DNA and RNA II

Yesterday we saw the basics of nucleosides, nucleotides, and nucleic acids. We finished with a brief description of DNA and RNA polynucleotides. We said that in both cases we have a repeating backbone, in which we can have ribose or deoxyribose, and bases coming out of the backbone. The backbone is polar due to the precense of negatively charges phosphates and the sugar, and the bases are relatively non-polar. We also have a C5'-end and C3'-end in each monomer, from which the previous and the subsequent nucleosides are attached.

Today we will discuss the structure and physiscal properties of DNA and RNA molecules, making emphasis on the structure of DNA, which was

The DNA double-helix

One of the first clues to the structure and composition of DNA molecules came in the 1940's. A researcher called Erwin Chagraff figured out that the ammount of nucleotides followed a simple relationship. He found that the [A] is equal to [T], and that [G] is equal to [C]. Furthermore, in any fragment of DNA, irrespective of the nature of the organism it was isolated from, the [A] + [G] was equal to [C] + [T]. That is:

These relationships are known as the Chagraff's rules. Although they did not know why at the time, this means that the ratio of A and G is equal to the ratio of C and T. Basically, for every A there has to be a T, and for every G there has to be a C. This was very important, because it tells us how (and which) bases interact with one another.

Second, X-ray structural studies by Rosalind Franklin indicated that DNA was an helical molecules, with a primary repeating pattern every 3.4 Å, and a secondary repeating pattern every 36 Å.

This last piece of data indicated that DNA was a repeating polymer, which had two types of repeating patterns. Furthermore, there was the A = T and G = C rule from Chagraff. Now the problem was to find a 3D structure that fitted this data.

Linus Pauling (the same guy that came up with the a-helix and b-sheets in peptides by model building), tought of an helical structure formed by two strands of DNA in which all the phosphodiesters bonds pointed inwards, and runned along the helix's axis. This, to me, is very strange, because considering all what Pauling new about molecular interactions, it is funny that he came up with a model that placed all the negatively charged phosphodiesters together.

Two very bright chaps, James Watson and Francis Crick, had been gathering information about all this, and were also trying to figure out the 3D structure of DNA to no avail - They had thought of a helix, but they could not figure out some things. They went to a meeting in which Pauling presented his data, and zonk!, they realized what was missing: A doule-helix was needed. So they went back to their offices (they did little experimental work), and they proposed the following model:

- The structure of a DNA molecule is a double-helix, in which one strand runs from the 5' to the 3' end, and the other one goes in the opposite direction. The two strands interweave, giving the whole molecule a right-handed helical twist. The twist of the helix makes the whole molecules to have a new turn every 36 Å.

- In the model, the phosphates and deoxyribose backbones point towards the outside of the axis of the double-helix. This makes sense, because since they are the most polar part of the molecule, there will be favourable h-bonding and interactions with the solvent. The negatively charged phosphate groups will also form salt bridges with Mg²⁺ and other ions.

- The bases are placed so that the aromatic planes are perpendicular to the direction of the axis, and pointing towards the inside of the double helix. This, as we will see, allows them to have p-p stacking interactions with the base before and the one after the strand, and to form h-bonds with bases from the other (antiparallel) strand. We will find the bases separated by 3.4 Å (the space between the planes of two bases).

- The h-bonding of the bases from the two strands is such that an adenine base can interact with a thyamine base, and a cytosine base with a guanosine. If we mismatch the bases, we have no h-bonds.

Here you can see a live model as a CHIME file. Finally, and as we will see below, tthis model also gave Watson and Crick a hint of how genetic material could be replicated.

Watson and Crick base-pairing and base stacking

Lets analyze now the h-bonding of the bases with a bit more detail. In order to obtain a regular, compact, structure, there is only one way in which we can put the DNA bases, or base-pair them. If we take the two purines, A and A or A and G, or G and G, we cannot form reasonable h-bonds. Furthermore, the two base pairs would push each strand far away from each other, and would leave holes if the next residue was a pyrimidine.

The same happens with two pyrimidines. We cannot base-pair C and C, C and T, or T and T, because the h-bond acceptors and donors in each base don't match. Furthermore, if we could, the two strands would be too close together, and we would not leave space for purine bases.

The only way in which we can do this is if we take a purine and a pyrimidine. This solves the 'spacing' problem. Furthermore, we need to have bases that have the h-bond donors and acceptors in a way that they can be shared by both, that is, they have to be complementary to one another. If you try all the possible combinations, the only way we can get this to work  is if we pick an A and a T or a G and a C. These base-pairs are complementary, and fit the model perfectly.

Since this was the original mode of base-pairing proposed by Watson and Crick in their model, it is called Watson and Crick base-pairing. Note that We have one more h-bond possible in the G-C pair than in the A-T pair. Will it be harder to destroy a DNA double-helix with high content of G and C than it is to destroy a double-helix with high content of A and T?

This, as we saw before, is one of the forces that plays an important role in DNA. As we saw, not only will it stabilize the DNA double-helix, it will also determine the complementarity of the strands, and, as we will see below, how the thing is replicated.

The other force that helped in stabilizing DNA was base stacking. This is easy to see in the CHIME model or in the picture above. You can clearly see that all bases are sandwiched between two other bases, the one preceding and the one after from the same strand. These interactions are not specific (that is, we don't need any particular base arrangement), but they are very important, and plays a role in DNA replication.

By looking ant the model of DNA, you can clearly see that there are two 'stairways' by which you could envision climbing the helix. Due to the disposition of the bases and the curling of the phosphodiester/sugar backbone, one is narrower than the other. These are called, respectivelly, the minor and major grooves of the DNA helix:

These two grooves are very important, because there are many proteins and small molecules that bind either the major or the minor grooves. Understanding the structures of the two is required in the design of drugs targeted against DNA replication.

DNA Replication

One more thing that derived from Watson and Cricks model of DNA is the way in which DNA is replicated. They proposed it as a model of transfer of genetic information, which was pretty much correct.

The idea is the following. Since the bases must be complementary to one another, we can use some sort of mechanism to copy them. The process involves the following:

- First, the two complementary DNA strands are untangled. This is done enzymatically, obviuosly.

- Now we have both strands separated. We can envision another enzyme bringing the right nucleotide with a base that is complementary to the base that is now exposed. The enzyme in charge of this is called polymerase.

- As polymerase extends the chains (from the 5'-end to the 3'-end), we get two new strands, each of which is complementary to one of the old strands that came from the original molecule of DNA.

The amazing thing is that Watson and Crick thought of this model decades before all this was actually figured out in the lab.

Structural forms of DNA

One thing that they did miss was other possible structures that DNA can adopt. They are all helical, but different to the model they proposed in 1953. The reason is that DNA is a molecule with several flexible (single) bonds. We have both phosphodiesters, wich can turn, the N-glycosidic bond to the base, which can turn, the sugar itself, which can bend back and forth, etc., etc. Therefore, under certain conditions, DNA will arrange in a structure different to the Watson and Crick model, which is also called B-DNA.

In one case, we have have A-DNA, which is favoured when there is little water content. In this helix, we have a smaller separation between the bases, 2.4 Å instead of 3.4 Å, and we have a shorter pitch per turn. Furthermore, the bases are not perpendicular to the axis of the helix, but perpendicular to the phosphodiester backbone:

It is still a right-handed double-helix. This structure of DNA puts more nucleotides in the same lenght than B-DNA, and is therefore stockier: Shorter and slightly thicker. A model of this can be found here as a CHIME file.

The other structure of DNA is called Z-DNA. It is a lot different from B-DNA. We have a left-handed instead of a right-handed helix, with a pitch of 3.8 Å. This means that it is a lot thiner than B- or A-DNA, and there are far less nucleotides per turn of the helix:

Again, a CHIME file for Z-DNA can be found here. It is beleived that the presence of Z-DNA plays a role in the regulation of the expression of some genes in prokaryotes and eukaryotes. DNA can adopt other structures, but we won't mention them here.

RNA structure

Although that RNA differs from DNA only in one base (uracil for thymine) and the precense of a 2'-OH group in all its sugar moieties, RNA rarely shows up as double stranded helices. We find RNA in single strands, or in base paired structures that are not double-helices.

RNA can be complementary to itself (A to U instead of T, and G to C), and it can also be complementary to DNA. This is the principle used in the transcription from the DNA message to the RNA message (which are the same, but using different molecules.

Therefore, we have different kinds of RNA. One is messenger RNA (mRNA), which is the one made by copying the DNA template in the nucleus. Then we have transfer RNA (tRNA), and ribosomal RNA. The ribosomal RNA forms large part of the ribosome. The transfer RNA is the one that carries and transfers the different amino acids into the ribosome. Each letter of the message of mRNA is compared to a code in tRNA. If they match, the polypeptide chain is elongated by one amino acid.

Becuase of this variety of functions, RNA cannot be as 'rigid' as DNA - We will have a lot more possible 3D structures. You will see all this in detail while studying CH 343...

Chemical modifications of DNA

We will finish today's class by briefly mentioning some of the possible things that can go wrong. Nucleotides, particularly those from DNA, can be modified non-enzymatically by chemical reactions or environmental stress. Many of the chemicals or stresses involved are things that we are normally exposed to. Since these reactions give products with no complementarity to the DNA nucleotides, they will create mismatches in the replication of DNA strands, and are called mutations.

One mutation is the loss of exocyclic amines in purines and pyrimidines:

This reaction is spontaneous in the body. Another important reactions is depurination of adenine and guanosine. The whole purine is hydrolyzed:

This reaction is called depurination, and it is catalyzed by acid in the laboratory. Under the action of UV radiation (sun-rays), we can have a variety of nasty reactions. One of them is a peryciclic [2+2] condensation of the double bonds in thymine:

This reaction happens when we have contiguous thymine residues in the same strand of the DNA double-helix. Finally, many chemicals react either with the carbonyl or the amine groups present in the purines and pyrimidines. One that is very dangerous is methylation, which happens with reagents like (CH₃O)₂SO (dimethylsulfate) or CH₃I (iodomethane). These compounds donate a methyl residue, which react with the tautomeric forms of purines and pyrimidines:

This is bad for two reasons. First, it may affect the replication of DNA itself. Secondly, methylation is used by the organism to regulate the expression of certain genes, and if we methylate DNA, we may trigger or shut down the expression of a gene at the wrong time.

Next time we'll see PCR briefly...

Prepared by Guillermo Moyna, 1999.