Antigen cross‐presentation: proteasome location, location, location

Abstract

Our understanding of the mechanisms by which peptides from proteins present in phagosomes and endosomes are processed and presented on MHC class I molecules, in a pathway called cross‐presentation, is still incomplete. One of the main questions arising from currently proposed models is how do proteins in the phagosome lumen reach the proteasome in the cytoplasm to be processed properly. In this issue of The EMBO Journal, Sengupta et al (2019) present evidence for a surprising turn of events where, in fact, the proteasome acts within the lumen of endosomes and phagosomes.

A critical function of the immune system is to distinguish “self” from “non‐self”, so that we are efficiently protected against infection by foreign microorganisms including bacteria, viruses, and parasites, while avoiding autoimmune responses. Furthermore, the immune system can be viewed as a quality control entity distinguishing normal from aberrant cell forms. Accordingly, infected cells and cells modified when their usual pattern of protein expression is altered during diseases such as cancer are recognized as aberrant and destroyed. The ability of the immune system to recognize such cells is made possible by the constant sampling and processing of cellular components, enabling the presentation of peptides on major histocompatibility complex (MHC) molecules at the cell surface. The display of these peptides informs the immune system about the “health” status of each cells, so that they can be eliminated when they deviate from their proper forms.

Two main pathways of antigen presentation have been established. Endogenous proteins, the ones made up by cells, including viral encoded proteins translated upon infection, are processed into peptides by the proteasomes in the cytoplasm. These peptides are then translocated into the endoplasmic reticulum (ER), through the TAP transporters, where they are loaded on MHC class I molecules for cell surface presentation. The display of “unusual” peptides through this pathway alerts a subtype of cytotoxic T lymphocytes, the CD8⁺ T cells, which recognize and kill cells in the organism displaying these peptides. On the other hand, materials originating from the extracellular milieu, such as microorganisms and neighboring cell debris, internalized through endocytosis and phagocytosis, are processed by proteases present in the endo‐lysosomal system and presented on MHC class II molecules, mainly engaging a CD4⁺ T‐cell response and the production of antibodies by B lymphocytes. The apparent segregation between the MHC class I and class II pathways would, in principle, limit the immune system to a single type of response, either humoral or cytotoxic, potentially weakening the destruction of infected cells or cells engaged in a tumorigenesis path. This is where cross‐presentation comes to the rescue. This process, originally described by Bevan (1976), enables components internalized from the external milieu (which would normally be presented on MHC class II molecules) to be presented on MHC class I molecules. This allows, for example, to engage a combined and much more efficient CD4⁺ and CD8⁺ T‐cell response during bacterial infections. The characterization of the mechanisms allowing cross‐presentation has been the object of an intense quest by scientists in the field of immunology. As we will see below, key aspects of this pathway remain to be fully understood. At this point, let us emphasize the fact that the proteasome is at the center of various models proposed to explain cross‐presentation.

Indeed, proteasomal processing is a key step to generate optimal peptides for MHC‐I antigen presentation. The presence of the proteasome in the cytoplasm ensures a rapid access to the soluble proteins of the cytosol. How proteins present in endosomes/phagosomes reach the proteasome for cross‐presentation is more complicated and still mostly unknown. The group of Kenneth Rock has shown that proteins can be “translocated” from the phagosome lumen to the cytoplasm in a pathway rightly named “phagosome‐to‐cytosol” pathway (Kovacsovics‐Bankowski & Rock, 1995; pathway 1 in Fig 1). The precise mechanisms driving this process are still a matter of debate. For the time being, scientists are often simply drawing an arrow on schematics and cartoons to indicate the transfer of proteins from the lumen of these organelles to the cytoplasm. Once in the cytoplasm, the proteins are believed to be handled by the proteasomes through the classical MHC I pathway. Because the proteasome is present in the cytoplasm, this pathway clearly involves three distinct cellular locations, the phagosome lumen, the cytoplasm, and the ER. The finding that the ER is involved in phagosome biogenesis (Gagnon et al, 2002) paved the way for an alternative model where two of the three aforementioned elements, the ER and phagosomes, form a hybrid organelle where direct processing of peptides and loading on MHC I molecules occurs (Ackerman et al, 2003; Guermonprez et al, 2003; Houde et al, 2003). The third element, the proteasome, is closely associated with the cytoplasmic side of the phagosome membrane (Houde et al, 2003; see pathway 2 in Fig 1). This model still requires a protein translocation step from the phagosome lumen to the cytoplasm to be fully functional.

