Abstract
Membrane contact sites (MCSs) are microdomains that exchange ions and lipids between the membranes of two organelles. They facilitate the exchange of metabolites and act as a site for intracellular communication through material transport. Because of the important physiological significance of MCSs in localizing the exchange of substances and metabolic regulation, they are considered to play an important role in cell biology. Understanding MCS structure is essential for analyzing how substances move to and from each organelle. Several methods have been developed to analyze MCS function, with electron microscopy (EM) being the predominant technique when structural detail is needed. In this review, we summarize the ultrastructure of MCSs and how EM can be used to determine their role in cell biology.
Keywords
Introduction
Membrane contact sites (MCSs) are an important method of communication between organelles. The MCSs are located where two organelles meet in very close proximity. The molecules that are actively involved in this process are still in the process of being identified and structurally analyzed. The proteins that maintain the MCS and proximity between organelles are known as tethers (Eisenberg-Bord 2016; Scorrano et al., 2019). Electron microscopy (EM) has been used not only to image these structures, in which the tethers maintain a very small gap, but also to measure the gaps accurately. It is important to note that the contacting membranes do not fuse with one another (Gatta and Levine, 2017; Sarhadi et al. 2023; Voeltz et al., 2024); This has been confirmed by much electron microscope data. Biochemistry and membrane cell biology show that MCSs are typically supported by many tethers. In general, the distance across the gaps forming the MCSs is 15–30 nm (Gatta and Levine, 2017; Wu et al., 2017); however, this distance appears to depend on the length of the tethers (Hoffmann and Kukulski, 2017). The distance between the two organelles is reported in some cases to be less than 10 nm, and here presumably the tethers are short. The surface area forming the MCS may vary considerably at different types of contact. This may depend on the type of molecular communication that occurs or the extent to which the structure is transient.
This short review describes the structure of MCSs as revealed by EM. When cells are observed using an electron microscope, many of the organelles that were previously considered to be floating in the cytoplasm were found to be in contact with other organelles (Figure 1). The recent discovery that close contact between organelles is important for cell function has prompted researchers to reevaluate the abundance of organelle contacts in electron micrographs, which had previously been largely ignored.
The organelle makes contact sites with various cellular compartments. (A–D) Electron micrographs showing the organelles in proximity. Yellow arrows indicate the membrane contact sites. Endoplasmic reticulum (ER), Lipid droplet, Endosome (En), Mitochondria (M), Lysosome (L). Scale Bars, 250 nm (A–C) or 200 nm (D).
MCS Analysis by EM
EM is a powerful tool for revealing the detailed structures of MCSs, with many different techniques available. Standard transmission EM (TEM) enables static, high-resolution observation of intracellular MCS structures (Sarhadi et al., 2023; Scorrano et al., 2019). Beyond standard TEM, there are several other approaches, each with their own advantages. Both immuno-gold labeling, which uses antibodies to detect specific membrane components, and correlative light electron microscopy (CLEM) can make significant contributions to organelle identification, which is important to understand MCS function. Ultrathin cryo-sectioning by the Tokuyasu method provides a different approach to immunodetection in EM without such strong fixation as standard immune-gold labelling. An additional technique uses ascorbate peroxidase (APEX) to visualize MCSs by labeling proteins present in close proximity to the APEX-tagged protein (range approximately 20 nm). Electron tomography (ET) is a method in which a large number of transmission images are taken of a single sample that is rotated; views of the same objects from different angles are used for three-dimensional (3D) construction of the sample. The information obtained with these techniques, especially ET and immuno-gold labeling, provides more comprehensive and complete insight than TEM (Mari et al. 2014; Nara et al. 2023) Another 3D reconstruction approach is Focused Ion Beam-Scanning Electron Microscopy (FIB-SEM), with processing technology to cut the desired portion inside the sample. This can overcome the tilt angle problem in ET analysis. Performing this under cryo-conditions (cryo-FIB) allows visualization of undisturbed in situ structures of organelles at high resolution.
