Circadian Rhythmicity - Intermittent Fasting (Re-)Sets the Peripheral Clock: Macros, Body & Liver Fat, AMPK & More

Fridge raiding in the middle of the night is just a side issue in today's installment of the Circadian Rhtythmicity Series that revolves around Intermittent Fasting with different macronutrient ratios and its effects on body weight, body fat, liver fat and the expression of circadian clock genes.

Everyone who's also following my posts on facebook will probably be aware that (a) I don't miss a single opportunity to kiddingly refer to scientists and journalists reading this series, whenever I point you towards a new paper or news item about circadian rhythmicity. He (or she) will yet also be aware that (b) one of the studies, I actually postponed to this installment of the series, made the rounds as of late, as the ScienceDaily piece on it - which was basically a "copy & paste job" of the EurekaAlert! press release - gave the impression that a "time-restricted" (others would call it "intermittent fasting") high fat diet would be superior to the carbohydrate-laden standard rodent chow, as far as the body compositional outcomes are concerned. I wrote and talked about this on facebook, as well as the SuppVersity Science News Roundup on Super Human Radio, last Thursday. Anyways,... today, I am going to deliver on my promise from two weeks ago and will eventually  discuss the results of the Sherman study (Sherman. 2012) and the influence of nutrient timing and composition on the  mammalian circadian rhythm in detail.

To fast or not to fast, this is not a question, anymore

As I have already pointed out in the "Break-Fast Installment" of this series, it is, at least from a circadian perspective, actually not a question of whether or not you should "fast", but more or less only one of how long this fast should last. An exact answer to this question has not yet been found, though, and I would even bet money that it never will, simply because it depends on too many confounding and highly individual (epi-)genetic and lifestyle factors. And still, based on what we know now, one thing can be said for sure: If you are standing up in the middle of the night either voluntarily or because whatever pathology may be driving you and have a protein shake or ransack the fridge, you have a problem. A psychological one in the first case, a physiological and probably clock-gene related one in the latter case.

*Always keep in mind: Mice are nocturnal animals, they start to party with pubertal humans (though this is meant as a joke, I suggest you take a look at my additions to figure 5, before you totally discard it as insignificant), when the lights go out. It is therefore only natural to restricting their food-intake to the so-called dark-phase - just as natural as not ransacking the fridge in the middle of the night would be for a normal human being!

That said, I don't think that the ad libitum (=whenever they wanted) fed mice in the Sherman study did have a fridge.  What's certain though is that the simple restriction of their food-access to a 4h time window that opened with the light being switched off* did have profound effects on the body composition of the mice in both, the low fat as well as the high fat groups. It did not, however, totally blunt the negative effects the above all energetically dense "high fat diet" had on the gynoid fat pads and the hepatic lipid content and thus sooner or later the liver function of the four-week-old male C57BL/6 mice (Sherman. 2012):

"[...] although body weight and epididymal fat mass were 20 and 48% lower, respectively, in the RF-HF group than in the AL-HF group, liver lipid content was not significantly different between these groups."

In view of the role a "fatty" and thus malfunctioning liver plays in the etiology of the metabolic syndrome, it can therefore not be said that the high fat diet, respectively the incarnation of the latter that was used in the study at hand, is either healthy or anti-obesogenic. After all, the increase in visceral and liver fat, of which only the former was slightly blunted in the study at hand, usually precede the development of the other classic features of a conglomerate of metabolic ailments, we usually refer to as "metabolic syndrome".

Less body weight, more liver fat - a pretty poor trade, I would say

Against that background, it is of questionable value that the overall body weight of the "intermittendly fasted" (=restricted feeding, short RF) rodents on the intentionally fattening "high fat diet" (HF) that contained 42% of the energy in the form of soybean oil and palm sterate, was lower than that of their peers on the ad-libitum (AL) high or low fat diets (AL-LF, AL-HF).

Figure 1: Comparison of gynoid (=epididymal) and liver fat (left) and body weight development (right) of male mice on ad-libitum (AL) or (time-)restricted (RF) low (LF) and high fat (HF) diets (based on Sherman. 2012)

Moreover, the significance of these allegedly "beneficial" effects, of which the author of the ScienceDaily article falsely states that they would also entail that the mice "on the scheduled high-fat diet [...] had a lower final body weight than the mice that ate an unscheduled low-fat diet" (Hebrew University. 2012), become even more questionable if we also take into account that ...

