DNA Introduction for Computer Engineers

1 DNA Introduction for Computer Engineers Phil Lucht Rimrock Digital Technology, Salt Lake City, Utah 84103 last update: Feb 17, 2015 The following was a lunch-time presentation I gave Feb 7, 2007 to a group of hardware and software engineers at Thomson Electronics in Salt Lake City. Search "phil lucht documents" for PDF. The graphics look ratty in Windows Adobe PDF viewers when not scaled up, but look just fine in this excellent freeware viewer: http://www.tracker-software.com/pdf -xchange- products-c omparison-chart . The Vocabulary Problem ......................................................................................................... .............. 2 The Cell ....................................................................................................................... ............................. 2 Hydrogen Bonds...................................................................................................................................... 3 Power Management, Reaction Ca scades, Temp erature and pH ..................................................... 4 About DNA: a first look....................................................................................................... .................. 4 The Central Dogma of Biology: DNA → mRNA → Protein ............................................................... 6 How proteins are built by tRNA and mRNA: codons and the genetic code ...................................... 8 The enzyme catalytic function of proteins: The Folding Problem.................................................... 12 How RNA and DNA evol ved (conjecture) .......................................................................................... 14 Backtrack now to take a more detailed look at DNA......................................................................... 15 DNA Polymerase ................................................................................................................. .................. 17 DNA Replication (in vivo, meaning in the living organism).............................................................. 17 How DNA is replicated in a test tube (in vitro meaning in glass) : PCR......................................... 20 Determining the DNA sequence................................................................................................... ........ 21 DNA Error Co rrection ......................................................................................................................... 22 Examples of some Special Purpose Protein Devices that help in DNA management ..................... 24 DNA Physical Ma nagement ................................................................................................................. 27 Conserved DNA Sequenc es and Evolution ......................................................................................... 28 Riff-lips (RFLP' s) CSI ........................................................................................................ ................. 28 More Detail on Protein Structure............................................................................................... ......... 28 Some interesting protein machin es : walker and el ectric motor ...................................................... 30 Reverse Genetics and DNA Recombinant methods ........................................................................... 32 Production of insulin and such things .......................................................................................... ....... 33 Control of Gene Expression and Cancer .......................................................................................... ..34 Some fancy cells............................................................................................................... ...................... 34 The Biologica l Hierarchy....................................................................................................... ............... 35 2 Almost all of the images below ar e taken from this unusual website: http://www.ncbi.nlm.nih.gov/ books/bv.fcg i?call=bv.View..Show TOC&rid=mboc4.TOC&depth=10 which contains the entire on-line version of this book: (and other books as well) Molecular Biology of the Cell (4th edition, 2002, the current edition) by: Bruce Alberts, Alexander Johnson, Julian Le wis, Martin Raff, Keith Roberts, Peter Walter In order to see any of the book, however, you have to search on a word or phrase! I have been reading the 3rd edition (1994) which I got at the Salt Lake City library sale for $1.00. One of the authors of the 3rd edition is James Watson who with Francis Crick a nd Rosie Franklin discovered DNA in 1953. If you want to learn biology, don't start with this book. Instead, start with Biology , by Neil Campbell and his heirs. The Vocabulary Problem Both above and below the cell boundary, huge vocabularies have developed to describe things, and this makes reading papers a bit difficult. There is no alte rnative to learning what the words mean. Here are some of the 10,000 examples one could come up with: above the cell: infarction, distal, medial, gymnosperm, angios perm, meristem, cartilage, trochanter, embolism in and below the cell: histone, meiosis, nucleotide, base pair, mito chondria, endoplasmic reticulum, Golgi apparatus On-line dictionaries like hyperdictionary.com are very useful. The Cell • The cell is a midpoint in the bi ological complexity tree. Evolution issues exist above and below it. No living thing is not made of cells. A Virus is not living and just contains some RNA or DNA strands and some protein. Neither DNA/RNA molecules nor proteins are alive, they are just chemicals. • Think of the cell as a large, fully operational city with thousands of workers all very busy. The cell is "alive" in that all its systems are operational including its replication systems. It consumes power and dissipates heat. It is immensely more complex than a Pentium chip and is best thought of as a highly parallel processor that exploits 3D geometry to perform its functions (as we shall see below). Although microscope