© Iwona Adamus
The immune system uses a sophisticated profiling method to identify and subdue foreign pathogens. Using small protein fragments from viruses or bacteria, antigen-presenting cells (APCs) prime armies of lymphocytes to look for matching traits elsewhere in the body. Whether APCs find such proteins within their cytoplasm or phagocytose them from outside usually determines which processing pathway is used and what kind of lymphocytes are called to service. But sometimes antigens switch tracks, moving from endocytic compartments to the cytosolic pathway. This is termed cross-presentation, and how proteins breach cellular barriers to access the inside track has been a big question in immunology.
A surprising and appealing answer came in 2003 with the publication of this issue's Hot Papers.12 The work was inspired by evidence from Michel Desjardins' lab at the University of Montreal showing that components of the endoplasmic reticulum (ER) are delivered to phagosomes during...
Data derived from the Science Watch/Hot Papers database and the Web of Science (Thomson Scientific, Philadelphia) show that Hot Papers are cited 50 to 100 times more often than the average paper of the same type and age.
In 2002, Desjardins proposed that the ER participates in forming the nascent phagosome, delivering both membrane and crucial components to transport proteins across the membrane.3 Proteomics analysis on a highly purified preparation of phagosomes revealed the presence of various key players: ER translocators, polyubiquitinating machinery of the proteasome, and major histocompatability complex (MHC) class I molecules. "All of a sudden," says Desjardins, "you bring all that you need for antigen cross-presentation in the phagosome." His working hypothesis: "The phagosome can be a self-sufficient organelle," says Desjardins, processing exogenous molecules and loading them onto MHC class I.
To confirm, investigators force-fed antigenic material such as ovalbumin to cultured APCs; phagocytosis is stimulated by coating the material onto latex beads. Using various visualization techniques, Desjardins and colleagues essentially took snapshots of the antigen along its cross-presentation journey.1 They identified it on the cytoplasmic side of the phagosome and showed that it was polyubiquitinated, evidence that the phagocytosed antigen was indeed crossing membranes and gaining entry to the cytoplasm. Peptides from ovalbumin were also found on MHC class I molecules.
A complementary paper from Sebastian Amigorena's group at INSERM in Paris showed that antigens not only get out of the phagosome, but they also come back in.2 Peter van Endert, a coauthor also at INSERM, says, "The ER-phagosome is perfectly able to transport antigenic peptides via the [cytosolic] pathway, and then to assemble it with a class I molecule." In addition, says van Endert, the process proceeds "more efficiently if the source protein was also in the phagosome."
"So it's a complex model we're proposing," says Desjardins. "Start in the lumen, go out to be further processed, then go back in the lumen." Soon after the two hot papers were published, Peter Cresswell's lab at Yale University reported that pinocytosis also engaged ER to allow cross presentation.4 "That's the nice thing," says Desjardins, "that three independent labs were coming to the same conclusion."
In the widely accepted direct presentation model an endogenous antigen in the cytosol is degraded by the proteasome. Antigenic peptides are transported to the lumen of the endoplasmic reticulum by the transporter for antigen processing (TAP) and loaded onto major histocompatibility complex (MHC) class I molecules. In the still controversial model of cross presentation, exogenous antigens are endocytosed and transported out of the phagosome into the cytosol, possibly by the Sec61 complex. In the cytosol, the antigen is degraded by the proteasome and then taken back up by an ER–phagosome hybrid, which contains MHC I molecules and components of the peptide loading complex, including TAP. (N.S. Wilson, J.A. Villadangos,
Of course, key pieces of the model remain obscure. Van Endert lists four unknowns: the identity of the translocator, how the proteasome associates with the phagosome, where the MHC class I molecules are coming from, and once loaded with antigen, how those class I molecules travel to the cell surface. And many are uncertain how this will translate in vivo, says Laurence Eisenlohr, an immunologist at Thomas Jefferson University in Philadelphia. "It's the old problem of: Just because it can happen, doesn't mean it does happen."
AN ER SCRAP
Cross presentation is not just an academic question. Because the inside track loads antigens onto MHC class I molecules, cytotoxic T lymphocytes (CTL) are called to service. This cell-mediated search-and-destroy mission, with CTLs lysing target cells, is thought to be reserved for infected APCs. If such power could be harnessed for vaccines and immunotherapies, it would mean major advances against infectious disease and cancer. The key is understanding how potentially antigenic molecules can access the inside track without intracellular infection. "It's a problem that requires a lot of cell biology and immunology," says Herman Eisen, professor emeritus at the Massachusetts Institute of Technology. "It's at the intersection of the two."
But the two disciplines haven't quite seen eye to eye on the cross presentation solution. Immunologists knew exogenous material was getting to the cytosolic pathway, and the presence of ER provided a retrotranslocation mechanism to accomplish that. Cresswell says, "I think people who were interested in this immunological problem immediately realized that here was a solution."
But while immunologists toil to address the details and to fill out the model, cell biologists remain disturbed by the idea that the ER fuses with the plasma membrane during phagocytosis. "I'm not really so sure the basic assumption upon which all this is based has really been that well established or tested," says Ira Mellman, a cell biologist at Yale University. The immunology community has charged ahead, he says, looking for functions of a presumed ER-phagosome that he isn't sure exists. Recently, Mellman, Sergio Grinstein of the University of Toronto, and two European groups have provided more substance for this skepticism.5
ER components observed in phagosomes using standard cell fractionation techniques are contaminants. Mellman says it's an uphill battle to purify "a low-abundance organelle" such as phagosomes (0.01% of total membrane) from "a ubiquitous membrane" such as ER (50%-80% of total membrane). One can never really get rid of it all, "even if you purify your stuff 1000-fold," he says.
Cresswell counters the contamination argument with kinetic data reported by Desjardins. With time, ER proteins disappeared from phagosomes and were replaced by lysosomal proteins. So given that ER contamination is always possible, "that you had this change in composition with time," he says, "That was the thing that convinced me."
Mellman and Grinstein used multiple techniques, from live-cell video imaging to quantitative high-resolution immunoelectron microscopy. They also repeated some of Desjardins' experiments. Evidence of an ER enzyme lining the phagosome, "was the most incontrovertible evidence that the ER was really there and really important," says Grinstein. But when he and Mellman repeated the experiment, they couldn't detect the enzyme. So, says Grinstein, if some evidence may result from contamination and other evidence is not repeatable, "then the theory is on shaky ground."
"We could find absolutely no evidence that there was an obligatory role for the ER in the initial uptake process," says Mellman. "That undoes the basic assumption."
Amigorena says it is critical to learn how the ER gets into the phagosome, suggesting that the direct fusion of ER with the plasma membrane is under dispute, but not the existence of ER proteins.
Cresswell argues that it's a matter of scale. "The other explanation is that yes, there's ER membrane there, but it's not anything like the amount that the Desjardins paper suggests," he says. Thus, ER components may be difficult to detect using quantitative methods. Measuring activity, on the other hand, is more sensitive than measuring molecules, he says.
Still, Cresswell admits, "It's a little bit up in the air. There's this data that says it seems to be going on, and then there's this negative data that says, if this is going on we should be able to see it and we can't." For his part, Mellman concurs. "I don't think our paper settles the issue. But it certainly brings us back down to a point where I think we have to be very, very careful about what assumptions we make."