When teaching human or animal physiology, it is very easy to come up with examples of ubiqutous negative feedback loops. On the other hand, there are very few physiological processes that can serve as examples of positive feedback. These include opening of the ion channels during the action potential, the blood clotting cascade, emptying of the urinary bladder, copulation, breastfeeding and childbirth. The last two (and perhaps the last three!) involve the hormone oxytocin. The childbirth, at least in humans, is a canonical example and the standard story goes roughly like this:
When the baby is ready to go out (and there's no stopping it at this point!), it releases a hormone that triggers the first contraction of the uterus. The contraction of the uterus pushes the baby out a little. That movement of the baby stretches the wall of the uterus. The wall of the uterus contains stretch receptors which send signals to the brain. In response to the signal, the brain (actually the posterior portion of the pituitary gland, which is an outgrowth of the brain) releases hormone oxytocin. Oxytocin gets into the bloodstream and reaches the uterus triggering the next contraction which, in turn, moves the baby which further stretches the wall of the uterus, which results in more release of oxytocin...and so on, until the baby is expelled, when everything returns to normal.
As usual, introductory textbook material lags by a few years (or decades) behind the current state of scientific understanding. And a brand new paper just added a new monkeywrench into the story. Oxytocin in the Circadian Timing of Birth by Jeffrey Roizen, Christina E. Luedke, Erik D. Herzog and Louis J. Muglia was published last Tuesday night and I have been poring over it since then. It is a very short paper, yet there is so much there to think about! Oh, and of course I was going to comment on a paper by Erik Herzog - you knew that was coming! Not just that he is my friend, but he also tends to ask all the questions I consider interesting in my field, including questions I wanted to answer myself while I was still in the lab (so I live vicariously though his papers and blog about every one of them).
Unfortunately, I have not found time yet to write a Clock Tutorial on the fascinating topic of embryonic development of the circadian system in mammals and the transfer of circadian time from mother to fetus - a link to it would have worked wonderfully here - so I'll have to make shortcuts, but I hope that the gist of the paper will be clear anyway.
There are two aspects to timing of birth. The first one is the determination how long the pregnancy will last, i.e., on which day the birth will occur. The second one determines at what time of day the labor will start. The first is not dependent on the circadian clock (e.g., this is not affected by lesions to the suprachiasmatic nucleus, SCN), while the second one is determined by the circadian clock. In all mammals studied to date, there is a strong circadian control of the timing of birth, likely related to the safest time of day for that particular species. My complete lack of interest in human chronobiology while in grad school makes my memory really hazy on this, but I think I remember that majority of human births (especially in more primitive societies without artificial light, epidurals, Ceasarians and induced labor) occurs somewhere around dawn (correct me if I am wrong, please).
Also, the current understanding is that the first one - duration of pregnancy - is determined by the fetus, while the second one - time of day of labor - is determined by the mother.
Surprisingly to me, the paper cites two earlier studies that show that oxytocin is NOT necessary for birth (but is necessary for milk ejection in nursing) in mice. I was not aware of this (should I blame my inability to access Closed Access literature now?). It is not sure that the same applies to humans, but let's look at mice first anyway. How do the little mice come out of their mother without oxytocin? Is the uterus contracting due to other influences or is it possible for mice to get born without uterine contractions? I don't know.
Both the release of oxytocin and the availability of receptors at the target cells shows robust circadian rhythms in mammals. It is thought that the clock uses oxytocin as a time-signal for the onset of birth, i.e., as a way for the central pacemaker in the SCN to either to entrain the peripheral clock in the uterus, or to "gate" the timing of birth, i.e., determine the permissible and impermissible times for the onset of labor. This paper was set out to test this hypothesis by comparing timing of births in oxytocin-deficient mice and wildtype mice. Both types of mice were kept in light-dark cycles and the cycle was either phase-advanced by 6 hours during the last few days of pregnancy, or phase-delayed by 6 hours, or left as it was initially.
The results are quite surprising and thought-provoking. In wildtype mice, those in which oxytocin is present, the births were clustered around a single time of day. However, the phase-shifting of the light cycle did not phase-shift the timing of birth! And it also did not shift the timing of oxytocin release either. This means that the circadian rhythm of oxytocin release is driven by a pacemaker that is separate and independent from the pacemaker governing the locomotor activity. The location of this pacemaker is yet to be elucidated, but it could be a subset of cell in the SCN, or a yet-unknown clock located in the paraventricular nucleus of the hypothalamus, or in the posterior pituitary itself. Anyway, the results support the hypothesis that oxytocin is in some way involved in the timing of birth.
In oxytocin-deficient (OT) mice, the births also clustered around the same time of day as in the wildtype mice. But, in OT mice that experienced a shift in the light cycle, the births were happening at random times of day - the shift somehow disrupted the circadian gating of labor (and in two cases even resulted in the death during labor). This suggests that oxytocin is involved, but that it is not the only factor that affects the timing of birth.
With oxytocin present, the timing of birth is determined and stable, even resistant to shifts in the light cycle. Without oxytocin, the clock that times labor still operates, but is now susceptible to shifts in light-cycle (transduced by as yet unidentified signal - melatonin?). This poses the question of the location of the clock that times the labor. Is it the same one that times oxytocin release? Or is it located in the fetus itself, challenging the notion that it is the mother who times labor? Is it some kind of a feedback loop between a mother clock and a fetal clock that determines the timing of birth? And how much is this result in mice relevant to humans?
I love papers like this! A short paper, describing an apparently simple experiment, yet it opens all sorts of cans of worms, provokes so many new questions and, hopefully, will result in a lot of interesting future research.
"My complete lack of interest in human chronobiology while in grad school makes my memory really hazy on this, but I think I remember that majority of human births (especially in more primitive societies without artificial light, epidurals, Ceasarians and induced labor) occurs somewhere around dawn (correct me if I am wrong, please)."
Yes, or rather, labor is more common at night (I don't know about trends for time of day at delivery off the top of my head). The adaptive argument for why has to do with higher concentrations of group members at night because humans are diurnal, so as to lessen predation risk during labor and delivery. This isn't to say that all women in traditional societies give birth in the middle of the night with group members huddled around--certain tribes (including the !Kung) are noted for unassisted childbirth away from the group--but it does seem to be an overall timing pattern (with wide variation) that is roughly consistent across groups, from what I know.
Fascinating! Thanks for blogging about this.
Very cool. I love this kind of paper too--short and simple, yet opening into a million new directions for thought/research. The kind of paper that sends me into a frenzy of geeky excitement that makes my boyfriend look at me like I have two heads.
There are actually quite a few positive-feedback phenomena in physiology (compartment syndrome comes to mind) but all of the ones I know are pathological.