slides make the city look larg ely empty and boring, that is just because the workers are too small to be seen, and tend to be transparent as well. Here is a picture showing the true density: 3 Hydrogen Bonds All bonding between atoms in chemistry and biology is electrical and is due the "valence" electrons which are the outermost electrons in atoms. Thus, life (as we know it) depends only on the so-called electromagnetic force and knows nothing about nuclear, weak, or gravitational forces. The strongest bonding between two atoms is ionic bonding as in Na + - Cl– making salt. The valence electron cloud is entirely on the chlorine and entirely removed from the sodium. Next strongest is covalent bonding as in O-O oxygen or C-O carbon monoxide or CH 4 methane. In these cases, the electron cloud is shared (by a variable amount) between the atoms and provides an attractive force. When the sharing is not completely symmetrical, the molecule is said to be "polar" . Next strongest is called the hydrogen bond which occurs often in this form: O----HO as, for example, in HHO --- HOH. This is the bond which makes water so cohesive and holds ice together. There are several kinds of hydrogen bonds. The first bond shown is the strongest: O—H...:N (29 kJ/mol) O—H...:O (21 kJ/mol) N—H...:N (13 kJ/mol) // used to hold base pairs in DNA together N—H...:O (8 kJ/mol) // used to hold base pairs in DNA together Biology uses mostly covalent and hy drogen bonding to accomplish its tasks, but there is an occasional ionic bond here and there. Covale nt bonds generally hold structures together which don't change too often, while the weaker hydrogen bonds are used more in dynamic operations where things need to move around quickly. 4 Power Management, Reaction Cascades, Temperature and pH The power to run cells comes fro m glucose sugar molecu les (delivered in humans by the blood). The glucose is "burned" with oxygen in the eucaryote power modules ( mitochondria ), the waste products being CO 2, water and heat. The energy is used to en ergize little "spring loading" molecules called ATP (billions of them per cell) which in turn go around and deliver energy wh ere it is needed within the cell. After an ATP unloads its energy, it becomes ADP and re turns to the power module to be recharged. This recharging is done by electric motors called ATPase (described below) which are located in the walls of the power modules. The power generation process is called respiration : the cells are breathing oxygen. Biological systems do not like large amounts of heat because heat wrecks delicate organic molecules. For this reason, chemical reactions which would give off large amounts of heat generally don't occur in a single reaction. They usually occur in a cascade of reactions with many intermediates, so that each reaction gives off only a little energy. Reactions which require energy also occur in this fashion, with ATP supplying a little energy to each stage of the cascade. Chemical reactions that require energy have a hump to get over called the activation energy. Biological systems cause the hump to be vastly lowered for desired reactions by the use of specific catalysts, which are proteins called enzymes. As discussed below, the main cell workers are proteins which are polymers that get folded into very specific 3D shapes. This shape changes with temperature or pH (acidity). For this reason, biological organisms have subsystems which main tain both temperature and pH in a very narrow range. If you leave this range, proteins stop working (humans: 98.6 F and pH = 7.35, slightly basic) In the larger picture, the glucose molecules which supply power to animal cells are generated in plant cells from solar energy. The entire biosphere is power ed by the sun. Plant cells have power generation subunits called chloroplasts where the glucose is created, after wh ich it is store in fruits and roots. About DNA: a first look DNA and RN A are "nucleic acids". The nucleic part is because DNA lies inside the cell nucleus, and the acid part just because it is slightly acidic in a pH se nse. RNA is "ribo-" nucleic acid because it contains little ribose sugar molecules in its backbone (see later). DNA is the same thing but has one O atom removed from that sugar, so D stands for "de-oxy-ribo-". Both DNA and RNA are simple polymer chain type molecules. They can be any length and in living cells tend to be millions of units long. Short DNA molecules are called oligonucleotides and you can order them on the internet, up to maybe 100 custom units long. There are only four possible building blocks for DNA or RNA. By contrast, proteins (which are also polymer chains) have twenty building blocks and are exponentially more complicated. • Long term information storage is in the DNA. Here is the basic idea: 5 The little lettered blocks are cova lently bonded to the orange "spine" pieces, but are bonded to each other by hydrogen bonds shown as red lines. The four letters A,C,G,T could have been given the names 0,1,2,3. You can think of each letter as a base-four bit, or as two regular bits such as A = 00, C = 01, G = 10, T = 11. The letters are not called bits in biology. Instead, they go by these names: " bases" , "nucleosides" or "nucleotides" . The word "base" arises because the chemicals are slightly alka line, not acidic. The word "nucleo" is because DNA is in the nucleus of modern cells (eucaryotes). The letters are the first letters of the chemical that makes up the base ( eg, C = Cytosine). The letters always match in pairs between the two DNA strands: A ↔ T and C ↔ G. Thus, the DNA double strand has 100% redundancy, it is a RAID 1 system -- a fully mirrored set. You cannot have A ↔ C because it does not fit, and the hydrogen bonds shown in red (attractive force) don't work right. The pairing A ↔ T or C ↔ G is called a "base pair" abbreviated bp. DNA lengths are measured in base pairs. Think of a base pair = 2 bits . • Facts about DNA information storage : human genome has 24 different kinds of DNA strands (24 chromosomes , including X and Y) total base pair count is 3 x 10 9 , but only 1.5% of it does anything, so 45 x 106 base pairs of interest. Including the waste area, person is 6 x 109 bits ~ 1 GB, can fit on a $10 keychain jump drive. Without the waste, person ~ 11.25 MB. genomes range from 106 (small bacteria) to 1011 (lilies) base pairs. • Information density is 1 bp per nm3 = 109 /μ3 = 1021 bp/cm3 ~ 1021 bits/cm3 = 1020 bytes/cm3 = 1011 GB/cm3 = 108 TB/cm3 = 100 million terabytes per cm3. This would make a great data storage system. DNA molecules are extremely stable in the right environment, a great archiving method (it has worked well for 3 billion years). One would use exis ting enzymes to do the read and write operations. 6 The Central Dogma of Biology: DNA → mRNA → Protein Consider the following picture which shows how prot eins are made using information in the DNA: The "workers" in the Cell City are mostly proteins , but some workers are made directly of RNA, in which case they are called rRNA (r = ribosomal, later). The Protein workers do the following tasks: (1) boring job assignment: some proteins are used to form physical structure (bone, skin) (2) others are reaction catalysts, known as enzymes, whose names end in -ase. (3) others are used for communication , or are just assistants in various processes (4) others have motor functions (walkers, muscle) or work in the defense industry (antibodies) The RNA shown above is a copy of a segment of one of the DNA strands. It contains a program subroutine known as a "gene" which contains the "code" to make a protein. The RNA is best thought of as a punched paper tape. RNA of this type is called messenger RNA or just mRNA . The paper tape is delivered to a machine (later) which converts the paper tape's copied DNA subroutine sequence into a specific protein worker, or "device". After the devices are built in the above manner, they often get assembled into more complex mechanical systems called "machines". When devices and machines have done their job, they are deallocated by garbage colle ctor enzymes and their resources are put back on the heap (which in a cell is small molecules fl oating around in solution, "the free list"). When a DNA subroutine is copied to make an mRNA, the process is called "transcription". When the mRNA is then used to program the protein-making machine, that is called "translation" . The combination of these two processes is called "expression" , as in "gene expression". That things work as shown above is called the Central Dogma of Biology . It was once thought that the complex proteins contained the information a nd DNA was boring, but that was wrong (DNA structure was figured out in 1953). It is true that proteins are much more complicated than DNA. There are about 20,000 protein types ("genes") that work in a human, and some are very large and complex. 7 Technical detail: RNA is like DNA, except the lette r (base) T gets replaced with the letter U, and one OH group on the sugar ring is replaced by H (the "de-oxy" part of DNA). Here is a more accurate picture of the Central Dogma: First, biology has two kinds of cells. The procaryotes are ancient and came first, they are bacteria. The modern cell is the eucaryote which has a nucleus where the DNA is stored. The machines that do the translation into proteins from the paper tapes are out on the factory floor ( cytoplasm ), outside the library (nucleus). This allows pre-processor programs to be run on the mRNA (called RNA splicing ), where the waste space introns are removed, and only the code-bearing exons are strung together to recover the true protein generating code sequence. There is much evidence that the eucaryote cell formed during evolution as an assembly of procaryotes that specialized their functions, just as workers in a society end up doing specialized work. One piece of evidence is that the eucaryote contains a power generation subunit called a mitochondrion that is the same size and shape as a procaryote and c ontains its own private and ancient DNA stored in a ring, just as in bacteria. The pictures above are not drawn to scale. The typical eucaryote is 10x larger than the bacteria and has 1,000 as much volume. Roughly the procaryote has a diameter of 1 μ, but the eucaryote is 10 μ. The processes in the two cases shown above are not quite the same. It is possible to find chemicals that disrupt the procaryote gene expression process, but do not block eucaryotic expression. These chemicals can be used as antibiotics : kill the bacteria, don't kill the people. Interestingly, these chemicals are in fact developed by bacteria as protection against other bacteria. 8 How proteins are built by tRNA and mRN A: codons and the genetic code First, there are 20 "amino acids" in biology. These are just simple chemical compounds that have names like Lysine. Just as DNA and RNA are made of sequences of "bases" attached to a backbone, so it is that proteins are made of sequences of amino acids attached to a different kind of backbone. In DNA the bases lie on the inside of the double backbone, but in proteins, the building bloc k amino acids stick out the sides of a single protein backbone. It is the properties of these amino acids that make the protein do its thing. Some amino acids like water, some don't, some are charged, some are polar, some make certain hydrogen bonds, and so on. Short proteins are called polypeptides. An important device used in protein manufacture is the so called transfer RNA device ( tRNA ) that looks like this: As shown on the right, the tRNA is a molecule of RNA that has lots of places where it binds to itself because it has sequences of (complementarily) matching bases. This is an example of a "worker" that is made of rRNA. Most workers are proteins as noted above; but this is a very ancient worker that goes way back to the RNA-only times. A sequence of three bases is called a codon, but we would call it a computer byte. However, since each "bit" of the codon has four values, you think of each letter as 2 bits, so a codon is 6 bits instead of 8. Think of codon = byte with 6 bits. Whereas a normal byte takes on 256 values, a codon can only take on 64 values = 4*4*4. In the central picture above, a codon UGG on the mRNA paper tape is being matched by the "anticodon" ACC on the tRNA. The other end of the tRNA holds the specific amino acid that is coded for by the codon UGG, in this case "tryptophan". Since there are 20 different amino acids, there are at least 20 different kinds of tRNA molecules. Sin ce they "translate" between the codon code and the corresponding amino acid, the process below is called "translation ". 9 Here then is how the tRNA's take part in the protein manufacturing process: The horizontal blue strip with small boxes is a piece of messenger mRNA "paper tape". The round numbered circles are the amino acids being strung toge ther to make a protein. The blue three-legged creatures are the tRNA discussed above. The two green blobs can be thought of as forming a "paper tape reader" that reads the mRNA paper tape. However, the tape stays fixed and the reader is what moves. 10 Here it is moving to the right and as it does so, it constructs the protein as a sequence of amino acids, here numbered 1,2,3,4,5. This is digital logic operating at the molecular level, no doubt about it. The green blobs that form the two halv es of the paper tape reader are called ribosomes. Ribo stands for the R in RNA (ribose sugar), and "-some" is Gree k for a "body", as in "chromosome" (a colored body seen in a microscope before DNA wa s known). Ribosomes are made of both proteins and rRNA, so they have evolved a bit from their ancient ancestors. Notice that there is a byte-framing issue implied by the above design. If you start one base off, all the bits are shifted and the codon bytes are misali gned and you end up making a junk protein that does nothing. This happens when you have a "point deletion" mutation of the DNA. Sometimes a cell needs a large amount of a protein to be created on short order. In this case, multiple ribosomes form an assembly line for more efficient manufacturing (just as in our world). Here is a picture: Here, the ribosomes are moving clockwise around a ring. The more clockwise you go, the longer is the protein string coming out. New ribosomes (floating in the cell solution) jump onto the mRNA tape as space opens up at the start. Notice the use of start and stop bytes in the mRNA subroutine sequence. At the right is a photograph (from EM) showing such a ribosome ring. Here are some pictures of the codon code table : 11 Since there are only 20 amino acids but 64 codes, many codes have redundant use. Certain codes are used as start and stop codes, but this table does not show which. Here is another view of the table: 12 This shows the stop codes, but not the fact that Met is the start code. The upper drawings show those tRNA things we already talked about. One wonders if some exo-life system might use more or less amino acids than ours (assuming it is at all similar). The enzyme catalytic function of proteins: The Folding Problem First of all, the hum an genome codes for about 20,0 00 proteins. This is much smaller than the number thought only a few years ago, and is similar in number to lower life forms. Simple proteins are maybe 300 amino acids long, but complex ones can be ~100,000 amino acids long. Moreover, proteins (devices) generally form multi-enzyme complexes (machines). [ The number of operational genes is not really known, it might be 10,000, or it might be 40,000. ] In chemistry, you have reactions like a + b → c + d. The reaction rate is pr oportional to the density of "a " and the density of "b" in your solution. In cells, all densities are very low and reaction rates are therefore 13 slow. The protein enzyme catalysts increase the rate of desired reactions by a factor of 106 by holding one of the reactants "a" in a "pocket" which exposes just the right part of "a" to the solution, so that when a "b" comes along, the reaction is instant. You don't n eed "fast b" molecules, an y will do. The molecule "a" is called a "substrate" . The catalyst lowers the "activation barrier" that normally limits the reaction rate of a + b. Here is an example: The long protein string folds into a specific shape (t he gray blob) which exposes certain amino acids in a specific 3D geometry (colored disks). Amino acids differ by their sticking-out "side chains" and these are shown as colored blobs on the right. The blue-lined molecule on the right is the "substrate" molecule being held in the pocket and being exposed for reaction with the "b" that comes along. The substrate is held in this example by both electric charge and by "hydrogen bonds" shown as red bars. In general, the shape of an enzyme is not static, but changes in response to the presence of some other molecule(s) at a "binding site" . The presence (or absence) of one of these control molecules at a binding site can cause the pocket to have the wrong shape and not work, so the control molecules form a regulation system. Typically the end product of a long chain of reactions acts as a negative control on the first enzyme, thus self-limiting the production of product -- a standard negative feedback system . The Folding Problem When the linear protein is created by the ribos ome (from the DNA program) as a sequence of amino acids, it "naturally" folds into a specific 3D sh ape as shown above. This shape requires a certain temperature and pH to be the right shape, as in th e above example. No one has yet written a computer program which can take a linear amino acid sequence as "input" and compute the final shape of the folded protein as "output", but this will no doubt be done re latively soon. People are working hard on this. The reverse folding problem is even more interesting. Suppose you want to design a "pocket" to act as a catalyst for some commercial chemical production reaction. Given the pocket shape you want, what linear amino acid sequence would create such a pocke t after the protein folds? Another unsolved problem. [ And what would be the shortest sequence to give such a pocket? ] 14 Sometimes proteins don't fold right. When this happ ens, certain folding manager proteins try to fix them up, and if the fix does not work, the mal -folded molecules are sent off to by recycled. How RNA and DNA evolved (conjecture) RNA w as there first, DNA came in a much later software release [ Life v1018.0 ] . The problem that RNA solved was how to make a chemical reaction one of whose products acted as its own reaction catalyst. Sort of a chicken and egg problem. The RNA structur e provides an extremely "rich" set of possibilities. For example, consider an RNA chain that is 132 bases long (not very long). How many such RNA chains are there: # possible 132-long RNA molecules = 4 132 = 1080 // since 80 = 132 log 10(4) This number is about equal to the number of pr otons in the known universe, an EXTREMELY large number. Most RNA molecules out of these 10 80 do nothing useful, but there are so many that you are bound to find some that do what you want, like cataly ze the chemical reaction that creates the molecule. The self-catalyzing RNA's cooperated and formed primitive ribosomes that could catalyze all their reactions in common. These primitive ribosomes produ ced RNA from RNA, not proteins from RNA. No one, however, has found an organism that works this way (I don't think), so it is just a theory. Proteins then started taking over and enhancing the RNA worker functions, and now most work is done by proteins, but RNA is always found in the bootstrap functional elements such as the ribosome and tRNA. The modern ribosomes and tRNA now are able to catalyze ALL chemical reactions of the form mRNA + amino acids → protein You can see that this "general purpose solution" of the catalyst problem is what we think of as a "computer", a sort of programmed milling machine. E volution found this solution just as human industrial evolution found milling machines. In human society, all work was originally done manually (RNA), but after society evolved, almost all work is done by machines (proteins), but there are still places where the work is done by a combination of machine and manual labor (ribosome). 15 Backtrack now to take a more detailed look at DNA Here is a blo wup of one DNA strand containing 4 bases: The actual "bases" are the small molecules shown in green. When you combine the blue ribose sugar ring with the base, you have a "nucleoside" , and when you add the phosphate PO 4 group to that, you get a "nucleotide" . Notice the letters G,A,T,C on the right. The DNA "backbone" is formed from the phosphate and ribose (missing one O) sugar molecules. Here is a detail showing how the "hydrogen bonds" hold the base pairs together: 16 It is because the bonds are in differe nt locations that you cannot get a T ↔C pairing, for example. Here is a 3D picture of the above situation: Look for the bases A,T,C,G and for the hydrogen bonds that stabilize them between the two spines that form the outer part of the helix. There are 10 base pa irs per turn of the helix. Diameter of the DNA helix is 2 nm which is 20 angstroms (A) or about 20 hydr ogen atoms placed side to side. Length of a typical human DNA is 5 cm = 50,000,000 nm. This is 5,000 times the nucleus diameter of 10,000 nm, which is why DNA has to be folded up a lot to fit in there. 17 DNA Polymerase This is a protein enzy me which is used to build the "other side" of a DNA chain if one side is present. It does a RAID 1 "rebuild". Here is how it does it: Here we are building the DNA top strand from the existing lower strand ( don't confuse this with earlier pictures showing how proteins are constructed). The little nucleotides just "float into position" in a pocket provided by the enzyme shown in green (called DNA polymerase ). Note that the DNA polymerase only works in one direction , it creates the second strand in what is known as the " 5' to 3' " direction. Also, it needs some existing DNA double strand in order to get going. If this is not present, some other device called DNA primase comes along and glues on a "primer" to allow DNA polymerase to work. In the picture on the right, a single strand comes in, a double strand goes out. DNA Replication (in vivo, meaning in the living organism) When DNA i s replicated (during cell division), replication happens at a replication fork . Here is what that looks like schematically: Because the DNA polymerase can only work in one direction, it works efficiently on the "leading strand" , but on the other "lagging" strand it has to wo rk backwards in little fo rward segments called 18 Okazaki fragments . But it gets the job done. Here is a more detailed (but still schematic) picture of the replication fork: Notice that the two DNA polymerase enzymes slid e along the DNA helix and are locked in place by sliding clamp rings (made of two proteins, put in place by clamp-loader devices). The DNA helicase at the fork has the job of untwisting the helix in to two single strands. The green enzyme is a primase which has to constantly lay down primers for the backward s Okazaki segments. The brown proteins are used to keep the lagging strand from curling up before it can be worked on. This picture shows how a "machine" is formed by combining multiple "devices". The m achine overall can do a job that the individual devices cannot do. Finally, here is how two forks operate at once to do the actual DNA replication: 19 20 How DNA is replicated in a test tube (in vitro meaning in glass) : PCR First, here is the picture: At the left we have a single piece of double stranded DNA (maybe from the hair of a criminal on CSI). It gets heated up and separates into two single strands. A primer enzyme (primase) puts primers on both single strands, and then the usual DNA polymerase repli cates each single strand into a double strand. The process is then repeated on each of these double strands. After 30 cycles, you end up with 2 30 identical chunks of DNA, which is a billion strands, enough th at you can do work with it (to do riff-lips, for example, see below). This process is called PCR for polymerization chain reaction . I think people were stunned at how easy the above process is (1983). Of course you have to use biologically created enzymes, they don't yet know how to make large enzymes by chemical synthesis. But why bother. 21 Determining the DNA sequence Again, first here is the picture: The idea is very clever. Consider the left column of orange boxes. Ignore the letters in these boxes, it is just a primer thing that lets you find fragments in th e final gel (I think it attaches to some chemical that makes the fragments "light up"). The DNA sequen ce being studied is this one: ATGTCAGTCCAG... You take a double strand DNA and heat it (as in the PCR reaction above) and it breaks into single strands. You then use the usual DNA polymerase to reform double strands from the single strands. BUT, you include a small amount of BOGUS "A" base in the soup. It looks like an A, feels like an A, but does not work like an A. When one of these bogus "red A's" is added to the chain, it can no longer grow because of the bogus quality of the A. This bogus A could hit at ANY of the A locations, and three are shown in the left column after the three orange boxes. So you get three fragments of three different lengths (weights). When these are put into a standard electrophoresis diffusion gel , the lightest one (just a single bogus A) diffuses the farthest and ends up at the bottom. The heavier fragments diffuse less in proportion to their size/weight. This then explains the thr ee horizontal pink bars on the left side of the gel. 22 You do the same experiment with bogus bases for the other three bases T, C and G. You end up with four lanes on the gel, and you then just "read off" the sequence from bottom up. This process is highly automated nowadays, and a machine th at does the above process is called a "gene sequencer ". The idea of using a "bogus base" has application in cancer chemotherapy . For example, the drug Gemzar consists of a bogus C base which blocks DNA replication just as described above. The only problem is that such "nucleoside" drugs tend to stop all DNA replication in the body, causing sickness. See http://en.wikipedia.org/wiki/Gemcitabine which you w ill find very readable. This kind of treatment is sometimes called "carpet bombing", but it is one of th e few that actually have some success. For sure, more specific treatments will appear. DNA Error Correction In biolog y it is called "proofreading ", and as usual, it is based on 3D shapes. If a wrong base pairing occurs, the DNA helix has an external bump in it. This is felt by a shape-checking drone protein which then does the error correction as follows: An entire region including the damage is chopped out so you get the second picture from the bottom on the left side. Then DNA polymerase comes along and regenerates the corrected DNA strand from the undamaged strand ( a RAID 1 local rebuild). The picture on the right is one of the error corrector proteins displayed in a certain graphical form we will discuss below. I have read that there are 50 different error corr ection mechanisms that are operational all the time. Here is a graphic that shows the above one in more detail, and a point-error fixer-upper: 23 On the left is a "single bit error" (single base error really) repair. It starts with a U accidentally taking the place of a C. The offending U is jerked out by one enzyme, the spine is "nicked" by another, and then DNA Polymerase does the repair (DNA ligase does the operational "close"). [ Nothing much happens in a cell without the aid of a custom enzyme. ] On the right is a "burst error" where a "whisker" has grown between an adjacent C and T. This whisker will cause a malfunction and needs to be repaired. As in the previous graphic, a nuclease (enzyme that can cut DNA or RNA) comes along and removes a whole segment of one strand after untwisting a piece of the DNA (done by DNA helicase ). Again, the polymerase and ligase enzymes then do the repair. Everything is based on the full RAID-1 mirror re dundancy. The DNA is so well protected that you generally don't get mutations that damage both sides of the helix. The error corrector knows which side of the helix to repair when it gets a mismatch, because it knows which side is being created by replication. Without error correction, the DNA replication base error rate is about 1 in 10,000. With error correction, it is about 1 in 1,000,000,000. No a ttempt is made to do error correction on RNA because a base error there will at worst create a non-functional rRNA or protein which will get trashed by the recycling staff. Obviously, it is in DNA → DNA + DNA that you want to focus your error correction efforts, since we are talking cell replication and propagation down the evolutionary channel. [ Cancer! ] 24 Examples of some Special Purpose Protein Devices that help in DNA management During replic ation at the replication fork, as the DNA is untwisted, the still-twisted part gets twisted tighter and tighter, just as what happens when you play with a curled phone cord. Eventually replication would stop or the DNA might break. A special device (protein) comes into play here, which I would have called "phone-cord-ase" or "un-twist-ase". Since this device is repairing a "topological problem" in 3D space, and since it does so by temporarily changing th e shape of something (isomer), the device has the name "topoisomerase I" . Here is how it works: 25 This green device opens one side of the DNA temporarily so the right side (here) can relieve tension by untwisting, then it reseals the opening and th e job is done! A twist-tension remover. Another problem is a more general one with DNA, as we shall see, in that it tends to get tangled up since it is so long and skinny, like fishing line. In the top picture, the DNA has locally threaded itself as part of some larger tangle. We want to unthread the loop to get the lower picture. 26 abovebelow So here is the solution, "topoisomerase II" , which takes two tangled strands and removes the tangle. The red and orange tubes are a place where the DNA is tangled in a loop, and we need to somehow move one piece through the other piece. I wish I could do this with my orange power cord. The orange strand is opened so the red strand can p ass through, then the orange strand is reconnected. This example shows a protein machine in a more typical light, where there is "action" , not just something fairly static like the enzyme pocket. The action is triggered by special sensor molecules not shown in this picture, and the action is power ed by ATP. If you do a Google image search on topoisomerase, you will see an impressive animation show ing how the real molecule does the above job. 27 DNA Physical Management A strand of DNA is 5,000 times as long as the nucleus is wide, so it has to be ma naged in some manner to avoid a hopeless tangled mess. ALSO, the DNA has to be dense-packed during cell division after DNA replication so the daughter cells each get a copy of the DNA library. These are big problems. Here is a graphic showing how the DNA is organized at least just prior to cell division: At the top is the DNA strand. It gets wound on little protein bobbins called nucleosomes with each bobbin getting two DNA wraps (160 base pairs). The bobbins are then close-packed as shown into what is called the "30 nm fiber ". This then goes into lampbrush loops which form more loops, and the DNA ends up in the famous chromosome structure shown (t hat looks here like a sideways H). As it says, the chromosome is 10,000 times shorter than the DNA, and is therefore manageable (and fits in the nucleus). The upper and lower half of the chromosome cont ain the two identical DNA double-strands (because chromosomes are formed after DNA replication). A m echanical system then grabs the two halves at the crossing point (the centromere ) and pulls them apart, and delivers one copy to each daughter cell. The entire process shown above is managed by protein devices called histones. 28 Conserved DNA Sequences and Evolution The m ouse and human genomes separated some 100 m illion years ago, so they are separated by an evolutionary distance of 200 million years. During this time, mutations occurred at known rates, and one can see that the non-coding portion of the two DNA genom es differ by the expected number of mutations. By itself, this is one argument for the fact that the man/mouse branching even occurred. But certain parts of the DNA are the same despite this long evolutionary separation. The reason is simple: critical genes cannot be altered because if they are, the organism dies, and no evolutionary pr opagation of that mutation survives. This histone above is one of the most conserved gene sequences observed. This is because DNA spatial management is a critical function. Riff-lips (RFLP's) CSI Due to mutations over long periods of time, peopl e have difference stuff in their uncoded DNA regions. You can think of these as being different allele s (gene subprograms), but these regions never make proteins so that name would not be correct. These variations are called polymorphisms. People who have different DNA will have different fragment size sets when their DNA is subject to breakdown by "restriction nuclease" enzymes, where the word r estrict means to break into pieces at well-defined code sequence sites. It is fairly easy to take some DNA, amplify it by PCR, apply restriction nuclease, then display the fragment sizes on a gel. If your fragment size distribution looks the same as that found in a hair you left at a crime scene, then you must be the criminal with very high degree of probability. This method is much faster and cheaper than actually seque ncing the DNA of the crime scene hair and that of various potential suspects. This method is called RFLP which stands for restricted fragment length polymorphism. ******************************************************************************* More Detail on Protein Structure The way proteins fold is very complicated, and has various levels. The linear am ino acid sequence is just the "primary structure". The next level is the "secondary structure" where pieces of the protein form little helices or little flat ribbons (called α helices and β sheets). Any polymer like DNA or a protein, given complete freedom, wants to form a helix because that is just what happens which you stack near-identical blocks together in a pile with a binding spine. Bu t when protein strands run parallel to each other, sometimes the hydrogen bonding between strands make them form ribbons instead: So here is the alpha helix protein secondary structure: 29 and here is the beta sheet: The next step in folding is the "tertiary structure" where "domains" are formed, here is an example: 30 The protein shown on the left has four domains , and you see the helix/ribbon notation used. When multiple domains are combined, you have "quaternary structure". The mess on the right is the real thing and is pretty much incomprehensible (the balls are amino acids). In general, action occurs where the protein has a bend and some amino acids "stick out ", and these bends normally occur on the outside surface of the molecule. A sticking-out (and in fact any) amino acid is called a residue . The above pictures should give some feel for the difficulty of "the folding problem" mentioned earlier. Some interesting protein machines : walker and electric motor Lest y ou doubt the veracity of this section, here is a web snippet: • Kinesins are a large family of motor proteins , most of which walk along microtubules toward the plus end , away from the centrosome (MTOC). • Dyneins walk along microtubules toward the minus end (toward the centrosome). The first machine below is called a "walker" and I have seen animations of these things, very amazing. The green blob is a protein (or maybe a set of proteins ), and the shape changes ar e induced by externally applied control molecules (not shown). Each shape cha nge can be thought of as a chemical reaction. As long as at least one of the reactions is "irreversible", the walker will make progress in the direction it is headed, rather than randomly bob back and forth. Wa lkers (motor proteins in general) are used as a delivery system in the cell, and they walk on small rails called microtubules which fill most of the cell. ATP→ADP provides the power. 