Many methods have been used over the years to find out what molecules are working together with PERIOD in <i>Drosophila</i>, and in another model system mice, and with <i>frq</i> in <i>Neurospora</i> as well as the feedback loop in plants. Mutant screens are still going on. Biochemistry approaches have been used. Bioinformatics has been applied, and there have even been discoveries with spontaneous mutants. Also, our concepts of transcriptional regulation have been evolving. Transcriptional regulators are usually huge complexes of proteins, many contributing distinct functions. Transcription factors are chemically modified in order to function. They're phosphorylated or tagged with phosphate groups, giving them a bulky charged side group. And it's increasingly recognized that part of regulating transcription is modifying the region around the promoter called the chromatin. This can result in physically bending or opening up the area so that transcription can occur or to stop it from occurring. Now with perfect hindsight we would predict that all of these molecular functions would be part of the molecular mechanism of the clock. First, lets fill in the feedback loop in animals. That would be flies and mice, which are the clock model systems that are used for genetics research. The first important point to note, is that many of the clock genes in flies have homologues or close relatives in mice. Mice have not only won but three period genes. All making PERIOD proteins and all expressed a bit differently. In flies, the TIMELESS protein is a corepressor with PERIOD. In mice, CRYPTOCHROME is the PERIOD corepressor. In addition, in mice there are two cryptochrome genes, and they're also expressed somewhat distinctly, allowing for fine tuning of their activity. This whole collection of repressors, <i>Drosophila</i> period and timeless, mouse period and crpytochrome are expressed rhythmically. The proteins that activate these genes are CLOCK and CYCLE in flies, and CLOCK and BMAL1 in mice. These proteins are obvious transcriptional activators in that they have DNA binding domains of the bHLH type. They bind to E-box sequences which are palindromic DNA sequences in many promoters that read CACGTG. Interestingly they also have PAS domains like the PERIOD protein. There's some evidence that the CLOCK protein in mice can be replaced by other proteins, at least in some cells, suggesting complexity on the mouse system also on the side of activators. In mice, CLOCK is not rhythmically expressed, while BMAL1 is. In flies, it's the other way around, with CLOCK expressed with a circadian rhythm and CYCLE, ironically not. The feedback loop in the fungus <i>Neurospora</i> is similarly constructed, in that you could superimpose the molecular structures on one another. But the fungal clock proteins, that are working like the <i>Drosophila</i> ones, are made up of unique sequences relative to the animal clock proteins. White collar 1 (WC1) and White collar 2 (WC2) are activators, and FRQ, and FRQ RNA helicase are the negative feedback elements. Plants have a highly complex feedback loop, again, with a distinct set of clock proteins relative to animals and fungi. The model plant for clocks genetics is <i>Arabidopsis</i>, and as we saw in mice, there can be multiple copies of the same gene, expressed at slightly different times. Recently the tiny unicellular alga <i>Ostreococcus</i> was introduced as a novel model for clock's research in plants. It has homologues in most of the plant clock genes except in single copies. This may simplify research on the green circadian clock mechanism. So far we've sketched out either relatively simple or more complex feedback loops that all use a similar formula. Same loop construction in plants, animals, and fungi but with different proteins. In this basic model the oscillations in the genes <i>frq</i> or <i>period</i> could occur without any other gene oscillating in amount. All they need to do is suppress their activators for a while, and then the protein is degraded to allow more of the RNA to be expressed, but this is not exactly what's observed. Very often, at least one of the activators also oscillates in amount. This is mediated by an additional feedback loop. In mice, the nuclear orphan receptor <i>rev erb alpha</i> was shown to act as a repressor of <i>bmal1</i> transcription. Another nuclear receptor <i>ror a</i> activates <i>bmal1</i> transcription. They bind specific sequences in the promoter of <i>bmal1</i> and BMAL1 binds E-boxes in their promoters leading to an auxiliary feedback loop. Thus a transcriptional feedback loop was folded into the one that was elaborated using the <i>period</i> gene. By now the molecular models for plants and fungi as well as for <i>Drosophila</i> also have an extra transcriptional feedback loops. [NOISE]
