News & Views | Published:


Professional secrets

Looking inside the compartments of certain immune cells — professional antigen-presenting cells — has revealed how the immune system can trigger a cell-killing response to extracellular pathogens.

Our immune system provides effective protection against most pathogens. It can mount a highly specific 'adaptive' response that distinguishes between different sorts of pathogen and activates the appropriate mechanisms to eliminate them. Writing in this issue, Amigorena and colleagues1 and Desjardins and colleagues2 suggest how an adaptive immune response typically reserved for pathogens that replicate inside immune cells can also be generated against pathogens that exist outside these cells.

Macrophages and dendritic cells are specialized cells that initiate adaptive immune responses. They are sometimes called professional antigen-presenting cells (APCs) because one of their primary duties is to degrade microbial proteins and display the resulting peptide fragments, or antigens, on their surface. The kind of immune response that is then produced depends on how the antigens are displayed by the APCs.

In general, peptide antigens from pathogens replicating inside APCs (endogenous antigens) are transported into a cellular compartment, the endoplasmic reticulum (ER), and bind to a protein complex called MHC class I. From the ER, the MHC class I–antigen complex is shuttled to the cell surface, where it stimulates antigen-specific cytotoxic T lymphocytes (CTLs) — killer cells that seek out and eliminate infected cells (Fig. 1).

Figure 1: Professional antigen-presenting cells process intracellular and extracellular pathogens differently.

In the endogenous pathway, proteins from intracellular pathogens, such as viruses, are degraded by the proteasome and the resulting peptides are shuttled into the endoplasmic reticulum (ER) by TAP proteins. These peptides are loaded onto MHC class I molecules and the complex is delivered to the cell surface, where it stimulates cytotoxic T lymphocytes (CTLs) that kill the infected cells. In contrast, extracellular pathogens are engulfed by phagosomes (exogenous pathway). Inside the phagosome, the pathogen-derived peptides are loaded directly onto MHC class II molecules, which activate helper T cells that stimulate the production of antibodies. But some peptides from extracellular antigens can also be 'presented' on MHC class I molecules. How this cross-presentation occurs has now been explained1,2. It seems that by fusing with the ER, the phagosome gains the machinery necessary to load peptides onto MHC class I molecules (see Fig. 4g on page 405).

Antigens from extracellular pathogens are usually processed differently from endogenous antigens, and with different results. Extracellular pathogens are engulfed by APCs and degraded in phagosomes — membrane-bound cellular compartments. Peptide fragments produced in phagosomes are not transported to the ER. Instead, they bind directly to a different set of MHC molecules, MHC class II, inside the phagosome and then return to the cell surface, where they stimulate a subset of helper T cells (Fig. 1). Unlike the CTLs, helper T cells do not kill other cells, but instead produce chemical signals. These trigger the production of antigen-specific antibodies and elicit an inflammatory response that eliminates the microbial pathogens.

So the rules governing antigen presentation seem to be simple — endogenous antigens are presented by MHC class I molecules to stimulate a cell-killing response, and exogenous antigens are presented by MHC class II molecules to stimulate a helper response. But in reality, the situation is not so straightforward. Professional APCs can break these rules and present exogenous antigens on MHC class I molecules3, in a process known as cross-presentation. The cross-presentation pathway seems to be required for CTL responses to certain exogenous antigens4,5. For example, some viruses do not infect professional APCs, but in order to generate a CTL response that will eliminate virus-infected cells, the APC must present exogenous viral proteins on MHC class I molecules. Cross-presentation might also be important for triggering CTL responses to some cancer cells.

So how do exogenous antigens cross over to the MHC class I pathway? Amigorena and colleagues had previously shown6 that exogenous proteins could escape from phagosomes and enter the cytosol (the intracellular space). So it was thought that cross-presentation might simply involve the transport of exogenous proteins from the phagosome into the cytosol, where they could join the pathway used to process endogenous antigens. This view was supported by data7,8 showing that cross-presentation depends on two components of the MHC class I pathway — the proteasome complex and the 'transporter associated with antigen presentation' (TAP). The proteasome complex degrades proteins in the cytosol and TAP shuttles the resulting peptide antigens into the ER for loading onto MHC class I molecules. Furthermore, cross-presentation is sensitive to the drug brefeldin A (refs 7, 8), which disrupts the transport of the MHC class I–antigen complex from the ER to the cell surface9.

