Monday, December 19, 2011

Too good to be to true

Currently, Foxp3+ CD4 T cells, also known as regulatory T cells (Tregs), are most studied and still least understood T cell subset. Why is that? The main challenge to Tregs/suppressor concept is that they do not fit in any mainstream theoretical framework of immune system organization. In absence of such working model, translation of basic Tregs research into clinical practice will look as sporadic efforts with no real progress and consensus.

If you are interesting to know more about Tregs, I will recommend reading following article published in journal Nature few weeks ago. This study by Michael Rosenblum et al. (1), exactly represents that confusion typical to Tregs research. First, this is a fairly simple study and we may applaud Nature for taking step to publish a research paper that does not include highly expensive (in many cases completely irrelevant or useless) experimental methods. Actually I see some pattern in this direction from Nature's editorial board. Nature started to publish simple papers (the most recent example of this kind is a research article about dendritic cells and lymph node-specific high endothelial venules). However, simplicity is not the same as significance or relevance.

In this new article by Michael Rosenblum et al., the authors created mice that will express OVA antigen in the skin only upon treatment with a doxycycline and crossed them to the mice transgenic for OVA-specific T cell. In absence of doxycycline, this modified mice have healthy skin, and harbor substantial population of OVA-specific Tregs (due to doxycycline-independent thymic expression of OVA). Within 5-10 days upon doxycycline treatment, the skin becomes highly inflamed due to OVA-specific T cell effector infiltration. However, within 40-60 days skin inflammation is completely resolved. How? The authors suggested that it has to do with Tregs, because Tregs depletion led to more severe skin inflammation. The problem is that even in condition of Tregs depletion, the skin inflammation is still resolved and almost within the same time frame as in presence of Tregs (Fig 2a, e, g). In general, Tregs depletion with anti-CD25 antibody may not eliminate all Tregs, so this particular experiments are not optimal approaches. This leave open the main question what is the role of Tregs in this process. In Fig. 4E, f, the authors showed that when expression of OVA is turned off for 20-30 days and then turned on again, skin inflammation that developed 2nd time is less severe and is resolved at faster rate compared to initial skin inflammation in the same mice. The authors showed that after resolution of initial skin inflammation, there were more Tregs retained in the skin and these Tregs were more potent in suppressing naïve T cells proliferation. The authors conclusion is that tissue (skin) maintains memory Tregs specific to initial antigen and their presence is required to resolve skin inflammation.

Essentially, the authors idea is that similar to effector T cells, Tregs also undergo the same phases of a typical immune response: activation, expansion, memory differentiation. In my opinion, the data to support this claim are weak. First, Tregs depletion should have been done with Foxp3-DTR mice to selectively deplete Foxp3+ T cells and second, no data is provided to show that initial OVA expression and inflammation does not select for skin epithelial cells expressing less of OVA that may have explained reduced inflammation upon re-expression of OVA in the skin.


Tuesday, December 13, 2011

does stress induce allergy?

Here is another example of what I call “right paper wrong journal”. This time we should thank journal Science for not disappointing us.

Type I allergic reaction is characterized by production of antigen-specific IgE. Upon re-encounter with the corresponding antigen, IgE molecules engage receptors on mast cells, eosinophils or basophils that results in release of histamine or other active enzymes. These substances induce smooth muscle contraction, capillary permeability, mucus production, leading to clinical signs of asthma, skin rash, itching, mucosal swelling. It seems that IgE responses are bad. So why we have it? In general, evolutionary explanation for IgE response was its role in defense against intestinal parasites (worms). However, today very diverse set of antigens (for example, some food) could induce IgE response. How they do it is not at all clear.

