Multi-Cellular Scripts

In this book so far we have used massive number of 'woulda' and 'coulda' qualifiers. That is pretty much the nature of any discussion of conditions 3 or 4 billion years ago, since you can't just fill a vat with organic compounds and watch it evolve. In a decade or two, further genetic research will probably allow us to fill in some of the blanks in the evolutionary story we have described so far. But even the simplest of cells are complex enough that the full story of their creation and daily functions is going to be an enormously large story, with huge amounts of detail. As we move on to multi-cellular organisms, the amount of supposition needs to increase still further. It takes mind-numbingly large quantities of information to specify us multi-cellular organisms-- so much detail that it's impossible to grasp it all at once. We humans contain about 3.3 billion base pairs in our DNA sequence, which is about 825 megabytes of information. Your camera probably stores more information than that, but on the other hand, your camera doesn't have that information arranged cleverly enough to self reproduce, let alone walk down the street to buy a frappacino and decide what pictures to take next. Figuring out the way those few billion nucleotides work is more complicated than solving a ten-thousand by ten-thousand Sudoku puzzle, and it will take swarms of geneticists scores of years to completely break the code, if it ever happens at all. So far in this story, we've only looked at some general and very simple aspects of the formation of our genetic code, and that is as far as we're going to get with understanding them right now. So it's time to shift away from the small DNA details entirely, and take a look at larger structures. Cell Differentiation The next evolutionary step that we'll examine is the complex cell differentiation found in plants, animals and fungi. At least from the point of view of large Eukaryotes, these multi-cellular organisms are the final step in our evolutionary history of DNA. Bacteria and other prokaryotes are capable of grouping into chains and mats with other bacterial cells, so it is likely that the early eukaryotes were also capable of similar clumping. However, the first step in creating true multi-cellular organisms would be to form different types of cells, with different functions, and then link them together. Volvox, a colonial flagellated amoeba, is about the simplest example of that. It contains a sphere of identical body cells, and then several reproductive cells in the center which eventually form into new Volvox colonies. This simple cell differentiation seems to use a simple protein repressor to shut off reproductive activity in the body cells. The next step up on the level of complexity are the Porifera or sponges, which contain many different cell types, but no organs or larger structures. This phylum is also the simplest to contain a homeobox gene . Homeobox Genes Homeobox genes were first discovered in 1983. They are developmental genes which create 'homeodomain' proteins, which play an important role in cell and tissue differentiation, although the process is still poorly understood. Mutations in homeobox genes tend to produce gross physical abnormalities in multicellular organisms. For example, a mutation in the Drosophila bithorax and postbithorax genes forms an additional set of wings instead of a haltere , and disruptions in the human Hoxd13 gene cause polydactyly (extra fingers). Porifera (sponges), one of the simplest phyla, seem to contains just one or a small number of homeobox genes . More complex organisms contain more homebox genes (humans appear to have at least hundreds of homeobox genes, though they were only discovered in 1983, and accurate counts are still not available). Remarkably, there are many similarities between homeobox genes in species as diverse as Drosophila (fruit flies) and humans, so they must have appeared extremely early in our evolution, and then remained relatively constant for hundreds of millions of years. Information Content Before we consider how homeobox genes might work, it will be useful to look briefly at the 'information content' of a multi-cellular organism such as we humans, and match it to the information content found in our DNA. Current estimates for the number of protein-coding genes in the human genome currently range from about 20,000 to 66,000, with a consensus of approximately 30,000 genes . The roundworm Caenorhabditis elegans contains about 19,000 genes to specify its 959 somatic cells, so that seems like a reasonable 'baseline' amount of genetic material to specify basic metabolism and cell differentiation for a simple multicellular organism. Are an extra 11,000 genes enough to specify the extra organs and structures in a human? Organs and Tissues Humans contain a few dozen major metabolic, secretory and sensory organs, each with a few dozen different types of cells within them. Each of those cells might contain a few dozen specialized proteins or cell structures that do something interesting that a protozoa wouldn't do. Translate that into some extremely ballpark math, that might mean 50 organs times 20 cell types, or 1,000 different cell types that would each need some kind of specs. With 20 or 30 extra details to consider within each cell, there are already too few genes to match up with the information required. However, there is a bigger chunk of data that is necessary to specify a functional human, and that is the physical layout of those organs and tissues. The size and placement of each tissue is important. The arrangement of different cell types within each tissue is also vitally important. Besides all