Figure 1. Models for the location of the proteasome in cross‐presentation.

The proteasome can be positioned in various locations within the cell to accommodate the need to process large polypeptides into the smaller peptides that fit best on the groove of MHC I molecules. (1) In the phagosome‐to‐cytosol pathway, large polypeptides are translocated from the lumen into the cytoplasm by an unknown transporter to reach the proteasome. Antigenic peptides enter the ER lumen through the TAP transporters of the classical MHC I pathway. (2) The proteasome is closely associated with the phagosome membrane (in the cytoplasmic side). Large polypeptides translocated as in 1 reach the proteasome and the peptides are moved back into the phagosome lumen through TAP to bind MHC I molecules. (3) The positioning of the proteasome in the phagosome lumen has the advantage to enable a very efficient system to generate peptides binding directly to MHC I molecules in the lumen (acquired through ER–phagosome interaction). This processing pathway, which is more efficient than the potential processing by hydrolases, would take place in a TAP‐independent process.

Sengupta et al (2019) introduce a surprising new concept stating that proteasomes are present in the lumen of endocytic and phagocytic organelles. They used immunoelectron, confocal, and time‐lapse microscopy, to confirm the presence of multiple subunits of the proteasome in the lumen of endosomes and phagosomes. Clearly, the key question is whether the proteasome is active in the lumen of these organelles. They used two approaches to address this question. They isolated phagosomes and solubilized their membrane (to which the proteasome could be associated, see Fig 1), leaving the latex bead that they contain. They then showed that the proteasomes stuck on the beads were able to cleave a fluorogenic substrate. Using a second method, they showed that ovalbumin present on opsonized latex beads in the lumen of phagosomes was cleaved by the proteasome after removal of the phagosome membrane. The development of assays in living cells to monitor the activity of the proteasome within the phagosome lumen should provide further evidence in the future.

The contribution of the ER to phagosome biogenesis and the presence of the proteasome in the lumen allow to propose a model where phagosomes are fully functional for cross‐presentation. This model precludes the need to translocate proteins across the phagosome membrane for processing. One of the main criticisms for such an integrated pathway is based on the fact that cross‐presentation is inefficient in cells lacking the TAP transporters, suggesting that processing has to occur in the cytoplasm prior to loading in the ER or phagosome lumen (see pathways 1 and 2 in Fig 1). The study of Sengupta et al (2019) addresses this issue by showing that one of the consequences of knocking‐out TAP1 is to decrease the availability of MHC I and its binding partner β2‐microglobulin. In this context, expressing β2‐microglobulin in bone marrow‐derived dendritic cells lacking TAP1^−/− increased the cross‐presentation of phagocytosed ovalbumin. They also showed that the hydrolytic properties of the endo‐lysosomal pathway can be modulated by altering phagosome maturation to increase cross‐presentation in the absence of TAP, suggesting that mechanisms exist to favor the production of peptides over the complete degradation of proteins into amino acids in these organelles. Thus, it is likely that proper conditions for proteasome activity and the loading of peptides on MHC class I molecules exist at some point during the maturation of early phagosomes into phagolysosomes to maximize the contribution of these organelles to cross‐presentation.

Future studies will have to address whether the new pathway for cross‐presentation proposed here is used in all antigen‐presenting cells and how it is finely regulated. The relative contribution of the proteasome in the cytoplasm associated with the phagosome membrane or present in the lumen will also have to be determined. More importantly, the mechanisms by which proteasomes reach the lumen of endosomes and phagosomes have not been addressed here and will have to be elucidated. Arguably, the fusion between phagosomes and autophagosomes could allow such a transfer. Fusion events between these organelles have been involved in the vacuolar processing and presentation of endogenous proteins on both MHC class I and class II molecules (Paludan et al, 2005; English et al, 2009). Interestingly, autophagy would not only capture proteasomes but also the ubiquitinating enzymes required for proteasomal degradation. Within the last 20 years, the view that we have on the phagosome has evolved from a simple organelle derived from the plasma membrane into a complex structure modified through series of interactions with other organelles to yield a self‐sufficient compartment specialized in antigen presentation. Integrating the proteasome in the phagosome lumen is perhaps the missing link.

The EMBO Journal (2019) 38: e102799