Most research papers on the circadian clock start out saying something like the clock is a transcriptional feedback loop and so on. Editorial warning here, I object to this simplification because the more we learn about how the molecular clock is constructed, the more we see how the transcriptional feedback loop is embedded into the cellular, molecular workings. And we see that there's much more to the clock beyond the loop. That said, the components of the transcriptional loop, period, cryptochrome, timeless, female one and clock and animals, freak and the white colors in fungi, LHS, CCA1, and the family of PPRs in plants. These proteins have a tremendous effect on many clock properties, hence statements like this is the clock. So if you'd like to find genes regulated by the clock, the transcriptional feedback loop is a pretty good starting place. The first clock controlled genes called CCGs were identified in Neurospora. The hypothesis behind the experiment went like this. Clock controlled genes, for instance, such as those that must be regulating clock controlled conidia formation in Neurospora, should be expressed at different times of day. The expression patterns of the first clock tunes themselves period and frequency in Neurospora, show that regulated RNA expression was a method that the clock used. So it stood to reason that other genes might also be expressed at higher or lower levels at opposite times of day. This might sound trivial since, in the last lecture, I told you that the clock mechanism involves transcription factors. But it was many years until this was broadly appreciated, namely that everything we think of as clock genes is regulating transcription in some way. So the hypothesis is that clock controlled genes will be abundant at one time of day and scarce at another. From there, the method is conceptually simple. RNA is collected at opposite times of day and then these two samples are subtracted from one another. This is technically possible because a DNA copy of the RNA can be made artificially in the lab. The DNA copy will be complementary to all of the RNAs in the sample that was used as a template. There will be more complementary DNAs corresponding to those genes that generated more RNA. Since RNA and DNA hybridize readily to their complementary sequences, the DNA and RNA from opposite times of day can be mixed together so that RNAs that are expressed at similar levels are made into double stranded hybrids of RNA and DNA. While species that are expressed at high levels at one time of day, but low levels at another time of day can be isolated just by separating out the single and double stranded RNAs. At that point it's a relatively simple matter of identifying what's left over, what was found in excess, and then amplify it via standard cloning methods. Once you have candidate genes, this brings up a philosophical issue. What are the rules for calling something a clock controlled gene? What standards would you use to say with certainty that this is a gene whose expression is clock controlled? In general, the idea is something that it's not a clock gene itself. That is it doesn't change the basic properties of circadian timing but it is expressed rhythmically with the same period as the circadian clock. This idea was used to challenge and to confirm the first new CCGs in Neurospora. First the timing of expression in normal wild type strains was monitored and shown to have the same free running period as conidia formation. Second, the same experiment was done, but with a clock mutant that had a long free running period. A clock controlled gene should oscillate with the same longer period in the mutant strain. These first two CCGs did so. Finally, the CCGs were knocked out of the genome. The genes were mutated so they no longer functioned. And then the rhythms and clock genes were followed to see if the gene and if the CCG, had an effect on the Circadian clock. In this case, there was no feedback onto the clock, so this was a clear demonstration of the concept of a clock controlled gene. The first clock controlled gene was identified in mammals with a very different hypothesis. Namely that the clock associated transcriptional activators might be regulating output gene expression directly. The neurotransmitter vasopressin was targeted for this experiment. It's expressed with a high amplitude rhythm in the suprachiasmatic nucleus in mice. And in arhythmic clock mutants the gene expression is flat and at a low level. Inspection of the promoter region revealed that it contains an e box, the preferred binding sequence of the clock female one heterodimer. A series of in vitro experiments were performed where various combinations of female one clock and the vasopressin promoter fused to a reporter gene were transformed into a cell. Only when all three are present is the reporter gene expressed. If the e box is mutated then the reporter is not expressed, pinpointing the e box as essential for this clock controlled gene expression pattern. A break through in the identification of clock controlled genes came with the advent of technology, namely DNA microarrays that could measure the transcriptome, the entire population of messenger RNAs produced in a cell or tissue at a given moment in time. That meant that rather than searching for one or two clock-regulated genes, one could search for nearly all clock-regulated genes. A number of studies jumped on this technology in the early 2000s and they came to similar conclusions, at least for eukaryotic organisms. In mice, tissues like the SCN, liver, heart and kidney were used to harvest RNA over one or two days and compared the entire compliment of messenger RNA produced, asking which genes' expression