Looking at the weaknesses and limitations of EM, one is that it is a very low throughput technique (Scorrano et al., 2019). This is addressed by the latest TEM and FIB-SEM approaches, which can capture a wide field of view. By combining whole-cell FIB-SEM images with deep learning-based segmentation, maps of thousands of structures packed within a single cell have been successfully obtained (Heinrich et al., 2021). Thus, sampling and training a random number of samples proved to be steps toward overcoming the low throughput problem. EM is also being used to develop a high-throughput MCS quantitation system (Liu et al., 2022). DeepContact precisely measures the proportion of contact between two organelle segmentations visualized in a high-resolution electron microscope image by determining the ratio of the contact length between two organelles to the organelle's perimeter length. ET and cryo-ET are laborious approaches and can be technically challenging (Sarhadi et al., 2023). ET imaging is made easier by the use of a computer driven tiltable stage. There are various software packages available to reconstruct the resulting tilt angle images, ranging from simple to professional versions. However, cell fixation with chemical agents may introduce artifacts (Collado et al., 2019). Thus, caution should be taken when interpreting absolute distances with accuracy on a nanometer scale with aldehyde-fixed tissues (Chung et al., 2022). The fixation method for EM is an important aspect of the procedure, thus optimal protocols should be evaluated. Lak et al. elaborated on the experimental fixation conditions to maintain antigenicity in Immuno-gold labeling (Lak et al., 2021).
ER-Mitochondria
The existence of physical contacts between the ER and mitochondria was revealed in the 1950s by EM (Bernhard and Rouiller, 1956; Porter and Palade, 1957); however, the details of the structure and the significance of these contacts, including the presence of tethers, remained unresolved. Cryo-CLEM analysis and other methods have shown that the intermembrane distance between smooth ER and mitochondria is on average approximately 24 nm (Table 1, Pacher et al., 2000; Csordás et al., 2006; Wozny et al., 2023). In mouse liver, the width of the gap and the length of the ER-mitochondria MCSs depend on the metabolic state (Giacomello and Pellegrini, 2016; Sood et al., 2014). Some observations indicate that ribosomes are excluded from MCSs formed on the surface of mitochondria (Giacomello and Pellegrini, 2016, Sood et al., 2014; Wu et al., 2018). Moreover, ERs without ribosomes can form tight MCSs with gaps of 10–30 nm from the surface of the mitochondria. However, there are also contacts between the rough-ER and the mitochondria (called ‘ribo-MERCs’) with the contact distance wider at 50–80 nm (Giacomello and Pellegrini, 2016; Hung et al., 2017; Lak et al., 2021); however, the tether remains indistinct, so it is unclear whether it is an MCS.
The Inter-Membrane Distance of MCSs and the Electron Microscope Technique Used for Observation.
The contacts between the ER and mitochondria are essential for proper mitochondrial metabolism (Kornmann et al., 2009) and for regulating mitochondrial Ca2+ levels (Hayashi et al., 2009). During mitosis, Ca2+ transfer from the ER to the mitochondria is enhanced. The length of contacts between ER and mitochondria is significantly longer in mitosis (284 nm) than in interphase (189 nm) in Jurkat T lymphocytes (Yu et al., 2024). Conversely, the intermembrane distance is not significantly changed between interphase and mitosis. A 3D reconstruction analysis revealed that during mitosis, the mitochondria were embedded in an ER basket because the contact has increased in extent, while in interphase the ER edges contacted the mitochondria (Yu et al., 2024).
ER-PM
In many eukaryotic cells, the ER exchanges molecules through contact with the plasma membrane. 3D reconstruction experiments by Pichler et al. using yeast showed that there are 1100 contacts between PM and ER that are within 30 nm of one another, a very significant number compared with the 80 ER-mitochondria proximities (Pichler et al., 2001). In yeast, ER-PM proximity covers approximately 40% of the surface area of the cytoplasmic side of the PM (Henne, 2016). In mammalian cells, ER-PM proximity account for 2%–5% of the PM's area (Henne, 2016). The ER-PM contacts in COS7 cells are formed by wide ER compartments, which are connected to the remainder of the ER by narrow tubules (Fernández-Busnadiego et al., 2015). The average distance between ER-PM contacts is 25.4 nm, and when E-Syt3 or E-Syt1 are overexpressed, the distance decreases to 18.8 and 21.8 nm, respectively (Fernández-Busnadiego et al., 2015).
