Does a universal mechanism for TLS exist Our previous
Does a universal mechanism for TLS exist? Our previous studies using the bacteriophage T4 DNA polymerase generated the model provided in Fig. 7 that quantifies the molecular forces influencing key steps in the polymerization pathway during the replication of a non-instructional versus miscoding DNA lesion. Results from this current study recapitulate the importance of nucleobase desolvation toward achieving optimal binding of a nucleotide opposite damaged DNA. In most biological systems, substrate desolvation is a significant barrier in achieving optimal binding and catalysis, since most enzymes rely heavily on the formation of hydrogen-bonding interactions between a substrate and enzyme. This barrier, termed “enthalpic–entropic compensation”, occurs as hydrogen bonds that exist in solution between water and a substrate must be removed so that new hydrogen bonds can form between the substrate and CI 976 within the enzyme\'s active site. During DNA polymerization, this barrier is especially relevant, as water molecules must be removed from the incoming nucleotide in order for the DNA polymerase to consummate the formation of correct hydrogen bonds between complementary base pairs within duplex DNA. The kinetic data presented here strengthen this model, as hydrophobic nucleotides bind to gp43 with lower Kd values compared to their unmodified counterparts. In particular, analogs with low solvation energies such N6-Me-dATP, O6-Me-dGTP, and 6-Cl-PTP bind 3- to 10-fold more tightly than dATP, which has a significantly higher solvation energy. Similar results were obtained with artificial nucleotide analogs such as 5-MeCITP during the replication of non-instructional and miscoding DNA lesions. For example, 5-MeCITP binds with low Kd values of 13μM and 57μM during the replication of an abasic site and 8-oxo-G, respectively . The importance of nucleobase desolavtion is further strengthened by weaker binding affinities measured for hydrophilic nucleotide analogs such as 5-CITP, which displays a high Kd value of 172μM during the replication of an abasic site . In this case, the high Kd value likely reflects an enthalpic penalty caused by the negative charge of the carboxyl group. While desolvation appears to play a universal role in the binding of the nucleotide substrate during TLS, the molecular forces regulating the polymerization step are more divergent as they depend upon the physical nature of the DNA lesion. For instance, π-stacking interactions play a large role in facilitating the polymerization step during the replication of non-instructional lesions such as abasic sites. This is based on the fact that artificial analogs such as 5-CITP and 5-MeCITP that possess significant π-electron density also display incredibly fast kpol values of 67s and 79s, respectively. In contrast, the rate constant for the polymerization step during the replication of the miscoding lesion, 8-oxo-G, depends more upon hydrogen-bonding interactions. This is evident as all of the nucleotide analogs tested here, which contain modifications to hydrogen-bonding groups, have lower kpol values compared to dATP. In fact, artificial analogs such as 5-CITP and 5-MeCITP that are incorporated opposite an abasic site with fast kpol values of ~70s are inserted opposite 8-oxo-G with lower kpol values of ~0.23s. Do other DNA polymerases utilize nucleobase desolvation during TLS? During chromosomal replication, high-fidelity DNA polymerases accurately and efficiently replicate undamaged DNA. In contrast, the activity of these DNA polymerases is significantly hindered when replicating damaged DNA. As a result, specialized DNA polymerases such as pol eta, pol kappa, and pol iota are recruited to participate more intimately in the efficient replication of unrepaired DNA lesions. However, the ability of specialized DNA polymerases to effectively perform TLS comes at a cost as they generally display reduced fidelity when replicating undamaged DNA. Current models attempting to explain this dichotomy are based primarily on structural differences that exist between the two classes of DNA polymerase , , . In general, both high-fidelity and specialized DNA polymerases possess a similar global architecture that resembles a right hand and contains elements corresponding to fingers, palm, and thumb domains , . However, close inspection reveals that the active sites of most specialized DNA polymerases are significantly larger than those of high-fidelity DNA polymerases. The expanded active site of specialized DNA polymerases is often used to explain how these polymerases can replicate large, bulky lesions, whereas the more constrained active site of high-fidelity polymerases hinders their ability to efficiently replicate damaged DNA. At face value, the results presented here using modified nucleotide analogs are consistent with the mechanism. However, we propose that nucleobase solvation also plays an important role in achieving nucleotide discrimination, especially during the replication of damaged DNA. An excellent example of this phenomenon comes from the kinetic studies here demonstrating that gp43 binds dATP very poorly when replicating 8-oxo-G. In this case, we propose that the weaker binding affinity reflects energetic penalties associated with stripping away water molecules that are bound to key hydrogen-bonding groups present on the natural nucleotide. The inference here is that the association of water molecules with these functional groups creates a solvation sphere around the nucleobase, which increases the overall size of the nucleotide. The resulting increase in size hinders efficient binding within the constrained active sites of high-fidelity polymerases. As demonstrated here, modifications such as alkylation that increase the overall hydrophobicity of the nucleobase also reduce the size of this solvation sphere. The biophysical consequence is that the smaller size of the nucleotide makes binding to the polymerase more efficient and lowers the energetic penalties required for complete desolvation of the incoming nucleotide.