Screening for Polymerases with Altered Substrate Tolerance

Over the past 15 years a group of chemists worked to develop a novel strategy for DNA sequencing that is based on fluorescence detection at the single molecule level [43-47]. The proposed ''single molecule sequencing'' requires the complete and faithful synthesis of DNA copies exclusively from labeled analogs of the four types of nucleobase, A, G, C, and T, and thus unusual polymerase activity. Most natural DNA polymerases are known to discriminate against bulky fluorescent nucleotide analogs which often also have a net charge differing from that of the natural substrates [48-51]. Until recently, only few mutant bacteriophage T4 DNA polymerases were known to have an increased capacity to incorporate modified monomers [52].

To gain access to one or more DNA polymerases with the abilities to:

1. incorporate a fluorophore-labeled nucleotide,

2. extend the terminus by addition of the next fluorophore-labeled substrate, and

3. retain sufficient incorporation fidelity we developed a functional screening approach and started the search by using the Klenow fragment (KF) of DNA polymerase I (Pol I) of Escherichia coli. Because elongation of a primer-template is not a single-step reaction that can be assessed with simple yes/no decisions, we needed a detection technique that yields quantitative information on molecular sizes, or fluorescence intensities, or both. Fluorescence correlation spectroscopy (FCS) [45, 53] is, in principle, a suitable technique for analysis of minute spontaneous fluctuations in the fluorescence emission behavior of small molecular ensembles that reflect inter- or intramolecular dynamics. FCS can, for example, give information about the velocity of trans-lationally diffusing molecules. Because this value is related to the molecular size, we assumed that, ideally, chain lengths of polymerization products can be deter-

Fig. 4.3.3. Elongation of the primer hybridized incorporation of sterically and electronically to the 5'-biotinylated template requires polymerization along a homopolymeric (dA)58 stretch. In the sole presence of fluorescence-labeled deoxynucleoside triphosphates (dye = tetramethylrhodamine, TAMRA), this selective constraint forces the multiple successive demanding substrates. Prior to fluorimetric assessment, reaction products can be immobilized on streptavidin-coated surfaces and purified from excessive fluorescent monomer

Fig. 4.3.3. Elongation of the primer hybridized incorporation of sterically and electronically to the 5'-biotinylated template requires polymerization along a homopolymeric (dA)58 stretch. In the sole presence of fluorescence-labeled deoxynucleoside triphosphates (dye = tetramethylrhodamine, TAMRA), this selective constraint forces the multiple successive demanding substrates. Prior to fluorimetric assessment, reaction products can be immobilized on streptavidin-coated surfaces and purified from excessive fluorescent monomer

mined. We thus applied series of primer extension reactions that required elongation of a primer-template along a homopolymeric (dA)58 stretch of the a template (Figure 4.3.3).

While setting up our screening system with the wild-type KF, we observed the prevalence of a single molecular species that was identified to be the full length reaction product. The surprising result suggested that the wild-type enzyme was at least capable of catalyzing the template-instructed polymerization of 58 tetrame-thylrhodamine-(TAMRA)-labeled deoxyuridylic acid residues. In pursuance of labeling long, natural DNA molecules, we then utilized KF in further primer extension reactions and substituted:

1. TTP by its analog TAMRA-dUTP

2. dCTP by rhodamine110-(R110)-dCTP, and

3. both TTP and dCTP by the respective analogs.

Diverse analyses demonstrated that we prepared heteroduplex DNA fragments of 7000-9000 base pairs (bp) with complete substitution of both pyrimidine nucleotides [54]. Although the labeled DNA differs very much from natural DNA (for example, with respect to solubility, melting behavior, and circular dichroism), we were also able to show that the high degree of label incorporation so far did not alter the fidelity of KF.

It should be mentioned that our screening approach, i.e. comparative evaluation of individual polymerase variants, required two purification steps [55]. On the one hand, polymerase mutants cloned and expressed by use of E. coli had to be separated from competing host polymerases; on the other hand, reaction products had to be separated from excess fluorescent monomers that could interfere during the FCS measurement. Although all of these purification steps were performed with one-step procedures using commercially available microwell formats, the throughput was comparatively low. Future improvements should therefore aim at circumventing as many purification steps as possible, for example by alteration of the expression system or by employing detection techniques that enable more efficient distinction between educts and products.

0 0

Post a comment