T-Cell Receptors

I've been thinking about T-cell receptors (TCRs) a lot lately, primarily because they're a delicious, tangled little knot of wonderful complexity. Some (Sciliz, I'm looking at you) call them fickle, but no, they're far tastier than that. Synaesthetic metaphors aside, TCRs are the gate-keepers between strong and solemn anti-pathogen protection and a raging inferno of doomy autoimmunity. More specifically, the genes that encode TCR are divided up into Lego-block-like segments [V, D, and J segments] that are sorted more-or-less-at-random*. This happens in each developing T-cell, and the genes undergo some further random mutagenesis via the RAG proteins to generate completely unique receptor specificities. In turn, this means that each TCR is unique, like a precious little snowflake or puppy nose, and it guards the body against its cognate antigen vigorously**. However, much unlike a snowflake or puppy nose, the mechanism by which TCRs generate useful signals and screen out non-specific noise is absolutely badass. In fact, rather than naming battleships after dead guys, who are no longer badass, perhaps we'd be better suited to name warships after T-cell receptor clone lines: Jurkat, D0.110,...others?

Anyway, I've been thinking a lot about the applications of network information theory to TCRs lately. But that's not what I'm going to talk about today, because in thinking about TCRs like that I've recently come to appreciate just how absolutely unrefined, perhaps even crass, my initial understanding of TCRs was.

Back when I first took an immunology course through my university's medical school extension thing, this was the impression I got of TCRs. G-protein coupled receptors (GPCRs) had been beaten into our skulls with the Alberts' Molecular Biology of the Cell text and as such I figured it worked a lot like a GPCR. This turned out to be rather incorrect. In fact, I was also under the impression that almost everything in the world was some sort of GPCR***. Fortunately, this too turned out to be incorrect.

The CD4/8 molecule pictured above is a rather handy cell-surface marker that is expressed, respectively, on CD4+ and CD8+ T-cells, which in turn respectively recognize antigen in MHC-II and MHC-I proteins. If that didn't make any sense, don't worry about it, what's important here is that CD4 or CD8 help the T-cells see the antigen that elicits a signal. CD4 and CD8 are co-stimulatory molecules without which antigen-presenting cells cannot effectively transmit information.

Anywho, as evolution would have it, there are more proteins involved in TCR signal transduction.
At the time I learned of the existence of CD3 and CD28, I had absolutely no clue what they did, so I merrily accepted that they were part of the TCR complex and moved on. Why I didn't look further into them is currently beyond me, but the most likely answer is that I was distracted at the time by free cookies. As such, I thought that signal transduction just carried on with CD3 and CD28 as handsome bookends.

Little did I know of the strange superpowers these two molecules have. Eventually, because science enjoys knocking over the obelisks of my ignorance despite how hard we've labored to keep them intact****, I gained that knowledge.
I couldn't believe it. It was like when I found out that I couldn't eat clouds. Signal transduction WITHOUT antigen stimulation? What sort of stochastic heresy was this? This shattered the beautifully deterministic world-view that my undergraduate education had instilled, that every receptor has a ligand that can be defined and that the cell is a place of beautiful complacency and order. I had known that ligands::signals weren't always 1::1, and I had accepted that, but this, this was somehow darker and more malevolent, and the knowledge that TCRs could be so deftly manipulated and easily fooled was tantalizing. This assumption, also, turned out to be wrong. TCRs are stubborn little bastards.

Nonetheless, I set aside my beloved copy of Janeway's Immunobiology and leaped into the scientific literature face first. I read cell biology papers, I read biochemistry papers, I read biophysics papers, I read autoimmune pathogenesis papers, I read genetics papers, I may have read some enzymology papers, and I know I read a bunch of computational biology papers. I emerged with a mouthful of bistable switches and nonlinear dynamics, and I believe that my understanding of molecular biology as a whole was greatly enhanced by it. Cells weren't orderly little machines, they were complex and messy tiny mechanisms that had error and chaos built in as part of their very functions. And that chaos and nonlinear responses can be harnessed by evolution to increase the sensitivity, response range, and efficiency of cells was absolutely beautiful in its inhumanity. Humans build orderly structures, boxes and flat surfaces and well-engineered machines, but cells don't given a damn! Cells get by, thrive, and capture the very entropy we try so hard to scrub out of our lives, and that, that right there was something that I didn't appreciate until I wound up climbing up atop the tallest stuff at hand one night (Halloween, exactly) and laid sprawled out on top of it watching the sky. Seeing the moon reel about the sky and the stars wheel away above it while the earth beneath me tangibly spun made me feel so very small and tiny and insignificant in the face of such a vast expanse as the universe. But then, at the same time I knew that I was composed of billions of microscopic cells to the point where we are complex molecular galaxies unto ourselves and that, too, made me feel small and insignificant. And lastly, knowing that the stars in the sky and the cells in my body will get by without giving a damn about whether I know they exist or appreciate them was a sort of beautiful thought that has remained with me, probably as the closest thing to faith that I've found.

Epic literature reading aside, the mechanisms of TCR activation as I now envision them are rather complex to draw, so I am going to use art instead.

Apotheose der während des Befreiungskrieges für das Vaterland gefallenen französischen Helden, by Anne-Louis Girodet-Trioson

This isn't meant to be tongue-in-cheek about it being epic*****, but more that it is incredibly crowded and busy and spatial location matters a LOT. You see, resting T-cells have TCRs scattered across their cell surfaces more or less at random, and each of those TCRs is identical. But once 1 of those is conjugated with the triggering antigen at just the right specificity, they all suddenly (within a matter of mere minutes) condense over to the site of that triggering and form an immunological synapse.