• the mice in the AL-LF (ad libitum, low fat) group consumed 6% more energy than their peers in the RF-HF group (restricted feeding, high fat) and
• the overall reduced body weight is unlikely to be solely the resul of lower body fat levels - specifically in the RF-HF group, where the fat is already beginning to clog the liver

And while the macronutrient and fatty acid composition of the diets does not allow for any definitive statements assertions about the value of real low carb diets with ~80% of the energy from fat, the data in figure 2 illustrates a couple of metabolically relevant differences that arose as a result of the different macronutrient composition of the diets (please remember that I plotted only values with significant inter-group differences between low and high fat groups, yet not those difference, which are attributable solely to the intermittent feeding / fasting regimen and have no relation to the macronutrient composition).

Figure 2: Metabolically relevant serum markers with statistically significant differences that can be attributed to the differences in macronutrient composition of the diets; arrows indicate changes that are more or less characteristic for "fasting" as in "not eating" or being in a energy deficit (data expressed rleative to AL-LF group; based on Sherman. 2012)

If you take a closer look at the data in figure 2 you will probably realize that the classic effects of fasting, i.e. a reduction in leptin, a reduction in insulin, a reduction in glucose, a reduction in triglycerides and increases in glucocorticoids (to squeeze the fat out of the cells and catabolize amino acids in order to stabilize the blood sugar via gluconeogenesis) and the "hunger hormone" ghrelin are either mitigated or almost blunted by the "high fat" approach to intermittent fasting (I repeat this once more: the "high fat" diet in this study has more in common with the standard American diet than with a ketogenic diet). This raises the question: Is this a good or a bad thing? Or, put another way: If this does not look like fasting does it have the same effects on the circadian clock as "real" fasting would? The study results suggest that this is the case - albeit with a small, but potentially relevant quantitative difference in the fasting induced phase shifts:

Figure 3: The RF-HF mice were the almost twice as active as the AL-HF mice and 42% more active than the RF-HF animals (of whom Sherman et al. write that they were about as active as the AL-LF mice)

• the casein kinase Iε (CkIε) oszillated robustly only in the low fat groups 
• "high fat" fasting reversed the phase advance of the circadian expression of Per1 and produced a phase advance of the previously phase-delayed expression of the Cry1 and Ror genes
• "low fat" fasting induced a phase advance in all measured clock genes
• compared to the high fat restricted feeding regimen the low fat restricted feeding group exhibited a phase delay with respect to the Bmal1 and Cry 1 genes

While the overall image that emerges here, i.e. the ability of a timed feeding regimen (aka intermittent fasting) to rectify and/or advance the peripheral (here, hepatic) shifts induced by the HF diet, is pretty obvious, our understanding of the interactions of the individual clock genes is still too premature to say something definitive about the implications of the registered changes.

"So what's that supposed to mean, now? Is fasting good? Is fat bad? ... or what?"

No effect of macros? In as much as these observations support that restricted feeding aka intermittent fasting alone can reverse the majority of negative effects on the peripheral expression of clock genes, it says nothing about their central expression (all values here were measured in the liver!) and the effects on body composition. The increase in pAMPK, for example speaks in favor of the low fat diet. The same is true for the potential longevity effects. Plus, it is essentially pointless to speculate about the best macronutrient ratios, as long as a clone of the fattening standard American diet, of which I do simply assume no one of you will truly believe that it would deliver better results than a low fat diet in a rodent model, is all we can compare the low fat approach to intermittent fasting to (read more on the practical side of things in the next installment).

If we use the results of Um et al., who found in 2007, already, that the expression of the aforementioned casein kinase Iε (CKIε) gene and its negative feedback mechanism on another clock gene, namely mPER2, are triggered by the AMPK promoting effects of the "wonder-drug" metformin (Um. 2007), the following observations Sherman made with respect to the expression of AMPK and other proteins that are involved in the fatty acid and glucose metabolism begin to make some sense:

• The RF-LF diet led to increased levels of pAMPK and pACC, indicating intracellular low energy levels, inhibition of fatty acid synthesis and increased fatty acid oxidation. 
• The AL-HF diet down-regulated AMPK, ACC and SIRT1 daily protein levels by 50% compared with AL-LF mice.
• The timed HF diet led to 37% lower levels of pAMPK than those in the RF-LF group and 62% increased pACC levels compared with the AL-LF group, indicating adequate energy levels but reduced fatty acid synthesis .
• The RF-HF diet also increased daily levels of PPARα mRNA.

