Lithium, Circadian Clocks and Bipolar Disorder

I have previously only touched on the immensely interesting topic of the possible connection between circadian clocks and the Bipolar Disorder. A recent paper prompted me to look into this in a little more detail.

Lithium Affects the Circadian Clock

First, let's go a little bit into the past, the early history of chronobiology. During the 1940s and 1950s, while the field was still in its pioneering spirit and little was known about the circadian clocks, many researchers were using survey (or shot-gun) approaches to the studies of biological rhythms: studying as many organisms as they could get their hands on in order to come up with generalities and evolutionary answers, surgically removing every possible organ or brain region in order to find locations of clocks in various organisms, exposing the organisms to every possible light regimen imaginable in order to study the oscillatory properties of biological clocks, etc.

One of the approaches was to administer to animals every chemical one could find on the lab-shelf to see how it affects the circadian rhythms. This line of work yielded a big surprise - biological clocks are amazingly resistant to pharmacological agents. The few substances that had an effect were hormone melatonin (naturally, as it is the main signaling molecule of the circadian system), heavy water (deuterium oxide) and lithium (a few others were found much later, including sex steroid hormones). Lithium had the same effect - slowing down the clock, i.e., increasing the period - in a number of philogenetically very distant organisms.

Lithium affects the Bipolar Disorder

At the same time, lithium was one of the most prescribed drugs for treating bipolar disorder (at that time usually called "manic-depressive disorder"). Soon enough, people started making the links between effects of lithium on bipolar dissorder and the effects of lithium on the circadian clock. Is the bipolar disorder essentialy a circadian clock disorder?

During periods of depression, the circadian rhythms are phase-advanced (click to enlarge):


Lithium is supposed to phase-delay the phase-advanced rhythms, i.e., bring them back to the normal phase. Here is an actograph of the sleep-wake cycle of a bipolar patient treated with various drugs, including lithium, as well as phase-shifts of the light-dark cycle, over a long period of time - click on the image to enlarge so you can read the text:



This does not appear to be a very efficient treatment by lithium in this particular patient, though.

Lithium Affects Circadian Pacemaker Cells in a dish

Much more recently, it was discovered that each individual pacemaker cell (in the suprachiasmatic nucleus of the hypothalamus) in the mammalian circadian system responds to lithium. In other words, the effects of lithium are not at the system level (e.g., interfering with cell-cell communication), but on the level of the cell. This suggests that lithium may act on a particular clock gene and the search for the gene in question commenced.



To make things easier, the candidate clock-gene target of lithium is likely not to be limited to mammals, or vertebrates, as lithium has the same effects on rhtyhms in other organisms, including the fruitfly Drosophila melanogaster. Thus, it is likely that the target clock gene is one that is shared by the circadian clocks in Invertebrates and Vertebrates, thus somewhat narrowing down the list of candidates.

Molecular Mechanism of Circadian Rhythm Generation in Mammals

Let me now try to explain how the mammalian circadian clock works on the molecular level in as simple way as possible, so the non-scientists reading this can - hopefully - understand. Biologists can follow the links for more detailed information if so inclined. In order to do this, I will first give a super-simple primer on molecular biology (I hope I don't make any stupid mistakes on this part as I type it very fast in order to get to the cool new stuff). This is an oversimplification, so I hope molecular biologists do not chastise me for omitting all the extraneous details, as much as they may be important. This is BIO 101.

We are all composed of billions of cells. All of the genetic material - DNA - is found in the nucleus of each cell. DNA is a very long linear molecule, built like a chain out of many, many links. The links in the chain are the nucleotides, each made of a sugar molecule, a phosphate and a nucleic acid. There are four types of nucleic acids in the DNA: adenine, thymine, cytosine and guanine (A, T, C and G). The order of links with different types of nucleic acids on the DNA chain is the "code".


A gene is a small string on the long DNA chain - a sequence of nucleotides that is transcribed as a unit. Transcription is the formation of an RNA molecule - also a chain - using the DNA as a template. Thus, transcription makes an RNA molecule that is a mirror image of the gene. Wherever in the DNA sequence there was C, in the RNA there will be G, and vice versa. Wherever there was T in the gene sequence, there will be A on the RNA transcript, and vice versa (with a little change here - RNA will have uracil - U - instead of T where appropriate).

