The developmental diversion of thymocytes

I like reports that make me learn something new and appreciate novel developments leading to more integral view of immunological concepts. My understanding of the thymic central tolerance process was that thymocytes receiving strong signals from tissue antigens through TCR undergo invariably the clonal deletion. But it looks like the clonal deletion of cells that can develop into potentially autoreactive T lymphocytes is not the only way which exists for such population in the thymus. I have read the paper that suggests that there may be actually two outcomes for thymocytes with self-reactive TCR – first is the clonal deletion whereas second the developmental diversion.

The link:

What is the developmental diversion, though? According to authors it is a process that happens when a thymocyte gets the signal through its autoreactive TCR but is not able to receive the costimulation with CD28 molecule. In such case it can enter a pool of DN cells (double negative for CD4 and CD8) and turn up in the intestinal epithelium where it re-expresses CD8 (in its αα form). Cells derived from the developmental diversion are anergic and when the clonal deletion is impaired (as for example in CD28 knockout mice) the efficiency of central tolerance is not reduced because autoreactive thymocytes have an substitute pathway that sequesters them from harmful and self-reactive mature population.

How the the developmental diversion was detected? The publication contains a lot of data, so I will focus on most crucial evidence. The initial observation made by investigators was that CD28 knockout mice (and also B7 double knockout with no CD80 and CD86 which are CD28 ligands) has unusually numerous population of DN thymocytes that express TCRαβ. In normal mice DN thymocytes are in their majority TCRαβ-negative. The DN population from mice deleted for CD28 contains also the similar proportion of autoreactive TCRs as pre-selection DP (double positive) thymocytes but mature C4 or CD8 T cells from the same strain are mostly deprived of self-reactive rearrangement. Therefore authors conclude that the clonal deletion of autoreactive thymocytes requires CD28 costimulation and in its absence such cells are diverted into the alternative developmental way.

Investigators follow this phenomenon by studying at what exact stage of thymocyte development the diversion may occur and what happens with diverted thymocytes once they leave the thymus (they end up in the intestinal epithelium as already has been remarked). The most interesting thing, however, is that the developmental diversion seems to take place in normal mice as well as TCRαβ+CD8ααintraepithelial lymphocytes from the wild type strain are enriched for autoreactive specificities. I definitely need to start following this story.

Pobezinsky LA, Angelov GS, Tai X, Jeurling S, Van Laethem F, Feigenbaum L, Park JH, & Singer A (2012). Clonal deletion and the fate of autoreactive thymocytes that survive negative selection. Nature immunology, 13 (6), 569-78 PMID: 22544394

B cells can secrete IL-6 and drive Th17 response in autoimmunity

The main task of B cells is to release protective immunoglobulins. Yet it is not their only role since they are apparently capable to take on the diverse array of activities that do not directly form a part of effector humoral responses. Instead of just secreting antibodies B cells can influence the outcome of an immune response by dictating the behavior of other cell types. It appears that such mechanism may underlie the development of autoimmunity in the central nervous system. I have found the publication that by combining the work on a mouse experimental model and the analysis of human patients presents the compelling evidence pointing to how B cells stimulate CD4 T cells into the pathogenic phenotype through the antibody-independent action.

The link:

The aim of the paper is to seek a clarification for the unexplained outcomes of various B cell depletion treatments during the course of multiple sclerosis or its mouse model – EAE. For example, the targeted reduction of antibody-secreting plasma cells results in the worsening of disease symptoms. Similarly, the broad B cell depletion leads to the improvement that precedes the drop in the level of autoantibodies. To understand whether B cells may have an additional antibody-independent role in the pathogenesis of multiple sclerosis or EAE investigators focus on IL-6 which is a pro-inflammatory cytokine and the essential factor in autoimmune conditions that afflict the central nervous system. They analyze how B cells can contribute to IL-6 secretion in the context of autoimmunity and assess if B cell-derived IL-6 may influence the conduct of CD4 T cells which are the main factor in multiple sclerosis/EAE development.