Desjardins and colleagues, meanwhile, had provided a clue to how exogenous antigens might escape from the phagosome10,11. In an effort to understand how phagosomes form and mature, these authors had catalogued the proteins contained on phagosomes that had been purified from macrophages. Surprisingly, they detected proteins that were thought to reside only in the ER10. Further analysis revealed that phagosomes fused with the ER during, or shortly after, pathogen engulfment at the cell surface11. This was an important finding because a protein-translocation channel called the Sec61 complex is embedded in the ER membrane. Sec61 both imports newly synthesized proteins into the ER and exports proteins from it, targeting the proteins for degradation by the proteasome12,13. So it was predicted that the Sec61 complex might be delivered to the phagosome after fusion with the ER. If true, it could be the missing link in the cross-presentation pathway — the transporter responsible for exporting exogenous proteins from the phagosome into the cytosol.

The two groups1,2 now provide evidence to support this hypothesis. Looking at phagosomes from dendritic cells1 and macrophages2, respectively, they showed that Sec61 is indeed present on the phagosome membrane. And both groups found that a fluorescently tagged version of the protein ovalbumin could be exported from the phagosome into the cytosol. But did protein export involve Sec61? To find out, Desjardins and colleagues looked at the export of the cholera toxin A1 subunit (CTA1). This protein moves from outside the cells to the ER and is then exported from the ER into the cytoplasm through the Sec61 channel14,15. The authors observed that CTA1 could also be exported from the phagosome into the cytoplasm, which strongly suggests that, after Sec61 is delivered to the phagosome, it can still export proteins.

What happens to proteins once they have been exported from the phagosome? First, they must be turned into peptide antigens by the proteasome. Desjardins and colleages showed that proteasomes are associated with the cytosolic side of the phagosome membrane. And both groups showed that TAP and the MHC class I molecules are delivered to the phagosomes and remain active. Their further analyses revealed that the peptide antigens are loaded onto MHC class I molecules inside the phagosome.

So it seems that after the exogenous proteins are exported from the phagosome through the Sec61 channel, they are degraded by the proteasome and the resulting peptide antigens are shuttled back into the phagosome by TAP. There, they are loaded onto MHC class I molecules on the inside of the phagosome membrane. The results confirm observations16 that the phagosome is a fully competent antigen-processing compartment for the MHC class I pathway.

The new studies provide strong support for a model of cross-presentation in which ER–phagosome fusion occurs. But they do not show that such fusion is necessary. Arguments have also been made that cross-presentation occurs only in cultured cells and might not be relevant in animals17. Now that a molecular mechanism governing cross-presentation has been proposed, experiments to test the importance of this immunological process are certain to follow. But the debate on cross-presentation continues — to resolve it, the professionals will have to reveal all of their secrets.


  1. 1

    Guermonprez, P. et al. Nature 425, 397–402 (2003).

  2. 2

    Houde, M. et al. Nature 425, 402–406 (2003).

  3. 3

    Bevan, M. J. J. Exp. Med. 143, 1283–1288 (1976).

  4. 4

    den Haan, J. M. & Bevan, M. J. Curr. Opin. Immunol. 13, 437–441 (2001).

  5. 5

    Sigal, L. J., Crotty, S., Andino, R. & Rock, K. L. Nature 398, 77–80 (1999).

  6. 6

    Rodriguez, A., Regnault, A., Kleijmeer, M., Ricciardi-Castagnoli, P. & Amigorena, S. Nature Cell Biol. 1, 362–368 (1999).

  7. 7

    Brossart, P. & Bevan, M. J. Blood 90, 1594–1599 (1997).

  8. 8

    Kovacsovics-Bankowski, M. & Rock, K. L. Science 267, 243–246 (1995).

  9. 9

    Nuchtern, J. G. et al. Nature 339, 223–226 (1989).

  10. 10

    Garin, J. et al. J. Cell Biol. 152, 165–180 (2001).

  11. 11

    Gagnon, E. et al. Cell 110, 119–131 (2002).

  12. 12

    Koopmann, J. O. et al. Immunity 13, 117–127 (2000).

  13. 13

    Romisch, K. J. Cell Sci. 112, 4185–4191 (1999).

  14. 14

    Roy, C. R. Trends Microbiol. 10, 418–424 (2002).

  15. 15

    Schmitz, A., Herrgen, H., Winkeler, A. & Herzog, V. J. Cell Biol. 148, 1203–1212 (2000).

  16. 16

    Ramachandra, L., Song, R. & Harding, C. V. J. Immunol. 162, 3263–3272 (1999).

  17. 17

    Zinkernagel, R. M. Eur. J. Immunol. 32, 2385–2392 (2002).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.