that, humans contain approximately 100 billion neurons, which are arranged into clever combinations that allow us to have a behavioral repertoire with many thousands of different actions. Our hearts beat, our intestines churn, our eyes blink automatically when something gets too close, and our salivary glands kick into gear when we smell Pad Thai or see golden arches. Each of those actions requires coordinated action from dozens of sensory, processing and motor neurons, and it seems highly implausible that such complex connections could be created from just a gene or two. There is need for more information. RAM Another way to look at the same problem is to consider the human genome as a 'program', complete with megabytes. Each of our 3.3 billion base pairs contains two bits of information, so each human is defined by about 825 megabytes of information. That is probably less data than what's in your camera . All of our proteins are specified by about 12 megabytes of DNA, but that's just the molecular coding for the raw materials in our bodies. A typical software program contains both executable code, and enormous amounts of data to specify dialogs, error messages and other details of the interface. Our DNA probably makes a similar split, with some sort of data encoding for the other details. How might that occur? Multi-Cellular Scripts Since there are approximately 2 billion base pairs in the repetitive portions of the human genome, they seem a likely candidate for the extra data needed to specify positions, placement and sizes for the various organs and tissues in our bodies. We've already talked about the advantages of script-based control of cellular structures via the Foxy genes and proteins, and it seems reasonable that a similar mechanism would have evolved for controlling multi-cellular tissues development. So let's imagine a new type of gene, similar to Foxy, but involved in multi-cellular development instead of the placement of organelles within cells. We'll call it a Moxy gene, which codes for a Moxy protein . Moxy might work along with homeobox genes, or it might be exactly the same thing. Let's look at some possibilities for Moxy, and how it would differ from Foxy. Moxy Structure Most likely a Moxy would have a Fred-derived structure that is very similar to Foxy-- with an elbow reading a genetic chain, and a knee that would change its conformation and do something. The biggest difference is that Moxy would probably read from DNA directly, instead of looking at a portable RNA script chopped out of a mRNA molecule. The reason? Moxy needs to act over several generations of cells, so it might take days or years to get through a script. RNA is the 'throw-away' molecule within cells, which is fine for a Foxy that is whipping out a cilia over the course of a few minutes or hours, but not good enough for the time scale that Moxy needs. The only molecule in the cell that stays constant over the lifetime of a cell is its DNA. The other difference is that the expression of the Moxy gene would affect multiple cells instead of just one. That means that Moxy has to work a little differently from Foxy, because differentiating cells are a moving target. Moxy Script Reading Since Moxy can't just carry around a strip of RNA like Foxy does, how would it read a script, in a way that could transcend cell divisions? There are five possible ways Moxy might manage to extend its script reading over a long period of time. Most likely, cells use some combination of these methods, depending on which suits the need of a particular Moxy system. Let's look at them now. Sticky Moxy In the 'sticky' version of Moxy, the protein would bind to the DNA chain more or less permanently, and read one element of a script at each cell division. For example, let's look step by step at how a Moxy protein might allow precise digital control of the color of each segment in a dividing feather cell or skin cell-- which would allow digital control over the individual's color pattern. In this case we'll use a highly simplified two-color model that would be very similar to our original Fred (and probably much simpler than what is actually used to control coloration in real feathers). A sticky Moxy coding feather colors might look something like this: 1. In the first generation cell, Moxy matches the proper ID sequence for some repetitive DNA, and changes its conformation based on the first base pair. This change triggers synthesis of a pigment in that cell. Moxy hangs on tight to the DNA chain. 2. At the next cell division, Moxy moves down the chain by one molecule. When it's time to produce pigment, Moxy reads the next available script nucleotide, and creates a cell with a different pigment. 3. Several generations of cells would go by, each with a color determined by the color data stored in the satellite DNA sequence. 4. Eventually we have a complete feather pattern. For a 'sticky' Moxy to work, it would need to remain attached to the same spot in the DNA strand through many cycles of mitosis. That means it would need to shift out of the way temporarily during DNA replication, and then return to the same spot. It would also need to 'jiggle' one base pair at each step-- which could happen during cell replication, or some other time. Finally, if Moxy were to affect all the daughter cells, it would need to be duplicated during each cell division, so each of the two new cells could also have a Moxy at the same location. That means the cell would need to create additional Moxy proteins just before each cell division, and add them to the right place just after each division. Taken together, it sounds like creating a 'sticky' Moxy is a daunting task that seems rather