The E-Syt yeast ortholog tricalbin forms a highly curved dome-shape locally in the ER and contributes to the reduction of the physical distance to the PM. This local curvature was observed as a singular peak, contributing to 0.15% of the ER surface area. The appearance of this peak was enhanced during heat stress (Collado et al., 2019). Alternatively, there is evidence that the curvature prevents contact surfaces from approaching one another (Hoffmann and Kukulski, 2017; West et al., 2011). Similar peak structures have been observed in ER-PM (Collado et al., 2019), in autophagosome-vacuole (Bieber et al., 2022), ER-mitochondria (Nara et al., 2023), and nucleus-mitochondria (Tracey-White and Hayes, 2024) contacts.
In studies using FIB-SEM on neurons to identify the contact structures between the ER and the PM, 12.5% of the ER was in contact with the PM (Wu et al., 2017). Of the ERs, 37.3% were “thin ER” with a narrow ER lumen, to which ribosomes were not bound. The thin ER may reside between the ER with its wide lumen and the PM, or it may line up with another thin ER in an adjacent neuron; however, the significance of the thin ER is not yet known and remains to be clarified in the future.
The ultrastructural analysis of the ER-PM contacts has remained unresolved in cell types other than yeast and neurons. In polarized epithelial cells, the lateral PM domain had 40- and 4-fold more ER contact areas compared with the other apical PM and basal domains, respectively (Chung et al., 2022). 3D reconstruction experiments with EM revealed that ER-PM contacts in the lateral domain were large and extensively occupied the PM, whereas the average length of the contact site between the contacts was 120 nm (Table 2). Extensive ER-PM contacts (>200 nm) were predominantly found in the lateral domain that forms intercellular junctions between two hepatocytes (Chung et al., 2022), suggesting that the formation of ER-PM proximity is tightly regulated.
Length of MCSs and the Electron Microscope Technique Used for Observation.
Surprisingly, few EM studies have revealed the presence of tether molecules in MCSs by labeling the tethers with antibodies. The oxysterol-binding protein-related proteins, ORP5 and ORP8, are ER-anchored proteins that have been implicated in MCS function but their localization differs in each MCSs. In immunogold labeling experiments performed on ultrathin frozen sections, approximately 60% of the EGFP-ORP5 was localized to areas of apposition between the EM and the PM, whereas EGFP-ORP8 was virtually absent (Galmes et al., 2016). In contrast, approximately 60% of the EGFP-ORP8 was located at the contact sites between the ER and mitochondria. Antibodies that can bind to tether proteins in the gaps and thus that can be used for immuno-electron microscopy (immuno-EM) are required.
ER-Endosomes/Lysosomes
EM analysis revealed that ∼50% of endosomes containing endocytosed EGFR were in contact with the ER, at distances of less than 20 nm, and tethered structures were also observed in mammalian HeLa cells (Eden et al., 2010). A single multivesicular body exhibits multiple contacts with the tubular ER as revealed by EM tomography (Friedman et al., 2013). The intermembrane distance between the ER and endosomes was ∼10 nm in HeLa cells (Alpy et al., 2013 and Di Mattia et al., 2020) and ∼13 nm in JEG3 placental cells (Nara et al., 2023).
The lipid transport protein Vps13C interacts with the ER protein VAP-B, whereas the ER protein PDZD8 interacts with endosomal Rab7. When Vps13C and PDZD8 are overexpressed simultaneously, the two proteins are localized to different regions of the ER-endosome MCSs. Correlative fluorescence microscopy and FIB-SEM analyses revealed that the distance to form the MCSs is distinctly shorter where PDZD8 is localized (Cai et al., 2022). Expression of a truncated form of Vps13C, in which a portion of lipid transfer channel is deleted, resulted in an intermembrane distance of ∼24 nm in the MCSs, which was ∼5 nm shorter compared with that of full-length Vps13C. Cryo-ET analysis also revealed that the rod-like tether structures were shorter (Cai et al., 2022). Thus, EM analysis directly demonstrates that different interacting molecules form MCSs at different distances.
