitochondria lipid exchange is mainly mediated by two proteins complexes, namely ERMES and Vps13-Mcp1, in yeast.Moreover, our data suggest that LTPs like ERMES and Vps13 might exhibit selectivity for specific fatty acids in phospholipids.For instance, the transport kinetics of the di-unsaturated PC 32:2 and PC 34:2 species was affected more compared to the mono-unsaturated PC 32:1 upon depletion of the ERMES complex, than upon VPS13 deletion.These observations raise the intriguing possibility that LTPs like ERMES and Vps13 potentially serve as upstream determinants of organelle identity by transporting specific lipid species.META-LIC can thus be used to systematically investigate the

Caveats in the current METALIC strategy
In the current version of METALIC, as one of the mass tags is introduced in the headgroup by the ERlocalized PEMT, the assessment of lipid transport is restricted to PE and PC.One way to expand the repertoire of the assessed phospholipids could be to skip the use of PEMT, and instead target CFAse to a 'donor' organelle of interest, biochemically fractionate the 'acceptor' organelle of interest and detect cyclopropane lipids irrespective of the head group in the 'acceptor' fraction.However, to unambiguously assess lipid transport, this approach entails that the purified fraction is homogenous without any traces of the 'donor' organelle.In this context, a recently developed method, termed MemPrep, which is based on immunoisolation of organelle membranes, bears immense potential to purify organelles to near-purity [28].
Assessing the directionality of interorganelle lipid transport using PEMT and CFAse is not straightforward as lipids can get their first mass-tag in either of the organelles and acquire their other mass tag in the second organelle upon transport (Fig. 2C).In this context, instead of detecting doubly mass-tagged lipids in whole cell extracts, the above-mentioned approach to use CFAse in solo mode followed by 'acceptor' organelle fractionation can also solve the directionality issue.In the case of ER-to-mitochondria, the directionality issue can be tackled by targeting CFAse to the ER and probing for cyclopropane fatty acids in the mitochondria-specific lipid, cardiolipin.
While METALIC can assess lipid transport between a chosen organelle of origin and an organelle of destination, it can't reveal if the transport is direct or indirect via another organelle.One way to address this issue would be to target CFAse to the suspected intermediate organelle and assess lipid transport in a separate experiment.On similar lines, METALIC cannot differentiate whether the transport occurred by vesicular or nonvesicular means.One can address this issue by knocking out the corresponding genes involved in the respective pathways and quantify their contribution to lipid transport.Vesicular transport being an essential process, the use of temperature-sensitive mutants in yeast or drugs like brefeldin A might be required.Taking the abovementioned factors into account, it is important to note that transport rates obtained from METALIC cannot be used as a direct measure of lipid exchange.However, the effect of perturbations on lipid traffic can be assayed with METALIC, as we have shown upon inactivation of LTPs involved in ER-mitochondria lipid transport [22].
An important aspect to be considered is whether the remodeling of phospholipids via the Lands cycle can interfere with the METALIC read-out.In Lands cycle, phospholipids acted upon by phospholipases are converted to a lyso-phospholipid with one acyl chain, which in turn can be re-acylated from acyl-CoA or from another phospholipid in a trans-acylation reaction by acyltransferases [29][30][31].Potential confusion may arise from the hydrolysis of cyclopropane lipids in their compartment of origin, transport of the generated cyclopropane fatty acid to the ER, and its subsequent reincorporation into a phospholipid molecule.Such a molecule might bear a double mass label while not having been subjected to phospholipid transport per se.It is therefore important to control for the presence and activity of phospholipases and acyltransferases in the donor and acceptor compartments.
Finally, using CFAse as a mass-tagging agent has a few limitations.As it acts on double bonds in the fatty acyl chain of phospholipids, METALIC cannot be used to assess the transport of saturated lipids or the structurally different cholesterol.Whether CFAse can act on sphingolipids with unsaturated fatty acids in human cells remains to be elucidated.Moreover, Fig. 2. Principle of the METALIC assay.(A) Enzyme 1, an ER-localized endogenous enzyme introduces a chemical modification (in pink) in the head group, which acts as a diagnostic 'mass tag.' Upon transport by a LTP to the chosen organelle of destination, the targeted Enzyme 2 introduces a second 'mass tag' (pink triangle).Detection of doubly mass tagged lipid by mass spectrometry serves a read-out for assaying lipid transport.(B) Assaying PC trafficking between ER and any organelle of interest using METALIC.To monitor kinetics, cells are metabolically labeled with deuterated methionine that is converted to deuterated S-adenosylmethionine (d-SAM), the common cofactor used by PEMT and CFAse, releasing S-adenosylhomocysteine (SAH).The labeling of the headgroup in the ER with deuterated methyl (-CD3) groups results in a +9 Da addition compared to unlabeled PC.The mass labeled PC upon reaching the acceptor organelle receives an additional +16 Da in the form of a deuterated methylene (-CD2) group.Thus, the detection of distinct +25 Da species is indicative of lipid exchange between ER and the targeted organelle.(C) Quantifying singly mass-tagged lipids serves as a read-out for enzyme activity as well as metabolic activity of the cell.While doubly mass-tagged lipids report on lipid transport, it can result either from a lipid first modified in the headgroup at ER or in the fatty acid tail at the CFAse-targeted organelle.
CFAse activity can be influenced by several factors.When the enzyme is targeted to the mitochondrial matrix or the lumen of organelles such as the ER or peroxisome, the availability of SAM is a limiting factor for enzyme activity.The alternative here is to target CFAse to the cytosolic face of organelles as we have demonstrated for the ER, outer mitochondrial membrane, peroxisome, plasma membrane and the vacuole/lysosome [22].In the current METALIC setup, CFAse often bears an organelle targeting signal at its N terminus and a mCherry tag at its C terminus to assess localization.It has been observed that optimal activity is dependent on its N-terminal domain and its ability to dimerize [32].Hence, it is possible that the targeting signal and the mCherry tag interfere with CFAse activity, which is in line with our observation that even at steady state CFAse modifies ~40% of phospholipids at best.However, though enzyme activity can be affected by different factors, the rate of mass-tagging by CFAse as well as PEMT can be monitored independently and used to normalize the rate of the doubly masstagged species (Fig. 2C), making METALIC a very useful method to compare the kinetics of interorganelle lipid transport in different genetic or environmental conditions.

Concluding remarks
Unlike PCR for DNA and GFP-tagging for proteins, which rapidly accelerated our understanding of these key biomolecules, lipid biology has lagged due to the dearth of equivalent tools.The fact that lipids are not genetically encoded but rather synthesized by different enzymes has traditionally limited our options for tagging lipids.In METALIC, we have exploited enzymatic synthesis of lipids as a mass-tagging tool, paving way to track interorganelle lipid flux in near-native conditions with minimal perturbations to the structural properties of lipids.The idea to mass-tag lipids by exogenous expression of lipid modifying enzymes from bacteria either alone or in combination with endogenous enzymes could be extended to other relevant enzyme candidates from prokaryotes to probe lipid trafficking in eukaryotic cells.Thus, in future versions of METALIC, there is a wider scope to track lipid transport between any pair of organelles without being dependent on the endogenous ER-localized PEMT.There is a great mismatch between the pace at which LTPs have been discovered and our understanding of their activity and regulation in vivo.METALIC, together with innovative click chemistry approaches that have been developed recently for visualizing phospholipid trafficking [21,33,34], has immense potential to narrow this gap and rapidly advance our knowledge on the mechanisms underlying lipid trafficking and their nexus to organelle physiology.

Fig. 1 .
Fig. 1.Lipid transport at membrane contact sites.(Left) Structural representation of a glycerophospholipid depicting a saturated (no double bonds) fatty acid and a mono-unsaturated (1 double bond) fatty acid linked by an ester bond at the sn-1 and sn-2 positions in the glycerol backbone, respectively.The head group can be substituted by different chemical moieties such as ethanolamine or choline, giving rise to PE or PC, respectively.(Right) Two kinds of LTPs (shuttle & bridge-like) are thought to facilitate lipid transfer at regions of close proximity (10-30 nm) between two organelle membranes.

1295FEBS
Letters 598 (2024) 1292-1298 ª 2024 The Authors.FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies.substratespecificities of LTPs at different organelle interfaces and understand how blocking the transport of specific lipid species can affect different aspects of organelle function.