1 MCB 502A-2015. Lecture #7 Chromosome replication: dna and pol mutants Isolation of a polA mutant Last time we finished with isolation of various dna(Ts) mutants. The mutants unable to synthesize their DNA at higher temperature belonged to two major interesting types (immediate-stop versus delayed-stop) and to at least ten different complementation groups (individual genes). Remarkably, none of the dna mutants have any change in the activity of Kornberg polymerase, suggesting its non-replicative nature. However, the last nail in the coffin of the idea that Kornberg DNA pol is the replicative enzyme was driven by the same John Cairns who earlier demonstrated theta-replication for the E. coli chromosome. Cairns reasoned that a replicative DNA polymerase should be an indispensable enzyme. If Kornberg DNA polymerase is dispensable, it cannot be the replicative DNA polymerase. In vitro properties of Kornberg polymerase made it unlikely that this is a replicative enzyme, — so, John Cairns convinced Paula De Lucia to try and isolate a mutant deficient in Kornberg DNA polymerase. This was not an easy task. Since all these enrichments for dna-minus mutants never yielded a mutant deficient in Kornberg enzyme, Cairns and De Lucia knew that this approach would not work. Therefore, they attempted a brute force screen: Paula heavily mutagenized a culture of E. coli and assayed individual clones, derived from this culture, for the activity of Kornberg DNA polymerase. This, of course, required simplification of the assay procedure, so eventually Paula was able to analyze 100 mutants a day (J. Cairns, personal communication). Paula analyzed 3,500 candidates before she was rewarded with a single mutant, clone #3478, that lacked the DNA polymerase activity in Kornberg’s assay conditions. Julian Gross suggested to call the gene for the Kornberg DNA polymerase polA (so that it could be pronounced “Paula”, in memory of the amount of work Paula De Lucia put into the isolation of the mutant), and the isolated mutant received the designation polA1. Actually, we know now that a complete polA deletion would have been inviable under the assay conditions. The polA1 mutation inactivated the polymerase activity of the enzyme, but left the 5'—>3' exonuclease intact, which was important for the robustness of the mutant. Interestingly, although activity of Kornberg DNA polymerase was reduced 100-fold in the mutant, the cells grew almost normally and appeared to replicate their DNA on time and without problems! The isolation of a mutant "with no detectable DNA polymerization activity" established the non-replicative nature of Kornberg DNA polymerase. This teaches us an important lesson: to never assume that the actual in vivo function of the enzyme is what it does in assays in vitro! As Genetics has to rely on Biochemistry for figuring out mechanisms of various processes, Biochemistry has to rely on Genetics for assigning the in vivo functions (“the biological reason”) to the in vitro mechanisms. Isolation and properties of DNA pol II and pol III Biochemists went back to work and used a combination of polA mutant and Kornberg polymerase-specific antiserum (to completely eliminate the activity of Kornberg polymerase, which was still 1% in the mutant) to look for novel DNA polymerase activities. They did isolate two new DNA polymerases, whose activities were masked in the extracts of WT cells by the powerful activity of Kornberg enzyme. Therefore, they called Kornberg polymerase DNA pol I, while the newly-isolated enzymes were called DNA pol II and III. Curiously, this work was done by Thomas Kornberg (the second son of Arthur Kornberg, not to be confused with the first son, Roger, who later on got his own Nobel prize for eukaryotic RNA polymerase characterization).2 The properties of the new polymerases in comparison with DNA pol I are listed in the Table (see next page). The optimal assay conditions of the new enzymes differed from those of Pol I, making it possible to assay for them directly, especially in the presence of Pol I antiserum. Soon the gene for pol II has been isolated and called polB. The double polA polB mutant was fully viable, suggesting that neither DNA polymerase is required to replicate the chromosome. The slow rate of DNA synthesis by Pol II immediately disqualified this enzyme as a replicative DNA polymerase, but Pol III looked promising. Then the available dna-heat-sensitive mutants were checked for heat-liable DNA Pol III activity. It was found that at 42°C, the Pol III activity from dnaE mutants became inactive. Therefore, dnaE must code for Pol III, and the gene was renamed polC. (However, the dnaE name also stuck, so both names are used these days in E. coli. Confusingly, some bacteria have both orthologs, so dnaE and polC are separate genes in them!) We will deal with the structure and function of DNA pol III later; what should concern us now are the two irritating seemingly universal properties of the E. coli DNA polymerases: 1) the inability to start DNA synthesis without a primer (this property, which seemed natural for the apparently DNA repair polymerases I and II, was not fit for the replicative enzyme Pol III); 2) the inability to synthesize DNA in the 3’—>5’ direction. Therefore, instead of solving the mystery of DNA synthesis at replication forks, characterization of the new DNA polymerases only deepened it, strengthening the two questions: 1) how is the DNA synthesis primed? 2) how is the 5’—>3’ DNA strand replicated? The “fork & knife” idea One ingenious proposal answered both questions about the chromosomal replication by taking into account the strange products of the in vitro DNA synthesis by DNA pol I. As was mentioned before, Kornberg enzyme, if "left to play unattended" with a nicked DNA duplex, synthesizes a suspicious DNA product: it looked like a regular dsDNA by spectrophotometry, but differed from it in the ability to instantly (though partially) renature and also had a branching appearance in the EM. Kornberg interpreted this result to mean that, when synthesizing DNA, his polymerase can displace the 5’-ending strand and, if the displaced strand grew too long, can switch the template (Draw like Fig. 4-32 Kornberg, p. 154). This observation allowed Guild to propose the following elegant model of DNA synthesis at the replication fork: (Draw like 8-25 Freifelder, p. 279). This model, known as the “fork &
View Full Document
Unlocking...