Both non-self and self-specific CD8 T cell clones are found in equal frequency in human PBMCs

Clonal selection model of immune system ontogeny is the most widely accepted concept of immunology. Simply put, it postulates that T cell clones (B cells too) specific for self-antigens (epitopes) are purged from adaptive receptor repertoire during in utero development.

At the time of its formulation, there was no direct proof for this model. It was mainly supported by observations derived from organ transplantation studies. Later, research on TCR transgenic mice confirmed that self-specific T cells were indeed purged from T cell compartment. 

However, more recent studies questioned the validity of results obtain from transgenic mice, mainly due to non-physiological timing of TCR expression in transgenic T cells in the thymus. 

In this regard, this new, short study in journal Immunity is of interest. Using sensitive, tetramer-specific T cell enrichment method, the authors showed that human PBMC surprisingly contains equal numbers of both non-self and self-specific CD8 T cell clones.  

The authors showed that similar to the frequency of non-specific CD8 T cell clones, healthy human PBMCs contain self-specific CD8 T cell clones with frequency range between 1:10⁴ to 1:10⁶ (see below too).

Next, the authors showed that CD8 T cells derived from both female and male donors contained functionally active male antigen, HY-specific CD8 T cells (though, here CD8 T cells were expanded using non-physiological stimuli, such as anti-CD3/anti-CD28 antibodies or PHA).

However, as expected, further experiments with the pool of enriched self-specific CD8 T cell clones revealed that these self-specific CD8 T cells were anergic when stimulated with specific peptide and anti-CD28 antibody (resembling more physiological stimulation, hopefully).

In summary, these results suggest the following: immune system contains lots of self-specific CD8 T cells. These cells are anergic, i.e. non-responders in an ordinary sense. However, they are capable of becoming functionally active when stimulated by non-conventional ways, like PHA or therapeutic antibodies targeting its signaling molecules (CD28, for example) or during the process of autoimmunity?

Why immune system harbors so many self-specific T cells? It was really surprising that on average, PBMCs from male donors still had 1/3 of number of HY-specific CD8 T cells found in PBMCs from female donors. Even though almost every cell in male donors can express HY antigen, the authors still found so many high-affinity HY-specific CD8 T cells in PBMCs from male donors. Staining for mouse HY-specific CD8 T cells showed the same results.

Another interesting observation was the finding that PBMCs from healthy donors contained very high frequency of pre-proinsulin-specific CD8 T cells (almost 1:10⁴) that were significantly increased in donors with T1D. Is it possible that selection of high frequency of pre-proinsulin-specific CD8 T cells has to due with its cross-reactivity for certain evolutionary relevant pathogen-specific epitope?

David Usharauli

         

Leukemia - when RAG genes go Rogue

Receptor diversity of the adaptive immune system (in both B and T cells) is driven by RAG genes. Additionally, B cells also express AID that is responsible for Ig somatic hyper-mutation (SHM) and class switch recombination (CSR). Basically, RAG and AID genes represent naturally occurring gene editing system in mammals.

However, evolution always come at a cost. New paper in Nature Immunology suggests that improperly expressed RAG and AID genes may contribute to the evolution to childhood leukemia.     

By looking at AID expression pattern during human or mouse bone marrow B cell lymphopoiesis, the authors observed that IL-7 signaling played a major role in preventing AID expression prior to small pre-BII cell stage.

In vitro experiments on pre-BII cells from AID-GFP mouse confirmed that withdrawal of IL-7, in combination with LPS (a surrogate for inflammation), induced co-expression of both AID and surface Ig  κ Light-chain (a surrogate marker for RAG gene activity).

Further experiments showed that transduction of mouse pre-BII cells with WT version of pre-leukemic genetic lesion ETV6-RUNX1 induced RAG gene expression.

Finally, adoptive transfer of ETV6-RUNX1 transduced WT pre-BII cells that underwent 5 cycle of in vitro IL-7 withdrawal in combination with LPS, produced leukemia in the NOD.SCID hosts (which lacks its own adaptive immune system). However, ETV6-RUNX1 transduced pre-BII cells from AID-KO or RAG-KO mouse failed to produce leukemia.

In summary, these results indicate that pre-BII cells are vulnerable for improper co-expression of AID and RAG genes due to combination of ETV6-RUNX1 genetic lesion and inflammatory milieu (that may diminish IL-7 signaling). Such condition may arise when the humans are exposed to childhood infections later in their life. It is believed that timely vaccination could reduce incidence of such leukemia in children.     

David Usharauli

Tumors repel immune system but attract lytic viruses

Clinically relevant tumors have two characteristics: uncontrolled proliferation (expansion) and immune evasion.

Surprisingly, several recent papers indicated that tumor transformations are associated with inhibition of cell-autonomous anti-viral/anti-modified RNA/anti-modified DNA recognition pathways such as RIG-I or STING.

New research article in Nature Medicine showed that exploiting this diminished anti-viral state in cancer cells could lead to the development of cancer cure based on lytic viruses with selective affinity for tumors.

First, the authors showed that cancer-associated fibroblast (CAF) showed [TGF-β dependent] increase in sensitivity to virus replication compared normal fibroblast harvested from the same cancer patient.

Next, the authors showed that cancer cells too become susceptible to viral infection when co-cultured with cancer associated fibroblast.

This viral titre enhancement was mediated by soluble factor.

Screen for active factors indicated that enhancing factor was fibroblast growth factor-2 (FGF-2). Indeed, inhibition of FGF-2 by RNAi reduced viral titre in tumor-CAF co-culture.

Finally, the authors showed that virus expressing FGF-2 could induce regression of established tumor in mouse model.

In summary, these results suggest that exploiting cancer vulnerability towards lytic viruses may be an alternative path for biological cancer therapy. 

Note: It is puzzling that virus targets only cancer cells. The authors explained such selective sensitivity of cancer cells based on reduced baseline anti-viral activity in cancer cells. However, we need to consider the fact that cancer patients maybe immune deficient in general, making them generally susceptible to viral infections.

David Usharauli