1 MCB 502A-2015. Lecture #6. Enzymatic synthesis of DNA. DNA polymerization directionality Is enzymatic DNA synthesis directional, and if yes, is there a preferred direction? As you know, DNA degradation can proceed in both the 3’→5’ and the 5’→3’ directions along the DNA strands. Moreover, there are enzymes that can degrade DNA strands in both directions, like ssDNA-specific ExoVII. There are even enzymes, like ExoV, that can degrade both strands of the same DNA duplex at the same time. If DNA polymerization is anywhere similar to DNA degradation, then a priory we can assume that there are also three possibilities for the DNA polymerization directionality: 3’→5’; 5’→3’; both. To determine the directionality of DNA polymerization, one needs primed DNA templates that can be extended in one direction only: in either 5’→3’ or in 3’→5’. There are at least two ways to generate such primed templates: 1) by terminal nucleotidyl transferase, which can synthesize long homopolymers, but can also sequentially synthesize combinations of different homopolymer stretches, for example, single long nucleotides of the following structure: 5’-dA4000dT100-3’ or 5’-dA100dT4000-3’. There will be 3’-end primers available in the former substrate, while 5’-end primers available in the latter. 2) by a partial hydrolysis of a duplex DNA with exonucleases of known polarities. For example, ExoVIII processively degrades the 5’-ending strands of duplex DNA and, thus, generates 3’-overhangs. Alternatively, ExoIII processively degrades the 3’-ending strands of duplex DNA and, thus, generates 5’-overhangs. The results of Kornberg polymerase reaction with the four templates are: Thus, whenever there is 3’-paired end, there is DNA synthesis, but whenever there is a 5’-paired end, Kornberg polymerase cannot synthesize the complementary DNA strand. Therefore, this enzyme can synthesize DNA in the 5’→3’ direction only. One could arrive at the same polarity indirectly, by determining the polarity of the strange exonuclease activity of Kornberg DNA polymerase that is stimulated by DNA synthesis. The nick-translation reaction propagates the nick along the DNA by employing simultaneous DNA degradation and re-synthesis going in the same direction. Please recall that this exonuclease activity, although much inhibited, is still detectable in the absence of triphosphates. Its polarity is 5’→3’ on dsDNA, so the polarity of DNA synthesis must be also 5’→3’. Why 5’→3’? Could we figure out the polarity of DNA synthesis by isolating precursors of DNA synthesis from cells (or determining them ourselves experimentally in vitro)? Monomeric precursors for any biopolymer are different from the monomer units within the polymer in that one of their side is "activated" by an energy-rich bond. In case of DNA, the monomer units within DNA are dNMPs (deoxynucleoside monophosphates), while DNA precursors are dNTPs (triphosphates), but which side of the dN is charged with PPP, — the 3’ or the 5’? There are no 3’-dNTP in vivo, only 5’-dNTP. Similarly, in vitro, only 5’-dNTP support DNA synthesis. Does the invariable 5'-position of the triphosphates on the DNA precursors commit DNA synthesis to the 5’—>3’ direction? Template + primer Paired end Action dA100dT4000 5’ no synthesis dA4000dT100 3’ synthesis ExoVIII-treated dsDNA 5’ little synthesis ExoIII-treated dsDNA 3’ significant synthesis2 Surprisingly, no. Mechanisms of various polymer chain growth may be grouped into two broad classes: tail growth and head growth. The tail growth mechanism involves an attack of the non-activated end of a polymer on the activated end of a precursor. Examples of tail growth include polymerization of DNA, RNA and some polysaccharides. Head growth involves attack of the non-activated end of a precursor on the activated end of a polymer. Examples of head growth include protein synthesis, fatty acid biosynthesis and synthesis of O-antigen (Fig. 3-6 of Kornberg). Successful examples of head growth argue that there is no big reason why DNA should be growing tail-first rather than head-first. There is one nuance that makes the 5’—>3’ DNA polymerization preferred if you have exclusively 5’-dNTPs in the cell. In order to make polymerization irreversible, the released pyrophosphate is hydrolyzed to keep its concentration low (otherwise pyrophosphorolysis becomes a problem). The 3’—>5’ chain growth is still an option. However, once in a while, for the chemical reasons beyond cell's control that will be discussed later, DNA polymerase inserts a wrong nucleotide, say T opposite G. Still not a big deal, since all DNA polymerases have the so-called proofreading exonuclease activity (that we will discuss next). The enzyme backs down one position and hydrolyzes the last incorporated nucleotide (dTMP). (Pyrophosphorolysis is not an option, because for pyrophosphorolysis to be efficient the enzyme should retain the outgoing pyrophosphate, which would make polymerization highly reversible, therefore inefficient). If the synthesis is tail-first, the enzyme is ready to go again. However, if the synthesis was head-first, the 5’-end is not ready for synthesis restart, since it now has a monophosphate on it, instead of the required triphosphate. The 5’-end needs to be charged with a pyrophosphate before the enzyme can continue, and this will require a separate enzyme and will translate into a significant delay. Therefore, with all DNA precursors 5’-dNTPs and the concentration of pyrophosphate low to make DNA replication irreversible, it is argued that the efficient synthesis can go only tail-first, that is, in the 5’—>3’ direction, because after proofreading, the restart requires no further action. By the way, the same 5’—>3’ direction of transcription, which has no proofreading, argues against the idea. The "transcription" argument against is a weak one, though, as the original replicative RNA-dependent RNA polymerases (in the original RNA-world) likely did have the proofreading function. Sadly, the notion that the 5'—>3' direction of DNA synthesis was selected because it accelerated proofreading, as many other notions pertaining to what happened some 3.5 By ago, is generally impossible to test, so this nice idea is another example of "non-science". The proofreading activity of Kornberg
View Full Document
Unlocking...