Rapid PCR in a continuous flow device

Masahiko Hashimoto, Department of Chemistry, Louisiana State University, Baton Rouge, LA 70803, USA.
Pin-Chuan Chen
Michael W. Mitchell
Dimitris E. Nikitopoulos
Steven A. Soper
Michael C. Murphy


Continuous flow polymerase chain reaction (CFPCR) devices are compact reactors suitable for microfabrication and the rapid amplification of target DNAs. For a given reactor design, the amplification time can be reduced simply by increasing the flow velocity through the isothermal zones of the device; for flow velocities near the design value, the PCR cocktail reaches thermal equilibrium at each zone quickly, so that near ideal temperature profiles can be obtained. However, at high flow velocities there are penalties of an increased pressure drop and a reduced residence time in each temperature zone for the DNA/reagent mixture, that potentially affect amplification efficiency. This study was carried out to evaluate the thermal and biochemical effects of high flow velocities in a spiral, 20 cycle CFPCR device. Finite element analysis (FEA) was used to determine the steady-state temperature distribution along the micro-channel and the temperature of the DNA/reagent mixture in each temperature zone as a function of linear velocity. The critical transition was between the denaturation (95 degrees C) and renaturation (55 degrees C-68 degrees C) zones; above 6 mm s(-1) the fluid in a passively-cooled channel could not be reduced to the desired temperature and the duration of the temperature transition between zones increased with increased velocity. The amplification performance of the CFPCR as a function of linear velocity was assessed using 500 and 997 base pair (bp) fragments from lambda-DNA. Amplifications at velocities ranging from 1 mm s(-1) to 20 mm s(-1) were investigated. The 500 bp fragment could be observed in a total reaction time of 1.7 min (5.2 s cycle(-1)) and the 997 bp fragment could be detected in 3.2 min (9.7 s cycle(-1)). The longer amplification time required for detection of the 997 bp fragment was due to the device being operated at its enzyme kinetic limit (i.e., Taq polymerase deoxynucleotide incorporation rate).