Authors show that B cells have the inherent ability to secrete IL-6 when they are stimulated with ligands engaging innate receptors (LPS or CpG plus anti-CD40 antibody) and attempt to prove that such B cells’ capacity may have the physiological importance.  To this end they demonstrate that B cells collected from mice that received EAE-driving immunizations are enriched for IL-6 mRNA compared to B cells from healthy mice. Additionally, the abrogation of IL-6 expression in B cells blunts the severity of neurological symptoms typical to EAE. The final proof indicating for the antibody-independent role of B cells in murine EAE pathogenesis comes from BCDT (B cell depletion therapy) experiments as such treatment is effective in alleviating the EAE-resulting damage only when B cells are competent to release IL-6. On the other hand, knocking out IL-6 expression in B cells does not influence their capacity to secrete antibodies. Investigators also pinpoint that marginal zone B cells are the subset which is most proficient in IL-6 release.

IL-6 is a cytokine involved in the formation of Th17 subset which is known for its participation in pathogenic responses pertinent to the autoimmunity driven by CD4 T cells. Because EAE stands as a representative such disorder, authors ask whether B cell-derived IL-6 could impact the development of Th17 population in the course of this disease. They demonstrate that it is indeed the case since the ablation of IL-6 expression in B cells is able to diminish the propagation of Th17 response during EAE. Remarkably, the B cell-derived IL-6/Th17 axis is operative regardless of antigen specificity as the Th17 population is also less numerous following immunizations with EAE-irrelevant OVA peptide. In the last part of paper devoted to human patients investigators show that the ability of B cells to control Th17 development through IL-6 release may be conserved across mammalian species.

The information that this report contains is obviously important because of its practical value but it also stimulates to ask broader questions. Why we are equipped with signaling pathways like the one described in the discussed report? To what end the apparently pathogenic (in the context of autoimmunity) B cell-derived IL-6/Th17 axis has evolved? Why B cells are so prone to act as autoimmunity mediators when stimulated with TLR ligands (especially nucleic acid-recognizing ligands)? The case of lupus and now that of multiple sclerosis have provided enough evidence for such ostensibly rebellious nature of B cells. Could potentially pathogenic pathways starting with the recognition of TLR ligands by B cells represent an evolutionary trade-off with the better control of gut commensal bacteria as an asset and the danger of autoimmunity as a liability? As a matter of fact MyD88 signaling in B cells has been shown to take part in accommodating the intestinal microbiota by preventing their systemic spread when colonic injury occurs (B Cell-Intrinsic MyD88 Signaling Prevents the Lethal Dissemination of Commensal Bacteria during Colonic Damage: Immunity. 2012; 36 (2): 228-38).

I have one more remark concerning the data that this paper presents. Authors show that CpG (TLR9 ligand) is more efficient that LPS (TLR4 ligand) in driving IL-6 secretion by B cells. Nucleic acid-recognizing TLRs are unusual because they are hidden from the cell surface to endolysosomal compartments. Apart from that, recent publications by Gregory Barton’s group reveal that TLR9, TLR7 and TLR3 (expressed by a macrophage cell line) require proteolytic processing prior to becoming functional detectors. Such feature is interpreted as an additional safety measure since ligands for these receptors are expressed both by host cells and pathogens. I am not really sure if it makes sense but nobody checked if the similar requirement for proteolytic cleavage exists in B cells.

Barr TA, Shen P, Brown S, Lampropoulou V, Roch T, Lawrie S, Fan B, O’Connor RA, Anderton SM, Bar-Or A, Fillatreau S, & Gray D (2012). B cell depletion therapy ameliorates autoimmune disease through ablation of IL-6-producing B cells. The Journal of experimental medicine, 209 (5), 1001-10 PMID: 22547654