The overall effect on the expression of AMPK, Sirt1 and ACC, i.e. the reversal of most of the negative effects of the high fat diet in response to the intermittent feeding regimen, is thus very similar to what we have seen in the case of the the clock genes (see previous list). Reason enough for Sherman et al. to conclude that "[t]hese results demonstrate again the dominance of the timed feeding over the HF diet." (Sherman. 2012) - or, as a take home message for you:

At least with respect to the circadian rhythm, it appears that...
when you eat is of greater importance, than what you eat!

Central vs. peripheral, master vs. slave, light vs. food - every orchestrate needs a director

Now that we have gained insight into the priority of when you eat over what you eat as far as their ability to correct, readjust and resynchronize the peripheral circadian clock are concerned, we are facing another, a follow up question that reads "Is when you eat also more important than what the light cues are telling your suprachiasmatic nucleus?" [(re-)read Part I & Part II of this series to learn all about the importance of light exposure]
 

Figure 5: Regardless of the model you prefer, the central clock and with it its exclusive direct regulator will always be the most important factor in the circadian master/slave or orchestrate system (based on Richard. 2012. Fig. 3)

In one of the most recent reviews of the currently available literature on circadian rhythmicity, Richard & Gymz write about the relation of central (=brain) and peripheral (=organs) aspects:

On a basic level, the circadian clock can be divided into 2 parts: the central clock, residing in the suprachiasmatic nucleus (SCN) of the brain, and the peripheral clocks that are present in nearly every tissue and organ system tested. Light enters through the retina of the eye, causing electrical signals to pass through the retinal hypothalamic tract, which are converted to chemical signals in the SCN. Light signals and other physiological factors, such as feeding cues, entrain the central circadian clock. There has been much debate among chronobiologists about the relationship between the central clock and the peripheral clock with 2 major theories emerging.

These two theories scientists have come up with do explain the complex interrelation between the mostly light-driven central and the various peripheral clocks based on either a ...

• "master-slave" model which gives complete synchronization power to the central clock and thus assumes that all peripheral clocks are centrally synchronized in the brain, or an
• "orchestra" model according to which the multiple peripheral clocks are like the members of an orchestra, with each of them playing its own "instrument"

If you don't want the metabolic concert of your body to end up as a cacophonous cat's concert as in "David Teniers the Younger Cat concert" (see image), you better make sure that all the member's of it's orchestrate are in time with the director, the central and light dependent circadian clock in the suprachiasmatic nucleus of your hypothalamus.

The "orchestra" model has the beauty of ascribing greater importance to the previously discussed effects of feeding-times and (macro-)nutrient composition - a position that is backed up by the majority of recent studies, by the way (Richard. 2012).

"Thus, each peripheral clock can adapt to its own external and internal stimuli, such as feeding cues for the liver, kidney, and pancreas, but is "conducted" by the light-dark cues sensed by the central clock." (Richard. 2012, my emphases)

In other words, while your "peripheral clock", e.g. the clock of your fatty acid metabolism in your liver can go wrong, you better make sure that it does not, because in the end, it does not really matter how "independent" each of the members of an orchestra may be. When they start playing, their synchronization by the director, or, analogously, appropriate light cues determines whether the overall outcome of this intricate metabolic concert is a symphony and their concert hall, i.e. you(!) lean & healthy or a cacophonous mess that's making you fat & sick!

Light over Food Timing & Food Timing over Food Types.
On that note: Screens out for today, folks!

References:

• Hebrew University of Jerusalem. "A carefully scheduled high-fat diet resets metabolism and prevents obesity, researchers find." ScienceDaily, 12 Sep. 2012. Web. 16 Sep. 2012.
• Richards J, Gumz ML. Advances in understanding the peripheral circadian clocks. FASEB J. 2012 Sep;26(9):3602-13. Epub 2012 Jun 1.
• Sherman H, Genzer Y, Cohen R, Chapnik N, Madar Z, Froy O. Timed high-fat diet resets circadian metabolism and prevents obesity. FASEB J. 2012 Aug;26(8):3493-502.
• Um, J. H., Yang, S., Yamazaki, S., Kang, H., Viollet, B., Foretz, M., and Chung, J. H. (2007) Activation of 5=-AMP-activated kinase with diabetes drug metformin induces casein kinase Iε
  (CKIε)-dependent degradation of clock protein mPER2. J. Biol. hem. 282, 20794–20798