Tracking deletion of autoreactive clones versus Treg generation for thymically expressed epitopes

So far 3 different outcomes have been identified for developing T cells in the thymus: to develop into naive T cells, get deleted or become Foxp3+ Treg. Both deletion and Treg path require the presence of specific epitopes. However, how a given T cell decides between these pathways is not well understood. 

Here is a new paper in PNAS that tries to tackle this question using the tetramer tracking approach. The authors are using PLP (brain-specific protein) as an endogenous antigen expressed in the thymus. Surprisingly both PLP^WT and PLP^KO mice showed near similar numbers of tetramer-positive T cells in peripheral tissue. However, as expected, only PLP^WT mice that express PLP epitopes in the thymus but not PLP^KO mice that do not express the same epitopes showed Treg development.

   


 Similar results were obtained when thymus tissue was analyzed.

  

To make tetramer tracking for reliable the authors used transgenic mice expressing a fixed TCR beta chain. These mice also showed a similar phenotype.  

As in PLP^WT and PLP^KO mice, fixed:TCR beta mice on PLP^WT but not on PLP^KO background harbored Tregs in the periphery. Notable, the rest of the tetramer-positive Foxp3-negative T cells displayed an anergic phenotype (CD73^HiFR4^Hi).

A similar phenotype was found in the thymus. Note, there was an unexpected and significant reduction of tetramer-positive T cells from the thymus to the periphery in fixed:TCR beta mice on PLP^KO background compared to fixed:TCR beta mice on PLP^WT background. 

So far these data indicated that there is almost no deletion of PLP specific T cells in the thymus on WT mice [compaed KO] but ~2-fold reduction in fixed:TCR beta mice on PLP^WT compared to KO. Almost half of the tetramer-positive T cells ended up in the Treg pool on the WT background. The remaining T cells showed an anergic phenotype. However the dramatic reduction of tetramer-positive T cells from the thymus to the periphery in KO mice raises some serious unanswered questions.

Finally, to find some correlation between TCR specificity and Treg/anergy/deletion phenotype, the authors selected 4 PLP-specific TCRs (denoted here as A, B, C, D). Their analysis showed that some (clone "A") but not other PLP-specific TCRs (clone "C") were able to generate Tregs in the thymus. Notable, TCR "C" displayed the highest affinity to PLP epitope. Also, there is a substantial reduction of clone "C" from the thymus to the periphery in the Foxp3-negative compartment. This possibly reflects the fact that most clones in "C" are anergic and slowly disappear from the periphery.  

In summary, this study re-confirms that tolerance to self-antigens is mostly controlled via Treg generation and that not all antigens/epitopes and their corresponding TCRs are able to participate in this process. There are few unexplained observations in this paper though as discussed above. 

posted by David Usharauli

An antigen from gut commensal Bacteroides thetaiotaomicron (B. theta) is recognized by Foxp3+ Tregs

1. B.theta-specific T cells when transferred into #antibiotic-treated and B.theta #colonized T cell-deficient mice (#RAG1-KO) develop into #Foxp3+ #Tregs as well as effector T cells. It is not clear if donor tg T cells were on RAG1-KO background as well. Host mice stay healthy. pic.twitter.com/TD0pnnxgAR

— David Usharauli (@3DiMMUNE) February 9, 2019

3. The authors were able to identify a #specific #epitope in B.theta recognized by B.theta-specific T cells. The #sequence, EEFNLPTTNGGHAT, were mapped in BT4295 predicted to be a SusE/SusF #lipoprotein in outer #membrane (OM). pic.twitter.com/rH5cQHQcVN

— David Usharauli (@3DiMMUNE) February 9, 2019

In #summary, this study showed that certain gut #commensal #microbiota species, in this case, #Bacteroides thetaiotaomicron (B. theta) are recognized by #Tregs and even induce #effector T cell differentiation though the host stays healthy because Tregs regulate effector T cells.

— David Usharauli (@3DiMMUNE) February 9, 2019

Another point is whether #microbiota-specific Tregs are #denovo induced or #amplified from existing ones. I personally think it is a latter. Most #Tregs develop from #CD25-negative #precursors poised for Foxp3 expression & such precursors are all #thymus derived, in my opinion.

— David Usharauli (@3DiMMUNE) February 9, 2019

A minor population of tumor infiltrating CD8+ T cells actually express tumor antigen specific TCR

Tumor immunotherapy with checkpoint inhibitors (anti-CTLA-4, anti-PD1/PDL1 antibodies) showed remarkable therapeutic effect in narrow slice of cancer patients (~20% - 25% of cases). It relies on reactivation of tumor infiltrating T cells called TILs (tumor-infiltrating lymphocytes). Obviously, the more we know about nature of these TILs the better medical approaches could be implemented.

A new study in Nature Medicine by Ton Schumacher's group in Netherlands analyzed single cell sorted TIL TCR specificity derived from 4 treatment-naive patients. Their limited and [technically inadequate, in  my view] analysis revealed that tumor specific T cells represent a minor population among TILs from such cancers as ovarian cancer (OVC) and microsatellite stable colorectal cancer (CRC). 

First, the authors validated their approach with T cells derived from melanoma samples. 88 CD8+ TILs were single cell sorted, TCR sequenced and TCR α/β chain pairs established. 15 such TCR pairs were trasduced into primary T cells and exposed to melanoma samples. 9 different TCR transduced T cells showed IFN-γ expression (60%). It is not clear what tissue they used as non-melanoma negative samples. They mentioned B cells in methods section though it is not obvious why would B cells act as a good negative samples for skin tissue. Also, I find it highly inadequate to limit tumor antigen-specificity detection to intracellular IFN-γ expression when using primary T cells as a TCR transduction carrier. Based on TCR affinity it might engage other types of response.

With these caveats in mind lets examine their results. When they tested single cell sorted TIL TCRs derived from ovarian cancer sample (from a single patient, OVC21), out of 20 TCR transduced T cells only 1 TCR showed IFN-γ response to tumor tissue. Another patient's TCR assay showed zero reactivity. Again, it is not clear whether low reactivity is a biological effect or simply technical deficiency as discussed earlier. We cannot even be sure whether that 1 TCR is actually tumor specific either. It also appears that pairing of TCR α/β chains is not a straightforward either because out of 95 sorted cells only 37 (39%) TCRα/β pairs could be identified.

A slightly more encouraging results were obtained with TCRs from one CRC sample (patient CRC11). Here 5 out 16 TCR tested showed IFN-γ response to cancer organoid tissue. however, similar test on another patient's TCRs showed zero response.

In summary, this study tried to show [but utterly failed in my view] that for some tumors tumor-reactive TCRs among TILs are quite rare. What are then such TILs' TCR specificity are unknown. Why are they recruited in tumor tissue is not known either. The authors have not even formally tested what neoantigens, if any, these tumors from those 4 patients actually expressed. Very unsatisfactory study.

posted by David Usharauli

Cross-reactivity between microbial-derived antigens and tumor neoantigens correlates with long-term survival

This is a very interesting paper published in Nature few months back. In this study the authors wanted to uncover immune correlates of long-term (>10 yrs) survival from pancreatic ductal adenocarcinoma that normally account for less than 2% of all patients. 

First they found that "patients with both the highest predicted neoantigen number and either the greatest CD3+CD8+, or polyclonal T-cell repertoire, but neither alone, exhibited the longest survival". It suggested that tumor neoantigen-derived epitope quality rather than simply quantity correlated with survival. 

More importantly, bioinformatics analysis of similarities (mimicry or cross-reactivity or poly-reactivity) and neoantigen fitness modelling between tumor neoatigens and microbial-derived antigens significantly stratified short- and long-term survivors independent of confounding factors and adjuvant chemotherapy. 

If these data will  hold true in other cancer settings, it will append how immunotherapy is applied to treat cancer patients. While this finding is potentially hugely important for immunotherapy, the authors tried not to make too big a statement about it (I would imagine it is a result of a typical reviewers conservatism), writing that "This hypothesis does not assume any associations between pre-existing antimicrobial immunity and survival, but rather aims to develop a strategy to identify candidate neoantigens based on defined immunogenic pathogen-derived epitopes" and in other place "Our results do not indicate causal associations of pre-existing microbial and anti-tumour immunity in LTSs [long-term survivors]. Instead, our data suggest that embedding microbial homology in the context of our neoantigen quality model can help to create an effective surrogate for immunogenic neoantigens." 

posted by David Usharauli

Outcome of TCR/pMHC interactions is presently unpredictable

This week journal Cell published a very important paper from KC Garcia's lab at Stanford. The focus of the study was to understand and determine, if possible, all the parameters that makes a given peptide/MHC and its similar looking variants stimulatory versus non-stimulatory to a given TCR molecule.

First, they compared two TCR molecules specific for class II HLA-DR4-HIV_{(Gag164–183)}. While both TCR expressing clones, TCR6 and TCR11, could bind pHLA tetramer, only TCR6 "proliferated, secreted cytokines, and downregulated TCR in response to stimulation with" HLA-DR4-HIV_{(Gag164–183)}. Both TCR clones could respond directly or after transducing them into TCR αβ-deficient cell line to non-specific stimulation such as anti-CD3 (OKT3).

Similar observation were made using TCRs specific for class I HLA-B35-HIV_{(Pol448–456)}. Here too, both TCR clones designated as TCR589 and TCR55 could bind specific pMHC tetramer but only TCR589 showed dose-dependent secretion of IL-2 in response to peptide-pulsed APCs.

Next, the authors determined TCR/pMHC affinity by surface plasmon resonance. For TCR55 (non-stimulatory) and TCR589 (stimulatory) clones it was 17_μM and 4_μM K_D, respectively. Surprisingly, for TCR6 (stimulatory) and TCR11 (non-stimulatory) clones affinity was 10_μM and 1_μM K_D, respectively.

Analysis of TCR55 (non-stimulatory) and TCR589 (stimulatory) TCR/pMHC interaction on lipid bilayers by total internal reflection fluorescence (TIRF) microscopy showed similar mean dwell times of 10.1 ± 2.0 s and 8.7 ± 2.1 s, respectively. Structural analysis of the crystal structures of TCR589 and TCR55 in complex with HLA-B35-HIV_{(Pol448–456)} also showed "no substantial differences in the conformations of the constant regions that would indicate differential allostery correlating with signaling."

Functional analysis of TCR proximal signaling molecules showed that stimulatory TCR589 but not TCR55 could up-regulate CD69, induce calcium release and recruit ZAP70 in response to HLA-B35-HIV_{(Pol448–456)}.

Next, they generated peptide library similar to HIV_{(Pol448–456)} with potencies "ranging from full agonism to partial agonism to no detectable signaling". Importantly, "the peptides that elicited these disparate signaling outcomes were remarkably similar in sequence."

One common feature of agonist peptides such as pep20 and SQL that could signal TCR55 was that such interaction excluded CD45 with varying degrees of efficiency compared to non-stimulatory peptide ligands in "a cell-free system using giant unilamellar vesicles (GUVs) to eliminate effects of signaling in live cells".

In addition, another common feature of stimulatory peptide ligands were so called catch bond formation between TCR and stimulatory pMHC "under conditions of force using a Biomembrane Force Probe (BFP) bond lifetime assay". In contrast, non-stimulatory ligands formed slip bond formations.

In summary, this study indicates that "catch bond formation and CD45 exclusion as principal emergent properties differentiating stimulatory from non-stimulatory TCR-pMHC interactions" but "the same pMHC can be an agonist or non-agonist for different TCRs, although it is difficult to assign causality to either the TCR or the peptide-MHC. The formation of a catch or slip bond is a collective property of the entire TCR-pMHC interface."

My opinion: one drawback of this study is that not many labs can verify its conclusions. I have not even heard of some of the techniques they used here. I could only trust that the authors tried to interpret their data as unbiased as it is rationally possible. Another weakness of the study is a failure to identify parameters that control CD45 exclusion or catch bond formation when TCR interacts with pMHC. Basically, with the exception of few very high affinity ligands (< 2_μM), we cannot predict in silico whether a given peptide will be stimulatory or non-stimulatory without actually doing an experiment on it.   

posted by David Usharauli

Effort to identify tumor-specific antigens: The University Industrial Complex study results

New study in journal Cell is prime example why utilization of sophisticated high-throughput methods and computer technologies does not guarantee generation of clinically useful results. I imagine the only reason this study was even accepted in Cell was the fact that list of authors included many well-known scientists with links to both academia and silicon valley (Stanford University School of Medicine, Chan Zuckerberg Biohub, Parker Institute for Cancer Immunotherapy).   

Idea of this study was to develop techniques to quickly identify tumor-specific antigens (most likely mutated antigens) that could be used in immunotherapy (though there is no evidence that any cancer vaccines based on mutant protein sequences actually work in humans using available practices). 

For this task, the authors took advantage of yeast-display library expressing random peptide covalently linked to the HLA-A*02:01 molecule, an allele which is present in up to 50% of a number of populations. The authors estimated that "approximately 400 million unique peptides ranging from 8 to 11 amino acids are represented in the combined [yeast-display] libraries."

To validate this approach, they used three recombinant 'blinded' positive control TCRs derived from a melanoma patient (their antigen specificity had been identified independently by exome sequencing, tetramer staining and binding prediction algorithms). However, antigen-specificity of only 1 TCR (NKI 2) could be validated using their yeast-display library. As the authors said "targets of NKI 1 and NKI 3 could not be unambiguously identified through this blinded validation."

Of note, in these validation experiments with NKI 2 (specific for ALDPHSGHFV, a peptide neoantigen derived from CDK4 and other DMF5 TCR specific for EAAGIGILTV derived from the MART-1 melanoma antigen, successful validation [specific enrichment + TCR staining] occurred when HA tagged 10-mer epitope library were used. 

