We first determined the ability of Phi29 DNA polymerase to selectively amplify a B. burgdorferi circular plasmid, p26 and a covalently-closed linear plasmid, lp54. B. burgdorferi B31 genomic DNA with or without human 293 T genomic DNA was amplified with Phi29 DNA polymerase and primers sets specific for either plasmid (). Real-time PCR was then performed on aliquots of amplified DNA using primers specific for genes located on each DNA (). As shown in , the concentration of gene BBB08, located on cp26, was increased dramatically in genomic DNA amplified with Phi29 DNA polymerase and primers specific for cp26. Importantly, specific amplification of BBB08 was observed in samples with or without human DNA ( vs. ). In contrast, little or no amplification was observed for lp54 DNA as evidenced by the failure to observe an increase in the amount of the lp54 gene BBA03 target region in samples amplified with primers specific for either linear plasmid lp54 or circular plasmid cp26. Furthermore, the BBB08 target was not amplified in Phi29DNA reactions which only included primers specific for the linear plasmid lp54. From these results we conclude that circular plasmids but not linear plasmids of B. burgdorferi DNA can be readily amplified with Phi29 DNA polymerase and that amplification occurs in the presence of a vast excess of human DNA.
Fig. 1 Phi29 DNA polymerase selectively amplifies B. burgdorferi B31circular plasmid cp26. B. burgdorferi B31 genomic DNA (250 ng) with (A) or without (B) human genomic DNA (1 μg) was amplified with Phi29 DNA polymerase using primers specific for linear (more ...)
To verify that amplification of circular plasmid as opposed to linear plasmids was a general phenomenon and that the products are indeed primer specific, we tested other circular and linear plasmids in the B. burgdorferi B31 genome. A mixture of B. burgdorferi B31 genomic DNA and human genomic DNA was amplified with Phi29 DNA polymerase using primers specific for circular plasmids cp26 and cp32-6 as well as the linear plasmids lp28-3 and lp36. Relative amplification efficiency was monitored by real-time PCR using gene-specific primers (BBB08 for cp26, BBM04 for plasmid cp32-6, BBH05/06 for lp28-3, and BBK21 for lp36). These data () confirm that Phi29 DNA polymerase specifically amplified circular plasmids cp26 and cp32-6 as evidenced by the accumulation of plasmid specific gene targets BBB08 and BBM04, respectively. As before, there was no amplification of the linear plasmids, lp36 and lp28-3 linear plasmids. Furthermore, the data also demonstrate that circular plasmid amplification by Phi29 DNA polymerase is primer specific, cp32-6 target gene BBM04 was not amplified in reactions using cp26 specific primers, and vise versa.
Fig. 2 Phi29 DNA polymerase specifically amplifies B. burgdorferi B31circular plasmids. B. burgdorferi B31 genomic DNA with human genomic DNA was amplified with Phi29 DNA polymerase using primers specific for circular plasmids cp26 and cp32-6, and linear plasmids (more ...)
To develop a high-throughput sequence-based assay to detect pathogen DNA, we combined plasmid-specific primer amplification with subsequent random hexamer primer amplification. First, we amplified DNA mixtures containing a constant amount of human DNA spiked with decreasing amounts of Borrelia DNA (250 ng–2.5 pg) using biotinylated primers specific to cp26. Following amplification, the products were captured on streptavidin-coated magnetic beads. Bound material was subjected to universal amplification using the Illustra GenomiPhi V2 DNA Amplification Kit and the products digested with HindIII. After shot-gun cloning of the digests into Escherichia coli, 10 plasmid clones were randomly selected from each amplification reaction and sequenced.
Results revealed that the number of clones containing Borrelia cp26 DNA was directly related to the amount of Borrelia genomic DNA in the amplification reactions. Of the 10 clones isolated from each sample, 10 of 10 were positive for cp26 sequence in reactions containing 250 ng Borrelia DNA, 9 of 10 were positive in reactions containing 25 ng, 5 of 10 were positive in 2.5 ng reactions while 2 of 10 contained cp26 sequence in reactions with 250 pg. We could; moreover, reproducibly detect cp26 DNA in clones (1 of 10) obtained from reactions containing as little as 25 pg of Borrelia DNA which represents approximately 10,000 molecules of B. burgdorferi. Interestingly, the Borrelia positive clones only contained cp26 sequences and these were randomly distributed around the circular molecule; all non-positive clones contained human DNA. Thus, by performing an initial reaction with biotinylated primers, we were able to separate the amplified product from the original DNA mixture and re-amplify it with Phi29 DNA polymerase to generate sufficient material for cloning and DNA sequencing. The amplification reactions can be performed in several hours and has the advantage that only minimal quantities of starting material are needed for analysis.
In conclusion, we were able to specifically amplify pathogen circular DNA in the presence of human genomic DNA using Phi29 DNA polymerase mediated rolling-circle amplification. Our inability to amplify linear plasmids may be due to their having covalently closed ends which are know to promote rapid snap-back annealing following denaturation. Self-annealing would prevent primer binding and Phi29 DNA amplification. This technology can ultimately be adapted as a sensitive method to detect specific DNA from both known and unknown pathogens in a wide variety of complex clinical and environmental samples.