This is the authors' abstract. We don't add key points for this paper.
The membrane-associated RING-CH (MARCH) family of E3 ubiquitin ligases is thought to perform predominantly immunoregulatory functions by controlling the localization and abundance of a diverse array of immune receptors. However, the physiological substrates of the MARCHs remain elusive. In this issue of Immunology and Cell Biology, the group of Evelina Gatti describes a novel function for MARCH9.1 The location where this enzyme carries out ubiquitination, its substrates and the functional implications of its activity are unexpected, but provide potential explanations for the ability of MHC class I molecules (MHC I) to perform cross-presentation. To carry out their function, membrane receptors must be located in the right subcellular compartment.2 This can be the plasma membrane, endosomes, the endoplasmic reticulum (ER) and so on. Some receptors function predominantly in only one location, for example, the plasma membrane in the case of the T or B-cell receptors, or endosomal compartments in the case of Toll-like receptor 3, 7 or 9. The role of other receptors requires them to visit and carry out specific tasks in more than one location. MHC I molecules are a case in point: their function is firstly to bind antigenic peptides, and secondly to present those peptides to T cells, but they perform these activities at two distinct locations. Acquisition of peptide cargo occurs in the lumen of the ER, but the resulting MHC I–peptide complexes have to leave this compartment, traverse the Golgi and become exposed on the plasma membrane to present the peptides to T cells (Figure 1).3 Furthermore, MHC I molecules can also enter endosomes, exchange peptides acquired in the ER for new peptides generated in the endosomal lumen and then traffic to the plasma membrane. This is one of the mechanisms that enable cells to perform ‘cross-presentation’, defined as the presentation by MHC I molecules of peptides derived from extracellular proteins internalized by endocytosis (Figure 1).4 It has been generally assumed that the cross-presenting MHC I molecules access endosomes passively by an indirect route, namely their inclusion in plasma membrane regions that form endocytic cups that eventually become the limiting membrane of endosomal compartments. This is the mechanism employed by the vast majority of plasma membrane proteins to access, and undergo degradation in, lysosomes at the end of their life cycle. Rigotti et al.1 now show that MHC I molecules can access the endocytic route directly from the Golgi, but such molecules must be ‘tagged’ in their cytosolic tail by MARCH9-mediated ubiquitination. Ubiquitination—the covalent addition of one or more units of the protein ubiquitin to other proteins—is one of the major mechanisms that enable cells to regulate the location and turn-over of proteins.5 Thus, cytosolic ubiquitinated proteins are degraded by the proteasome6 or autophagy.7 Membrane proteins can also be ubiquitinated in their cytosolic tails, ‘tagging’ them for delivery to endosomal intra-luminal vesicles by the endosomal sorting complexes required for transport (ESCRT) system and, subsequently, lysosomal degradation.8 The MARCH family appears to be specialized at ubiquitinating immunoregulatory membrane proteins. Of its 11 members, all but two (MARCH7 and 10) are themselves anchored to membranes.9 The best characterized are MARCH1, which ubiquitinates MHC II10 and CD8611 in bone marrow-derived cells, and MARCH8, which does the same to MHC II but in non-hematopoietic cells.12, 13 Expression of MHC II and CD86 increases 5- to 10-fold in cells deficient in MARCH1, illustrating the role of ubiquitination on membrane receptor expression and function. Studies analyzing cell lines overexpressing MARCH1 and MARCH8 have proposed other receptor substrates,14 but most if not all of these are not ubiquitinated in cells expressing physiological MARCH levels (Mintern and Villadangos, unpublished data). Similarly, MARCH9 has also been proposed to ubiquitinate a range of receptors, but how many of these are ubiquitinated in physiological conditions remains uncertain.14 Rigotti et al.1 wondered whether MHC I might be a MARCH9 substrate. This was not a random choice. MHC I is a target of herpesvirus homologs of the MARCH family, K3 and K5.15 Virus-infected cells suffer MHC I ubiquitination in the ER, which targets the molecules for destruction via the ER-associated degradation (ERAD) pathway. The concomitant reduction in MHC I antigen presentation contributes to the ability of the virus to evade immunity. However, nobody had yet shown MHC I ubiquitination in normal, uninfected cells. This was the first significant finding of the Rigotti et al.1 manuscript: MARCH9 ubiquitinates MHC I.1 The authors provide an explanation as to why the reaction had not been observed previously. The MARCHs are expressed at very low levels, but MARCH9 expression increases in cells activated by pathogen-associated signals, especially in dendritic cells. Surprisingly though, MARCH9 ubiquitination of MHC I does not occur in the ER and does not cause degradation. Instead, it affects a proportion of MHC I molecules en route to the plasma membrane at another site where membrane protein ubiquitination is known to take place, namely the trans-Golgi network.8 The cohort of MHC I molecules tagged by ubiquitination are delivered to endosomal compartments, and the authors provide evidence supporting the notion that such MHC I molecules contribute to cross-presentation.1 In addition, they show that the same mechanism assists delivery of the MHC I-like molecule, CD1a to endosomal compartments for acquisition of antigenic cargo, which in this case does not consist of peptides but lipids.1 The authors conclude that pathogen detection upregulates a MARCH9-mediated mechanism for enhancement of MHC I cross-presentation and CD1a lipid presentation. The results in this manuscript reveal MARCH9 as one more component of the complex machinery that regulates antigen presentation and opens a series of interesting questions for immunologists. What is the relative contribution of this mechanism to cross-priming and CD1a-mediated immune responses in vivo? Is this pathway used specifically in some cell types and not others? Is MARCH9 upregulated in response to any type of immunological challenge, or only in response to particular types of infection? Are there more MARCH9 substrates, and do these also have immunological functions? Can the answers to these questions be exploited to inhibit or enhance this pathway for therapeutic purposes? Cell biologists will also ask some questions posed by this study. Why are the substrates ubiquitinated by MARCH1 or MARCH8 targeted for endosomal destruction, but those ubiquitinated by MARCH9 are tagged for delivery to endosomes but not degradation? Is it because the type of ubiquitination mediated by these closely related E3 ligases is different? Characterization of the MARCH family has only just started; description of their true substrates and functional consequences of their activities is likely to yield many more surprising and fascinating findings in the future. The authors declare no conflict of interest. Peptide antigens generated in the cytosol by proteasomal degradation are loaded on MHC I molecules in the ER lumen. The MHC I–peptide complexes follow the default secretory pathway through the Golgi to the plasma membrane to display the peptides to T cells. Rigotti et al.1 show that a cohort of MHC I molecules are ubiquitinated in the trans-Golgi network by MARCH9. The ubiquitinated molecules are delivered to endosomes, where they can bind peptides derived from extracellular proteins degraded by endosomal proteases and cross-present these on the cell surface.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Mintern et al. (2017) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: