I am currently reading the book Genome by Matt Ridley, and I’ve found it to be quite interesting. However, it raised a question in my mind that I’ve had since my high-school biology days. Back then, I wondered how a nerve cell knew it was supposed to be a nerve cell if all the genes are exactly the same as the white blood cell whizzing by it in a nearby blood vessel or the bone cell in the skull.
It seems to me that all of this so-called “junk DNA” might contain information about the blueprint of the species in question, so that in humans it might only be used from fertilization until the shape of the fetus is fully formed.
I’ve also wondered how a cell knows how to work; that is, if it’s producing proteins from transcribed DNA, how does it know how many to produce? What’s to prevent it from making too much or too little of a protein if the the RNA is just floating around the cell and hitting ribosomes randomly? (Or, if the RNA is destroyed after one use, what’s to prevent a particular section of DNA from being transcribed too many times?) It’s been awhile since I’ve ready biology texts, so I guess I could be a little rusty…
I realize some of these questions may not be answered yet, but I’m interested in hearing the existing theories.
basically, what determines what a cell will be doing is what the cells around it are doing. one example that’s popping to mind is something called ‘sonichedgehog’. i can’t remember if that’s the name of the gene or the protein. it has a lot to do with orientation of the body. how do the cells of your big toe ‘know’ to grow on the inside of your foot? what makes your liver grow on your right side?
to see (very rudimentarily- not because i can’t dumb it down enough for you, but because i know little or nothing about cellular biology and i vaugely rememberthis from scientific american or somesuch) how they know this, consider your thumb. with your right hand palm down, you can make a three dimensional grid. when your hand was first developing, lets say that the cells to the right (pinky side) grew first. they secrete some of this protein. the further away from those cells you get, the lower the concentration of sonic hedgehog. so the cells determine their position based on concentrations of proteins. i believe that there is one version for the x axis, one for the y, and one for the z.
now each cell is secreting a great number of proteins. some code for structure, some for position, others for cell type, etc.
if you want an anthropomorphic example, consider cheerleaders forming words on the ‘grass’ at superbowl halftime. each cheerleader only knows where he or she is supposed to be based on relation to other cheerleaders. bonnie knows that fred should be in back of her, and mindy and cindy (the twins) should flank her. if the cheerleaders were to be blindfolded, they could still make their letters with some signals other than visual. they could tell eachother where they were, or whateva. the main thing is that you could tell them who they had to stand near and where (next to, behind) and they could form words without knowing at all the larger structure (in this case, maybe ‘too many roman numerals this year’).
was that to muddled? if so, ask somebody else, cuz my brain is starting to hurt. damned insomnia.
Junk DNA is still a mystery, but it doesn’t have as important a function as the one you have described. Labs like Harvard, Baylor and Penn State have been working on “knock out” (I’ll call it KO) experiments with mice. they check for a specific area of DNA, KO it in a litter of mice to be and see what happens. The reproductive cycle of the mice is so short they can work through cause and effect pretty quickly. I think they intend to match this up to the human genome later one. Anyway, junk DNA is not essential to life and while it may have some influence on health and disease and other things, it is not the program director.
Is just a hypothesis, or is it scientifically verified theory? How does the cell interpret these proteins? Are there some groups of regulating enzymes that monitor these “neighbor proteins” and direct protein manufacture based on this? It seems there would have to be some instructions in the DNA somewhere to control the interpretation… How does the cell know which part of the DNA to read under those specific circumstances?
Actually I remembered most of that, but it brought up more questions… How do ribosomes interpret RNA and actually manufacture the proteins? Do they have some sort of interpretative codebook for the amino acids? What role does the endoplasmic reticulum play in this? Are there any good, modern cellular biology books that I should just read?
Understanding ontogeny and the process of cell differentiation is still very much a work in progress. To put it very briefly, there are huge numbers of regulatory pathways and proteins in the human body. A typical example would involve a protein binding to a regulatory sequence of a gene, turning it on or off. Many of the regulatory sequences are areas of DNA that were considered “junk” before their functions were known, because they aren’t in the genes themselves. I suspect as we learn more and more, the amount of junk DNA will shrink, though I’m sure some of it is actually junk. Anyway, the known pathways number in the thousands, so I won’t even try to start. As for how it all starts - how you get so many different types out of one cell - that’s not well understood, AFAIK. For a rather philosophical view of the process, read At Home in the Universe by Stuart Kaufmann. There’s a chapter in there on ontogeny. Not the most detailed explanation around, but it’s the only one I can think of off the top of my head.
Now for some of the other questions:
Yes, they do. Every three bases (a codon) in the RNA corresponds directly to a specific amino acid. This code was worked out back in the 50s or thereabouts. Simply, the ribosome binds to the RNA and scans along until it hits a start codon. tRNAs bring amino acids to the ribosome. Each tRNA is specific for both a codon and an amino acid. Eventually, the ribosome hits a stop codon and falls off.
None, really. I know you’re thinking of rough ER, which has ribosomes associated with it. ER is used to export secreted proteins out of the cell. If a protein needs to be secreted, the front end consists of a signal peptide. What happens is the ribosome begins translating the RNA into protein, making the signal peptide. Then it gets stuck, so you have the signal sticking out of the ribosome, with the rest of the RNA waiting to be translated. The signal is attracted to a protein gate on the ER. It drags the ribosome and RNA along to the ER. Once at the gate, the protein sticks through like a thread through a needle, the ribosome’s block is released, and the rest of the protein is made, with it getting shoved through into the ER lumen. Often, the signal peptide is removed at this point. That explanation wasn’t as lucid as I hoped. I hope it makes sense.
Yes, there are. I don’t know of any titles offhand, though.
OK, since I work on fruit flies, here is a simple example :
How does the embryo know its head from its ass? (briefly)
The egg develops in a group of cells called polar cells, which are connected by the future anterior (head end) to the developing egg. These pump mRNAs of a gene called bicoid into the embryo. These filter back to form a gradient, with the peak being at the head and the least being at the tail. This is done by the mRNA binding a protein in the egg. Most binds when it first enters, little makes it down to the ass.
Next, fertilization and all that goes with it. The fruit fly embryo basically partitions into cells around a central cavity. Genes in the embryo come on. bicoid and genes of that class (including genes localized to the ass entirely) induce or repress their downstream targets. For instance, bicoid turns on hunchback, which forms a sharper peak in the embryo. hunchback and bicoid in turn turn on a group of genes called “gap genes.” These are expressed in large broad stripes anterior to posterior. These in turn induce/repress another group which form smaller stripes called “pair-rule” genes, and these go to “segment polarity.” The group of segment polarity genes (coupled with the other guys) can tell a cell precisely where it is on the anterior-posterior axis of the fruit fly. More importantly, these gene products all bind DNA and induce or repress downstream targets turning on genes required for forming a specific segment (tell the heart to be the heart, the gut to be the gut).
The Nobel prize went for this (at least partially). Nusslein-Volhard and Wieschaus.
Most of this is oversimplified, but the effect of so called “morphogens” (genes expressed in a gradient over a body area like bicoid) are common in early embryogenesis. This is not to say all development works like this. The earlier example of Shh (Sonic hedgehog, in the family of genes with Indian hedgehog, which are vertebrate homologs of the Drosophila hedgehog gene also isolated by Nusslein-Volhard and Wieschaus) in limb patterning works exactly this way.
I am still in school so this is stuff I live every day.
RNA->protein is a process highly regulated by cells.
Promoter activity (often responsive to gene product) turns on and off at the drop of the hat. It can go from cranking out buckets to shutting off completely really quickly.
mRNA is really labile in the cell. The cell has many enzymes and pathways to ensure this. I’d bet most mRNA doesn’t make it to the ribosomes. Valuable mRNA again gets targetted and coated in proteins sometimes to preserve it (like some of the mRNAs in the Drosophila embryo.
Once the mRNA gets to the ribosomes, many, many ribosomes read the same message (this is called a polyribosome).
Protein again is easily degraded, and the amino acids recycled.
One thing to remember is our cells live in an abundance of energy. If they make too much of one thing, they just degrade it. They have no qualms about doing so.
Next, junk DNA consists of repetitive elements such as telomeric (end of chromosome) and centromeric (middle of chromosome) repeats (sometimes dinucleotide, sometimes bigger) which can go on for millions of bases. Other bits of junk DNA include endogenous replication-defective retrovirus bits on the order of 100,000s/genome. It is not that interesting, trust me. People try hard to avoid it, even though bits may be very important (telomeres in aging, centromeres in chromosome stability, repetitive elements in packaging, few expressed genes in repetition or endogenous retroviruses, etc.).
Don’t be bashful, Edwino, give up this school stuff and hang out on the board answering questions.
I try hard to keep all of the students on the board busy answering questions, some of them take time off to drink, study or date, but if you hang out with fruit flies you probably are very entertained right there in lab.
Yes, yes, I know about the code, but is the specific chemical translation process known? How does the ribosome actually do the decoding? Where does it get “CAG” = specific amino acid? Seems to me this is a pretty important question.
edwino:
OK, if I understand correctly this bicoid RNA is distributed at different concentrations by the parent’s polar cells when the new organism is dividing for the first few times (blastula formation?), and when the ball of cells is large enough, they begin to interpret the DNA differently, leading to cell differentiation. Is it understood precisely how the DNA is bound by the mRNA and how it knows which specific genes to turn on or off? What controls the unzipping of the correct DNA strand and the correct piece to translate? Is there any sort of indentity so the enzymes can say, “Yep that’s chromosome 18, I need code group 23,243…”?
If I were to use a computer analogy, where the DNA is the operating system on the hard drive, RNA is the RAM, ribosomes are the CPU’s, and proteins are output, what is the equivalent of ROM that tells the cell the “boot sector”? (Yes, I realize this almost certainly a very bad analogy, but it seems to me DNA is not entirely unlike a very simple machine language.)
On an even more puzzling note, how could a single-celled organism evolve into a multicellular one? I would assume evolutionary theory on this might have something to do with specialized cells somehow exchanging genetic information via RNA in a symbiotic relationship. Eventually this exchanged RNA somehow is incorporated into the other cell’s DNA, yes? How did polar cells and bicoid RNA evolve? Are there any multicellular species that lack polar cells?
Jois:
Thanks for the info, but it seems to me I’m asking questions that require smaller details than contained in those lectures.
This is all done by the tRNA. The mRNA exits the nucleus and binds to a ribosome. The ribosome/mRNA is recognized by an amino-acyl-tRNA with an anticodon directly paired to the codon in the mRNA.
There are 64 possible codons, with 3 stop and 1 start, and there are a bunch of tRNAs (not 64 because tRNA binding is kind of sloppy) with specific ones for start and stop. There is an enzyme called amino-acyl tRNA transferase which previous to this recognizes the appropriate “uncharged” tRNA and a corresponding appropriate free amino acid and “charges” the tRNA by attaching the amino acid to the tRNA. This is what then goes to the ribosome, which transfers the AA on the tRNA to the “nascent” or growing polypeptide chain.
Sorry for quoting so much but I need it to address all the points here.
Drosophila embryos are a little different than the ones you have described, which is more like mouse/frog/sea urchin/human. A fly one is basically one big cell, with one fertilized nucleus, which divides a bunch of times before the nuclei move to the edge of the big cell and partition into smaller cells. mRNA is pumped in from the nurse cells at the future anterior end. This mRNA is called “maternal effect” - it is made by the mother’s polar cells. It is translated after fertizilation, but gene expression (new mRNA) in the embryo does not start until the nuclei have partitioned after many divisions. So, to answer a later question, if the DNA is the OS, the maternal effect mRNA is the bootstrap. After this, mRNA and protein is made by the new organism, which acts as the first kernel instructions if you wish.
The binding is as precisely understood as anything in biology. bicoid and all the other genes I described are called “homeotic” genes – they contain a protein motif called a “homeobox.” This is a specific 2 coil protein motif (a “helix-turn-helix”) which identifies a well characterized DNA sequence. Binding goes through a whole bunch of related proteins which stabilize the DNA/protein structure. Presumably, once binding occurs, like most gene activators, it interacts with transcription (RNA making) apparatus. These include proteins to unwind the DNA, recognize the start of the gene, and make the RNA. This process right now is not known, but if you propose a nice experiment, the NIH would gladly give you a grant.
In terms of chromatin organization and recognition : Yeah it may happen – the chromatin is somehow organized in the nucleus. Whether it is a “kinetics” thing so that a transcription factor recognizes a gene randomly or a “guided” thing in that 1 protein recognizes the chromatin area and brings the appropriate transcription factor into the gene, that is still a heated question in molecular genetics.
This question is being nicely answered (hopefully) by several avenues of research. There is a wonderful model organism called Dictyostelium discoideum, which spends its life happily as a free-living amoeba given enough food. As soon as a colony starts to starve, it moves together in an aggregate and differentiates into a stalk with a little pod of spores on top of the stalk. The mechanism of this aggregation deals with cAMP, a signaller in many uses in the cell, but often a signal of starvation. RNA is not necessarily a good mechanism for genetic change – it is very unstable out of a cell, and we really don’t see exchange of genetic material like that between cells.
More likely, cells living in a cluster (like many bacteria, which are single-celled) refined mechanisms that single-cellular organisms already possess (like sensing cell density and amount of food around) so that living in a cluster was better than living alone. Eventually, if selected for this long enough, they lost the ability to live alone. After this, it is easy to imagine that they started to get assigned different roles. After all, we all go through dozens of cell divisions before our cells (in the gastrula, after gastrulation during embryogenesis) lose the potential to be any type of cell in the body. That’s how monozygotic (identical) twins happen.
For bicoid specifically, we are looking at a system with many levels of complexity. Before you have polar cells and mRNA gradients, you need germ-line cells. These maybe evolved soon after multi-cellular blobs first started to differentiate. They may have co-evolved with sexual reproduction (as yeasts and bacteria have sex but no germline cells). They started perhaps as cells which stayed undifferentiated through the development process. After this, you can imagine all types of scenarios where they are maintained by support cells, and the positive selection for good maintanence of the germline cells.
It looks as though the most primitive type of signalling across a multicellular organism is concentration gradients. Dicty senses cAMP gradients to find its way to aggregates. Dealing with concentration is a fairly standard thing to do, as the biological system is equipped well to sense different concentrations of things – Kms and binding efficiencies of proteins, as well as dissociation constants and the like all give a concentration sensor.
edwino covered it nicely, but you asked me, so I’ll answer as well.
The ribosome actually has very little to do with the decoding process - it hosts the process and keeps everything in sync by moving exactly 3 bases at a time.
As was mentioned, the real magic happens with the amino-acyl tRNA transferase. That’s the enzyme that does the matching of the codon with the amino acid. How, I hear you cry? Well, basically, one part of the enzyme recognizes a specific codon and another part recognizes a specific amino acid. Poof.
It can actually get a bit more complicated, since there are far more codons than amino acids. Therefore, some of the transferases are sloppy - they only look at the first two bases, not all three. That’s how you sometimes get 4 codons coding for the same amino acid. There are variations on that theme as well.
The other interesting thing about this is what happens with mutations. Sometimes, a mutant transferase will bind the wrong aa with the wrong codon. Of course, if all the transferases for an aa were messed up, death would pretty certainly result. However, since most (if not all) of the transferase genes are present multiple times in the genome, you can get low-level screwups, resulting in an organism that occasionally substitutes one aa for another. This can be useful in some experiments.
OK, it’s lunchtime, and I’m getting the chance to read this through in detail…
I doubt it… I haven’t the experience or know-how in the field–I have to read each of these paragraphs a couple times to absorb the unfamiliar terminology. How do biochemists manipulate and understand these reactions and behaviors on a molecular level?
In computer terms, this is sequential searching vs. indexed searching; you can probably guess which is usually more efficient with large buckets of data.
As for the “magic” of transferase, I actually understand that part now–the different types of enzymes are the code book. Put simply, each type has chemical bonds on one side that match the tRNA codon, and on the other side for a particular amino acid. I was under the mistaken impression the ribosome did the translation–stupid high school biology.
Ok, I was reading this thread, being both educated and entertained when I ran into the long scientific name of some creature listed above.
Now, I’m not too familiar with this organism in particular, but the neat lifecycle and signalling mechanism sounds exactly like that used by Myxomycetes. Is it a member of that class?
If it is, people wanting to see this weird animal can just look around in moist woods around their place for an orange blob of slime, probably on some decaying log.
For those unfamiliar (like me) with official nomenclatures, these things are commonly known as slime molds, which is actually pretty close to the latin for their class.
The ribosome does the translation, in the fact that it is an adapter that brings the whole reaction together. The tRNA binds to the codon, but the ribosome is responsible for holding everything in place while the amino acid is added to the peptide. The AA-transferase just recognizes tRNA by its end bit, and reconjugates it with a free AA.
Kyberneticist :
I’m using the taxonomy browser on NCBI but I am having a hard time figuring it out. It appears the Dicty is in a Eukarya class of Dictyosteliidae. Myxogastria which it says equals myxomycetes, is another class, just like Entamoebidae and Microsporidae which both are free living amoeba (that cause all types of human problems).
The interesting thing about your story is that in the Dicty lab I worked in, we were encouraged to go outside and bring in a handful of soil. Since Dicty eats bacteria, we mixed the soil with bacteria, and put it on non-nutritional agar (so that the only way these things could grow was to eat the bacteria, and once the bacteria were gone, they would starve and develop). We then put the plates in the dark with a slit of light shining at one end. Dictyosteliods eat the bacteria, starve, and then crawl to the light in slug form. When they stop, they form a fruiting body. So I went and got a tube of soil from next to the parking garage here in the middle of Houston, and guess what? Dictyosteliods! It was nice to be able to isolate your model organism – I now work with fruit flies and they are so far removed from the things flying around your trash it isn’t funny. And transgenic mice are about the stupidest thing you could possibly imagine. Just try to picture the result of a brother x sister cross * 40 generations.
Another look at the taxonomy browser makes me think the Dicty is actually in its own phylum. That’s a little hard to believe, but Metazoa (animals) is listed on the same level as Dictyosteliidae, and I know that is a phylum.