unlikely to evolve. Shrinking Moxy Perhaps a better way to implement a Moxy-based script would be to permanently remove one or more DNA base pairs from a script, shortly before or after each cell division. Moxy could then read the current base pair as part of the expression of that new cell's development. That way the gene itself would keep track of each cell generation, and there would be no need for the Moxy protein itself to stick around for long periods of time. Here's how Moxy would work if it removed base pairs: 1. In the first generation cell, Moxy matches the proper ID sequence for some repetitive DNA, and changes its conformation based on the first base pair. This change triggers synthesis of a pigment in that cell. When Moxy leaves, it removes the first base pair. 2. In the second generation cell, Moxy matches the same gene ID sequence, grabs the next available script nucleotide, and creates a cell with a different pigment. 3. Several generations of cells would go by, each with a color determined by the color data stored in the satellite DNA sequence. 4. Eventually we have a complete feather pattern. Meanwhile, Moxy has removed the last satellite base pair from the chromosome. The chain is completely gone. The shrinking version of Moxy has many advantages: it doesn't interfere with DNA replication, it doesn't require any molecules to stick around for very long, and it can advance along the script at any pace it likes (there is no need to do it just during cell divisions). The biggest problem with the shrinking system is that once a script is read, it's permanently gone. Some other gene will not be able to use the same script later. This creates some built-in aging, unless the cells can be somehow replaced with new cells that have the original sequence. That may not be such a big problem, as long as there is a distinction between body cells (which may age and lose script data) and germ cells or stem cells (which keep the original copy of the genome). There is evidence that some DNA is lost, as cells age. So 'shrinking' Moxy appears rather plausible. Offsetting Moxy A third possibility for Moxy control is to use an external counting device, and use it to find a new script value during a during each cell division. It might work something like this: 1. In the first generation cell, Moxy matches the proper ID sequence for some repetitive DNA, and changes its conformation based on the first base pair. This change triggers synthesis of a pigment in that cell. 2. At the next cell division, Moxy checks an 'age counter' located somewhere else on the chromosome. The counter knows this is the second cell division, so Moxy offsets by one base pair, and reads the next RNA base pair. That produces a different color in the next feather barbule. 3. Several generations of cells go by. Moxy shifts one additional base pair for each generation. 4. Eventually we have a complete feather pattern. For an 'offset' Moxy to work, it would need some sort of counter to read somewhere else to know the number of cell divisions which have occurred. One possible source for that info are the telomeres at the ends of each chromosome-- they reduce in length by several base pairs after each cycle of mitosis. Marker Moxy A variation of the 'offsetting' Moxy would be a similar system that used some sort of marker on the DNA strand itself. That might be a methylated nucleotide, a temporary change in the nucleotide sequence, or the placement of the DNA strand on its nucleosome. With a marker, Moxy would find the script by its ID as usual, and then 'walk' down the script until it hit the marked base pair. The only real challenge for the marker system would be to preserve the mark during cell replication, and pass the marker along to both of the daughter cells. Master Protein Moxy A final way that Moxy might preserve a script through many cell generations is to act just like a Foxy, and carry an RNA chain to use as a script. In that case the Moxy protein would need to resist any auto-digesting enzymes within the cell that might try to recycle the protein or its chain back to their components. It would also need to have a way to pass along to daughter cells (unless the script only needed to be in one of the daughters). Cells could manage that for a few cell generations if they made many copies of Moxy in the first cell, but eventually some of the daughters would end up without a copy of the Moxy and its script. Moxy Programming By the time Moxy came on the scene, there had already been many millions of years of single-cell eukaryotes. That means that Moxy scripts would have fit into an already complex system of gene regulation. Any Moxy protein would probably have used all sorts of regulatory proteins, hormones, transmitters and other metabolic tools to actually manage whatever structure it specified. Moxy and Foxy The biggest job in specifying a multi-cellular organism is deciding which type of cell to have where. That probably means that many Moxy genes would have acted by controlling a Foxy gene, which would the control the setup of specific organelles or other structures within the cell. Once a cell was formed, it could follow its own internal script sequence (mostly implemented by Foxy genes) to determine its own structures, whether they would change with time, whether the cell would multiply, and even when the cell would die. All it would really take is a 'library' of different cell types for the cell to choose from, and a Moxy telling it which to use. Moxy and Moxys The first multi-cellular organisms probably used just a single level of Moxy controls to guide some simple differentiation of cells during embryonic growth. However, as plants and animals evolved