The synapse polarizes the TCRs all to 1 side of the cell and clusters them together. Actin also gets rearranged. This provides a cell surface anchor point for intracellular proteins to begin forming a complex cascade. The TCR proteins themselves change shape and CD3's intracellular domains are released from the cellular membrane where they were held. CD3 can then serve as a scaffold for effector kinases such Lck and Fyn. Some JAK/STAT scaffolding proteins eventually transmit the signal from Lck/Fyn/Zap70 to ERK and JNK, which in turn then affect transcription of NF-kB and other response genes such as IL-1b. At the same time, there is a massive calcium flux across the cellular membrane.

Like this. I had to take considerable liberties in omitting details to make this fit on one slide.

However, the synapse is presently more interesting than its effector functions. The immunological synapse is actually composed of 2 parts, the central supramolecular activation complex (cSMAC) and the peripheral supramolecular activation complex (pSMAC). As you might have already imagined, the cSMAC is surrounded by the pSMAC. Within the synapse, many T-cell surface proteins are able to interact with their ligands on the surface of the antigen-presenting cell and vice versa. The SMAC structures are stable (relatively for T-cells) and allow the antigen-presenting cells to properly stimulate the T-cell with both its cognate antigen and some pro-proliferation cytokines such as IL-2.

Now here's where I really have no idea how to visually illustrate it. Within the cSMAC, TCRs are systematically obliterated as the activated, presumably phosphorylated?, TCRs get pulled into the cell and chewed through proteasomes to allow fresher TCRs to get at the antigen-presenting action. This degradation of activated TCRs from the cSMAC is essential for proper T-cell activation (if you block it the T-cells keel over and die, as they are wont to do). And this is what is really frackin' cool: bistable switch behavior in the activation kinetics of TCRs is mediated first by formation of the immunological synapse, and then sustained by endocytosis and degradation of activated TCRs.

Toaster, what the blugoon is a bistable switch?

Hysteresis can be found in a lot of situations, including magnetization, memristors, and a whole lot of very useful electronics and materials sciences.

A bistable switch exhibits the property of hysteresis, which is rather awesome. Hysteresis essentially means that a signal going 1 way across some response action (sensor) doesn't necessarily get the opposite response when the signal goes the other way.

Hystersis can be found in TCRs at several levels. First, it is part of the kinetic proofreading that occurs at TCRs prior to cSMAC formation. TCRs have to be extraordinarily sensitive to their cognate antigen and must be vigilant in screening out false positives. Therefore it takes a stronger signal to activate TCRs than it does to deactivate them. Second, use of hysteresis in this is an efficient way of screening out false positives; the formation of the cSMAC and subsequent degradation of activated TCRs both help to sustain hysteresis behavior. Condensation of TCRs into a cSMAC requires a stronger activating signal than deactivating signal, which means it will remain active at a signal weaker than that which activated it. Thirdly, degradation of activated TCRs helps sustain hysteresis by ensuring that the signal is propagated into the cell at a proper rate. Too fast and the T-cell's built-in anti-autoimmunity mechanisms will kill the whole cell, and too slowly risks death by anergy, while still the net signal from an antigen presenting cell must be higher to activate the T-cell than to deactivate it.

There are many finer details of TCRs that I did not cover here, and still more that I probably don't yet know about. T-cells are an essential part of the immune system, able to differentiate into several different useful phenotypes. CD8+ cytotoxic T-cells can circulate around the body and survey for viral infection or cancerous abnormalities. CD4+ Th1 cells secrete massive amounts of IFNg and help direct innate immune effector cells to combat bacterial infections. CD4+ Th2 cells secrete lots of IL-4 and are involved in battling away parasites or, more relevantly in the Western world, causing allergies and asthma. CD4+ CD25+ regulatory T-cells shut down the immune response by secreting IL-10 after the pathogens have been cleared away to prevent Th1 T-cells from tearing up the place (see: cytokine storm). Memory T-cells lurk around and provide lasting immunity to previously encountered pathogens. And Th17 T-cells do something, they seem to be involved in protection against autoimmunity but we're not quite sure yet. Each of these T-cell phenotypes relies upon the TCR to detect their cognate antigen and help defend the body against the invading microbes that find its squishy, nutrient-rich nooks so very appealing. At the population level, I sometimes think of T-cells as a library of exquisitely-finely-tuned peptide detection machinery that we all carry around with us, and I find that chaos they usefully harness in hysteresis to be beautiful and captivating. There may be other, as of yet, undiscovered T-cell phenotypes and functions despite an already broad and deep literature, and that is an exciting prospect. T-cells and the exact mechanisms of their TCRs remain a relatively open biological frontier, and the best we can do is to dive in face-first.

*Not quite, but explaining it goes beyond the scope here. Look here for more information.
**If it doesn't get killed off first, as the grand majority of developing lymphocytes do, due to anergy or too strong a reaction against self.
***I really liked imagining that there were little Ggamma subunits shuttling around everywhere, all the time, in absolutely everything making a "BlootablootalootaLOO!" burbling sound as they went, with tiny ADP bubbles behind them.
****This is an obligatory joke notification footnote.
*****Also to help break up large blocks of text.