The effect of MyD88 deletion on autoimmunity driven by Foxp3 inactivation

Over last 10 years few subjects in immunology have received more attention than regulatory CD4 T cells called in abbreviation Tregs. Tregs are considered to have the potent suppression activity over adaptive immune responses and their lack may result in the autoimmunity development. The central trait pertinent to Tregs is the expression of the transcription factor Foxp3.  The essential role of Foxp3 in Tregs’ life is underscored by the fact that Foxp3-deficient mice or human patients with mutations in the respective gene acquire massive systemic autoimmunity due to the absence of Tregs and generally do not fare well. The publication I have found adds yet another twist to Tregs and Foxp3 story. It turns out that the concurrent to Foxp3 deletion of MyD88 (the crucial adaptor protein linking the innate recognition of microbial signature patterns to the expression of genes involved in defense mechanisms) imparts effects that are not identical between major environmental surfaces (skin, gastrointestinal tract or lungs) and the systemic compartments.

The link:

Foxp3-deficient mice suffer from the advanced inflammatory skin condition and as a result have grossly increased skin pathology indicators like dryness, loss of hair and bleeding. Apart from that their ears and tails are seriously necrotized. However, animals deleted for both Foxp3 and MyD88 show many substantial improvements. Authors demonstrate that the removal of MyD88 from Foxp3-deficient background diminishes immune infiltration to the epidermis and locally deactivates molecular pathways involved in the amplification of inflammatory signals and cellular trafficking (NF-κB translocation to the nucleus, the expression of ICAM-1 on keratinocytes). Additionally, the skin level of numerous cytokines is reduced in doubly deficient animals compared to mice with the single Foxp3 deletion.

What is remarkable, MyD88 deletion on Foxp3-deficient background has also significant systemic effects because such mice grow to much bigger size than visibly runted Foxp3 single mutants. Therefore investigators analyze the extend of immune infiltration in multiple organs of double Foxp3/MyD88 mutants and find out that they have decreased inflammation scores and the expression of pro-inflammatory cytokines not only in the skin but in the small intestine and lungs as well. However, the alleviating consequences of MyD88 removal are restricted to environmental surfaces as the symptoms characteristic for Foxp3 deletion continue unabated in the liver and the pancreas of Foxp3/MyD88-deficient animals (and are even enhanced in their salivary glands). Moreover, the detailed examination of spleen and lymph nodes (authors indicate that they focus on skin draining lymph nodes and mesenteric lymph nodes) shows that cellular counts, proliferation indicators and the expression of various cytokines are elevated in Foxp3/MyD88-deficient mice compared to animals with Foxp3 deletion.

Such difference between the systemic compartments and environmental surfaces could be explained by several factors. Authors show that introducing MyD88 deletion to Foxp3-deficient background disrupts the chemokine gradient between lymph nodes and effector tissues. They also demonstrate that homing ability of CD4 T cells to lungs is incapacitated and as a consequence lymphocytes may accumulate in draining lymph nodes. Finally, in a series of adoptive transfer experiments it is established that the protective effect of MyD88 deletion acts at the level of target tissue and is independent on whether CD4 T cells express MyD88.

In an interesting, although not entirely conclusive part of the paper authors follow the hypothesis that the protective influence of MyD88 deletion in this model may be due to the removal of capability to process activation signals derived from microbiota. To prove such concept they attempt to mimic the effect of MyD88 ablation by purging commensal bacteria from gastrointestinal tracts of Foxp3-deficient animals with two different antibiotic treatments. The first such treatment includes two antibiotics (doxycycline and cotrimoxazole) and indeed relieves some symptoms of Foxp3 inactivation in the skin and lungs. However, the second regimen comprising four antibiotics (kanamycin, vancomycin, metronidizol, and amphotercin-B) actually worsens the state of Foxp3-deficient animals and accelerates their death. I would be interesting to know what part of commensal microflora can be hold responsible for either such protective or detrimental effects in the context of ongoing autoimmunity.