into greater complexity, the 'programming language' of Moxy would have also grown, just as it did with Foxy genes. One Moxy protein could easily regulate another Moxy, which might regulate a descending cascade of other Moxys and Foxys that might be dozens or hundreds of layers deep. In a way, Moxy provides a 'programming language' for multi-cellular development, and it probably included similar subroutines and master routines. What this probably means for human evolution is that some of the early, basic Moxy homeodomains will be extremely well conserved, with scripts that are highly resistant to change. Smaller Moxy functions would be much less conserved, and changing them would have a smaller effect that would be less 'risky' in an evolutionary sense. Moxy Tricks Since Moxy appeared in relatively sophisticated organisms, it would have had a good repertoire of biochemicals available, which would have allowed it to develop many clever variations. Some of these features might have already appeared in Foxy genes (which could then have easily been converted to Moxy format). Timed Moxys As with Foxy, some Moxys might read a script with the help of a timer, so they could take certain actions at certain regular times regardless of outside conditions. That would be an effective way to 'guide' the growing axon of a neuron cell along a specific path, or to change the behavior of a homeodomain during different periods of embryonic growth. A timed Moxy would be almost identical to a timed Foxy protein, but it would probably act more slowly, and affect multi-cellular features instead of the contents of a single cell. Cell ID Since every cell contains the entire set of DNA, it also contains all of the genetic information that it would need to become any type of cell. That means that it would be possible to have a 'master Foxy' gene which could turn that cell into any possible cell type, just by reading a 'master script' that specified exactly what the cell should contain. It would also be possible to set up a library of scripts for every possible cell type, and give each of them an ID sequence that would locate that particular bit of satellite DNA. The net result would be very interesting, from a design point of view-- a Moxy could put any cell type at any location, simply by passing the right ID sequence to the master Foxy protein. Reading Frames The example we just used had a single-nucleotide reading frame, and coded for just two values just like the original Fred. Of course by the time Moxy came on the scene, Foxy had already been around for millions of years, and had probably already evolved the use of different reading frames for different tasks. Consequently, it is likely that Moxy would have soon used the same variations to suit different specifying tasks for tissues and structures. Moxy could read single nucleotides and make 4 choices via conformational changes. It could also read gene ID sequences (and then trigger different genes at different times). Or it might have a more sophisticated system, using two or more nucleotides at a time. Size and Length Specification A Moxy specifying the shape of an insect mouth part might read chunks of a script and pass them along to daughter cells as a way to precisely and digitally control its 2D or 3D shape. Here's how that might work, step by step: 1. A Moxy in the first parent cell matches with a match sequence (not shown). 2. The Moxy reads along the sequence, cuts it at a stop sequence, splices the chain and carries away the first fragment . 3. That short piece of script controls the number of cell divisions at right angles to the structure. 4. The next cell division grabs the next available chunk of script . 5. It also controls the number of cell divisions sideways. 6. The process continues, and a structure is formed with precise digital control of its shape. It might be an insect mandible, or fern leaves, or any other structure where a comb-like shape would be useful to the organism. The same sort of shape-specifying could also take place in a non-destructive way. In that case, Moxy would make a RNA copy of each portion of the sequence, and then place a marker so it could return to the correct place in the script at the next cell division. Moxy Evolution Moxy was huge. By putting multi-cellular expression under easily changed genetic control, it greatly increased the chances of beneficial mutations occurring, and greatly sped up the pace of multi-cellular evolution. The fastest changes would happen in simple scripts controlling some simple property such as the shape of a leaf, a wing, a mouth part or a neuron path. Such changes would rarely be lethal, and occasionally they would produce a positive change that would push that organism on bit of the way along its evolutionary path. Simple scripts could control the personality and appearance quirks of an organisms with a fairly low risk of lethality, and with subtle but effective impact on survival value. Larger, but slower, changes would happen when mutations occurred in older scripts controlling other scripts. They might result in a size change in an organ, a change in developmental timing, or something else major. Such changes would more often be lethal, but they might also lead occasionally to a major breakthrough on the species level. The largest changes would occur when an entirely new system of expression evolved, complete with a new set of scripts and new Foxy, Moxy and Hox genes to use them for a new function. That kind of change would happen rarely, and it would frequently be lethal. However, an occasional beneficial script would result in a major innovation that might emerge as a new genera or a new class. Previous Home Next