Mitochondria-Endosome/Lysosomes
Mitochondria-endosome/lysosome contacts have been detected in several specialized cell types. In developing erythroid hemoglobin-produced cells, HRP-Transferrin-labeled endosomes contact mitochondria to deliver endocytosed Fe (Sheftel et al., 2007). In melanocytes, early melanosomal marker Pmel17-labeled melanosomes, which are specialized lysosome-related organelles of pigment cells, contact mitochondria (Daniele et al., 2014). 3D tomographic analysis revealed that several tethers were present between the melanosomes and mitochondria, with a length of 20–30 nm (Daniele et al., 2014).
It has been suggested that contacts between endosomes and mitochondria may provide a circuitous route for lipid transport from the ER to the mitochondria (Gatta and Levine, 2017). Deletion of NPC1, a cholesterol sensor protein present in endosomal membranes, was found to eliminate MCSs between the ER and endosomes, but it elevates cholesterol in mitochondria. In cells treated with the drug U18666A, which mimics NPC1 deficiency, ET analysis revealed that lysosomes labeled with incorporated HRP were engulfed by mitochondria (Höglinger et al., 2019). Lysosomes are captured as if embraced by the mitochondrial outer membrane, suggesting tight contact. Furthermore, double knockdown of NPC1 and MLN64 reduced mitochondrial lysosomal MCS, indicating that MLN64 is required for the formation of this MCS. Mitochondria in placental cells require cholesterol to produce steroid hormones needed to maintain pregnancy. The mitochondria of placental cells are in close proximity to endosomes, forming an MCS with a gap of approximately 19.1 nm as measured by immuno-EM analysis. When the endosomal protein MLN64/Stard3 was knocked down, the contacts separated to a distance of 38–154 nm (Nara et al., 2023). These findings indicate that the endosome-mitochondrial contact structures have a lipid exchange function similar to other contact structures.
Other MCSs
Although there has been extensive research on communication between the nucleus and the mitochondria, there have been few examples of physical contact until recently. Because the outer membrane of the nucleus is connected to the membrane of the endoplasmic reticulum, it was difficult to distinguish this from the ER-mitochondrial contact. Eisenberg-Bord et al. successfully obtained electron micrographs of nuclear ER-mitochondrial contacts in yeast and showed that the intermembrane distance was ∼20 nm (Eisenberg-Bord et al., 2021). A study using A549 cells showed similar quantitative results with an intermembrane distance of 19.5 nm (Zervopoulos et al., 2022). In yeast, overexpression of Cnm1, which is localized to the nuclear membrane, caused mitochondria to make contact around the nucleus periphery, approximately twofold closer compared with the controls (Eisenberg-Bord et al., 2021). In the basal epithelial cells of larval Zebrafish skin, a large number of mitochondria were found in the vicinity of the nuclear periphery and contact with each other at a length of approximately 263 nm (Tracey-White and Hayes, 2024).
Autophagy is another process implicated in MCS formation. When autophagy is induced by starvation, cup-shaped structures known as phagophores are formed, which grow into autophagosomes that subsequently fuse with vacuoles to degrade their contents. Correlative cryo-ET analysis revealed that the contacts between the phagophores and vacuoles may be observed on the side and back of the phagophores (>90%) in yeast. On the other hand, there is almost contact at the curved rim of the phagophores with ER or Lipid droplet. The contacts between the phagophores and vacuoles may be observed on the side and back of the phagophores (>90%), whereas there is almost no contact at the highly curved rim of the phagophores (Bieber et al., 2022). The phagophores have locally projecting membrane peaks 3.6–33 nm in height and 16–32 nm in width (Bieber et al., 2022). Detailed analysis of organelle shape either side of an MCS which has its morphology preserved by cryo-EM/ET can be used to identify where forces are transmitted in vivo, which in turn indicates a strong contact.
Conclusion and Perspective
The main advantage of MCS analysis by EM is that it not only produces images with high resolution but also allows detailed measurement of distances between organelles. However, it may be that the intermembrane gap between membranes may vary depending on the expression of different tether molecules. In addition, the question remains as whether all the electron-dense bridging structures we have observed connecting organelles are tethers. It is expected that not only immuno-EM, but also CLEM techniques will facilitate further detailed MCS analyses.
Footnotes
Acknowledgment
Author Contributions
The authors contributed to this review.
Consent to Participate
Not applicable.
Consent for Publication
Not applicable.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethical Considerations
This article does not contain any studies with human or animal participants.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
