Summary: arguing that cell-intrinsic encoding of antigen-agnostic memory will strictly limit its durability, we propose a cell-extrinsic mechanism wherein the memory durability is no longer constrained by per-cell dynamics, tuned instead by the overall immune circuit dynamics.

Summary: arguing that cell-intrinsic encoding of antigen-agnostic memory will strictly limit its durability, we propose a cell-extrinsic mechanism wherein the memory durability is no longer constrained by per-cell dynamics, tuned instead by the overall immune circuit dynamics.

Using computer simulations and a machine learning approach, we show that feedback strength is a key determinant of memory durability. Importantly, strong enough feedback can overcome the constraints from cell turnover / epigenetic state transmission to drive long-lasting memory!

Positive feedback in bystander activation dynamics involving IL-15 / IFN-gamma signaling is well-known; we find evidence for positive feedback in trained immunity as well, driven by cytokines including IL-6 and IL-1beta. How does this effect memory durability?

How, then, can we have long-lasting bystander activation memory and trained immunity as reported across studies in mice and humans? Short answer: positive feedback in immune cell circuits!

With trained immunity, the durability is further constrained by epigenetic state transmission fidelity during self-renewal / differentiation of immune cells. Crucially, even moderately long-lasting memory (a few weeks) requires very high fidelity which is biologically unlikely!

In this preprint, we developed mathematical models of bystander activation and trained immunity dynamics and used model simulations to show that memory durability is ordinarily constrained by immune cell turnover and cytokine dependence.

While both these memory modalities have been shown to effect the response to a subsequent inflammatory challenge, dynamics of such memory and the immune parameters that determine its durability remain largely unknown.

Antigen-agnostic memory includes bystander activation memory, encoded by counts of memory T cells not specific to the antigen, and trained immunity, encoded by the altered epigenetic state of innate immune cells and hematopoietic progenitors.

Experiments have shown that in addition to antigen-specific memory involving memory T and B cells, infection / vaccination can induce various antigen-agnostic changes— in immune cell counts and in the epigenetic state of immune cells— that encode memory of past exposure.

The topology of this circuit, along with the relative bias of the Treg TCR repertoire towards self-antigen recognition, could be sufficient to explain how Treg cells can restrain autoimmune responses at homeostasis / during infection and restrain immunopathology.