were found to change overtime. Very generally speaking, about 10% of the genome in each tissue was rhythmically expressed. Were these same genes in all tissues? Only in part, the same clock genes were found oscillating in all tissues. But the genes that are presumably clock controlled, rather than the clock genes involved in the clock mechanism itself, they tend to be tissue specific. In the liver, this means enzymes and proteins involved in oxidative phosphorylation and energy metabolism. In the SCN, oscillating genes included these involved in synapses. The implication here is that part of normal development specification is to organize and specify clock controlled gene expression. Different outputs in different cells, this strongly suggests a role for Epigenetic mechanisms whereby local regions of the DNA are modified so that the gene is more or less available for transcription. Epigenetic mechanisms can effectively signal if a given set of transcriptional factors such as clock and BML1 can have access to a gene's promoter region. Epigenetic regulation in the form of differential post translation modification of histones, acetylation and methylation has been described as oscillating over 24 hours at clock regulated regions. However, the implication of the finding I described concerning CCG's is that a stable modification occurs in development that specifies which set of clock controlled genes will be expressed in a given cell type. These early findings were updated recently with adding even more tissues. Now, 12 tissues in a single study. In this work, the same themes emerge but are clarified. Some tissues have more rhythmic transcripts and some have fewer. Actually, some as low as 3% to 4%, but the average is still around 10%. The phases of clock regulated genes can vary a lot between tissues and there can be many different phases of clock regulated gene expression within a single tissue. Per two and BMAL1 are expressed at opposite times of day, we've know this for quite some time, with per two activated by clock and BMAL1 while BMAL1 expression is coordinated by REV-ERBa and RORa. Clock controlled genes could be regulated by either of these transcription factors, or by D-Box Binding transcription factors, or by still other, yet to be identified transcription factors that are themselves clock-regulated. The Eskin-o-gram could be modified to show many possible models for clock controlled gene regulation. The mechanisms that can lead to rhythmic RNA expression were further explained in several ways. DNA bound by the clock protein produces a burst of transcription, essentially at one time of day, despite that the clock protein regulated genes have RNAs that can oscillate at different times of day. This implies posttrancriptional mechanisms are used to yield rhythms in RNA expression. A non-transcriptional mechanism used to regulate RNA levels is implemented with complimentary or antisense RNA. When RNA forms a double stranded complex in the cell, it's a signal for degradation or translational silencing. These antisense, non-coding RNAs are produced by many organisms and are now recognized as a major mechanism for cells to fine tune their RNA levels. The circadian clock apparently also has non-coding RNAs in its tool kit. Experiments that sequenced all RNAs identified many rhythmically produced Non-coding RNAs. Alternatively, non-coding RNAs were detected that are complimentary for rhythmically produced messenger RNAs. Taken together, these new mechanisms will be important for fine-tuning the timing of CCG expression. I mentioned that in eukaryotes, the number of clock-controlled RNAs in various model organisms is similar, coming in at around 10%. This holds for Drosophila, plants, and fungi, as well as the various mouse tissues. A contrast to these observations is the prokaryotic model system cyanobacteria. In this photosynthetic bacteria, practically all genes show circadian rhythms in their expression pattern. This has been linked to a phenomenon of the chromosomal DNA opening up and closing on a massive scale according to time of day, meaning that genes can or cannot be expressed according to the state of the chromosome. Coordination of this phenomenon with several specific transcription factors allows for expression at several times of day, although the end of the subjective night is when the majority of genes are expressed. The state of chromatin is more and more recognized as regulating clock-controlled gene expression even in eukaryotes. It's just that it only occurs on selected genes and the mechanism of making the DNA accessible is somewhat different from what we heard about in prokaryotes. An important caveat when discussing clock-regulated RNAs is to note that in most cases we want to know if a rhythmic protein is made. The RNA is usually just the way to get to the active protein. When scientists have looked at the entire collection of rhythmic proteins, the circadian proteion, they find about 10% rhythmic proteins like with the RNA transcriptone. Interestingly, only some of these proteins come from rhythmically expressed RNAs. Some rhythmic RNAs lead to proteins expressed at constitutive levels. This can happen when the protein is so stable that a small addition, once a day, doesn't make enough of a difference to be detected. This is the case for albumen. In other cases, rhythmic proteins are generated from non-rhythmic RNAs. This can occur due to regulation of the translational machinery, which does occur to some degree or for regulation of the protein's stability. If a protein degrades more rapidly at one time of day relative to another, and if the production of the protein occurs at a stable rate, then the amount of protein will oscillate.