31 32 The second machine is an rotating motor /charger called ATPase: This motor penetrates the membrane of the cell's power generation module called the mitochondrion. That power module takes in glucose (energy) and conver ts it to a gradient in proton density across the membrane. The protein gradient creates a backflow of protons through the membrane which causes the above motor armature to rotate freely. The bottom end then converts the rotational energy into activation of ATP molecules, and these in turn float around in the cell and power all reactions that need power. When an ATP molecule gives up its energy, it beco mes ADP and returns to the above charger to get activated again. Amazingly, someone has actually take n a movie of the actual motor turning. Notice that its diameter is 10 nm which is 100 angstroms or 100 H atom diameters. To see it rotate, they had to attach a long light bar to the rotor and they had to tag that bar with something that "lights up" so you can see it. The actual motor design is shown on the right: notice th e regions of alpha helix and beta sheet described above. The above is really an "electric motor" since it is run on an electric current. The current happens to be protons instead of electrons. Sometimes the motor is run in reverse and becomes a generator which consumes ATP energy and creates a proton current to increase the membrane proton gradient. The motor is made of several kinds of proteins as shown by the colors. Biology does not bother to put shiny metallic covers on its motors, so the devices al ways look a little strange to us. Reverse Genetics and DNA Recombinant methods Of the ~20,000 hum an proteins, I think only about 5 00 are "understood" inasmuch as it is known more or less what they do. In regular genetics, the idea w as to do a lot of breeding and wait for a mutation to occur, as indicated perhaps by a weird blue eye color on a fruit fly. You would then conclude that that mutation occurred in the gene that determines eye colo r. Maybe you could then locate that gene on the DNA somehow. 33 This brings up the subject of alleles , which are different forms of the same gene. Usually the good form is called A, and the mutated damaged form (usua lly it is just inoperative) is called a. Perhaps the good protein form creates blue eyes instead of defau lt red eyes. An animal gets one gene from each parent. If the animal is Aa or aA or AA, it makes the blue eye protein and so organism has blue eyes. But the 1 of 4 offspring that are aa will have red eyes (perhaps color of blood). When the gene in question performs some vital body function, then the person who comes up aa has a "genetic disease" (there are very many such diseases). The aA and Aa people are then carriers because their kids might have the disease. Organisms that are AA or aa are said to by homozygous , while aA and Aa are heterozygous. Sometimes A and a are called dominant and recessive . So we come to reverse genetics. The idea here is to intentionally damage a protein's gene, and then see what that does. You don't do such experiments with people. You take a piece of DNA, you alter it using genetic engineering techniques, then you put the altered piece into a replication factory (bacteria), and then you inject the final resulting DNA in to a test organism, which is now called a transgenic organism . You look to see what is broken in that organi sm. This is certainly an arduous task. Obviously if you hit something important, the organism dies. Y ou generally have to breed the organism until the aa homozygote is formed to see the problem. That is to say, organism was AA, but you made some Aa which you bred to makes some aa. Presumably lots of people are working hard to figure out what all the human proteins actually do, but they have to use other methods. Production of insulin and such things The idea here is to find the insulin gene in the DNA, cause that gene to be inserted into a bacteria DNA ring, and then let the bacteria breed like mad. As they breed, they replicate the DNA including your piece of interest, and they express the protein you want, li ke insulin. Here is the plan for the DNA insertion: The product like insulin is then harvested from the re sulting bacteria. If you could make a plant gene that creates methane gas (from solar power), you might solv e the energy crisis (but not the carbon problem). 34 Better to have your plant create a macroscopic organi zed proton gradient and take off electric power directly, and have an associated plant fix the generated carbon (CO 2) back into glucose. When you blend DNA as shown above, you are doing DNA recombination , and the technology is recombinant DNA technology. Another approach is to alter plant genes to improve plant quality, done with the same techniques. I think this is why we have tasteless red rubber tomatoes. The whole GM area. Control of Gene Expression and Cancer This is a hot topic. It is well known that early stem cells are capable of expressing all the proteins that are coded in the DNA library, but cells somehow differentiate and then can no longer express certain genes. They specialize. The question is: how is this control exercised? Some of it seems to be signaling protein messages floating around, some of it seems to be the shape of the bulk DNA (see loop pictures above). Some diseases occur when genes can no longer be expressed, and some (like cancer) occur when the expression of genes (due to growth factors) cannot be turned off. Some fancy cells Procary ote organisms always have one cell, but eu caryote organisms can have one or many cells. This picture shows some of the amazing organisms that are made of a single eucaryotic cell: (protists) 35 The Biological Hierarchy quarks or whatever ele mentary particles exist \\ physics protons, neutrons, electrons \\ physics a t o m s \ \ p h y s i c s simple molecules (inorganic) \\ physics → chemistry complex organic molecules (like DNA, proteins) \\ chemistry → biochemistry small scale structures (like cell wall) \\ → molecular biology larger pieces of the cell (organelles, like the power module) cells \\ cell biology tissues // heart muscle \\ anatomy, medicine organs // heart organ systems // circulatory system organisms // ape \\ psychology, behavior, naturalists communities // families of apes living together \\ sociology, politics ecosystem of multiple communities // the biosphere \\ ecology issues multiple ecosystems // other life-bearing planets \\ exobiology