Your body keeps its own time. It has an internal 24-hour “circadian clock” that drives the rise and fall of many molecules. Everything from brain activity to hormone levels waxes and wanes according to these molecular metronomes, which dictate how hungry, hot and sleepy we are.
They also affect how well we respond to medicine. Since the late 1980s, scientists have shown that drugs work better at certain times of the day. For example, the cancer drug cisplatin is more effective and less toxic if it’s given in the evening. Adriamycin is more of a morning drug. In another cancer trial, tailoring chemotherapy to these daily rhythms—a practice known as chronotherapy—made the same drugs more effective and reduced the frequency of toxic side effects.
Chronotherapy would seem to be a no-brainer but it hasn’t caught on widely. That may be partly due to scepticism, but there’s a more practical reason: it’s hard to read a person’s body clock. Some people are larks, others are owls. The ticks and tocks of the clock vary depending on age, sex, health, employment, and more. The clocks of two people can be half a day apart. How do you administer a drug at the right time if you can’t tell that time?
The conventional way would be to take blood samples every hour or so for 24 hours, and measure the concentrations of melatonin—a hormone that rises in darkness and falls in light. Melatonin can be detected in saliva samples but because the hormone is found in such low concentrations, the process can’t be automated. As such, it’s labour-intensive work that takes days and tightly controlled environmental conditions. If you have patients to treat, you rarely have such luxuries.
Takeya Kasukawa and Masahiro Sugimoto from the RIKEN Center for Developmental Biology have a better way. Their team have developed a “metabolite timetable” that plots how dozens of molecules rise and fall in relation to one another. With this timetable, they could accurately read a person’s internal clock with just two blood samples, taken 12 hours apart.
It was inspired by flowers. In 1751, Swedish biologist Carl Linnaeus (the father of biological classification) suggested that a carefully planted garden could double as a clock. Since certain flowers open or close at particular times of the day, you could tell the time by planting dozens of species and checking how open they were. Each individual flower could be inaccurate, since weather and seasons would change their rhythms. But look at dozens, and you iron out the variation.
The metabolite timetable works on the same principles. Back in 2004, Hiroki Ueda, who led the new study, showed that this approach could work based on the activity of genes in mice. In 2009, he showed that metabolites could be used in the same way. Now, he has transferred this technique, which had only worked in mice, into humans.
The team worked with six volunteers who spent 2 weeks in a special sleep laboratory – a small apartment kept at a constant temperature and humidity, and completely devoid of clocks or any other time-keepers. They woke, slept and ate at specified times, and seven researchers kept a constant eye on them. At the start and end of their two-week stay, they spent 39 hours awake in a reclining chair, while Ueda’s team took blood samples through a catheter every two hours.
The researchers looked at the first 39-hour samples from three of their volunteers, whose rhythms best agreed with each other. They found more than 4,000 molecules whose levels waxed and waned, and whittled this list down to 58 that showed the strongest and most regular signals. These included a mix of hormones, fats, and amino acids.
They created their timetable by calibrating the levels of these substances against typical markers, and tested it using their remaining 39-hour samples – the second batch from same three volunteers, and both sets from the other three. By evaluating at the levels of the 58 chosen chemicals, the team managed to estimate the volunteers’ body time to within three hours of their actual body time (and mostly, to within two hours).
Michael Terman, a psychiatrist who works on body clocks, is not impressed. “The range of error is nothing to brag about for a system with inherent precision measured in minutes,” he says, although he notes that more samples would improve accuracy. Terman also says that many of the molecules included in the timetable aren’t independent. For example, cortisol is on the list, as are several hormones that are produced by metabolising cortisol.
John O’Neill, a body clock researcher at the University of Cambridge, says that Ueda’s new study isn’t a technical advance over his earlier work but it “represented an important step forward in its emphasis on real world application.
“In terms of buzzwords, “chronotherapeutics” is going to become the new “pharmacogenomics”, says O’Neill. The latter term refers to tailoring treatments according to a person’s genome. It’s all the rage, but hampered by an incomplete understanding of the interactions between drugs and different genetic backgrounds. Chronotherapy could provide a simpler way of tailoring treatments and making them more effective. “A lot of big pharma research and investment is going into exploring whether, for example, a drug which failed phase III clinical trials, and therefore represents millions of dollars of lost investment, might pass if it were dosed at a different time of day,” says O’Neill.
But this approach will only work if we have an easy, reliable way of reading a person’s biological clock. Ueda’s study edges us closer to that. “It is [now] essential to see whether this metabolite timetable holds true for humans living their normal lives outside of the controlled laboratory environment, and to what extent the metabolite profiles may be affected by environment, age, sex and genetic background,” says O’Neill.
Ueda acknowledges that. His current timetable is a proof-of-principle, built using only three select people. But he thinks it should be possible to scale that up to create a schedule that would apply to a large population. He also wants to make the timetable so accurate that you can read a person’s body clock from just one blood sample, rather than the current two.
But even if this approach was refined, Terman is unconvinced that it would be useful in a clinical setting. “When a clinician meets a patient in need of chronotherapeutic intervention, the goal is immediate decision making,” he says. “Subjecting the patient to a day-long diagnostic procedure under controlled environmental conditions entails treatment delay, and may be infeasible in many cases.”
Terman says it would be just as useful, and much simpler, to work out a person’s chronotype—whether they’re a night person (owl) or a day one(lark). These chronotypes correlated with the substances that Ueda identified, and Terman says, “[They] can be “measured in minutes with a paper-and-pencil questionnaire, with an error range no different from using the metabolite timetable.”
Reference: Kasukawa, Sugimoto, Hida, Minami, Mori, Honma, Honma, Mishima, Soga & Ueda. 2012. Human blood metabolite timetable indicates internal body time. PNAS http://dx.doi.org/10.1073/pnas.1207768109
Image by Paco
More on body clocks:
- As oxygen filled the world, life’s universal clock began to tick
- Transplanted arteries stiffen up if their clocks are broken
- Genetic snooze button shows that broken sleep impairs memories
- Pocket Science – chameleons hunt with cold-proof tongues and zebrafish babies go blind at night
- Clock gene and moonlight help corals to co-ordinate a mass annual orgy