Rivas MN, Koh YT, Chen A, Nguyen A, Lee YH, Lawson G, & Chatila TA (2012). MyD88 is critically involved in immune tolerance breakdown at environmental interfaces of Foxp3-deficient mice. The Journal of clinical investigation, 122 (5), 1933-47 PMID: 22466646

The early microbiota species composition may influence the B cell development in children

We never walk alone; each of us is colonized by scores of bacterial species that live in our gastrointestinal tract. The colonization event occurs right after birth and has profound effects on our immune system – this sentence appears in every modern immunology textbook. There is, however, no such thing like one human microbiome as the species composition in our gut depends on many aspects. Diet, geographical region we live in or civilization level of society we belong to may influence to what species we are exposed and most probably dictate also many qualitative features of human immune responses. But we do know yet many technical details of the interactions between the human microbiome species arrangement and our immune system characteristics. I have found the publication that studies what impact on the development of B cell populations in the childhood may have the early microbiota species composition in the gut.

The link:

The work is performed on Swedish children and it contains two parts. In the first authors follow the development of B cells in the analyzed cohort of individuals along the period that ranges from right after birth to three years of age. B cells from peripheral blood (defined as CD20+ cells from the lymphocyte gate) are subdivided into CD5+ population (secreting mostly polyreactive IgM antibodies) and CD27+ subset (whose majority are class-switched memory cells). The first population forms the bulk of B cells in children, it expand soon after birth and gradually contracts to reach low level in adults. The second subset grows much slower; it attains its plateau at 18 months of life and continues at similar numbers at 36 months (although there is still much more CD5+ B cells than CD27+ B cells at that time). In the adult blood CD27+ population prevails over CD5+ B cells.

For the second part authors collect fecal samples in the same cohort of children during the early weeks of their life (1, 2, 4 and 8) and study the microbiota species composition in obtained samples. To detect bacterial groups present in the analyzed material they use combined microbiological (growth conditions), biochemical and molecular approaches that identify Escherichia coliEnterobacteria that are not E.coliStaphylococcus aureusBacterioidesEnterococci, BifidobacteriaLactobacilli and Clostridia. Next investigators attempt to correlate quantitative features of B cell development (counts of CD20+, CD5+ and CD27+ cells) along the infancy period with qualitative aspects of microbiota composition (presence or absence of given bacterial groups) during the early weeks of life. The major conclusion of this paper is that the early presence of species or bacterial groups like E.coli and Bifidobacteria in children guts associates with the increased numbers of CD27+ B cells at 4 and 18 months of their life. However, the colonization with S.aureus correlates inversely with CD27+ B cell counts at 4 month.

Obviously correlations do not necessarily indicate for the casual bond but I think the data presented in this report are interesting especially when you combine them with the information that authors provide in the discussion part. According to that E.coli and Bifidobacteria are thought to be classical infantile bacterial species but S.aureus is not. This organism is considered to be a skin bacterium; however, it colonizes the gastrointestinal tract of Swedish children with high penetration (~70%) and may slow down the development of effective immune responses. Such publications are important because the incremental appreciation of how microbiota species composition regulates human immunity may let us use this knowledge for practical purposes in the future.

Lundell AC, Björnsson V, Ljung A, Ceder M, Johansen S, Lindhagen G, Törnhage CJ, Adlerberth I, Wold AE, & Rudin A (2012). Infant B cell memory differentiation and early gut bacterial colonization. Journal of immunology (Baltimore, Md. : 1950), 188 (9), 4315-22 PMID: 22490441

The intestinal role of NLRC4

The May issue of Nature Immunology contained an article that has described an intriguing mechanism of tolerance to microbiota without losing the ability to detect invading intestinal pathogens and switching on protection mechanisms when it comes to the defending. According to that report macrophages residing in the colon (but not from the bone marrow) remain hyporesponsive to TLR stimulation provided by commensal microbiota. However, the infection with Salmonella can provoke the same macrophages to process and secrete IL-1β cytokine in the manner that is dependent on NLRC4 inflammasome and caspase -1 activation (I have discussed this publication in one of my previous posts: The report I am discussing today is related to the above paper because it systematically analyzes the intestinal role of NLRC4. It turns out that there are substantial differences between consequences of deleting NLRC4 compared to effects of TLR5 deletion in conditions when both strains are presumed to harbor the same microbiota (both NLRC4 and TLR5 recognize flagellin – a dominant immune activator in the gut). These differences can be visible either in the healthy colon or in colitis development but not during Salmonella infection.

