NK Memory in Contact Hypersensitivity
First described (to my knowledge) in Nature on May 2006, Ulrich von Andrian's group was experimenting in hapten-induced contact hypersensitivity (CHS) when they noticed a recall response in the absence of B and T cells (via a RAG2 knock-out mouse). They pinpointed NK cells as the source and performed follow-up experiments to support NK cell memory
Introduction
Contact hypersensitivity (such as responses to poison ivy) occurs when the immune system is primed against foreign antigen, inducing specific T and B cell responses (T cells are more important). Upon secondary encounter, memory T cells are activated and begin secreting cytokines such as interferon gamma. These cytokines in turn activate macrophages and other cells, which ultimately induce inflammation: pain, heat, swelling, redness (and the godawful itch).
Contact hypersensitivity can be measured in a variety of ways, but in the von Andrian's paper, it was mostly measured by ear-swelling and by identifying the type and number of cellular infiltrates in the ear.
The authors' used three different chemicals in order to induced CHS. In most of their experiments, they used 2,4-dinitrofluorobenzene (DNFB), but oxazolone (OXA) and picryl chloride was also used. These are small molecule haptens which react with cellular proteins to create new, foreign epitopes for which the immune system can respond.
Secondary-response to DNFB in RAG2-/- mice
The authors' original intent was to examine various tissues in skin-induced CHS. They were characterizing the role of T cells in hypersensitivity of the bladder when they discovered that RAG2-/- mice have a recall response to antigen: furthermore, the magnitude of the response was similar to WT mice. They wanted to know if hypersensitivity in the bladder was T cell dependent, but what they found is that other cell types might have memory-like properties.
Their work in the bladder gave an indication of what cell type might be involved. In WT sensitized mice, T cells predominate in the infiltrate. However, NK cells are a close second. It should also be mentioned that PMNs [mostly neutrophils] are also high; but as they live for only a few days, these can be ruled out. But before they could begin to work on NK cells, they had to prove that antigen specificity actually occurs.
Antigen specificity of recall response in CHS using RAG2-/- mice
The "recall response" could have been nothing more than innate inflammation that was stronger the second time around (given the environment produced a month previously). In order to show antigen specificity, mice were sensitized to one hapten, like DNFB, and then the recall response to an unrelated antigen (OXA) was assessed. In both WT and RAG2-/- mice, increased ear swelling was shown only when the mouse was primed, and then boosted with, the same antigen. This satisfies one requirement of memory-antigen specificity, and suggested the involvement of a memory-like cell in the absence of a B or T cell response.
Increase in NK cell numbers after challenge
As mentioned previously, NK cells were shown to be present in large numbers in CHS, second only to T cells. The numbers of NK cells were assessed both during sensitization and after challenge. As expected, in RAG2-/- mice, there was an increase in NK cells after challenge with hapten, as measure by flow cytometry using the markers NK1.1 and CD45. Immunofluoresence also shown increased numbers of NK cells. But again, it wasn't clear if the increase in NK cells are a byproduct of the model or the mediators.
Hypersensitivity in the absence of NK cells?
If NK cells are the population mediating this recall response, RAG2-/- mice that have a defect in NK cells shouldn't have any secondary response. Thus, the authors chose two different mouse models in which to perform their studies, only one of which I'll talk about here. RAG2-/- mice were crossed with mice lacking the common gamma chain, resulting in a double knock out mouse (RAG2-/-IL2Rgamma-/-). The common gamma chain is shared by a family of cytokine receptors and results in defective responses to multiple cytokines, some of which are also important for NK cell development. Thus, this double knockout has no B, T, and NK cells. As expected, without NK cells the challenge response to haptens was greatly reduced (when compared to a RAG2-/-). This experiment increases the argument for NK cell memory, but there are a few caveats: other cell types might also use these cytokines and there is probably reduced activation in these cells. Therefore, in addition to using another model, SCID x beige mice, further experiments had to be performed.
NK cells were also depleted using specific antibody(such as NK1.1) Again, when compared to WT secondary responses to haptens, inflammation in RAG-/- mice was much lower.
Adoptive transfer of naive and sensitized NK cells
Because other cell types might have been affected by lacking the common gamma chain, adoptive transfer assays were used to confirm NK cells as the mediator of this memory-like phenotype. Compared to naive NK cells or sensitized cells depleted of NK cells, previously sensitized NK cells adopted into naive mice had greater hapten-specific ear swelling.
Other
The authors also perform other experiments in order to narrow-down the exact subset of NK cells mediating this effect. However, this isn't as convincing. There's nothing wrong with the data, which narrows down the NK cell population to that in the liver, but not enough was done to confirm that the subset actually comes from the liver. For example, activation of NK cells might have increased localization to the liver. Or, because they were only comparing NK cells between lymph node, liver, and spleen, they might have missed an increase in the skin population.
Likewise, the attempt to characterize the NK population using expression of activating receptors (via flow cytometry) has the same problem. Without knowing the exact receptor the NK cell uses to respond to the hapten, it's like looking for a needle in a haystack.
Conclusion
Based on these experiments, it was clear NK cells mediate a memory-cell like function. The authors demonstrate increased recall responses after 4 weeks in RAG2-/- mice, antigen specificity(in the broad sense), and adoptive transfer of antigen-experienced NK cells mediating this function.
That being said, there were still many things left to be proved. The lack of a specific receptor-ligand pair hindered quantification of these memory-like NK cells, without which the scientific community was apt to be skeptical. In addition, without a better model (ear swelling isn't the greatest assay in a non-CHS setting) future experiments having to do with pathogen models and vaccine development couldn't be accomplished.
The next paper deals with a specific receptor-ligand pair that was able to extend the knowledge on these memory-like NK cells. In the meantime, checkout the 2006 paper and see for yourself:
O'Leary JG, Goodarzi M, Drayton DL, and von Andrian UH. (2006) T cell-and B cell-independent adaptive immunity mediated by natural killer cells. Nature 7(5): 507-14
Pubmed
http://www.ncbi.nlm.nih.gov/pubmed/16617337
No comments:
Post a Comment