The authors anyway went ahead with this "less than perfect" approach to try to identify tumor antigen specificity of T cells derived from 2 patients with colorectal adenocarcinoma and homozygous for the HLA-A*02 allele. The authors focused on 20 TCR most enriched in tumor tissues (based on frequency of occurrence of the same TCR genes). 

Out of these 20, only 4 TCRs could enrich peptide from the library (only with c-Myc tagged 9-mer epitope library) and only 3 TCR could stain yeast samples.  

Next, the authors try to identify epitopes from potential landscape of sequences for each TCR. Several algorithms were deployed (at least 3 or more such as a modified variant of the previous statistical method using a position weight matrix and a method utilizing a two-layer convolutional neural network). They found 1 peptide sequence EYGVSYEW, which closely matches the peptide motif for TCR 1A, however, neither this exome peptide or the anchor-modified exome peptide (EMGVSYEM), nor the human peptide predictions stimulated the cell line modified to express the TCR 1A. TCR 4B was stimulated with several peptides and as the authors write "true in vivo specificity cannot be unambiguously identified without additional tumor information". Regarding TCR 2A and 3B, only 1 peptide stimulated cell line expressing these TCRs. This peptide was MMDFFNAQM, which is derived from U2AF2, a protein involved in an RNA splicing complex. However, in both patients, no mutations were found in U2AF2.

In summary, the authors wrote "although we cannot definitively determine an immune response targeting the peptide derived from U2AF2, the evidence from the yeast-display screen, prediction algorithm, and in vitro stimulation identify this peptide as the likely target". However, when reading this study it is clear that none of the components worked: yeast-display screen performed suboptimally, prediction algorithms provide little clue and in vitro stimulation made it even more confusing. So, what have we learned from all of these? I would say maybe don't do what they did.

posted by David Usharauli    

Tolerance to insulin is maintained by Foxp3+ Tregs

A new study in Journal of Immunology suggests that tolerance to insulin is maintained by Foxp3+ Tregs rather than by deletion of insulin-reactive T cell clones. 

 

Here, the authors reconstituted mice with T cells on scid background transduced either with high (4-8) or low (12-4.1) affinity TCR specific for native insulin peptide (insulin epitope B:9–23). In addition, each of TCR construct were fused with either native insulin (INS) or modified insulin carrying super-affinity peptide (R22E). All mice expressing either INS or R22E but not irrelevant HEL were protected from developing diabetes.

The authors showed that while R22E deleted developing insulin-specific T cell clones in the thymus, native INS did not.

In fact, the authors showed that if the T cells also lacked Foxp3 molecule (scid-scurfy), then protection against diabetes was lost in mice exposed to native INS.

This study could be interpreted to show that with the exception of  epitopes which are able to delete (purge) cognate T cell clones in the thymus, tolerance to self in the periphery is maintained by thymic-derived Foxp3+ Tregs.

posted by David Usharauli 





    

TCR derived from gut Foxp3+ Tregs can generate two distinct T cell phenotypes

This is my first review of a paper published in new journal from Science family called Science Immunology.

This paper deals with the relationship between TCR and origin of peripheral [gut-residing] Foxp3+ regulatory T cells. Specifically, the authors have used method called somatic cell nuclear transfer (SCNT) that allowed generation a transnuclear (TN) mouse that carries a TCR cloned from gut tissue Foxp3+ Tregs. 

First surprising observation was that these Foxp3+ TCR transnuclear but otherwise unmanipulated mice (called pTreg TN/RKO mice, on RAG KO base) had no Foxp3+ Tregs neither in the thymus or the peripheral lymph nodes (mLNs).

Second, the authors found that transnuclear mice instead harbored specialized IFN-γ⁺CD8αα⁺ intraepithelial lymphocyte (IEL) subset.

Indeed, adoptive transfer Foxp3+ TCR transnuclear T cells into WT host revealed that donor T cells could differentiate either into CD8αα⁺ intraepithelial lymphocyte (IEL) subset or acquire Foxp3 marker as representative of peripheral Treg subset. Of note, development of both of these subsets dependent of gut microbiota.

Finally, the authors showed that donor Foxp3 transnuclear T cells transferred into T cell-deficient hosts did not induce gut inflammation even if they were derived from scurfy background (lacking ability to express functional Foxp3 protein), implying that functionality of CD8αα⁺ intraepithelial lymphocyte (IEL) derived from Foxp3+ TCR T cells did not cause any tissue pathology.

In summary, this study showed that depending on circumstances TCR derived from intestine Foxp3+ Tregs can drive development of two distinct subsets with immunoregulatory role: CD8αα⁺ intraepithelial lymphocyte (IEL) subset or peripheral Foxp3+ Treg subset.  

David Usharauli