The link:

Authors have shown earlier that the deletion of TLR5 caused severe changes in the interactions between host and intestinal commensals even in the absence of any challenge. Such perturbations can lead to the greater bacterial burden in the gut followed by the increase in pro-inflammatory indicators, colitis and eventually the tendency to develop metabolic syndrome diseases (Metabolic syndrome and altered gut microbiota in mice lacking Toll-like receptor 5: Science. 2010; 328:228-31). Deletion of NLRC4, however, does not procure any spontaneously occurring physiological changes.

In addition, NLRC4 knockout mice do not display any major effect when they are given injections with anti-IL-10R antibody (the cytokine IL-10 is very important player in maintaining the intestinal tolerance and ablating IL-10 signaling often results in colitis). In contrast, animals with TLR5 or double TLR5/NLRC4 deletions develop colitis upon such treatment. Another way to inflict colitis is to expose the intestinal epithelium to a chemical called DSS which introduces damages to the gut barrier. Investigators show that NLRC4 knockouts similarly to mice with deleted TLR5 are more sensitive to DSS-driven colitis that the wild type strain.

The paper also examines the response against Salmonella infection in mice with NLRC4, TLR5 and double TLR5/NLRC4 deletions. Authors use two strategies – the first applies the procedure resulting in disease that is similar to infection in humans (by pretreating mice with streptomycin prior to bacterial exposure) and serves to examine early inflammatory events in colon and cecum. The second approach includes low-dose oral infection with several Salmonella strains to assess the effect of analyzed deletions on mice survival.

NLRC4 is involved the early detection of Salmonella, although it seems to work together with TLR5. Data show that the single removal of either NLRC4 or TLR5 does not render mice incapable to mount the inflammatory response to this pathogen. Only the double NLRC4/TLR5 deletion strain does not react to Salmonella. The secretion of IL-1β cytokine is also abolished only in the double deletion animals. Interestingly, all double deletion effects (no inflammation and no IL-1β secretion) are mirrored by MyD88 deletion which ablates the signaling to the majority of TLR receptors and the signaling to cytokines processed by NLRC4 inflammasome – IL-1β and IL-18. Authors use this fact to underscore the notion that flagellin (ligand for TLR5) is indeed the dominant immune activator in the gut. Finally, the protective role of NLRC4 is demonstrated by showing that NLRC4 deletion impairs survival to a flagellate strain of Salmonella. In contrast, no any difference between mice with NLRC4 knockout and wild type strain is seen when animals are infected with aflagellate bacteria.

It is hard to directly compare both papers because they have quite different scopes – the first studies only one colonic cell population whereas the other analyzes the overall immune response in the colon. But I think that they corroborate each other versions because unlike TLR5 NLRC4 looks indeed like the immune receptor which is not involved in dealing with microbiota under normal circumstances. However, it is the TLR5/NLRC4 axis that seems be critical in mounting the efficient defense against intestinal pathogens (because only the double TLR5/NLRC4 makes no response to Salmonella). The first paper shows that intestinal macrophages are hyporesponsive to TLR stimulation but they maintain the constitutive expression of pro-IL-1β (which is somehow diminished in germ-free mice). TLR5 is expressed on the surface of antigen presenting cells but also by intestinal epithelial cells whereas NLRC4 is present exclusively inside intestinal macrophages. Could colonic epithelial cells prime these macrophages to express pro-IL-1β when receiving signals through TLR5 from microbiota?

Carvalho FA, Nalbantoglu I, Aitken JD, Uchiyama R, Su Y, Doho GH, Vijay-Kumar M, & Gewirtz AT (2012). Cytosolic flagellin receptor NLRC4 protects mice against mucosal and systemic challenges. Mucosal immunology, 5 (3), 288-98 PMID: 22318495