The study I was referring to earlier was recently published in journal Science (1). This paper by Jessica Strid et al., examined an effect of innate T cell activation on IgE response. This innate T cells, called γδ T cells express NKG2D molecules that recognize tissue stress-induced ligand, Rae-1. By analyzing mice where expression of RAE-1 could be manipulated (on/off), the author showed that if skin application of nominal (neutral) antigen, OVA, was coincided with NKG2D-Rae-1 interaction, this led to allergy-prone OVA-specific IL-13 and IgE production from adaptive immune system. They showed that the imprinting of this allergy-prone responses were reduced in γδ T cell-deficient mice, NKG2D-deficient mice or in mice deficient in only subset of γδ T cell residing in mouse skin. The authors concluded that tissue-stressed induced innate T cell activation promotes allergy-prone IgE adaptive immune response.

This paper is very good paper for journal of immunology, for example. It's main strength is the idea that links environmental toxins (chemical pollutants) to allergy development. It's fundamental weakness is its reliance on IgE or IL-13 production as a sole signature of allergic response. They do not provide any data that would suggest that upon antigen re-encounter, this increase in OVA-specific IgE production, as observed in this study, could induce any clinically relevant sign of allergy. Shortly, this study lacks "science-level" relevance.

I personally find it difficult to believe that tissue-surveillance program through NKG2D-Rae-1 interaction should lead to IgE response. After all, tissue-surveillance program was originally thought to be primarily directed against tumors, where protection is mediated  through IFN-γ pathway, not IL-13/IgE pathway.


Sunday, December 4, 2011

trust but verify

Innate immune system detects the pathogen invasion and alerts adaptive immune system to its presence. However it is not always clear whether adaptive immune system should be alerted to the presence of any dose of pathogenic inoculum. May be innate immune system can handle low dose of pathogenic inoculum alone? However, what if innate immune system tries to handle the pathogen alone, fails doing it and alerts adaptive immune system too late when pathogenic onoculum become too large? The exact balance that will increase protective immunity and at the same time reduce tissue damage will depend on many variables, many of them still unknown.

If you are interested to know more about the complex interaction between innate and adaptive immune systems, I will recommend reading the following research article published in Nature Immunology a few weeks ago. In this study by Nadine Honke et al. (1), the authors first showed that in mice, intravenously injected (i.v.) vesicular stomatitis virus (VSV) is rapidly captured and cleared from the bloodstream by tissue macrophages. However, interestingly, the authors could detect the presence of live virus only in spleen tissue. Specifically, live virus was maintained in specialized macrophages called marginal zone macrophage that are identified by staining for CD169 marker. In contract, live virus could be recovered from all tissue examined from wild-type mice depleted of macrophages or from mice deficient in IFN-α receptor. These results suggested that CD169+ macrophages have some unique properties that allow them to maintain live virus in wild-type mice. The authors showed it may be related to higher expression of IFN-α signaling inhibitor Usp18 in CD169+ macrophages as compared to other tissue macrophages, for example, F4/80+ macrophages. Indeed, expression of VSV was absent in CD169+ macrophages from Usp-18 deficient mice. In other words, CD169+ macrophages “intentionally” inhibit IFN-α signaling in order to maintain live virus. Why is that? The authors explained this puzzle by observation that VSV-specific T cell and IgG response were reduced in Usp-18 deficient mice. Moreover, Usp-18 deficient mice became very susceptible to i.v. Injected VSV. Usp-18 deficient mice harbor far less live virus in the spleen at early time point post infection (at 7h) and harbor far more live virus in brain at late time point post infection (at 7 days). In addition, the authors showed that live virus primed VSV-specific adaptive immune response more efficiently compared to inactivated VSV. The authors concluded that “deliberate” maintenance of replication-competent virus in CD169+ macrophages gives enough time or antigenic material for efficient priming of protective adaptive immune response, while inactive virus is far inferior in providing signals necessary for efficient priming of protective adaptive immune response.

However, there are several holes in the story. The authors did not show results whether (1) VSV-specific T cells response was reduced in mice depleted of CD169+ macrophages or (2) whether mice depleted of CD169+ macrophages became more susceptible to VSV. Without these controls the whole concept of the paper will be misleading.