Sunday, 10 August 2008

Cross presentation: presenting unexpected antigens to the cell-mediated immune response.

Please cast your minds back to basic immunology. I am sure you remember there is the innate immune response (boring) and the acquired/adaptive immune response (very interesting). Perhaps the most interesting part of it is the gene re-arrangement to generate antibodies and T cell receptors (TCRs). But this re-arrangement is random; how can they tell self from non-self? The very basic explanation is seems to be that antigen (Ag) presenting cells (APCs) shows lymphocytes these Ag in the local draining lymph nodes when the body knows it is ill (ie the innate immune system is acting up). (Sorry for this long introduction to the topic but I need to remind myself of this while I type on the train myself.)

A problem comes when we remember a bit of basic immunology. Peptide Ag presented on MHC class II molecules to CD4+ T cells (T helper cells) are taken up by APCs and are exogenous Ag. However, peptides shown on MHC class I to CD8+ T cells are generated from endogenous proteins. So how can a virus that infects only the liver like hepatitis or the respiratory system (like rhinovirus or flu) show their peptide Ag to naive CD8+ T cells (cytotoxic T cells or CTLs for short) which live in the lymph organs. Unless all viruses also affect the APCs as well then it seems impossible! Let’s forget how ridicules that idea is first and remember CTLs also attack tumour cells expressing mutant proteins. Cross presentation appears to be the answer. Somehow, APCs such as dendritic cells (DC) take up proteins from other cells and process these in a manner that means they are treated like endogenous proteins and therefore processed to be peptide Ag loaded onto MHC class I. This is called cross presentation, when CTLs are activated this way they are cross primed. The fine details of this appear to be missing but a lot of evidence is mounting that shows this must be the case. Other than this I can see no other way apart from the naive CTLs circulating the body and activating their; but no evidence has been found for this that I can see (I am pretty sure this also flies in the face of most accepted ideas in immunology). There are four ways for DC to capture extracellular proteins; (i) endocytosis (ii) pinocytosis (cell drinking) (iii) phagocytosis and (iv) macropinocytosis. For example, phagocytosis can be the uptake of a bacterium or cellular debris such as an apoptotic body. When cells are signalled to undergo apoptosis the dying cell starts to bleb and releases intact fragments of the cell called apoptotic bodies that express ‘eat-me’ signals. It is not a huge leap of the imagination to think this could be a major way tumour proteins are presented to naive CTLs.

So how do these exogenous proteins within sub-cellular compartments such as the endosome or phagosome get loaded onto the MHC class I molecules? It is a good question and various routes have been suggested and evidence for some of these have been found. In DC exogenous proteins have been shown to be exported into the cytosol before they are degraded in the lysosome. This means they can be substrate for the proteasome and broken down to short peptides to be exported across the ER membrane by TAP. When here they can bind to the peptide binding cleft of MHC class I molecules and move to the cell surface like normal endogenous proteins in the classical presentation pathway. Another pathway acts in the endosome. This TAP/proteasome independent pathway uses cathepsin S to generate some peptide Ag. It is unclear whether other peptides here also act but currently appears unlikely. What I find amazing is how this protease produces peptides of the correct size of 8 or 9 amino acids residues long to load onto MHC class I molecules. This pathway even produces the same peptide Ag as the much more complex proteasome based pathway. When both pathways are knocked out most of the cross-presentation in vivo is lost.
Interestingly, tolerance to an Ag can be generated using these mechanisms when no stimulator of the immune system is present, and this has been called cross tolerance. It is now clear cross presentation is not some strange phenomenon like originally thought when discovered but is a key part of the immune system. When you knockout these pathways for cross-presentation none occurs. The next step is to manipulate this with vaccines to stimulate the cell-mediated immune response. Often only antibodies are produced because the antigens are not cross-presented.

A short and sweet review:
Brode and Macary (2004) Cross-presentation: dendritic cells and macrophages bite off more than they can chew!

A full and great review:
Rock and Shen (2005) Cross-presentation: underlying mechanisms and role in immune surveillance.

2 comments:

Catarina Vicente said...

Oh dear, immunology is so interesting but so incredibly confusing. You are very brave James to write a post on it! Now, I have a question, but I apologise if it is a really silly one. Regarding cross presentation, do you mean that DC 'eat' infected cells or bacteria or something of the kind, process it somehow inside and present the resulting antigens on their membrane. Then they travel to the lymphoid nodes, present the antigens to the T cells, and that is how they activate the adaptive immune system without any naive T cells leaving the nodes. Is that the basic idea or am I just getting everything wrong?

James Lloyd said...

You got the point exactly. The reason it is so interesting for immunologists (and me) is because the classical pathway for presenting to CTLs is all about taking cell proteins and adding them to MHC class I. DC and to a lesser extent macrophages eat cellular components or bacteria or take up other sources of proteins and process it so they can present it. When they do take up antigens and are stimulated by pathogens they development and move to the lymph nodes. Then they can bind to the T cells that have TCR that matches the MHC I and peptide on it and is CD8+. The APC also expresses co-stimulators that cause the CTL to become active. I hope that explains it a bit better.