Biophysicist Maxim Frank-Kamenetskii on DNA supercoiling, topoisomerases, and enigmatic organisation of chromosomes
A DNA molecule is a double helix consisting of two polymeric chains, which consist of monomeric units, nucleotides. A nucleotide consists of three parts: a phosphate group, a sugar, and a base. The phosphate group and the sugar form a sugar-phosphate backbone of both strands so DNA consists of two sugar-phosphate strands to which bases are attached. As it’s well known, there are four bases — A, T, G, C — and they form pairs: A pairs with T and G pairs with C. This pairing leads to the formation of the double helix: two strands anneal by A-T and G-C pairs to form a duplex. It is also important to note that each strand has a chemical direction and in a double helix they are anti-parallel: one strand goes in one direction, while the other strand goes in the opposite direction.
The double-strand DNA model was postulated by Watson and Crick in 1953 and it marked the beginning of a new era of molecular biology and later of biotechnology. This was the beginning of modern biology — the discovery of the double helix. In the beginning, not everybody embraced the model. Very soon, however, it became clear to many researchers that this model was the reality. Although it was not based on an absolutely solid ground and originally, it wasn’t unambiguously determined, many people believed in it.
That’s when the belief in the model and the interest in doing more research started. Eventually, the original model of Watson-Crick has been proven, but it happened 30 years later after the initial discovery. The most direct proof of any molecular structure is the X-ray crystallography.
The first determined DNA structure — of short DNA molecules that were investigated using X-ray crystallography – has proven not to be Watson-Crick’s double helix. The Watson-Crick double helix is technically called a B-form DNA. However, when Alex Rich and his group at MIT determined the structure of the first short DNA molecule consisting only of 6 base pairs, to their astonishment, it turned out to be a totally different structure. It was not a right-handed helix, but a left-handed helix and there were many other differences as well. The DNA structure they determined was called Z-DNA. Therefore, there are some alternative structures of DNA other than Watson-Crick’s DNA. Later, the researchers understood what happened was because it was a very special sequence, and under very special conditions. And for this sequence of DNA under these conditions, the DNA adopts a different structure. The B-form structure was solved by X-ray crystallography soon after this. In the cell, DNA is predominantly in the B-form although some special sequences form slightly different structures even under normal conditions. This different structure is called the B’-form. It has some distinctions from B-DNA. Normally, people ignore this fact but, in reality, some regions in DNA are not exactly in the B-form.
DNA supercoiling arises mostly in prokaryotic cells, in bacteria, because in eukaryotes like humans and other multicellular organisms DNA molecules are linear and also are very long: they have ends, they are not circular. By contrast, in prokaryotes, DNA is always circular and so this circularity leads to a lot of different properties of DNA, it leads to the notion of DNA topology. This is because when you have a circular molecule, it can form supercoiling and it can form knots, so this circularization of DNA in cells leads to different properties of DNA.
Topology, or as you can call it, pseudo-topology, plays an important role in linear DNAs as well. If DNA is not circular, but very long, in many respects it behaves as if it were circular. Thus, topological issues are very important in both prokaryotes and eukaryotes. This is emphasized by the fact that there are special enzymes called topoisomerases, which change DNA topology. They are present in eukaryotes and prokaryotes and if you shut down or silence the genes of topoisomerases in eukaryotes, all major processes, such as replication and transcription, in the cell will stall. This shows the importance of topological properties, or pseudo-topological properties, in eukaryotic cells.
Topoisomerases are necessary for normal replication because, in any process involving the separation of complementary strands of DNA, the issue of supercoiling immediately arises. Whenever you separate the strands, you unwind DNA. Since DNA is a double helix and it is a very highly wound helix, when you separate the strands and unwind these turns, this unwinding changes the supercoiling in double-stranded DNA ahead of the replication fork. And so, if you don’t remove the supercoiling, the replication just stalls, it cannot continue. This happens with replication and also with transcription. This was all very well studied, the formation of positive and negative supercoiling and the removal of supercoiling was demonstrated in both processes, replication and transcription.
In the case of eukaryotes, it is the same story. Although strictly speaking, it isn’t topology because molecules are linear and in principle they can relax the supercoiling, if you can wait for long enough time. But cells have very long DNAs and so to relax all of the supercoiling, you will need to rotate the molecule many-many times, and this is just impossible in a ‘crowded’ environment within the cell. In reality, although there is no strict notion of topology for linear DNAs (it is pseudo-topology), the relaxation of supercoiling cannot happen in the reasonable time scale within the cell. Thus, the cell accelerates this process using topoisomerases and deals with DNA as if it were circular.
If eukaryotes didn’t use topoisomerases, they couldn’t have long genomes and having long genomes is, of course, crucial for the eukaryotes’ evolution.
If we talk about how something was discovered and not about how things exist in nature we are mostly talking about prokaryotes because there, at least with circular DNA, you can study it in a test tube, where there is no crowded situation. This way it proved to be much easier to study the enzymes and their properties dealing with circular DNA. To understand the mechanisms of topoisomerases’ actions, short circular DNA molecules were used. It is not even bacterial DNA, which is also too long, but short circular molecules that were used to figure out how topoisomerases work. When you already know what these enzymes do, you can just knock out the corresponding genes and see what happens. This is how the importance of topoisomerases for eukaryotic cells was proved.
Over the years we have learned a lot about DNA topology and many important questions have been resolved. One thing that has been discovered already 20 years ago, was that these topoisomerases, at least some of them, which work on double-stranded DNA, are actually capable of disentangling DNA molecules. They disentangle knots and they disentangle molecules that are linked to each other. That was a very interesting discovery and it is still an issue under investigation because it is not totally clear how an enzyme, which is much smaller than circular DNA itself, can distinguish between knotted molecule and unknotted molecule. But the fact is that somehow it disentangles knots. In the area of DNA topology theory actually plays a crucial role. Currently, studies are mostly theoretical on the ability of so-called topoisomerase type 2 to disentangle knots. There are other issues which are being studied but, in principle, many questions in this area have already been understood, so we know a lot about topoisomerases, about their structure, and we know in detail how these molecules work. It is an extremely interesting area.
Of course, there are still open questions left. The main one is how DNA topology influences gene expression in eukaryotes. It has been a long story for many years: there were indications that there are so-called topological domains exist in DNA. It is also thought that DNA in the chromosome is arranged in the form of loops, which are not circular DNA but behave exactly like circular DNA since these loops are somehow locked by special proteins. These proteins are important for the formation of domains since they do not allow two DNA parts to rotate around each other and in this respect, a pseudo-circular DNA domain is formed. This domain can be supercoiled, topoisomerases can work on it and so forth.
For many researchers, it looked very attractive that such domains could be regulated with respect to supercoiling, and depending on supercoiling, gene expression may be affected. This issue has been raised many times in the literature and by different authors but there is still no clear answer to whether it is true, whether such domains are real. Another unresolved issue is that we know very little about the chromosomal organization, apart from nucleosomes. Nucleosomes provide a very generic organization of DNA in the chromosome, but it is only very first level of DNA organization within chromosomes. We still don’t know about higher levels of DNA compactization in chromosomes. It is very hard to study chromatin because it is a very unstable, changing entity. So in the future, we may have some fantastic discoveries in this respect if somebody is able to really pinpoint this issue of topological domains. It is a possibility but it has not yet been proven and studied properly because it is very hard to study. This is an open question and from time to time, a new generation of molecular biologists try to attack this issue but there is still no clear answer.
Edited by Vera Vasilieva