bistable circuits, hysteresis, and cell fate

i've been obsessed with posting about this, and i don't even know why, but here goes.

in biology, cells receive inputs from other cells, or the environment, and they integrate those inputs in order to make decisions - whether to produce a hormone, or to differentiate along one pathway or another, whether to live or die, etc. what's interesting about this situation is that the inputs are quite commonly analog - that is, you can linearly vary the amount of stimulus - but the required outputs are often binary - they are decisions, with pretty much an 'on' or 'off' choice. in other words, you don't want a cell to ever decide to 'kind of' die, or 'kind of' differentiate - you want all the way or nothing.

here's where positive feedback loops come into the picture. a positive feedback loop will operate such that you activate some mediator A. A then actives mediator B, which then goes back and activates more of mediator A. Now, positive feedback loops operate such that the loop won't kick in until you pass a certain level/frequency of stimulus, but after that, you'll 'jump' to a second level of signal (because the feedback has kicked in). the system will then equilibrate at one level or the other, depending on the initial stimulus - this is basically what's known as a bistable circuit. you can already see how this might be helpful to executing an on/off decision.

so where does hysteresis come in? well, hysteresis refers to the fact that, once the feedback loop has kicked in, and you have 'jumped' to the second level of activity, if you stop giving input, and then restart your input, you will more easily jump to a second level of activity - aka, the threshold for achieving a highly activated state is lowered once you've passed the initial threshold. this is important for signal integration in cells receiving discontinuous stimuli. without hysteresis, a cell receiving a super-threshold stimulus will get very little output from subsequent sub-threshold stimuli, whereas hysteresis allows subsequent stimuli to have a much larger impact in the cell's integrated output. this has the effect of making a cell far more sensitive to inputs. it should be mentioned that one of the negative aspects of hysteresis is irreversibility - in which the loop can generate so much of a feedback that achieving the "off state" is no longer possible - the cell is permanently 'on'. This loss of plasticity can, of course, be damaging.

why is this all important? Arup Chakraborty and colleagues modeled a system like this in lymphocytes, and then experimentally determined that this is, in fact, how Ras signaling occurs in lymphocytes. Now this is important because there are a number of aspects of lymphocyte activation that should be digital - thymic selection and t helper cell differentiation to name just a few. its also pretty cool that he used mathematical modeling to simulate a complex phenomenon and direct his experimental research, which validated his models. What's also interesting is that the digital behavior applies to Ras, whose signaling paradigm is critical to central tolerance of CD8 T cells (described by Ed Palmer, left).

im sure i've bored you all to death.