Saturday, December 9, 2017

From microbiome-association studies to causal microbe identification

As new Nature study says "the general inability to move beyond correlations and address causation has been the Achilles heel of microbiome research." Well said. Indeed, notwithstanding of so many microbiome studies, the number of identified microbiota species specifically and reproducibly linked to a particular [medical] condition is still zero.  

Is there a way to overcome this challenge? Maybe one could compare microbiome content of the hosts displaying variable clinical phenotype [DSS colitis, in this case] and find one microbial species, if you are lucky, that control that phenotype in every host? But isn't this approach exactly what all other studies have done all along? 

And actually, how strong is the data in support of this approach? The authors showed that germ-free mice and germ-free mice colonized with mouse microbiota derived from SPF mice (MMb mice) were highly sensitive to DSS colitis induction compared to SPF mice or germ-free mice colonized with human microbiota (HMb mice). Of note, observed difference in survival between MMb and SPF is strange since both should have SPF microbiota.

When the authors compared several pairs of mice strains housed separately or co-housed, they found that sensitivity to DSS colitis segregated with presence of Lachnospiraceae species (Clostridium immunis).       

Indeed, colonization of germ-free mice with human microbiota enriched with Clostridium immunis could improve mice survival in DSS colitis model.

However, again, survival data between different mouse strains did not correlate with level of Lachnospiraceae species, questioning simplicity of one-on-one relationship between Clostridium immunis and colitis score (HMb and SPF have similar survival curve but vastly differ in Lachnospiraceae content).

In summary, without more data and confirmation by other labs I will view these data as very preliminary and unverified with lots of caveats. Surely, not a Nature material in my opinion.

posted by David Usharauli

Saturday, December 2, 2017

Dual TCR expressing T cells could drive autoimmunity

An article in Cell Host and Microbe (CHM) caught my attention. In proposed that dual TCR expressing T cells are responsible for autoimmune phenotype in K/BxN mice

K/BxN mice develop spontaneous arthritis thought to driven by Vβ6+ KRN T cells recognizing glucose-6-phosphate isomerase (GPI), the self-Ag presented by MHC class II Ag7 molecules. Ordinarily, it is thought that pathogen cross-reactive to self antigen, in this case GPI, could initiate autoimmune disease (theory of molecular mimicry).

However, recently the role of endogenous microbiota in driving autoimmune arthritis received new attention. Here, the authors showed segmented filamentous bacteria (SFB) was required to initiate autoimmune arthritis in K/BxN.

However, since SFB-derived peptide recognized by T cells required Vβ14+ TCR it was unclear how SFB was mediating autoimmunity against GPI recognized by completely different TCR made of Vβ6+ chain. Further analysis showed that some T cells in K/BxN express dual TCRs expressing both Vβ6+ and Vβ14+ chains. Indeed, sorted T cells expressing dual TCR, but not Vβ6+ chain alone, recognized SFB-derived peptide (A6).

In vivo, adoptive transfer of monoclonal KRN T cell population on RAG KO background that prevents expression of other Vβ or Vα chains (only expressing Vβ6+ KRN T cells), could not mediate arthritis in T cell-deficient host (harbor normal B cells also required for arthritis development). Of note, other paper in 1999 however found no difference in arthritis development between WT and RAG KO KRN T cells.

In summary, the authors think that first dual TCR T cells get activated by SFB-derived epitope via Vβ14+ TCR, indirectly prime Vβ6+ TCRs, on the same T cells, that then actually mediates autoimmunity against self antigen GPI. However, it is not clear from this study whether endogenous Vα chains could contribute to cross-reactivity between SFB and GPI when recombined with Vβ6+ or Vβ14+ chains in WT KRN T cells.

posted by David Usharauli


Sunday, November 26, 2017

Allergen-sensitized mothers transfer protection against allergy to offspring through milk

In this study the authors showed that mouse pups born to mothers sensitized to allergen were significantly protected from developing allergic response to the same antigen.    

Protection in offspring was associated with the generation of antigen-specific Foxp3+ Tregs as observed in proliferation suppression assay or following short-term Treg depletion by DT (however, the authors did not analyze antigen-specificity of Tregs by tetramer staining). 

Further experiments showed that mother's milk contained allergen-specific antibodies and immune complexes (IC) and breastfeeding by allergen-sensitized mother (irrespective of birth mother status) was sufficient to transfer allergen protection to offspring.

In summary, this study suggests that breastfeeding by allergen-sensitized mothers can benefit offspring by preventing development of allergic response to the same allergen. However, it is not clear how exactly the authors see this mechanism working in humans. In mice, mothers were intentionally sensitized with allergen using epicutaneous (skin) application that supposed to mimic how humans with skin barrier dysfunction get sensitized to allergens. But the authors have not tested if milk from atopic human mothers can have the same effect on their offspring. For some reason the authors tested milk from nonatopic human mothers and showed that it 'worked' when fed to mice but did not provide any explanation why healthy, nonatopic human mother milk should contain any "protection" against allergen when mothers themselves are not sensitized as experiments in mice showed they must be for a milk derived immune complexes to work. So lots of unknowns and contradictions.

posted by David Usharauli