Unlike DNA, RNA is capable of exiting the nucleus of the cell and entering the cytoplasm. It goes to a tiny little spherical organelle called the ribosome. There, aided by a bunch of enzymes (which are proteins) and some other types of short chains of RNA, the genetic trasncript gets tranlated into protein. The order of three consecutive nucleotides (a triplet) has a chemical meaning: it is a code for a particular amino-acid. The order of triplets, thus, determines the order of amino-acids placed in the chain.


Once the whole RNA sequence is translated, the chain of amino-acids is further modified by other enzymes - they change its shape, add little molecules to it, etc. These modifications are key to the proper function of the protein. For instance, adding an ion of iron to the hemoglobin makes it possible for this molecule to transport oxygen to every cell in the body. Adding a phosphate group gives the protein extra energy. Adding a short chain of sugars assigns the protein its "zip-code", i.e., tells other proteins in the cell where to take this protein to, so they can shuttle it across the cell along microtubules, to its destination where it will perform its fuction.


Some of the proteins (called "transcription factors") have a specific role to go back into the nucleus, find particular genes (they use particular gene sequences to find and recognize them), and bind to them. The binding has an effect in either stimulating or inhibiting the transcription of that gene into RNA. Thus, the protein of that gene will or will not be synthetized in that particular cell.

Genes involved in the generation of circadian rhythms can be loosely classified into core clock genes and associated clock genes. The core clock genes are almost all transcription factors. Their proteins act by inhibiting or stimulating transcription of other core clock genes (as well as regulating expression of other - downstream - genes that serve as functional outputs of the cell, i.e., telling the body when to relase a hormone and when not, when to sleep, when to wake up, etc.).

If core clock genes were all there is, the circadian cycle would last only a couple of hours, at best. That is how long it takes for all the players to switch on and off each other once. In order to prolong the cycle to be closer to 24 hours, oter genes are associated with the clock. Their protein products act as modifiers - they may add or remove phosphate groups on core clock genes, inhibit or stimulate expression of some of the core clock genes, degrade the core clock proteins either spontaneusly or upon receiving a signal that the retinae have perceived light, etc.

Here is a schematic of the mammalian circadian clock. Genes called Period, Cryptochrome, Clock and Bmal (or MOP) are the core clock genes in the mammals:


It is similar in other organisms, with some changes, and you can also watch a great animation movie here.

How lithium affects the molecular clock?

A couple of years ago, it was proposed that the protein involved in the clock mechanism that is sensitive to lithium is not one of the core clock genes, but one of the accessory genes - namely Glycogen Synthase Kinase 3ß (GSK3), which, in turn, acts on Rev-Erb, which in turn acts on Bmal.

Now, a new paper came out with more evidence that this is so:
Nuclear Receptor Rev-erb{alpha} Is a Critical Lithium-Sensitive Component of the Circadian Clock by Lei Yin, Jing Wang, Peter S. Klein and Mitchell A. Lazar. You can find the press-release and excellent media commentary here, here, here, here, and here.

According to this paper, lithium inhibits GSK3. GSK3 normally protects Rev-Erb from destruction. Rev-Erb normally inhibits expression of the core-clock gene Bmal (and perhaps also Period). Thus, when lithium is present, there is no GSK3 to protect Rev-Erb from being broken down. Without Rev-Erb, Bmal and Period get expressed again.

Perhaps this all means that in the Bipolar Disorder the clock gets "stuck" in some way. Perhaps Rev-Erb accumulates and stops the clock from running. Lithium indirectly aids the distruction of Rev-Erb, thus allowing the circadian cycle to proceed.

As they say:

"These results point to Rev-erb as a lithium-sensitive component of the human clock and therefore a possible target for developing new circadian-disorder drugs. Some patients taking lithium have developed kidney toxicity and other problems. Lazar surmises that new treatments that lead to the destruction of Rev-erb would have the potential of providing another point of entry into the circadian pathway."