Exploring the Human Chronobiome and Precision Medicine Approaches
Exploring the Human Chronobiome and Precision Medicine Approaches
Complete the form below and we will email you a PDF version of "Exploring the Human Chronobiome and Precision Medicine Approaches"
Complete the form below to unlock access to this Audio Article: "Exploring the Human Chronobiome and Precision Medicine Approaches"
Disease symptoms, treatment efficacy and many physiological processes within the body vary depending on the time of day. The molecular circadian clock coordinates the body’s rhythms for each individual person. The resulting unique pattern of physiological traits within a 24-hour period can be described as their “chronobiome”. We recently spoke with Carsten Skarke, MD, the McNeil Fellow in Translational Medicine and Therapeutics at the University of Pennsylvania to learn more about his work on the human chronobiome, its relevance in terms of timing the delivery of therapeutics and developing precision medicines.
Joanna Owens (JO): Could you explain what you mean by the “chronobiome”?
Carsten Skarke (CS): We first coined the term “chronobiome” in 2017. It’s based upon the idea that you take what is already a very comprehensive assessment of the human condition – integrating a deep multi-omics characterization of an individual at a molecular level to study how these molecular networks are influenced by behavior, environmental stimuli or stressors – and then you take time of day into account. The sum of that is the chronobiome.
JO: How do you approach studying that complexity?
CS: Most of what we know about circadian rhythms has been teased out from studying model organisms, which is wonderfully helpful to inform hypotheses. But how well does that knowledge translate to the human domain? In Germany in the 1960s, they studied the effects of lack of light on the human circadian rhythm of people living in bunkers. But how would you do those studies at scale? We started with a small pilot study in which we followed six healthy urban young professionals. We asked, if we take participants and study them while they're following their individual schedule, would we be able to pick up time-of-day dependent differences in molecular networks and behavior? Usually in an experimental setup, you control the experiment to limit the noise. In our setup, we allowed this noise to enter our experimental data. We just recorded it.
JO: What did your pilot study find?
CS: Medicine has been established by taking cohorts of either healthy volunteers or patients to look at drug effects, and then a physician treats or selects a treatment based on an average signal from these cohorts. That average signal puts less weight on the between-person difference in how we are. Nonetheless, in our pilot we took a group of healthy male subjects, and we studied them for two 48-hour periods, spaced two weeks apart. I was expecting that the metabolic and molecular signatures we found would be similar between week one and week two, for each individual person. And this is true for some metabolites (an estimated 5–6% of what we measured), some proteins and a few gut bacteria. But for the majority, they were starkly different – i.e., high levels of a metabolite in the morning and low in the evening in the first study period, and then a markedly different pattern or scale of response two weeks later. What we learned, from multiple measurements in six healthy volunteers, is that maybe we are molecularly very noisy.
JO: What are the implications of this finding?
CS: It is early days, we need to expand this to take into consideration more people, including women, as well as look at the impact of age and seasonal differences. But the potential implications of such variation are huge. If you look at the literature, you will find many cases where it is argued that disruption of circadian clocks or biorhythm is contributing to the development of a particular disease. And of course, we know that if you carry out several decades of chronic shift work, you have a dose-dependent increase in risk of developing chronic diseases like hypertension, diabetes and cancer. But that suggests a picture that people are very disrupted in their biological rhythms when they're sick, so it looks very noisy, and conversely, when you are healthy, you should look very “clean”, with a regular rhythmic signal.
But from our pilot data, we can say that for some molecular signals things do look very rhythmic, but that is not the case for many others. So, if you can be healthy and be molecularly “noisy”, then maybe the difference in risk of disease isn’t whether it disrupts “normal” rhythms, but perhaps instead it might be how you respond to stress or to a particular stimulus. There might be a difference, for example, in the way a molecular network in someone with a complex disease reverts back to baseline compared with someone who is fit and well. We are already seeing this in areas like the microbiome; if you give someone an antibiotic that eradicates the healthy gut microbiome, it recovers, but many studies suggest that the recovered microbiota is slightly different than that observed pre-treatment. In that same sense, maybe what differentiates you as a healthy person versus a person at risk of disease, is the difference in your molecular response to stressors in a time of day-dependent way.
JO: How will understanding the human chronobiome help to resolve this?
CS: We need to have a reference data set, a reference of a healthy population to discern in a time-of-day-dependent context what we consider as healthy and what we consider as associated with disease. Because again, in the literature, you will find many different physiological markers associated with a disease. But my expectation is that having a better understanding of the healthy chronobiome will let us better understand whether signals measured in a patient are out of the normal value range because of the disease or just variability or “noise” as observed in healthy people.
JO: What do you see as the next steps for this research?
CS: To fully understand the human chronobiome we need to study it in healthy people and then use stressors to evoke phenotypes, e.g., by giving a fatty meal or using lipopolysaccharide to induce an inflammatory response.
We are also in wonderful times right now because of all the wearables and the technology we use to analyze physiology is so scalable. Take single-cell transcriptomics for example, we can hone extremely deep into the molecular underpinnings of cells but can also analyze cells at an unprecedented scale. One vision is to use what I call “easily collectible” data. Now we all have smartphones, which are essentially mini-computers, and we can get an awful lot of valuable information from these phones coupled to wearables like accelerometers or electrocardiogram monitors. This establishes a framework where we need to do little in order to generate data.
The ultimate question for us is, how would a framework using these approaches trigger a visit to the hospital for a blood sample, for example. For patients, it could mean that their phone constantly monitors their physiology, and you have already defined variability thresholds so that when that extends a certain threshold, it triggers a visit to a clinician for a deeper assessment. That's where I think this research is headed. Take Alzheimer's disease (AD) for example, the healthcare system as it is currently, is set up is to be mostly responsive, rather than preventive. The growing incidence of AD and other chronic diseases will challenge the system massively. We need to make the switch to a preventive approach, and I think this needs to happen on a global scale.
JO: How long do you think it will be before we fully understand the chronobiome?
CS: We need to get to a point where we know the extent of the diurnal variability in these molecular networks and then we need to bridge that to treatment effects, and ultimately to outcome. It will undoubtedly be a long road as there are certainly more questions than answers currently. In biology we are taking in all kinds of measurements, the more the better. And this is because we don't know which is the most important signal to which we can associate any future outcome. Ultimately the question will become “what is the minimum set of longitudinal assessments in a patient to give certain answers we need for individualizing a therapeutic approach?”
Carsten Skarke was speaking with Joanna Owens, PhD, Director of Faraday Science Communications.